Dye-sensitized Solar Cells [1 ed.] 9782940222360, 9781439808665, 9780429156359

With contributions by: Michael Bertoz, Juan Bisquert, Filippo De Angelis, Hans Desilvestro, Francisco Fabregat-Santiago,

491 81 39MB

English Pages [621] Year 2010

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Dye-sensitized Solar Cells [1 ed.]
 9782940222360, 9781439808665, 9780429156359

Table of contents :
Cover
Half Title
Title Page
Copyright Page
Table of Contents
PREFACE
1: PHOTOCHEMICAL AND PHOTOELECTROCHEMICAL APPROACHES TO ENERGY CONVERSION
1.1 The sun as an abundant energy resource
1.2 Photochemical conversion and storage of solar energy (artificial photosynthesis)
1.3 Photographic sensitization
1.4 Photoelectrochemical conversion of solar energy
1.4.1 Photogalvanic cells
1.4.2 Generations of photovoltaic solar cells
1.4.3 Photoelectrochemical solar cells with liquid junctions
1.4.4 Photoredox reactions of colloidal semiconductors and particulates
1.5 Dye sensitization of semiconductors
1.5.1 Dye sensitization of bulk semiconductor electrodes
1.5.2 Dye-sensitized solar cells – an overview
1.5.3 Sequence of electron-transfer steps of a DSC
1.5.4 Key efficiency parameters of a DSC
1.5.5 Key components of the DSC
1.5.6 Quasi-solid state DSCs with spiro-OMeTAD
1.5.7 Improvement in efficiency through the nanostructuring of materials
1.5.8 Dye solar cells based on nanorods/nanotubes and nanowires
1.5.9 Sensitization using quantum dots
1.5.10 semiconductor-sensitized ETA solar cells
1.5.11 DSCs based on p-type semiconductor
1.6 Conclusions
1.7 References
2: TITANIA IN DIVERSE FORMS AS SUBSTRATES
2.1 Titania: fundamentals
2.2 Electrochemistry of titania: depletion regime
2.2.1 Photoelectrochemistry under band-gap excitation
2.2.2 In-situ FTIR spectroelectrochemistry in the depletion regime
2.2.3 Photoelectrochemistry under sub-band-gap excitation
2.3 Electrochemistry of titania: accumulation regime
2.3.1 Capacitive processes
2.3.2 Li-insertion electrochemistry
2.3.3 Spectroelectrochemistry of titania in the accumulation regime
2.4 Titania photoanode for dye sensitized solar cells
2.4.1 Non-organized titania made by decomposition of Ti(IV) alkoxides
2.4.2 Electrochemical deposition of titania
2.4.3 Aerosol pyrolysis
2.4.4 Organized nanocrystalline titania
2.4.5 Single-crystal anatase electrode
2.4.6 Other methods of producing titania electrodes for DSC
2.4.7 Multimodal structures
2.5 Conclusion
2.6 Acknowledgements
2.7 References
3: MOLECULAR ENGINEERING OF SENSITIZERS FOR CONVERSION OF SOLAR ENERGY INTO ELECTRICITY
3.1 Introduction
3.2 Ruthenium sensitizers
3.2.1 Effect of protons carried by the sensitizers on the performance
3.2.2 Effect of cations in the ruthenium sensitizers on the performance
3.2.3 Device stability
3.2.4 Effect of alkyl chains in the sensitizer on the performance
3.2.5 Effect of the p-conjugation bridge between carboxylic acid groups and the ruthenium chromophore
3.2.6 High Molar Extinction Coefficient Sensitizers
3.2.7 Tuning spectral response by thiocyanato ligands
3.2.8 Non-thiocyanato ruthenium complexes
3.3 Organic sensitizers
3.3.1 High efficiency organic sensitizers
3.3.2 Near-IR absorbing sensitizers
3.4 References
4: OPTIMIZATION OF REDOX MEDIATORS AND ELECTROLYTES
4.1 Introduction
4.2 Charge transfer processes in DSCs
4.3 Electrolyte Components and their roles in the DSCs
4.3.1 Organic solvents
4.3.2 Cations
4.3.3 Additives
4.3.4 Electron mediators
4.4 Ionic liquid, quasi-solid and solid electrolytes
4.4.1 Ionic liquid electrolyte
4.4.2 Active iodide molten salts
4.4.3 Nonactive iodide molten salts
4.4.4 Additives in ILEs
4.4.5 Quasi-solid electrolyte
4.5 Remarks and prospects
4.6 References
5: PHOTOSENSITIZATION OF SnO2 AND OTHER OXIDES
5.1 Dependence of the Sensitization Efficiency on the Energy Difference
5.2 Coupled semiconductor Systems
5.3 SnO2-C60-Ru(bpy)23+ System
5.4 Probing the Interaction of an Excited State Sensitizer with the Redox Couple
5.5 Sensitization of Nanotube Arrays
5.6 Charge Separation of Organic Clusters at an SnO2 Electrode Surface
5.7 Concluding Remarks
5.8 Acknowledgements
5.9 References
6: SOLID-STATE DYE-SENSITIZED SOLAR CELLS Inc.RPORATING MOLECULAR HOLE-TRANSPORTERS
6.1 Introduction
6.2 Spiro-OMeTAD-based solid-state dye-sensitized solar cell
6.3 The influence of additives upon the solar cell performance
6.4 Charge generation: Electron Transfer
6.5 Reductive quenching
6.6 Charge generation: Hole-transfer
6.7 Charge transport in molecular hole-transporters
6.8 Hole mobility in spiro-OMeTAD
6.9 Influence of charge density on the hole-mobility in molecular semiconductors
6.10 The influence of chemical p-doping upon Conductivity and hole-mobility
6.11 The influence of ionic salts on conductivity and hole-mobility
6.12 Current collection
6.13 TiO2 pore filling with molecular hole-transporters
6.14 Charge recombination: The influence of additives
6.15 Charge recombination: Ion solvation and immobilization
6.16 Charge recombination: Controlling the spatial separation of electrons and holes at the heterojunction
6.17 Enhancing light capture in solid-state DSCs
6.18 Alternative structures for mesoporous and nanostructured electrodes in solid-state DSCs
6.19 Outlook for hole-transporter based solid-state DSCs
6.20 References
7: PACKAGING, SCALE-UP AND COMMERCIALIZATION OF DYE SOLAR CELLS
7.1 Introduction
7.2 From cells to panels
7.2.1 Definitions
7.2.2 Designs
7.2.3 Materials
7.2.4 Module performance - experiment vs. modeling
7.3 Long-term stability - the key to industrial success
7.3.1 Single cells
7.3.2 Modules
7.3.3 Panels
7.4 Scaling up to commercial production levels
7.4.1 Material costs and availability
7.4.2 Manufacturing
7.5 Commercial applications
7.6 Conclusions
7.7 Acknowledgements
7.8 References
8: HOW TO MAKE HIGH-EFFICIENCY DYE-SENSITIZED SOLAR CELLS
8.1 Introduction
8.2 Experimental considerations
8.2.1 Preparation of screen-printing pastes
8.2.2 Synthesis of Ru-dye
8.2.3 Porous-TiO2 electrodes
8.2.4 Counter-Pt electrodes
8.2.5 DSC assembling
8.2.6 Measurements
8.3 Results and discussion
8.3.1 TiCl4 treatments
8.3.2 Effect of the light-scattering TiO2 layer
8.3.3 Thickness of the nanocrystalline TiO2 layer
8.3.4 Anti-reflecting film
8.3.5 Reproducibility of DSC photovoltaics
8.4 Conclusion
8.5 Acknowledgements
8.6 References
9: SCALE-UP AND PRODUCT-DEVELOPMENT STUDIES OF DYE-SENSITIZED SOLAR CELLS IN ASIA AND EUROPE
9.1 Introduction
9.2 Scaling up of laboratory cells to modules and panels
9.3 DSC development studies in various european laboratories
9.3.1 Energy Research Centre of the Netherlands (ECN)
9.3.2 Fraunhofer Institute for Solar Energy Systems (Fraunhofer ISE)
9.3.3 G24 Innovation
9.3.4 3GSolar, Israel
9.4 DSC development studies in various laboratories of Japan
9.4.1 Aisin Seiki Co. Ltd. and Toyota Central R&D Laboratories
9.4.2 Fujikura Ltd. (Japan)
9.4.3 Peccell Technologies, Inc. (Japan)
9.4.4 Sharp Co. Ltd. (Japan)
9.4.5 Sony Corporation Ltd. (Japan)
9.4.6 Shimane Institute for Industrial Technology (Japan)
9.4.7 TDK Co., Ltd. (Japan)
9.4.8 Eneos Co. Ltd. (Japan)
9.4.9 NGK Spark Plug Co., Ltd. (Japan)
9.4.10 Panasonic Denko Co. Ltd. (Japan)
9.4.11 Taiyo Yuden Co., Ltd. (Japan)
9.4.12 Dai Nippon Printing Company
9.4.13 Mitsubhishi Paper Mills and Sekisui Jushi Corporation
9.4.14 J-Power Co. Ltd. (Japan)
9.5 DSC Development Work in Korea and Taiwan
9.5.1 Korean Institute of Science and Technology (KIST)
9.5.2 Electronics and Telecommunications Research Institute (ETRI), Korea
9.5.3 Samsung SDI, Korea
9.5.4 Industrial Technology Research Institute of Taiwan (ITRI)
9.5.5 J Touch Taiwan
9.6 DSC development work in Australia and China
9.6.1 Dyesol, Australia
9.6.2 Institute of Plasma Physics, Chinese Academy of Sciences
9.7 Conclusion
9.8 Acknowledgement
9.9 References
10: CHARACTERIZATION AND MODELING OF DYE-SENSITIZED SOLAR CELLS: A TOOLBOX APPROACH
10.1 Introduction
10.2 Theoretical background
10.2.1 Interfacial electron transfer processes in the DSC
10.2.2 Electron trapping in the DSC
10.2.3 Electron transport in the DSC
10.3 The toolbox
10.3.1 Determination of injection efficiency and electron diffusion length under steady-state conditions
10.3.2 Electrochemical and spectrolectrochemical techniques to study the energetics of the oxide/dye/electrolyte interface
10.3.3 Electrochemical measurements with thin layer cells
10.3.4 Small-amplitude time-resolved methods
10.3.5 Methods based on frequency response analysis
10.3.6 Photovoltage decay
10.3.7 Determination of density of trapped electrons in DSCs
10.3.8 Measuring the internal electron quasi Fermi level in the DSC
10.3.9 Determining the electron diffusion length using IMVS and IMPS
10.3.10 Photoinduced absorption spectroscoy (PIA)
10.3.11 Conclusions
10.4 Acknowledgments
10.5 Appendix 1 Analytical IMPS solutions
10.6 Appendix 2 Numerical solutions of the continuity equation [10.115]
10.7 References
11: DYNAMICS OF INTERFACIAL AND SURFACE ELECTRON TRANSFER PROCESSES
11.1 Introduction
11.2 Energetics of charge transfer reactions
11.2.1 Mesoscopic metal oxide semiconductors
11.2.2 Dye sensitizer
11.3 Kinetics of interfacial electron transfer
11.3.1 Charge injection dynamics
11.3.2 Charge recombination
11.4 Electron transfer dynamics involving the redox mediator
11.4.1 Kinetics of interception of dye cations by a redox mediator
11.4.2 Conduction band electron – oxidized mediator recombination
11.4.3 Rlectron transport in nanocrystalline TiO2 films
11.5 References
12: IMPEDANCE SPECTROSCOPY: A GENERAL INTRODUCTION AND APPLICATION TO DYE-SENSITIZED SOLAR CELLS
12.1 Introduction
12.2 A basic solar cell model
12.2.1 The ideal diode model
12.2.2 Physical origin of the diode equation for a solar cell
12.3 Introduction to IS methods
12.3.1 Steady state and small perturbation quantities
12.3.2 The frequency domain
12.3.3 Simple equivalent circuits
12.4 Basic physical model and parameters of IS in solar cells
12.4.1 Simplest impedance model of a solar cell
12.4.2 Measurements of electron lifetimes
12.5 Basic physical models and parameters of IS in dye-sensitized solar cells
12.5.1 Electronic processes in a DSC
12.5.2 The capacitance of electron accumulation in a DSC
12.5.3 Recombination resistance
12.5.4 The transport resistance
12.6 Transmission line models
12.6.1 General structure of transmission lines
12.6.2 General diffusion transmission lines
12.6.3 Diffusion-recombination transmission line
12.6.4 Parameters of the diffusion-recombination model
12.6.5 Effect of boundaries on the transmission line
12.7 Applications
12.7.1 Liquid electrolyte cells
12.7.2 Experimental IS parameters of DSCs
12.7.3 Nanotubes
12.7.4 Effects of the impedance parameters on the j-V curves
12.8 Acknowledgments
12.9 Appendix: properties of measured DSCs
12.10 References
13: THEORETICAL AND MODEL SYSTEM CALCULATIONS
13.1 Introduction
13.2 Theoretical and computational methods
13.2.1 Density Functional Theory (DFT)
13.2.2 Basis sets
13.2.3 The Car-Parrinello method
13.2.4 Solvation effects
13.2.5 Excited states
13.2.6 Nonadiabatic method
13.3 Dye sensitizers
13.3.1 Ruthenium(II)-polypyridyl sensitizers
13.3.2 Calculations on N3
13.3.3 Calculations on other Ru(II)-dye sensitizers
13.3.4 Trans-complexes
13.3.5 Organic sensitizers
13.3.6 Squaraine dyes
13.4 Studies of the TiO2 substrate
13.4.1 TiO2 models
13.5 Dye sensitizers on TiO2
13.5.1 Organic dyes on TiO2: adsorption and electron dynamics
13.5.2 Inorganic dyes on TiO2: adsorption and excited states
13.6 Conclusions and perspective
13.7 References
INDEX

Citation preview

DYE-SENSITIZED SOLAR CELLS

Fundamental Sciences

Chemistry

DYE-SENSITIZED SOLAR CELLS Edited by K. Kalyanasundaram

With contributions by: Michael Bertoz, Juan Bisquert, Filippo De Angelis, Hans Desilvestro, Francisco Fabregat-Santiago, Simona Fantacci, Anders Hagfeldt, Seigo Ito, Ke-jian Jiang, K. Kalyanasundaram, Prashant V. Kamat, Ladislav Kavan, Jacques-E. Moser, Md. K. Nazeeruddin, Laurence Peter, Henry J. Snaith, Gavin Tulloch, Sylvia Tulloch, Satoshi Uchida, Shozo Yanagida and Jun-ho Yum Forewords by: Michael Grätzel and Shozo Yanagida

EPFL Press A Swiss academic publisher distributed by CRC Press

Taylor and Francis Group, LLC 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487 Distribution and Customer Service [email protected] www.crcpress.com Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress.

This book is published under the editorial direction of Professor Hubert Girault (EPFL). The publisher, editor and authors of this book would like to thank the Swiss Federal Institute of Technology (EPFL) for its generous support towards the publication of this book and are grateful to the following industrial sponsors for their participation that helped make this project possible: Dyesol Group, Sefar A.G. and enerStore Consulting, Ltd.

is an imprint owned by Presses polytechniques et universitaires romandes, a Swiss academic publishing company whose main purpose is to publish the teaching and research works of the Ecole polytechnique fédérale de Lausanne. Presses polytechniques et universitaires romandes EPFL – Rolex Learning Center Post office box 119 CH-1015 Lausanne, Switzerland E-mail : [email protected] Phone : 021 / 693 21 30 Fax : 021 / 693 40 27 www.epflpress.org © 2010, First edition, EPFL Press, Lausanne (Switzerland) ISBN 978-2-940222-36-0 (EPFL Press) ISBN 978-1-4398-0866-5 (CRC Press) Printed in France All right reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprint, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publisher.

FIRST FOREWORD It is a pleasure and honor to contribute this Foreword to the special monograph covering the latest progress in dye sensitized solar cells (DSCs), which have been invented in our laboratory and remain the focus of intense investigations on the international level. The DSC is a unique photovoltaic converter in as much as it is the only solar cell that mimics natural photosynthesis. Like green plants and algae it uses a molecular absorber, the dye, to harvest sunlight and generate electric charges, achieving for the first time successfully the separation of the two functions of light harvesting and charge-carrier transport. This opens up many new options for practical applications, such as the realization of transparent glass panels that generated electric power from the sun. The DSC also employs a radically new concept for solar cells introducing three-dimensional (bulk) interpenetrating network junctions instead of the customary flat p-n layer embodiment by conventional solar cells. Thus in several regards the DSC is a unique photovoltaic converter that uses a approach in the harvesting of solar light that is disruptive with regards to conventional semiconductor technology. The thirteen chapters written by leaders in this rapidly developing field address the salient features of these novel type of solar cells, starting from the fundamental level to industrial applications. They provide a comprehensive picture of the latest advances of this technology and an in-depth analysis of the elementary process that lead to electric-power production from sunlight. Thus the reader obtains an excellent overview on the most relevant aspects of the operation of DSCs, starting from the quantum-mechanical modeling of light absorption and charge-carrier generation, to a detailed analysis of the electron-transfer and charge-carrier-conduction processes involved in the light-to-electric power-conversion process. Apart from these fundamental aspects, the reader is shown how far the practical development of new photovoltaic contender has progressed since it inception some 18 years ago. I am extremely grateful to the authors that have contributed to this book for providing such superb overviews, reflecting the fascinating research results that this new field has been creating. Michael Grätzel Lausanne, February 2010

SECOND FOREWORD It is a great privilege and honor to write a foreword to the monograph on the dyesensitized solar cell (DSC). Firstly, I must mention the recent achievement in the laboratories of Professor Michael Grätzel of the EPFL; a new-dye-based DSC has given light-to-energy conversion in lab-size cells (≤ 1 cm2) of about 12 %. To my understanding, this feat has been achieved through molecule- and ion-interface science and electron-transfer science applied to the vicinity of the new dye molecules adsorbed on mesoscopic nano-crystalline TiO2 (nc-TiO2). The design of DSCs – with their mesoscopic structures of dye-adsorbed nanocrystalline-TiO2 film infiltrated with iodide/iodine electrolytes – was inspired by photosynthesis. Since then, tremendous progress has been made thanks to the knowledge and wisdom of coordination chemistry, inorganic material chemistry and electrochemistry. These cells, once referred to as photosynthetic solar cells or dyesensitized ceramic solar cells, have developed thanks to the innovation and enthusiasm of research teams with the aim to find an efficient method for generating clean and renewable energy. The conviction is that solar energy will be able to supply all our energy needs, which, currently is on the order of 13 terawatts. Dye Sensitized Solar Cells provides a comprehensive overview, bringing together the fundamentals of DSCs from the materials, performance and mechanistic aspects; it is also an advanced level monograph that summarizes key advances and technical issues. As explained in this book, the principle of the DSC is quite different from that of crystalline-material-based solar cells. At the beginning, a few experts in the field of semiconductor physics were very skeptical about the potential efficiency of electron injection and charge separation at the interface between nano-crystalline TiO2 and dye molecules. In addition, some experts in the field of photocatalysis chemistry questioned the long-term stability of DSCs under solar light irradiation. The monograph poses the foundation for the current paradigm in semiconductor photovoltaics, i.e., built-in electrical field to separate the photo-generated electron and hole pair is not operative in DSCs. In this sense, the DSC may be regarded as a mesoscopic heterojunction solar cell or as a mesoscopic electron-injection solar cell. The unfortunate observation of photo-degradation of dye molecules on nano-crystalline TiO2 may be rationalized as the consequence of the inevitable photochemical reactions of the dye molecules with co-adsorbed oxygen and water molecules on nc-TiO2. The DSC invention and their progress in the laboratories of the EPFL can be admired by referring to a proverb which states that “Prosperity comes from the prepared mind”. The contributions of the leading DSC research teams can be also admired from the perspective of a phrase that I found at an Institute in Trivandrum

viii

Second Foreword

(India): “Determination is a vital and essential factor in achieving success. The joy of success has encouraged many worthy researchers and engineers of DSCs”. Cadmium-telluride solar-cell modules, with a price of less than 1 $/watt, are now growing only in the solar power-plant market, but they have a limitation due to the producer liability issues arising from toxicity of materials. Readers of the monograph will come to realize that DSC modules, by achieving high conversion efficiency and long-term stability, will lead to the creation of new jobs. Thus, investment in DSC research and in the development of functional materials for DSC modules continues, for example in such areas as cost-effective panchromatic dyes; transparent and highly conductive sheet materials; and mesoscopic metal oxides with electron-active morphology. It is not unrealistic to hope that scientists and engineers who work on DSC technology, through their mission to improve solar technology, will have a real impact on our world in very near future. Shozo Yanagida, Emeritus Professor Osaka University, January, 2010.

CONTENTS PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii 1

2

PHOTOCHEMICAL AND PHOTOELECTROCHEMICAL APPROACHES TO ENERGY CONVERSION . . . . . . . . . . . . K. Kalyanasundaram 1.1 The sun as an abundant energy resource. . . . . . . . . . . . . 1.2 Photochemical conversion and storage of solar energy (artificial photosynthesis) . . . . . . . . . . . . . . . . 1.3 Photographic sensitization . . . . . . . . . . . . . . . . . . . . 1.4 Photoelectrochemical conversion of solar energy . . . . . . . . 1.4.1 Photogalvanic cells . . . . . . . . . . . . . . . . . . . 1.4.2 Generations of photovoltaic solar cells. . . . . . . . . 1.4.3 Photoelectrochemical solar cells with liquid junctions 1.4.4 Photoredox reactions of colloidal semiconductors and particulates. . . . . . . . . . . . . . . . . . . . . 1.5 Dye sensitization of semiconductors . . . . . . . . . . . . . . 1.5.1 Dye sensitization of bulk semiconductor electrodes . . 1.5.2 Dye-sensitized solar cells – an overview . . . . . . . . 1.5.3 Sequence of electron-transfer steps of a DSC . . . . . 1.5.4 Key efficiency parameters of a DSC . . . . . . . . . . 1.5.5 Key components of the DSC . . . . . . . . . . . . . . 1.5.6 Quasi-solid state DSCs with spiro-OMeTAD . . . . . 1.5.7 Improvement in efficiency through the nanostructuring of materials . . . . . . . . . . . . . . 1.5.8 Dye solar cells based on nanorods/nanotubes and nanowires . . . . . . . . . . . . . . . . . . . . . 1.5.9 Sensitization using quantum dots . . . . . . . . . . . 1.5.10 Semiconductor-sensitized ETA solar cells . . . . . . . 1.5.11 DSCs based on p-type semiconductor . . . . . . . . . 1.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . 1 . . . 1 . . . . . .

. . . . . .

. 2 . 5 . 6 . 6 . 7 . 11

. . . . . . . .

. . . . . . . .

. 14 . 16 . 16 . 17 . 18 . 19 . 21 . 32

. . . 33 . . . . . .

. . . . . .

. 34 . 35 . 36 . 37 . 38 . 38

TITANIA IN DIVERSE FORMS AS SUBSTRATES . . . . . . . . . . . . 45 Ladislav Kavan 2.1 Titania: fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . 45

x

Dye-Sensitized Solar Cells

2.2

2.3

2.4

2.5 2.6 2.7 3

Electrochemistry of titania: depletion regime . . . . . . . . . . . 2.2.1 Photoelectrochemistry under band-gap excitation . . . . 2.2.2 In-situ FTIR spectroelectrochemistry in the depletion regime . . . . . . . . . . . . . . . . . . . . . 2.2.3 Photoelectrochemistry under sub-band-gap excitation. . Electrochemistry of titania: accumulation regime . . . . . . . . . 2.3.1 Capacitive processes . . . . . . . . . . . . . . . . . . . 2.3.2 Li-insertion electrochemistry. . . . . . . . . . . . . . . 2.3.3 Spectroelectrochemistry of titania in the accumulation regime . . . . . . . . . . . . . . . . . . . . . . . . . . Titania photoanode for dye sensitized solar cells . . . . . . . . . 2.4.1 Non-organized titania made by decomposition of Ti(IV) alkoxides . . . . . . . . . . . . . . . . . . . . . 2.4.2 Electrochemical deposition of titania . . . . . . . . . . 2.4.3 Aerosol pyrolysis . . . . . . . . . . . . . . . . . . . . . 2.4.4 Organized nanocrystalline titania . . . . . . . . . . . . 2.4.5 Single-crystal anatase electrode . . . . . . . . . . . . . 2.4.6 Other methods of producing titania electrodes for DSC . 2.4.7 Multimodal structures . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

MOLECULAR ENGINEERING OF SENSITIZERS FOR CONVERSION OF SOLAR ENERGY INTO ELECTRICITY . . . . Jun-ho Yum and Md. K. Nazeeruddin 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Ruthenium Sensitizers . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Effect of protons carried by the sensitizers on the performance . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Effect of cations in the ruthenium sensitizers on the performance . . . . . . . . . . . . . . . . . . . . 3.2.3 Device stability . . . . . . . . . . . . . . . . . . . . . 3.2.4 Effect of alkyl chains in the sensitizer on the performance . . . . . . . . . . . . . . . . . . . . 3.2.5 Effect of the p-conjugation bridge between carboxylic acid groups and the ruthenium chromophore . . . . . 3.2.6 High Molar Extinction Coefficient Sensitizers. . . . . 3.2.7 Tuning spectral response by thiocyanato ligands . . . 3.2.8 Non-thiocyanato ruthenium complexes . . . . . . . . 3.3 Organic sensitizers . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 High efficiency organic sensitizers . . . . . . . . . . . 3.3.2 Near-IR absorbing sensitizers . . . . . . . . . . . . . 3.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . 48 . . 49 . . . . .

. 52 . 52 . 55 . 56 . 57

. . 59 . . 60 . . . . . . . . . .

. 61 . 62 . 63 . 64 . 71 . 73 . 74 . 76 . 76 . 76

. . . 83 . . . 83 . . . 84 . . . 85 . . . 86 . . . 88 . . . 89 . . . . . . . .

. . . . . . . .

. 92 . 96 . 99 101 102 102 109 113

Dye-Sensitized Solar Cells

4

5

6

xi

OPTIMIZATION OF REDOX MEDIATORS AND ELECTROLYTES . Ke-jian Jiang and Shozo Yanagida* 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Charge transfer processes in DSCs . . . . . . . . . . . . . . . . 4.3 Electrolyte components and their roles in the DSCs . . . . . . . 4.3.1 Organic solvents . . . . . . . . . . . . . . . . . . . . . 4.3.2 Cations . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 Additives . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.4 Electron mediators . . . . . . . . . . . . . . . . . . . . 4.4 Ionic liquid, quasi-solid and solid electrolytes . . . . . . . . . . 4.4.1 Ionic liquid electrolyte . . . . . . . . . . . . . . . . . . 4.4.2 Active iodide molten salts . . . . . . . . . . . . . . . . 4.4.3 Nonactive iodide molten salts . . . . . . . . . . . . . . 4.4.4 Additives in ILEs . . . . . . . . . . . . . . . . . . . . . 4.4.5 Quasi-solid electrolyte . . . . . . . . . . . . . . . . . . 4.5 Remarks and prospects . . . . . . . . . . . . . . . . . . . . . . 4.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 117 . . . . . . . . . . . . . . .

PHOTOSENSITIZATION OF SnO2 AND OTHER OXIDES . . . . Prashant V. Kamat 5.1 Dependence of the Sensitization Efficiency on the Energy Difference . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Coupled Semiconductor Systems . . . . . . . . . . . . . . . 5.3 SnO2-C60-Ru(bpy)32+ System . . . . . . . . . . . . . . . . . . 5.4 Probing the Interaction of an Excited State Sensitizer with the Redox Couple. . . . . . . . . . . . . . . . . . . . . 5.5 Sensitization of Nanotube Arrays . . . . . . . . . . . . . . . 5.6 Charge Separation of Organic Clusters at an SnO2 Electrode Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7 Concluding Remarks. . . . . . . . . . . . . . . . . . . . . . 5.8 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 5.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

SOLID-STATE DYE-SENSITIZED SOLAR CELLS INCORPORATING MOLECULAR HOLE-TRANSPORTERS . . Henry J. Snaith 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Spiro-OMeTAD-based solid-state dye-sensitized solar cell . 6.3 The influence of additives upon the solar cell performance . 6.4 Charge generation: Electron Transfer . . . . . . . . . . . . 6.5 Reductive quenching. . . . . . . . . . . . . . . . . . . . . 6.6 Charge generation: Hole-transfer . . . . . . . . . . . . . . 6.7 Charge transport in molecular hole-transporters. . . . . . . 6.8 Hole mobility in spiro-OMeTAD . . . . . . . . . . . . . .

. . . . . . . .

117 118 121 121 121 123 125 128 128 132 135 139 139 141 142

. . . 145

. . . 146 . . . 147 . . . 149 . . . 151 . . . 153 . . . .

154 156 156 156

. . . . 163 . . . . . . . .

. . . . . . . .

. . . . . . . .

163 165 166 168 171 171 174 175

xii

Dye-Sensitized Solar Cells

6.9 6.10 6.11 6.12 6.13 6.14 6.15 6.16 6.17 6.18 6.19 6.20 7

8

Influence of charge density on the hole-mobility in molecular semiconductors . . . . . . . . . . . . . . . . . . . . . . . . The influence of chemical p-doping upon conductivity and hole-mobility. . . . . . . . . . . . . . . . . . . . . . . . . . The influence of ionic salts on conductivity and hole-mobility Current collection . . . . . . . . . . . . . . . . . . . . . . . TiO2 pore filling with molecular hole-transporters . . . . . . Charge recombination: The influence of additives. . . . . . . Charge recombination: Ion solvation and immobilization . . . Charge recombination: Controlling the spatial separation of electrons and holes at the heterojunction . . . . . . . . . . Enhancing light capture in solid-state DSCs. . . . . . . . . . Alternative structures for mesoporous and nanostructured electrodes in solid-state DSCs . . . . . . . . . . . . . . . . . Outlook for hole-transporter based solid-state DSCs . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . 175 . . . . . .

. . . . . .

. . . . . .

177 180 181 187 192 193

. . . 194 . . . 195 . . . 198 . . . 203 . . . 203

PACKAGING, SCALE-UP AND COMMERCIALIZATION OF DYE SOLAR CELLS . . . . . . . . . . . . . . . . . . . . . . . . . . Hans Desilvestro, Michael Bertoz*, Sylvia Tulloch and Gavin Tulloch 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 From cells to panels . . . . . . . . . . . . . . . . . . . . . . . 7.2.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 Designs . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.3 Materials . . . . . . . . . . . . . . . . . . . . . . . . 7.2.4 Module performance - experiment vs. modeling. . . . 7.3 Long-term stability - the key to industrial success . . . . . . . 7.3.1 Single cells . . . . . . . . . . . . . . . . . . . . . . . 7.3.2 Modules . . . . . . . . . . . . . . . . . . . . . . . . 7.3.3 Panels . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Scaling up to commercial production levels . . . . . . . . . . . 7.4.1 Material costs and availability . . . . . . . . . . . . . 7.4.2 Manufacturing . . . . . . . . . . . . . . . . . . . . . 7.5 Commercial applications . . . . . . . . . . . . . . . . . . . . 7.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.7 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 7.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

HOW TO MAKE HIGH-EFFICIENCY DYE-SENSITIZED SOLAR CELLS . . . . . . . . . . . . . . . . . . . . . . . . Seigo Ito 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . 8.2 Experimental considerations. . . . . . . . . . . . . . 8.2.1 Preparation of screen-printing pastes. . . . . 8.2.2 Synthesis of Ru-dye . . . . . . . . . . . . .

. . . .

. . 207 . . . . . . . . . . . . . . . . .

207 211 211 211 214 218 224 224 228 230 231 231 237 240 245 246 246

. . . . . . . 251 . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

251 252 252 253

Dye-Sensitized Solar Cells

8.3

8.4 8.5 8.6 9

8.2.3 Porous-TiO2 electrodes . . . . . . . . . . . 8.2.4 Counter-Pt electrodes. . . . . . . . . . . . 8.2.5 DSC assembling . . . . . . . . . . . . . . 8.2.6 Measurements . . . . . . . . . . . . . . . Results and discussion . . . . . . . . . . . . . . . . 8.3.1 TiCl4 treatments . . . . . . . . . . . . . . 8.3.2 Effect of the light-scattering TiO2 layer . . 8.3.3 Thickness of the nanocrystalline TiO2 layer 8.3.4 Anti-reflecting film . . . . . . . . . . . . . 8.3.5 Reproducibility of DSC photovoltaics . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . .

xiii

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

SCALE-UP AND PRODUCT-DEVELOPMENT STUDIES OF DYE-SENSITIZED SOLAR CELLS IN ASIA AND EUROPE . . . . K. Kalyanasundaram, Seigo Ito, Shozo Yanagida and Satoshi Uchida 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Scaling up of laboratory cells to modules and panels . . . . . . 9.3 DSC development studies in various European laboratories . . 9.3.1 Energy Research Centre of the Netherlands (ECN) . . 9.3.2 Fraunhofer Institute for Solar Energy Systems (Fraunhofer ISE) . . . . . . . . . . . . . . . . . . . . 9.3.3 G24 Innovation . . . . . . . . . . . . . . . . . . . . . 9.3.4 3GSolar, Israel . . . . . . . . . . . . . . . . . . . . . 9.4 DSC development studies in various laboratories of Japan . . . 9.4.1 Aisin Seiki Co. Ltd. and Toyota Central R&D Laboratories . . . . . . . . . . . . . . . . . . . . . . 9.4.2 Fujikura Ltd. (Japan) . . . . . . . . . . . . . . . . . . 9.4.3 Peccell Technologies, Inc. (Japan) . . . . . . . . . . . 9.4.4 Sharp Co. Ltd. (Japan) . . . . . . . . . . . . . . . . . 9.4.5 Sony Corporation Ltd. (Japan) . . . . . . . . . . . . . 9.4.6 Shimane Institute for Industrial Technology (Japan). . 9.4.7 TDK Co., Ltd. (Japan) . . . . . . . . . . . . . . . . . 9.4.8 Eneos Co. Ltd. (Japan) . . . . . . . . . . . . . . . . . 9.4.9 NGK Spark Plug Co., Ltd. (Japan) . . . . . . . . . . . 9.4.10 Panasonic Denko Co. Ltd. (Japan) . . . . . . . . . . . 9.4.11 Taiyo Yuden Co., Ltd. (Japan) . . . . . . . . . . . . . 9.4.12 Dai Nippon Printing Company . . . . . . . . . . . . . 9.4.13 Mitsubhishi Paper Mills and Sekisui Jushi Corporation. . . . . . . . . . . . . . . . . . . . . . . 9.4.14 J-Power Co. Ltd. (Japan) . . . . . . . . . . . . . . . . 9.5 DSC Development Work in Korea and Taiwan . . . . . . . . . 9.5.1 Korean Institute of Science and Technology (KIST). . 9.5.2 Electronics and Telecommunications Research

. . . . . . . . . . . . .

. . . . . . . . . . . . .

254 258 258 260 260 260 262 263 263 264 265 266 266

. . 267 . . . .

. . . .

267 268 271 271

. . . .

. . . .

273 278 280 281

. . . . . . . . . . . .

. . . . . . . . . . . .

281 287 290 293 295 296 297 300 301 303 305 306

. . . .

. . . .

306 308 308 308

xiv

Dye-Sensitized Solar Cells

Institute(ETRI), Korea . . . . . . . . . . . . . Samsung SDI, Korea . . . . . . . . . . . . . . Industrial Technology Research Institute of Taiwan (ITRI) . . . . . . . . . . . . . . . . . 9.5.5 J Touch Taiwan . . . . . . . . . . . . . . . . . DSC development work in Australia and China . . . . . 9.6.1 Dyesol, Australia . . . . . . . . . . . . . . . . 9.6.2 Institute of Plasma Physics, Chinese Academy of Sciences . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . 9.5.3 9.5.4

9.6

9.7 9.8 9.9 10

. . . . . . 311 . . . . . . 311 . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

312 313 313 313

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

317 318 319 319

CHARACTERIZATION AND MODELING OF DYE-SENSITIZED SOLAR CELLS: A TOOLBOX APPROACH . . . . . . . . . . . . . Anders Hagfeldt and Laurence Peter 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Theoretical background . . . . . . . . . . . . . . . . . . . . . 10.2.1 Interfacial electron transfer processes in the DSC . . . 10.2.2 Electron trapping in the DSC. . . . . . . . . . . . . . 10.2.3 Electron transport in the DSC . . . . . . . . . . . . . 10.3 The toolbox . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3.1 Determination of injection efficiency and electron diffusion length under steady-state conditions . . . . . 10.3.2 Electrochemical and spectrolectrochemical techniques to study the energetics of the oxide/dye/electrolyte interface . . . . . . . . . . . . . 10.3.3 Electrochemical measurements with thin layer cells. . 10.3.4 Small-amplitude time-resolved methods . . . . . . . . 10.3.5 Methods based on frequency response analysis . . . . 10.3.6 Photovoltage decay . . . . . . . . . . . . . . . . . . . 10.3.7 Determination of density of trapped electrons in DSCs. . . . . . . . . . . . . . . . . . . . . . . . . 10.3.8 Measuring the internal electron quasi Fermi level in the DSC . . . . . . . . . . . . . . . . . . . . . . . 10.3.9 Determining the electron diffusion length using IMVS and IMPS . . . . . . . . . . . . . . . . . . . . 10.3.10 Photoinduced absorption spectroscopy (PIA). . . . . 10.3.11 Conclusions . . . . . . . . . . . . . . . . . . . . . . 10.4 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . 10.5 Appendix 1 Analytical IMPS solutions . . . . . . . . . . . . 10.6 Appendix 2 Numerical solutions of the continuity equation [10.115] . . . . . . . . . . . . . . . . . . . . . . . . 10.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . 323 . . . . . .

. . . . . .

323 324 324 328 331 336

. . 336

. . . . .

. . . . .

342 354 357 362 374

. . 376 . . 383 . . . . .

. . . . .

386 388 395 396 396

. . 397 . . 399

Dye-Sensitized Solar Cells

11

12

DYNAMICS OF INTERFACIAL AND SURFACE ELECTRON TRANSFER PROCESSES . . . . . . . . . . . . . . . . . . . . Jacques-E. Moser 11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Energetics of charge transfer reactions . . . . . . . . . . 11.2.1 Mesoscopic metal oxide semiconductors . . . . 11.2.2 Dye sensitizer. . . . . . . . . . . . . . . . . . . 11.3 Kinetics of interfacial electron transfer . . . . . . . . . . 11.3.1 Charge injection dynamics . . . . . . . . . . . . 11.3.2 Charge recombination . . . . . . . . . . . . . . 11.4 Electron transfer dynamics involving the redox mediator . 11.4.1 Kinetics of interception of dye cations by a redox mediator . . . . . . . . . . . . . . . . . . 11.4.2 Conduction band electron – oxidized mediator recombination . . . . . . . . . . . . . . . . . . 11.4.3 Electron transport in nanocrystalline TiO2 films . 11.5 References . . . . . . . . . . . . . . . . . . . . . . . . .

xv

. . . . . 403 . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

403 406 406 414 416 416 430 440

. . . . . 441 . . . . . 449 . . . . . 450 . . . . . 453

IMPEDANCE SPECTROSCOPY: A GENERAL INTRODUCTION AND APPLICATION TO DYE-SENSITIZED SOLAR CELLS . . . Juan Bisquert and Francisco Fabregat-Santiago 12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 12.2 A basic solar cell model . . . . . . . . . . . . . . . . . . . . 12.2.1 The ideal diode model . . . . . . . . . . . . . . . . 12.2.2 Physical origin of the diode equation for a solar cell 12.3 Introduction to IS methods. . . . . . . . . . . . . . . . . . . 12.3.1 Steady state and small perturbation quantities . . . . 12.3.2 The frequency domain . . . . . . . . . . . . . . . . 12.3.3 Simple equivalent circuits . . . . . . . . . . . . . . 12.4 Basic physical model and parameters of IS in solar cells . . . 12.4.1 Simplest impedance model of a solar cell . . . . . . 12.4.2 Measurements of electron lifetimes . . . . . . . . . 12.5 Basic physical models and parameters of IS in dye-sensitized solar cells . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.5.1 Electronic processes in a DSC . . . . . . . . . . . . 12.5.2 The capacitance of electron accumulation in a DSC . 12.5.3 Recombination resistance . . . . . . . . . . . . . . 12.5.4 The transport resistance . . . . . . . . . . . . . . . 12.6 Transmission line models . . . . . . . . . . . . . . . . . . . 12.6.1 General structure of transmission lines. . . . . . . . 12.6.2 General diffusion transmission lines . . . . . . . . . 12.6.3 Diffusion-recombination transmission line . . . . . 12.6.4 Parameters of the diffusion-recombination model . . 12.6.5 Effect of boundaries on the transmission line . . . .

. . . 457 . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

457 461 461 463 466 467 469 470 480 480 486

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

486 486 488 491 500 508 508 512 515 519 520

xvi

Dye-Sensitized Solar Cells

12.7

Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.7.1 Liquid electrolyte cells . . . . . . . . . . . . . . . . . 12.7.2 Experimental IS parameters of DSCs . . . . . . . . . 12.7.3 Nanotubes . . . . . . . . . . . . . . . . . . . . . . . 12.7.4 Effects of the impedance parameters on the j-V curves 12.8 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . 12.9 Appendix: properties of measured DSCs . . . . . . . . . . . . 12.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

THEORETICAL AND MODEL SYSTEM CALCULATIONS . Filippo De Angelis and Simona Fantacci 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 13.2 Theoretical and computational methods . . . . . . . . . 13.2.1 Density Functional Theory (DFT) . . . . . . . 13.2.2 Basis sets . . . . . . . . . . . . . . . . . . . . 13.2.3 The Car-Parrinello method . . . . . . . . . . . 13.2.4 Solvation effects . . . . . . . . . . . . . . . . 13.2.5 Excited states . . . . . . . . . . . . . . . . . . 13.2.6 Nonadiabatic method . . . . . . . . . . . . . . 13.3 Dye sensitizers . . . . . . . . . . . . . . . . . . . . . . 13.3.1 Ruthenium(II)-polypyridyl sensitizers . . . . . 13.3.2 Calculations on N3 . . . . . . . . . . . . . . . 13.3.3 Calculations on other Ru(II)-dye sensitizers . . 13.3.4 Trans-complexes . . . . . . . . . . . . . . . . 13.3.5 Organic sensitizers . . . . . . . . . . . . . . . 13.3.6 Squaraine dyes . . . . . . . . . . . . . . . . . 13.4 Studies of the TiO2 substrate. . . . . . . . . . . . . . . 13.4.1 TiO2 models . . . . . . . . . . . . . . . . . . 13.5 Dye sensitizers on TiO2 . . . . . . . . . . . . . . . . . 13.5.1 Organic dyes on TiO2: adsorption and electron dynamics . . . . . . . . . . . . . . . . . . . . 13.5.2 Inorganic dyes on TiO2: adsorption and excited states . . . . . . . . . . . . . . . . . . . . . . 13.6 Conclusions and perspective. . . . . . . . . . . . . . . 13.7 References . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

523 523 526 540 543 548 549 550

. . . . . . 555 . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

555 557 557 558 558 558 559 559 560 560 560 563 565 566 567 568 568 574

. . . . . . 575 . . . . . . 579 . . . . . . 588 . . . . . . 589

INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593

PREFACE The last century witnessed an incredible number of technological advances that have changed our lifestyle considerably. The extensive use and growing dependence on electrical and electronic equipment have increased the energy/power requirements on a global scale. With dwindling fossil-fuel reserves, there is an urgent need to find alternative energy resources to meet the growing demand. Alternate energy resources must be efficient, cost-effective and ecologically friendly. The harnessing of solar energy, in this context, becomes a very attractive proposition. The sunlight reaching the earth’s surface every day far exceeds the annual demand. A moderately efficient solar cell array (with 8-10 % efficiency) covering a limited area of the earth’s surface would be able to provide an enormous amount of electric power and thus help reduce greenhouse-gas emissions. Chemists have been interested for a long time in the harnessing of sunlight, either to drive useful chemical transformations or to convert the light directly into electrical energy. Two short publications in Nature by Honda, Graetzel and coworkers have had a dramatic impact on the focus of research for those chemists interested in photochemical and conversion and storage of solar energy (A. Fujishima and K. Honda, Nature, 238, 37 (1972) and B. O’Regan and M. Graetzel, Nature 353, 737 (1991)). The first publication demonstrated the possibility of the photo-decomposition of water into its constituent elements through the irradiation of semiconductor electrodes such as TiO2 immersed in aqueous electrolyte. The second publication described two important variants of this photo-electrochemical cell, specifically the use of high-surface-area mesoporous materials for the oxide substrate, and the application of dye molecules to harvest the sunlight. Both these propositions have proven to be seminal to a new field of scientific research. Interest in the research and development of DSCs is now spread across numerous academic and industrial laboratories. Over six thousand research publications have appeared in the primary scientific literature on the performance features, and the number of patents being filed in this area is growing exponentially (already more than 300 in 2009 for the DSC area alone). The overall solar-to-electrical conversion efficiency has surpassed 10 % for lab-size cells (under areas of 1 cm2) and 8 % for modules (25 - 100 cm2). In recognition of the pioneering contributions made by the Swiss group, the DSC is already referred to as Graetzel Cell. The secondary literature on DSCs (reviews) is rather limited, most often covering the work of specific research groups or conference presentations. The DSC is an important contemporary technology, and one that is rapidly evolving. This monograph presents a comprehensive introduction to this new emerging area. Indeed, the DSC is the outcome of the cross

xviii

Dye-Sensitized Solar Cells

fertilization of concepts used in photovoltaic solar cells and nanoscience, nanotechnology and light-induced electron transfer reactions. Many features of DSC are unique and advantageous over the solar cells based on crystalline or amorphous silicon. Nearly all the components of the DSC are “tunable”, including the semiconducting oxide substrate, the dyes, the electrolytes, the redox mediator and the counter electrode. This has opened great opportunities for chemists and material scientists. Transparency and multi-color design alone offer huge potential for the integration of DSCs as part of the building architecture. The book is organized broadly in two parts. The first half is an overview of the material choices and performance features of all key components of the DSC. The second half covers several experimental techniques that help decipher the functioning of the DSCs in more detail, as well as theoretical calculations that help understand the key parameters that characterize the performance of the solar cells in quantitative terms. Nearly all the mechanistic studies to quantify parameters that control the overall performance of the solar cells are discussed. For completeness, the monograph includes chapters dealing with the scaling-up issues that must be faced to take lab-cell studies that are academic in nature to the commercialization of the technology in the form of large-area solar panels and numerous electronic gadgets. The book benefits from an excellent team of authors, all of whom are experts with long hands-on experience in various aspects of the DSC technology and have made seminal contributions to our understanding on how these solar cells operate. The book is suitable as a text for a one-semester advanced-level course for upper-level undergraduates and graduate students; it will also serve as a reference work for selfstudy for active researchers in the field. In view of the interdisciplinary nature of DSC science, the book should be of interest to those working in the fields of chemistry, physics, material science and engineering. It is a pleasant task to thank all the contributing authors who were kind enough to spare time from their busy schedule to write the chapters and thus share their expertise with the scientific community at large. At a personal level, it has been a great privilege for me to be associated in photochemistry research with Prof. Dr. Michael Graetzel for nearly three decades, sharing both the excitement and agony during the long period as DSCs matured from one of academic curiosity to an important member of the family of “third generation solar cells”, ready for commercialization in the near future. Special thanks also go to Dr. Fred Fenter of the EPFL Press for all his help in putting together this volume. K. Kalyanasundaram Lausanne, February 2010

CHAPTER 1

PHOTOCHEMICAL AND PHOTOELECTROCHEMICAL APPROACHES TO ENERGY CONVERSION K. Kalyanasundaram

1.1

THE SUN AS AN ABUNDANT ENERGY RESOURCE

A striking feature of contemporary society is the life-style based on machines and gadgets that consume power. Currently the energy-requirement (consumption) estimate for seven billion people worldwide is about 13 terawatts (TW) and this is expected to go up by another 10 TW in 40 years time. Available fossil fuel resources are limited and depleting rapidly. Hence, there is increased global awareness concerning the urgent need to find alternative energy resources to meet our requirements. Three viable options are being discussed: carbon-fuel-based sources, nuclear power and renewable sources, such as solar. The main criticism against carbon-based energy is its impact on the environment; its use will lead to a substantial increase in atmospheric CO2 levels, provoking catastrophic climate changes. On the nuclear front, power needs would require hundreds of gigawatt (GW)-level nuclear power stations to be built, and yet no viable method has been found to dispose of the dangerous nuclear fuel wastes. The third choice, that of renewable energy based on the sun as the source is therefore very attractive and promising for several reasons. Sunlight is an abundant energy resource freely available, supplied directly to our home. The amount of solar energy reaching the surface of the earth is 120,000 TW. Even if a small fraction of sunlight could be converted to alternative and usable energy forms, there would be no worry about the energy supply line. In this monograph, we review in depth one specific emerging technology that will permit direct conversion of sunlight to electricity. Devices that do this type of conversion are called photovoltaic solar cells, the term “voltaic” having its origin in the chemical potential differences that occur in materials following absorption of sunlight. After a historical review of various approaches that chemists have examined for the photochemical conversion and storage of solar energy, an overview of the

2

Dye-Sensitized Solar Cells

dye-sensitized solar cell (DSC), the topic of this monograph is presented. The objective here is to provide an overview of various ancient and contemporary photochemical energy conversion technologies studied to place the scope of dye-sensitized solar cells in the right context. Functional features of various key components of the DSC are elaborated to prepare the setting (background) for more in-depth discussions on these by leading experts in the following Chapters.

1.2

PHOTOCHEMICAL CONVERSION AND STORAGE OF SOLAR ENERGY (ARTIFICIAL PHOTOSYNTHESIS)

Scientists have been interested in the idea of harnessing the sun for a long time [1.1-1.7]. They noted several centuries ago that distinct chemical transformations occur when materials are exposed to the sunlight. Much of this interest has been focused on photosynthesis, the process by which green plants utilize sunlight to decompose water to its constituent elements H2 and O2 and subsequently use them to convert atmospheric CO2 to carbohydrates (sugar): 6 CO2 + 6H2O (+ sunlight) → C6H12O6 + 6 O2

(1.1)

The process of photosynthesis takes place in the chloroplasts, using chlorophyll, the green pigment [1.8]. Extensive research effort has been put in to understand and mimic the primary processes of the photosynthesis mentioned earlier. The efficiency of conversion of solar radiation to useful biomass (chemical energy) for photosynthesis is very modest, about 3-6 %. An important aspect of the chloroplast apparatus is the use of two types of chlorophyll molecules (as antenna and reaction centers) in an optimized configuration that maximizes the efficiency of the process at both high and low light conditions. At the turn of 20th century, Giacomo Ciamician, a chemistry professor at the University of Bologna, Italy, was fascinated by the ability of plants to harness sunlight. He was the first scientist to investigate photochemical reactions in a systematic way. He used the open balcony of the building as a “photoreactor” where hundreds of bottles and glass pipes containing various substances and mixtures were exposed to sunrays. At the 1912 meeting of the International Congress of Applied Chemistry, Ciamician proposed replacing “fossil energy (coal)” with the natural solar radiation reaching the earth [1.9]. Approaches to “artificial photosynthesis” can be broadly classified to three types [1.2-1.6]: (i)

homolytic bond fission reactions;

(ii) molecular energy conversion-storage systems; and (iii) light-induced electron transfer reactions. The homolytic fission of a chemical bond (reaction 1.2) AB → A· + B·

(1.2)

Photochemical and photoelectrochemical approaches

3

is a particularly simple photochemical reaction and is always endergonic. The primary products are highly reactive “free radicals”, and they do undergo rapid secondary reactions utilizing full or part of the energy stored in reaction (1.2). Another difficulty is that the absorbed photons must have energy greater than the AB bond energy. For homolytic fission reactions to occur with sunlight photons in the visible (near UV) wavelength region, the AB bond energy must be less than 300 kJ mole−1, which rules out any chemical bond. Only a handful of reactions do fit to the above criteria, e.g., photolysis of NOCl to give NO and 1/2 Cl2 and that of FeBr2. The quantum yields of these reactions are very low and hence of no practical value. Molecular energy storage reactions are those chemical reactions that lead to net molecular energy storage. Formation of new bonds, isomerisation or reorganization of existing bond framework of the molecule in a unimolecular fashion, or in bimolecular reactions with potential substrates, are some examples. The reactions studied so far almost all involve unsaturated organic molecules. The endergonic or energy-storage feature of these reactions often arises from excessive “strain” induced in the product molecule, or loss of resonance energy. One system that has been studied extensively is the photo-isomerisation of derivatives of norbornadiene to quadricyclane (reaction 1.3). Solar energy

(1.3) Norbornadiene (NBD)

Thermal energy (96 kJ/mol)

Quadricyclane (QC)

The above reaction has been shown to be photosensitized using benzophenone or copper halides. High energy requirement of such isomerisation reactions demand photosensitizers that absorb in the UV region. Since the number of such high-energy photons in the solar radiation is very small, overall solar conversion efficiency of such systems tend to be very low, typically less than 1 %. The third type of reactions, referred to as photoinduced electron transfer reactions, involve the transfer of one or more electrons between two reactants following the absorption of light by one: D + A (+ sunlight) → D+ + A−

(1.4)

From now on we use the terms dye (D) and photosensitizer (S) for those reactants in a multi-component system that absorb the light and initiate the energy-conversion processes. Either the donor or acceptor can act as the light absorber, or light absorption is achieved by a third component. Absorption of light by molecules and complexes S raises them to higher electronically excited state S*, where the light energy is transformed and stored in the form of enhanced reactivity in the electronically excited state S*. Numerous studies have established that molecules in the electronically excited state S* are distinct, with their own structural and electronic properties and enhanced reactivity [1.10-1.12]. The molecule in the excited state S* readily undergo

4

Dye-Sensitized Solar Cells

bimolecular electron-transfer reactions with suitable donors or acceptor molecules, as shown in the reactions given below: S* + A → S+ + A−

(1.5)

S* + D → S + D

(1.6)



+

With reference to the sensitizer S that absorbs the light, reactions (1.5) and (1.6) are labeled as oxidative and reductive quenching of the excited state. Light-induced electron-transfer reactions are also referred to as photoredox reactions. Photoredox reactions of this type convert a major part of the light energy of the absorbed photon into chemical energy stored in the products. The electron-transfer products, rich in energy, have a tendency to undergo back-electron reactions, resulting in rapid reestablishment of the ground states of the reactants S and D or A: S+ + A− → S + A + heat

(1.7)

S− + D → S + D + heat

(1.8)

Weller, Mataga and coworkers first demonstrated the occurrence of this kind of excited-state electron transfer in organic molecules in the 1960s. The pioneering work of Rehm and Weller established free-energy relationships to quantitatively explain rates and efficiencies of photo-induced electron-transfer reactions. The Rehm-Weller relationship is used widely to explain orders of magnitude variation in the rate constants of electron-transfer processes and yields of redox products in terms of the driving force of the reactions of interest. The reason quenching rate constants approach a plateau value at the diffusion controlled limit at large driving force (exothermic) is explained within Marcus theory as due to behavior in the inverted region. Soon thereafter, Balzani, Sutin, Meyer and others demonstrated similar photoredox processes in numerous transition metal complexes. The most well studied paradigm is the tris(2,2′bipyridine) complex of Ru(II), Ru(bpy)32+ and their variants [1.13]. A long-term objective of research into photochemical redox reactions is to obtain overall generation of fuels such as H2, CH4, and CH3OH. Table 1.1 lists a Table 1.1 Some of the chemical reactions with net storage energy [1.2–1.6]. Reaction 1 H2O(l) → H2(g) + O2 (g) 2 1 CO2(g) → CO(g) + O2(g) 2 1 CO2(g) + H2O(l) → HCOOH (l) + O2(g) 2 CO2(g) + H2O(l) → HCHO(g) + O2(g) 1 CO2(g) + 2H2O (l) → CH3OH(l) + O2(g) 3 CO2(g) + 2H2O(l) → CH4(g) + 2O2 (g) 1 N2(g) + 3H2O(l) → 2NH3(g) + O2 (g) 2 CO2(g) + H2O(l) → 1/6 C6H12O6(s) + O2 (g)

# Electrons

ΔE (V)

2

1.23

2

1.33

2

1.48

4

1.35

6

1.21

8

1.06

6

1.17

4

1.24

Photochemical and photoelectrochemical approaches

5

number of target chemical reactions that generate fuels via light-induced electrontransfer processes. Within the framework of “artificial photosynthesis”, a reaction of utmost importance is the decomposition of water to its elements H2 and O2 and reduction of CO2. Most of the reactants listed are transparent to solar radiation. Hence one needs to use an “external” photosensitizer to achieve the overall conversion. A second serious problem in achieving the listed reactions is the “multi-electron transfer” nature of these processes. Without the use of a suitable “catalyst” to mediate, the desired products are not obtained in reasonable yield. Water photolysis to its constituents H2 and O2 is the Holy Grail for chemists working in this field. But there are other solar-chemical conversions that are less challenging. For example decomposition of HX (X = Br, I) to H2 and X2 using solar radiation would be industrially more beneficial than the decomposition of water.

1.3

PHOTOGRAPHIC SENSITIZATION

Becquerel laid the foundations of the field of photoelectrochemistry way back in 1839 with his observation of measureable current passing between two platinum electrodes in the presence of sunlight when the electrodes are immersed in an electrolyte containing metal halide salts [1.14]. Moser reported on photosensitization effects in Silver halide grains in 1877 [1.15]. Silver halide-based photography since then has evolved as the biggest application of photosensitization phenomenon with several billion dollars global market (until the recent development of digital photography) [1.16]. A typical photographic film contains tiny crystals of very slightly soluble silver halide salts such as silver bromide (AgBr), commonly referred to as “grains.” The grains are suspended in a gelatin matrix and the resulting gelatin dispersion (commonly referred to as an “emulsion”) is melted and applied as a thin coating on a polymer base or, as in older applications, on a glass plate. When light or radiation of appropriate wavelength strikes one of the silver halide crystals, a series of reactions begins that produces a small amount of free silver in the grain. Initially, a free bromine atom is produced after the bromide ion absorbs the photon (reactions 1.9 and 1.10): AgBr (crystal) + hν (radiation) → Ag+ + Br + e−

(1.9)

The silver ion can then combine with the electron to produce a silver atom. Ag+ + e− → Ag0

(1.10)

Ag3+, Association within the grains produces aggregated species such as 0 Ag 4. The free silver produced in the exposed silver halide grains constitutes what is referred to as the latent image, which is later amplified by the development process. The grains containing the free silver in the form of Ag04 are readily reduced by chemicals referred to as developers, forming relatively large amounts of free silver; the deposit of free silver produces a dark area in that section of the film. The developer under the same conditions does not significantly affect the unexposed grains. Once the developed image is obtained, a large amount of unexposed and undeveloped silver halide remains in the emulsion. If that silver halide is not removed

Ag30, Ag+4 , and

Ag+2 ,

Ag02,

6

Dye-Sensitized Solar Cells

before the image is exposed to radiation capable of producing a latent image, the image will continue to darken. The process of removing the residual silver halide from the image is called fixing. The silver halides are only slightly soluble in water; therefore, to remove the material remaining after development it is necessary to convert it to soluble complexes which can he removed by washing. Sodium thiosulfate, commonly termed hypo, has been used for this purpose since 1839. The radiation sensitivity of silver halides ends for all practical purposes at about 525 nm. The sensitivity of the silver halides may be extended to radiation of longer wavelengths by the addition of dyes or “color sensitizers.” Development of digital cameras at low cost without compromise on the quality has reduced the practice of classical photography to the professionals.

1.4

PHOTOELECTROCHEMICAL CONVERSION OF SOLAR ENERGY

Photoelectrochemical solar cells (also called photovoltaic cells) are designed to convert solar radiation directly to electricity. Once stored, electrical energy, as current, can be used for many different electrical appliances, including electrolysers for oxidation or reduction of chemicals. Photovoltaic cells are the most efficient routes to solar-energy conversion and storage. Herein we review some of the approaches that have been examined in the past: Photogalvanic cells using metal electrodes, photovoltaic solar cells based on semiconductor electrodes (solid-state devices) and liquid-junction photoelectrochemical cells where the semiconductor electrodes are immersed in redox electrolytes. Even though practical applications of photogalvanic and “wet” photoelectrochemical systems have not been realized until this date, they have helped identify critical factors that are to be controlled for all successful applications. The best solar-conversion efficiency for single junction (bandgap) and multijunction solar cells are obtained in solar cells made up of ultra-pure materials.

1.4.1

Photogalvanic cells

Soon after the recognition that exposure of certain chemicals to sunlight can cause oxido-reduction in the 19th century, attempts have been made to capture the energy stored in such electron-transfer processes. A simple approach is to introduce two metal electrodes in a solution of suitable “dye” with an electron acceptor and try to capture the electron transfer products induced by light, prior to their recombination: D+ + e− → D (cathode)

(1.11)

A− → A + e− (anode)

(1.12)

Generation of potential difference is known as “galvanic effect” and hence the term photogalvanic cell is used to describe such photo-electrochemical devices [1.17, 1.18]. Figure 1.2 shows schematically the principles of operation of a photogalvanic cell. The net effect of such mediated reduction would be the driving of the electron through an external load and hence the overall conversion of light to electricity. As

Photochemical and photoelectrochemical approaches

7

A

D D

A hν

Fig. 1.2 Schematic presentation of a photogalvanic cell.

early as 1940, Rabinowich proposed such photogalvanic cells for solar energy conversion to electricity [1.17]. One extensively studied example of photogalvanic cell is the photolysis of thionine and similar xanthene dyes in the presence of ferric (Fe3+) ions in an electrochemical cell using two Pt electrodes. Calculations by Albery et al. [1.18] showed that it is possible to obtain an overall rate for conversion of light energy to electricity in the 5-9 % range. However in spite of focused efforts over several decades, the best efficiency obtained was only 0.03 %. Two key factors are responsible for the observed low efficiency. First, the metal electrodes employed did not have any selectivity for the reaction desired (reduction of D+ at the cathode and not the oxidation of A− or vice versa for the anode). Indiscriminate electrode processes at the metal electrodes effectively reduce the number of electrons that can flow over the external circuit. Secondly, only a small fraction of the dye present in the vicinity of the cathode contributed to the measured photocurrent. Electron-transfer products formed in the bulk of the solution underwent recombination before reaching the two electrodes. There are also additional problems such as self-quenching of the excited state in concentrated dye solutions, limited solubility, and thermal- and photo-instability during extended photolysis. Time-resolved studies showed that the rate of back electron transfer was much faster than the rate in the forward direction. This results in only a small shift of the equilibrium towards products and hence the amount of scavengeable high energy products. To a limited extent, using scavenging reagents it is possible to intercept back-reactions kinetically. Attempts to improve the selectivity of the two electrodes and thinner path length cells produced only marginal improvements. Though attempts of solar light conversion to electricity using photogalvanic cells did not yield meaningful results, studies nevertheless helped identify key factors that are to be controlled if photoredox reactions of dye molecules are to be viable light-energy-conversion systems. We will return to this subject below.

1.4.2

Generations of photovoltaic solar cells

The need for power in the outer space to run communication and military satellites provided NASA and major American industries extensive funding to develop highly efficient solar-to-electrical conversion devices based on semiconductor electrodes.

8

Dye-Sensitized Solar Cells

Load and available-area constraints demand solar panels for satellites to have the highest possible conversion efficiency, even if the costs are very high [1.19-1.31]. The understanding of photo-processes involving semiconductors require, in turn, an understanding of the primary mechanisms of charge-carrier generation and mobility in these materials. Herein we briefly review the very basic points; the reader is referred to specialized monographs for an in-depth discussion. In a bulk crystalline semiconductor, the highest occupied and lowest unoccupied molecular orbitals (HUMO and LUMO) of constituent atoms or molecules converge into valence and conduction bands. In the absence of dopants, the energy level (Fermi level) of the semiconductor lies half-way between the separation gap of the valence and conduction bands. Doping with electron-donors (n-doping) makes the material electron-rich, and the Fermi level moves closer to the conduction band. Similarly, doping with electronacceptors (p-doping) depletes the number of electrons available and the Fermi level moves closer to the valence band. Optical excitation of the semiconductor with light of energy higher than the bandgap separation of the semiconductor leads to generation of free charge carriers, electrons (e−) and holes (h+). In a sandwich structure composed of an n-doped and p-doped semiconductor, charge separation occurs due to bending of the bands in the vicinity of the interface (see Fig. 1.3). With light, additional carriers are created, and the single Fermi level splits into two quasi-Fermi levels in the n-type or p-type region respectively. These quasi-Fermi (1)

p-type

n-type

electron hole (2)

depletion layer

(3)

electric field diffusion potential

EC Ef Ev

(4) Vo

EC Ef light Ev

Fig. 1.3 Band Picture of n- and p-type semiconductors with the indication of the Fermi level (Ef) before (scheme 1) and after joining (scheme 2) resulting in a p-n junction.

Photochemical and photoelectrochemical approaches

9

levels are now split; the higher the light intensity the more they split. Close to the electrode both quasi-Fermi levels collapse toward the majority quasi-Fermi level, where they are connected. This shift of the Fermi levels in the electrodes represents the open circuit voltage, which can be approximated by the shift of the minority quasi-Fermi levels. Such separation of the charge carriers permits selective collection at the collector electrodes and a net conversion of sunlight to electric power. Based on the nature of the material, maximum conversion efficiency obtainable, and the associated cost of photovoltaic power, Martin Green has grouped various photovoltaic solar cells in three major categories (see Fig. 1.4). First generation Photovoltaics use the highest purity materials with least structural defects (such as single crystals). The highest power-conversion efficiencies obtained to date are in first generation PVs. Due to high labor costs for the material processing and the significant energy input required, cost per watt is also the highest. It is very unlikely that these systems will allow photoelectric power conversion for less than US $1/watt. In addition to single component Si and layered semiconductors, binary semiconductors of II-IV and III-V have been examined. Second generation devices are based on low-energy, intensive preparation techniques such as vapor deposition and electroplating. Since it is difficult to prepare systems without defects, maximum power conversion is lower. Nearly all thin film photovoltaics fall in this category, and the power cost can be less than 1 $US/watt. Most efficient examples are solar cells made up of multi-crystalline or amorphous US 0.10$

US 0.20$ US 0.50$

Max. thermodynamic limit (multijunction or 3G)

Efficiency (%)

80

60 US 1.00$

40 Single Junction max US 3.50$

20

0

100

200

300

400

500

Cost (US $/m2) Fig. 1.4 Classification of photovoltaic solar cells into three categories, based on the nature of the materials used and associated cost of electric power generation.

10

Dye-Sensitized Solar Cells

Si, CdTe and Cd-In-Ga-Se (CIGS). Table 1.5 provides a summary of the state-of-the art conversion efficiency reported for various semiconductor-based solar cells of the first and second generations. These materials are applied as a thin film to a supporting substrate such as glass or ceramics, reducing material mass and therefore costs. There have been several theoretical calculations on maximum power conversion obtainable using solar radiation. The most popular calculation is that of Shockley and Queisser. Considering photovoltaic solar cells as a one-photon giving one-electron threshold device, these authors estimate 31 % as maximum under 1 sun illumination and 40.8 % under maximal concentrated solar light (46,200 suns). During the past decade several approaches have been suggested to cut down the energy losses and increase the overall conversion. In one classification, all photovoltaic systems that can potentially give power conversion efficiency over and above the Shockley and Queisser limit are labeled as third generation photovoltaics. Advances in our understanding of solid-solid and solidliquid interfaces of various kinds permit now usage of wide variety of quasi-crystalline and even amorphous materials made out of monodispersed colloids, polymers, gels and electrolytes. Since there is excellent potential for these photovoltaic systems, based on Table 1.5 Confirmed solar conversion efficiency for various photovoltaic systems (single junction), measured under AM 1.5 (100 mW/cm2) at 25 °C. [31]. PV material

Efficiency (%)

Area (cm2)

VOC (V)

Si (crystalline) Si (multicryst.) Si (thin film) Si (thin film) GaAs (cryst.) GaAs (thin film) GaAs (multicryst.) InP (cryst.) CIGS (cell) CIGS (sub-module) CdTe (cell) Si (amorph.) Si (nanocryst.) Dye-sensitized (sub-module) Dye-sensitized (sub-module) Dye-sensitized (sub-module) Org. polymer Organic (sub-module)

25.0 ± 0.5 20.4 ± 0.4 16.7 ± 0.4 10.5 ± 0.3 26.1 ± 0.8 26.1 ± 0.8 18.4 ± 0.5 22.1 ± 0.7 19.4 ± 0.6 16.7 ± 0.4 16.7 ± 0.5 9.5 ± 0.3 10.1 ± 0.2 10.4 ± 0.3

4.00 1.002 4.017 94.0 0.998 1.001 4.011 4.02 0.994 16.0 1.032 1.070 1.199 1.004

8.2 ± 0.3 8.2 5.15 1.1

ISC (ma/cm2)

FF (%)

Lab, year

0.705 0.664 0.645 0.492 1.038 1.045 0.994 0.878 0.716 0.661 0.845 0.859 0.539 0.729

42.7 38.0 33.0 29.7 29.7 29.5 23.2 29.5 33.7 33.6 26.1 17.5 24.4 22.0

82.8 80.9 78.2 72.1 84.7 84.6 79.7 85.4 80.3 75.1 75.5 63.0 76.6 65.2

Sandia 1999 NREL 2004 FhG-ISE 2001 FhG-ISE 2007 FhG-ISE 2007 FhG-ISE 2008 NREL 1995 NREL 1990 NREL 2008 FhG-ISE 2000 NREL 2001 NREL 2003 JQA 1997 AIST-Sharp 2005

25.45

0.705

19.1

61.1

AIST–Sharp 2007

18.50

0.659

19.9

62.9

AIST-Sony 2008

62.5 51.2

NREL 2006 NREL 2008

1.021 0.876 232.8 29.3

9.39 0.072

FhG-ISE: Fraunhofer Institute for Solar Energy Systems in Freiburg, Germany; NREL: National Renewable Energy Laboratory, Golden CO, USA; AIST: National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan; JQA: Japanese Quality Assurances Association.

Photochemical and photoelectrochemical approaches

11

novel materials and nanotechnology, to deliver solar electric power at very low costs (0.1-0.5 $US/watt), solar cells based on dye-sensitization, polymer organic-bulk heterojunctions, and quantum dots are also referred to as third-generation PV systems. Third generation solar cells for the highest possible conversion Examination of the loss mechanisms that lead to the Shockley and Queisser limit in single-junction solar cells leads to the identification of two effects: the non-usage of the photons of energy below the bandgap, and the thermalization losses that occur with photons of energy much higher than the bandgap. Based on theoretical analysis, several approaches have been proposed to overcome these: hot carrier cells, up/down convertors, multiple excitonic charge-carrier generation and multi-junction (tandemtype) solar cells. The hot carrier cell tackles the major PV loss mechanism of thermalisation of carriers. The underlying concept is to slow the rate of photoexcited carrier cooling, caused by phonon interaction in the lattice, to allow time for the carriers to be collected whilst they are still “hot”, thus enhancing the voltage of a cell. Luminescent materials are being investigated that either absorb one high-energy photon and emit more than one low-energy photon just above the band gap of the solar cell (downconverters); the other approach is a material that absorbs more than one low-energy photon below the band gap of the cell and emits one photon just above the band gap (up-converters). The important property for these devices is high quantum efficiency, meaning that they must be very radiatively efficient. Multiple excitonic charge-carrier generation refers to the formation of multiple excitons per absorbed photon, which can happen when the energy of the photon absorbed is far greater than the semiconductor band gap. This phenomenon does not readily occur in bulk semiconductors, where the excess energy simply dissipates away as heat before it can cause additional electron-hole pairs to form. In semi-conducting quantum dots, the rate of energy dissipation is significantly reduced, and the charge carriers are confined within a minute volume, thereby increasing their interactions and enhancing the probability for multiple excitons to form. A quantum yield of 300 percent has been reported, for example, for 2.9-nm-diameter PbSe (lead selenide) quantum dots when the energy of the photon absorbed is four times that of the band gap. But multiple excitons start to form as soon as the photon energy reaches twice the band gap. Quantum dots made of lead sulphide (PbS) have also shown the same phenomenon.

1.4.3

Photoelectrochemical solar cells with liquid junctions

Chemists have studied the behavior of semiconductor electrodes immersed in electrolytes containing suitable electron donor or acceptor molecules. As in the case of the semiconductor, the electrolyte solution has its own chemical potential defined by the nature and composition of additives. In the situation of physical contact, a spacecharge layer is built up and the Fermi levels of the semiconductor are influenced, leading to “band bending”. In the case of a photoanode, band-bending in the depletion region drives any electron that is promoted into the conduction band into the interior of the semiconductor, and holes in the valance band are driven toward the electrolyte,

12

Dye-Sensitized Solar Cells

Load

Load

e–

e–

OX

e– hν

OX

Red



Electrolyte n-type

e–

Red Electrolyte

CE

p-type

CE

Fig. 1.6 Principles of operation of liquid-junction solar cells based on n- and p-type semiconductor electrodes.

where they participate in an oxidation reaction. Electrons through the bulk drive an external load before they reach the counter electrode or storage electrode, where they participate in a reduction process. Figure 1.6 shows schematically various electron transfer steps involved. Under illumination and open-circuit conditions, a negative potential is created at a photoanode, and as a result the Fermi level for the photoanode shifts in the negative direction, thus reducing the band-bending. Under illumination with increasing intensity, the semiconductor Fermi level shifts continually toward negative potentials until the band-bending effectively reduces to zero, which corresponds to the flat-band condition. At this point, a photoanode exhibits its maximum photovoltage, which is equal to the barrier height. With the help of a second electrode, it is possible to affect the regeneration of the mediator oxidized (or reduced) at the illuminated semiconductor. The net effect of the illumination in the photoelectrochemical cell is conversion of light energy to electrical energy in two steps, initially as modified chemical potential of the electrolyte and then as electric power (electrons driven through the external circuit). Under such conditions, the redox mediator is recycled and the electrochemical photovoltaic cell is referred to as regenerative solar cell. These photovoltaic cells are also referred to as liquid-junction solar cells to differentiate them from pure solid-state devices. There have been a number of studies on the mechanistic aspects of regenerative solar cells. Pioneering work in this area has been by Gerischer, Memming, Bard, Wrighton, Heller, and their coworkers [1.32-1.35]. Early studies of regenerative solar cells were carried out in aqueous media. A serious problem often faced by those who use semiconductors is the rapid photodecomposition of the semiconductor in aqueous electrolytes. Photogenerated holes are strong oxidants capable of oxidizing the semiconductor itself (photocorrosion) or oxidizing water, resulting in the formation of a thin oxide-insulating layer at the electrolyte interface. Both the thermodynamic and kinetic factors involved in stability of the semiconductor have been investigated. The problem appears more acute with n-type materials, where the photogenerated holes, which move to the interface, are

Photochemical and photoelectrochemical approaches

13

capable of oxidizing the semiconductor itself. For example, with ZnO in an aqueous solution at pH 0 the half-reaction (E0D = + 0.9 V vs. NHE) ZnO + 2h+ → Zn2+ + 1/2 O2

(1.13)

can occur readily with holes produced at the potential of the valence band edge (−3.0 V vs. NHE). Thus irradiation of a ZnO electrode in an aqueous solution will cause at least partial decomposition of the semiconductor electrode. The II-IV group semiconductors of the chalcogenide family is also susceptible to serious photocorrosion problems. The decomposition potential for CdX (X = S or Se) has been determined to be within the bandgap. Thus photocorrosion becomes a thermodynamically favoured reaction for photogenerated holes (reaction 1.14): CdX + 2h + → Cd2+ + X

(1.14)

Attempts were made to stabilize the semiconductor photoelectrodes with nonaqueous solvents in order to dissolve the electrolyte and redox species. Early work in non-aqueous solvents, such as acetonitrile and methanol have resulted in cells with low efficiencies for several reasons, including the higher resistivity of the solventelectrolyte; the limited solubility of redox species; and the poor bulk and surface properties of the semiconductor. However, the coating of the semiconductor with a layer of conducting polymer has been found to be effective. For the cadmium chalcogenides, photocorrosion may be suppressed by kinetic competition in the presence of efficient hole scavengers, such as polysulfide and ferrocyanide. In the absence of such species, photocorrosion is the only reaction. Formation of Cd2+ and Se has been reported and verified by various experimental techniques. In aqueous KOH, Cd(OH)2, Se, SeO32−, and SeO42− have to be considered as potential corrosion products. Table 1.7 gives representative data on the light-conversion efficiency of regenerative solar cells based on single-crystal electrodes. Polycrystalline materials are important to consider due to their potentially low fabrication and materials costs. However, trapping or recombination of charge carriers Table 1.7 Light-conversion efficiency of regenerative solar cells based on semiconductor electrodes immersed in redox electrolyte solutions. Semiconductor

n-GaAs p-InP n-GaAs0.72P0.28 n-WSe2 n-CuInSe2 n-CuInSe2 n-MoSe2 n-CdSe n-WS2 n-CdSe

Aqueous redox electrolyte

1 M K2Se, 0.01 M K2Se2, 1 M KOH 0.3 M V3+, 0.05 M V2+, 5 M HCl 1 M K2Se 1 M KI, 0.01 M KI3 6 M I−, 0.1 M Cu2+, 0.1 M In3+ I 3− / I − mixture, Cu+ 1 M KI, 0.01 M KI3 1 M Na2S2, 1 M NaOH 1 M NaBr, 0.01 M Br2 Fe(CN)64+

Solar conv. efficiency (%)

Stability (C/cm2 )

Ref.

12.0

35,000

[1.36]

11.5 11.0 10.2 10.1 9.5 9.4 7.2 6.0 12.4

27,000 3000 40,0000 15,000 70,000 50,000 20,000

[1.37] [1.38] [1.39] [1.40] [1.41] [1.42]

unstable

14

Dye-Sensitized Solar Cells

occurs readily at defect sites, causing a dramatic decrease in light-energy conversion. The magnitude of the anodic photocurrent has been found to depend upon a number of experimental variables; light intensity, electrolyte concentration, surface roughness and solution pH all influence the rate of charge injection. Best results are obtained with the use of single crystals. Most often they are even etched to take away few layers just prior to measurements. Texas Instruments has developed [1.43] a large-scale solar-energy chemicalconvertor SCC (TISES) for the photo-electrolysis of HBr to Br2 and H2. The project has an associated cost of US$ 1000 million and uses Si-based solar cells to generate electricity, which is then used to run a classical electrolyser. Solar-to-chemical efficiencies of 8.6 % have been both predicted and measured for the electrolysis of 48 percent HBr to hydrogen and bromine by a full anode/cathode array. An individual cathode solar-to-hydrogen efficiency of 9.5 % has been obtained. Semiconductor surfaces have been modified to protect low-band-gap materials against photocorrosion [1.44, 1.45]. A self-driven photoelectrochemical cell consisting of Pt-coated p-InP and Mn-oxide-coated n-GaAs has been demonstrated to operate at 8.2 % maximum efficiency to generate H2 and O2 under simulated sunlight [1.46]. More recently, a two-band gap cell in a tandem arrangement has been used to split water at 12 % efficiency [1.47]. A multi-junction GaAs-Si cell has been recently used to drive water splitting with over 18 % solar-to-electrical conversion efficiency [1.48].

1.4.4

Photoredox reactions of colloidal semiconductors and particulates

Since the 1970s, there have been numerous studies of light-induced electron-transfer processes involving finely divided semiconductor materials such as colloids or particulates in aqueous media. In contrast to bulk semiconductor-electrode-based systems, both forms of photo-generated charge carriers (electrons and holes) reach the surface and are available for suitable electron-transfer processes. Figure 1.8 schematically illustrates the situation with the colloids and particulate systems. Most often, aerated (or oxygenated) solutions are employed, where the electrons are rapidly scavenged by molecular oxygen to form peroxides. Particulate systems have been studied as a potential low cost and efficient means of degrading toxic industrial wastes. Advances in colloid and sol-gel chemistry permit facile synthesis of well-defined monodisperse colloids of various semiconductors. Their translucent nature beyond the band– A/A–

– et

CB hυ +

+

ht

D/D+

VB

Fig. 1.8 Schematic illustration of possible electron-transfer processes in semiconductor colloid and particulate systems.

Photochemical and photoelectrochemical approaches

15

gap absorption permit detailed time-resolved studies of excited-state charge transfer involving valence-band holes or conduction-band electrons. In these systems involving finely divided semiconductors, such as monodispersed colloids and particulate systems, it is known that reducing the size to few nanometers has important and dramatic consequences. A priori, finely divided semiconductors may appear as “potentially very inefficient systems”, given the serious limitations identified in studies with single-crystal electrodes in terms of their trap and defect sites. In particulate systems, the available surface area increases by several orders of magnitude, and hence there is a distinct possibility that this huge enhancement of the surface area could be counter-productive. In reality, it has been found that light-energy conversion processes can indeed be efficient in these finely divided systems. Indeed, there are many distinct differences in the details of the evolution of charge-carrier transport. Most of these arise from the fact that the smaller, finely divided particles are of comparable dimension with respect to critical parameters that control the efficiency, such as charge-carrier diffusion length. Thus, the serious limitation of recombination of charge carriers in the bulk of the semiconductor is nearly non-existent in nano-sized systems. It is possible to scavenge quantitatively all the photogenerated charge carriers. Titanium dioxide TiO2 is a low-cost readily available semiconductor with a bandgap of 3.2 eV. When the particulates of this semiconductor are illuminated with light of wavelength less than 385 nm in aerated solutions, hydroxyl radicals (•OH) and superoxide ions (O2−) are produced as the net reactants from the photogenerated holes and electrons respectively. Both these are strong oxidants capable of oxidizing other organic and inorganic molecules present in the solution. This process of heterogeneous catalytic photooxidation has been developed extensively as an industrial process for the decontamination of toxic wastes and pollutants [1.49-1.59]. This area is known as heterogeneous photocatalaysis. Systems based on this approach are used to remove or destroy low-level pollutants in air and water. The oxidation potential of hydroxyl radical (•OH) is 2.8 V relative to the normal hydrogen electrode (NHE), much higher than that of other substances used for water disinfection: ozone (2.07 V), H2O2 (1.78 V), HOCl (1.49 V) and chlorine (1.36 V). Employed as a heterogeneous catalyst, titanium dioxide is readily recoverable for recycling. Using aqueous colloidal suspensions of semiconductors, Grätzel, Kamat and coworkers studied mechanistic details of interfacial electron-transfer processes of semiconductors with various redox reagents present in the solution [1.60-1.65]. With excitation by a short laser pulse of energy greater than the bandgap energy of the semiconductor, it is possible to generate charge carriers in the colloidal particles and follow their dynamics via time-resolved measurements. These studies have also been extended to dye sensitization of semiconductor colloids. Dynamics of charge injection from the excited state of various xanthene and porphyrin dyes have been probed. With dyes that are ionic, electrostatic forces (attraction or repulsion) play important roles in determining the degree of association of the dye to the semiconductor surface. Oxide semiconductors are amphoteric (due to the presence of large number of ionizable groups OH−/O−/OH2+) and the overall charge on the particle can be positive or negative depending on the solution pH with respect to the point of zero charge (zeta potential). In some aspects, these fundamental studies of electron-transfer processes

16

Dye-Sensitized Solar Cells

involving colloidal semiconductors have laid the foundation for the adoption of these processes into a viable dye-sensitized solar cell.

1.5 1.5.1

DYE SENSITIZATION OF SEMICONDUCTORS Dye sensitization of bulk semiconductor electrodes

Sensitization of large band-gap semiconductors is a logical extension of the numerous studies made earlier on the fundamentals of the photographic process, on photogalvanic cells and on light-energy conversion using liquid-junction or regenerative solar cells. Studies of dye sensitization of bulk semiconductor electrodes in turn have laid the foundations for the development of dye-sensitized solar cells. Numerous authors have contributed to our understanding, but seminal contributions have been made in particular by Gerischer, Calvin, Tributsch, Willig, Spitler, Parkinson, and coworkers [1.66-1.73]. In this work, metal-oxide semiconductors such as ZnO, TiO2, SnO2, In2O3, and SrTiO3 were sensitized with ruthenium polypyridyl complexes or organic dyes such as rhodamine B, rose bengal (xanthenes), fluorescein, and alkylthiacarbocyanines. Through extensive studies of the charge-injection processes under different dye conditions, mechanistic details have been established. In particular, luminescence studies on dyes adsorbed onto semiconductor electrodes have shown that the excited states could be efficiently quenched on these surfaces. Photosensitization in general can occur via transfer of the excitation energy to a suitable state/energy level of the acceptor or by electron transfer. With semiconductors, oxidation of the dye takes place through transfer of an electron from a molecule’s excited energy level to the conduction band of the semiconductor. In an electrochemical cell using semiconductor as bulk electrodes, the excited-state charge injection manifests itself as photocurrents, measurable quantitatively under anodic polarisation. Reduction of the excited state of the dye is also known to occur through a valence-band mechanism, requiring cathodic polarisation of the electrode for detection. This process has been observed with semiconductors with high hole mobility, such as GaP or SiC. Tributsch, Gerischer and Calvin pioneered the field when they examined photosensitization of ZnO using chlorophyll derivatives as a model system for the primary process in photosynthesis. As part of his doctoral thesis work, Spitler later studied the excited-state charge injection of rose bengal onto a single crystal TiO2 (rutile) electrode [1.64]. A quantum efficiency of 4 × 10−3 was measured for the electron injection from the excited rose bengal dye to the conduction band of the semiconductor. Parkinson has extended these studies to oxide surfaces of low-index faces of the anatase and rutile forms of TiO2 by covalently attaching dyes to the surfaces. A frequent observation made in the study of the dye sensitization of photocurrent at semiconductor electrodes is that the quantum efficiency for conversion of absorbed photons to electrons is low, usually on the order of a few percent. In several recent examinations of this problem, it was concluded that inefficient sensitization could be attributed to states at the electrode surface that facilitate the return of the transferred electron from the solid to the oxidized dye layer, thus quenching the production of

Photochemical and photoelectrochemical approaches

17

photocurrent. In some cases these states have been attributed to hydrolyzed surface of oxide electrodes such as ZnO or SnO2. It is fairly clear from these examples that any surface layer on a semiconductor electrode could lead to the efficient quenching of the photocurrent through a back reaction. Such surface layers are absent at electrodes made of the group-VI dichalcogenides. The van der Waals (001) surfaces of these layered semiconductors do not oxidize nor interact strongly with solvents, and therefore they provide an abrupt interface between the electronic states of an adsorbed dye and the energy bands of the semiconductor. The saturation of the bonding on these surfaces also prevents any chemical reactions or hydrogen bonding interactions with the surface. The van der Waals surfaces of the layered semiconductors n-WS2 and n-WSe2 can be sensitized with high quantum yields with the infrared absorbing thiapentacarbocyanine dye. A quantum yield of electrons per absorbed photon of 0.6-0.8 has been measured [1.721.73]. The surface-dye concentration dependence of the sensitized photocurrent has also been studied in the presence and absence of the supersensitizer hydroquinone. Adsorbed dye aggregates could be identified and selectively photooxidized. This has yielded sensitized photocurrent densities in excess of 40 µA/cm.

1.5.2

Dye-sensitized solar cells – an overview

In this section we present a broad overview of the dye-sensitized solar cell (DSC), covering the basic principles of operation and the various key components that are currently being optimized with respect to solar-to-electrical conversion [1.74-1.91]. Figure 1.9 shows schematically the basic architecture. A DSC is basically a thin-layer solar cell formed by sandwich arrangements of two transparent conducting oxides (TCO) electrodes. The main highly colored electrode has a few-micron-thick mesoporous TiO2 layer coated with a photosensitizer. The counter-electrode is composed of islands of finely divided Pt deposited onto another TCO. The inter-layer space is filled with an organic electrolyte containing a redox mediator, usually a mixture of iodine and iodide in a low viscosity organic solvent such as acetonitrile. Best solar conversion efficiency obtained for this type of DSC is in the range of 11-12 % for laboratory scale cells (area < 1 cm2) and around 8.5 % for large-area modules (100 cm2). Glass/TCO (CE) Pt layer Redox electrolyte

2–6 mm

e

Dye-coated TiO2



Glass/TCO (WE) Fig. 1.9 A schematic representation of a dye-sensitized solar cell (DSC).

18

Dye-Sensitized Solar Cells

h+

Au

Au

Dye-coated TiO2

200 nm

1mm e– TiO2 underlayer

Glass/TCO

Fig. 1.10 A schematic representation of a quasi-solid-state version of the dye-sensitized solar cell using a hole-transport material co-deposited onto the mesoporous oxide layer.

A quasi-solid state variation of the DSC has also been developed. This uses an organic hole transport material, such as a triarylamine, to transport charges between the photoanode and the counter-electrode. Figure 1.10 shows schematically the organization of various components in the quasi-solid-state version. The hole transporter can be an inorganic p-type semiconductor such as CuSCN or an organic donor molecule. They can be considered as intermediate between the DSCs based on organic electrolytes and polymer organic solar cells using bulk heterojunctions. To avoid the undesirable situation of the hole transporter reaching the back collector electrode, quasi-solid-state versions of the DSC employ a thin underlayer, also made of oxides. The quasi-solid-state version can be easily adapted for portable electronics. For example it is possible to use TCO layers deposited onto flexible organic polymer substrates in order to manufacture a fully flexible lightweight version of the DSC.

1.5.3

Sequence of electron-transfer steps of a DSC

Exposure of this solar-cell assembly to visible light leads to a sequence of reactions. Figure 1.11 shows schematically these processes. We first consider the reactions that take place at the anode, where the absorption of the light by the dye S leads to formation of its electronically excited state S*: S + hν → S* (photoexcitation)

(1.15)

The molecule in the excited state can decay back to the ground state or undergo oxidative quenching, injecting electrons into the conduction band of TiO2. S* → S + hν′ (emission)

(1.16)

S* → S + e-cb (TiO2 charge injection)

(1.17)

+

The injected electrons travel through the mesoporous network of particles to reach the back-collector electrode to pass through the external circuit. The oxidized dye is reduced rapidly to the ground state by the donor (iodide) present in the electrolyte: 2S+ + 3I− → 2S + I3− (regeneration of S)

(1.18)

Photochemical and photoelectrochemical approaches

19

e– (S+/S*)

–0.9 –0.7 ΔV



e–

e– –

0.2

– 3

(I / I )

0.8

+

(S /S) dye e– electrolyte

TiO2

e– counter electrode

load e–

2.5 V, vs SCE

Fig. 1.11 Schematic drawing of a DSC showing the principles of operation.

In the absence of a redox mediator to intercept and rapidly reduce the oxidized dye (S+), recombination with the electrons of the titania layer takes place, without any measurable photocurrent: S+ + e− (TiO2) → S (recombination)

(1.19)

The electrons reaching the counter-electrode through the external circuit reduce in turn the oxidized iodide (I−) so that the entire sequence of electron transfer reactions involving the dye and the redox mediator ( I2-I−) is rendered cyclic: I3− + 2e− → 3I− (regeneration of I−)

(1.20)

If cited reactions alone take place, the overall effect of irradiation with sunlight is to drive the electrons through the external circuit, i.e., direct conversion of sunlight to electricity.

1.5.4

Key efficiency parameters of a DSC

The spectral response of the dye-sensitized solar cell depends on the absorption properties of the dye. Characterization of the cell depends on a number of experimentally accessible parameters, including the photocurrent and photopotentials measured under different conditions (open and closed circuit, under monochromatic light or sunlight illumination): Ioc, Voc, Isc and Vsc. The term incident photon-to-electricalconversion efficiency (IPCE) is a quantum-yield term for the overall charge-injection collection process measured using monochromatic light (single wavelength source).

20

Dye-Sensitized Solar Cells

The photocurrent measured under closed circuit Isc is the integrated sum of IPCE measured over the entire solar spectrum: ∞

I SC = ∫ IPCE(l) ⋅ I sun (l)d l

(1.21)

0

Thus IPCE(l) can be expressed as IPCE(l) = 1240 (Isc / lϕ )

(1.22)

where l is the wavelength, Isc the current at short circuit (mA/cm2) and ϕ is the incident radiative flux (W/m2). The overall sunlight-to-electric-power conversion efficiency of a DSC is given by the following expression: h=

P max ISC ⋅ VOC ⋅ FF = Pin Pin

(1.23)

Maximum power obtainable in a photovoltaic device is the product of two terms Imax and Vmax. The value of Imax gives the maximum photocurrent obtainable at some “maximum power point”. The Fill Factor FF is defined as the ratio of (ImaxVmax / IscVoc). The four values Isc, Voc, FF and h are the key performance parameter of the solar cell. The overall efficiency (hglobal) of the photovoltaic cell can be calculated from the integral photocurrent density (Iph), the open-circuit photovoltage (Voc), the fill factor of the cell (FF) and the intensity of the incident light (Is = 1000 W/m2) hglobal = ( Iph ⋅ Voc ⋅ FF ) / Is

(1.24)

The measured photocurrent will depend on the light intensity; the efficiency of charge injection in the excited-state quenching process; the degree of recombination of electrons with the oxidized dye (S+); and the efficiency of charge transport in the titania films to the counter-electrodes. Maximum photovoltage obtainable in such sensitized solar cells is the energy gap between the chemical potential level of the mediating redox electrolyte and the conduction band level of TiO2. Figure 1.12 shows representative photocurrent-voltage data for a dye-sensitized solar cell based on the Ru-bpy sensitizer N945 measured under AM 1.5 sunlight illumination (100 mW cm−2) (red line), along with the dependence of the power conversion efficiency for monochromatic light (blue line). The cell active area is 0.158 cm2. The inset shows a photocurrent action spectrum. For efficient DSCs with overall conversion efficiency >10 %, at 1 sun (AM 1.5) irradiation, typical photocurrent is of the order of 15-20 mA/cm2, the photovoltage in the range 650-750 mV, with a fill factor of 0.65-0.80. The overall sunlight-to-electric-power conversion efficiency of a DSC can be expressed as the product of three key terms: H = habs · hinj · hcoll

(1.25)

Where habs is the efficiency of light absorption by the dye; hinj is the efficiency of charge injection from the excited state of the dye; and hcoll is the efficiency of charge collection in the mesoporous oxide layer. An ideal photosensitizer will be the

Photochemical and photoelectrochemical approaches

21

20

15

80

IPCE, %

J, mA (cm2)

100

10

5

60 40 20

0

0 400

500 600 700 800 Wavelength, nm

–5 0

200

400

600

800

E (V) Fig. 1.12 Photocurrent-voltage curve of a solar cell based on the Ru-bpy sensitizer N945 measured under AM 1.5 sunlight illumination (100 mW cm−2) (red line). The cell active area is 0.158 cm2. Insert in blue shows the monochromatic wavelength dependence of the photocurrent.

one that absorbs all sunlight in the visible-near IR region with high absorption crosssection (coefficient). The efficiency of charge injection hinj depends on the number of low-lying electronic excited states below the conduction-band edge of the oxide semiconductor and the ability of these states to undergo electron transfer with the titania in preference to other decay channels of the excited state. For efficient electron transfer with the titania, a good electronic coupling between the electron-acceptor level of titania and the highest occupied molecular orbitals (HOMO) of the dye is required. As has been shown for many photoredox reactions of molecules and coordination compounds in solution, the energetics of the charge-injection step can be determined by considering the redox potentials of the dye excited state relative to the acceptor (conduction band) level of the oxide substrate. A moderate driving force of approximately 200 mV ensures that the excited-state electron transfer occurs rapidly and quantitatively.

1.5.5

Key components of the DSC

Dye-sensitized solar cells have many components that have to be optimized, both individually and then again as a component of a highly interactive assembly that includes substrate glass with the transparent conducting oxide TCO layer; mesoporous titania TiO2 layer; underlayer(s); dye; electrolyte solvent; redox mediator; and a counter-electrode. In the case of solar modules, different modes of interconnection of individual cells must be taken into account to optimize the active area for power generation. Here we provide a broad overview of the progress made in each of these key areas to set the stage for more extensive discussions in later Chapters.

22

Dye-Sensitized Solar Cells

Substrates for the DSC As mentioned earlier, the DSC has a sandwich structure involving two transparent conducting oxide-glass substrates. The requirements for the TCO substrate are low sheet resistance (nearly temperature independent to the high temperatures used for sintering of the TiO2 layer, 450-500 °C); and a high transparency to solar radiation in the visible-IR region. Typical sheet resistance of the TCO used is 5-15 Ω/square. The cost of TCOs rises steeply with lower sheet resistance and better light transmittance. The cost of these two substrates for the electrodes account for nearly half of the total cost of the solar cell. Both indium-doped tin oxide (In:SnO2, ITO) and fluorinedoped tin oxide (F:SnO2, FTO) have been employed. Use of glass substrates confers good protection against penetration of oxygen or water. But the heavy weight of the glass renders this form of DSC non-portable, restricting its uses to terrestrial power generation. Indium-doped tin-oxide TCO is the most common substrate used in many photonic and optoelectronic devices. This justifies its mass-production on the industrial scale. Unfortunately it has been found that the thermal stability of the ITO glass is not good, with layers peeling off the glass and/or formation of defect sites on the surface that can reduce the solar conversion efficiency of the DSC. Hence, the currently preferred TCO for DSC application is fluorine-doped tin oxide (F:SnO2). There have been several studies directed towards improving the efficiency-loss problems related to the usage of ITO for DSCs. Only two companies (Pilkington, USA and Asahi, Japan) produce F:SnO2 coated glass at the moment. The price largely depends on the total area that is produced; thus the price is highly uncertain and dependent on the commitment of the two companies to scale up their production. Today, it is asserted that the price might be approximately 10 $US/m2, which would correspond to 20 cents/Wp. Since the ohmic resistance of the TCO is too high (typically 10 Ω/square), additional current collectors like silver fingers have to be applied in modules. These have to be shielded from the electrolyte by some sealant. Such sealants reduce significantly (by 25 % or more) the effective area of the cell exposed to sunlight. In order to obtain optimal adhesive and rheological properties of a mesoporous film with the TCO substrate, suitable binders are added to the colloidal solutions of TiO2 prior to film deposition by doctor-blade techniques or screen printing. Sintering of the oxide layer at 450-500 °C gives the film two important properties: the sintering brings the individual colloidal particles to come into close contact so that the conductance and charge collection properties of the titania layer are improved; and the aerial oxidation at elevated temperature removes all the organic matter (potential trap sites) from the mesoporous film. There have been several studies on the performance of titania layers deposited using different precursors and coating procedures. Conducting oxide layers can be deposited on a wide variety of substrates, including polymer-based plastics. Advantages of such DSCs are low weight, flexibility and preparation protocols that are amenable to established industrial methods such as roll-to-roll printing. ITO-coated polyethyleneterephalate (ITO-PET) and polyethylenenaphthalate (ITO-PEN) are well known examples. The disadvantages of plastic substrates include very limited temperature tolerance (max 150-160 °C),

Photochemical and photoelectrochemical approaches

23

comparatively higher sheet resistance (60 ohm/square for ITO-PET) and permeability of the plastics to humidity (water and oxygen) over extended outdoor exposure. A third attractive substrate is thin metal foils, such as titanium or stainless steel. They have essentially same advantages as the polymer-based and plastic substrates. With metal substrates, care must be taken against corrosion of the metal by the electrolyte. Also light transmittance will be a serious issue, limiting exposure of the solar cell to sunlight from only one side. TiO2 as the photoelectrode Many wide-bandgap oxide semiconductors (TiO2, ZnO, SnO2, …) have been examined as potential electron acceptors for DSCs. TiO2 turned out to the most versatile, delivering the highest solar-conversion efficiency. TiO2 is chemically stable, nontoxic and readily available in vast quantities. It is the basic component of white paints. Particulate dispersions of TiO2, known as P25, is produced by Degussa by the ton for the paint industry. Thousands of publications have appeared on the preparation of colloidal particles of titania by the sol-gel hydrolysis route using different precursors. An important requirement for the semiconductor is high transport mobility of the charge carrier to reduce the electron-transport resistance. ZnO, SnO2, Nb2O5 and many titanates of the general formula MTiO3 have been examined as alternative oxides for the DSCs. Some of these studies are reviewed in Chapter 5 of this volume. TiO2 has many crystalline forms, with anatase, rutile and brookite being the easily accessible ones. Rutile has a slightly lower bandgap as compared to anatase and can absorb a few percent of sunlight in the near-UV region. In the standard version of DSCs, typical film thickness is 2-15 µm, and the films are deposited using nanosized particles of 10-30 nm. The highest solar-conversion efficiency is obtained in doublelayer structures, where an underlayer of thickness 2-4 µm is first deposited using larger (200-300 nm) size particles. While the beneficial effects of the underlayer has been unambiguously established for the quasi-solid-state version of DSC that uses hole transport materials (HTM), clear trends of beneficial effects have yet to be observed for the common liquid-electrolyte-based DSCs. A number of experimental approaches have been used for the deposition of the mesoporous film on the TCO substrate. The most commonly used are doctor-blading, screen printing and spray-drying. Procedure for the preparation of DSCs For the key anode component, a mesoporous film of TiO2 (of several micron thickness) is deposited using colloidal particles prepared by the controlled sol-gel hydrolysis of Ti-alkoxides. This electrode acts as a sponge and can readily take up a variety of organic and inorganic dye molecules. The dye is deposited by immersion in a stock solution for less than an hour. Depending on the method of preparation and material processing, the oxide layer can be highly translucent (ideal for integration as part of the building architecture) or opaque. The mesoporous layer consists of well interconnected colloidal particles in the size range of 15-30 nm, and the layer thickness is in the range of 5 to 15 µm. Depending on the particle size, the effective surface area (for dye adsorption), porosity and pore volume (for penetration of the redox electrolytes) of the oxide layer can vary significantly. For the best photovoltaic performance, post

24

Dye-Sensitized Solar Cells

treatment of the mesoporous oxide layers with TiCl4 is also applied and enhanced haze obtained by optimization of the light scattering properties of the mesoporous layer (Chap. 8 of the present volume). Consequences of fractal aspects of mesoporous films It was mentioned earlier that, in earlier studies of dye sensitization of semiconductor single-crystal electrodes, the sunlight-conversion efficiency obtained was quite low (≤1 %) due to two important factors: the excited-state injection process is efficient only at the monolayer coverage level, and there is poor absorption efficiency (cross section) for visible light by a monolayer of the dye. An important breakthrough, leading to an enhancement in the conversion efficiency, came about when the available surface area for dye loading was increased considerably in the mesoporous oxide form of thin films. Figure 1.13 illustrates this difference by comparing the measured IPCE for sensitization of TiO2 in two distinct cases: when TiO2 is present as a single crystal and then as a mesoporous oxide layer. Two distinct improvements can be observed in this figure. The monochromatic conversion efficiency IPCE increased significantly from 10 to greater than 90 %, due to more efficient light absorption properties of the dye distributed over the larger area within the mesoporous structure. Secondly, the photocurrent generation is much more efficient in the low-energy region of the dye absorption - a feature unique to the mesoporous nature of the nanocrystalline layers. Because the overall photocurrent, Isc, is given by the area under the IPCE vs. wavelength spectrum, the results shown in the figure represent an increase by an order of magnitude. The term roughness factor or fractal dimension is used to refer to this increased performance of the mesoporous film; this value can be simply calculated as the ratio of the measured photocurrent in the two cases cited above, with the assumption that only those molecules that are in direct contact with the oxide surface are photoactive, and that the remainder merely filter the light. Apart from poor light harvesting, a compact semiconductor film would need to be n-doped to conduct electrons. In this case energy-transfer quenching of the excited sensitizer by the electrons in the semiconductor would inevitably reduce the photovoltaic conversion efficiency. Dyes for DSC photosensitizers Along with the mesoporous oxide layer, a key component of the DSC is the photosensitizer (“dye”) that absorbs the solar radiation and injects electrons into the conduction band of the oxide substrate. Thousands of small-, medium- and large-sized molecules have been examined as potential candidates during the past two decades. In addition to organic molecules, coordination complexes of transition metals with polypyridine or porphine ligands have been studied. With reference to the simplified DSC model, it is possible to list a number of desirable properties for a photosensitizer: strong light absorption in the visible and nearIR region (for efficient light harvesting); good solubility in organic solvents (for facile deposition from stock solutions in few hours or less); presence of suitable peripheral anchoring ligands such as −COOH (to promote the effective interaction of the dye with the oxide surface and thus the coupling of donor and acceptor levels); suitable disposition of the HOMO and LUMO of the dye molecule (to permit quantitative injection of

Photochemical and photoelectrochemical approaches

25

0.15

IPCE (%)

0.10

0.05

0.00 300

400

500 600 Wavelength (nm) (a)

700

800

400

500 600 Wavelength (nm) (b)

700

800

100

IPCE (%)

80

60

40

20

0 300

Fig. 1.13 A comparison of the monochromatic power conversion efficiency for the same Rudye N719 deposited (a) on the single crystal TiO2 (anatase) electrode and (b) on a mesoporous oxide based DSC.

26

Dye-Sensitized Solar Cells

COOH COOH COOH

HOOC N

COOTBA

TBA OOC

N

N

N Ru

N

N SCN

Ru

N NCS

SCN COOH

NCS NCS

749 "black" dye

N-3 "red" dye

Fig 1.14 Chemical drawings of “red” and “black” dyes used in DSCs. Table 1.15 The highest power-conversion efficiencies obtained using different polypyridine complexes of Ru in standard DSCs using mesoporous TiO2 layers and iodide-triiodide redox electrolyte [1.98]. Dye N-719 N-749 N-749 N-719 N-3 K-19 N-945 N-621 Z-907 J-13

Surface area (cm2)

h (%)

VOC (V)

> L, where Ln is the diffusion length and L the thickness of the layer, for reflecting boundary conditions at the back contact. Curve (2) is the case of a short diffusion length, Ln