Hide Notes 

Slide 1: Solar Hydrogen Generation
Slide 2: Solar Hydrogen Generation Toward a Renewable Energy Future Edited by Krishnan Rajeshwar University of Texas at Arlington, TX, USA Robert McConnell Amonix, Inc., Torrance, CA, USA Stuart Licht University of Massachusetts, Boston, USA 123
Slide 3: Editors Krishnan Rajeshwar Department of Chemistry & Biochemistry University of Texas, Arlington Arlington TX 76019-0065 USA rajeshwar@uta.edu Robert McConnell Amonix, Inc. 3425 Fujita St. Torrance, CA 90505 bob@amonix.com Stuart Licht Chemistry Division National Science Foundation 4201 Wilson Blvd. Arlington, VA 022230 USA slicht@nsf.gov Department of Chemistry University of Massachusetts, Boston 100 Morrissey Blvd. Boston MA 02135 USA stuart.licht@umb.edu ISBN: 978-0-387-72809-4 e-ISBN: 978-0-387-72810-0 Library of Congress Control Number: 2007943478 c 2008 Springer Science+Business Media, LLC All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper 987654321 springer.com
Slide 4: Dedication Krishnan Rajeshwar To the three girls in my life, Rohini, Reena, and Rebecca: I could not have done this without your love and support Robert McConnell To my wife Suzie Star whose love and support made this possible. To my children and especially my grandson Tharyn. My hope for them is to live in a cleaner world powered by renewable energy and hydrogen. Stuart Licht To my children: Reeva, Gadi, Ariel, Jacob and Dov; I hope to open a path to a sustainable energy future for them. To my wife Bregt, this is here because you are here.
Slide 5: Preface This book examines ways to generate hydrogen from sunlight and water. It largely arose out of a desire to bring all the disparate ways to accomplish this goal within the confines of a single edited volume. Thus we are aware of many books and reports discussing the pros and cons of a hydrogen economy but none, that we are aware of, that focus on the science and technology of generating hydrogen from sunlight and water. While renewable hydrogen currently remains an elusive goal, at least from a cost perspective, the scientific principles behind its generation are well understood. Thus over and above reviewing this substantial fundamental database, part of the incentive for creating this book was to hopefully inspire future generations of scientists and engineers to respond to the grand challenge of translating the impressive laboratory advances and prototype demonstrations to a practical renewable energy economy. Much of this daunting hurdle has to do with optimizing the efficiency and hence the cost-effectiveness of hydrogen producing solar energy systems. History certainly is on our side in meeting this challenge. Many early civilizations used the sun, water, and the wind to meet basic needs. Even geothermal heat was used by North American Indians some 10,000 years ago for cooking. The ancient Greeks used hydro power to grind flour and the Persians used windmills to pump water in the first millennium. The human race is very good at solving technological problems and we can certainly wean ourselves from fossil fuels if we collectively put our minds to it. But cost is certainly going to be a driver and no amount of civic sense is going to render the hydrogen economy practically realizable if a gallon of gasoline continues to be substantially cheaper than a kilogram of hydrogen. Unfortunately however we can only give short shrift to the issue of economics in this book because of the rapidly shifting landscape of assumptions that an evolving technology brings with it. Nonetheless, the concluding chapter of this book examines investments, levelized hydrogen prices, and fuel cycle greenhouse gas emissions of a centralized electrolytic hydrogen production and distribution system powered by photovoltaic electricity.
Slide 6: viii Preface Another important and related topic, not specifically addressed in this book, concerns the issue of how to store hydrogen, especially in a mobile transportation application. We felt that this topic was specialized and wide ranging enough to warrant a separate volume to be created by scientists and engineers far more qualified and knowledgeable than us. While fuel cells are briefly introduced in Chapter 1, how hydrogen is to be utilized to generate power is again left to many other excellent treatises in the literature; some of these are cited in what follows. Every effort was made to remove redundancy and add homogeneity to the material in this multi-author volume. Indeed, the more authoritative level of discussion afforded by having specialists write each chapter will have hopefully overridden any “rough edges” that remain from chapter to chapter. Undoubtedly, many flaws remain for which we as editors are wholly responsible; we would welcome feedback on these. A project of this magnitude could not have been completed without the collective contributions of many people, some of whom we wish to acknowledge at this juncture. First, Ken Howell deserves special thanks for his many useful suggestions. His patience as this book production went through countless delays is also much appreciated. Don Gwinner, Al Hicks and their production team at NREL managed to create quality illustrations from the drawings and graphs (many in primitive form) that were furnished to them. Maria Gamboa is thanked for very capably doing the pre-print lay-out of the various manuscripts. Finally we offer simple thanks to our families for their love and support and for putting up with the many weekends away spent in putting this volume together. Krishnan Rajeshwar Arlington, Texas Robert McConnell Torrance, CA Stuart Licht Washington, DC
Slide 7: Contents Preface ...................................................................................................................... vii 1. Renewable Energy and the Hydrogen Economy ............................................... 1 Krishnan Rajeshwar, Robert McConnell, Kevin Harrison, and Stuart Licht 1 Renewable Energy and the Terawatt Challenge ............................................ 1 2 Hydrogen as a Fuel of the Future .................................................................. 3 3 Solar Energy and the Hydrogen Economy .................................................. 11 4 Water Splitting and Photosynthesis ............................................................. 12 5 Completing the Loop: Fuel Cells................................................................. 14 6 Concluding Remarks ................................................................................... 16 References ................................................................................................... 16 2. The Solar Resource ............................................................................................ 19 Daryl R. Myers 1 Introduction: Basic Properties of the Sun .................................................... 19 2 The Spectral Distribution of the Sun as a Radiation Source ........................ 20 3 The Earth's Atmosphere as a Filter .............................................................. 22 4 Utilization of Solar Spectral Regions: Spectral Response of Materials ...................................................................................................... 25 5 Reference Spectral Distributions ................................................................. 32 6 Summary ..................................................................................................... 38 References ................................................................................................... 38 3. Electrolysis of Water .......................................................................................... 41 Kevin Harrison and Johanna I vy Levene 1 Introduction ................................................................................................. 41 2 Electrolysis of Water ................................................................................... 43
Slide 8: x Contents 3 4 5 6 7 2.1 Alkaline .......................................................................................... 44 2.2 Proton Exchange Membrane........................................................... 45 Fundamentals of Water Electrolysis ............................................................ 50 3.1 First Principles ................................................................................ 50 3.2 Overpotentials................................................................................. 52 Commercial Electrolyzer Technologies ...................................................... 54 Electrolysis System ..................................................................................... 55 5.1 Energy Efficiency ........................................................................... 56 5.2 Electricity Costs.............................................................................. 58 Opportunities for Renewable Energy .......................................................... 59 Conclusions ................................................................................................. 60 References ................................................................................................... 61 4. A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen ............ 65 Robert McConnell 1 Direct Conversion of Concentrated Sunlight to Electricity ......................... 65 2 The CPV Market ......................................................................................... 66 3 Higher and Higher Conversion Efficiencies ................................................ 69 4 CPV Reliability ........................................................................................... 72 5 Following in Wind Energy’s Footsteps ....................................................... 73 6 Low-Cost Hydrogen from Hybrid CPV Systems ........................................ 75 7 Describing the Hybrid CPV System ............................................................ 76 8 Discussion ................................................................................................... 81 9 Hydrogen Vision Using Hybrid Solar Concentrators .................................. 82 10 Conclusions ................................................................................................. 83 Acknowledgements ..................................................................................... 84 References ................................................................................................... 84 5. Thermochemical and Thermal/Photo Hybrid Solar Water Splitting ............ 87 Stuart Licht 1 Introduction to Solar Thermal Formation of Hydrogen............................... 87 1.1 Comparison of Solar Electrochemical, Thermal & Hybrid Water Splitting................................................................................ 87 2 Direct Solar Thermal Water Splitting to Generate Hydrogen Fuel ............. 90 2.1 Development of Direct Solar Thermal Hydrogen ........................... 90 2.2 Theory of Direct Solar Thermal Hydrogen Generation .................. 91 2.3 Direct Solar Thermal Hydrogen Processes ..................................... 92 3 Indirect (Multi-step) Solar Thermal Water Splitting to Generate Hydrogen Fuel ............................................................................................. 94 3.1 Historical Development of Multi-Step Thermal Processes for Water Electrolysis .................................................................... 94 3.2 Comparison of Multi-step Indirect Solar Thermal Hydrogen Processes ........................................................................ 96 3.3 High-Temperature, Indirect-Solar Thermal Hydrogen Processes......................................................................................... 96
Slide 9: Contents xi 4 5 Hybrid Solar Thermal/Electrochemical/Photo (STEP) Water Splitting ....................................................................................................... 99 4.1 Historical Development of Hybrid Thermal Processes .................. 99 4.2 Theory of Hybrid Solar Hydrogen Generation ............................... 99 4.3 Elevated Temperature Solar Hydrogen Processes and Components .................................................................................. 111 Future Outlook and Concluding Remarks ................................................. 116 References ................................................................................................. 116 6. Molecular Approaches to Photochemical Splitting of Water ....................... 123 Frederick M. MacDonnell 1 Scope ......................................................................................................... 123 2 Fundamental Principles ............................................................................. 124 3 Nature's Photosynthetic Machinery ........................................................... 125 4 Design of Artificial Photosystems ............................................................. 129 5 The Ideal Sensitizer: Does Rubpy Come Close? ...................................... 133 5.1 Stability ....................................................................................... 133 5.2 Photophysics and Photochemistry ............................................... 136 6 Supramolecular Assemblies: Dyads, Triads and Beyond .......................... 138 6.1 Energy Transfer Quenching: Antenna Complexes ....................... 138 6.2 Bichromophores: Increasing Excited-State Lifetimes .................. 140 6.3 Reductive and Oxidative Quenching: Dyads and Triads with Donors and Acceptors .......................................................... 142 6.4 Single versus Multi-Electron Processes ........................................ 145 7. OER and HER Co-Catalysts ...................................................................... 150 7.1 Mimicking the Oxygen Evolving Center: Water Oxidation Catalysts ....................................................................................... 150 7.2 The Hydrogen Evolving Reaction (HER): Hydrogen Evolution Catalysts ....................................................................... 153 8. Future Outlook and Concluding Remarks ...................................................... 154 Acknowledgements ................................................................................... 156 References ................................................................................................. 156 7. Hydrogen Generation from Irradiated Semiconductor-Liquid Interfaces .......................................................................................................... 167 Krishnan Rajeshwar 1 Introduction and Scope .............................................................................. 167 2 Types of Approaches ................................................................................. 170 3 More on Nomenclature and the Water Splitting Reaction Requirements ............................................................................................. 172 4 Efficiency of Photoelectrolysis .................................................................. 178 5 Theoretical Aspects ................................................................................... 180 6 Oxide Semiconductors............................................................................... 183 6.1 Titanium Dioxide: Early Work .................................................... 183
Slide 10: xii Contents 7 8 9 10 11 12 13 14 15 Studies on the Mechanistic Aspects of Processes at the TiO2-Solution Interface ................................................................ 186 6.3 Visible Light Sensitization of TiO2............................................... 186 6.4 Recent Work on TiO2 on Photosplitting of Water or on the Oxygen Evolution Reaction ......................................................... 187 6.5 Other Binary Oxides .................................................................... 190 6.6 Perovskite Titanates and Related Oxides..................................... 192 6.7 Tantalates and Niobates ................................................................ 197 6.8 Miscellaneous Multinary Oxides .................................................. 198 Nitrides, Oxynitrides and Oxysulfides ...................................................... 200 Metal Chalcogenide Semiconductors ........................................................ 202 8.1 Cadmium Sulfide .......................................................................... 202 8.2 Other Metal Chalcogenides .......................................................... 204 Group III-V Compound Semiconductors .................................................. 205 Germanium and Silicon ............................................................................. 206 Silver Halides ............................................................................................ 208 Semiconductor Alloys and Mixed Oxides ................................................. 208 12.1 Semiconductor Composites .......................................................... 208 Photochemical Diodes and Twin-Photosystem Configurations for Water Splitting .......................................................................................... 210 Other Miscellaneous Approaches and Hydrogen Generation from Media Other than Water ............................................................................ 211 Concluding Remarks ................................................................................. 213 Acknowledgments ..................................................................................... 213 References ................................................................................................. 213 6.2 8. Photobiological Methods of Renewable Hydrogen Production .................... 229 Maria L. Ghirardi, Pin Ching Maness, and Michael Seibert 1 Introduction ............................................................................................... 229 2 Green Algae............................................................................................... 230 2.1 Mechanism of Hydrogen Production .................................. ..........230 2.2 Hydrogenase-Catalyzed H2 Production........................................ 233 2.3 [FeFe]–hydrogenases. ................................................................... 234 3 Cyanobacteria ............................................................................................ 235 3.1 Mechanisms of Hydrogen Production ......................................... 235 3.2 Hydrogenase-Catalyzed H2 Production ....................................... 236 3.3 [NiFe]-Hydrogenases.................................................................... 238 3.4 Nitrogenase-Catalyzed H2 Production ......................................... 240 3.5 Nitrogenases ................................................................................. 241 4. Other Systems............................................................................................ 242 4.1 Non-Oxygenic Purple, Non-Sulfur Photosynthetic Bacteria ........ 242 4.2 Mixed Light/Dark Systems .......................................................... 243 4.3 Bio-Inspired Systems.................................................................... 244 5 Scientific and Technical Issues.................................................................. 245 5.1 General ......................................................................................... 245 5.2 Oxygen Sensitivity of [FeFe]-Hydrogenases ................................ 246
Slide 11: Contents xiii 5.3 5.4 6 Oxygen Sensitivity of [NiFe]-Hydrogenases ................................ 248 Competition between Different Pathways for Photosynthetic Reductants ........................................................... 249 5.5 Down-Regulation of Electron Transport Rates............................. 250 5.6 Low-Light Saturation Properties of Photosynthetic Organisms .................................................................................... 251 5.7 Photobioreactor and System Costs ............................................... 252 5.8 Genomics Approaches. ................................................................. 254 Future Directions ....................................................................................... 254 Acknowledgments ..................................................................................... 255 References ................................................................................................. 255 9. Centralized Production of Hydrogen using a Coupled Water Electrolyzer-Solar Photovoltaic System ......................................................... 273 James Mason and Ken Zweibel 1 Introduction ............................................................................................... 273 2 Description of a PV Electrolytic H2 Production and Distribution System ....................................................................................................... 274 3 Capital Investment and Levelized Price Estimates .................................... 281 4 Sensitivity Analysis: H2 Production and PV Electricity Prices ................. 285 5 Economic Analysis of Second Generation (Year 31–Year 60) H2 Systems...................................................................................................... 289 6 Life Cycle Energy and GHG Emissions Analyses .................................... 294 6.1 Life Cycle Analysis Methods ....................................................... 294 6.2 Life Cycle Energy and GHG Emissions Analyses Results ........... 296 7 System Energy Flow/Mass/Balance Analysis ........................................... 296 8 Conclusions: Summary of Results and Suggestions for Future Analysis. .................................................................................................... 298 Appendices ................................................................................................ 305 1 Energy Units and CO2 Equivalent Emissions Estimates............... 305 2 Levelized Price Estimates Derived by Net Present Value Cash Flow Analysis ...................................................................... 305 3 Adiabatic Compression Formula ................................................. 307 4 Deviations from DOE H2A Assumptions ................................... 308 5 Summary of Reviewer Comments with Responses ...................... 309 References ................................................................................................. 312 Index ....................................................................................................................... 315
Slide 12: Contributors Maria L. Ghiradi, National Renewable Energy Laboratory 1617 Cole Blvd., Golden, CO 80401 marie_ghiradi@nrel.gov Kevin Harrison National Renewable Energy Laboratory NREL MS3911 1617 Cole Blvd., Golden, CO 80401 Ph: 303-384-7091, F:303-384-7055, Kevin_Harrison@nrel.gov Johanna Ivy Levene National Renewable Energy Laboratory NREL MS3911 1617 Cole Blvd., Golden, CO 80401 johanna_levene@nrel.gov Stuart Licht Chemistry Division National Science Foundation 4201 Wilson Blvd., Arlington, VA 022230 Ph: 703-292-4952, slicht@nsf.gov Chemistry Department 100 Morrissey Boulevard University of Massachusetts, Boston, MA 02135-3395 Ph: 617-287-6156, stuart.licht@umb.edu
Slide 13: xvi Contributors Frederick M. MacDonnell Department of Chemistry and Biochemistry The University of Texas at Arlington, Arlington, TX 76019-0065 Ph: 817-272-2972, F:817-272-3808, macdonn@uta.edu Pin Ching Maness National Renewable Energy Laboratory 1617 Cole Blvd., Golden, CO 80401 pinching_maness@nrel.gov James Mason Hydrogen Research Institute 52 Columbia St., Farmingdale, NY 11735 Ph: 516-694-0759, E: hydrogenresearch@verizon.net Robert McConnell Amonix, Inc. 3425 Fujita St., Torrance, CA 90505 Ph: 310-325-8091, F: 310-325-0771, E: bob@amonix.com Daryl Myers Electric System Center NREL MS3411 1617 Cole Blvd., Golden, CO 80401 Ph: 303-384-6768, F:303-384-6391, E: daryl_myers@nrel.gov, W:http://www.nrel.gov/srrl Krishnan Rajeshwar College of Science, Box 19065 The University of Texas at Arlington, Arlington, TX 76019 Ph: 817-272-3492, F:817-272-3511, E: rajeshwar@uta.edu, Michael Seibert National Renewable Energy Laboratory 1617 Cole Blvd., Golden, CO 80401 Ph: 303-384-6279, F: 303-384-6150, mike_seibert@nrel.gov Ken Zweibel Primestar Solar Co., Longmont, CO ken.zweibel@primestarsolar.com
Slide 14: Biographical Sketches of Authors Maria L. Ghirardi is a Senior Scientist at NREL and a Research Associate Professor at the Colorado School of Mines. She has a B.S., an M.S. and a Ph.D degree in Comparative Biochemistry from the University of California at Berkeley and has extensive experience working with photosynthetic organisms. Her research at NREL involves photobiological H2 production and covers metabolic, biochemical and genetic aspects of algal metabolism, generating over 60 articles and several patents. Kevin W. Harrison is a Senior Engineer in the Electrical Systems Center at NREL. He received his Ph.D. at the University of North Dakota and leveraging management, automated equipment design and quality control experience, gained while working for Xerox Corporation, he joined NREL in 2006. At NREL he leads all aspects of the renewable hydrogen production task whose objective is to improve the efficiency and reduce the capital costs of a closely coupled wind to hydrogen demonstration project. Generally speaking his research interests are in reducing the environmental impact of the world’s energy use by integrating and utilizing renewable energy for electricity and transportation fuels. Johanna Ivy Levene is a Senior Chemical Applications Engineer at NREL. She specializes in the technical and economic analysis of electrolysis systems, and her current focus is the production of fuels from renewable resources. Prior to her work at NREL, Johanna has worked as a process control engineer, a database administrator, a systems administrator and a programmer. Results from her work have been published in Solar Today and Science. Stuart Licht is a Program Director in the Chemistry Division of the National Science Foundation (NSF) and Professor of Chemistry at the University of Massachusetts, Boston. His research interests include solar and hydrogen energy, energy storage, unusual analytical methodologies, and fundamental physical chemistry. Prof. Licht received his doctorate in 1986 from the Weizmann Institute of Science, followed by a Postdoctoral Fellowship at MIT. In 1988 he became the first Carlson Professor of Chemistry at Clark University, and in 1995 a Gustella
Slide 15: xviii Biographical Sketches of Authors Professor at the Technion Israel Institute of Science, in 2003 became Chair of the Department of Chemistry at the University of Massachusetts Boston, and in 2007 a Program Director at the NSF. He has contributed 270 peer reviewed papers and patents ranging from novel efficient solar semiconductor/electrochemical processes, to unusual batteries, to elucidation of complex equilibria and quantum electron correlation theory. F. M. MacDonnell is Professor of Inorganic Chemistry at the University of Texas at Arlington (UTA). He received his PhD at Northwestern University in 1993. After a postdoctoral stint at the Chemistry Department of Harvard University, he joined the Chemistry and Biochemistry Department at UTA in 1995. His research interests are in the design of photocatalysts for light harvesting and energy conversion and has published over 100 articles in these areas. Pin Ching Maness is a Senior Scientist at NREL. She received her Masters Degree in 1976 at Indiana State University, Terre Haute, IN. She worked as a Research Specialist at the University of California, Berkeley, CA from 1976 to 1980, before joining NREL in 1981. Her research interests are in studies of the physiology, biochemistry, and molecular biology of various biological H2-production reactions in cyanobacteria, photosynthetic bacteria, and cellulolytic fermentative bacteria. James M. Mason is Director of the Hydrogen Research Institute in Farmingdale, New York. He received his PhD at Cornell University in 1996. His research interests are the economic modelling of centralized hydrogen production and distribution systems using renewable energy sources. Robert D. McConnell recently joined Amonix, Inc., a concentrator photovolatics (PV) company located in Torrance, CA as Director of Government Affairs and Contracts. He earned his PhD at Rutgers University in Solid State Physics following a Bachelor’s degree in Physics at Reed College in Portland, Oregon. After a postdoctoral stint at the University of Montreal and employment at the research institute of the electric utility, Hydro Quebec, he joined NREL in 1978. He has authored numerous papers and edited or co-edited five books and chaired four international conferences on centrator PV. His technology interests include concentrator PV, future generation PV concepts, hydrogen, superconductivity, and wind energy. He has served as Chairman of the Energy Technology Division of the Electrochemical Society and is presently Convener of the international working group developing concentrator PV standards under the aegis of the International Electrotechnical Commission located in Geneva, Switzerland. Daryl R. Myers is a Senior Scientist at NREl. In 1970 He received a Bachelor of Science in Applied Mathematics from the University of Colorado, Boulder, School of Engineering. Prior to joining NREL in 1978, he worked for four years at the Smithsonian Institution Radiation Biology Laboratory in Rockville Maryland, and is a Cold War veteran, serving as a Russian linguist in the United States Army from 1970 to 1974. He has over 32 years of experience in terrestrial broadband and spectral solar radiation physics, measurement instrumentation, metrology
Slide 16: Biographical Sketches of Authors xix (calibration), and modelling radiative transfer through the atmosphere. Daryl is active in International Lighting Commission (CIE) Division 2 on Physical Measurement of Light and Radiation, the American Society for Testing and Materials (ASTM) committees E44 on Solar, Geothermal, and Other Alternative Energy Sources and G03 on Weathering and Durability, and the Council for Optical Radiation Measurements (CORM). Krishnan Rajeshwar is a Distinguished Professor in the Department of Chemistry and Biochemistry and Associate Dean in the College of Science at the University of Texas at Arlington. He is the author of over 450 refereed publications, several invited reviews, book chapters, a monograph, and has edited books, special issues of journals, and conference proceedings in the areas of materials chemistry, solar energy conversion, and environmental electrochemistry. Dr. Rajeshwar is the Editor of the Electrochemical Society Interface magazine and is on the Editorial Advisory Board of the Journal of Applied Electrochemistry. Dr. Rajeshwar has won many Society and University awards and is a Fellow of the Electrochemical Society. Michael Seibert is a Fellow at the National Renewable Energy Laboratory in Golden, CO, USA. He received his Ph.D. in the Johnson Research Foundation at the University of Pennsylvania and then worked at GTE Laboratories before joining NREL (formerly the Solar Energy Research Institute) in 1977. His research has resulted in over 180 publications and several patents in the areas of materials development for electronic microcircuits, primary processes of bacterial and plant photosynthesis, cryopreservation and photomorphogenesis of plant tissue culture, water oxidation by photosystem II in plants and algae, microbial H2 production, hydrogenase structure and function, genomics of Chlamydomonas, and computational approaches for improving H2 metabolism in algae. He also holds a concurrent position as Research Professor at the Colorado School of Mines and is a Fellow of the AAAS. Ken Zweibel is President of PrimeStar Solar, a CdTe PV company located in Colorado, USA. He graduated in Physics from the University of Chicago in 1970. He was employed for 27 years at SERI and then NREL in Golden, CO, where he worked on the development of CdTe, copper indium diselenide, and amorphous and thin film silicon. When he left in December 2006, he was manager of the Thin Film PV Partnership Program. He has published numerous papers and articles, and two books on PV, the most recent being, “Harnessing Solar Power: The PV Challenge.” Besides the success of PrimeStar Solar, he is interested in solar policy and solutions to climate change and rising energy prices.
Slide 17: 1 Renewable Energy and the Hydrogen Economy Krishnan Rajeshwar,1 Robert McConnell,2 Kevin Harrison,2 and Stuart Licht3 1 2 University of Texas at Arlington, Arlington, TX NREL, Golden, CO 3 University of Massachussetts, Boston, MA 1 Renewable Energy and the Terawatt Challenge Technological advancement and a growing world economy during the past few decades have led to major improvements in the living conditions of people in the developed world. However, these improvements have come at a steep environmental price. Air quality concerns and global climate impact constitute two major problems with our reliance on fossil energy sources. Global warming as a result of the accumulation of greenhouse gases such as CO2 is not a new concept. More than a century ago, Arrhenius put forth the idea that CO2 from fossil fuel combustion could cause the earth to warm as the infrared opacity of its atmosphere continued to rise.1 The links between fossil fuel burning, climate change, and environmental impacts are becoming better understood.2 Atmospheric CO2 has increased from ~275 ppm to ~370 ppm (Figure 1); unchecked, it will pass 550 ppm this century. Climate models indicate that 550 ppm CO2 accumulation, if sustained, could eventually produce global warming comparable in magnitude but opposite in sign to the global cooling of the last Ice Age.3 The consequences of this lurking time bomb could be unpredictably catastrophic and disastrous as recent hurricanes and tsunamis indicate. Every year, a larger percentage of the 6.5 billion global population seeks to improve their standard of living by burning ever-increasing quantities of carbon-rich fossil fuels. Based on United Nations forecasts, another 2.5 billion people are expected by 2050 with the preponderance of them residing in poor countries.4 Coupled with this growing population’s desire to improve their quality of life are the developed countries already high and rising per capita energy use which promises to add to the environmental pressure. Oil, coal, and natural gas have powered cars, trucks, power plants, and factories, causing a relatively recent and dramatic buildup of greenhouse gases in the atmos-
Slide 18: 2 Krishnan Rajeshwar et al. Fig. 1. Atmospheric carbon dioxide record from Mauna Loa. Data courtesy of C. D. Keeling and T. P. Whorf. phere, most notably CO2. The anthropogenic buildup of heat-trapping gases is intensifying the earth’s natural greenhouse effect, causing average global temperatures to rise at an increasing rate. We appear to be entering into a period of abrupt swings in climate partially due to buildup of human-released CO2 in the atmosphere. Most alarming is not the fact that the climate is changing but rather the rate at which the buildup of CO2 is occurring. Ice core samples from Vostok, Antarctic, look back over 400,000 years before present at atmospheric CO2 levels by examining the composition of air bubbles trapped in the polar ice buried over 3623 m (11,886 ft) deep.5 These data show that the range of CO2 concentrations over this time period have been relatively stable, cycling between about 180 and 300 parts per million by volume (ppmv). According to the World Meteorological Organization the CO2 concentration in 2005 reached an unprecedented 379.1 ppmv.6 This environmental imperative requires us to quickly come to terms with the actual costs, including environmental externalities, of all of our energy use. Only then will the economic reality of energy consumption be realized and renewable sources expand through true market forces. That is not to say that fossil fuels like oil, natural gas, and coal do not have a future in helping to meet this growing demand. However, it should go without saying that all new sources of CO2 should be captured and stored (i.e., sequestered). Although integrating the systems required to safely and economically storing CO2 deep underground have not been realized. More than ever, CO2 released into the atmosphere by coal-fired power plants must be addressed
Slide 19: Renewable Energy and the Hydrogen Economy 3 to effectively deal with global climate change. In addition to greenhouse gas emissions, destructive extraction and processing of the fuel, fine particulates of 2.5 micrometers (μm) released from coal-fired power plants are responsible for the deaths of roughly 30,000 Americans every year.7 Even notwithstanding this climate change and global warming concern are issues with the supply side of a fossil-derived energy economy. Gasoline and natural gas supplies will be under increasing stress as the economies of heavily-populated developing countries (such as India and China) heat up and become more energy intensive. It is pertinent to note that this supply problem is exacerbated because the United States alone consumes a disproportionately higher fraction (more than the next five highest energy-consuming nations, Ref. 8) of the available fossil fuel supply. There are no signs that the insatiable energy appetite of the U. S. and other advanced parts of the world are beginning to wane. While there is considerable debate about when global oil and natural gas production is likely to peak,9 there is no debate that fossil fuels constitute a non-renewable, finite resource. We are already seeing a trend in some parts of the world (e.g., Alberta, Canada) of a switch to “dirtier” fossil fuels, namely, coal, heavy oil or tar sand as petroleum substitutes. This switch would mean an increase in CO2 emissions (note that the carbon content of these sources is higher than gasoline or natural gas), a greater temperature rise than is now being forecast, and even more devastating effects on the earth’s biosphere than have already been envisioned.10 Currently, renewable energy only constitutes a very small fraction of the total energy mix in the U. S. and in other parts of the world (Figure 2). For example, in 2000, only about 6.6 quads (one quad is about 1018 J) of the primary energy in the U. S. came from renewables out of a total of 98.5 quads.11 Of this small fraction supplied by renewable energy, about 3.3 quads were from biomass, 2.8 from hydroelectric generation, 0.32 from geothermal sources, 0.07 from solar thermal energy and 0.05 quads from wind turbines.8 This profile would have to switch to an energy mix that resembles the right-side panel in Figure 2 if the CO2 emissions are to be capped at environmentally safe levels. This is what the late Professor Rick Smalley, winner of the Nobel Prize in Chemistry, referred to as the Terawatt Challenge. Recent analyses12 have posited that researching, developing, and commercializing carbon-free primary power to the required level of 10-30 TW (one terawatt = 1012 W) by 2050 will require efforts of the urgency and scale of the Manhattan Project and the Apollo Space Program. This book examines the salient aspects of a hydrogen economy, particularly within the context of a renewable, sustainable energy system. 2 Hydrogen as a Fuel of the Future Jules Verne appears to be one of the earliest people to recognize, or at least articulate, the idea of splitting water to produce hydrogen (H2) and oxygen (O2) in order to satisfy the energy requirements of society. As early as 1874 in The Mysterious Island, Jules Verne alluded to clean hydrogen fuels, writing:
Slide 20: 4 Krishnan Rajeshwar et al. Fig. 2. The terawatt renewable energy challenge; the energy mix has to switch from the panel on the left to the panel on the right to cap CO2 levels at safe limits. Data from the International Energy Agency. "Yes, my friends, I believe that water will someday be employed as fuel, that hydrogen and oxygen, which constitute it, used singly or together, will furnish an inexhaustible source of heat and light….I believe, then, that when the deposits of coal are exhausted, we shall heat and warm ourselves with water. Water will be the coal of the future." Remarkable words indeed from a prophetic visionary who foresaw also the technological development of spacecraft and submarines. Hydrogen gas was first isolated by Henry Cavendish in 1766 and later recognized as a constituent of water by Lavoisier in 1783.13 The production of hydrogen and oxygen by the electrolytic decomposition of water has been practiced since the year 1800, when the process was first discovered by Nicholson and Carlisle.14 Since then, the idea of society using hydrogen as a primary energy carrier has been explored and refined. In the late 1920s and the early 1930s a German inventor, Rudolf A. Erren, recognized and worked towards producing hydrogen from off-peak electricity and modifying the internal combustion engine to run on hydrogen.15 Erren’s primary objective was to eliminate pollution from the automobile and reduce oil imports. In the 1970s Derek Gregory appears to have been one of the leading advocates in creating the case for a hydrogen-based economy.13,15,16 The literature suggests that the term hydrogen economy may have been coined by H. R. Linden, one of Gregory’s colleagues at the Institute of Gas Technology, in 1971.13 Gregory points to hydrogen’s environmental benefits and recognizes that, while fossil fuels are inexpensive, requiring the atmosphere to assimilate the byproducts of their combustion is not without consequence.
Slide 21: Renewable Energy and the Hydrogen Economy 5 The water electrolyzer industry grew substantially during the 1920s and 1930s, as elaborated later in Chapter 3. This included products from companies such as Oerlikon, Norsk Hydro, and Cominco in multi-megawatt sizes.14,17,18 Most of these installations were near hydroelectric plants that supplied an inexpensive source of electricity. As more hydrogen was needed for industries, steam reforming of methane gradually took over as the hydrogen production process of choice because it was less expensive. Hydrogen is often blamed for the 1937 Hindenburg disaster. The shell of the German airship was a mixture of two major components of rocket fuel, aluminum and iron oxide, and a doping solution which was stretched to waterproof the outer hull. Researchers concluded that the coating of the Hindenburg airship was ignited by an electrical discharge and the ensuing explosion to be inconsistent with a hydrogen fire.19 It turns out that 35 of the 37 people who died in the disaster, perished from jumping or falling from the airship to the ground. Only two of the victims died of burns, and these were from the burning airship coating and on-board diesel fuel.20 Modern laboratory tests confirmed that the 1930s fabric samples to still be combustible. “Although the benefits of the hydrogen economy are still years away, our biggest challenges from a sustainability standpoint are here today,” said Mike Nicklas, Past Chair of the American Solar Energy Society, during his opening comments at the first Renewable Hydrogen Forum in Washington, D.C., in April 2003.21 Hydrogen (H) is the simplest of atoms, consisting of one proton and one electron also called a protium. As atoms, hydrogen is very reactive and prefers to join into molecular pairs (H2) and when mixed in sufficient quantities with an oxidant (i.e., air, O2, Cl, F, N2O4, etc.) becomes a combustible mixture. Like all other fuels, H2 requires proper understanding and handling to avoid unwanted flammable or explosive environments. Hydrogen is not a primary source of energy; rather it is an energy carrier much like electricity. Therefore, energy is required to extract hydrogen from substances like natural gas, water, coal, or any other hydrocarbon. At 25 °C and atmospheric pressure the density of air is 1.225 kg m-3 while hydrogen is 0.0838 kg m-3, making it 14.6 times lighter than air. This is an important safety consideration in that a hydrogen leak will dissipate quickly. Hydrogen’s positive buoyancy significantly limits the horizontal spreading of hydrogen that could lead to combustible mixtures. Hydrogen is the lightest (molecular weight 2.016) and smallest of all gases requiring special considerations for containing and sensing a leak. Figure 3 shows the two types of molecular hydrogen distinguished by the spin, ortho- and para-hydrogen. They differ in the magnetic interactions as ortho-hydrogen atoms are both spinning in the same direction and in para-hydrogen the protons are spinning anti-parallel. At 300 K, the majority (75%) is ortho-hydrogen, while at 20 K 99.8% of the hydrogen molecules are para-hydrogen. As the gas transitions from gas to liquid at 20 K heat is released and ortho-hydrogen becomes unstable.22 Hydrogen becomes a liquid below its boiling point of −253 °C (20 K) at atmospheric pres-
Slide 22: 6 Krishnan Rajeshwar et al. Fig. 3. Ortho- (left) and para-hydrogen (right). sure. Pressurization of the hydrogen to 195 pisg (13 barg) increases the boiling point to −240 °C (−400 °F), pressures above that don’t return a significant improvement.22 At ambient temperature and pressure hydrogen is colorless, odorless, tasteless and nontoxic. However, leaks of hydrogen (or any gas for that matter) can displace oxygen and act as an asphyxiant. Any atmosphere with less than 19.5% oxygen by volume in considered oxygen deficient and asphyxiation can lead to physiological hazards. The primary hazard associated with gaseous hydrogen is the unintentional mixing of the fuel with an oxidant (typically air) in the presence of an ignition source. Hydrogen fires and deflagrations have resulted when concentrations within the flammability limit were ignited by seemingly harmless ignition sources. Ignition sources include electrical, mechanical, thermal and chemical. For example; sparks from valves, electrostatic discharges, sparks from electrical equipment, mechanical impact, welding and cutting, open flame, personnel smoking, catalyst particles and lightning strikes in the proximity of hydrogen vent stacks.23 With the exception of helium, hydrogen has the lowest boiling point at atmospheric pressure of where it becomes a transparent and odorless liquid. Liquid hydrogen has a specific gravity of 0.071, which is roughly 1/14th the density of water and is neither corrosive nor reactive. The low specific gravity of liquid hydrogen further reveals hydrogen’s low volumetric energy density in that a cubic meter of water contains more hydrogen (111 kg) than a cubic meter of pure hydrogen in liquid state (71 kg). The values of the main physical properties of gaseous hydrogen are shown in Table 1. Leaking hydrogen gas and (once ignited) its flame are nearly invisible. The pale blue flame of a hydrogen fire is barely visible and is often detected by placing a standard household wicker broom in the path of the suspected hydrogen flame. The hydrogen flame temperature in air (2045 C, 3713 F) releases most of its energy in the ultraviolet (UV) region requiring UV sensors for detecting the presence of a flare or fire. The UV radiation from a flaring hydrogen fire can also cause burns akin to over-exposure to the sun’s damaging UV radiation.
Slide 23: Renewable Energy and the Hydrogen Economy Table 1. Selected properties of gaseous hydrogen at 20 °C and 1 atm. Physical Property Molecular weight Density Specific gravity Viscosity Diffusivity Thermal conductivity Expansion ratio Boiling point (1 atm) Specific heat, constant pressure Specific heat, constant volume Specific volume Diffusion coefficient in air Enthalpy Entropy 2.016 0.0838 0.0696 8.813 x 10-5 1.697 0.1825 1:848 -253 (-423) 14.29 10.16 11.93 6.10 4098 64.44 Units kg/m3 (Air = 1) g/cm sec m2/hr W/m K Liquid to gas °C (°F) J/g K J/g K m3/kg cm2/sec kJ/kg J/g K 7 The amount of thermal radiation (heat) emitted from a hydrogen flame is low and is hard to detect by feeling (low emissivity). Most commercially available combustible gas detectors can be calibrated for hydrogen detection. Typically alarms from these sensors are set by the manufacturer between 10%–50% of the lower flammability limit (LFL) of hydrogen to avoid the presence of an unwanted flammable environment. Table 2 compares the same fuels as above and reports their volumetric energy density in kg m-3. Hydrogen has the highest energy content per unit mass than any fuel making it especially valuable when traveling into space. As mentioned earlier, hydrogen suffers volumetrically when compared with traditional fuels making storing sufficient on-board terrestrial vehicles an engineering challenge. The LFL of hydrogen represents the minimum concentration required below which the mixture is too lean to support combustion.24 Hydrogen has a wide flammability range of (4%–75%) while gasoline is (1.5%–7%) when mixed with air at standard temperature (25 ºC) and pressure (1 atm). Hydrogen in oxygen has a slightly wider flammability range (4%–95%). Table 3 summarizes a selected number of important combustion properties of hydrogen. Table 2. Comparing hydrogen properties with other fuels. Based on LHV and 1 atm, 25 °C for gases. Hydrogen -3 Methane Gasoline Diesel Methanol 799 14,500 4030 5.0 Density, kg m 0.0838 0.71 702 855 Energy density, MJ m-3 10.8 32.6 31,240 36,340 Energy density, kWh m-3 3.0 9.1 8680 10,090 Energy, kWh kg-1 33.3 12.8 12.4 11.8 *Energy density = LHV ∗ density ( ), and the conversion factor is 1 kWh = 3.6 MJ.
Slide 24: 8 Krishnan Rajeshwar et al. Table 3. Selected combustion properties of hydrogen at 20 oC and 1 atm.a Combustion Property Flammability limits in air Flammability limits in oxygen Detonability limits in air Detonability limits in oxygen Minimum ignition energy in air Auto ignition temperature Quenching gap in air Diffusion coefficient in air Flame velocity Flame emissivity Flame temperature a From Ref. 19. 4 – 75 4 – 95 18 – 59 15 – 90 17 585 (1085) 0.064 0.061 2.7 – 3.5 0.1 2045 (3713) Units vol% vol% vol% vol% J °C (°F) cm cm2/sec m/s °C (°F) Each fuel is limited to a fixed amount of energy it can release when it reacts with an oxidant. Every fuel has been experimentally tested to determine the amount of energy it can release and is reported as the fuel’s higher heating value (HHV) and lower heating value (LHV). The difference between the two values is the latent heat of vaporization of water, and the LHV assumes this energy is not recovered.22 In other words, LHVs neglect the energy in the water vapor formed by the combustion of hydrogen in the fuel because it may be impractical to recover the energy released when water condenses. This heat of vaporization typically represents about 10% of the energy content. It is often confusing to know which heating value to use when dealing with similar processes such as electrolysis and fuel cells. The appropriate heating value depends on the phase of the water in the reaction products. When water is in liquid form, the HHV is used; if water vapor (or steam) is formed in the reaction, then the LHV would be appropriate. An important distinction is that water is produced in the form of vapor in a fuel cell as well as in a combustion reaction and, therefore, the LHV represents the amount of energy available to do work. Table 4 shows both the LHV and the HHV for common fuels. Obviously, the most important virtue of using hydrogen as a fuel is its pollutionfree nature. When burned in air, the main combustion product is water with O2 in a fuel cell to directly produce electricity; the only emission is water vapor. Indeed this Table 4. HHVs and LHVs at 25 °C and 1 atm of common fuels, kJ g-1 a Fuel Hydrogen Methane Gasoline Diesel Methanol a From Ref. 22. HHV 141.9 55.5 47.5 44.8 20.0 LHV 119.9 50.0 44.5 42.5 18.1
Slide 25: Renewable Energy and the Hydrogen Economy 9 Fig. 4. Decarbonization of the energy source over the centuries. fuel cell product is clean enough to furnish drinking water to the crews in spacecraft! Crucially, the use of hydrogen completes the decarbonization trend that has accompanied the evolution of energy sources for mankind over the centuries (Figure 4). The combustion of H2, unlike fossil fuels, generates no CO2. Unlike fossil fuels, however, hydrogen is not an energy source but is an energy carrier since it almost never occurs by itself in nature, at least terrestrially. (The atmospheres of other planets, e.g., Mars, are rich in hydrogen. Should space travel prove to be economical and accessible in the future, we may have a viable means to "mine" H2 as we are doing for petroleum and coal these days!) In the interim timeframe: Where is the H2 to come from? Historically, H2 has been used for energy since the 1800s. It is a major constituent (up to ~50% by volume) of syngas generated from the gasification of coal, wood, or municipal wastes. Indeed, syngas was used in urban homes in the U. S. for heating and cooking purposes from the mid-1800s until the 1940s and is still used in parts of Europe, Latin America and China where natural gas is unavailable or too expensive. Most of the H2 manufactured these days comes from the steam reforming of methane (see above). Other processes for making H2 from fossil fuel sources include the water gas shift reactions. Neither of these approaches is carbon-neutral in that significant amounts of CO2 are generated in the H2 manufacture process itself. The ultimate goal would be to produce H2 with little or no greenhouse gas emissions. One option is to combine H2 production from fossil fuels with CO2 sequestration. Carbon sequestration, however, is as yet an unproven technology. Another approach is biomass gasification––heating organic materials such as wood and crop wastes so that they release H2 and carbon monoxide. This technique is carbonneutral because any carbon emissions are offset by the CO2 absorbed by the plants during their growth. A third possibility is the electrolysis of water using power generated by renewable energy sources such as wind turbines and solar cells. This approach is discussed in Chapters 3 and 4. Although electrolysis and biomass gasification involve no major technical hurdles, they are cost-prohibitive, at least at present: $6–10 per kilogram of H2 pro-
Slide 26: 10 Krishnan Rajeshwar et al. Fig. 5. The water splitting/hydrogen fuel cycle without (left panel) or with (right panel) inclusion of solar energy input. duced.The goal is to be able to develop and scale-up technologies to afford a pump price for H2 of $2–4 per kilogram. In such a scenario, hydrogen in a fuel cell powered car would cost less per kilometer than gasoline in a conventional car today. Clearly, water would be the ideal and most sustainable source for H2 and this H2 generation concept dates back two centuries. Table 5 summarizes various schemes for generating H2 via splitting of water and Figure 5 depicts the water splitting/hydrogen fuel cycle without (left panel) or with (right panel) inclusion of solar energy input. The approaches considered in Table 5 and Figure 5 form the topics of discussion in Chapters 4 through 7 of this book. The power needed for water electrolysis could come from nuclear energy although producing H2 this way would not be significantly cheaper than using renewable power sources. Nuclear plants can generate H2 in a non-electrolytic, thermal mode because of the intense heat generated in a thermonuclear reaction. This apTable 5. The ability of nuclear and various renewable energy sources to meet the 14-20 TW demand of carbon-free power by 2050.a Source Biomass Wind on land Power available TW 7–10 2.1 Comments Entire arable land mass of the planet must be used excluding the area needed to house 9 billion people Would saturate the entire Class 3 (wind speed at 5.1 m/s at 10 m above ground) global land mass with windmills Requires the construction of 8000 new nuclear power plants Would require damming of all available rivers Nuclear Hydroelectric a 8 1.5 From Ref. 26
Slide 27: Renewable Energy and the Hydrogen Economy 11 proach, while potentially cost-effective, has not been demonstrated yet. It must be noted that any option involving nuclear power has the same hurdles that have dogged the nuclear electric power industry for decades, namely those of waste disposal problems, proliferation concerns and lack of public acceptance. (This contrasts with the success of the nuclear power industry in some countries, e.g., France.) Producing 10 TW of nuclear power would require the construction of a new 1-GWe nuclear fission plant somewhere in the world every other day for the next 50 years!25 3 Solar Energy and the Hydrogen Economy Solar energy is a virtually inexhaustible and freely available energy source. More sunlight (~1.2 × 105 TW) falls on the earth's surface in 1 h than is used by all human activities in 1 year globally. The sun is earth's natural power source, driving the circulation of global wind and ocean currents, the cycle of water evaporation and condensation that creates rivers and lakes, and the biological cycles of photosynthesis and life. It is however a dilute energy source (1 kW/m2 at noon, Chapter 2); about 600–1000 TW strikes the earth's terrestrial surfaces at practical sites suitable for solar energy harvesting.27 Covering 0.16% of the land on earth with 10% efficient solar conversion systems would provide 20 TW of power,28 nearly twice the world's consumption rate of fossil energy and an equivalent 20,000 1-GWe nuclear fission plants. Clearly, solar energy is the largest renewable carbon-free resource amongst the other renewable energy options. Consider the total amounts possible for each in the light of the 14–20 TW of carbon-free power needed by 2050. Table 5 provides a summary;26 clearly the additional energy needed per year over the 12.8 TW fossil fuel energy base is simply not attainable from biomass, wind, nuclear and hydroelectric options. The answer to this supply dilemma must lie with solar energy. Chapter 2 provides an overview of the solar energy resource with particular emphasis on the solar spectrum. Solar energy can be harnessed in many ways25 but three routes of particular relevance to the theme of this book rely on electrical, chemical, and thermal conversion. Thus the energy content of the solar radiation can be captured as excited electron-hole pairs in a semiconductor, a dye, or a chromophore, or as heat in a thermal storage medium. Excited electrons and holes can be tapped off for immediate conversion to electrical power, or transferred to biological or chemical molecules for conversion to fuel. Solar energy is "fixed" in plants via the photosynthetic growth process. These plants are then available as biomass for combustion as primary fuels or for conversion to secondary fuels such as ethanol or hydrogen. All of these possibilities are addressed in more detail in the Chapters that follow. While there is tremendous potential for solar energy to contribute substantially to the future carbon-free power needs, none of the routes listed above are currently competitive with fossil fuels from cost, reliability, and performance perspectives. Photovoltaic solar cells have been around for decades and have been widely deployed in space vehicles. Terrestrially, their utilization thus far has been limited to niche applications or remote locales where less expensive electricity is not available. Costs for turnkey installations were 6–10 times more expensive in 1999 for solar
Slide 28: 12 Krishnan Rajeshwar et al. electrical energy than for electricity derived from coal or oil. The present cost of photovoltaic (PV) modules is ~$3.50/peak watt. Considering the additional balance of system costs (land, maintenance, etc.) this translates to an energy cost of ~$0.35/kWh. The target at present is ~$0.40/peak watt corresponding to electricity at $0.02/kWh or H2 produced by PV hybrid water electrolyzers at $0.11/kWh. Major advances in electrolyzer technology could bring this hydrogen cost to $0.04/kWh,29 which is about the present cost of H2 from steam reforming of natural gas. These issues are further elaborated in Chapters 2, 3, and 9. A cost goal of $0.40/peak watt requires solar photovoltaic conversion at a total cost of $125/m2 combined with a cell energy conversion efficiency of ~50%. Such combinations of cost and efficiency require truly disruptive photovoltaic technologies. Many such approaches are being actively pursued in research laboratories around the world. A critical discussion of outstanding issues, including dispelling the seven myths of solar electricity may be found in Refs.25, 29, and 30. The economic outlook for the other two solar approaches is not much rosier, at least at present. Solar fuels in the form of biomass produce electricity and heat at costs that are within the range of fossil fuels, but their production capacity is limited. The low efficiency with which plants convert sunlight to stored energy means large land areas are required. To produce the full 13 TW of power used by the planet, nearly all the arable land on earth would need to be planted with switchgrass, the fastest growing energy crop. Artificial photosynthetic systems, however, are more promising (see next Section) and these are discussed in Chapter 6. Solar thermal systems provide the lowest-cost electricity at the present time, but require large areas in the Sun Belt in the U. S. and continuing advances in materials science/engineering. 4 Water Splitting and Photosynthesis The decomposition of liquid water to form gaseous hydrogen and oxygen: H 2O(l ) → H 2(g ) + O 2(g ) 2 1 (1) is a highly endothermic and endergonic process with ΔH° = 285.9 kJ/mol and ΔG° = 237.2 kJ/mol. This reaction may be driven either electrochemically or thermally via the use of solar energy. The standard potential ΔE° for Reaction 1 corresponding to the transfer of two electrons is given by: ΔE° = −ΔG°/2F = −1.23 V (2) In Eq. 2, F is the Faraday constant (96485 C mol−1) and the negative sign denotes the thermodynamically non-spontaneous nature of the water splitting process. The actual voltage required for electrolysis will depend on the fugacities of the gaseous products in Reaction 1 as well as on the electrode reaction kinetics (overpotentials)
Slide 29: Renewable Energy and the Hydrogen Economy 13 Fig. 6. Pourbaix diagram of water.31 along with the Ohmic resistance losses in the cell. In practice, steady-state electrolysis of water at 298 K requires the application of ~1.50 V. Figure 6 contains a Pourbaix diagram for water;31 the zones in this diagram are labeled by the formulas for the predominant species at the electrode potential and pH indicated on the axes. Thus the threshold (thermodynamic) potentials for the decomposition of water via: H 2O (l ) → 4 e − + 0 2(g ) + 4 H + (aq ) (2a) (2b) (3a) (3b) or 4 OH − (aq ) → 4 e − + O 2 (g ) + 2 H 2O (l ) and 2 H + (aq ) + 2 e − → H 2(g ) or 2 H 2O(l ) + 2 e − → H 2 (g ) + 2 OH − (aq ) clearly depend on solution pH and they vary at a Nernstian rate of −0.059 V/pH at 298 K. Optimizing the rates of the electrochemical processes (Reactions 2 and 3) constitute much of the R&D focus in electrochemical or photoelectrochemical splitting of water. Two-compartment cells are also employed to spatially separate the evolved gases with special attention being paid to the proton transport membranes (e.g., NafionR). Chapter 3 provides a summary of the progress made in water electrolyzer technologies.
Slide 30: 14 Krishnan Rajeshwar et al. Water is transparent to the wavelengths constituting the solar spectrum. Therefore, photocatalytic or photoelectrochemical splitting of water requires an agent (semiconductor, dye, or chromophore) capable of first absorbing sunlight and generating electron-hole pairs. Molecular approaches are discussed in Chapter 6 and semiconductor-based approaches are described in Chapter 7. Thermochemical splitting of water involves heating water to a high temperature and separating the hydrogen from the equilibrium mixture. Unfortunately the decomposition of water does not proceed until temperatures around 2500 K are reached. This and other thermal routes are discussed in Chapter 5. Solar thermal processes are handicapped by the Carnot efficiency limits. On the other hand, solar photonic processes are limited by fundamental considerations associated with bandgap excitation; these have been reviewed in Refs.32 and 33. The water splitting reaction, Eq. 1, have been stated here as the Holy Grail21 of hydrogen generation using solar energy. However other chemical reactions have been investigated and include, for example:34,35 2 HBr ⎯⎯→ H 2 + Br2 2 H 2O + 2 Cl − ⎯⎯→ 2 OH + Cl 2 + H 2 hυ − hυ (4) (5) However, these alternative schemes are fraught with problems associated with the generation and handling of toxic or hazardous by-products such as Br2 and Cl2. Turning to photobiological schemes for producing H2 (Chapter 8), a complex reaction scheme uses solar energy to convert H2O into O2 and reducing equivalents which appear as NADPH. In photosystem 1, the reducing equivalents in NADPH are used to reduce CO2 to carbohydrates: 6 H 2O + 6 CO 2 + 48 hυ → C8H12O6 + 6 O 2 (6) or in bacteria, used directly as a reductive energy source.36,37 In artificial photosynthesis, the goal is to harness solar energy to drive high-energy, small-molecule reactions such as water splitting (Reaction 1) or CO2 reduction, Reaction 7:38 2 H 2O + 2 CO 2 + 4 hυ → 2 HCOOH + O 2 (7) Photobiological processes for H2 production are considered in Chapter 8. 5 Completing the Loop: Fuel Cells The high-energy chemicals such as H2 that form in the reactions considered in the preceding reaction, can be recombined in fuel cells to extract the stored chemical energy as electricity. A fuel cell is an electrochemical device that converts the chemical energy in a fuel (such as hydrogen) and an oxidant (oxygen, pure or in air) directly to electricity, water, and heat. Fuel cells are classified according to the electrolyte that they use (Table 6). For automobile applications, the polymer-electrolytemembrane (PEM) type of fuel cell is the leading candidate for developing zeroemission vehicles. Other types of fuel cells (e.g., solid oxide fuel cells or SOFCs)
Slide 31: Renewable Energy and the Hydrogen Economy Table 6. Types of fuel cells.a 15 Operating temperature, °C Polymer-electrolyte membrane (PEM) Sulfuric acid impregnated in membrane 60–80 Alkaline KOH 70–120 Phosphoric acid Phosphoric acid 160–200 Molten carbonate Lithium/potassium carbonate 650 Solid oxide Yttria-stabilized zirconia 1,000 a Adapted from Ref. 39. Type Electrolyte have been considered for stationary power needs. Figure 7 contains the schematic diagram of a PEM fuel cell.39 The major virtue of a fuel cell, other than its clean emissions, is its high electrical conversion efficiency. This is not Carnot-limited (unlike in heat engines) and for an ideal hydrogen-oxygen fuel cells, can approach an impressive 83%.40 In practical devices, up to 60% of the energy content in H2 can be converted to electricity, the remainder being dissipated as heat. For comparison, practical internal combustion engines using H2 fuel achieve efficiencies of only 45%.40 The principle of fuel cells has been known since 1838 thanks to William Grove. However, widespread deployment did not begin till the 1960s and 70s when fuel cells were used in space and for military (e.g., submarine) applications. Nowadays, fuel cells are being considered for low-polluting co-generation of heat and power in buildings and for transportation applications. As with the technologies considered earlier, the main deterrent is cost. Today's fuel cell demonstration cars and buses are custom-made prototypes that cost about $1 million apiece.41 Economies of scale in mass manufacture would bring this cost to a more reasonable $6,000-10,000 range. This translates to about $125 per kilowatt of engine power, which is about four times as high as the $30 per kilowatt cost of a comparable gasoline-powered internal combustion engine.41 A major cost component in the PEM fuel cell is the noble metal (usually Pt) electrocatalyst. Efforts are underway in many laboratories to find less expensive substitutes (see for example, Refs. 42–44). Other technical hurdles must be overcome to make fuel cells more appealing to automakers and consumers. Durability is a key issue and performance degradation is usually traceable to the proton exchange membrane component of the device. Depending on the application, 5,000–40,000 h of fuel cell lifetime is needed. Chemical attack of the membrane and electrocatalyst deactivation (due to gradual poisoning by impurities such as CO in the feed gases) are critical roadblocks that must be overcome. High temperature membranes, that can operate at temperatures above 100 °C, are desirable to promote heat rejection, speed up electrode reaction rates, and to improve tolerance to impurities. This is an active area of materials research. Unfortunately, space constraints preclude a detailed description of fuel cell technologies and the underlying issues. Instead, the reader is referred to excellent reviews and books that exist on this topic.45-47
Slide 32: 16 Krishnan Rajeshwar et al. Fig. 7. Schematic diagram of a PEM fuel cell. Reproduced from Ref. 39. Copyright (2004), by permission of The Electrochemical Society. 6 Concluding Remarks If renewable energy economy based on hydrogen were to become a reality, a nexus of three technologies, namely solar energy (thermal and photovoltaic), hydrogen production, and fuel cells will have to occur. However, many grand challenges remain in overcoming the technical and cost hurdles associated with each of these technologies. Many of these have been outlined above but are also elaborated in the Chapters that follow. Nonetheless, it is interesting to note, at this juncture, that all three technologies are poised at a very interesting stage of development in their translation from the R&D laboratory to the commercial world. How soon will they reach the marketplace will depend on many factors, some more tangible than others. References 1. S. Arrhenius, Phil. Mag. 41 237 (1896). 2. M. I. Hoffert et al., Science 298 981 (2002). 3. M. I. Hoffert and C. Covey, Nature 360 573 (1992).
Slide 33: Renewable Energy and the Hydrogen Economy 17 4. R. D. McConnell, J. A. Turner, J. B. Lasich, and D. Holland, Concentrated solar energy for the electrolytic production of hydrogen., in International Solar Concentrator Conference for the Generation of Electricity or Hydrogen, Alice Springs, Australia, 2004, p. 24. 5. J. R. Petit, et al., Vostok Ice Core Data for 420,000 Years, NOAA/NGDC Paleoclimatology Program, IGBP PAGES/World Data Center for Paleoclimatology Data Contribution Series #2001-076, Boulder, CO, 2001. 6. WMO Greenhouse Gas Bulletin 2005: Atmospheric Carbon Dioxide Levels Highest On Record, World Meteorological Organization, Retrieved on November 5, 2006, from http://www.wmo.int/web/arep/gaw/ghg/PR_762_E.doc, 2006. 7. C. G. Schneider, Death Disease and Dirty Power: Mortality and Health Damage Due to Air Pollution from Power Plants, Clean Air Task Force, 8. Sources: Energy Information Administration, U. S. Department of Energy, Rep. DOE/EIA-0384 (2000), Wash. D. C., August, 2001. International Energy Agency, http://omrpublic.iea.org. 9. J. Rifkin, The Hydrogen Economy, Jeremy P. Tarcher/Putnam, New York, 2002. 10. D. L. Albritton et al., Summary for Policy Makers: Climate Changes 2001: A Report of Working Group I of the Intergovernmental Panel on Climate Change, IPCC, 2001. www.earth.usgcrp.gov/ipcc/wg1spm.pdf. 11. S. F. Baldwin, Physics Today, April 2002, p. 62. 12. M. I. Hoffert et al., Nature 395, 881 (1998). 13. D. Gregory, A Brief History of the Hydrogen Energy Movement, Symposium Papers: Hydrogen for Energy Distribution, Institute of Gas Technology, Chicago, 1978. 14. D. H. Smith, Industrial water electrolysis, in Industrial Electrochemical Processes, edited by A. T. Kuhn, Elsevier Publishing Company, 1971, pp. 127–157. 15. D. Gregory, The hydrogen economy, in Scientific American. 228 (1) 13 (1973). 16. D. Gregory and J. B. Pangborn, Hydrogen energy, in Hydrogen for Energy Distribution, Institute of Gas Technology, 1978, pp. 279–310. 17. K. E. Cox, J. K.D. Wiliamson, Hydrogen: Its technology and implications, Production Technology, CRC Press, 1 (1977). 18. A. Konopka, D. Gregory, Hydrogen production by electrolysis: Present and future, in 10th Intersociety Energy Conversion Engineering Conference, IEEE Cat. No. 75CHO 983-7 TAB, 1975. 19. Safe Use of Hydrogen and Hydrogen Systems, NASA Training Center, 2006. 20. J. C. Bokow, Fabric, Not Filling, to Blame Hydrogen Exonerated in Hindenburg Disaster, National Hydrogen Association, http://www.hydrogenus.com/advocate/ad22zepp.htm,, 1997. 21. R. W. Larson, The right future? ASES and the renewables community examine renewable hydrogen's potential benefits — and weigh growing concerns, in Solar Today, 2004. 22. Hydrogen Properties, College of the Desert, http://www.eere.energy.gov/hydrogenand fuelcells/techvalidation/pdfs/fcm01r0.pdf, December 2001. 23. P. M. Ordin, Safety Standard for Hydrogen and Hydrogen Systems, l. NSS 1740.16: NASA, Office of Safety and Mission Assurance, 1997. 24. OSH Answers: Compressed Gases-Hazards, Canadian Centre for Occupational Health & Safety (CCOHS), Retrieved on March 25, 2006, from http://www.ccohs.ca/oshanswers/ chemicals/compressed/compress.html, 2005. 25. Basic Research Needs for Solar Energy Utilization, Report of the Basic Energy Sciences Workshop on Solar Energy Utilization, April 18-21, 2005, Office of Science, U. S. Department of Energy, Wash. D. C. See also http://www.sc.doc.gov/bes/reports/files/ SEU_rpt.pdf. 26. Prof. Nate Lewis’ website: http://nsl.caltech.edu/energy.html. 27. A. J. Nozik, Inorg. Chem. 44, 6893 (2005).
Slide 34: 18 Krishnan Rajeshwar et al. 28. J. A. Turner, M. C. Williams, K. Rajeshwar, The Electrochemical Society Interface, Fall 2004, p. 24. 29. The Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs, The National Academies Press, Washington, D. C., 2004. 30. Solar Electricity: The Power of Choice, Second Quarter, 2001; www.nrel.gov/ncpv. 31. K. B. Oldham, J. C. Myland, Fundamentals of Electrochemical Science, Academic Press, San Diego, 1994, p. 129. 32. M. D. Archer, J. R. Bolton, J. Phys. Chem. 94 8028 (1990). 33. J. R. Bolton, Solar Energy 57 37 (1996). 34. A. J. Bard, M. A. Fox, Acc. Chem. Res. 28 141 (1995). 35. J. S. Kilby, J. W. Lathrop, W. A. Porter, U. S. Patents 4 021 323 (1977); 4 100 051 (1978); 4 136 436 (1979). 36. R. E. Blankenship, Molecular Mechanisms of Photosynthesis, Blackwell Science, Oxford, U. K., 2002. 37. R. D. Britt, ed., Oxygenic Photosynthesis: The Light Reactions, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1996. 38. J. H. Alstrum-Acevedo, M. K. Brennaman and T. J. Meyer, Inorg. Chem. 44, 6802 (2005). 39. V. Ramani, H. R. Kunz and J. M. Fenton, The Electrochemical Society Interface, Fall 2004, p. 17. 40. J. M. Ogden, Physics Today, April 69 (2002). 41. J. M. Ogden, Sci. Amer. 295 94 (2006). 42. D. Berger, Science 286 49 (1999). 43. R. Bashyam and P. Zelenay, Nature 443 63 (2006). 44. C. He, S. Desai, G. Brown, and S. Bollepalli, The Electrochemical Society Interface Fall 41 (2005). 45. M. F. Mathias, R. Makharia, H. A. Gasteiger, J. J. Conley, T. J. Fuller, C. J. Gittleman, S. S. Kocha, D. P. Miller, C. K. Mittelsteadt, T. Xie, S. G. Yan, and P. T. Yu, The Electrochemical Society Interface Fall 24 (2005). 46. W. Vielstich, A. Lamm, and H. A. Gasteiger, Ed., Handbook of Fuel Cells – Fundamentals, Technology, and Applications, John Wiley & Sons, Chicester, U. K., 2003. 47. H. A. Liebhafsky and E. J. Cairns, Fuel Cells and Fuel Batteries: A Guide to Their Research and Development, John Wiley & Sons, New York, 1969.
Slide 35: 2 The Solar Resource Daryl R. Myers NREL, Golden, CO 1 Introduction: Basic Properties of the Sun The sun is class G2-V yellow dwarf star of radius 6.95508 x 107 km and surface area of 6.087 x 1022 cm2. It emits radiation produced by the internal conversion of matter into radiation into the entire 4-pi steradian solid angle (sphere), with the sun at the center. The mean radiation intensity, or radiance of the solar surface is 2.009x107 watts per square meter per steradian (Wm-2 sr-1), or a total of 2.845x1026 watts. The Earth's orbit is elliptical with an eccentricity of 0.0167 (1.4710x109 km at perihelion, 1.5210x109 km at aphelion). At the mean Earth-Sun distance, the sun subtends a solid angle of 9.24 milliradians or 0.529°.1 Thus the sun is not truly a point source, and the rays from the sun are not truly parallel, but diverge into a cone with half angle of about 0.529°. At the mean distance of the Earth from the sun of 1.495979 x 109 km, the solar radiation reaching the top of the Earth's atmosphere is 1366.1 Wm-2 ± 7.0 Wm-2 or 1.959 calories cm-2 minute-1.2 The Earth's elliptical orbit causes the distance between the Earth and the Sun (the Earth's radius vector) to vary by 3.39% from perihelion (closest) to aphelion (farthest). These variations in distance cause the intensity of solar radiation at the top of the atmosphere to vary as 1/R2, where R is the radius vector. Thus the solar input at the top of the atmosphere varies from 1414 Wm-2 (in December) to 1321 Wm-2 (in July). Additional variations in solar intensity, or brightness, result from the solar sunspot cycle, and even solar oscillations. These slight variations in the solar output are usually accounted for in the calculation of solar energy available at the top of the atmosphere, or the total extraterrestrial solar radiation, referred to as ETR. The ETR has only been monitored from space since the early 1970's, or almost three solar sunspot cycles. Excellent histories of ETR measurements and analysis are provided in Frohlich3 and Gueymard.4
Slide 36: 20 Daryl R. Myers 2 The Spectral Distribution of the Sun as a Radiation Source In this chapter we briefly describe the solar spectral distribution, or distribution of energy with respect to wavelength, over the region of the electromagnetic spectrum of use to renewable energy systems. The sun radiates energy at wavelengths ranging from the X-ray and gamma ray spectral region out into the very long wavelength radio spectral region. We will restrict our discussion, for the most part, to solar energy in the wavelength region between the ultraviolet (UV) of wavelength 250 nanometers (nm) and the near infrared (NIR) with wavelength of 4000 nm or 4.0 micrometers. The Planck theory of blackbody radiation provides a first approximation to the spectral distribution, or intensity as a function of wavelength, for the sun. The blackbody theory is based upon a "perfect" radiator with a uniform composition, and states that the spectral distribution of energy is a strong function of wavelength and is proportional to the temperature (in units of absolute temperature, or Kelvin), and several fundamental constants. Spectral radiant exitance (radiant flux per unit area) is defined as: M (λ ) = 2πc 2h λ5 (e hc / λkT − 1) (1) where λ is wavelength (in meters), h is Planck's constant = 6.626196x10-34 Joule seconds (J s), c is the velocity of light in vacuum = 2.9979250x108 meter per second (ms-1), k is Boltzman's constant = 1.3806x10-23 Joule per Kelvin (J K-1), and T is absolute temperature in Kelvins. The sun is not a "perfect" radiator, nor does it have uniform composition. The sun is composed of about 92% hydrogen, 7.8% helium. The remaining 0.2% of the sun is made up of about 60 other elements, mainly metals such as iron, magnesium, and chromium. Carbon, silicon, and most other elements are present as well.1 The interaction of the atoms and ions of these elements with the radiation created by the annihilation of matter deep within the sun modifies and adds structure to the solar spectral distribution of energy. Astrophysicists such as Kurucz have used quantum calculations and the relative abundance of elements in the sun to compute the theoretical spectral distribution from first principles.5 Figure 1 shows a plot of the Kurucz computed spectral distribution at very high resolution (0.005 nanometer at UV) as well as an inset showing much lower resolution (0.5 nanometer in UV to 5 nm in IR) plot. Figure 2 is a plot of the low resolution ETR spectrum compared with the Planck function for a blackbody with a temperature of 6000 Kelvin. The differences in the infrared, beyond 1000 nanometers are small. The larger differences in the shortwavelength region are due to the absorption of radiation by the constituents of the solar composition, resulting in the "lines" observed by Fraunhofer and named after him.
Slide 37: The Solar Resource 21 Fig. 1. The theoretical extraterrestrial solar spectral distribution (at the top of the Earth's atmosphere at the mean Earth-Sun distance of one astronomical unit) of Kurucz at high and low spectral resolution. Fig. 2. The low-resolution ETR spectral distribution (gray jagged curve) and the 6000-Kelvin blackbody spectral distribution.
Slide 38: 22 Daryl R. Myers 3 The Earth's Atmosphere as a Filter Above, we described the solar resource at the top of the atmosphere. The atmosphere acts as a continuously variable filter for the ETR radiation. The atmosphere has stable components of 78% nitrogen, 21% oxygen, and 1% argon and other "noble" gases. There are also variable components in the atmosphere, such as water vapor (0% to 2% of the total composition), and gases dumped into the atmosphere by manmade and natural process, such as carbon dioxide, (0.035%), methane, and nitrous oxides.6 At a concentration of only 0.3 parts per million, Ozone absorbs and attenuates the dangerous UV radiation below 280 nm. Water vapor absorbs mainly in the infrared, contributing to the heating of the atmosphere. Similarly, small concentrations of "greenhouse" gases such as carbon dioxide and methane absorb in the infrared, but with such strength that their increasing concentration may pose a threat to the stability of the Earth's climate. Suspended particulates such as aerosols, dust and smoke, as well as condensed water vapor (clouds) also strongly modify the solar resource as the atmosphere is traversed by photons from the top of the atmosphere to the surface. As the photons propagate through the atmosphere, radiation in the narrow cone of light from the solar disk (the direct beam) interacts with the atmosphere by being absorbed or scattered out of the beam. Molecules of atmospheric gases, aerosols, dust particles, and so on, do the absorption and scattering. Scattered radiation contributes to the bright blue of the clear sky dome, or the dull gray of overcast skies. The scattered radiation also illuminates clouds, which reflect most of the wavelengths of light in the visible region, making them appear white. Figure 3 schematically shows this process of atmospheric sorting of the radiation into three components: the direct beam, the scattered radiation (called diffuse radiation) and the combination of the direct and diffuse radiation, called the total hemispherical, or global radiation from the entire sky dome. As a result of the absorption and scattering processes in the atmosphere, the ETR spectral distribution is significantly modified. Figure 4 shows the effect of the atmosphere on the ETR spectral distribution for a very specific solar geometry, referred to as Air Mass 1.5. As indicated in Fig. 3, Air Mass 1.0 occurs when the sun is directly overhead. The angle between the horizon and the observer (solar elevation) is then 90°, and the angle between the zenith (overhead point) and the observer is 0°. The relative position of the sun with respect to the horizon or zenith is specified as the elevation angle, ε, or zenith angle, z, respectively. The term Air Mass refers to the relative path length through the atmosphere, with respect the minimum path length of 1 for zenith angle 0°. Geometrically, Air Mass M is: M = 1 / cos( z ) or M = 1 / sin (ε ) (2) Figures 5 and 6 show how the direct beam and global horizontal spectral distributions are modified on a clear day as a function of increasing Air Mass. Note how, as the Air Mass increases, the direct beam spectra change more than the global sky
Slide 39: The Solar Resource 23 Fig. 3. Scattering of the direct beam photons from the sun by the atmosphere produces diffuse and global sky irradiance. Fig. 4. The attenuation and absorption of the ETR spectral distribution by the atmosphere. Top curve is the ETR spectral distribution. In decreasing order, the global, direct, and diffuse spectral distributions at the bottom of the atmosphere (at sea level) for the sun at zenith angle 48.2° (Air Mass 1.5) are shown.
Slide 40: 24 Daryl R. Myers Fig. 5. Progressive reduction in direct beam spectral distributions as air mass in increased. The plots representing uniformly increasing 10° steps in zenith angle from 0° (top curve) to 80° (bottom curve). Fig. 6. Progressive reduction in global total hemispherical spectral distributions as Air Mass is increased, as in Fig. 5.
Slide 41: The Solar Resource 25 spectra. This is because the energy scattered out of the direct beam is "transferred" into the global spectra, as increasing contributions to the diffuse sky radiation. Figure 5 in particular illustrates the shift of the spectral peak to longer (red) wavelengths associated with red skies at sunset and sunrise. Figure 7 schematically shows the relationship of absolute temperatures, and color as perceived by the human eye for various outdoor natural conditions, and artificial indoor sources. Our eyes adapted to take advantage of the energy peak in solar spectral distribution between 400 nm and 700 nm, and evolved under the spectral distribution of our sun. Therefore, the match (or mismatch) of artificial source spectral distributions with the solar spectral distribution is important for both natural and artificial lighting applications. As mentioned above, other elements besides the gases in the atmosphere interact with the ETR spectral distribution. One of the most important of these other atmospheric constituents are small particles called aerosols. Particles scatter radiation most efficiently when the wavelength of the radiation is smaller than the particle size. Many of the particles that work their way into our atmosphere (dust, decaying organic material, smoke from fires, etc.) have diameters that scatter shortwave (UV) and visible light very efficiently. This removes a great deal of energy from the direct beam, and redistributes the energy over the sky dome.7 A measure of the scattering power of aerosols is the amount of energy removed from (or the attenuation of) the beam radiation. The Beer-Bouger-Lambert law for the attenuation of a beam of intensity Io to intensity I, resulting from the amount, x, of a material is given by: I/Io = e-τ x or I = Io e-τ x where τ is the attenuation coefficient, called the aerosol optical depth , or AOD. The attenuation by aerosols is both exponential, and a strong function of wavelength, implying a large impact on the direct beam solar spectral distribution. Anders Angstrom first proposed a relation between the wavelength and τ, dependent on two parameters and a reference point at a wavelength of 1000 nm as τ= β λ−α. β is related to the size of the particles, and ranges from about 1 to 2. α is related to the scattering properties of the particles, and ranges from about 0.001 to 0.5. Most often, AOD is referred to a specific wavelength, usually 500 nm, or occasionally 550 nm. An AOD of 0.01 at 500 nm represents very clean, pristine, clean atmosphere. Values of AOD at 500 nm of 0.1 to 0.2 are quite typical of average conditions. Values of 0.4 or greater represent a heavy aerosol load and very hazy skies. Figure 8 shows the clear sky direct beam spectral distribution at Air Mass 1.5 for a range of AOD from 0.05 to 0.40 From the discussion above and Figs. 5 to 7, the fluctuations in the solar spectrum at the Earth's surface are dependent on many factors. These variations can be characterized with measurements and models to account for their impact on solar renewable energy technologies. 4 Utilization of Solar Spectral Regions: Spectral Response of Materials Why this extended discussion of the solar spectral distribution? The primary reason is provided by the example discussing the sensitivity of our eyes, in the previous
Slide 42: 26 Daryl R. Myers Fig. 7. Temperature and color relationships for various natural and artificial sources of optical radiation.
Slide 43: The Solar Resource 27 Fig. 8. Decreasing direct beam spectral irradiance as aerosol optical depth at 500 nm increases from 0.05 (top curve) to 0.04 (bottom curve). Fig. 9. Human eye daytime (photopic) relative spectral response (right axis) and solar spectrum (jagged curve).
Slide 44: 28 Daryl R. Myers Section. Figure 9 shows the wavelength region where the human eye responds, the spectral response of the eye, overlaying an Air Mass 1.5 global solar spectrum. The peak of this response is at 555 nm, corresponding to the color green. Many of the materials used in renewable, and specifically, solar energy system applications have a significant response to the solar spectrum over a limited spectral interval. The green color of most plant leaves is the result of absorption of blue and red light by chlorophyll, which reflects almost all of the light in the region around 550 nm, as shown in Fig. 10. Our skin contains compounds that absorb ultraviolet light with wavelengths shorter than 400 nm. These compounds react with the UV photons to produce suntans, sunburns, or even cancer. Similarly, many semiconductors, such as silicon, germanium, etc. produce a flow of electrons (the photovoltaic effect) when photons of a certain wavelength interact with the materials. Figure 11 shows Air Mass 1.5 direct normal, diffuse sky, and total global solar spectral distributions, with indications of the spectral regions where our vision, plants, and various photovoltaic materials interact with the solar spectral distribution. Fig. 10. Blue and red light absorbed by chlorophyll (types a and b shown) used in photosynthesis. Green (around 550 nm) is reflected and not used.
Slide 45: The Solar Resource 29 Fig. 11. Solar spectral distributions and the various regions of spectral sensitivity for vision, plant photosynthesis, and photovoltaic conversion technologies. Figure 12 shows that various combinations of photovoltaic materials can be constructed to respond over different spectral ranges, utilizing more or less of the solar spectral distribution.8 The different response regions shown in Fig. 12 are the result of the band gap between bound electrons in the material and the conduction band for electrons (and holes) in terms of energy (in electron volts, eV). The energy, E, of a photon is related to the wavelength as E = (h c) / λ. In semiconductors suitable for Photovoltaic applications, the band gaps are relatively close together, so photons with relatively low energy (0.5 to 1.5 eV) can stimulate electrons to enter the conduction band. We can convert the power versus wavelength plot to one of number of photons versus energy, as on the left of Fig. 13. On the right of Fig. 13 we show the relationship between the available solar energy and projected conversion efficiencies of some PV materials can approach 50%, if concentrated (focused by lens or mirrors) solar radiation is used in conjunction with future generation materials.
Slide 46: 30 Daryl R. Myers Fig. 12. Spectral response regions for various photovoltaic technologies. CIG stands for cadmium indium gallium selenide. Fig. 13. Correlation of available solar spectral photon energies with band gap of present and future generation photovoltaic materials. New represents some new combination of materials optimized or tailored for a specific band gap.
Slide 47: The Solar Resource 31 As Fig. 13 shows, future generation materials with "designer" band gaps can produce higher efficiency devices to generate more electricity with the same solar spectrum. By stacking the available materials, additional components of the solar spectrum contribute to the overall production of conduction electrons, or electric current. Optimization of the performance of solar energy systems, as well as building thermal performance (heating and cooling loads), and daylighting (window performance) all require knowledge of the terrestrial solar spectral distribution. Optical properties of materials such as transmittance, reflectance, and absorption are always dependent on the wavelength of the incident radiation. For example, Fig. 14 shows the properties of several components of a window for building applications. The top left panel shows the properties of a single pane of glass which permits thermal infrared radiation into a room; high inside reflectance keeps the thermal energy trapped, offsetting the need for more heating energy in a cold climate. In the lower panels of the figure, properties of each pane of a double pane structure are shown. Low IR transmittance (lower left) of the outer pane keeps thermal infrared solar radiation from entering the building. The broad transmittance band of the inner layer and low inside reflectance of the inner pane allows thermal infrared energy to escape. This structure reduces the cooling load in a sunny environment. Fig. 14. Examples of optical properties of materials (reflectance, transmittance) for window structures. When used in conjunction with solar spectral distributions, energy savings can be computed.
Slide 48: 32 Daryl R. Myers Fig. 15. Spectral response of thermopile pyranometer measuring total solar radiation is shown with thick black line. Spectral radiance (brightness) of the sky dome (blue line). The cut-off at 3000 nm means the radiometer will not respond to the infrared sky radiation that peaks at 7000 nm (7 micrometers). Similar principles can be used for designing selective absorbers, where the goal is absorb as much of the solar spectrum as possible and convert that absorbed energy into heat, or thermal energy. Conversely, reflective materials can be designed to select only visible (cold mirror) or infrared (hot mirrors) to isolate and direct selected portions of the solar spectrum for various applications. Knowledge of the optical properties of materials in relation to the solar spectrum is also important in measuring broadband solar radiation. For instance, a pyranometer used to monitor total solar radiation for a renewable energy system has a spectral response (due to the special glass dome protecting the detector) that does not respond to the thermal infrared radiation of the sky beyond 3000 nm, as shown in Fig. 15. However, there will be thermal infrared radiation exchanged between the radiometer and the sky dome, which will influence the measurement performance of the pyranometer.9 5 Reference Spectral Distributions From the discussion in the preceding Sections, even without addressing the influence of clouds, it is clear that the terrestrial solar spectrum is highly variable. So, how can
Slide 49: The Solar Resource 33 we relate the spectral responses and spectral optical properties of materials to each other when this variability is present? The answer is to establish a standard spectral distribution with which to compute performance. Then comparisons can be made based on standard set of conditions. Furthermore, if measurements are made under conditions, deviations from the standard conditions can be computed, documented, and in most cases corrected or scaled to the reference conditions. Several national and international consensus standards organizations, such as the American Society for Testing and Materials (ASTM) and the International Standards Organization (ISO) have adopted a reference standard extraterrestrial spectral distribution (ASTM E490-00a), and terrestrial reference spectral distributions for direct beam and total hemispherical (on a 37° tilted south facing surface) spectra at a prescribed air mass of 1.5 (ASTM G173-03).2,10,11 The extraterrestrial reference spectral distribution was assembled from a number of recent diverse space-based (satellite) measurement sources. The ETR reference spectrum is normalized to a total integrated irradiance of 1366.1 Wm-2. The terrestrial reference spectra required a more extensive set of criteria to be met. Specifically, a set of reasonable conditions that could occur rather commonly in nature should be used. The conditions for the terrestrial reference spectra were chosen to meet the following criteria:12 • • • • • • • Air mass 1.5 represents the condition where approximately 1/2 of the total available solar energy is available for air masses greater than and less than this condition, respectively. For solar thermal and photovoltaic systems using flat plate collectors, energy collection is optimized for collectors tilted south (in the northern hemisphere) at approximately the latitude of the site. The mean latitude of the contiguous 48 United States is approximately 37° North. Standard test conditions prescribed in photovoltaic standards require a total hemispherical irradiance on flat plate collectors of 1000 Wm-2 (a value that can obtained on a clear day around noon). Concentrating solar collector systems that utilize the direct beam radiation would be more likely deployed in areas with relatively low aerosol optical depth, to maximize direct beam utilization.13 The terrestrial reference should be easily reproducible, preferably by a simple (but accurate) model calculation, and the model should be easily maintained and updated, and in the public domain.14 The terrestrial reference spectrum should have uniform wavelength increments for ease of computation and comparison with measured spectral data. These criteria resulted in the choice of a relatively simple, but accurate spectral model of Gueymard (SMARTS: Simple Model for Atmospheric Transmission of Sunshine) to compute the reference standard terrestrial spectra.7 The philosophy behind the SMARTS model is to parameterize the band model transmittance functions used by a very complex (50,000 line of FORTRAN code and about 200 subroutines) MODTRAN (MODerate resolution TRANSmittance) code15 developed by the Air Force Geophysics Laboratory, for the most important at-
Slide 50: 34 Daryl R. Myers Table 1. Transmission expressions developed for SMARTS model. Absorption Mechanism Rayleigh Scattering Ozone Nitrogen Dioxide (NO2) Mixed and Trace Gases Water Vapor Aerosol Transmittance Expression Tr(λ) = exp{(P/Po)/[a0(λ/λo)4+a1(λ/λo)+a2+a3(λ/λo)-2]} To(λ) = exp [–mo uo Ao(λ) ] Tn(λ) = exp [–mn un An(λ) ] Tg(λ) = exp [–( mg ug Ag(λ))] Tw(λ) = exp[–( mw uw)Bw(λ)Bm(λ)Bp(λ)Baw(λ) Aw(λ)] Ta(λ) = exp [–ma βi (λ/λ1)-αi] mospheric constituents, at a resolution of 0.5 nm in the ultraviolet less than 400 nm, 1 nm between 400 nm and 1700 nm, and 5 nm between 1700 nm and 4000 nm. These parameterized transmittance functions were developed to account for Rayleigh scattering (Tr), ozone (To), mixed gas (Tg), nitrogen dioxide (Tn), water vapor (Tw), and aerosol (Ta) transmission of the direct beam irradiance using Eq. 3: E(λ) = Eo(λ) Tr(λ) To(λ) Tg(λ) Tn(λ) Tw(λ) Ta(λ) (3) at each wavelength (λ, in nm), where E is the terrestrial spectral irradiance, Eo is the extraterrestrial spectral irradiance, and the spectral transmittances are defined above. Table 1 summarizes the form of transmittance functions developed for the SMARTS model. Table 1 expression parameters are; P = station pressure, Po = standard pressure, ai = fitting coefficients, m = air mass correction for path length, u = absorber abundances, A = absorption coefficients, a = absorber/wavelength dependent, Bs = water vapor band, airmass, and pressure scaling factors, αi βi, = Ångstrom parameters, i = 1 for λ < 500 nm, i = 2 for λ ≥ 500 nm,* λ1: Reference wavelength (usually 1000 nm or 1 μm) Figures 16 and 17 show percent difference between SMARTS MODTRAN results and one of many comparisons of measured and SMARTS spectral data. Agreement within the uncertainty limits of spectral irradiance measurements (1% in the visible, and 3% to 5% in the ultraviolet and infrared), is achieved. Version 2.9.2 of the SMARTS spectral model used to generate the spectral reference standard, the users manual for the model, and a list of references, can be downloaded free of charge from the National Renewable Energy Laboratory Renewable Resource Data Center at the following URL: http://rredc.nrel.gov/solar/models /SMARTS/. A CD-ROM adjunct to the ASTM G-173-03 standard, with a copy of the model, manual, and reference material is available for purchase from ASTM. The SMARTS input file is a straightforward assembly of fifteen to twenty parameters arranged as in a stack of input cards. Table 2 is an annotated input file used to generate the ASTM G173 standard spectra on a 37° tilted south facing plane. Note that we do not show all possible input combinations. Figures 18 and 19 portray the ETR and terrestrial standard reference spectra. * The common assumption the α1 = α2 = α can lead to errors for the urban, maritime, and rural aerosol profiles. The Angstrom exponents are determined as a function of aerosol type and relative humidity (cf. Appendix B of Gueymard.)7
Slide 51: The Solar Resource 35 Fig. 16. Percent difference between MODTRAN and SMARTS spectral results, for ASTM reference spectra conditions. Largest differences due to SMARTS trace gases. Fig. 17. SMARTS model results (lines) and measurements (symbols) at 5 nm resolution for 3 air masses at NREL, Sep 18, 2001.
Slide 52: 36 Daryl R. Myers Table 2. SMARTS version 2.9 input file for ASTM reference spectra G173-03. Card ID 1 2 2a 3 3a 4 5 6 7 7a 8 9 9a 10 10b 10c 11 12 12a 12b 12c 13 13a 14 15 16 17 17a Value ASTM_G173_Std_Spectra 1 1013.25 0. 1 USSA 1 1 1 370 1 S&F_RURAL 0 0.084 38 1 38 37 180 280 4000 1.0 1367.0 2 280 4000 .5 2 89 1 0 2.9 0 0 0 0 2 1.5 Parameter/Description/Variable name Comment line Pressure input mode (1 = pressure and altitude): ISPR Station pressure (mb) and altitude (km): SPR, ALT Standard Atmosphere Profile Selection (1 = use default atmosphere): IATM1 Default Standard Atmosphere Profile: ATM (one of eleven choices, including user defined) Water vapor input (1 = default from Atmospheric Profile): IH2O (may be user specified) Ozone calculation (1 = default from Atmospheric Profile): IO3 (may be user specified) Pollution level mode (1 = standard conditions/no pollution): IGAS (for 10 pollutant gases) Carbon monoxide volume mixing ratio (ppm): qCO2 Extraterrestrial spectrum (1 = SMARTS/Gueymard): ISPCTR (one of seven choices) Aerosol profile to use: AEROS (one of 10 choices, including user specified) Specification for aerosol optical depth/turbidity input (0 = AOD at 500 nm): ITURB Aerosol optical depth @ 500 nm: TAU5 Far field spectral Albedo file to use (38 = Light Sandy Soil): IALBDX (on of 40 choices, including user defined) Specify tilt calculation (1 = yes): ITILT Albedo and Tilt variables—Albedo file to use for near field, Tilt, and Azimuth: IALBDG, TILT, WAZIM Wavelength range—start, stop, mean radius vector correction, integrated solar spectrum irradiance: WLMN, WLMX, SUNCOR, SOLARC Separate spectral output file print mode (2 = yes): IPRT: Spectral & broadband files Output file wavelength—Print limits, start, stop, minimum step size: WPMN, WPMX, INTVL Number of output variables to print: IOTOT (up to 32) Code relating output variables to print (8 = Hemispherical tilt, 9 = direct normal + circumsolar): OUT(8), OUT(9) [up to 32 spectral parameters available for output] Circumsolar calculation mode (1 = yes): ICIRC Receiver geometry—Slope, View, Limit half angles: SLOPE, APERT, LIMIT Smooth function mode (0 = none): ISCAN (Gaussian and triangle filter shapes can be specified) Illuminance calculation mode (0 = none): ILLUM (Luminance and efficacy may be selected) UV calculation mode (0 = none): IUV (UVA, UVB, action weighed dosages available) Solar geometry mode (2 = Air Mass): IMASS (zenith and azimuth, date/time/lat/long available) Air mass value: AMASS
Slide 53: The Solar Resource 37 Fig. 18. ASTM E490-00a extraterrestrial reference spectrum. The actual spectral data go out to 100,000 nm (100 microns). Inset shows details in the 250 nm to 2000 nm region. Fig. 19. ASTM G173-03 Terrestrial reference spectra for Air Mmass 1.5, conditions specified in Table 2.
Slide 54: 38 Daryl R. Myers 6 Summary The basic properties of the sun and the solar radiation received at the top of the Earth's atmosphere have been described in this chapter. The extraterrestrial solar spectral distribution is modified through interactions with the gases and particles in the atmosphere to produce terrestrial spectral distributions that vary with respect to both amplitude and wavelength over a very wide range. The solar resource to a specific solar renewable energy technology depends upon the spectral response of the systems and materials involved. This is true whether the technology addresses biomass (photosynthesis), daylighting, building heat loads, thermal energy conversion, or photovoltaic production of electricity. Each application utilizes one or mores specific regions of the terrestrial solar spectrum, either in isolation or in various combinations. The solar renewable energy community has developed a set of tools and standards to assist in the design and optimization of solar renewable systems, including hybrid systems that may combine solar and other renewable technologies, such as wind energy generation. The present set of extraterrestrial and terrestrial solar spectral standards have evolved over the past 30 years to keep abreast of the requirements that new, innovative renewable energy systems researchers, designers, and manufacturers require to meet their customers’ needs. References 1. 2. 3. Cox, A. N., ed. Allen's Astrophysical Quantities. 4th ed., AIP Press, Springer Verlag, New York, NY., 1999. ASTM, Standard Solar Constant and Zero Air Mass Solar Spectral Irradiance Tables, Standard E490-00a, American Society for Testing and Materials, West Conshohocken, PA, 2000. C. Frohlich and J. Lean, Total Solar Irradiance Variations: The Construction of a Composite and it's Comparison with Models, International Astronomical Union Symposium 185: New Eyes to See Inside the Sun and Stars, Dortrect, The Netherlands, Kluwer Academic, 1998. C. A. Gueymard, The sun's total and spectral irradiance for solar energy applications and solar radiation models, Solar Energy,. 76(4) 423 (2004). R. L. Kurucz, Synthetic Template Spectra. Highlights of Astronomy, L. Appenzeller, ed., Vol. 10, The Hague, Netherlands, Aug 15-17, 1994, Kluwer Acad. (1995) pp. 407–409. W. M. Farmer, The Atmospheric Filter, Vol. I., JCD Publishing, Winter Park, FL, 2001, p. 273. C. Gueymard, Parameterized transmittance model for direct beam and circumsolar spectral irradiance, Solar Energy 71(5) 325 (2001). H. Field, Solar Cell Spectral Response Measurements Related to Spectral Bandwidth and Chopped Light Waveform, 26th IEEE Photovoltaic Specialists Conference, Institute of Electrical and Electronic Engineers, Anaheim, CA, 1997. I. Reda, J. Hickey, C. Long, D. Myers, T. Stoffel, S. Wilcox, J. J. Michalsky, E. G. Dutton, and D. Nelson, Using a blackbody to calculate net-longwave responsivity of shortwave solar pyranometers to correct for their thermal offset error during outdoor calibration using the component sum method, Journal of Atmospheric and Oceanic Technology 22 1531 (2005). 4. 5. 6. 7. 8. 9.
Slide 55: The Solar Resource 39 10. ASTM, Standard Tables for Reference Solar Spectral Irradiance at Air Mass 1.5: Direct Normal and Hemispherical for a 37° Tilted Surface, Standard G177-03. 2003 American Society for Testing and Materials, West Conshohocken, PA. 11. ISO, Solar energy—Reference solar spectral irradiance at the ground at different receiving conditions, pt. 1. International Standard 9845-1, International Organization for Standardization, 1992. 12. C. Gueymard, D. Myers, and K. Emery, Proposed reference irradiance spectra for solar energy systems testing, Solar Energy,. 73(6) 443 (2002). 13. S. Kurtz, D. Myers, T. Townsend, C. Whitaker, A. Maish, R. Hulstrom, and K. Emery, Outdoor rating conditions for photovoltaic modules and systems. Solar Energy Materials Solar Cells 62 379 (2000). 14. Myers, D., K. Emery, C. Gueymard, Revising and Validating Spectral Irradiance Reference Standards for Photovoltaic Performance Evaluation. ASME Journal of Solar Energy Engineering, 2004. 126: p. 567-574. 15. G. P. Anderson, A. Berk, P. K. Acharya, M. W. Matthew, L. S. Bernstein, J. H. Chetwynd, Jr., H. Dothe, S. M. Adler-Golden, A. J. Ratkowski, G. W. Felde, J. A. Gardner, M. L. Hoke, S. C. Richtsmeier, B. Pukall, J. B. Mello, and L. S. Jeong, MODTRAN4: Radiative Transfer Modeling for Remote Sensing, in Optics in Atmospheric Propagation and Adaptive Systems III, Society of Photo-Optical Instrumentation Engineers Bellingham, WA., 1999.
Slide 56: This page intentionally blank
Slide 57: 3 Electrolysis of Water Kevin Harrison and Johanna Ivy Levene NREL, Golden, CO 1 Introduction Hydrogen energy systems, based on renewable energy (RE) sources, are being proposed as a means to increase energy independence, improve domestic economies, and reduce greenhouse gas emissions from stationary and mobile fossil-fueled sources. In 2003, the United States consumed roughly 84.3 billion m3 (7.6 billion kilograms) of hydrogen, the majority of which was produced via the widely established thermal process known as steam methane reforming (SMR).1. The electrolytic production of hydrogen, while not economically competitive today with SMR, is positioned to become the preferred method due to the inevitable price increase of natural gas and as environmental, social, and economic factors are weighed. SMR constitutes roughly 50% of the 450–500 billion m3 yr-1 (38–42 billion kg yr-1) of global production of the gas.1,2. SMR, like hydrogen production from all fossil fuels, suffers from supply issues and climate-altering carbon-based pollution. The reforming process generates CO2 as well as carbon monoxide (CO), which is poisonous to humans because the oxygen-transporting hemoglobin has 200 times the affinity to CO than O2.3 Electrolysis currently supplies roughly 4% of the world’s hydrogen. If hydrogen is to be used as a transportation fuel, the United State could conceivably replace the 140 billion gallons per year (gal yr-1) of gasoline consumed in 2004 with domestically produced hydrogen. The energy equivalent of this much gasoline is 17.3x1015 BTU, assuming approximately 5.2 million BTU bbl-1 of motor gasoline.4 The environmental gains hoped for by the transition to a hydrogen economy can only be achieved when renewable sources are ramped up to produce an increasing amount of the hydrogen gas.
Slide 58: 42 Kevin Harrison and Johanna Ivy Levene From the early 1800s to the mid 1900s town gas was comprised of roughly 50% hydrogen that brought light and heat to much of America and Europe and can still be found in some parts of Europe, China and Asia. Due to hydrogen’s thermal conductivity and low density the gas is being used to cool many large thermal electrical power generators. Hydrogen is used in a wide variety of applications:5 • • • • • Chemicals Ammonia and fertilizer manufacture Synthesis of methanol Sorbitol production General pharmaceuticals and vitamins Electronics Polysilicon production Epitaxial deposition Fiber optics Metals Annealing/heat treating Powder metallurgy Fuels Petroleum refinement Liquid rocket fuel Some use in fuel cells Food and float glass Fats/fatty acids Blanketing Renewable sources of electricity and off-peak hydroelectric can be used to produce a sustainable supply of hydrogen for transportation, peak-shaving applications and in some special cases to smooth the variability in the renewable source. Powering millions of hydrogen internal combustion engines and/or fuel cell vehicles with hydrogen generated with traditional fossil fuel sources (without carbon dioxide (CO2) capture and storage or geological sequestration) is merely transferring the pollution from the tailpipe to the stack pipe. In the case of SMR, liquid natural gas imports would increase to replace today’s 12.9 million bbl day-1 of oil imports here in the U.S.4.4 As developing countries fall in love with motorized transportation, much like the developed countries already have, transportation’s contribution to greenhouse gas emissions will grow from the 25% it holds today.6 Still today, the electrolytic production of hydrogen using renewable sources is the only way to produce large quantities of hydrogen without emitting the traditional byproducts associated with fossil-fuels. The electrolysis of water is an electrochemical reaction requiring no moving parts and a direct electric current, making it one of the simplest ways to produce hydrogen. The electrochemical decomposition of water into its two constituent parts has been shown to be reliable, clean and with the removal of water vapor from the product capable of producing ultra-pure hydrogen (> 99.999%).
Slide 59: Electrolysis of Water 43 The primary disadvantage of electrolysis is the requirement of high-quality of electrical energy needed to disassociate the gas. Electricity is a convenient energy carrier as it can be transported to loads relatively easily. However, locating and constructing new transmission and distribution power lines is challenging and expensive. The cost of transporting electricity along power lines can constitute greater than 50% of the total cost at the point of end-use.7 Historically, hydrogen production via electrolysis has only been viable where large amounts of inexpensive electricity have been available or the high purity product gas was necessary in a downstream process. The potential environmental benefit of a hydrogen-based economy is hinged to a large degree on the ability to generate the gas from renewable resources in a costeffective manner. An apparently ideal solution is to use wind-generated electricity to electrolyze water. Today, hydrogen production via electrolysis only meets the U.S. Department of Energy (DOE) goals of $2–$3 per kilogram (kg) in large installations where electrolyzer capital costs are low, less than $800 per kilowatt (kW), and those having access to inexpensive electricity, less than $0.04 per kilowatt-hour (kWh).8 Electricity from large-scale wind farms in Class 4 or better resource can be generated in the range of $0.05–$0.08 kWh-1, not including today’s $0.019 kWh-1 Federal production tax credit.9,10 The out-of-pocket cost of fossil-fuels, whether for electricity production or as transportation fuels, has remained relatively low; limiting the expansion of renewable forms of energy. For example, if the external costs of production were taken into account the cost of coal-generated electricity would rise an additional $0.03–$0.06 kWh-1.11Further limiting market penetration of renewable sources is that fossil fuels continue to receive the bulk of tax incentives here in the U.S.12 The term renewable defines these technologies as driven by natural and sustainable processes which are inherently variable, not intermittent. Natural processes vary over time but are not subject to the on-off switching that, for example, a light bulb connected to a switch is subjected to. Advocates may want to begin training themselves to describe RE as variable, not intermittent, to better describe their naturally occurring behavior. RE sources of energy can provide cost-effective, emission-free electricity with zero- or low-carbon impact making it one of the preferred methods for supplying energy to society. The large-scale wind energy facilities being installed throughout the world are a testament to the growing demand, environmentally preferred and cost-effectiveness of this RE technology. 2 Electrolysis of Water “Personally, I think that 400 years hence the power question in England may be solved somewhat as follows. The country will be covered with rows of metallic windmills working electric motors, which in their turn supply current at a very high voltage to great electric mains. At suitable distances there will be great power stations where during windy weather the surplus power will be used for the electrolytic decomposition of water into hydrogen and oxygen…. In times of calm, the gases will be recombined in explosion motors working
Slide 60: 44 Kevin Harrison and Johanna Ivy Levene dynamos which produce electrical energy once more or more probably in oxidation cells.” Haldane, in his talk entitled, Daedalus or Science and the Future, Cambridge University, 1923.13 Hydrogen as an energy carrier and potentially widely-used fuel is attractive because it can be produced easily without emissions by splitting water. In addition, the readily available electrolyzer can be used in a home or business where off-peak or surplus electricity could be used to make the environmentally preferred gas. Electrolysis was first demonstrated in 1800 by William Nicholson and Sir Anthony Carlisle and has found a variety of niche markets ever since. Two electrolyzer technologies, alkaline and proton exchange membrane (PEM), exist at the commercial level with solid oxide electrolysis in the research phase. Electrolysis is defined as splitting apart with an electric current. Decomposition of the water occurs when a direct current (DC) is passed between two electrodes immersed in water separated by a non-electrical conducting aqueous or solid electrolyte to transport ions and completing the circuit. The voltage applied to the cell must be greater than the free energy of formation of water plus the corresponding activation and ohmic losses before decomposition will proceed. Ion transport through the electrolyte is critical as the purest of water would only contain small amounts of ions making it a poor conductor. Ideally, 39 kWh of electricity and 8.9 liters of water are required to produce 1 kg of hydrogen at 25 °C and 1 atmosphere pressure. Typical commercial electrolyzer system efficiencies are 56%–73% and this corresponds to 70.1–53.4 kWh/kg.14 The U.S. consumes somewhere between 140–150 billion gallons of gasoline per year equating to the same number of kilograms if we were to use only hydrogen for transportation. This would result in needing 330 billion gallons of water to make that much hydrogen. If the hydrogen were used in a fuel cell that is two times as efficient as an internal combustion engine in a car the amount of water required would be half. For comparison, gasoline production uses 300 billion gallons per year, domestic water use tops 4800 billion gallons per year and thermal electric power generation 70 trillion gallons per year.15 When comparing literature from fuel cell (FC) models with water electrolysis work it is important to remember the differences between the system anode and cathode. This basic understanding may be trivial to most but is many times confused when switching between the two processes. The anode is always the electrode at which oxidation occurs, where electrons are lost. The cathode is defined as the electrode at which electrons enter, where reduction takes place. In electrolysis the cathode is the electrode where H2 gas is created, in FC systems the anode is the electrode where H2 is introduced. 2.1 Alkaline The alkaline electrolyzer is a well-established technology that typically employs an aqueous solution of water and 25–30 wt.% potassium hydroxide (KOH). However, sodium hydroxide (NaOH), sodium chloride (NaCl) and other electrolytes have also been used. The liquid electrolyte enables the conduction of ions between the elec
Slide 61: Electrolysis of Water 45 trodes and is not consumed in the reaction but does need to be replenished periodically due other system losses. Typically commercial alkaline electrolyzers are run with current densities in the range of 100–400 mA cm-2. The reactions for the alkaline anode and cathode are shown in Eqs. 1 and 2 respectively, showing the hydroxyl (OH-) ion transport. 4 OH–(aq) O2(g) + 2 H2O(l) H2(g) + 2 OH–(aq) (1) (2) 2 H2O(l) + 2e– The first water electrolyzers used the tank design and an alkaline electrolyte.20 These electrolyzers can be configured as unipolar (tank) or bipolar (filter press) designs. In the unipolar design (see Figure 1), electrodes, anodes, and cathodes are alternatively suspended in a tank. In this design, each cell is connected in parallel and the entire system operated at 1.9–2.5 Vdc. The advantage to the unipolar design is that it requires relatively few parts, is extremely simple to manufacture and repair because individual cells can be taken offline while the remaining cells remain productive. The disadvantage is that it usually operates at lower current densities and lower temperatures.16 More recent unipolar designs include operation at high hydrogen pressure outputs (up to 6,000 psig). The bipolar design (Fig. 2), often called the filter-press, has alternating layers of electrodes and separation diaphragms that are clamped together. The cells are connected in series and result in higher stack voltages. Since the cells are relatively thin, the overall stack can be considerably smaller than the unipolar design. The advantages to the bipolar design are the reduced stack footprints, higher current densities, and its ability to produce higher pressure gas. The disadvantage is that it cannot be repaired without servicing the entire stack.16,17 Fortunately, it rarely needs servicing. Previously asbestos was used as a separation diaphragm, but manufacturers have replaced or are planning to replace this with new polymer materials such as Ryton®.18 2.2 Proton Exchange Membrane A second commercially available electrolyzer technology is the solid polymer electrolyte membrane (PEM). PEM electrolysis (PEME) is also referred to as solid polymer electrolyte (SPE) or polymer electrolyte membrane (also, PEM), but all represent a system that incorporates a solid proton-conducting membrane which is not electrically conductive. The membrane serves a dual purpose, as the gas separation device and ion (proton) conductor. High-purity deionized (DI) water is required in PEM-based electrolysis, and PEM electrolyzer manufacturer regularly recommend a minimum of 1 MΩ-cm resistive water to extend stack life. PEM technology was originally developed as part of the Gemini space program.16 In a PEM electrolyzer, the electrolyte is contained in a thin, solid ion-conducting membrane rather than the aqueous solution in the alkaline electrolyzers. This allows the H+ ion (proton) or hydrated water molecule (H3O+) to transfer from the anode side of the membrane to the cathode side, and separates the hydrogen and oxygen
Slide 62: 46 Kevin Harrison and Johanna Ivy Levene Fig. 1. Unipolar (tank) electrolyzer design. Fig. 2. Bipolar (filter-press) electrolyzer design.
Slide 63: Electrolysis of Water 47 Fig. 3. PEM cell components and reaction showing the positive anode and negative cathode electrodes. gases. Oxygen is produced at the anode side and hydrogen is produced on the cathode side. The most commonly used membrane material is Nafion® from DuPont. PEM electrolyzers use the bipolar design and can be made to operate at a high differential pressure across the membrane. DI water is introduced at the anode of the cells, and a potential is applied across the cell to dissociate the water. The protons (H+) are pulled through the membrane under the influence of an electric field and rejoin with electrons being supplied by the power source at the cathode to form hydrogen, H2, gas. PEM electrolyzers are operated at higher current densities (> 1600 mA cm-2) almost an order of magnitude higher than their alkaline counterparts. Stack efficiency decreases as current density increases but is necessary to increase hydrogen production to offset the higher capital costs of PEM cells. PEM advantages over alkaline include the ability to maintain a significant differential pressure across the anode and cathode avoiding the risk of high pressure oxygen. In addition, PEM electrolysis requires DI water but avoids the hazards surrounding KOH. The PEM anode and cathode reactions are described in Eqs. 3 and 4, respectively, and shown in Figure 3, 2 H 2 O → 4 H + + 4e _ + O 2 4 H + + 4e _ → 2 H 2 (3) (4) Figure 4 shows the major water and hydrogen components inside Proton Energy Systems HOGEN 40RE ® including the heart of the system: the PEM stack in front
Slide 64: 48 Kevin Harrison and Johanna Ivy Levene Fig. 4. Internal components of HOGEN 40RE. center. The compartment behind these systems (not viewable) contains the AC/DC power converter, ventilation fan, 24 Vdc power supply, system controller, radiator, and control relays. The RE version contains a DC/DC power converter and DC disconnects used to interconnect to a PV array. The combustible gas detector monitors hydrogen levels in this compartment and the oxygen phase separator. Figure 5 shows the step currents from the power supplies and the resulting hydrogen flow in standard cubic feet per hour (scfh) from the system. Hydrogen production ripple is caused by the internal hydrogen phase separator pumping down the
Slide 65: Electrolysis of Water 49 Fig. 5. Sample current step waveform from external power supplies and resulting hydrogen mass flow from HOGEN 40RE. accumulated water and desiccant drying tube crossover. These system functions cause a drop in system pressure resulting in varying hydrogen production output. The system efficiency (Eq. 5) is calculated using both the ancillary losses plus the stack energy. The system efficiency uses the higher heating value of hydrogen (39 kWh kg−1), the energy consumed by the stack (kWh), efficiency of the DC power supplies, and the balance of plant ancillary loads like pumps, valves, sensors and controller (kWh). Stack efficiency (Eq. 6) is determined by calculating the ideal cell potential at the operating temperature and pressure multiplied by the number of cells in the stack and then divided by the measured stack voltage, ⎛ kWh ⎞ HHV⎜ ⎜ kg ⎟ ⎟ ⎝ ⎠ System Efficiency = ⎛ Stack Input Energy (kWh ) ⎞ ⎜ ⎜ Power Supply Efficiency ⎟ + Ancillary Losses (kWh ) ⎟ ⎝ ⎠ HydrogenProduced (kg ) Stack Efficiency = Ideal Stack Potential Actual Stack Potential (5) (6) The HHV of hydrogen is 39 kWh kg−1 and the ideal stack potential is a function of temperature and pressure. All efficiencies are referenced to the HHV of hydrogen. The minimum amount of energy that must be consumed to split water into hydrogen
Slide 66: 50 Kevin Harrison and Johanna Ivy Levene Table 1. Constants for heat capacities of gases in ideal state and liquid water. A 3.249 3.639 8.712 B 0.422x10-3 0.506x10-3 1.25 x10-3 C 0 0 –0.18x10-6 D 0.083x105 –0.227x105 0 H2 O2 H2O and oxygen is known as the heat of formation (enthalpy) and corresponds to the HHV of hydrogen. 3 Fundamentals of Water Electrolysis 3.1 First Principles The overall reaction of the electrolysis cell (Eq. 7) provides the required stoichiometric coefficients for the products and reactant used in Eq. 8. The sign convention is positive for products and negative for reactants with analogous definitions for ΔB, ΔC and ΔD. Data for the constants A, B, C and D are thermodynamic properties and are reproduced in Table 1 from,19 H 2O + 2e _ → H 2 + ΔA → AH 2 1 O2 2 (7) 1 AO 2 − AH 2 O 2 (8) The model uses the specific heat capacity of water and gases in the ideal state to determine the standard Gibbs free energy of reaction, ΔG°, i.e., Eq. 9, o ⎡ ΔG o − ΔH o ΔH o 1 T ΔC o T ΔC P dT ⎤ o P 0+ 0+ ⎥ dT − ΔG = RT ⎢ 0 T0 R RT0 RT T T0 R T⎥ ⎢ ⎦ ⎣ ∫ ∫ (9) where ΔH0o and ΔG0o are the standard enthalpy and Gibbs energy of formation of liquid water, respectively, at reference temperature T0. The integrals of Eq. 9 take into account the temperature dependency of the heat capacities of the products and reactants and are reduced to Eqs. 10 and 11, ∫T ΔC p dT = (ΔA)T0 (τ − 1) + o 0 T ΔB 2 2 ΔC 3 3 ΔD ⎛ τ − 1 ⎞ T0 τ − 1 + T0 τ − 1 + ⎜ ⎟ (10) 2 3 T0 ⎝ τ ⎠ ( ) () ∫ o ⎡ ⎛ ΔC P dT ΔD ⎞⎛ τ + 1 ⎞⎤ 2 = ΔA ln τ + ⎢ΔB T0 + ⎜ ΔC T0 + 2 2 ⎟⎜ ⎟⎥ (τ − 1) ⎜ T0 R T ⎢ τ T0 ⎟⎝ 2 ⎠⎥ ⎝ ⎠ ⎣ ⎦ T (11)
Slide 67: Electrolysis of Water 51 where T is the reaction temperature (K), T0 is the reference temperature (298 K), R is he universal gas constant (8.314 J mol-1 K-1) and tau (τ) is defined as τ ≡ T/T0. Enthalpy is an intrinsic property of a substance and a function of temperature and pressure.19 In practice the Gibbs free energy is the net internal energy available to do work, less work done by changes in pressure and temperature.20 Exergy, on the other hand, is defined as the total amount of work that can be harnessed and becomes more relevant in high-temperature and high-pressure electrolysis. At standard temperature and pressure (STP, 25 °C, 1atm) Gibbs free energy of formation is defined as the point of zero energy and is used to calculate the change in energy of a system. The reversible (i.e., the minimum) voltage required to electrolyze water into hydrogen and oxygen is determined by the change of Gibbs free energy of formation between the products and reactants. As described above, the Gibbs free energy of formation is not constant; it changes with temperature and state (liquid or gas),20 Eo = − ΔG o zF (12) where Eo is the theoretical minimum reversible voltage of a cell, z is the number of electrons (2) taking part in the reaction and F is Faraday’s constant (96,485 Coulomb mol-1). The actual voltage (Vcell) required to decompose water at any significant rate will require Vcell be greater than Eo. The difference between the voltages is known as overpotential, polarization or simply losses. The Nernst potential (Vn) of Eq. 13 accounts for changes in the activity of the reaction and for nonstandard conditions, 1/2 ⎞ ⎛ RT ⎜ PH 2 PO 2 ⎟ Vn = Eo + ln 2 F ⎜ PH 2 O ⎟ ⎜ ⎟ ⎝ ⎠ (13) where PH2, PO2 and PH2O represent the partial pressures of hydrogen, oxygen and water respectively. The partial pressure of water is determined with the empirical formula from Ref. 21 and is shown in Eq. 14. This relationship enables the partial pressure of water to be calculated from experimental data as the temperature of the DI water into the stack anode varies, ⎤ ⎡ Tc PH 2 O = 610.78 exp ⎢ (17.2694)⎥ ⎦ ⎣ (Tc + 238.3) (14) where Tc is the temperature into the stack in °C and should not be confused with T and T0 from earlier that have units of Kelvin. The partial pressures of hydrogen and oxygen are determined using measurements from the stack cathode and anode and Eqs. 15 and 16, PH 2 = PC − PH 2 O (15)
Slide 68: 52 Kevin Harrison and Johanna Ivy Levene PO 2 = PA − PH 2 O (16) where PC and PA are the experimental pressures (atm) of the cathode and anode respectively. 3.2 Overpotentials Water electrolysis is an electrochemical reaction where water is split into hydrogen and oxygen in the presence of a catalyst and applied electric field. As current density increases the cell losses due to membrane, electrode, and interfacial resistances dominate and are referred to as ohmic overpotential. At equilibrium (i.e., no current) there exist dynamic currents, measured in amps, at each electrode and are a fundamental characteristic of electrode behavior. The anode and cathode exchange current densities can be defined as the rate of oxidation and reduction respectively. The exchange current density is a measure of the electrode’s ability to transfer electrons and occurs equally in both directions resulting in no net change in composition of the electrode.22 A large exchange current density represents an electrode with fast kinetics where there is a lot of simultaneous electron transfer. A small exchange current density has slow kinetics and the electron transfer rate is less. The anode and cathode exchange current density’s can be fitted exponentially as a function of temperature. Experimentally is has been determined and intuition suggests that as temperatures increase the faster a chemical reaction will proceed. Arrhenius was the first to recognize that the higher kinetic energy due to higher temperature results in lowering the activation potential.22 Lower activation losses reduces the amount of energy for the reaction to proceed, thus increasing stack efficiency. The conductivity coefficient was fitted linearly because it is primarily a function of current. (i) Activation overpotential Electrochemical reactions possess energy barriers which must be overcome by the reacting species. This energy barrier is called the ‘activation energy’ and results in activation overpotential, which are irreversible losses (heat) in the system. Activation energy is due to the transfer of charges between the electronic and the ionic conductors. The activation overpotential is the extra potential necessary to overcome the energy barrier of the rate-determining step of the reaction to a value such that the electrode reaction proceeds at a desired rate.23 The anode (Eq. 17) and cathode (Eq. 18) activation overpotentials, ηA and ηC, represent irreversible losses of the PEM stack and dominate the overall overpotential at low current densities: ηA = ⎛i RT ln⎜ α a zF ⎜ ia, o ⎝ ⎞ ⎟ ⎟ ⎠ (17)
Slide 69: Electrolysis of Water 53 ηC = ⎛i⎞ RT ⎟ ln⎜ α c zF ⎜ ic, o ⎟ ⎝ ⎠ (18) where αa and αc are the anode and cathode electron transfer coefficients respectively, ia,o and ic,o are the anode and cathode exchange current densities (A cm-2) respectively and i is the current density of the stack (A cm-2). The electron transfer coefficient is a measure of the symmetry of the activation energy barrier and can range from zero to unity.22 The higher the exchange current density the easier it is for reaction to continue when current is supplied to the stack. The cathode exchange current density is thus not the limiting parameter of the activation overpotential term and is often ignored. The current density (i) normalizes the stack current (I) to the active area of the cell. (ii) Ohmic Ohmic losses occur because of resistance to the flow of ions in the solid electrolyte and resistance to flow of electrons through the electrode materials. Because the ionic flow in the electrolyte obeys Ohm’s law, the ohmic losses can be expressed by Ohm’s law. The ohmic overpotential, ηo of Eq. 19 is a function of the stack current density (i), membrane thickness (ϕ), and the conductivity of the stack (σ), ηo = ϕ i σ (19) where σ (Siemen cm–1) represents the sum of membrane resistance to ion transfer and bundled electrical resistances of electrodes and interconnections within the stack. As the stack ages, internal polarization losses increase and stack voltage will increase for a given current. (iii) Anode exchange current density Small exchange current densities exhibited by the anode give rise to slow charge transfer that is in turn an activation-controlled process. Unfortunately, for the case of both the anode and cathode exchange current densities, the range of values varies dramatically from author to author because of the various operating conditions, cell construction, and stack configurations. Typical values found in the literature for the anode exchange current density range from 10–7–10–12 A cm–2.24,25 (iv) Cathode exchange current density The cathode exchange current density is typically four orders of magnitude greater than the anode exchange current density and supported by Choi and Berning in Ref. 24 and 25. The anode side is therefore limiting the reaction and dominates the activation overpotential.
Slide 70: 54 Kevin Harrison and Johanna Ivy Levene (v) Conductivity Specifically speaking, membrane conductivity represents only the membrane’s resistance to flow of protons (H+) and is highly dependant on its thickness (ϕ) and water content. Electrical resistance of electrodes, cell interconnects, and the formation of any insulating layer on the electrode surface are all bundled under the conductivity term. Voltage decreases for a given current as temperature increases and can be controlled to improve stack efficiency. 4 Commercial Electrolyzer Technologies Electrolyzers produced in 2006 range in sizes from less than 0.1 kg day-1 to over 1000 kg day-1. The smallest systems, in the under 0.5-kg day-1 range, are used for the production of hydrogen at the lab scale, and are designed as hazard-free alternatives to high pressure gas cylinders.26 Electrolyzers are sized to meet the system requirements of the existing high purity hydrogen markets. However, only systems above the 0.55-kg day-1 production rate are viable for producing hydrogen as a transportation fuel, and the transportation fueling market could use systems larger than the existing 1000-kg day-1 unit. A 0.055-kg day-1 system would produce 200 kg year-1, enough to fuel a single car. A 1000-kg day-1 system would produce 365,000 kg year-1, enough to fuel 1,800 vehicles. Both of these values assume a 100% capacity factor on the electrolysis system. Typically, these units are capable of operations in the high 90% range.27 The number of cars served by a system was determined by calculating that a car requires approximately 200 kg of hydrogen year-1. This 200-kg requirement assumes that on average a car travels 12,000 miles year-1, and that a vehicle will travel 60 miles kg-1 of hydrogen.28 Electrolysis systems that could be used for transportation fuel production can be categorized into five different size ranges: home, small neighborhood, neighborhood, small forecourt and forecourt.28 The term forecourt refers to a refueling station. The number of cars served and hydrogen production rate for each size are as follows: • • • • • The home size will serve the fuel needs of 1–5 cars with a hydrogen production rate of 200–1000-kg hydrogen year-1. The small neighborhood size will serve the fuel needs of 5–50 cars with a hydrogen production rate of 1000–10,000-kg hydrogen year-1. The neighborhood size will serve the fuel needs of 50–150 cars with a hydrogen production rate of 10,000–30,000-kg hydrogen year-1. The small forecourt size, which could be a single hydrogen pump at an existing station, will serve 150–500 cars with a hydrogen production rate of 30,000–100,000-kg hydrogen year-1. A full hydrogen forecourt will serve more then 500-cars per year with a hydrogen production rate of greater then 100,000-kg hydrogen year-1. A sampling of electrolyzer manufacturers in 2006 are presented in Table 2.
Slide 71: Electrolysis of Water Table 2. Commercial electrolyzer manufacturers and selected performance data. Manufacturer AccaGen SA Upper Pressure Lower capacity capacity -1 -1 (kg day ) (kg day ) Alkaline, acid 0.043 215.7 up to 200 barg and PEM Alkaline 0.75 300. up to 6500 psig Alkaline 6.47 1639.4 up to 30 barg Unknown 25.9 1078.6 30 barg Technology Location Switzerland Connecticut, USA Germany Germany 55 Ref. 29 6 5 30 Avalance ELT Gesellschaft fur Hochleistungselktrolys eure zur Wasserstofferzeugung Giner PEM Hamilton Sundstrand Hydrogenics 11.8 129.3 10.8 1639.4 10.8 129.4 0.01 2.2 1.1 0.005 0.647 6.5 2.6 11.8 129.3 129.4 1639.4 539.3 1046.2 0.01 30.2 12.9 0.026 PEM PEM and alkaline Industrie Haute Techno- Alkaline logie SA Linde Unknown Norsk Hydro Alkaline Peak Scientific Ion exchange membrane PIEL division of ILT Alkaline Technology s.r.l. Proton Energy System PEM Schmidlin-DBS AG Siam Water Flame Co. Teledyne Energy Systems Treadwell Corporation Membrane Alkaline Alkaline PEM 3000 psig Massachusetts, USA up to 100 psig Connecticut, USA up to 363 psig Canada up to 32 barg Monthey, Switzerland 25 barg Germany 15 barg Norway 0-100 psig Scotland 3 barg Italy up to 218 psig 1–155 psig Connecticut USA Neuheim, Switzerland and Padova, Italy Bangkok, Thailand Maryland, USA Connecticut, USA 31 32 33 34 35 36 30 31 32 37 38 39 40 0.647 unknown 129.4 up to 115 psig 22.0 up to 1100 psi 5 Electrolysis System The system used to produce hydrogen via electrolysis consists of more than just an electrolyzer stack. A typical electrolysis process diagram is shown in Fig. 6.45 The primary feedstock for electrolysis is water. Water provided to the system may be stored before or after the water purification unit to ensure that the process has adequate feedstock in storage in case the water system is interrupted. Water quality requirements differ between electrolyzers. Some units include water purification inside their hydrogen generation unit, while others require an external purification unit, such as a deionizer or reverse osmosis unit, before water is fed to the cell stacks. The high purity water will be mixed with KOH if the system is an alkaline system before being introduced to the hydrogen generation unit. Note that PEM units will not a KOH feed, as no electrolytic solution is needed. Each system has a hydrogen generation unit that integrates the electrolysis stack, gas purification
Slide 72: 56 Kevin Harrison and Johanna Ivy Levene Fig. 6. Process flow diagram for a water electrolyzer system. and dryer, and heat removal. Electrolyte circulation is also included in the hydrogen generation unit in alkaline systems. The hydrogen generation system is usually en closed in a container or is installed as a complete package. Oxygen and purified hydrogen are produced from the hydrogen generation unit. If desired, a compressor, hydrogen storage, and oxygen storage can be added to the system. A second feedstock needed for electrolysis is electricity. Typically electricity is not considered a feedstock but a utility; however it is a critical component in the splitting of the water molecule into hydrogen and oxygen. An electrolyzer typically will convert supplied AC to DC, as the stack requires DC to split water. Typical utilities that the electrolysis systems need include electricity for other peripheral equipment; cooling water for the hydrogen generation unit; prepressurization gas; and instrumentation gas (Fig. 6). 5.1 Energy Efficiency Energy efficiency is defined as the higher heating value (HHV) of hydrogen divided by the energy consumed by the electrolysis system per kilogram of hydrogen pro-
Slide 73: Electrolysis of Water Table 3. Efficiency of selected electrolyzers in the market. Energy required system (kWh/kg) AccaGen SA Hydrogenics IHT PIEL division of ILT Technology s.r.l. 74.5–52.4 53.4 46.7–51.2 77.9 System efficiency (%) 52–74 73 84–76 50 Production pressure Up to 200 bar 363 psi Up to 32 bar 3 bar 57 Ref. 29 37 29 31 duced. See Eqs. 5 and 6 for details on this calculation. HHV is used as opposed to the lower heating value (LHV) because in commercial electrolyzers in 2005, the water electrolyzed is in a liquid state. To further clarify why HHV is used, the reaction of the formation of water is: H2 + ½ O2 H2O + energy (20) At 25 °C and 1 atm, the heat of formation of liquid water, or the energy released when water is formed in the reaction above is 39 kWh kg-1 of hydrogen. This value is the higher heating value (HHV) of hydrogen. The heat of formation of steam is 33 kWh/kg of hydrogen, and is the lower heating value (LHV) of hydrogen. The electrolysis reaction is the opposite of the formation of water reaction: H2O + energy H2 + ½ O2 (21) As a result, the amount of energy needed to create hydrogen from liquid water using electrolysis is 39 kWh kg-1. The reason this distinction is important is because if using the lower heating value the efficiency of electrolyzers is misrepresented. If LHV is used to calculate electrolyzer efficiencies, the maximum hydrogen system efficiency is 33/39 or approximately 85%. Thus a 100% efficient electrolysis system on an LHV basis is actually thermodynamically impossible if you are electrolyzing liquid water. That is to say that an electrolyzer that converts every kWh of input energy into hydrogen energy will have only 85% efficiency, even though there are no losses.41 The energy efficiency of several electrolyzers is shown in Table 3. The energy efficiency ranges of commercial systems ranges from 47–77 kWh/kg (83–51%). An efficiency goal for electrolyzers in the future has been reported to be in the 46.9 kWh kg-1 range, or a system efficiency of 83%.42 This 83% includes compression of the hydrogen gas to 6000 psig. Currently most electrolyzers reach a pressure ranging from 0–500 psig for the power requirements presented, with a few research stage electrolyzers reaching pressures in the 3000–6500-psig range. So most electrolyzers would need additional energy input beyond what is presented in the table below to compress to fueling pressures. Note that the above values are energy requirement of the entire electrolysis system, excluding any additional compression beyond what the stack produces. This is an appropriate way to calculate system efficiency. As an example, the electrolyzer
Slide 74: 58 Kevin Harrison and Johanna Ivy Levene Fig. 7. Hydrogen costs via electrolysis with only electricity costs considered. stack alone for a Hydrogenics system requires 46.8 kWh kg-1 (4.2 kWh Nm-3), which corresponds to 83% efficiency when you divide the HHV of hydrogen by the electrolyzer power requirement. However, when you include the rectifier and auxiliaries the energy requirement becomes 53.4 kWh kg-1 (4.8 kWh Nm-3) or 73% efficient. 5.2 Electricity Costs Electricity costs are a key component when producing hydrogen via electrolysis. A boundary analysis was completed to determine the effects of electricity price on hydrogen costs, and the results are shown in Fig. 7.28 For each electrolyzer, the specific system energy requirement is used to determine how much electricity is needed to produce hydrogen; no capital, operating or maintenance costs are included in the calculation. The system energy requirement used is the lowest energy requirement reported for each manufacturer. This graph shows that, at current electrolyzer efficiencies, in order to produce hydrogen at lower than $3.00 kg-1, electricity costs must be between 4 and 5.5¢ kWh-1. In order to produce hydrogen for less than $3.00 kg-1 with a system that is 100% efficient, electricity prices must be less than 7.5¢ kWh-1. The U.S. Department of Energy’s Energy Information Administration (EIA) reports 2002 industrial, commercial, and residential electricity prices at 4.83, 7.89, and 8.45¢ kWh-1, respectively.28 Thus, if only electricity costs were incurred, current electrolyzers could produce hydrogen for $3.00 kg-1 at industrial electricity prices; an ideal system could produce hydrogen for $3.00 kg-1 at slightly lower then commercial prices. This analysis shows that regardless of any additional cost elements, electricity costs will be a major price contributor.
Slide 75: Electrolysis of Water 59 6 Opportunities for Renewable Energy Integrating electrolyzers with renewable energy system can present challenges as well as unique benefits. Currently most renewable energy systems produce power and interconnect with thte electrical grid via some form of power electronics (PE). To use electrical grid power, today’s commercial electrolyzers also have some type of power electronics interface that can represent a significant portion of the overall system cost.43 The power electronics convert alternating current (AC) from the grid to direct current (DC) power required by the electrolysis cell stack. In addition to the DC requirements of the stack, the system also consumes additional AC power for the balance of plant or ancillary loads. At least one electrolyzer manufacturer offers a version of an electrolyzer that can accommodate a connection to photovoltaic (PV) panels in addition to having the standard AC to DC converter for utility operation.44 The additional power electronics, incorporating maximum power point tracking (MPPT), converts all available DC power from the PV array to run the electrolysis stack. This system appears to be one of the firsts to incorporate dedicated PE to interface with a PV source. In addition to using PV systems as electricity sources, wind energy can also be used. Today, the majority of wind to hydrogen demonstration projects are focused on installing commercially available electrolyzers and powering them from the AC power from wind turbines.45–47 In these projects the AC from the wind turbines is sent out onto the grid and the electrolyzers tied into the grid achieving a loose coupling of source and load. Scheduling the power to the electrolyzer, based on an output signal from the wind turbines, is relatively straightforward in this case. Capital costs of electrolysis equipment range from just under $1000 kW-1 for the largest alkaline systems to over $10,000 kw-1 for small proton exchange membrane (PEM) electrolyzers.28,48 Merely taking an off-the-shelf wind turbine with its own PE and commercial electrolyzer with its own PE reduces overall energy transfer from the wind to hydrogen system. The potential exists to characterize electrolyzer performance under varying input power and design a single PE package and intelligent controller to achieve direct coupling between the stack and wind turbine output. This topology would not only eliminate the redundancy of power electronics that exists in the wind turbine and electrolyzer but also achieve gains in system cost and robustness. Characterizing the system demands of renewable energy sources and the requirements of the hydrogen-producing stack appears to have synergistic benefits. Ultimately, the detailed understanding of both systems and design of the directly coupled wind to electrolysis would reduce the cost of renewably generated hydrogen. In renewable-based energy systems PEM electrolysis seems to have an advantage over alkaline in that the thin membrane and ion transport mechanism can react to nearly instantaneously with the rapidly changing energy output of renewable sources, especially wind. Stacks involving the circulation of a liquid electrolyte have inherently more inertia in the transport of ions in solution than the PEM systems. On the turbine side, variable-speed wind turbines (which will soon be the norm as a result of enhanced energy capture relative to constant-speed machines) rely on power electronics to convert the variable frequency, variable voltage AC produced at the generator to DC. Small turbines used in battery-charging applications stop here;
Slide 76: 60 Kevin Harrison and Johanna Ivy Levene however, larger turbines used to connect to the grid must then convert the DC back to AC at grid frequency: 60 Hertz (Hz). It is important to note that because of the economies of scale, it is the large wind turbines that are achieving highly competitive energy costs and will likely be the device of choice in large-scale wind-to-hydrogen operation. The small wind-to-hydrogen systems (< 20 kW) being studied today are systems incorporating a common DC bus fixed with a battery bank to which the wind turbine and electrolyzer as well as fuel cells and PV panels are connected. Typically, the wind turbine is of the battery-charging type, which requires connection to a constant voltage DC bus (hence, the battery bank) and incorporates power electronics to convert wild AC to DC and to regulate power output. The electrolyzer stack accepts DC power input but the system would also include power electronics to regulate power input and possibly convert DC at one voltage level to another. There are a number of weaknesses with this configuration, namely a redundancy of power electronics leading to increased cost and potential for failure. The inability to match wind turbine power output to electrolyzer power requirements because of separate power electronic controllers ultimately results in reduced energy capture. An advanced topology would be the direct coupling of an electrolyzer with a wind turbine. This would allow hydrogen production that is proportional to the available wind energy and reduce electricity storage requirements.. The single point of control will allow the matching of wind turbine and electrolyzer electrical characteristics, thereby increasing the energy capture of the wind turbine. Finally, this solution will eliminate the need for a constant voltage DC bus and provide a true test of electrolyzer operation under fluctuating power-input conditions. Renewable electrolysis can help overcome one of the key barriers to realizing a hydrogen-based economy by replacing the carbon-intensive one that exists today. There is an excellent opportunity for research in renewable hydrogen production both in terms of understanding the operation of the electrolyzer under variable sources and optimizing, in terms of efficiency, cost, and robustness, the link between a renewable source and electrolyzer stack. 7 Conclusions There exists an opportunity to change the face of our energy consumption from one of polluting our air, water, and land to one more in harmony with the environment. The environment and the economy are often at odds for resources, but it does not have to be that way. Renewable hydrogen seems to possess the ability to transition the world’s carbon-based economy into a near-carbon-free economy. Hydrogen can be extracted from all fossil fuels as well as split from water using the electricity from RE sources. However, without sequestering the climate-altering CO2 produced using fossil fuels, the environmental benefits are completely lost and may be even worsened by the transition to hydrogen as an energy carrier. If the environmental benefits of the long-term development of the hydrogen economy are to be realized, the production of hydrogen via electrolysis from RE sources will be a vital component. Today’s commercially available electrolyzers are
Slide 77: Electrolysis of Water 61 designed to use grid electricity, produce well regulated DC power to the electrolysis stack, and condition the output gas for applications different than that required by PEM fuel cells. The key element in hydrogen production from any electrical source is the electrolyzer stack that converts water and electricity into hydrogen, oxygen and heat. The electrolyzer stack is inherently a nearly constant, low-voltage, DC device requiring some form of control system and power electronics to connect it to a highvoltage, AC source of power. The primary intent of this work is to design, build and verify a system capable of accurately varying important system variables that are normally strictly monitored and controlled by the commercial electrolyzers containing the same PEME stack. The goal of the experimental characterization of the stack, under varying conditions and power, is to enable an optimized interconnection between the stack and RE source. Such a coupled system specifically designed with the RE source in mind would reduce the overall cost of independent stand-alone systems and may eliminate the need for electrical storage components. Electrical power provided to the electrolyzer in such a system would be controllable with excess power provided to the grid. Thus, a combined system would have more dispatch-ability than a wind-electric turbine or PV array alone. Such dispatchability might be used to provide the utility with a measure of control over the renewable energy systems total output that does not exist in current renewable based, gridconnect only systems. Using a variable RE source, like wind or PV, to generate the hydrogen gas will guarantee this energy carrier will be produced with nearly zero emissions. References 1. B. Suresh, S. Schlag, and Y. Inogucji, Chemical Economics Handbook Marketing Research Report, SRI Consulting, 2004. 2. M. Momirlana and T.N.Veziroglub, The properties of hydrogen as fuel tomorrow in sustainable energy system for a cleaner planet, International Journal of Hydrogen Energy 30, 795 (2005). 3. Hemoglobin, in Wikipedia, the Free Encyclopedia, Retrieved on June 22, 2006 from http://en.wikipedia.org/wiki/Hemoglobin. 4. Annual Energy Review 2004, EIA, http://www.eia.doe.gov/emeu/aer/pdf/aer.pdf, Report No. DOE/EIA-0384 (2004), August 2005. 5. Safe Use of Hydrogen and Hydrogen Systems, NASA Training Center, 2006. 6. J. B. Heywood, Fueling our transportation future, Scientific American. 295 60 (2006). 7. W. E. Winshe, K. C. Hoffman, and F. J. Salzano, Hydrogen: Its future role in the nation's energy economy, Science 180 1325 (1973). 8. J. Levene, B. Kroposki, and G. Sverdrup, Wind Energy and Production of Hydrogen and Electricity - Opportunities for Renewable Hydrogen, NREL Report No. CP-560-39534, 2006. 9. Comparative Cost of Wind and Other Energy Sources, American Wind Energy Association (AWEA), http://www.awea.org/pubs/factsheets/Cost2001.PDF, 2001. 10. The Economics of Wind Energy, American Wind Energy Association, http://www.awea.org/pubs/factsheets/EconomicsOfWind-Feb2005.pdf, February 2005.
Slide 78: 62 Kevin Harrison and Johanna Ivy Levene 11. R. L. Ottinger, D. Wooley, D. R. Hodas, N. A. Robinson, and S. E. Babb, Pace University Center for Environmental Legal Studies; Environmental Costs of Electricity, Oceana Publications, New York, 1990. 12. K. Silverstein, Clean tech goes mainstream, EnergyBiz Insider, http://www.energy central.com/centers/energybiz/ebi_detail.cfm?id=164, CyberTech Inc., 2006. 13. J. B. S. Haldane, DAEDALUS or Science and the Future, E. P. Kutton & Company, New York, 1923. 14. Technology Brief: Analysis of Current-Day Commercial Electrolyzers, NREL, Golden, CO NREL/FS-560-36705, September 2004. 15. J. A. Turner, Sustainable hydrogen production, Science. 305 972 (2004). 16. A. Konopka and D. Gregory, Hydrogen Production by Electrolysis: Present and Future, in 10th Intersociety Energy Conversion Engineering Conference, IEEE Cat. No. 75CHO 983-7 TAB, 1975. 17. W. Kincaide, Alkaline Electrolysis: Past, Present and Future, in Hydrogen for Energy Distribution, Institute of Gas Technology, 1978. 18. Ryton® PPS - Chevron Phillips Chemical Company LLC, Retrieved on June 29, 2006, from http://www.cpchem.com/enu/ryton_pps.asp, 2006. 19. J. M. Smith, H. C. Ness, and M. M. Abbott, Introduction to Chemical Engineering Thermodynamics, 6th ed., Mc Graw Hill, New York, 2001. 20. J. Larminie and A. Dicks, Fuel Cell Systems Explained, 2nd ed., John Wiley and Sons, Ltd., West Sussex, England, 2002. 21. T. Padfield, Moisture in air, Equations describing the physical properties of moist air, retrieved on January 9, 2006, from http://www.natmus.dk/cons/tp/atmcalc/atmoclc1.htm, 1996. 22. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Fundamentals and Applications, 2nd ed., John Wiley & Sons, Inc., New York, 2001. 23. S. H. Chan, K. A. Khor, and Z. T. Xia, A complete polarization model of a solid oxide fuel cell and its sensitivity to the change of cell component thickness, Journal of Power Sources 93 130 (2001). 24. T. Berning and N. Djilali, "Three-Dimensional Computational Analysis of Transport Phenomena in a PEM Fuel Cell — A Parametric Study," Journal of Power Sources, vol. 106, pp. 284-292, 2003. 25. P. Choi, D. G. Bessarabov, and R. Datta, A simple model for solid polymer electrolyte (SPE) water electrolysis, Solid State Ionics, 175 535 (2004). 26. UHP Zero Air and Hydrogen Generators for Fuel Gas, 2006 <http://www.chromtech. com/online_catalog/instruments/gas_gen/Hydrogen_ZeroAir2.pdf>, p. 3. 27. Peak Scientific: The Future of Gas Generation, 2007, Peak Scientific, 2006. <http://www. peakscientific.com/peak_products/product_detail.asp?GasID=2&ApplicationID=11& ProductID=25>. 28. J. Ivy, Summary of Electrolytic Hydrogen Production: Milestone Completion Report, NREL, Golden, CO, NREL/MP-560-35948, April 2004. 29. AccaGen SA – Homepage, Vol. 2006, AccaGen SA, 2006 <http://www.accagen.com/>. 30. GHW - Gesellschaft für Hochleistungselektrolyseure zur Wasserstofferzeugung mbH, 2006 <http://www.ghw-mbh.de/english/01_home/index.html>. 31 Welcome to Giner Inc..Vol. 2006, Giner, Inc. and Giner Electrochemical Systems, LLC 2006 <http://www.ginerinc.com/>. 32. Hamilton Sundstrand - System Solutions, Vol. 2006, Hamilton Sundstrand, 2006 <http://www.snds.com/ssi/ssi/SystemSolutions/h2gen.html>. 33. Hydrogenics, On-site hydrogen generation stations, hydrogen storage and compression, Vol. 2006, 2006 <http://www.hydrogenics.com/onsite/products.asp>. 34. IHT, Clean hydrogen solutions. Vol. 2006, 2006 <http://www.iht.ch/>.
Slide 79: Electrolysis of Water 63 35. Linde, Hydrogen Solutions - Supply > On-Site > Ecovar® | Linde Gas Division, Vol. 2006, 2006 <http://www.linde-gas.com/international/web/lg/com/likelgcom30.nsf/>. 36. Hydro, Hydrogen Technologies, Vol. 2006, 2006 <http://www.hydro.com/electrolysers/ en/>. 37. On-site hydrogen generation stations, hydrogen storage and compression, Vol. 2006: Hydrogenics Corporation, 2006 <http://www.hydrogenics.com/onsite/products.asp>. 38. Clean hydrogen solutions, Vol. 2006, IHT, 2006 <http://www.iht.ch/>. 39. Hydrogen Solutions - Supply > On-Site > Ecovar® | Linde Gas Division, Vol. 2006, Linde, 2006 <http://www.linde-gas.com/international/web/lg/com/likelgcom30.nsf/>. 40. Hydrogen Technologies, Vol. 2006, Norsk Hydro Electrolysers AS, 2006, <http://www. hydro.com/electrolysers/en/>. 41. R. Merer, RE: H2A Update, personal e-mail, 17 Mar. 2004. 42. 3.1 Hydrogen Production, Multi-Year Research, Development and Demonstration Plan: Planned program activities for 2003-2010, Washington DC, US Department of Energy, Energy Efficiency and Renewable Energy, January 21, 2005, p. 51. 43. S. Hock, C. Elam, and D. Sandor, Can we get there? Technology advancements could make a hydrogen electric economy viable—and expand opportunities for all renewables, Solar Today, May-June 2004, p.24-28. 44. Proton Energy Manufactures Three Families of HOGEN® Hydrogen Generation Systems, Proton Energy Systems - Products - HOGEN H Series, Hogen S Series, Hogen GC, retrieved on July 6, 2006, from http://www.protonenergy.com/products.html, 2005. 45. Hybrid Wind Energy System, Retrieved on June 26, 2006, from http://energy.coafes.umn. edu/windenergy, University of Minnesota, Research and Demonstration Center, 2005. 46. Basin electric joins pilot project to marry wind, hydrogen, Energy Services Bulletin, retrieved on May 22, 2006, from http://www.wapa.gov/es/pubs/esb/2004/December/ dec045.htm, 2006. 47. G. Schroeder, Transition to the Hydrogen Age Transition to the Hydrogen Age: Myths and Realities, retrieved on June 29, 2006 from http://fcgov.com/utilities/pdf/eps06-fuelhydrogen.pdf, 2006. 48. A. F. G. Smith and M. Newborough, Low-Cost Polymer Electrolysers and Electrolyser Implementation Scenarios for Carbon Abatement, Heriot-Watt University, Edinburgh, Report to the Carbon Trust and ITM-Power PLC, November 200.
Slide 80: This page intentionally blank
Slide 81: 4 A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen Robert McConnell NREL, Golden, CO 1 Direct Conversion of Concentrated Sunlight to Electricity Concentrating sunlight through the use of mirrors or lenses is historically associated with the generation of heat. Legend has it that Archimedes used mirrors and the sun’s energy to set attacking Roman ships on fire.1 Children often discover that magnifying lenses can burn paper or tree leaves, sometimes after first burning their fingers. At the turn of the 19th century, several inventors and engineers used heat from solar concentrators to operate steam engines to pump water and later to generate electricity by means of rotating machinery.2 Several solar concentrator technologies being developed today use heat and rotating machinery. Large systems based on this technology have been generating electricity successfully in California since the 1980s. With the invention of the modern solar cell in 1955, scientists and engineers began developing a revolutionary new technology—photovoltaics (PV)—for converting sunlight directly into electricity. Photovoltaic technologies are based on hightechnology semiconductors in which the sun’s photons liberate an electric charge within the semiconductor and that charge is driven by an internal electric field to electrodes connected to an external load. In the 1960s and 1970s, these marvelous solar batteries were the only reliable power sources providing electricity for the first space satellites of the Cold War, as well as for later communications satellites. In the 1970s, engineers demonstrated that concentrating sunlight and focusing the equivalent of hundreds of suns onto a solar cell could generate hundreds of times more electricity.2 However, not all of the sunlight is converted to electricity, and engineers designed heat sinks to transfer heat away from the solar cells or actively cooled the solar cells using a cooling fluid because efficiency decreases when the solar cells heat up. So, for efficient electricity production, this solar heat was wasted. As we shall see, that ordinarily wasted heat can be used to augment the electrolytic
Slide 82: 66 Robert McConnell production of hydrogen above the already dramatically high efficiencies of concentrator solar cells. We will describe how solar concentrator photovoltaic (CPV) systems can produce hydrogen from water at high efficiency. But before we do, we need to understand more about the characteristics of CPV systems and why they are just now entering into the world’s energy markets. Unlike flat-plate PV systems seen on roofs around the world today, solar concentrators need to track the sun. To focus sunlight onto a solar cell throughout the day, a tracking mechanism points the solar concentrator structure at the sun as it crosses the sky. Electrical output drops dramatically if the sun is not focused on the cell or if clouds block the sun. The resulting system consists of a solar concentrator using mirrors or lenses, a tracking mechanism, solar cells, and a heat sink. As shown in Figs. 1 and 2, these CPV systems are quite different from the flat-plate PV panels generating electricity throughout the world today. Sun-tracking also increases the daily energy production above that of non-tracking flat-plate PV panels. Utilities are interested in all solar tracking technologies because of this additional value in energy production. 2 The CPV Market Sometimes technologies are developed that few people buy. For decades, CPV appeared to be one of those technologies with no market and no customers. Of some 1,500 megawatts (MW) of PV sold throughout the world in 2005, less than 1 MW were CPV systems. Although most of the world’s PV installations in 2005 were on rooftops, CPV systems had not been developed for roofs. As the photos show, typical CPV systems are large and more suitable for a utility customer, although several Fig. 1. Several 35-kilowatt (kW) CPV systems built by Amonix in Torrance, California, are installed at an Arizona Public Service power plant. The system uses Fresnel lenses to concentrate sunlight. The pickup truck in the shade gives an idea of size.
Slide 83: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 67 Fig. 2. Several 25-kW CPV systems built by Solar Systems in Hawthorn, Australia, and installed on aborigine lands. These systems use mirrors for concentration (see www.solarsystems.com.au). Note the people in the foreground for an idea of size. companies are now developing smaller CPV products for rooftop markets. CPV systems generate little electricity in areas with cloud cover and, not surprisingly, CPV researchers often live in sunny areas such as the southwestern United States, Israel, Spain, and Australia. It is frequently stated that CPV will be competitive only in these sunny, cloudless regions; however, studies of CPV in less sunny locations suggest that the costs could still be competitive with those of other PV technologies if the solar cell efficiencies are high enough.3,4 Nevertheless, CPV systems will certainly penetrate their first markets in these sunny areas—just as the first wind systems were installed in very windy locations before going into less windy sites as their costs declined. In the 1980s, the U.S. Department of Energy (DOE) and the Electric Power Research Institute (EPRI), the research organization for electric utilities, funded CPV projects for utility applications. Both organizations curtailed their CPV studies in the early 1990s as rooftop PV markets started to become dominant. Recently, however, two companies—Amonix in California and Solar Systems in Australia—found customers for their systems shown in Figs. 1 and 2. Amonix now has a 10 MW/year production facility in a joint venture with the developer, Guascor, in Spain. And Solar Systems has been installing hundreds of kilowatts of CPV systems in Australian outback locations where electricity is expensive due to high transport cost of diesel fuel for diesel generators.5 For almost two decades, these two companies have been persistent and innovative in developing several generations of CPV designs leading to their present products. But more than technology development is needed for a new product to enter energy markets. Market incentives can be critical, especially for new technologies struggling to compete with deeply entrenched conventional energy technologies. The justification for society to provide market incentives can be the benefits of clean air, combating global climate change, providing local energy production and jobs, as well as avoiding the often-ignored problems of mining and waste removal associated
Slide 84: 68 Robert McConnell with large-scale, conventional energy sources. Further, today’s conventional energy sources have a long history of government incentives and support for justifiable reasons. As renewable technologies mature and energy needs increase, governments around the world are finding renewable energy market incentives both justifiable and effective in responding to society’s energy concerns. For PV systems, two main types of market incentives exist. Most government market support for PV in the United States is in the form of money refunded for the purchase and installation of a PV system. Therefore, many dollars per installed PV watt are returned to the customer, who, in turn, hands the money over to companies providing and installing the systems. These rebates were designed for companies and customers wanting to install small flat-plate PV systems for rooftops, which is the principal market for PV systems. Such rebates have been successful in developing PV markets for rooftops in Japan, as well as in the United States. Almost 20 states have some form of rebate for PV systems that can be combined with the new federal rebate approved by the U.S. Congress in 2005. However, there is an issue with most rebates in that they are paid at or soon after the PV installation, with little or no requirements that the system perform well in 2, 5, or even 20 years from the time of sale. Addressing such a situation, Germany developed an effective feed-in tariff program that pays, at a declining rate, for the energy produced over 20 years. The State of Washington and Spain recently initiated their own programs for feed-in tariffs and California is beginning to move in this direction. These programs express a commitment by the governments to honor energy purchase agreements for as long as 15 or 20 years. The U.S. rebate for PV systems is presently planned to be available for only 2 years. Feed-in tariffs can be very important market incentives, especially ones designed to reward investors for energy production, to reduce their risk in recovering their investment and to promote longterm system reliability. Such tariffs have been instrumental in the market success of wind energy systems, presently totaling about 10 times more electricity generating capacity than the world’s PV systems. In the case of CPV systems, feed-in tariffs open a market door for a technology that maximizes electricity production because CPV systems produce more kilowatt-hours (kWh) per kW than flat-plate PV systems. An attractive feed-in tariff provided the economic justification for the recent Amonix-Guascor CPV joint venture in Spain. As Fig. 3 shows, an advantage exists today for CPV systems using highefficiency solar cells in terms of energy produced for the same amount of capital invested in different PV systems.6 This is a very simple comparison between total project cost and annual energy produced for the different systems. It avoids the many assumptions required in other techno-economic analyses, such as the levelized cost of electricity. The comparison is made between a typical non-tracking flat-plate PV system, a single-axis-tracking flat-plate system, a CPV system using standard CPV silicon technology, and a CPV system with today’s new high-efficiency solar cells. A $1000 investment in a technology using today’s high-efficiency CPV cells could yield 450 kWh per year—almost 2-1/2 times more electricity than that generated by $1000 paid for fixed flat-plate PV systems. The increased bang for the buck is huge for investment in CPV technologies using new high-efficiency cells.
Slide 85: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 69 Fig. 3. CPV systems using new high-efficiency solar cells generate considerably more electricity for the same amount of money than do the alternatives.6 3 Higher and Higher Conversion Efficiencies With the advent of funding from the DOE in the late 1970s and early 1980s came plans and goals to develop PV technologies through improving performance (efficiency), reducing cost, and assuring reliability of operation. CPV systems offered the possibility of lower cost because expensive solar cells are replaced with less costly structural steel holding mirrors or lenses. However, early CPV systems showed the importance of optical efficiencies as optical losses typically reduced the CPV system efficiency by 15% to 20%. To compensate for optical losses, CPV systems needed the highest-quality, highest-performing solar cells to compete with flat-plate PV systems. Early PV researchers, principally Martin Green in Australia and Richard Swanson and Vahan Garboushian in the United States, developed innovative designs for crystalline silicon solar cells, leading to the record efficiencies of the 1980s and 1990s. Today’s CPV systems using high-efficiency crystalline silicon solar cells have system efficiencies approaching 20%. Installed CPV system costs are comparable today with those of utility-scale flat-plate PV systems at about $6/watt. 5 But the
Slide 86: 70 Robert McConnell Fig. 4. The highest-efficiency solar cells, both crystalline silicon and multijunction concentrator devices, have been most suitable for solar concentrator systems. Replace with graph having the new 40.7 % result in 2006. dramatically higher efficiency solar cells—above 40% now—as shown in Fig. 4, are creating considerable excitement about CPV systems.5 Research on multijunction solar cells began in the 1980s as part of a DOE effort to explore new solar cell materials and new solar conversion processes to improve cell efficiency. A single-junction solar cell is tuned to just one wavelength of the solar spectrum so that maximum efficiency occurs only at that color. (A semiconductor junction refers to an interface between a p-type semiconductor material and an ntype material. P and n refer to semiconductor charge carriers and are a reminder that solar cells behave like batteries in that they have positive terminals, negative terminals, and generate direct current.) Early solar cell researchers calculated that an infinite number of junctions would be the most effective means to harvest energy from each and every color in the solar spectrum and that such a stacked set of junctions could theoretically convert more than 80% of the sunlight into electricity. Yet, the first monolithic two-junction solar cell, made almost three decades after the discovery of the modern solar cell, demonstrated efficiencies less than that of a singlejunction cell. The materials and chemical science difficulties encountered in making the first monolithic two-junction solar cells were significant. These multijunction PV technologies are based on elements in columns III and V of the Periodic Table, and they are often referred to as III-V solar cells. Soon thereafter, two-junction III-V solar cells were developed with efficiencies higher than those of the best silicon solar cells.
Slide 87: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 71 Table 1. Benchmark (10-MW) System parameters and impact of multijunction (III-V) solar cell efficiency on a CPV utility reference system.7 High-efficiency solar cells are installed in essentially identical solar concentrator structures, and the cost per watt drops from about $6/watt to well under $2/watt while electricity costs fall below 10 cents per kWh. Higher production levels can lead to even lower levelized costs of energy (LCOE).7 System size Module price Cell efficiency Module size Module efficiency Installed system price LCOE MW $/Wdc % kWpdc % $/Wdc $/kWhac 10 4.13 26 (Si) 40 20 5.95 0.15–0.27 12.5 3 32 (III-V) 50 25 4.3 0.10–0.15 16 1.56 40 (III-V) 64 32 2.52 0.06–0.11 The U.S. Department of Defense recognized the potential of these new solar cells for powering satellites and supported the development of their manufacturing processes. Again, a new PV technology found a commercial niche in space power markets. Today, almost every commercial and defense satellite—as well as the Mars Rover instrumentation packages—use multijunction III-V solar cells for their electrical power sources. Just before the turn of the century, collaborative research and development by the National Renewable Energy Laboratory and Spectrolab, a division of Boeing, demonstrated a three-junction solar cell with a higher efficiency than that of two-junction cells. Efficiencies are now over 40% in the laboratory, with reasonable quantities of 35% cells available from suppliers. This technology may be coming back to earth more quickly than the early silicon cell technology did as today’s governments and investors respond to the world’s demands for more and cleaner energy sources. As shown in Table 1, performance pays.7 Capturing those economic benefits involves replacing the crystalline silicon solar cells in essentially identical solar concentrator structures with new high-efficiency III-V multijunction cells. The pioneering companies of Amonix and Solar Systems developed their CPV structures around crystalline silicon solar cells, but both are rapidly incorporating the new high-efficiency multijunction cells into CPV products that they expect to have available in the near future—within 2 to 5 years. Can the companies making multijunction III-V solar cells (e.g., Spectrolab and Emcore in the United States) meet this new and imminent market demand? Today’s annual manufacturing capacity for multijunction solar cells is about 1 MW under 1-sun illumination. Remember that these multijunction cells are used in non-concentrator versions in space. And as the market for satellites undergoes its own demand cycles, there are periods in which substantial portions of the production facilities are available for other markets, such as the terrestrial CPV market. Concentration provides a huge lever to this production capacity. A solar concentration ratio of 1000 suns means that a manufacturing capacity of 1 MW of flat-plate space PV panels could be the solar power sources for 1000 MW of CPV systems. The total production capacity throughout the world for III-V multijunction solar cells is already about 1 MW/year. The potential capacity there-
Slide 88: 72 Robert McConnell fore exists for CPV technology to make a dramatic leap from megawatts to gigawatts in the market in the very near future. However, companies want to be sure that these new multijunction solar cells will operate reliably in their CPV systems. After all, the new solar cells typically operate at higher voltages, generate higher current, and behave differently under environmental conditions of temperature cycles and humidity than do crystalline silicon solar cells. And there is a long history for crystalline silicon operation on earth, whereas very few multijunction III-V solar cells have been deployed in field installations. However, early demonstrations are promising. One CPV company, Concentrating Technologies, has operated Spectrolab’s triple-junction solar cells for more than one year at an Arizona Public Service test site. Nevertheless, companies integrating these new solar cells into their solar concentrator structures expect it will take 2 to 5 years to assure reliable products for the marketplace.5 4 CPV Reliability Today, flat-plate crystalline silicon technologies are renowned for their reliability in generating electricity for decades. What is often forgotten is that before flat-plate PV test standards were established in the early 1980s, large projects of flat-plate PV systems sometimes failed catastrophically. Standards organizations provide an important service for all technology development activities by providing a forum for companies, customers, and independent engineers to create a set of agreed-upon tests for identifying weaknesses in products before they go to market. Test standards, especially military standards, were critical to the success of space solar cells developed for defense satellites in the 1960s and 1970s. However, with the first efforts to bring space PV down to earth, project leaders discovered that the test standards for space solar cells were inadequate for terrestrial PV systems. Programs begun in the early 1980s at the Jet Propulsion Laboratory led to the successful development of qualification standards for crystalline silicon flat-plate PV technologies. Today, crystalline silicon solar cells are renowned for their long-term reliability, and few people are aware of the early disasters. Many early CPV systems suffered the same fate in that reliability was a serious issue. Professor Charles Backus, a CPV pioneer and mechanical engineer, noted that electrical or electronic engineers developing PV systems and PV standards were unaccustomed to solving mechanical engineering problems or developing standards for large mechanical structures.2 In the late 1980s, Sandia National Laboratories developed a set of stress tests (accelerated environmental testing) for CPV systems based on their early CPV field tests funded by the DOE. This work served as the basis for the first CPV qualification standard developed in the late 1990s and finally published by the International Electrical and Electronics Engineers (IEEE) standards organization in 2001. This first IEEE standard was most suitable for CPV systems using Fresnel lenses, typical of many U.S. CPV designs. The International Electrotechnical Commission (IEC), based in Geneva, Switzerland, is voting on the first international CPV draft standard suitable for testing all of the CPV geometries and technologies. The IEC standard builds on the concept of
Slide 89: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 73 testing representative sample assemblies for key elements of different designs under conditions of high temperature, low temperature, temperature cycling, humidity, electrical performance under wet and humid conditions, and outdoor performance. This new IEC standard is expected to play an important role as companies work rapidly to integrate III-V multijunction solar cells into their solar concentrator structures.8 5 Following in Wind Energy’s Footsteps Wind energy is one renewable energy technology developed successfully by mechanical engineers. In the 1970s, wind systems and PV systems started out on nearly the same footing: only a few experimental systems for each were installed around the world. Today, there are roughly 10 times more wind energy systems than PV systems installed—50,000 MW of wind systems versus 5,000 MW of PV systems. Why were wind energy technologies able to surpass PV systems? One reason was that wind developers were able to quickly demonstrate economies of production just as a market opportunity appeared. The state of California offered long-term standardoffer contracts from 1985 to 1989 to purchase the electricity over 20 years from large-scale renewable energy projects. These long-term contracts were similar in many ways to the successful European feed-in tariffs. Solar concentrators producing heat to drive electric generators—called concentrating solar power (CSP) systems— also took advantage of the California opportunity; almost 400 MW of CSP systems were installed in the 1980s, and they have been generating solar electricity ever since then. Fabrication facilities are relatively inexpensive for both wind and CSP systems when compared with PV manufacturing facilities. Wind production facilities resemble automobile assembly lines.9 PV production facilities, although not as complex or costly as those of the integrated-circuit industry processing semiconductor silicon, still cost roughly 10 times more than wind production facilities. For early investors, this is an important issue. Consider the investment choice. Crystalline silicon and amorphous thin-film flatplate PV production facilities both cost about $100 million or more for 100 MW per year manufacturing plants.10,11 A wind production facility of the same size might cost $10 million for the same annual production. Investment in production facilities for new technologies entails significant risk, and the lower risk for investing in wind facilities was one reason investors provided funds for large 1,000 MW wind projects in the 1980s. PV was not able to demonstrate the economies of production quickly enough to take advantage of the small window of opportunity provided by California’s standard-offer contracts. In the 1980s, the very-high-efficiency solar cells needed by CPV systems were still in the research laboratory. The California market incentives helped wind and CSP developers and investors move their technologies forward, reducing cost and acquiring valuable operational experience that improved reliability. Wind engineers developed their qualifications standards during this same period; like early PV technologies, wind systems often suffered from poor reliability until their certification standards were established and required in the marketplace.
Slide 90: 74 Robert McConnell Fig. 5. An Amonix production facility in Los Angeles is strikingly different from a flat-plate PV manufacturing facility. This difference results in much lower capital costs for the facility. Some noteworthy similarities exist between wind energy systems and CPV systems.9 They both employ relatively common materials, particularly steel. Wind system costs are typically less than $1 per watt; they depend mainly on the cost of steel, whereas flat-plate PV is linked to the availability and cost of expensive semiconductor silicon. But solar concentrator structures are also amenable to an auto-assembly type of production (see Fig. 5), and CPV developers estimate CPV production facility costs are much closer to those of wind systems than to those of flat-plate PV production facilities. In early EPRI cost studies, CPV production facility costs were estimated (on the same costing basis as the crystalline and amorphous silicon facilities) to be about $28 million for a 100 MW per year installation—about one-quarter the cost of the conventional silicon PV facilities.10 These lower investment costs can lead to a faster scale-up of manufacturing facilities because investor risk is relatively smaller than the risks entailed in investing in conventional PV production facilities. Further, cost studies in Spain and Israel estimate CPV installed system costs will, like wind systems costs, finish below $1 per watt when gigawatt levels of CPV production are reached.7,12 Both CPV and wind energy technologies are modular, like flat-plate PV modules, but the sizes are different. Wind units are now megawatts in size whereas CPV units range from kilowatts to tens of kilowatts. Flat-plate PV
Slide 91: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 75 modules are usually less than 100 watts. And, obviously, both wind and CPV systems have moving parts, yet moving parts have not limited the success of wind systems. Finally, wind systems first penetrated the energy marketplace in sites with very high and steady winds, whereas CPV systems will almost certainly enter markets in locations with considerable sunlight and almost no clouds, similar to the climates of the southwestern United States, Spain, Australia, and Israel. 6 Low-Cost Hydrogen from Hybrid CPV Systems We are now in a position to understand why this new high-efficiency solar electric technology provokes a fresh look at the challenge of generating hydrogen from water using sunlight. In addition to generating solar electricity at low cost, CPV systems have the potential to produce hydrogen through an electrolysis process. The generation of electrolytic hydrogen from solar energy is critically important to the world’s long-term energy needs for several reasons. The feedstock (water) and supplied energy (solar) are inherently carbon free so that on a life cycle basis the total carbon emissions will be significantly less than from fossil-based options for generating hydrogen. And there is the potential to generate hydrogen near its markets, thus minimizing transportation costs. In the past, the principal criticism of photovoltaics for generating hydrogen has been the high cost of PV electricity and the inefficiencies of the conversion processes, particularly the PV process. As we have seen, CPV systems have the potential for generating lower-cost electricity, primarily due to developing high-efficiency multijunction III-V solar cells with efficiencies above 40%. But it is the heat boost from CPV systems that can dramatically improve and enhance the electrolysis efficiency of water in a hightemperature solid-oxide electrolyzer. This heat boost—40% was measured in the 1990s by the company Solar Systems in Australia above 1100oC13,14—has been substantiated in recent theoretical analyses.15 This new pathway provides significant engineering and economic benefits for generating electrolytic hydrogen from solar energy, thereby creating opportunities for PV to contribute to future transportation markets directly with low-cost hydrogen or by producing liquid hydrogen-carrier fuels such as methanol.16 Solar-to-hydrogen conversion efficiencies of 40%, including optical losses, are attainable in the near-term (within the next few years) using high-efficiency III-V multijunction solar cells, whereas efficiencies of 50% and higher are realistic targets within 5 to 10 years. These efficiencies are dramatically higher, by roughly a factor of 3 or 4, than those of any of the other methods previously considered for generating electrolytic hydrogen from solar electricity.16 These results, based on the longterm potential for CPV systems to be mass produced at costs of less than $1/W, lead to hydrogen production costs comparable with the energy costs of gasoline— recognizing that 1 kg of hydrogen has the energy equivalent of one U.S. gallon of gasoline.17,18
Slide 92: 76 Robert McConnell Fig. 6. Schematic of system shows sunlight reflected and focused on the receiver, with reflected infrared directed to a fiber-optics light pipe for transport to a high-temperature solid-oxide electrolysis cell. Solar electricity is sent to the same electrolysis cell, which is able to use both heat and electricity to split water. 7 Describing the Hybrid CPV System This approach first proposed by Solar Systems in Australia employs a dish concentrator that reflects sunlight onto a focal point (see Fig. 6). At the focal point is a spectral splitter (heat mirror) that reflects infrared solar radiation and transmits the visible sunlight to high-efficiency solar cells behind the spectral splitter. Figure 7 schematically shows the transmission across the solar spectrum wavelengths. The reflected infrared radiation is gathered by a fiber-optics “light pipe” and conducted to the high-temperature solid-oxide electrolysis cell. The electrical output of the solar cells also powers the electrolysis cells. About 120 megajoules are needed—whether in electrical or thermal form, or both—to electrolyze water and generate 1 kg of hydrogen. The result is that more of the solar energy is used for
Slide 93: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 77 Fig. 7. Transmittance (and reflectance) of a spectral splitter mirror as a function of solar wavelength in microns. This response depicts that of a “hot mirror” in which light is transmitted in the visible region and reflected in the infrared. hydrogen production. And we shall see that the additional costs for the hybrid solar concentrator components—the spectral splitter and fiber-optics light pipe—are relatively small compared with the boost in hydrogen production. The testing of components shown in Fig. 6 occurred in the mid-1990s and has been described previously13,14,17 on a scale considerably smaller than that of Fig. 2. The solar concentrator was a paraboloidal dish 1.5 m in diameter, with two-axis tracking, and is capable of more than 1000-suns concentration (Fig. 8). The full dish was not needed and most of it was shaded appropriately for use with the small electrolysis cell. At that time, the solar cell was a GaAs cell with an output voltage of 1 to 1.1 V at maximum power point, with a measured efficiency of about 19%. The voltage was an excellent match for direct connection to the electrolysis cell when operating at 1000 oC. The tubular solid-oxide electrolysis cell was fabricated from yttria-stabilized zirconia; the cell had platinum electrodes because the test temperature was higher than that of typical solid oxide cells. Figure 9 shows a schematic of the solid-oxide electrolysis cell operation. A metal tube surrounded the cell to uniformly distribute the solar flux over the cell’s surface. The test occurred during a 2-hour period of operation, with an excess of steam applied to the electrolysis cell. The output stream of unreacted steam and generated hydrogen was bubbled through water and the hydrogen was collected
Slide 94: 78 Robert McConnell Fig. 8. This photo, taken in the 1990s, shows John Lasich demonstrating how cool his concept is for conducting infrared energy through a fiber-optics light pipe. The dish reflects sunlight to a “heat mirror” that reflects long-wavelength solar radiation to the fiber-optic bundle along the axis of the parabolic dish. The visible light seen at the end of the light pipe in Lasich’s hand is a result of partial reflection of visible light by the heat mirror. and measured. During a definitive 17 minutes of system operation in steady state, 80 mL of hydrogen were collected. The ratio of the thermoneutral voltage of 1.47 V to the measured electrolysis cell voltage of 1.03 V was 1.43, corresponding to a boost of more than 40% in hydrogen production due to the input of thermal energy. This was also confirmed by energy balance. Combining the optical efficiencies of the concentrator dish (85%), solar cell efficiency, and thermal-energy boost, the total
Slide 95: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 79 Fig. 9. Schematic of high-temperature electrolysis in a solid-oxide cell. The geometry can be planar or tubular as in the case of the first demonstration of the hybrid solar concentrator PV system. Operating the electrolysis cell in reverse corresponds to electricity and heat production in solid-oxide fuel cell operation. system efficiency was 22% for conversion of solar energy to hydrogen. At the time of these measurements in the mid-1990s, the efficiency was almost three times better than that recorded for any other technology converting solar energy to hydrogen. These early tests were not conducted with the most efficient solar cells available at that time. The record efficiency then was about 30% for a laboratory cell (see Fig. 4) and those cells were not easily obtainable. Today’s record efficiency is 40.7%, and 35%-efficient cells are commercially available.18 Therefore, 40% solar-tohydrogen efficiency is expected in the near term assuming a heat boost of 40%, a multijunction solar cell efficiency of 35%, and an optical efficiency of 85%. A 40% multijunction solar cell would yield a solar-to-hydrogen conversion efficiency of almost 50%. Nevertheless, electrochemical theoretical results calculated by Licht, shown in Figure 10, are consistent with these predictions based on Solar Systems’ early experiments.15 Two cost analyses have been reported for this concept.17,19 Although the resulting hydrogen costs agreed within their costing uncertainties, the hydrogen generation plants were quite different in nature, as were the financial assumptions in their cost analyses. The first analysis was conducted in 2004 and reported in 2005, and it used a set of financial and plant assumptions developed by the DOE Hydrogen Program.19,20 Because of the complexity of the DOE H2A plant assumptions, a back of the envelope calculation with simplified financial assumptions was made to highlight key cost elements.17 This analysis is presented below and compared with the H2A analysis. The principal difference between the two analyses is the additional costs of operation, transportation, storage and distribution in the H2A analysis. The largest cost for the hybrid solar concentrator system will be for the dish concentrator and PV receiver, shown in Fig. 6. Algora recently completed an extensive cost analysis based on previously collected data for CPV systems.7 Many of the
Slide 96: 80 Robert McConnell Fig. 10. Energy conversion efficiency of solar-driven water splitting to generate H2 as a function of temperature and photovoltaic conversion efficiency at AM1.5 insolation, at pH2O = 1 bar. Reprinted with permission from J. Phys. Chem. B 2003, 107, 4253-4260. Copyright 2003 American Chemical Society.” project costs came from installed costs for the 480-kW reflective CPV system in Tenerife, Spain. The analysis included a wide range of parameters, including cumulative production of 10 MW for present-day systems to cumulative production of 1000 MW for the mid-term systems where learning cost reductions are incorporated. Concentrations ranged from 400 to 1000 suns, with solar cell efficiencies ranging from 32% to 40%. Module efficiencies ranged from 24.8% to 32.2%, and the plant’s AC annual efficiency ranged conservatively from 18.2% to 23.6%. Present-day base costs were 2.34 euro/W (almost $3/W with today’s exchange rate). The lowest projected system costs ranged from 0.5 to 1 euro/W for efficiencies of 40%, 1000-suns concentration, and cumulative production of 1000 MW. We wanted to compare the results of our simplified engineering cost analysis of hydrogen generated by this hybrid solar concentrator system with those of more extensive cost analyses: 1. the electrolytic generation of hydrogen by wind systems, where cumulative production of this highly developed technology is approximately 50 gigawatts (GW); and,
Slide 97: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 81 Table 2. Component and system costs for 10-MW hybrid CPV project for solid-oxide electrolytic production of hydrogen. Component costs assuming 1000-MW technology ($/kW) 800 15 25 400 1240 Concentrator PV Spectral splitter Fiber optics Electrolysis cell Total System Cost 2. the conventional production of hydrogen by reforming natural gas. So, for our analysis, we used cost estimates for mature CPV technology. Cost studies for conceptual high-temperature nuclear reactors (projected for mature 600-MW designs) suitable for high-temperature electrolysis cells face similar problems because both the hybrid solar concentrator and high-temperature nuclear reactor are in early stages of exploratory research and development for hydrogen generation. Further, high-temperature solid-oxide electrolysis cells will be required in large sizes (500 kW to 500 MW) for integration with nuclear reactors.21 Unit sizes ranging from 20 to 50 kW could be used with hybrid solar concentrators. Although solid-oxide fuel cells are commercially available, solid-oxide electrolysis cells are beginning development. It is important to note that solid-oxide electrolysis cells have been demonstrated to date in small sizes equivalent to hundreds of watts. The modular character and size of the hybrid CPV system is commensurate with the development of solid-oxide electrolysis cell technology, also in early stages of development. Using a set of assumptions for a well-developed technology, we acquired costs in $/kW for solid-oxide electrolysis cells from a developer of solid-oxide electrolysis cells.22 Table 2 summarizes the cost data for a well-developed technology (1000-MW cumulative production) for the hybrid CPV system and high-temperature solid-oxide electrolysis cell. Table 3 summarizes the hydrogen production costs for a 10-MW project built with the well-developed technology assuming a 20% rate of return per year not including operating, storage, transmission or distribution costs.19 It also contain the estimated costs from the H2A analysis that includes the additional operating, storage, transmission and distribution costs expected for a distant, centralized hydrogen generation plant.17 Table 4 compares these production costs with those of other hydrogen production technologies. 8 Discussion The literature contains many cost analyses for hydrogen production, but the assumptions behind the analyses vary dramatically. The DOE, through its Hydrogen Program, is establishing a cost-analysis structure for comparing different hydrogen and
Slide 98: 82 Robert McConnell Table 3. Hydrogen production data for mature 10-MW plant.17 The estimated hydrogen cost of $2.48/kg has considerable uncertainty related to technology immaturity and simplistic assumptions. The H2A costs include operating, storage, transmission and distribution costs.19 Hydrogen cost data for mature technology 10 12.4 106 2.48 3.18 Plant size (MW) Plant cost ($ million) H2 produced (kg/yr) Hydrogen cost ($/kg)17 H2A Hydrogen cost ($/kg)19 fuel cell technologies within a common set of assumptions. The analysis in Table 3 is a preliminary study needing additional work to fit within that framework. CPV systems are just beginning to enter the energy market, so cost uncertainties are significant compared with those of highly developed wind systems with a worldwide installed capacity approaching 50 GW. Nevertheless, these preliminary hydrogen costs are comparable with hydrogen costs from wind electrolysis, so additional cost studies are warranted. Today, wind system costs are in the $800/kW range—as are the estimated costs for highly developed CPV systems—whereas wind electrolysis does not have an opportunity for a heating boost in electrolysis efficiency. Assuming these cost analyses continue to be positive, it will be worth demonstrating this hybrid solar concentrator technology on a larger scale. The uncertainties in this cost analysis arise principally from the early stage of technology development for solar concentrators, high-efficiency solar cells, and solid-oxide electrolysis cells. There are many positive indications that these technologies can progress and achieve their performance and cost potentials, but additional work will be needed. 9 Hydrogen Vision Using Hybrid Solar Concentrators The U.S. National Research Council and National Academy of Engineering believes that one of the four most fundamental technological and economic challenges for the hydrogen economy is: “To reduce sharply the costs of hydrogen production from renewable energy sources over a time frame of decades”23 Table 4. Cost comparison for the hybrid CPV production of electrolytic hydrogen. Note that 1 kg of hydrogen has the energy equivalent of one U.S. gallon of gasoline. Process Gas reformation20 Wind electrolysis20 Hybrid CPV electrolysis17 (approximating distributed generation) Hybrid CPV electrolysis19 (assuming centralized generation of H2A analysis) Hydrogen Production Cost ($/kg) 1.15 3.10 2.48 3.18
Slide 99: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 83 Wind electrolysis is a strong renewable energy option, and this study indicates hybrid CPV electrolysis could be another. Also, the solar energy resource is considered larger and more widely distributed than that of wind energy. And totally new system configurations may be possible with hybrid solar concentrator electrolysis. Small 50kW systems could be part of hydrogen filling stations, reducing hydrogen distribution costs. Systems could incorporate backup-heating sources, probably natural gas in the near term, to improve the electrolysis system capacity factor.22 Electricity providers throughout the world are considering large-scale CPV projects, some in the range of 100s of MW. The hybrid CPV system could generate both electricity and hydrogen for future electric utilities. With low-cost tank storage on utility land, the solid-oxide electrolysis cell could be designed to operate in a regenerative mode, producing electricity from hydrogen during non-solar periods. This design would greatly increase the value of solar electricity to utilities. Probably the most dramatic impact of this study has been the realization that the hybrid CPV system is a PV option that could provide transportation fuel on a large scale. In a scenario where hydrogen is used in fuel cell vehicles—which can have double the efficiency of standard internal combustion cars—the effective cost of solar hydrogen would be half, i.e., $1.24/kg. For customers paying $3 per U.S. gallon for gasoline, which is more than twice the effective hydrogen cost, the potential for a very large market clearly exists. To determine the final price of solar hydrogen to the customer, we would need to factor in the additional costs of operation, distribution, retailing, and taxes, as well as consider the society cost benefits due to the clean and renewable value of solar hydrogen. The International Energy Agency’s Photovoltaic Power Systems Program recently published a study on the feasibility of very large-scale PV systems.24 Entitled Energy from the Desert, the study explores the concept of using the world’s deserts to provide electricity at terawatt (1012 watt) levels of production. While it is evident that CPV systems could be a major contributor to such large projects, the hybrid solar concentrator opens the possibility for a new vision, one where the world’s oceans could be harvested for hydrogen. Such a vision would again follow in the footsteps of wind technology as wind projects continue to appear off of the world’s coastlines. The hybrid solar concentrator is a potential leap frog technology that may rapidly lower the cost of clean hydrogen in light of the following: the imminent market entry of CPV systems for electricity production; solar cell efficiencies above 40%, with clearer ideas for 50%-efficient solar cells; and the opportunity to use wasted solar heat for augmenting solar electrolysis. 10 Conclusions An innovative hybrid CPV electrolysis technology has been described that offers a potential cost of hydrogen lower than that from wind electrolysis and in the same range as gasoline for much of the world’s population. The analysis is preliminary, but additional cost analysis and technology demonstrations are warranted.
Slide 100: 84 Robert McConnell This study has described several reasons supporting the argument that this is a technology on the horizon that could be significant throughout the world. The reasons include the following: • • • • • The existing world’s annual production capacity for manufacturing highefficiency III-V multijunction solar cells is enough today for 1000 MW annually of CPV systems. This level provides a good jumpstart into the market. Society’s recent concerns about energy security, global climate change, clean air, and high-technology economic opportunities are leading to the appearance of market openings based on feed-in tariffs in regions with solar resources suitable for CPV systems. The CPV community has completed the first CPV qualification standards in time to respond to market opportunities, although additional safety, performance, and tracker standards are still needed. The early success of wind energy technologies, similar in several respects to CPV technologies, augurs rapid and dramatic success for CPV systems. The possibility of efficiently producing hydrogen by splitting water with low-cost solar electricity opens a new pathway for production of transportation fuels. This innovative renewable energy technology could leap frog other renewable energy technologies for electrolytic production of hydrogen—a potentially important transportation fuel for our future. Acknowledgements The author acknowledges the valuable work of many CPV pioneers who were responsible for the research progress described in this Chapter. At the risk of overlooking others, the author has particularly benefited from numerous articles by and discussions with Charles Backus, Andreas Bett, Vahan Garboushian, Martin Green, Richard King, Sarah Kurtz, John Lasich, Antonio Luque, Jerry Olson, Gabriel Sala, Richard Schwartz, Richard Swanson, and Masafumi Yamaguchi. Tomorrow’s CPV companies will be building on the successes of these early leaders in developing high-efficiency solar cells and solar CPV technologies. In the field of hydrogen and electrolysis, the author acknowledges valuable discussions with John Turner, Krishnan Rajeshwar, Joe Hartvigsen, S. Srinivansan, and again, John Lasich, the originator of the hybrid solar CPV concept for electrolytic hydrogen production. References 1. http://web.mit.edu/2.009/www/experiments/deathray/10_ArchimedesResult.html, October 2005. 2. C. E. Backus, A Historical Perspective on Concentrator Photovoltaics, Proceedings of the International Solar Concentrator Conference for the Generation of Electricity or Hydrogen, Alice Springs, Australia, November 2003.
Slide 101: A Solar Concentrator Pathway to Low-Cost Electrolytic Hydrogen 85 3. R. M. Swanson, Straight Talk About Concentrators, Future Generation Photovoltaic Technologies: First NREL Conference, Denver, Colorado, American Institute of Physics Conference Proceedings #404, October 1997, p. 277-284. 4. R. M. Swanson, The promise of concentratiors, Prog. Photovolt. Res. Appl. 8, John Wiley and Sons, Ltd., Hoboken, NJ (2000) pp. 93-111. 5. R. McConnell, S. Kurtz, and M. Symko-Davies, Concentrating PV technologies: Review and market prospects, ReFOCUS, Elsevier Ltd, , July/August 2005, p. 35. 6. NREL Frequently Asked Questions (FAQs): What's New in Concentrating PV?, Report No. FS-520-36542; DOE/GO-102005-2027, February 2005. 7. C. Algora, Next Generation Photovoltaics, Chapter 6, Ed. by A. Marti and A. Luque, Institute of Physics Publishing, Bristol and Philadelphia, 2004. 8. L. Ji and R. McConnell, New Qualification Test Procedures for Concentrator Photovoltaic Modules and Assemblies, Proceedings of the 2006 IEEE 4th World Conference on Photovoltaic Energy Conversion, Waikoloa, Hawaii, May 2006. 9. R. McConnell, Large-Scale Deployment of Concentrating PV: Important Manufacturing and Reliability Issues, Proceedings of the First International Conference on Solar Electric Concentrators, New Orleans, Louisiana, NREL/EL-590-32461, May 2002. 10. R. Whisnant, S. Wright, P. Champagne, and K. Brookshire, Photovoltaic Manufacturing Cost Analysis: A Required-Price Approach, Vols. 1 and 2, EPRI AP-4369, Electric Power Research Institute, Palo Alto, CA, 1986. 11. R. Whisnant, S. Johnston, and J. Hutchby, Economic analysis and environmental aspects of photovoltaic systems, Ch. 21, Handbook of Photovoltaic Science and Engineering, Ed. by A. Luque and S. Hegedus, John Wiley and Sons, Ltd., 2003. 12. D. Faiman, D. Raviv, and R. Rosenstreich, The Triple Sustainability of CPV with the Framework of the Raviv Model, Proceedings of the 20th European Photovoltaic Solar Energy Conference and Exhibition, Barcelona, Spain, June 2005. 13. J. Lasich, U.S. Patent No. 5658448, August 19, 1997. 14. J. Lasich, U.S. Patent No. 5973825, October 26, 1999. 15. S. Licht, J. Phys. Chem. B 107, 4253–4260, 2003 (also see Chapter 5 in this book). 16. N.Lewis, http://www7.nationalacademies.org/bpa/SSSC_Presentations_Oct05_Lewis.pdf, August 2006. 17. R. D. McConnell, J.B. Lasich, and C. Elam, A Hybrid Solar Concentrator PV System for the Electrolytic Production of Hydrogen, Proceedings of the 20th European Photovoltaic Solar Energy Conference and Exhibition, Barcelona, Spain, June 2005. 18. R. McConnell, M. Symko-Davies, and D. Friedman, Multijunction Photovoltaic Technologies for High Performance Concentrators, Proceedings of the 2006 IEEE 4th World Conference on Photovoltaic Energy Conversion, Waikoloa, Hawaii, May 2006. 19. J. Thompson, R. McConnell, and M. Mosleh, Cost Analysis of a Concentrator Photovoltaic Hydrogen Production System, NREL/CD-520-38172, Proceedings of the International Conference on Solar Concentrators for the Generation of Electricity or Hydrogen, Scottsdale, Arizona, May 2005. 20. D. Mears, M. Mann, J. Ivy, and M. Rutkowski, Overview of Central H2A Results, U.S. Hydrogen Conference Proceedings, April 2004. 21. R. Anderson, S. Herring, J. O’Brien, C. Stoots, P. Lessing, J. Hartvigsen, and S. Elangovan, Proceedings of the National Hydrogen Association Conference, 2004. 22. J. Hartvigsen, private communication; also see www.ceramatec.com 2006. 23. National Research Council and National Academy of Engineering, in The Hydrogen Economy, National Academies Press, Washington DC, 2004.
Slide 102: 86 Robert McConnell 24. Photovoltaic Power Systems Executive Committee of the International Energy Agency, Energy from the Desert: Feasibility of Very Large Scale Photovoltaic Power Generation Systems, Ed. by K. Kurokawa, James and James, London 2003; also see http://www.ieapvps.org/products/rep8_01s.htm.
Slide 103: 5 Thermochemical and Thermal/Photo Hybrid Solar Water Splitting Stuart Licht University of Massachusetts, Boston, MA 1 Introduction to Solar Thermal Formation of Hydrogen 1.1 Comparison of Solar Electrochemical, Thermal & Hybrid Water Splitting Solar electrochemical, solar thermal,1,2 and solar thermal/electrochemical hybrid3 hydrogen generation are introduced in this Section. Water electrolysis and electrolysis using solar concentrator technology were discussed in Chapters 3 and 4. The thermal and the hybrid processes will be discussed in depth in subsequent Sections of this chapter. At high temperatures (> 2000 °C), water chemically disproportionates to hydrogen and oxygen. Hence, in principle, by using solar energy to directly heat water to very high temperatures, hydrogen and oxygen gases can be generated. This is the basis for all direct thermochemical solar water splitting processes.1 However, catalysis, gas recombination, and containment materials limitations above 2000 °C have led to very low solar efficiencies for direct solar thermal hydrogen generation. In another approach, the utilization of a multi-step, indirect, solar thermal reaction processes to generate hydrogen at lower temperatures has been extensively studied, and a variety of pertinent reaction processes considered.2 These reactions are conducted in a cycle to regenerate and reuse the original reactions, ideally, with the only net reactant water, and the only net products hydrogen and oxygen. However, such cycles suffer from challenges often encountered in multi-step reactions.4 While these cycles can operate at lower temperatures than the direct thermal chemical generation of hydrogen, efficiency loses can occur at each of the steps in the multi-step sequence, resulting in low overall solar to hydrogen energy conversion efficiencies. Electrochemical water splitting, generating H2 and O2 at separate electrodes, largely circumvents the gas recombination and high temperature limitations occurring in thermal hydrogen processes. Thus a hybrid of thermal dissociation and elec-
Slide 104: 88 Stuart Licht trolysis provides a pathway for efficient solar energy utilization. The hybrid method expands on existing solar electrochemical processes, which are therefore discussed briefly here. There has been significant, ongoing experimental5-15 and theoretical.5,16,17 interest in utilizing solar generated electrical charge to drive electrochemical water splitting (electrolysis) to generate hydrogen. In each of the above referenced studies, water electrolysis occurs at, or near, room temperature. Photoelectrochemical models predict a maximum ~30% solar water splitting conversion efficiency by eliminating • • • • the linkage of photo to electrolysis surface area, non-ideal matching of photo and electrolysis potentials, and incorporating the effectiveness of contemporary electrolysis catalysts, and efficient multiple bandgap photoabsorbers (semiconductors).18 However, these models did not incorporate solar heat effects on the electrolysis energetics as elaborated below. The UV and visible energy rich portion of the solar spectrum is transmitted through water (Chapters 1 and 2). Therefore a mediator for light absorption, such as a semiconductor, is required to drive the electrical charge for the water-splitting process. The PV (photovoltaic) process refers to a solar panel connected ex situ to electrochemically drive water splitting, e.g., an illuminated semiconductor-based photovoltaic device wired to an electrolyzer. On the other hand, the photoelectrochemical process refers to in situ immersion of the illuminated semiconductor in a chemical solution (electrolyte) to electrochemically drive water splitting, as described in Chapter 7. The significant fundamental components of PV and photoelectrochemical hydrogen generation are identical, but from a pragmatic viewpoint the PV process seems preferred, as it isolates the semiconductor from contact with and corrosion in the electrolyte. The UV and visible energy rich portion of the solar spectrum is transmitted through H2O. Semiconductors, such as TiO2, can split water, but their wide bandgap limits the photoresponse to a small fraction of the incident solar energy. Solar photoelectrochemical attempts to split water have utilized TiO2,20 InP,21 and also multiple bandgap semiconductors.19,22,23 Photoelectrochemical water splitting studies have generally focused on diminishing the high bandgap apparently required for solar water splitting, by tuning (decreasing) the bandgap of the semiconductor, Eg, to better match the water splitting potential, EH2O. Multiples of electrolyzers and photovoltaics can be combined to produce an efficient match of the generated and consumed power, as shown in Fig. 1. Also multiple bandgap semiconductors can be combined to generate a single photovoltage well-matched to the electrolysis cell, and over 18% conversion energy efficiency of solar to hydrogen was demonstrated, albeit at room temperature (still without the potential benefits of hybrid thermal hydrogen generation.)23 Unlike room temperature solar PV and photoelectrochemical electrolysis, the hybrid approach utilizes energy of the full solar spectrum, leading to substantially higher solar energy efficiencies. The IR radiation is energetically insufficient to drive conventional solar cells, and this solar radiation is normally discarded (by reflectance or as re-radiated heat.) On the other hand, in the hybrid approach, as seen in Fig. 2
Slide 105: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 89 Fig. 1. Alternate configurations varying the number of photo harvesting units and electrolysis units for solar water splitting.3 The photoconverter in the first system generates the requisite water electrolysis voltage and in the second system generates twice that voltage, while the photoconverter in the third and fourth units generate respectively only half or a third this voltage.
Slide 106: 90 Stuart Licht Fig. 2. Schematic representations of solar water electrolysis improvement through excess solar heat utilization.3 and as described in a latter Section of this chapter, the IR wavelengths are not discarded, but instead utilized to heat water. This in turn substantially decreases the necessary electrochemical potential to split the water, and substantially increases the solar hydrogen energy conversion efficiencies. 2 Direct Solar Thermal Water Splitting to Generate Hydrogen Fuel 2.1 Development of Direct Solar Thermal Hydrogen The direct thermochemical process, to generate hydrogen by splitting water involves heating water to a high temperature and separating the hydrogen from the equilibrium mixture. Although conceptually simple, the single-step thermal dissociation of water has been impeded by the need for a high-temperature heat source to achieve a reasonable degree of dissociation, and––to avoid ending up with an explosive mixture––by the need for an effective technique to separate H2 and O2. Unfortunately, the decomposition of water does not proceed substantially until the temperature is very high. Generally temperatures of 2500 K have been considered necessary for direct thermal water splitting. The Gibbs function (ΔG, or free energy) of the gas reaction H2O ↔ H2 + ½ O2, does not become zero until the temperature is increased
Slide 107: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting Table 1. The pressure equilibrium constants of the water dissociation reaction.24 Temperature (K) 3000 8.56x10-3 1.57x10-2 3.79x101 7.68x101 91 K1 K2 K3 K4 2500 1.34x10-4 4.22x10-4 1.52x103 4.72x103 3500 1.68x10-1 2.10x10-1 2.67x100 4.01x100 to 4310 K at 1 bar pressure of H2O, H2 and O2.3 At a water pressure of 0.1 bar or greater, significant mole fractions of hydrogen are not spontaneously formed at temperatures below 2200 K. The entropy (ΔS), driving the negative of the temperature derivative of the Gibbs function change, is simply too small to make direct decomposition feasible at this time.4 2.2 Theory of Direct Solar Thermal Hydrogen Generation In the high temperature gas phase equilibrium of water, in addition to H2O, H2 and O2, the atomic components H and O must be considered. These components are relatively insignificant at temperatures below 2500 K, as the pressure equilibrium constants for either diatomic hydrogen or oxygen formation from their atoms are each greater than 103 at T < 2500 K. However, the atomic components become increasingly significant at higher temperatures. The pressure equilibrium constants of the water dissociation reaction are summarized in Table 1 for water splitting at temperatures at which significant, spontaneous formation of H2 occurs: H2O ↔ HO + H HO ↔ H + O 2 H ↔ H2 2 O ↔ O2 K1 K2 K3 K4 (1) (2) (3) (4) Kogan has calculated that, at a pressure of 0.05 bar, water dissociation is barely discernible at 2000 K.1 By increasing the temperature to 2500 K, 25% of water vapor dissociates at the same pressure. A further increase in temperature to 2800 K under constant pressure causes 55% of the vapor to dissociate.1 These basic facts indicate the difficulties that must be overcome in the development of a practical hydrogen production by solar thermal water splitting: (a) attainment of very high solar reactor temperatures, (b) solution of the materials problems connected with the construction of a reactor that can contain the water spitting products at the reaction temperature, and (c) development of an effective method for in situ separation of hydrogen from the mixture of water splitting products.
Slide 108: 92 Stuart Licht 2.3 Direct Solar Thermal Hydrogen Processes The problems with materials and separations at such a high temperature make direct decomposition not attractive at this time. The production of hydrogen by direct thermal splitting of water generated a considerable amount of research during the period 1975–1985. Fletcher and co-workers stressed the thermodynamic advantages of a one-step process with heat input at as high a temperature as possible.25-27 The theoretical and practical aspects were examined by Olalde,28 Lede,29-31 Ounalli,32 Bilgen33,34 and by Ihara.35,36 However, no adequate solution to the crucial problem of separation of the products of water splitting has been worked out so far. Effort was spent to demonstrate the possibility of product separation at low temperature after quenching the hot gas mixtures by heat exchange cooling,8 by immersion of the irradiated, heated target in a reactor of water liquid,9,10 by rapid turbulent gas jets,29,30 or rapid quench by injecting a cold gas.31 Based on a theoretical evaluation, Lapique18 concluded that by quenching under optimal conditions it should be possible to recover up to 90% of the hydrogen formed by thermal water splitting. However, the quench introduces a significant drop in the efficiency and produces an explosive gas mixture.2 To attain efficient collection of solar radiation in a solar reactor operating at the requisite 2500 K, it is necessary to reach a radiation concentration of the order of 10000 suns. This is a rather stringent requirement. By way of example, a 3-MW solar tower facility, consisting of a field of 64 slightly curved heliostats, has each heliostat capable of concentrating solar radiation approximately by a factor of 50. Even by directing all the heliostats to reflect the sun rays towards a common target, a concentration ratio of only 3000 may be obtained.37 It is possible, however, to enhance the concentration ratio of an individual heliostat by the use of a secondary concentration optical system, and such systems have been explored.37,38 Ordinary steels cannot resist temperatures above a few hundred degrees centigrade, while the various stainless steels, including the more exotic ones, fail at less than 1300 K. In the range 3000–1800 C alumina, mullite or fused silica may be used. A temperature range of about 2500 K requires use of special materials for the solar reactor. However, higher melting point materials can have additional challenges; carbide or nitride composites are likely to react with water splitting products at the high temperatures needed for the reaction. A list of candidate materials of high temperature oxide, carbide and nitride ceramics, is presented in Table 2. Separation of the generated hydrogen from the mixture of the water splitting products, to prevent explosive recombination, is another challenge for thermochemically generated water splitting processes. Separation of the thermochemically generated hydrogen from the mixture of the water splitting products by gas diffusion through a porous ceramic membrane can be relatively effective. Membranes that have been considered include commercial and specially prepared porous zironias, although sintering was observed to occur under thermal water splitting conditions,1,39 and ZrO2-TiO2-Y2O3 oxides.40 In such membranes, it is necessary to maintain a Knudsen flow regime across the porous wall.24 The molecular mean free path λ in
Slide 109: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting Table 2. Melting points of refractory materials, modified from Ref. 1. Material SiO2 Quartz TiO2 Cr2O3 Al2O3 UO2 Y2O3 BeO CeO2 ZrO2 MgO HfO2 ThO2 SiC B4C WC TiC Electrolytic graphite HfC Si3N4 BN Type oxide oxide oxide oxide oxide oxide oxide oxide oxide oxide oxide oxide oxide carbide carbide carbide carbide carbide carbide nitride nitride 93 Melting Point (°C) 1720 1610 1840 1990–2200 2050 2280 2410 2550 2660–2800 2715 2800 2810 3050 2200(decomp) 2450 2600(decomp) 3400–3500 3650(subl) 4160 1900 3000(decomp) the gas must be greater than the average pore diameter, φ. By kinetic theory, λ = √πνd2; where d = molecular diameter (cm), ν = molecular density = 273.15pνo/T, p = pressure (bar), T = temperature (K), and νo = 2.685x1019 molecules/cm3.1 A double-membrane configuration has been suggested as superior to a singlemembrane reactor.41 In recent times, there have been relatively few studies on the direct thermochemical generation of hydrogen by water splitting1,24,39-45 due to continuing high temperature material limitations. Recent experimental work has been performed by Kogan and associates1,37,40-41 and the cross section of one of their solar reactors is shown in Fig. 3. This reactor consists of a cylindrical zirconia housing of 10-cm inside diameter and 20-cm length, and insulated by 2-in thickness of Zircar felt and board. One end of the housing is closed by a circular disc with a central aperture 3 cm in diameter. A zirconia crucible having a porous wall is installed at the opposite end of the housing, and sintering of the crucible considerably limited performance. In 2004, Bayara reiterated that conversion rates in direct thermochemical processes are still quite low, and new reactor designs, operation schemes and materials are needed for new breakthroughs in this field.45
Slide 110: 94 Stuart Licht Fig. 3. Example of a solar reactor for direct thermochemical water splitting and solar hydrogen generation. Reprinted with permission from Ref. 1. Copyright (1998) International Journal of Hydrogen Energy. 3 Indirect (Multi-step) Solar Thermal Water Splitting to Generate Hydrogen Fuel 3.1 Historical Development of Multi-Step Thermal Processes for Water Electrolysis As mentioned earlier, direct thermal dissociation of water requires temperatures above approximately 2500 K. Since there are not yet technical solutions to the materials problems, the possibility of splitting water instead, by various reaction sequences, has been probed. Historically, the reaction of reactive metals and reactive metal hydrides with water or acid was the standard way of producing pure hydrogen in small quantities. These reactions involved sodium metal with water to form hydrogen or zinc metal with hydrochloric acid or calcium hydride with water. All these
Slide 111: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 95 Fig. 4. Temperature variation of the free energy for several decomposition reactions pertinent to hydrogen generation. Reprinted with permission from Ref. 72. Copyright (2004) International Journal of Hydrogen Energy. methods are quite outdated and expensive, including the reaction of metallic iron or ferrous oxide with steam at elevated temperatures to produce hydrogen. The possibility to produce hydrogen in multi-reaction processes from water, with a higher thermal efficiency has been extensively studied. As summarized by Perkins and Weimer, and as shown in Fig. 4, a variety of pertinent, spontaneous processes can be considered that have a negative reaction free energy at temperatures considerably below that for water. These reactions are conducted in a cycle to regenerate and reuse the original reactions, ideally, with the only net reactant water, and the only net products hydrogen and oxygen. However, such cycles suffer from challenges often encountered in multi-step reactions. While these cycles operate at much lower temperatures than the direct thermal chemical generation of hydrogen, conversion efficiencies are insufficient and interest in these cycles has waned. Efficiency losses occur at each of the steps in the multiple step sequence, resulting in low overall solar to hydrogen energy conversion efficiencies. Interest in indirect thermal chemical generation of hydrogen started approximately 40 years ago with an average of less than a handful of publications per year in the decade starting 1964. A dramatic upsurge in interest occurred in the subsequent years with an average of over 70 papers per year from 1975 through 1985. Following that time, and because of the lack of clear successes in the field, interest waned in subsequent years and has averaged only ca.10 publications per year.4
Slide 112: 96 Stuart Licht 3.2 Comparison of Multi-step Indirect Solar Thermal Hydrogen Processes Early studies performed on H2O-splitting thermochemical cycles were mostly characterized by the use of process heat at temperatures below about 1200 K, available from nuclear and other thermal sources. These cycles required multiple steps (more than two) and had inherent inefficiencies associated with heat transfer and product separation at each step. An overview of indirect thermochemical processes for hydrogen generation using more than two steps has been presented by Funk,4 and several of these cycles are summarized in Table 3. An example includes cycle No. 2 in Table 3, which utilizes the following reaction steps: CaBr2 + 2 H2O → Ca(OH)2 + 2 HBr 2 HBr + Hg → HgBr2 + H2 HgBr2 + Ca(OH)2 → CaBr2 + HgO + H2O HgO → CaBr2 + Hg + ½ O2 T = 1050 K T = 450 K T = 450 K T = 900 K 46,47 (5) (6) (7) (8) Status reviews on multiple-step cycles have been presented, and include the leading candidates GA’s 3-step cycle based on the thermal decomposition of H2SO4 at 1130 K,48 and the UT3’s 4-step cycle based on the hydrolysis of CaBr2 and FeBr2 at 1020 and 870 K:49 This process involving two Ca and two Fe compounds has received some attention.4 The process is operated in a cyclic manner in which the solids remain in their reaction vessels and the flow of gases is switched when the desired reaction extent is reached: CaBr2(s) + H2O(l) → CaO(s) + 2 HBr(g) CaO(s) + Br2(g) → CaBr2(s) + O2(g) T = 973–1050 K (9) T = 773–8 K (10) Fe3O4(s) + 8 HBr(g) → 3FeBr2(s) + 4 H2O(g) + 2 Br2(g) T = 473–573 K (11) 3 FeBr2(s) + 4 H2O(g) → Fe3O4(s) + 6 HBr(g) + H2 (g) T = 823–873 K (12) 3.3 High-Temperature, Indirect-Solar Thermal Hydrogen Processes More recently, higher temperature processes have been considered (at T > 2000 K), such as two-step thermal chemical cycles using metal oxide reactions.2 The first step is solar: the endothermic dissociation of the metal oxide to the metal or the lowervalence metal oxide. The second step is non-solar, and is the exothermic hydrolysis of the metal to form H2 and the corresponding metal oxide. The net reaction is H2O = H2 + 0.5 O2, but since H2 and O2 are formed in different steps, the need for high-temperature gas separation is thereby eliminated:
Slide 113: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 97 Table 3. Summary of multi-step chemical cycles for indirect thermochemical hydrogen generation, from Ref. 4. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Elements in cycle Hg,Ca,Br Hg,Ca,Br Cu,Ca,Br Hg,Sr,Br Mn,Na,(K) Mn,Na,(K),C V,Cl,O Fc,Cl,S Hg,Ca,Br,C Cr,Cl,Fe,(V) Cr,Cl,Fe,(V)Cu Fe,Cl Fe,Cl Fe,Cl Mn,Cl Fe,Cl I,S,N S (hybrid) I,S,N,Zn Br,S (hybrid) Fe,Cl Fe,Cl S,I S,I Maximum temperature, K 1050 1050 1070 1070 1070 1120 1070 1070 1120 1070 1070 1070 1070 1120 1120 920 1120 1120 1120 1120 920 920 1120 1120 Total reaction steps in cycle 4 4 4 3 3 4 4 4 5 4 5 5 5 5 3 3 6 2 4 3 5 4 3 3 1st step (solar): MxOy → xM + y/2O2 (13) (14) 2nd step (non-solar): xM + yH2O → MxOy + yH2 where M is a metal and MxOy is the corresponding metal oxide. Such a two-step cycle was originally proposed50 using the redox pair Fe3O4/FeO. The solar step, i.e., the thermal dissociation of magnetite to wustite at above 2300 K, has been thermodynamically examined51 and experimentally studied in a solar furnace.52,53 It was found necessary to quench the products in order to avoid re-oxidation, but quenching introduces an energy penalty of up to 80% of the solar energy input. The redox pair TiO2/TiOx (with x < 2) has been considered.54,55 Solar experiments on the thermal reduction of TiO2, conducted in an Ar atmosphere up to 2700 K, experienced losses due to the chemical conversion limited by the interfacial rate at which O2 diffuses. Other redox pairs, such as Mn3O4/MnO and Co3O4/CoO have also been considered, but the yield of H2 in the reaction has been too low to be of any practical interest.53 H2 may be produced instead by reacting MnO with NaOH at above 900 K in a 3-step cycle.56 Steinfeld further suggests2 that partial substitution of iron in Fe3O4 by other metals (e.g., Mn and Ni) forms mixed metal oxides of the type (Fe1-xMx)3O4 that may be reducible at lower temperatures than those required for the reduction of Fe3O4, while the reduced phase (Fe1-xMx)1-yO remains capable of splitting water.57-59
Slide 114: 98 Stuart Licht Fig. 5. Schematic of a rotating-cavity solar reactor concept for the thermal dissociation of ZnO to Zn and O2 at 2300 K, modifed from Ref. 2. It consists of a rotating conical cavity-receiver (#1) that contains an aperture (#2) for access of concentrated solar radiation through a quartz window (#3). ZnO particles are continuously fed by means of a screw powder feeder located at the rear of the reactor (#4). The gaseous products Zn and O2 continuously exit via an outlet port (#5) and are quenched. One of the most actively studied candidate metal oxide redox pair for the 2-step cycle, is ZnO/Zn. As reviewed by Steinfeld,2 several chemical aspects of the thermal dissociation of ZnO have been investigated.55,60,61 The reaction rate law and Arrhenius parameters for directly irradiated ZnO pellets has been derived.62 The condensation of zinc vapor in the presence of O2 by fractional crystallization in a temperaturegradient tube furnace was studied.63 Alternatively, electro-thermal methods for in situ separation of Zn(g) and O2 at high temperatures have been experimentally demonstrated to work in small-scale solar furnace reactors.64-67 High-temperature separation further enables recovery of the latent heat of the products (e.g., 116 kJ/mol during Zn condensation). Figure 5 shows the schematic configuration of a solar chemical reactor concept that features a windowed rotating cavity-receiver lined with ZnO particles that are held by centrifugal force.68 In this arrangement, ZnO is directly exposed to high-flux solar irradiation and serves simultaneously the functions of radiant absorber, thermal insulator, and chemical reactant. Solar tests carried out with a 10 kW prototype subjected to a peak solar concentration of 4000 suns proved the low thermal inertia of the reactor system. The ZnO surface temperature reached 2000 K in 2 s, and was resistant to thermal shocks.2 Cycles incorporating ZnO continue to be of active research interest.69–73
Slide 115: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 99 4 Hybrid Solar Thermal/Electrochemical/Photo (STEP) Water Splitting 4.1 Historical Development of Hybrid Thermal Processes The solar driven room temperature electrolysis of water, discussed in the introduction to this chapter, can be substantially enhanced by heating the water with excess solar thermal energy. Nicholson and Carlisle first generated hydrogen by water electrolysis in 1800. Modifications, such as high temperature electrolysis of steam74 or water electrolysis by photo-illuminated semiconductors75 had been reported by the 1970s. With increasing temperature, the quantitative decrease in the electrochemical potential necessary to split water to hydrogen and oxygen had been well known by the 1950s76 and as early as in 1980, Bockris had noted, that solar thermal energy could decrease the necessary energy for the electrolytic generation of hydrogen.77 Over the ensuing two decades, designs were intermittently introduced to utilize this principle.78–87 However, the process combines elements of solid state physics and electrochemical theory, complicating rigorous theoretical support of the process. The thermodynamic feasibility of the solar thermal electrochemical generation of hydrogen was initially shown in 2002.88,89 The theory combined photodriven charge transfer, with excess sub-bandgap insolation to lower the water potential and demonstrated water splitting efficiencies in excess of 50%. In 2004, experimental support, which is described in the latter Sections, was provided in support of this theory.90 4.2 Theory of Hybrid Solar Hydrogen Generation Thermally assisted solar electrolysis consists of (i) light harvesting, (ii) spectral resolution of thermal (sub-bandgap) and electronic (super-bandgap) radiation, the latter of which (iiia) drives photovoltaic or photoelectrochemical charge transfer V(iH2O), while the former (iiib) elevates water to temperature T, and pressure, p; finally (iv) V(iH2O) driven electrolysis of H2O(T,p). A schematic representation for this solar thermal water electrolysis (photothermal electrochemical water splitting) is presented in Figure 2, and rather than a field of concentrators, systems may use individual solar concentrators. This hybrid process provides a pathway for efficient solar energy utilization. Electrochemical water splitting, generating H2 and O2 at separate electrodes, largely circumvents the gas recombination limitations of direct solar thermochemical hydrogen formation and the multiple-step Carnot losses of indirect thermochemical processes. Photodriven charge transfer through a semiconductor junction does not utilize photons which have energy below the semiconductor bandgap. Hence a silicon photovoltaic device does not utilize radiation below its bandgap of ~1.1 eV, while an AlGaAs/GaAs multiple bandgap photovoltaic device does not utilize radiation of energy less than the 1.43-eV bandgap of GaAs. As will be shown, this unutilized, available long wavelength insolation represents a significant fraction of the solar spectrum. This long wavelength insolation can be filtered and used to heat water prior to electrolysis. The thermodynamics of heated water dissociation are more
Slide 116: 100 Stuart Licht favorable than that room temperature. This is expressed by a free-energy chemical shift and a decrease in the requisite water electrolysis potential, which can considerably enhance solar water splitting efficiencies. The spontaneity of the H2 generating water splitting reaction is given by the free energy of formation, ΔG°f, of water and with the Faraday constant, F, the potential for water electrolysis: H2O → H2 + ½ O2 0 0 − ΔGsplit = ΔGf, H 0 2O (15) (16) where ΔGf, H O (25 °C, 1 bar, H2Oliq) = –237.1 kJ mol-1, and 2 0 EH O = 2 0 ΔGf, H O 2 2F (17) where EH O (25°C, 1 bar, H2Oliq ) = 1.229 V 2 Reaction 15 is endothermic and the electrolyzed water will undergo self-cooling unless external heat is supplied. The enthalpy balance and its related thermoneutral potential, Etneut, are given by: − ΔH split = ΔH f, H 0 where ΔH f, H 2 O liq 2 O liq 0 (18) (25 °C, 1 bar, H2Oliq) = –285.8 kJ mol-1, and − ΔH f, H O 2 2F Etneut = (19) 0 where Etneut (25 °C, 1 bar, H2Oliq ) = 1.481 V. The water electrolysis rest potential is determined from extrapolation to ideal conditions. Variations of the concentration, c, and pressure, p, from ideality are respectively expressed by the activity (or fugacity for a gas), as a = γc (or γp for a gas), with the ideal state defined at 1 atmosphere for a pure liquid (or solid), and extrapolated from p = 0 or for a gas or infinite dilution for a dissolved species. The formal potential, measured under real conditions of c and p can deviate significantly from the (ideal thermodynamic) rest potential, as for example the activity of water, aw, at, or near, ambient conditions generally ranges from approximately 1 for dilute solutions to less than 0.1 for concentrated alkaline and acidic electrolytes.91–93 The potential for the dissociation of water decreases from 1.229 V at 25 °C in the liquid phase to 1.167 V at 100 °C in the gas phase. Above the boiling the point, pressure is used to express the variation of water activity. The variation of the electrochemical potential for water in the liquid and gas phases are given by:
Slide 117: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 101 RT γ H 2 pH 2 (γ O 2 pO 2 )1 / 2 0 EH 2 O liq = EH O + ln 2 liq 2F aw (20) RT γ H 2 pH 2 (γ O 2 pO 2 )1 / 2 0 EH 2 O gas = EH O + ln 2 gas 2F γ H 2 O pH 2 O (21) The critical point of water is 374 °C and 221 bar. Below the boiling point, E0H2O is similar for 1 bar and high water pressure, but diverges sharply above these conditions. Values of E0H2O include at pH 2 O = 1 bar: 1.229 V (25 °C), 1.167 V (100 °C), 1.116 V (300 °C), 1.034 V (600 °C), 0.919 V (1000 °C), 0.771 V (1500 °C), and at pH 2 O = 500 bar: 1.224 V (25 °C), 1.163 V (100 °C); 1.007 V (300 °C); 0.809 V (600 °C); 0.580 V (1000 °C). Due to overpotential losses, ζ, the necessary applied electrolysis potential is: 0 VH 2 O (T ) = EH 2O 0 (T ) + ζ anode + ζ cathode ≡ (1 + ζ ) EH 2O (T ) (22) The water electrolysis potential energy conversion efficiency occurring at temperature, T, is ηechem(T) ≡ E0H2O(T)/VH2O(T). Solar water splitting processes utilize ambient temperature water as a reactant. An interesting case occurs if heat is introduced to the system; that is when electrolysis occurs at an elevated temperature, T, using water heated from 25 °C. The ratio of the standard potential of water at 25° C and T, is r = E0H2O(25 oC) / E0H2O(T). As shown in Fig. 6, E0H2O(T) diminishes with increasing temperature, as calculated using contemporary thermodynamic values summarized in Table 4.94,95 In this case, an effective water splitting energy conversion efficiency of η’echem > 1 can occur, to convert 25 °C water to H2 by electrolysis at T: 0 0 EH O (T ) EH O (25 o C) ' 2 2 ηechem = r ⋅ ηechem (T ) = r ⋅ = VH 2 O (T ) VH 2 O (T ) (23) For low overpotential electrolysis, VH2O(T > 25 °C) can be less than E0H2O (25 °C), resulting in η’echem > 1 from Eq. 23. Whether formed with pn or Schottky type junctions, the constraints on photovoltaic (solid state) driven electrolysis are identical to those for photoelectrochemical water splitting, although the latter poses additional challenges of semiconductor/electrolyte interfacial instability, area limitations, catalyst restrictions, and electrolyte light blockage. The overall solar energy conversion efficiency of water splitting is constrained by the product of the available solar energy electronic conversion efficiency, ηphot, with the water electrolysis energy conversion efficiency.5 For solar photothermal water electrolysis, a portion of the solar spectrum will be used to drive charge transfer, and an unused, separate portion of the insolation will be used as a thermal source to raise ambient water to a temperature T:
Slide 118: 102 Stuart Licht Fig. 6. Thermodynamic and electrochemical values for water dissociation to H2 and O2 as a function of temperature.3 The curves without squares are calculated at one bar, for liquid water through 100 °C and for steam at higher temperatures. The high pressure utilized in this additional curve (pH2O = 500 bar; pH2 = pO2 = 1 bar) is of general interest as (i) the electrolysis potential is diminished compared to that of water at 1 bar, (ii) the density of the high pressure fluid is similar to that the liquid and (iii) may be generated in a confined space by heating or electrolyzing liquid water. ηsolar = ηphot ⋅ r ⋅ ηechem = ηphot ⋅ EH 0 2O EH 0 (T ) 2O (25 C) ⋅ E o (T ) 1.229 = ηphot ⋅ VH O (T ) VH O (T ) 2 2 0 H 2O (24) Conditions of ηsolar > ηphot can be shown to place specific restrictions on the photoabsorber. When VH2O < Etneut, heat must flow to compensate for the self-cooling which occurs at the electrolysis rate. That is, for an enthalpy balanced system any additional required heat must flow in a flux equivalent to iheat = iH2O, and at an average power Pheat, such that: + Pheat V Etneut = H 2 O iH 2 O (25)
Slide 119: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 103 Table 4. Thermodynamic free energy and enthalpy of water formation for (a) all constituents at 1 bar, and (b) 500 bar water and 1 bar H2 and O2. P of H2,gas = 1 bar P of O2,gas = 1 bar P of H2O = 1 bar H2O ∆G°f ∆H°f kJ/mol state kJ/mol liquid liquid liquid gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas gas 237.1 236.8 225.2 225.2 223.9 219.1 214.0 208.8 203.5 198.1 192.6 187.0 181.4 175.7 170.1 164.4 158.6 152.9 147.1 118.0 88.9 59.6 30.3 0.0 285.8 285.8 280.2 239.5 243.0 243.8 244.8 critical point 245.7 246.5 247.2 247.9 248.4 248.9 249.4 249.9 250.2 250.5 250.8 251.1 252.1 252.8 253.5 254.2 255.1 P of H2,gas = 1 bar P of O2,gas = 1 bar P of H2O = 500 bar PH2O H2O ∆G°f (bar) state kJ/mol 500 500 liquid liquid 236.2 235.9 T(K) PH2O (bar) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ∆H°f kJ/mol 285.0 285.0 298 300 373 373 400 500 600 647 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 2300 2800 3300 3800 4310 500 liquid 220.1 204.9 190.5 176.9 164.6 153.2 142.0 130.9 119.9 108.8 97.8 86.8 75.8 64.8 53.8 282.0 278.8 274.9 268.3 258.1 254.6 253.2 252.5 252.1 252.0 251.9 251.9 251.9 252.0 252.0 500 liquid 500 liquid P = 221 bar 500 super critical 500 super critical 500 500 500 500 500 500 500 500 500 500 super critical super critical super critical super critical super critical super critical super critical super critical super critical super critical A photoelectrolysis system can contain multiple photo-harvesting units and electrolysis units, where the ratio of electrolysis to photovoltaic units is defined as R. Efficient water splitting occurs with the system configured to match the water electrolysis and photopower maximum power point. This is illustrated in Fig. 1 representing the photosensitizers as power supplies driving electrolysis with a photodriven charge from a photon flux to generate a current density (electrons per unit area) to provide the two stoichiometric electrons per split water molecule. For example, due to a low photopotential, a photodriven charge from three serial arranged Si energy gap devices may be required to dissociate a single room temperature water molecule, as described in the lower right portion of the figure. Alternately, as in a multiple bandgap device such as AlGaAs/GaAs, the high potential of a single photodriven charge may be sufficient to dissocate two room temperature water molecules, as described in the upper right portion of the figure. In the figure, consider, four different photoelectrolysis systems, each functioning at the same efficiency for solar
Slide 120: 104 Stuart Licht conversion of electronic power, ηphot, and the same efficiency for solar conversion of thermal power,ηheat. Whereas the photoconverter in the first system generates the requisite water electrolysis potential, that in the second system generates twice that potential (albeit at one half the photocurrent), while the photoconverter in the third and fourth units generate respectively only half or a third this potential (albeit at twofold or threefold the photocurrent to retain the same efficiency). The harvested photon power for electronic energy per unit insolation area will be the same in each of the four cases. Furthermore, the number of harvested photons for thermal energy, and the total thermal power available to heat water, will be the same of the four cases. For example in Case II, although twice the number of electrolysis units are utilized, each operates at only half the hydrolysis current compared to Cases I, III and IV, splitting the same equivalents of water. For solar driven charge transfer, this maximum power is described by the product of the insolation power, Psun, with ηphot, which is then applied to electrolysis, ηphotPsun = Pechem = iH2OVH2O. Rearranging for iH2O , and substitution into Eq. 25, yields for heat balanced solar electrolysis at conditions of T and p, initiating with 25 °C, 1 bar water: Etneut = 1.481 V = VH 2 O (T , p ) 1 + Pheat ηphot Psun (26) As also elaborated in Chapter 2, Fig. 7 presents the available insolation power, Pλmax (mW cm–2) of the integrated solar spectrum up to a minimum electronic excitation frequency, νmin (eV), determined by integrating the solar spectral irradiance, S(mWcm–2nm–1), as a function of a maximum insolation wavelength, λmax (nm). This Pλmax is calculated for the conventional terrestial insolation spectrum either above the atmosphere, AM0, or through a 1.5 atmosphere pathway, AM1.5. Relative to the total power, Psun, of either the AM0 or AM1.5 insolation, the fraction of this power available through the insolation edge is designated Prel = Pλmax / Psun. In solar energy balanced electrolysis, excess heat is available primarily as photons without sufficient energy for electronic excitation. The fraction these sub-bandgap photons in insolation is αheat = 1 – Prel , and comprises an incident power of αheatPrel. Figure 8 presents the variation of the minimum electronic excitation frequency, νmin with αheat, determined from Prel using the values of Pλmax summarized in Fig. 7. A semiconductor sensitizer is constrained not to utilize incident energy below the bandgap. As seen in Figure 8 by the intersection of the solid line with νmin, over one third of insolation power occurs at νmin < 1.43 eV (867 nm), equivalent to the IR not absorbed by GaAs or wider bandgap materials. The calculations include both the AM0 and AM1.5 spectra. In the relevant visible and IR range from 0.5 to 3.1 eV (±0.03 eV) for both the AM0 and AM1.insolation spectra, νmin(αheat) in the figure are well represented (R2 > 0.999) by polynomial fits. When captured at a thermal efficiency of ηheat, the sub-bandgap insolation power is ηheatαheatPsun. Other available system heating sources include absorbed superbandgap photons which do not effectuate charge separation, Precomb, and noninsolation sources, Pamb, such as heat available from the ambient environment heat
Slide 121: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 105 Fig. 7. The solar irradiance (mW cm–2 nm–1) in the figure inset, and the total insolation power (mW cm–2) in the main figure of the solar spectrum;3 see also Chapter 2. sink, and Precov, such as heat recovered from process cycling or subsequent H2 fuel utilization. The power equivalent for losses, such as the low power consumed in delivering the heated water to electrolysis, Ppump, can also be incorporated. In Figs. 9 to12, determinations of the solar water splitting energy conversion are summarized, calculated using the E0H2O(T,p) data in Fig. 8, and for various solar water splitting system's minimum allowed bandgap, Eg-min(T,p) for a wide temperature range. Figures 9 and 10, or 8 and 11, are respectively modeled based on AM1.5, or AM0, insolation. Figures 9 and 11 are calculated, at various temperatures for pH2O = 1 bar; while Figures 10 and 12 repeat these calculations for pH2O = 500 bar. The ηsolar-max value is significantly greater for higher pressure photoelectrolysis (pH2O = 500 bar). However as seen comparing the minimum bandgap in these figures (or in Figures 11 and 12 in the analogous AM0 models), at these higher pressures, this higher rate of efficiency increase with temperature is offset by lower accessible temperatures (for a given bandgap). Larger ζ in VH2O will diminish ηsolar, but will extend the usable small bandgap range. Together, Figs. 9–12 show the constraints on ηsolar from various values of ηphot.
Slide 122: 106 Stuart Licht Fig. 8. αheat, the fraction of solar energy available below the minimum sensitizer insolation frequency used to drive charge transfer, νmin.3 The αheat is determined as 1– Prel, with Prel = Pλmax/Psun, and using values of Pλmax summarized in Fig. 7. The available incident power below νmin will be αheatPsun. The bandgaps of various semiconductors are superimposed as vertical lines in the figure. The high end of contemporary experimental high solar conversion efficiencies ranges from 100% ηphot = 19.8% for multicrystalline single junction photovoltaics to 27.6% and 32.6% for single junction and multiple junction photovoltaics.97 The efficiency of solar thermal conversion tends to be higher than solar electrical conversion, ηphot, particularly in the case of the restricted spectral range absorption used here with values of ηheat = 0.5, 0.7 or 1. While a small bandgap, Eg < 1.23 eV, is insufficient for water cleavage at 25 °C, its inclusion in Figs. 9–12 is of relevance in two cases, 1. 2. where high temperature decreases VH2O(T) below Eg, and where this Eg is part of a multiple bandgap sensitization contributing a portion of a larger overall photopotential.
Slide 123: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 107 Fig. 9. Energy conversion efficiency of solar driven water splitting to generate H2 as a function of temperature for AM1.5 insolation, with the system minimum bandgap determined at pH2O = 1 bar.3 The maximum photoelectrolysis efficiency is shown for various indicated values of ηphot.
Slide 124: 108 Stuart Licht Fig. 10. The energy conversion efficiency of solar driven water splitting to generate H2 as a function of temperature at AM1.5 insolation, with the system minimum bandgap determined at pH2O = 500 bar.3 The maximum photoelectrolysis efficiency is shown for various indicated values of ηphot.
Slide 125: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 109 Fig. 11. The energy conversion efficiency of solar driven water splitting to generate H2 as a function of temperature at AM0 insolation, with the system minimum bandgap determined at pH2O = 1 bar.3 The maximum photoelectrolysis efficiency is shown for various indicated values of ηphot.
Slide 126: 110 Stuart Licht Fig. 12. The energy conversion efficiency of solar driven water splitting to generate H2 as a function of temperature at AM0 insolation, with the system minimum bandgap determined at pH2O = 500 bar.3 The maximum photoelectrolysis efficiency is shown for various indicated values of ηphot.
Slide 127: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 111 Fig. 13. VH2O, measured in aq. saturated or molten NaOH, at 1 atm.90 CO2 is excluded by argon purge. The molten electrolyte is prepared from heated, solid NaOH with steam injection. O2 anode is 0.6-cm2 Pt foil. IR and polarization losses are minimized by sandwiching 5 mm from each side of the anode, two interconnected Pt gauze (200 mesh, 50 cm2 = 5 cm x 5 cm x 2 sides) cathodes. Inset: At 25 °C, 3 electrode values at 5 mV/s versus Ag/AgCl, with either 0.6cm2 Pt or Ni foil, and again separated 5 mm from two 50-cm2 Pt gauze acting as counter electrodes. It is also noted that Eg > 3.0 eV is in adequate for efficient use of the solar spectrum. Representative results from Fig. 12 for solar water splitting to H2 systems from AM1.5 insolation include a 50% solar energy conversion for a photoelectrolysis system at 638 °C with pH2O = 500; pH2 = 1 bar and ηphot = 0.32. 4.3 Elevated Temperature Solar Hydrogen Processes and Components Fletcher, repeating the fascinating suggestion of Brown that saturated aqueous NaOH will never boil, hypothesized that a useful medium for water electrolysis might be saturated, aqueous solutions of NaOH at very high temperatures.98 These do not reach boiling point at 1 atm due to the high salt solubility, binding solvent, and changing saturation vapor pressure, as reflected in their phase diagram.98 This domain is considered here, and also electrolysis in an even higher temperature domain
Slide 128: 112 Stuart Licht Fig. 14. Measured VH2O (30 mA cm–2) in aq. saturated or molten NaOH compared to thermodynamic EH2O values.90 above which NaOH melts (318 °C) creating a molten electrolyte with dissolved water, resulting in unexpected VH2O. Figure 13 summarizes measured VH2O(T) in aqueous saturated and molten NaOH electrolytes. As seen in the inset, Pt exhibits low over potentials to H2 evolution, and is used as a convenient quasi-reference electrode in the measurements which follow. As also seen in the inset, Pt exhibits a known large overpotential to O2 evolution as compared to a Ni electrode or to E0H2O (25 °C) = 1.23 V. This overpotential loss diminishes at moderately elevated temperatures, and as seen in the main portion of the figure, at 125 °C there is a 0.4 V decrease in the O2 activation potential at a Pt surface. Through 300 °C in Fig. 13, measured VH2O remains greater than the calculated thermodynamic rest potential. Unexpectedly, VH2O at 400 °C and 500 °C in molten NaOH occurs at values substantially smaller than that predicted. These measured values include voltage increases due to IR and hydrogen overpotentials, and hence provide an upper bound to the unusually small electrochemical potential. This phenomenon is summarized in Fig. 14, in which even at relatively large rates of water splitting (30 mA cm-2) at 1 atm, a measured VH2O below that predicted by theory is observed at temperatures above the NaOH melting point. Theoretical
Slide 129: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 113 calculations over an expanded temperature range are presented in Fig. 8, with calculations described in that Section. As seen in Fig. 14, the observed value at high temperature of VH2O approaches that calculated for a thermodynamic system of 500 bar, rather than 1 bar, H2O. A source of this anomaly is described in Fig. 15. Shown on the left hand is the single compartment cell utilized here. Cathodically generated H2 is in close proximity to the anode, while anodic O2 is generated near the cathode. Their presence will facilitate the water forming back reaction, and at the electrodes this recombination will diminish the potential. In addition to the observed low potentials, two observations support this recombination effect. The generated H2 and O2 is collected, but is consistent with a Coulombic efficiency of ≈50% (varying with T, j, and interelectrode separation.) Consistent with the right hand side of Fig. 15, when conducted in separated anode / cathode compartments, this observed efficiency is 98%–100%. Here however, all cell open circuit potentials increase to beyond the thermodynamic potential, and at j = 100 mA cm-2 yields measured VH2O values of 1.45V, 1.60V, 1.78V at 500°, 400°, and 300°, which are approximately 450 mV higher than the equivalent Fig. 13 values for the single configuration cell. The recombination phenomenon offers advantages (low VH2O), but also disadvantages (H2 losses), requiring study to balance these competing effects to optimize energy efficiency. In molten NaOH, the effects of temperature variation of ΔG0f (H2O) and the recombination of the water splitting products can have a pronounced effect on solar driven electrolysis. As compared to 25 °C, in Fig. 13 only half the potential is required to split water at 500 °C over a wide range of current densities. The unused thermal photons which are not required in semiconductor photodriven charge generation, can contribute to heating water to facilitate electrolysis at an elevated temperature. The characteristics of one, two, or three series interconnected Fig. 15. Interelectrode recombination can diminish VH2O and occurs in open (left) but not in isolated (right) configurations; such as those examined with or without a Zr2O mix fiber separator (ZYK-5H, from Zircar Zirconia, FL, NY) situated between the Pt anode and cathodes.90
Slide 130: 114 Stuart Licht Fig. 16. Photovoltaic and electrolysis charge transfer for thermal electrochemical solar driven water splitting.90 Photocurrent is shown for one, two or three 1.561 cm2 HECO 335 Sunpower Si photovoltaics in series at 50 suns. Photovoltaics drive 500-°C molten NaOH steam electrolysis using Pt gauze anode and cathodes. Inset: electrolysis current stability. solar visible efficient photosensitizers, in accord with the manufacturer’s calibrated standards, are presented in Fig. 16. These silicon photovoltaics are designed for efficient photoconversion under concentrated insolation (ηsolar = 26.3% at 50 sun). Superimposed on the photovoltaic response curves in the figure are the water electrolysis current densities for one, or two series interconnected, 500 °C molten NaOH single compartment cell configuration electrolyzers. Constant illumination generates for the three series-connected cells, a constant photopotential for stability measurements at sufficient power to drive two series molten NaOH electrolyzers. At this constant power, and as presented in the lower portion of Fig. 16, the rate of water splitting appears fully stable over an extended period. In addition, as measured and summarized in the upper portion of the figure, for the overlapping region between the solid triangle and open square curves, a single Si photovoltaic can drive 500 °C water splitting, albeit at an energy beyond the maximum power point voltage, and therefore at diminished efficiency. This appears
Slide 131: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 115 to be the first case in which an external, single, small bandgap photosentizer can cleave water, and is accomplished by tuning the water splitting electrochemical potential to decrease to a point below the Si open circuit photovoltage. VH2O-tunned is accomplished by two phenomena: 1. 2. the thermodynamic decrease of EH2O with increasing temperature, and a partial recombination of the water splitting products. VH2O-tunned can drive system efficiency advances, e.g., AlGaAs/GaAs, transmits more insolation, EIR < 1.4 eV, than Si to heat water, and with ηphoto over 30%, prior to system engineering losses, calculates to over 50% ηsolar to H2. Without inclusion of high temperature effects, we had already experimentally achieved ηsolar > 0.18, using an ηphot = 0.20 AlGaAs/Si system.5 Our use of more efficient, ηsolar = 26.3% at 50 suns, and inclusion of heat effects and the elevated temperature decrease of the water electrolysis potential, substantially enhances ηsolar.90 Existing, higher ηphot (= 0.28 to 0.33) systems should achieve proportionally higher results. Experimental components, for example as described in Fig. 2, for solar driven generation of H2 fuel at 40-50% conversion efficiencies appear to be technologically available. In the high efficiency range, photoelectrochemical cells tend to be unstable, which is likely to be exacerbated at elevated temperatures, and the model system will be particularly conducive to photovoltaic, rather than photoelectrochemical, driven electrolysis. As has already been elaborated in Chapter 4, the photovoltaic component is used for photodriven charge into the electrolysis component and does not contact the heated electrolyte. Stable photovoltaics are commonly driven with concentrated insolation97 and specific to the system model here, heat will be purposely filtered from the insolation prior to incidence on the photovoltaic component. Dielectric filters used in laser optics split insolation without absorption losses (Chapter 4). For example, in a system based on a parabolic concentrator, a casegrain configuration may be used, with a mirror made from fused silica glass with a dielectric coating acting as band pass filter. The system will form two focal spots with different spectral configuration, one at the focus of the parabola and the other at focus of the casegrain.99 The thermodynamic limit of concentration is 46000 suns, the brightness of the surface of the sun. In a medium with refractive index greater than one, the upper limit is increased by two times the refractive index, although this value is reduced by reflective losses and surface errors of the reflective surfaces, the tracking errors of the mirrors and dilution of the mirror field. Specifically designed optical absorbers, such as parabolic concentrators or solar towers, can efficiently generate a solar flux with concentrations of ~2000 suns, generating temperatures in excess of 1000 °C.100,101 Commercial alkaline electrolysis occurs at temperatures up to 150 °C and pressures to 30 bar,96 and super-critical electrolysis to 350 °C and 250 bar.102 Although less developed than their fuel cell counterparts which have 100 kW systems in operation and developed from the same oxides,103 zirconia and related solid oxide based electrolytes for high temperature steam electrolysis can operate efficiently at 1000 °C,104,105 and approach the operational parameters necessary for efficient solar
Slide 132: 116 Stuart Licht driven water splitting. Efficient multiple bandgap solar cells absorb light up to the bandgap of the smallest bandgap component. Thermal radiation is assumed to be split off (removed and utilized for water heating) prior to incidence on the semiconductor and hence will not substantially effect the bandgap. Highly efficient photovoltaics have been demonstrated at a solar flux with a concentration of several hundred suns. AlGaAs/GaAs has yielded a ηphot efficiency of 27.6% and a GaInP/GaAs cell 30.3% at 180 suns concentration, while GaAs/Si has reached 29.6% at 350 suns, InP/GaInAs 31.8%, and GaAs/GaSb 32.6% with concentrated insolation.97 5 Future Outlook and Concluding Remarks A hybrid solar thermal/electrochemical process combines efficient photovoltaic devices and concentrated excess sub-bandgap heat into highly efficient elevated temperature solar electrolysis of water to generate H2 fuel. Efficiency is further enhanced by excess super-bandgap and non-solar sources of heat but diminished by losses in polarization and photo-electrolysis power matching. As also shown earlier in Chapter 4 and elaborated further here, solar concentration can provide the high temperature and diminish the requisite surface area of efficient electrical energy conversion components. High temperature electrolysis components are commercially available, suggesting that highly efficient solar generation of H2 will be ultimately attainable. References 1. A. Kogan, Direct solar thermal splitting of water and on-site separation of the products. II. Experimental feasibility study, Int. J. Hydrogen Energy, 23 89-98 (1998). 2. A. Steinfeld, Solar thermochemical production of hydrogen-a review, Solar Energy, 78 603–615 (2005). 3. S. Licht, Solar water splitting to generate hydrogen fuel–A photothermal electrochemical analysis, Int. J. Hydrogen Energy, 30 459–470 (2005). 4. J. E. Funk, Thermochemical hydrogen production: past and present, Int. J. Hydrogen Energy, 26 185–190 (2001). 5. S. Licht, B. Wang, S. Mukerji, T. Soga, M. Umeno, and H. Tributsh, Over 18% solar energy conversion to generation of hydrogen fuel; theory and experiment for sefficient solar water splitting, Int. J. Hydrogen Energy, 26 653–659 (2001). 6. K. Agabossu, R. Chahine, J. Hamelin, F. Laurencelle, A. Anouar, J.-M. St-Arnaud, and T. K. Bose, Renewable energy systems based on hydrogen for remote applications, J. of Power Sources 96 168–172 (2001). 7. T. Ohmori, H. Go, N. Yamaguchi, A. Nakayama, H. Mametisuka, and E. Suzuki, Photovoltaic water electrolysis using the sputter-deposited a-Si/c-Si solar cells, Int. J. of Hydrogen Energy 26 661–664 (2001). 8. T. Tani, N. Sekiguchi, M. Sakai, and D. Otha, Optimization of solar hydrogen systems based on hydrogen production cost, Solar Energy 68 143–149 (2000).
Slide 133: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 117 9. P. Hollmuller, J.-M. Jouibert, B. Lachal, and K. Yvon, Evaluation of a 5-kWp photovoltaic hydrogen production and storage installation for a residential home in Switzerland, Int. J. of Hydrogen Energy 25 97–109 (2000). 10. S. Schulien, G. Sandstede, and H. W. Hahn, Hydrogen and carbon dioxide as raw materials for ecological energy – technology, Int. J. of Hydrogen Energy 24 299–303 (1999). 11. C. Meurer, H. Barthels, W. A. Brocke, B. Emonts, and H. G. Groehn, Phoebus – and autonomous supply system with renewable energy: six years of operational experience and advanced concepts, Solar Energy 67 131–138 (1999). 12. A. Szyszka, Ten year of solar hydrogen demonstration project at Neunburg vorm Wald, Germany, Int. J. of Hydrogen Energy 23 849–860 (1998). 13. P. A. Lehman, C. E. Chamberlin, G. Pauletto, and M. A. Rocheleau, Operating experience with photovoltaic-hydrogen energy system, Int. J. of Hydrogen Energy 22 465–470 (1997). 14. S. Galli and M. Stefanoni, Development of a solar-hydrogen cycle in Italy, Int. J. of Hydrogen Energy 22 453–458 (1997). 15. J. W. Hollenberg, E. N. Chen, K. Lakeram, and D. Modroukas, Development of a photovoltaic energy conversion system with hydrogen energy storage, Int. J. of Hydrogen Energy 20 239–243 (1995). 16. E. Bilgen, Solar hydrogen from photovoltaic-electrolyzer systems, Energy Conversion & Management 42 1047–1057 (2001). 17. M. P. Rzayeva, O. M. Salamov, and M. K. Kerimov, Modeling to get hydrogen and oxygen by solar water electroclysis, Int. J. of Hydrogen Energy, 26 195–201 (2001). 18. S. Licht, Multiple bandgap semiconductor/electrolyte solar energy conversion, J. Phys., Chem. B, 105 6281–6294 (2001). 19. Semiconductor Electrodes and Photoelectrochemistry, Edited by S. Licht, Wiley-VCH, Weinheim, 2002. 20. A. Fujishima and K. Honda, Nature 37 238 (1972). 21. A. Heller, E. Asharon-Shalom, and W. A. Bonner, B. Miller, Hydrogen-evolving semiconductor photocathodes: nature of the junction and function of the platinum group metal catalyst, J. Am. Chem. Soc. 104 6942–6948 (1982). 22. O. Khaselev and J. A. Turner, A monolithic photovoltaic-photoelectrochemical device for hydrogen production via waer splitting, Science, 280 425–427 (1998). 23. S. Licht, B. Wang, S. Mukerji, T. Soga, M. Umeno, and H. Tributsch, Efficient Solar Water Splitting, Conversion, J. Phys., Chem., B, 104 8920–8924 (2000). 24. H. Ohya, M. Yatabe, M. Aihara, Y. Negishi, and T. Takeuchi, Feasibility of hydrogen production above 2500 K by direct thermal decomposition reaction in membrane reactor using solar energy, Int. J. Hydrogen Energy, 27 369–376 (2002). 25. E. A.Fletcher and R. L.Moen, Hydrogen and oxygen from water, Science, 197 105 (1977). 26. J. E. Noring, R. B. Diver and E. A. Fletcher, Hydrogen and oxygen water V. The ROC system, Energy, 6 109 (1981). 27. R. B. Diver, S. Pederson, T. Kappauf, and E. A. Fletcher, Hydrogen and oxygen from water: VI. Quenching the effluent from a solar furnace, Energy 8 947 (1983). 28. G. Olalde, D. Gauthier, and A. Vialaron, Film boiling around a zirconia target. Application to water thermolysis, Adv. Ceramics, 24 879–883 (1988). 29. J. Lede, F. Lapigque, J. Villermaux, B. Cales, A. Ounalli, J. F. Baumard, and A. M. Anthony, Production of hydrogen by direct thermal decomposition of water: Preliminary investigations, Int. J. Hydrogen Energy, 7 939–950 (1982). 30. F. Lapigque, J. Lede, L. Villermaux, A. Cales, J. Baumard, A. M. Anthony, E. Abdul Aziz, D. Peuchberty, and M. Ledoux, Entropie, 110 42 (1983).
Slide 134: 118 Stuart Licht 31. J. Lede, J. Villermaux, R. Ouzane, M. A. Hossain, and R. Ouahes, Production of hydrogen by simple impingement of a turbulent jet of steam upon a high temperature zirconia surface, Int. J. Hydrogen Energy, 12 3–11 (1987). 32. A. Ounalli, B. Cales, K. Dembrinski, and J. F. Baumard, C. R. Acad. Sci. Paris, 292(11) 1185 (1981). 33. E. Bilgen, Solar hydrogen production by direct water decomposition process: a preliminary engineering assessment, Int. J. Hydrogen Energy, 9 53–48 (1984). 34. E. Bilgen, M. Duccarroir, M. Foex, F. Silieude, and F. Trombe, Use of solar energy for direct and two-step water decomposition cycles, Int. J. Hydrogen Energy, 2 251–257 (1977). 35. S. Ihara, Feasibility of hydrogen production by direct water splitting at high temperature, Int. J. Hydrogen Energy, 3 287–296 (1978). 36. S. Ihara, On the study of hydrogen production from water using solar thermal energy, Int. J. Hydrogen Energy, 5 527–534 (1980). 37. A. Yogev, A. Kribus, M. Epstein and A. Kogan, Solar “Thermal Reflector” systems: A new approach for high-temperature solar plants, Int. J. Hydrogen Energy 26 239–245 (1998). 38. A. Kribus, P. Doron, R. Rubin, J. Karni, R. Reuven, S. Duchan, and E. Taragan, A multistage solar receiver: the route to high temperature, Solar Energy, 67 2–11 (2000). 39. A. Kogan, Direct solar thermal splitting of water and on-site separation of the products. IV. Development of porous ceramic membranes for a solar thermal water-splitting reactor, Int. J. Hydrogen Energy, 25 1043–1050 (2000). 40. H. Naito and H. Arashi, Hydrogen production from direct water splitting at high temperatures using a ZrO2-TiO2-Y2O3 membrane, Solid State Ionics, 79 366–370 (1995). 41. R. P. Omorjan, R. N. Paunovic, M. N. Tekic, and M. G. Antov, Maximal extent of an isothermal reversible gas-phase reaction in single- and double-membrane reaction; direct thermal splitting of water, Int. J. Hydrogen Energy, 26 203–212 (2001). 42. A. Kogan, Direct solar thermal splitting of water and on-site separation of the products. I. Theoretical evaluation of hydrogen yield, Int. J. Hydrogen Energy, 22 481–486 (1997). 43. A. Kogan, E. Spiegler, and M. Wolfshtein, Direct solar thermal splitting of water and onsite separation of the products. III. Improvement of reactor efficiency by steam entrainment, Int. J. Hydrogen Energy, 25 739–745 (2000). 44. S. Z. Baykara, Hydrogen production by direct solar thermal decomposition of water, possibilities for improvement of process efficiency, Int. J. Hydrogen Energy, 29 1451–1458 (2004). 45. S. Z. Baykara, Experimental solar water thermolysis, Int. J. Hydrogen Energy, 29 1459– 1469 (2004). 46. N. Serpone, D. Lawless, and R. Terzian, Solar fuels: status and perspectives, Solar Energy, 49 221–234 (1992). 47. J. Funk, Thermochemical hydrogen production past and present, Int. J. Hydrogen Energy, 26 185–190 (2001). 48. D. OKeefe, C. Allen, G. Besenbruch, L. Brown, J. Norman, R. Sharp, and K. McCorkle, Preliminary results from bench-scale testing of sulfur-iodine thermochemical watersplitting cycle, Int. J. Hydrogen Energy, 7 381–392 (1982). 49. M. Sakurai, E. Bilgen, A. Tsutsumi, and K. Yoshida, Solar UT-3 thermochemical cycle for hydrogen production, Solar Energy, 57 51–58 (1996). 50. T. Nakamura, Hydrogen production from water utilizing solar heat at high temperatures, Solar Energy, 19 467–475 (1977). 51. A. Steinfeld, S. Sanders, and R. Palumbo, Design aspects of solar thermochemical engineering, Solar Energy, 65 43–53 (1999). 52. A. Tofighi, Ph.D. Thesis, L’Institut National Polytechnique de Toulouse, France, 1982.
Slide 135: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 119 53. F. Sibieude, M. Ducarroir, A. Tofighi, and J. Ambriz, High-temperature experiments with a solar furnace: the decomposition of Fe3O4,Mn3O4, CdO, Int. J. Hydrogen Energy, 7 79– 88 (1982). 54. R. D. Palumbo, A. Rouanet, and G. Pichelin, The solar thermal decomposition of TiO2 above 2200 K and its use in the production of Zn from ZnO, Energy - Int. J., 20 857– 868 (1995). 55. R. Palumbo, J. Lede, O. Boutin, E. Elorza Ricart, A. Steinfeld, S. Moeller, A. Weidenkaff, E.A. Fletcher, and J. Bielicki, The production of Zn from ZnO in a single step high temperature solar decomposition process, Chem. Eng. Sci., 53 2503–2518 (1998). 56. M. Sturzenegger and P. Nuesch, Efficiency analysis for a manganese-oxide-based thermochemical cycle, Energy, 24 959–970 (1999). 57. K. Ehrensberger, A. Frei, P. Kuhn, H.R. Oswald, and P. Hug, Comparative experimental investigations on the water-splitting reaction with iron oxide Fe1-yO and iron manganese oxides (Fe1-xMnx) 1-yO, Solid State Ionics, 78 151–160 (1995). 58. Y. Tamaura, A. Steinfeld, P. Kuhn, and K. Ehrensberger, Production of solar hydrogen by a novel, 2-step, watersplitting thermochemical cycle, Energy, 20 325–330 (1995). 59. Y. Tamaura, M. Kojima, Y. Sano, Y. Ueda, N. Hasegawa, and M. Tsuji, Thermodynamic evaluation of watersplitting by a cation-excessive (Ni, Mn) ferrite, Int. J. Hydrogen Energy, 23 1185–1191 (1998). 60. A. Weidenkaff, A. Reller, A. Wokaun, and A. Steinfeld, Thermogravimetric analysis of the ZnO/Zn water splitting cycle, Thermochim. Acta, 359 69–75 (2000). 61. A. Weidenkaff, A. Reller, F. Sibieude, A. Wokaun, and A. Steinfeld, Experimental investigations on the crystallization of zinc by direct irradiation of zinc oxide in a solar furnace, Chem. Mater., 12 2175–2181 (2000). 62. S. Moeller and R. Palumbo, Solar thermal decomposition kinetics of ZnO in the temperature range 1950–2400 K, Chem. Eng. Sci., 56 4505–4515 (2001). 63. A. Weidenkaff, A. Wuillemin, A. Steinfeld, A. Wokaun, B. Eichler, and A. Reller, The direct solar thermal dissociation of ZnO: condensation and crystallization of Zn in the presence of oxygen, Solar Energy, 65 59–69 (1999). 64. E. A. Fletcher, Solar thermal and solar quasi-electrolytic processing and separations: zinc from zinc oxide as an example, Ind. Eng. Chem. Res., 38 2275–2282 (1999). 65 E. A. Fletcher, F. Macdonald, and D. Kunnerth, High temperature solar electrothermal processing II. Zinc from zinc oxide, Energy, 10 1255–1272 (1985). 66. D. J. Parks, K.L. Scholl, and E.A. Fletcher, A study of the use of Y2O3 doped ZrO2 membranes for solar electro-thermal and solar thermal separations, Energy, 13 121 –136 (1988). 67. R. D. Palumbo and E. A. Fletcher, High temperature solar electro-thermal processing. III. Zinc from zinc oxide at 1200–1675 K using a non-consumable anode, Energy, 13 319–332 (1988). 68. P. Haueter, S. Moeller, R. Palumbo, and A. Steinfeld, The production of zinc by thermal dissociation of zinc oxide - solar chemical reactor design, Solar Energy 67 161–167 (1999). 69. H. Aoki, H. Kaneko, N. Hasegawa, H. Ishihara, A. Suzuki, and Y. Tamaura, The ZnFe2O4/(ZnO+Fe3O4) system for H2 production using concentrated solar energy, Solid State Ionics, 172, 113-116, 2004 70. M. Inoue, N. Hasewaga, R. Uehara, N. Gokon, H. Kaneko, and Y. Tamaura, Solar hydrogen generation with H2O/ZNO/MnFe2O4 system, Solar Energy, 76 309–315 (2004). 72. C. Perkins and A. W. Weimer, Likely near-term solar-thermal water splitting technologies, Int. J. of Hydrogen Energy, 29 1587–1599 (2004). 73. H. Kaneko, N. Gokon, N. Hasewaga, and Y. Tamaura, Solar thermochemcial process for hydrogen production using ferrites, Energy, 30 2171–2178 (2005).
Slide 136: 120 Stuart Licht 74. P. Blum, Cell for electrolysis of steam at high temperture, U.S. Patent 3, 993,653, Dec. 9, 1975. 75. D. I. Tcherev, Device for solar energy Conversion by photo-electrolytic decomposition of water, U.S. Patent 3, 925,212, Nov. 23, 1976. 76. A. J. DeBethune, T. S. Licht, and N. S. Swendemna, The temperature coefficient of Electrode Potentials, J. Electrochem. Soc., 106 618–625 (1959). 77. J. O'M. Bockris, Energy Options, Halsted Press, NY, 1980. 78. D. E. Monahan, Process and apparatus for generating hydrogen and oxygen using solar energy, U.S. Patent 4,233,127, Nov. 11, 1980. 79. L. E. Crackel, Spectral converter, U.S. Patent 4,313,425, Feb. 2, 1982. 80. C. Alkan, M. Sekerci, and S. Kung, Production of hydrogen using Fresnel lens-solar electrochemical cell, Int. J. of Hydrogen Energy, 20 17–20 (1995). 81. C. W. Neefe, Passive hydrogel fuel generator, U.S. Patent 4,511,450, April 16, 1985. 82. D. E. Soule, Hybrid solar energy generating system, U.S. Patent 4,700,013, Oct. 13, 1987. 83. G. Tindell, Electrical energy production apparatus, U.S. Patent 4,841,731, June 27, 1989. 84. J. B. Lasich, Production of hydrogen from solar radiation at high efficiency, U.S. Patent 5,973,825, Oct. 26, 1999. 85. S. R. Vosen and J. O. Keller, Hybrid energy storage systems for stand-alone electric power systems: optimization of system performance and cost through control strategies, Int. J. of Hydrogen Energy, 24 1139–1156 (1999). 86. J. Padin, T. N. Veziroglu, and A. Shahin, Hybrid solar high-temperature hydrogen production system, Int. J. of Hydrogen Energy, 25 295–317 (2000). 87. H. Izumi, Hybrid solar collector for generating electricity and heat by separating solar rays into long wavelength and short wavelength, U.S. Patent 6,057,504, May 2, 2000. 88. S. Licht, Efficient solar generation of hydrogen fuel - a fundamental analysis, Electrochemical Communications, 4 790–795 (2002). 89. S. Licht, Solar water splitting to generate hydrogen fuel: Photothermal electrochemical analysis, J. Phys. Chem. B, 107 4253–4260 (2002). 90. S. Licht, L. Halperin, M. Kalina, M. Zidman, and N. Halperin, Electrochemical Potential Tuned Solar Water Splitting, Chemical Communications, 3006-3007 (2003). 91. S. Licht, pH measurement in conentrated alkaline solutions, Anal. Chem., 57 514–519 (1987). 92. T. S. Light, T. S, Licht, A. C. Bevilacqua, and Kenneth R. Morash, Conductivity and resistivity of ultrapure water, Electrochem. Solid State Lett., 8 E16–E19 (2005) 93. S. Licht, Analysis in highly concentrated solutions: Potentiometric, conductance, evanescent, densometric, and spectroscopic methodolgies, in Electroanalytical Chemistry, Vol. 20, Edited by A. Bard and I. Rubinstein, Marcel Dekker, NY, 1998, pp. 87–140. 94. M. W. Chase, J. Phys. Chem. Ref. Data 14, Monograph 9 (JANF Thermochemical Tables, 4th edition), 1998. 95. M. W. Chase, J. Phys. Chem. Ref. Data Supplement No. 1 to 14, (JANF Thermochemical Tables, 3rd edition), 1986. 96. W. Kreuter and H. Hofmann, Electrolysis: The important energy transformer in a world of sustainable energy, Int. J. Hydrogen Energy, 23 661–669 (1998). 97. M. A. Green, K. Emery, D. L. King, S. Igari, and W. Warta, Solar Efficiency Tables (Version 17), Progr. Photovolt, 9 49–56 (2001). 98. E. Fletcher, J. Solar Energy Eng, 123 143 (2001). 99. A. Yogev, Quantum Processes for Solar Energy Conversion, Weizmann Sun Symp. Proc., Rehovot, Israel, 1996. 100. R. Kribus, J. Doron, P. Rubin, J. Karni, R. Reuven, S. Duchan, and T. Tragan, A multistage solar receiver, Solar Energy, 67 3–11 (1999).
Slide 137: Thermochemical and Thermal/Photo Hybrid Solar Water Splitting 121 101. E. Segal and M. Epstein, The optics of the solar tower reflector, Solar Energy, 69 229– 241 (2001). 102. B. Misch, A. Firus, and G. Brunner, An alternative method of oxidizing aqueous waste in supercritical water: oxygen supply by means of electrolysis, J. Supercritical Fluids, 17 227–237 (2000). 103. O. Yamamoto, Solid oxide fuel cells: fundamental aspects and prospects, Electrochimica Acta, 45 2423–35 (2000). 104. D. Kusunoki, Y. Kikuoka, V. Yanagi, K. Kugimiya, M. Yoshino, M. Tokura, K. Watanabe H. Miyamoto, S. Ueda, M. Sumi, and S. Tokunaga, Development of Mitsubishi planar reversible cell - Fundamental test on hydrogen-utilized electric power storage system, Int. J. Hydrogen Energy, 20 831–834 (1995). 105. K. Eguchi , T. Hatagishi, and H. Arai, Power generation and steam electrolysis characteristics of an electrochemical cell with a zirconia- or ceria-based electrolyte, Solid State Ionics, 86-8 1245–1249 (1996).
Slide 138: This page intentionally blank
Slide 139: 6 Molecular Approaches to Photochemical Splitting of Water Frederick M. MacDonnell University of Texas at Arlington, Arlington, TX 1 Scope This chapter focuses on the progress and challenges in the field of photocatalysis as applied towards the water splitting reaction (Eq. 1.). More specifically, homogeneous molecule-based systems that mimic the natural photosynthetic system are examined for their potential to drive reaction 1. A number of molecules, including porphyrins, metalloporphyrins and phthalocyanines,1-7 transition metal complexes of Ru, Os, Re, Rh, Pt, Cu,8-11 and acridine and flavin derivatives,12-14 have been examined as the chormophores and sensitizers for light-driven processes. 2 H2O O2 + 2 H2 ΔG = +474 kJ/mol = 4.92 eV (1) Of these sensitizers, Ru(II)-polypyridyl complexes are among the best explored due to their excellent chemical stability, broad absorption spectrum in the visible and favorable photophysical and redox properties. In this chapter, progress in the use of ruthenium polypyridyl-based molecular and supramolecular assemblies as sensitizers for 'artificial photosynthesis' is reviewed. Most of the comprehensive reviews on the application of such complexes towards artificial photosynthesis are over a decade old,8,15,16 excepting an recent review by T. J. Meyers and co-workers17 which was part of a 'Forum on Solar and Renewable Energy' thematic issue of Inorganic Chemistry (Oct 3, 2005).18 Included in this thematic issue were additional reviews by G. J. Meyer19 and M. Gratzel20 on the application of ruthenium polypyridyl complexes as sensitizers for photovoltaics based on semiconducting nanoparticles such as TiO2 which is a topic covered in other chapters of this book. Included in this issue is also a review of ruthenium polypyridyl complexes sensitizers for O2 evolution in heterogeneous systems.21 The focus of this chapter is largely on advances in homogeneous molecular systems that can be applied to the water-splitting problem.
Slide 140: 124 Frederick M. MacDonnell 2 Fundamental Principles In photosynthesis and also as elaborated further in Chapter 8, nature has developed a molecular-based system able to capture light and transiently store it as a reducing potential (Eq. 2) and as high energy molecules (Eq. 3).26 This energy is either quickly used or stored in the form of reduced CO2 NADP+ + 2e- + H+ ADP + Pi NADPH (2) (3) (4) ATP + H2O C6H12O6 + 6 O2 6 CO2 + 6 H2O products (e.g., glucose, see Eq. 4) which are more stable forms of stored energy. Water is the ultimate source of electrons for reactions 2 and 4 and dioxygen is the byproduct that is lost to the atmosphere. The development of artificial photosynthetic systems that would mimic the natural process, at least in basic function, is a challenging yet realistic chemical problem with obvious long-term benefits for mankind. As seen in Eq. 1, the water-splitting reaction has an overall energy requirement of 4.92 eV per O2 molecule formed (or +474.7 kJ/mol O2 formed). The most abundant solar radiation to strike the earth's surface falls in the visible range (750–400 nm) and fortunately, these photons are energetic enough (1.65–3.1 eV)27 so that as little as two photons are required to drive this process thermodynamically. When broken down into redox half-reactions (5 and 6), the multi-electron nature of reaction 1 is readily apparent. 2 H+ + 2 e2 H2O H2 + 79.9 kJ/ mol H2, pH 7 + 314.9 kJ/ mol O2, pH 7 (5) (6) O2 + 4 H+ + 4 e- These two reactions are often referred to as the hydrogen evolving reaction (HER) and oxygen evolving reaction (OER), respectively. The problem of driving these two half-reactions with light is two-fold. First, water does not absorb light in the visible and a chromophore is needed to capture and concentrate the solar energy. Second, the absorption of a photon by a chromophore is typically associated with the excitation of a single electron in the molecule. Even if this electron has sufficient reducing potential for reaction 5, the electron stoichiometry is not met. Similar arguments apply to the holes generated and their ability to drive reaction 6. Nature has addressed these problems by developing enzymes proficient at stepwise storage of multiple redox equivalents (electrons or holes) until the appropriate redox stoichiometry is met. For example, the oxygen-evolving center (OEC) in photosystem II is composed of a tetramanganese cofactor that does not evolve O2 until 4 electrons have been removed.28 In general, we observe that the transformation of many small molecule substrates into desirable products, such as H2O to O2 and H2, CO2 to C6H12O6, N2 to NH3, etc. are multi-electron processes and require not only the appropriate driving force but cofactors that enable the appropriate redox stoichiometry to be met. Artificial photosynthetic systems will similarly require
Slide 141: Molecular Approaches to Photochemical Water Splitting 125 entities/catalysts capable of driving multi-electron transfer (MET) reactions and proton-coupled electron transfer (PCET) reactions.29-31 3 Nature's Photosynthetic Machinery Natural photosystems have been extensively studied and the machinery of photosynthesis found to be highly modular, organized on multiple scale levels and compartmentalized.32 Importantly, all natural photosystems are molecule-based and their function can be understood both at the schematic level and, in many cases, at the molecular-level. Highly specialized components interact in controlled manner to ultimately deliver a product (reduction equivalents) and to effectively deal with waste (oxidation equivalents). The bacterial photosystem functions without dioxygen production which simplifies the task at hand. Namely, electrons are obtained from more easily oxidized terminal electron donors such as H2S instead of water. Nonetheless, the basic design needed to transform solar energy into stored chemical energy is present. The protein subunits and cofactors that comprise the photosystem in purple bacteria, such as Rhodobacter (Rb.) sphaeroides and Rhodopseudomonas (Rps.) viridis,33 are shown schematically in Fig. 1 which is based on a crystal structure of this assembly.34 The sensitizer in this photosystem, P865, is a symmetric bacteriochlorophyll dimer (labeled DM and DL in Fig. 1) which has a strong absorption at 865 nm corresponding to 1.43 eV. In the trans-cellular membrane assembly, the photoexcited P865* initiates charge separation by electron transfer down just one side of the photosynthetic assembly (the L side) as indicated by the arrows in Fig. 1. The reason for this asymmetry in electron transfer is still unclear, however, it is clear that electron transfer through a series of acceptor units: BL to φL to QA and finally QB, leads of a charge separated complex with very slow (~ 1 sec) back rates. The P865+ cation is a modest oxidant, (Ered = +0.45 V, and is reduced to the starting state by external reductants such as H2S. The reducing potential stored in QB is utilized in the cell (localized on the cytoplasmic side of the photosynthetic complex) to convert NADP+ to NADPH. Thus, the net chemical reaction is: H2S + NAD+ S + NADH + H+ (7) The structure and function of this bacterial photosystem reveals important principles for the design of artificial photosystems. First, the sensitizer needs to be positioned close to secondary acceptors and donors which themselves are spatially isolated from each other such that photoexcitation leads to rapid spatial separation of the electron-hole pair. Second, compartmentalization of the photosynthetic assembly is likely to be necessary so as to prevent wasteful back reactions. For water-splitting, a system in which H2 and O2 are generated in separate compartments would have both safety and efficiency advantages. In green plants and certain algae, the photosynthetic machinery is elaborated over those found in purple bacteria and is now able to reach the high potentials needed to oxidize water to dioxygen. This oxygenic photosystem is comprised of two photosynthetic reaction centers (sensitizer assemblies), photosystem I (PS I) and
Slide 142: 126 Frederick M. MacDonnell Fig. 1. Schematic of photosystem in purple bacteria photosystem II (PS II) and thus requires more photons to drive the overall reaction.While the details have changed upon moving from the bacterial photosystem to PS I, the principle remains that PS I is unable to reach the potentials required to oxidize water. Figure 2 shows the classic Z scheme, first proposed in 1960 by Hill and Bendall35 and subsequently elaborated many times,36-38 which indicates how the two photosystems work together. The two sensitizers or reaction centers for PS I and PS II are again chlorophyll dimers such as found in bacteria, however these chlorophylls absorb at higher energies, 700 and 680 nm, and transiently store more energy. The special pairs for PSI and PSII are denoted P700 and P680, respectively, and excitation of these reaction centers either by direct light absorption or by energy-transfer from an antenna assembly results in charge separation. A schematic view of the cofactor arrangement in PSII is shown in Fig. 3.39 The reducing potential of the P680* complex (~ –0.58 V) is used to shuttle an electron down a chain of redox acceptors in the PS II complex (pheotyphytin (phe/phe– at –0.42 V) to quinone A (QA/ QA- at –0.08 V) to quinone B (QB/ QB– ))40 and onto an
Slide 143: Molecular Approaches to Photochemical Water Splitting 127 Fig. 2. The classic Z scheme for PSI and PSII. external protein acceptor, plastoquinone. Reduced plastoquinone feeds another redox chain in which the reducing potential of these electrons are used to translocate protons across the membrane and thus generate a proton gradient to be used by the cell to generate ATP from ADP.41 The oxidizing potential of the P680+ cation is estimated at +1.25 V40,42 which is highly oxidizing and rapidly oxidizes the nearby tyrosine Z (Yz/Yz+ at +1.21 V). Ten microseconds after excitation, the electron-hole pair is mostly localized as [Yz+/QA-]. The tyrosine radical cation is positioned near the tetranuclear manganese oxygen evolving center (OEC) and possibly even coordinates to the Ca2+ site based on recent X-ray structural data.43 Four sequential excitations (the 'S-state cycle') removes four electrons from the OEC with the last oxidation concommitent with the evolution of O2 and regeneration of the starting state of the OEC.44 The structure of the OEC in PSII has recently been determined at resolutions as low as 3.2 A43,45,46 which is where the structural 'cubane' model43 shown in Figure 3 originates. The structure must be viewed with caution as there is some criticism of the co-factor structure due to the resolution of the data47 and the possibility of X-ray damage to the OEC cofactor during data collection.48 Nonetheless, this cubane structure incorporates most of the spectroscopic and compositional data known for PSII and serves as a useful model for the active site of oxygen evolution. Photosystem I forms the second light-absorbing component in the Z scheme for green plants and algae and like PS II, the structure of the protein complexes has been determined by X-ray crystallography.34,49-51 PS I is similar in function to that of the purple bacterial photosystem in that the oxidation potential generated is modest (P700+/P700 at ~ +0.5 V), however, the primary function of this photosystem is to
Slide 144: 128 Frederick M. MacDonnell Fig. 3. Schematic view of co-factor arrangement in PS II. generate a reducing potential for the generation of NADPH. The electrons for this process come from plastocyanin that is the terminal reduced product from PS II. Both PSI and PSII are membrane proteins which span the thylakoid membranes in chloroplasts.52 As indicated in Fig. 2, the luminal side of the membrane is where water is oxidized to O2 and protons are generated. The energy dependent translocation of these protons to the stromal side occurs as the electrons flow down the redox gradient indicated in the Z scheme. Ultimately, the protons are either consumed in the production of NADPH on the stromal side or their energy (in the form of a proton gradient) is used to produce ATP.38 The membrane is also the site in which the antenna complex, consisting of hundreds of chlorophylls are organized such that light excitation of any chlorophyll quickly results in energy transfer to the special pair in PS I or PS II for charge-separation.32
Slide 145: Molecular Approaches to Photochemical Water Splitting 129 Fig. 4. Schematic of artificial photosynthetic system. 4 Design of Artificial Photosystems From the natural photosystems, it is clear that a number of essential molecular components must be present and be organized in a supramolecular fashion such that the photogenerated electron-hole pair is quickly and efficiently separated and their potential energy delivered to functional co-catalysts. An artificial water-splitting photosystem will also require the controlled interaction of multiple subunits.17 A schematic of such a system containing the minimum number of necessary components is given in Fig. 4. An efficient and robust sensitizer absorbs photons and generates long-lived electron-hole pairs that are coupled to appropriate multi-electron co-catalysts for oxygen and hydrogen evolution. As mentioned previously, this chapter will focus largely on the use of Ru(II)polypyridyl complexes as sensitizers to drive the water-splitting reaction. While we will examine the various properties of 'Rubpy' complexes in more depth, an initial analysis of the energetic of water splitting with [Ru(bpy)3]2+ will set the stage for the subsequent material. As seen in the modified Latimer diagram,17,53 shown in Fig. 5, the photoexcited complex, denoted [Ru(bpy)3]2+* or more simply Ru2+*, is both a good oxidant and a good reductant. Electron transfer either to or from the excited state complex traps the ruthenium complex as [Ru(bpy)3]+ or [Ru(bpy)3]3+, respectively.54 These photoproducts are potent reductants ([Ru(bpy)3]+) and oxidants ([Ru(bpy)3]3+) which can drive subsequent redox reactions. These various species can be used to drive a catalytic reaction such as that shown in Eq. (8), A+D A- + D+ (8) by either an oxidative or reductive quenching pathway. These two catalytic manifolds are shown in Fig. 6, where A and D are generic acceptor and donor molecules. While the two manifolds can store the same amount of energy, they encompass different potential ranges and therefore differ with respect to the types of acceptor and donor molecules that may be used or the driving force that may be applied to any
Slide 146: 130 Frederick M. MacDonnell Fig. 5. Latimer diagram of the excited- and ground-state redox processes for [Ru(bpy)3]2+. Potentials given in V vs. NHE reference electrode. given step in the cycle. If we analyze the energetics of the water-splitting reaction within these two manifolds, as shown in reactions 9 to 17, we can see that by designing a system properly we can deliver more driving force where it is most needed. All the reactions are normalized to a 4-photon, 4-electron process as presumably (though not necessarily) would be required and a pH of 7.0 is assumed. The first three photoreactions, 9–11, summarize the capture of light energy (at ~ 450 nm or 2.75 eV) to form the excited singlet molecule [Ru(bpy)3]2+**. This molecule rapidly loses some energy upon intersystem-crossing to the triplet state and relaxation to vibronically cooled E00 triplet state, [Ru(bpy)3]2+*.17 The useable energy of this species is ~ 2.1 eV (~ 200 kJ/mol). Light-capture: 4Ru2+ (λ ~ 450 nm) 4 Ru2+** 4 Ru2+* 4 Ru2+* 4 Ru+ + O2 + 4 H+ 4 Ru2+ + 2 H2 O2 + 2 H2 + 4 Ru2+ 4 Ru3+ + 2 H2 4 Ru2+ + O2 + 4 H+ O2 + 2 H2 + 4 Ru2+ 4 Ru2+** ΔE = +1060 kJ/mol ΔE = –250 kJ/mol ΔE = +810 kJ/mol ΔG = –76 kJ/mol ΔG = –260 kJ/mol ΔG = –336 kJ/mol ΔG = –62 kJ/mol ΔG = –274 kJ/mol ΔG = –336 kJ/mol (9) (10) (11) (12) (13) (14) (15) (16) (17) 4 Ru2+ + 4 hν 4 Ru2+* + 2H2O 4 Ru+ + 4H+ Reductive quenching water-splitting path: 2 H2O + 4 Ru2+* 4 Ru2+* + 4 H+ 4 Ru3+ + 2 H2O 2 H2O + 4 Ru2+* Oxidative quenching water-splitting path:
Slide 147: Molecular Approaches to Photochemical Water Splitting 131 Fig. 6. Photoreactions based on reductive quenching of the photoexcited [Ru(bpy)3]2+*. As can be seen both pathways are exothermic by –336 kJ/mol, however the exothermicity of the HER, Eq. (13), can be favored in the reductive quenching path at the expense of the OER, Eq. (12). The reverse is true for the oxidative quenching pathway as may be expected. Of course, the solution pH is a major factor that can be tuned to favor hydrogen reduction or water oxidation. Application of the Nernst equation for the two manifolds yields expressions 18–21: reductive quenching pathway: ΔGOER = 84 − 22.85(pH) (kJ/mol) (kJ/mol) (18) (19) ΔGHER = −420 + 22.85(pH ) oxidative quenching pathway: ΔGOER = − 116 − 22.85(pH) ΔGHER = −220 + 22.85(pH ) (kJ/mol) (kJ/mol) (20) (21) From these expressions, we find the OER and HER are isoenergetic at pH 11 and 2.3 for the reductive and oxidative quenching pathways, respectively. More importantly, the pH range in which water splitting can occur can be determined, assuming the two half-reactions are not energetically coupled, i.e., the free energy of HER is not used to help drive the OER (or vice versa). Following the reductive quenching pathway, both the OER and HER are only spontaneous above pH 3.7. Conversely, the oxidative quenching pathway requires a pH below 9.6 for both to be spontaneous. Thus the quenching pathway plays an important role in determining optimum pH conditions. These energies and pH ranges are specific for [Ru(bpy)3]2+assuming a 4 pho-
Slide 148: 132 Frederick M. MacDonnell Table 1. Room temperature electrochemical and photophysical data for homoleptic rupolypyridyl complexes. Complex [Ru(bpz)3] 2+ E(Ru3+/2+)a 2.10 1.64 1.53 1.51 1.50 1.10 E(Ru2+/+) –0.56 –0.83 –1.09 –1.11 –1.05 –1.37 E(Ru3+/2+*) –0.06 –0.42 –0.57 –0.68 –0.58 –0.94 E(Ru2+*/ +) 1.60 1.23 1.02 1.08 1.03 0.69 λem, nm (τ, ns)b 627 (720) 632 (480) 610 (890) 604 (460) (<250 ps) 631 (335) φem 0.024 [Ru(4,4'-Cl2bpy)]2+ [Ru(bpy)3]2+ [Ru(phen)3]2+ [Ru(terpy)2]2+ [Ru(4,4'- Me2bpy)3]2+ d – 0.073 0.028 0.014 0.84 –1.07 700 (130)c 0.010c [Ru(4,4'(Et2N)2bpy)3]2+ a As measured in acetonitrile unless otherwise noted. Electrochemical measurements were done in the presence of 0.1-M Bun4NPF6 or 0.1-M Bun4NClO4. All potentials are quoted relative to NHE using the correction factor of +0.247 V for SCE and +0.225 for Ag/Ag+ where required. b degassed. c in MeOH/EtOH. d in H2O. bpz = 2,2'-bipyrazine. Data obtained from Ref. 54. ton process but are tunable by modifying the complex structure. This is most commonly achieved by modification of the diimine ligand with electron withdrawing or donating groups. As seen in Table 1, ligand substitution can shift the relevant redox potentials by nearly a volt in either direction, however any gain in the oxidizing potential is offset by a loss in reducing potential or vice-versa. An extensive listing of the ground and excited state redox data for Ru(II) and Os(II) diimine complexes can be found in a review by Balzani and coworkers.54 Also, Vlcek et. al. have reported that in many cases the excited state energies can be predicted simply from ligand redox parameters.55 This above analysis assumes that the free-energies of the HER and OER cannot be coupled which need not be true. As we observe in the natural photosystems, the free energy of one reaction may be used to change local pH by the translocation of protons across a membrane. The resulting pH change could easily be applied to increase the exothermicity of both O2 and H2 evolution in a straightforward manner and would be highly desirable. However, such a system requires an additional layer of sophistication in the design of the photosynthetic machinery. That said, artificial photodriven proton pumps are a reality as Gust, Moore and coworkers have shown that a membrane-bound carotenoid-porphyrin-quinone triad can actively transport protons against a gradient upon visible irradiation.56 There are additional aspects that must be considered for artificial or man-made photosystems. Nature's photosynthetic machinery is easily repaired or regenerated by the living organism. Artificial systems are likely to be harder to repair and therefore should be constructed of the most robust components available. The various
Slide 149: Molecular Approaches to Photochemical Water Splitting 133 components are likely to be subject to relatively harsh environmental conditions including high photon flux, thermal stress, aqueous solutions with variable pH and ionic strengths and oxygen. Only a few molecular species can meet all, or even most, of these conditions and while it is feasible to encapsulate or otherwise protect less stable molecular components, it is preferable that they be as robust as possible. No doubt the popularity of metalloporhyrins and ruthenium polypyridyl complexes as sensitizers is due, in large part, to their exceptional stability and favorable photophysical properties. Below we review the properties of ruthenium polypyridyl complexes that warrant all this attention, the current state of the art in 'wiring-up' these sensitizer to efficiently undergo energy or electron-transfer over large distances, the current ability to store multiple redox equivalents, and our ability to generate longlived charge-separated states. Finally the development of efficient co-catalysts for water oxidation (OER) and hydrogen evolution (HER) are reviewed as are aspects of linking these catalysts with the sensitizer as indicated in Fig. 4. 5 The Ideal Sensitizer: Does Rubpy Come Close? Rubpy is an often used shorthand for the well-studied [Ru(bpy)3]2+ complex cation and is also used more loosely to refer to any Ru(II) polypyridyl complex with similar properties. As shown in Fig. 7, the three best studied members of the Rubpy family are the parent complex, [Ru(bpy)3]2+, the phenanthroline analogue, [Ru(phen)3]2+, and the bis terpy complex, [Ru(terpy)3]2+, which has less optimal photophysical properties but has a number of synthetic and stereochemical advantages. The synthesis, stability and photophysics of Rubpy complexes has been reviewed many times17,54,57 and the highlights of these properties as related to artificial photosynthetic assemblies is summarized here. 5.1 Stability The stability of Rubpy complexes is a result of a number of factors including: 1. the large electronic stabilization found in these octahedral low spin d6 second row transition metal complexes (while the symmetry of the complexes in Fig. 7 is not Oh, to a first approximation an octahedral coordination environment adequately describes many of the electronic properties); the considerable covalent character of the bond as reflected in the similar electronegativities for Ru (χ 2.2) and N (χ 3.0), and the multiple-bond character associated with π bond formation between the full Ru t2g orbitals and the π-accepting orbitals on the polypyridyl ligands. 2. 3. These factors and the associated kinetic inertness of the low spin d6 electronic configuration make these complexes exceptionally stable with respect to ligand loss or decomposition.
Slide 150: 134 Frederick M. MacDonnell Fig. 7. Chemical structures of three commonly used Ru(II) polypyridyl complexes. [Ru(bpy)3]2+ and [Ru(phen)3]2+ are modestly sensitive to photochemical degradation via photolabilization of the ligands upon excitation of the MLCT band.58,59 The 3 MLCT state that is utilized in most light-to-energy conversion schemes, is populated with essentially a 100% quantum yield.60 At room temperature, however, thermal population of ruthenium-based d-d states, only ~ 43 kJ/mol higher in energy than the 3 MLCT state, can occur resulting in decomposition via ligand dissociation.58,59 Quantum yields for photochemical decomposition are generally small (φ ~ 0.04) and highly dependent on the solvent conditions and temperature.58 Aqueous solutions of varying ionic strength and acidity show quantum decomposition yields between 0.3% to 0.001% at 343 K.59 Dichloro methane solutions are considerably more reactive with decomposition quantum yields between 6 and 1 % at 298 K with the nature of the counterion playing an important role as small coordinating anions, such as halides, increasing the decomposition yield.58 In dichloromethane, the increased decomposition is attributed to more significant ion pairing in this non-polar solvent, which leads to more effective trapping of the 5-coordinate intermediate, and stabilization of the product by anation. Photoracemization of optically pure Δ-[Ru(bpy)3]2+ in aqueous solutions was also observed with a similar activation energy to the photodecomposition reaction and is presumed to occur via the same 5-coordinate intermediate.61 Photochemical decomposition of the ruthenium polypyridyl chromophore has not been significant obstacle to their use in energy conversion schemes because in most cases the 3MLCT state is rapidly quenched and the resulting photoproduct is stable with respect to ligand dissociation. For example, addition of the oxidative quencher, methylviologen (MV2+), greatly reduces the photodecomposition of [Ru(bpy)3]2+* as the MV2+ rapidly reacts with the 3MLCT state by electron transfer yielding [Ru(bpy)3]3+ and MV+.62 [Ru(bpy)3]3+ and [Ru(bpy)3]+ are commonly produced as photoproducts in the photochemistry of [Ru(bpy)3]2+ and, as noted, are potent oxidants and reductants, respectively. Barring a redox reaction, these complexes show good stability towards both substitution and racemization,63 however Ru(III) complex is not indefinitely stable in aqueous solutions. Over time spontaneous reduction of [Ru(bpy)3]3+ to [Ru(bpy)3]2+ is observed with some sacrificial degradation of the bpy ligands