iea hydrogen implementing agreement task 26: advanced ......a2 database development c2 advanced...

-

IEA Hydrogen Implementing Agreement

Task 26: Advanced Materials for Waterphotolysis

Final Report

Operating Agent: Dr. Eric L. Miller

U.S Department of Energy

Washington D.C., USA

March 2013

ISBN 978-0-9815041-6-2

2

IEA Hydrogen Implementing Agreement

Task 26: Advanced Materials for Waterphotolysis

TABLE OF CONTENTS

EXECUTIVE SUMMARY ...................................................................................................... 3

INTRODUCTION .................................................................................................................. 4

TASK DESCRIPTION ............................................................................................................ 4

Sub Task -M: PEC Focus Materials ................................................................................. 6

Sub Task -C: PEC Materials / Interface Characterizations ......................................... 8

Sub Task -S: PEC Standardized Materials Testing / Screening ................................. 8

Sub Task -T: PEC Materials Theory ................................................................................ 9

Sub Task -D: New PEC Materials Discovery ................................................................ 9

Sub Task -E: Techno-Economics Analyses .................................................................... 9

Sub Task -A: PEC Materials R&D Administration .................................................... 10

PRIMARY TASK PARTICIPANTS .................................................................................... 10

EFFECTIVENESS OF TASK PARTICIPATION .............................................................. 11

IMPORTANT ACCOMPISHMENTS ................................................................................ 11

MAJOR COLLABORATIVE PRODUCTS ........................................................................ 12

APPENDIX A: PEC Materials White Papers ............................................................... 13

APPENDIX B: Technoeconomic Analysis of PEC Hydrogen Production .............. 13

APPENDIX C: PEC H2O Splitting Standards, Experimental Methods, and

Protocols ............................................................................................................................ 13

APPENDIX D: 2010 Book ON SOLAR HYDROGEN& NANOTECHNOLOGY ... 13

APPENDIX E: 2012 Book PHOTOELECTROCHEMICAL HYDROGEN

PRODUCTION ................................................................................................................. 13

REFERENCES ....................................................................................................................... 14

ACKNOWLEDGEMENTS .................................................................................................. 14

ISBN 978-0-9815041-6-2

3

IEA Hydrogen Implementing Agreement Task 26: Advanced Materials for Waterphotolysis

Final Report Dr. Eric L. Miller, Operating Agent

EXECUTIVE SUMMARY Photoelectrochemical (PEC) hydrogen production, using sunlight to directly split water, is one of the paramount enabling technologies for a future where hydrogen is widely deployed as an energy carrier. The “traditional” semiconductor-based PEC material systems studied to date, in particular the simple metal oxides such as TiO2, WO3 and Fe2O3, however, have been unable to meet all the performance, durability and cost requirements for practical hydrogen production. Technology-enabling advances have been needed in the development of new and improved materials systems. Toward this end, the IEA Hydrogen Implementation Agreement Task 26, working in close conjunction with the U.S. Department of Energy’s (DOE) “Working Group on PEC Hydrogen Production” has brought together international experts in analysis, theory, synthesis and characterization from the academic, industry and national laboratory research sectors across the world, with exciting and important results on several fronts. A critical initial emphasis of this joint effort was the establishment of “Sub Tasks” focused on advancing critical PEC materials theory, synthesis and characterization capabilities for application in the research and development of broad-ranging materials systems of promise. Such material classes have included complex metal-oxide, -sulfide, and -nitride compounds; amorphous silicon alloys; III-V semiconductors; and the copper chalcopyrites. Key supporting activities in Task 26 have included establishing standardized testing and screening protocols for candidate PEC materials systems and performing techno-economic analyses of competing PEC production systems based on the new materials being developed. Task 26 was initiated with the IEA-HIA Executive Committee Meeting held at Brisbane Australia in June 2008. Over the course of its tenure, the Task’s “Program of Work” was expanded and refined to better foster international participation in activities initiated by the US DOE PEC Working Group. Numerous “Experts Meetings” were held in conjunction with high-impact international conferences related to photoelectrochemical conversion processes (typically also in conjunction with the DOE Working Group Meetings). These important joint meetings, dubbed “Hu’a Iki” from the Hawaiian expression for “tiny bubbles” (i.e., of hydrogen), served as a successful platform for coordinating and developing the international collaborative products from Task 26. The major products from this task, which are expected to benefit the PEC research community at large, have included: “White Papers” on PEC focus materials which document the current understanding of the challenges and potential of each materials class, and include the most up-to-date references for each; a technoeconomic analysis report on PEC hydrogen production projecting technical feasibility of different feasible approaches; a published document on standard methods in the characterization and reporting of PEC materials; and two seminal books on PEC hydrogen production comprised of chapter contributions from the international experts in this field. These Task 26 products are documented in this report.

4

INTRODUCTION

Photoelectrochemical (PEC) hydrogen production using sunlight to directly split water is one of the paramount enabling technologies for a future where hydrogen is widely deployed as an energy carrier. The “traditional” semiconductor-based PEC material systems studied to date, in particular the simple metal oxides such as TiO2, WO3 and Fe2O3, however, have been unable to meet all the performance, durability and cost requirements for practical hydrogen production. Technology-enabling advances are needed in the development of new, advanced materials systems. Toward this end, the IEA Hydrogen Implementation Agreement Task 26, working in close conjunction with the U.S. Department of Energy’s (DOE) “Working Group on PEC Hydrogen Production” brought together international experts in analysis, theory, synthesis and characterization from the academic, industry and national laboratory research sectors across the world, with exciting and important results on several fronts, as described in this report. It is recognized that PEC hydrogen production is at relatively early stage of development. Though it is one of the most promising approaches for practical conversion of sunlight to chemical energy, and much progress has been made in recent years, the technology readiness (TRL) level of various PEC technologies remains at TRL 1 and TRL 2. The specific technical goal of this Task 26 has been the research and development of new semiconductor materials for stable and efficient PEC hydrogen production systems. In order to achieve this goal Task 26 has formulated a comprehensive “Task” structure, as described below.

TASK DESCRIPTION

The main goal of the IEA-HIA Task 26 has been to seamlessly extend the excellent R&D efforts made under previous PEC Annexes -14 and -20 toward practical material and systems solutions for water-photolysis (e.g., solar water splitting). In this continued research, photon conversion efficiency and durability have been judged as the main measures of success in the development of new PEC materials. The four overarching objectives of the Task-26 program comprised:

– Intensification of international collaboration, making use of extended fields of expertise in areas of materials theory, synthesis and characterization, as well as data and data-base management;

– Advancement of photoelectrode materials science, particularly addressing the discovery of new practical materials, with bulk and surface properties specifically engineered to meet the requirements for efficient and stable PEC water splitting;

– Demonstration of stable and efficient water splitting in the leading materials systems, using standardized performance characterizations and round-robin testing procedures;

– Promotion of photolysis of water through publications, education and outreach program.

The specific technical goal of this Task has been the research and development of new semiconductor materials for stable and efficient PEC hydrogen production systems. Key to

5

the success of PEC, the candidate materials have to be functional (1) as a photoactive layer, absorbing a significant fraction of the incident light; (2) as a photoelectrochemical junction with the electrolyte; and (3) as a facilitator of the gas evolution reaction (either hydrogen or oxygen, depending on the n- or p-type nature of the semiconductor, in conjunction with the specific integrated device configuration). The requirements on the material include adequate light absorption over the solar spectrum, high carrier collection efficiency, stability in suitable aqueous electrolytes, and favorable kinetics for the electrode reaction. As candidate materials with suitable properties emerge, additional requirements for the photoelectrode semiconductor device integration become increasingly important, such as process compatibility of the complete multi-junctions devices, as well as long durability and low material cost. Task 26 has promoted and coordinated collaborative efforts among the international research participants to approach this daunting problem. The collaborative approach has been aimed at integrating and coordinating the available state-of-the-art theoretical, synthesis and characterization techniques to identify and develop the most promising materials classes to meet the PEC challenges in efficiency, stability and cost. From the application of density-functional theory to calculate band-structures and effects of introduced “impurities” on valence band maximum and conduction band minimum positions; through the use of diverse synthesis techniques, including combinatorial methods, to create tailored materials; and by employment of microstructural, electron spectroscopic, and electrochemical characterization techniques, a comprehensive picture of the materials properties and resulting performance can be developed. Through coordinated feedback between the theoretical, synthesis and characterization efforts, classes of “focus materials” deemed of particular interest for PEC applications by Task 26 participants have emerged for continued investigations. These “focus materials” classes have included tungsten-based, iron-based, silicon-based, III-V-based and copper chalcopyrite-based compounds (as well as others delineated in the following sections). In order to achieve its goals, Task 26 formulated a comprehensive “Sub Task” structure, as illustrated below in Figure 1, serving as the central organizational framework for Task activities. Task 26 relied heavily on activities in the individual Sub Tasks to coordinate collaborative efforts in international PEC R&D and facilitate the PEC materials advancement process.

6

Figure 1: Sub Task definitions for the IEA-HIA Task-26

Sub Task -M: PEC Focus Materials

The primary objective of the materials research efforts has been the development improved materials which meet photocurrent and durability goals, and which are compatible with device fabrication. The most promising candidate materials were identified, with the short-term goal of demonstrating laboratory-scale water-splitting devices, and with a long-term goal of transferring the fabrication processes toward the commercial scale. Significant R&D has continued on many photo-anode and photo-cathode materials fabricated via traditional routes such as PVD, CVD, spray pyrolysis, electrodeposition, etc. The current focus materials classes under investigation by Task 26 participants has included metal oxide semiconductors such as titania and tungsten-based compounds (metal and mixed-metal oxides, oxy-nitrides, oxy-sulfides, etc.), silicon-based compounds (such as silicon carbide and silicon nitride), copper-chalcopyrite compounds (including CIGS, CGS, etc.), III-V compounds, nano-structured MoS2 and WS2, among others. Current activities focusing on specific materials classes have included: Activity M1: Tungsten-oxide and Related Modified Compounds Tungsten oxide, particularly in thin-film and nano-particle forms, has been a workhorse in photoelectrochemical applications for years. It is inexpensive and stable, but its high bandgap (~2.6 eV) is limiting to PEC performance. Photocurrent densities of approximately 3 mA/cm2 have been achieved, with STH efficiencies over 3% in tandem configurations. To break the performance barrier, current research is focused on reducing bandgap through ion incorporation into the WO3 structure, and further integration in multi-junction devices.

A1 Research Coordination C1 Standard Measurements

A2 Database Development C2 Advanced Solid-State

A3 Reporting C3 Advanced In-Situ Methods

M1 Modified WO3 Films D1 Combinatorial Discovery

M2 Modified Fe2O3 Films D2 Auxiliary Materials

M3 Silicon Alloy Films

M4 Copper Chalcopyrite Films

M5 Nanostructured Sulfides

M6 III-V Semiconductors E1 System Configurations

M7 Modified TiO2 Films E2 Baseline TE Analyses

M8 Vanadate Semiconductor

M9 others

T1 Solid-State Calculations

T2 Interface Calculations

S1 Standard Testing Protocols

S2 Standard PEC Screening

S3 Standards Reviewing

A: PEC R&D Administration

M: Focus Materials

S: Standardized Testing/Screening

C: Materials Characterizations

D: Materials Discovery

E: Techo-Economics Analyses

T: Materials Theory

7

Activity M2: Iron-oxide and Related Modified Compounds Iron-oxide is abundant, stable, inexpensive and has a near-ideal bandgap (~2 .1 eV) for PEC applications. Unfortunately, it’s poor absorption, photo-carrier lifetime and transport properties have been prohibitive to practical water-splitting. Current research to overcome these barriers have been encouraging, with recent progress in thin films and nano-structured materials. Iron oxide in tandem configurations may also be of interest for practical solar water-splitting. Activity M3: Amorphous Silicon Compounds Amorphous silicon compounds have recently demonstrated interesting performances in PEC applications. The progress of this material class in PEC applications has benefitted from decades of research in the PV community. Technical barriers remain in PEC stability and interface properties, and electrolyte and surface modification studies could help overcome these barriers. With material and interface improvements, monolithically fabricated multi-junction devices using amorphous silicon compound films have practical appeal for PEC water splitting. Activity M4: Copper Chalcopyrite Compounds Copper chalcopyrite thin films are among the best absorbers of solar energy. As a result, chalcopyrite alloys formed with copper and gallium, indium, sulfur and selenium have been widely characterized in the PV world. A great advantage of this material class for PEC applications is the bandgap tailoring based on composition, with bandgaps ranging from 1.0 eV in CuInSe2 to 1.6 eV in CuGaSe2, and up to 2.43 eV in CuGaS2.. The CuGaSe2 bandgap is attractive for PEC applications. Photo-current densities exceeding 13 mA/cm2 have been demonstrated with this material in biased PEC cells. Stability, surface kinetics and surface energetics remain as barriers, but if research can successfully address these, high STH efficiency could be achievable in low-cost thin-film copper chalcopyrite systems. Activity M5: Tungsten- and Molybdenum- Sulfide Nano-structures As bulk materials, tungsten- and molybdenum-sulfides are excellent hydrogen catalysts, but their bandgaps (below 1.2 eV) are too low for PEC water-splitting. Quantum confinement using nano-structuring, however, can increase the bandgap up to 2.5 eV. Current studies in nanostructured MoS2 are focused on stable synthesis routes and integration of the nano-structures into practical bulk PEC devices. Activity M6: III-V Semiconductor Classes High-quality crystalline semiconductor compounds of gallium, indium, phosphorous and arsenic have been studies for decades. In PEC experiments to date, STH efficiencies between 12 and 16 percent have been demonstrated in GaInP2 /GaAs hybrid tandem photocathodes. High cost and limited lifetime have been barriers to practical PEC hydrogen production, and breakthroughs in synthesis and in surface stabilization are being pursued. Activity M7: Titanium-oxide and Related Modified Compounds Although TiO2 has been the most widely studied semiconductor material system for PEC applications, its wide bandgap has continued to limit its practical application. Methods to modify the band structure and surface properties of this material system are continuing

8

Activity M8: Vanadates and Related Modified Compounds Vanadate material systems, such as BiO2 are being investigated, with interesting recent results. Sub Task -C: PEC Materials / Interface Characterizations

Materials structure and composition play a key role in PEC cells performances. In fact, several studies based on both material and device characterizations have shown that the optimum PEC cells working point can be obtained only with specific material properties which require fine process tuning. Nevertheless, good material intrinsic properties do not guaranty optimal PEC cells performance. The integrated cell performance is closely tied to both physical phenomena localized at the solid-solid and the solid-liquid interfaces. As a result, the choice of materials requires careful consideration of the intrinsic bulk properties as well as the detailed properties. The development of new and effective PEC materials has therefore required experimental studies based on materials, interface and device characterizations. Specific characterization activities under this Sub Task have included: Activity C1: “Standard” Material Characterizations Comprehensive solid-state characterizations of PEC semiconductors have been the focus of this activity. This includes application at Task member facilities of standard and novel techniques, including, for example: (1) surface potential and morphology analyses using Kelvin probe Force Microscopy; (2) Hall measurement of bulk carriers’ mobility; (3) Spectrophotometric measurement of optical absorption; (4) Electron Back Scattered Diffraction and XRD analysis as well as (5) Transmission Electron Microscopy, and others. Activity C2: “Advanced” Materials and Interface Characterizations This materials characterization activity has employed the most advanced microstructural, optoelectronic, and electrochemical characterization techniques available to paint a more comprehensive picture of the materials and interface properties in relation to PEC performance. Example techniques include ex-situ X-ray photoelectron spectroscopy (XPS), ultraviolet photoelectron spectroscopy (UPS), and Auger, Inverse photoemission spectroscopy (IPES). Activity C3: Advanced In-Situ PEC Characterizations The solid-liquid interface plays a key role in PEC hydrogen production and should be characterized with advanced in-situ techniques using immerged tunneling microscopy, Kelvin probe techniques. XPS, UPS, etc. This activity focused on the achieving the technology breakthroughs needed to implement critical in-situ PEC characterizations . Sub Task -S: PEC Standardized Materials Testing / Screening

Development of standardized testing and reporting protocols for evaluating and ultimately screening candidate PEC materials systems is key to the success of PEC R&D. In the past, the lack of standardized conditions and procedures for reporting PEC results has greatly hampered research progress across the board. In this Sub Task, the international PEC

9

Experts were tasked with developing standardized testing and screening protocols and reporting procedures for publication and dissemination of the best data and results. Sub Task -T: PEC Materials Theory

Theoretical predictive guidance and understanding of synthesized materials is viewed as critical to the development of new PEC materials. The insight gained from calculations of relevant semiconductors can augment experimental results, and guide further experimental work depending on favorable or unfavorable outcomes of the theoretical studies. Critical materials properties such as band-gap, band-edge positions, optical absorption, and stability can be understood through calculation of band-structure, band-alignment, optical absorption coefficient, and total energies. Ab-initio methods such as density functional theory (DFT) have been employed previously for such tasks. Less computing-intensive methods such as semi-empirical calculations may also be employed for certain tasks. Another theoretical aspect of PEC materials research is the modeling of semiconductor-electrolyte junctions. One-dimensional models similar to those used in the field of photovoltaics may be developed by incorporating the relevant characteristics of the semiconductor, the electrolyte and the solid/electrolyte interfaces. The aims of such work include the calculation of current-voltage characteristics and other measurable quantities. Comparison of calculated and measured data is used in refining models of photoelectrode device operation. For example, first-principles DFT is known to provide the most accurate prediction and understanding for atomic and electronic structure and optical properties of materials; DFT models have been specifically developed to facilitate R&D of Task 26 focus materials classes. In modeling PEC materials, band-gap, band-edge positions, optical absorption, and stability are all critical issues which can be understood through DFT calculation of band-structure, band-alignment, optical absorption coefficient, and total energies. Sub Task -D: New PEC Materials Discovery

The primary objective of this Sub Task was the discovery of new material systems that can accelerate the state of the art in PEC hydrogen production devices Novel semiconductor, catalyst and other auxiliary materials including contact and interface layer materials have been investigated based on the theory, synthesis and characterization progress in the other Sub Tasks. Sub Task -E: Techno-Economics Analyses

Development and refinement of techno-economic analyses of PEC hydrogen production systems incorporating performance and processing cost feedback from the broader materials R&D efforts are needed. The objective of this task was to provide a basis for evaluating the long-term feasibility of large-scale PEC production technologies in comparison with other renewable approaches. This Sub Task focused on investigating material and device-level requirements of different systems configurations for PEC hydrogen production; for example, system configurations incorporating functionalized semiconductor particles versus functionalized semiconductor films. It also focused on establishing a baseline analysis of semiconductor-based solar hydrogen production systems, focusing on comparisons of the different systems configurations. This has been an important activity for establishing the practicality of PEC hydrogen production vis-à-vis other renewable routes.

10

Sub Task -A: PEC Materials R&D Administration

This critical Sub Task was responsible for organizing and coordinating the materials theory, synthesis and characterization efforts of all of the participating Task 26 members; and for maintaining dynamic communications among international experts in the PEC R&D community.

PRIMARY TASK PARTICIPANTS

Recruitment of International PEC Experts remained in full force during the tenure of Task 26. “Primary Experts” were identified from over 10 countries, and dozens of technology-support “Contributing Experts” were recruited from the participating countries, each with specific expertise in different areas of materials research for PEC Hydrogen Production. A list of the Task 26 “Primary Experts” and “Contributing Experts” is included in Figure 2 below:

Participating Countries and Primary Experts USA Drs. John Turner, Eric McFarland, Thomas Jaramillo, Heli Wang

EU Dr. Michael Graetzel Netherlands Drs. Roel van de Krol and Fatwa Abdi

Switzerland Drs. Artur Braun, Scott Warren and Kevin Sivula Germany Drs. Bjorn Marsen, Marcus Baer and Lothar Weinhardt UK Dr. Upul Wijayantha Poland Dr. Jan Augistinski

Australia Drs. Ian Plumb, Andreas Luzzi and Janus Nowotny

Japan Drs. Kazuhiro Sayama, Kazunari Domen, and Hironori Arakawa

S. Korea Drs. Jae Sung Lee, and Hyunwoong Park International Contributing Experts M. Al-Jassim, Z. Chen, T. Deutsch, H. Dinh, N. Gaillard, R. Garland, K. George, C.

Heske, J. Hu, M. Huda, W. Ingler Jr, B. James, J. Kaneshiro, J. Leisch, N. Lewis, A.

Madan, S. Mahendra, J. Melman, M. Misra, T. Ogitsu, B. Parkinson, C. Pendyala, R.

Perret, E. Selma, K. Sivula, M. Sunkara, H. Wang, S. Warren, Y. Yan, Y. Zhang, S.

Warren, L. Vayssieres, A.Kleiman-Shwarscten, B. Wood and others.

Figure 2: Task 26 Prim ary Experts as w ell as a sam pling of Contributing Experts

11

EFFECTIVENESS OF TASK PARTICIPATION

Central objectives of the IEA-HIA Task 26 have included international outreach and expansion of cooperative and collaborative activities between the US DOE PEC Working Group and the international research community in areas related to PEC materials discovery and development. Combining ever-advancing world-class capabilities in analysis, theory, synthesis and characterization is the surest path to the needed scientific breakthroughs in PEC semiconductor materials. With this in mind, Task 26 over its tenure has continued to expand its international base, and to hold Task meetings in conjunction with major international meetings relevant to photoelectrochemical processes. Synergistic activities among the US DOE Working Group projects, European PEC projects such as “NanoPEC”, and Asian research projects such as those in Japan and Korea has paid great dividends through Task 26 participation. As good examples of successful collaboration, a 2012 book compiled of chapters from PEC Experts edited by Michael Graetzel (Switzerland) and Roel van de Kroel (Netherlands) has been published; Additionally, the 2010 book, On Solar

Hydrogen and Nanotechnology edited by Lionel Vayssieres (Japan), also comprised of PEC Expert contributions, has been well received in the R&D community. Other successful examples of collaborative products related to Task 26 activities have included the “White Papers” on focus PEC materials classes, a technoeconomic analysis report on PEC hydrogen production, and a document of standard procedures and protocols for the characterization and reporting of PEC materials.

IMPORTANT ACCOMPISHMENTS

Over its tenure, the research methodology of the IEA-HIA Task-26 integrating collaborative tasks in PEC materials theory, synthesis, characterization and analysis has paid dividends in terms of technical achievements in PEC materials research and development. The tools in the IEA-HIA Task-26 research arsenal developed in Task A, C, S and T activities have been successfully utilized in effecting advances in PEC materials systems (Task M activities). Major accomplishments resulting from these activities have been documented in Materials “White Papers”, included in this report. Renewable solar hydrogen production via Photoelectrochemical (PEC) water splitting has been successfully demonstrated on the laboratory scale using current materials systems, but challenges remain in meeting all DOE targets in solar-to-hydrogen (STH) conversion efficiency, durability and cost. The research methodology of the IEA-HIA Task-26 integrating collaborative tasks in PEC materials theory, synthesis, characterization and analysis have led to some significant advances in the PEC state of the art over the course of the Task. Important accomplishments have been achieved over a broad spectrum of PEC materials R&D through the duration of Task-26, which are being thoroughly documented in the White Paper updates. Some of the notable achievements have included:

Achievement of new benchmark performance in oxide-based materials, specifically in

iron-oxide based material systems as a result of the EU NanoPEC Project [1].

12

Achievement of new benchmark performance levels in III-V materials multi-junction photoelectrodes at NREL as a result of the US PEC Working Group efforts [2, 3].

Achievement of new benchmark performance levels in copper-chalcopyrite thin film materials in multi-junction photoelectrode configurations at UH/MVSystems as a result of the US PEC Working Group efforts [4].

Successful demonstration of Z-scheme photocatalyst systems and screening of numerous photocatalyst materials in research institutes in Japan, including AIST and TUS [5,6,7,8].

Successful demonstration of enhanced performance heterojunction thin film material systems in South Korea at POSTECH [9].

Continued work in the development and publication of a series of “White Papers” on the most interesting PEC focus materials by the PEC experts; this effort was greatly facilitated by utilizing the DOE Working Group / IEA-HIA Task-26 SharePoint project site. [ https://photoelectrochemical.sharepointsite.net/default.aspx]

Continued work on the refinement of the “Standardized Methodologies for PEC Measurements and Reporting” effort. The international peer review process was completed, and Springer has agreed to publish the documents in book form with assigned editors (Huyen Dinh of NREL, Eric Miller of DOE, and Zhebo Chen of Stanford University).[www2.eere.energy.gov/hydrogenandfuelcells/pec_standards_review.html#

standards] Publication of more than 50 PEC articles per year in scientific journals and in

conference proceedings by Task-26 Experts and associated research groups (including US Experts with past or current affiliation with the DOE Solar Fuels Hub Project [10]).

Two important books on PEC have been published with major contributions from the Task-26 Experts:

− Photoelectrochemical Hydrogen Production, R. Van de Krol, M. Gratzel editors, Springer, 2011 [11].

− On Solar Hydrogen & Nanotechnology, L.Vayssieres editor, Wiley, 2010 [12].

MAJOR COLLABORATIVE PRODUCTS

The international PEC experts, in addition to achieving specific technical milestones (such as those described above), have produced a number of collaborative products which are expected to benefit the PEC research community at large. These have included: a series of “White Papers” on PEC focus materials which document the current understanding of the challenges and potential of each materials class, and include the most up-to-date references for each; a technoeconomic analysis report on PEC hydrogen production projecting technical feasibility of different feasible approaches; a published document on standard methods in the characterization and reporting of PEC materials; and two seminal books on PEC hydrogen production comprised of chapter contributions from the international experts in this field. These Task 26 collaborative products are documented in the appendices of this report as follows:

http://photoelectrochemical.sharepointsite.net/default.aspx

http://www2.eere.energy.gov/hydrogenandfuelcells/pec_standards_review.html

http://www2.eere.energy.gov/hydrogenandfuelcells/pec_standards_review.html

13

APPENDIX A: PEC Materials White Papers

As a culmination of recent activities in PEC Materials Development (e.g., Sub Task “M”), international experts worked collaboratively to develop a series of “White Papers” on the most interesting PEC focus materials. These White Papers document the current understanding of the challenges and potential of each materials class, including the most up-to-date references. Plans are underway to publish the White Papers online in a “Wiki” format to allow for continued expansion based on feedback from the PEC R&D community at large. The current set of White Papers, included in their current versions in Appendix A, include:

III-V crystalline semiconductors material systems Fe2O3 based thin-film materials WO3 based thin-film materials I-III-VI2 thin-film semiconductor materials Molybdenum disulfide (MoS2) nanostructured photocatalysts Bismuth vanatade (BiVO4) materials Tantalum oxi-nitride (TaON) materials PEC Materials Theory Updates New oxide materials and material systems

APPENDIX B: Technoeconomic Analysis of PEC Hydrogen Production

Under contract to the US Department of Energy, and in conjunction with the participation of PEC experts (e.g. Sub Task “T” activities), Directed Technologies Inc. (DTI) conducted a techno-economic evaluation of conceptual PEC hydrogen production systems and produced a comprehensive report (available to the public online) to document the analysis. Appendix B summarizes the contents of the DTI report, including the reports executive summary.

APPENDIX C: PEC H2O Splitting Standards, Experimental Methods, and Protocols

PEC experts participating in Task 26 activities (e.g. Sub Tasks “C” and “S”) in conjunction the US Department of Energy PEC Working Group developed a series of documents establish standardized protocols in the characterization and reporting of PEC materials systems. These documents have been assembled, and are being published collectively as a Springer Brief. Appendix C summarizes the contents of this important document, including the preface and introduction.

APPENDIX D: ON SOLAR HYDROGEN& NANOTECHNOLOGY (2010)

The 2010 seminal book on PEC hydrogen production “On Solar Hydrogen & Nanotechnology” published by Springer included major contributions from the Task-26 Experts. Appendix D summarizes the contents of this important publication, including the contributor list and the books preface.

APPENDIX E: PHOTOELECTROCHEMICAL HYDROGEN PRODUCTION (2012)

The 2012 seminal book “PEC Hydrogen Production” published by Wiley included major contributions from the Task-26 Experts. Appendix E summarizes the contents of this important publication, including the contributor list and the books preface.

14

REFERENCES

Please note additional references included with each of the Materials White Papers in

Appendix A

1. NanoPEC Project Publications website: http://nanopec.epfl.ch/publications 2. T. Deutsch, Semiconductor Photoelectrodes for Direct Water Splitting, Pacifichem

2010 Congress, Honolulu, HI. December 15-20, 2010. 3. J. A. Turner and T. Deutsch, Semiconductor Materials for Photoelectrolysis, US

D.O.E. Hydrogen Program Annual Merit Review Meeting 2011, Arlington, VA, May 9-13, 2011: http://www.hydrogen.energy.gov/pdfs/review11/pd035_turner_2011_o.pdf

4. A. Madan, J. Kaneshiro, et al., US D.O.E. Hydrogen Program Annual Merit Review Meeting 2011, Arlington, VA, May 9-13, 2011: http://www.hydrogen.energy.gov/pdfs/review11/pd053_madan_2011_o.pdf

5. Arai, T., Konishi, Y., Iwasaki, Y., Sugihara, H., and Sayama, K.: High-Throughput Screening Using Porous Photoelectrode for the Development of Visible-Light-Responsive Semiconductors J. Comb. Chem. 9, 574–581 (2007)

6. Kusama, H., Wang, N., Miseki, Y., and Sayama, K.: Combinatorial Search for Iron/Titanium-Based Ternary Oxides with a Visible-Light Response. J. Comb. Chem. 12, 356–362 (2010)

7. Higashi, M., Abe, R., Ishikawa, A., Takata, T., Ohtani, B., and Domen, K.: Z-scheme Overall Water Splitting on Modified-TaON Photocatalysts under Visible Light ( < 500 nm). Chem. Lett. 37, 138-139 (2008)

8. Arakawa, H., Zou, Z., Sayama, K., and Abe, R.: Direct Water Splitting By New Oxide Semiconductor Photocatalysts Under Visible Light Irradiation. Pure Appl. Chem. 79, 1917–1927 (2007)

9. POSTECH PEC Group website: http://ecocat.postech.ac.kr/ 10. DOE Solar Fuels Innovation Hub at JCAP Website: http://solarfuelshub.org/ 11. BOOK: Photoelectrochemical Hydrogen Production, R. Van de Krol, M. Gratzel

editors, Springer, 2012. 12. BOOK: On Solar Hydrogen & Nanotechnology, L.Vayssieres editor, Wiley, 2010.

ACKNOWLEDGEMENTS

I would like to acknowledge the International Energy Agency Hydrogen Implementing Agreement and the US Department of Energy for all their support for Task 26 activities. Special thanks go out to all the excellent, hard-working and dedicated PEC experts across the world that make this such an exciting and rewarding field of research and development.

http://www.hydrogen.energy.gov/pdfs/review11/pd035_turner_2011_o.pdf

http://www.hydrogen.energy.gov/pdfs/review11/pd053_madan_2011_o.pdf

http://ecocat.postech.ac.kr/

http://solarfuelshub.org/

APPENDIX A

US DEPARTMENT OF ENERGY PEC WORKING GROUP WHITE

PAPERS ON PHOTOELECTROCHEMICAL MATERIALS

http://www1.eere.energy.gov/hydrogenandfuelcells/m/photoelectrochemical_group.html#whitepapers

US DOE PEC Working Group White Papers: October 2013 Drafts

A-1

White Papers on Materials for Photoelectrochemical Water Splitting

CONTENTS

III-V Semiconductor Systems for High-Efficiency Solar Water Splitting Applications

Todd Deutsch, Heli Wang, Zhebo Chen, Shane Ardo, Shu Hu, Mahendra Sunkara, Dan Esposito, Yat Li, Shannon Boettcher

I-III-VI2 Copper Chalcopyrites for Photoelectrochemical Water Splitting Jess Kaneshiro, Todd Deutsch, Nicolas Gaillard, Zhebo Chen, Alan Kleiman-Shwarsctein, Feng Zhu, Michael Weir

The Viability of Using Tungsten Oxide Based Compounds as a Photoelectrode for the Solar Production of Hydrogen

Nicolas Gaillard, Yat Li, Heli Wang

Molybdenum Disulfide for Photoelectrochemical Water Splitting Z. Chen, J.D. Benck, T.F. Jaramillo

Engineered Ternary and Quaternary Oxide Minerals with Optimal Absorption Characteristics for Solar-Assisted Low-Cost Hydrogen Production

Nicolas Gaillard, Muhammad N. Huda

BiVO4 as a Photoanode for Photoelectrical Water Splitting Kyoung-Shin Choi, Roel van de Krol

Hematite as a Photoelectrode for Photochemical Hydrogen Production Heli Wang, Isabell Thomann, Arnold J. Forman, Yat Li, Mahendra Sunkara, Moreno de Respinis

Photovoltage of α-Fe2O3 Shannon W. Boettcher, Arnold J. Forman, Muhammad N. Huda, Heli Wang

The Viability of using Amorphous Silicon Carbide (a-SiC) as a Photoelectrode for PEC Hydrogen Production

Jian Hu, Feng Zhu, Ilvyda Matulionis, Josh Gallon, Nicolas Gaillard, and Todd Deutsch

Appendix: Current Matching Jian Hu, Feng Zhu, Ilvyda Matulionis, Josh Gallon, Nicolas Gaillard, and Todd Deutsch

Appendix: Energetic Mismatch Heli Wang, Arnold J. Forman, Moreno de Respinis, Nicolas Gaillard, Shannon Boettcher

Appendix: Design and Characterization of Photoelectrodes from First Principles Tadashi Ogitsu, Brandon Wood, Wooni Choi, Muhammad N. Huda, Su-Huai Wei



A-2

PEC White Papers: III-V Semiconductors for PEC

III-V semiconductor systems for high-efficiency solar water splitting applications

Todd Deutsch1, Heli Wang1, Zhebo Chen2, Shane Ardo3, Shu Hu3, Mahendra Sunkara4, Dan Esposito5, Yat Li6, Shannon Boettcher7 1National Renewable Energy Laboratory, 2Stanford University, 3California Institute of Technology, 4University of Louisville, 5National Institute of Standards and Technology, 6University of California-Santa Cruz, 7University of Oregon Introduction

Semiconductors composed of group IIIA and VA elements, commonly referred to as III-V’s, represent a material class that demonstrates unparalleled photovoltaic (PV) and photoelectrochemical (PEC) conversion efficiencies. The current 43.5% PV efficiency1 and 12.4% PEC water splitting efficiency2 records were set with III-V semiconductor materials. These high efficiencies are a result of direct transition optical band gaps and the ability to grow low-defect epitaxial films that have long charge carrier lifetimes and high mobilities. Because III-V’s have composition-dependent band gaps well matched to the solar spectrum (1-2 eV), they allow for multijunction configurations that can exceed the Shockley-Queisser efficiency limit3 for single junction photovoltaics. Multijunction cells also generate higher voltages, making them suitable for water-splitting applications where potential differences in excess of 1.7 V are required under operating conditions. Technical Challenges 1. Interfacial band edge mismatch One technical barrier is that the valence band edge of most III-V’s, at least the ones that absorb visible light, is pinned at a potential that is insufficient (too negative) to drive the water oxidation half reaction4. This barrier can be addressed by incorporating a tandem architecture, by coupling with a separate photoanode, or by creating a buried photoactive junction such that the solution-absorber interface energetics become less important 2,5,6. The first two of these strategies increase the oxidation potential of the hole through photoexcitation in a secondary space charge region in the bulk or at a counter electrode while the last uses an ohmic contact at the semiconductor-electrolyte interface to unpin the band edges. 2. Stability

Material stability in aqueous electrolyte is the greatest challenge preventing implementation of III-V’s in commercial solar-hydrogen photoreactors. The III-V materials are generally susceptible to corrosion via oxidation of the semiconductor components into solvated ions in the electrolysis bath7. Operating the semiconductor as a photocathode can provide cathodic protection from oxidation due the reducing environment and can extend the lifetime of III-V’s in contact with an electrolyte, but unprotected surfaces can only withstand a few days of continuous operation. Economical PEC production of hydrogen demands the semiconductor operate without a significant loss in performance for several



A-3


thousand hours, a formidable challenge considering the inherent instability of the III-V’s at the electrolyte interface. 3. Cost-effective high-volume synthesis Technical challenges relating to the synthesis of III-V based materials in a cost-effective manner is another key issue that must be addressed in order to commercialize III-V PEC photoreactors. III-V’s are typically synthesized by metal organic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE). MOCVD relies on highly toxic, pyrophoric, and expensive precursors. MBE is a slow ultra-high vacuum process. The multilayer device architectures and abrupt interfaces require these precision epitaxial deposition techniques, which utilize expensive single-crystal growth templates that are difficult to reuse. Therefore, although the active film thickness for III-Vs is typically 1-2 μm (due to the high absorption coefficient associated with a direct band gap), the growth substrate that the absorber resides on is typically several hundred μm-thick. The confluence of specialized batch synthesis and high substrate costs can make the finished semiconductor prohibitively expensive. Currently a 6” (182cm2) GaAs substrate suitable for epitaxy costs about $180, making the substrate contribution alone about $10,000/m2. The highest PEC water-splitting efficiency was measured under 12 suns2 demonstrating that III-V’s can be used with moderate light concentration. A photoreactor design using moderate light concentration (10x) could thus require only about 1/10th the absorber material per reactor unit area, leading to material economy. Despite incurring additional balance of systems costs, including tracking systems, as well as the loss in the ability to utilize much of the diffuse solar irradiation, technoeconomic analysis8 of a 10x concentrator reactor suggests that hydrogen would be commercially viable from a stable and efficient material with an absorber cost of $150/m2. Light emitting diode III-V synthesis has seen a dramatic reduction in semiconductor cost by maximizing the yield per batch through increased wafer size and multiple wafers per reactor.

There are a few techniques that exist that offer low-cost alternatives to MOCVD and MBE. Vapor-liquid-solid (VLS) growth mechanism allows for seeding of single-crystal semiconductor wires off arbitrary substrates. Usually, (111)-oriented single crystal substrates are used to grow vertically-aligned single crystal Si or GaAs wire arrays. Textured polycrystalline films with (111)-preferred orientation deposited on low-cost glass or flexible sheets can be used as growth templates for assembling aligned nanowire or microwire light absorbers9,10. Anther potential approach is to utilize nanowire to microwire arrays as substrates for growing thick epitaxial layers thus reducing the need for expensive, single crystal substrates11. The wire geometry is more forgiving of strain from near-epitaxial growth in comparison to planar materials, and epitaxy is preferred for tandem device fabrication. The processes for nanowire array substrates are scalable for large areas; however, contacting a mat of randomly oriented nanowires can be nontrivial. Nanosphere lithography allows for wafer-scale synthesis of nanowires12. Scaling-up of MOCVD process and large-scale nano-patterning techniques (e.g. nano-imprint lithography) is currently being pursued by the semiconductor manufacturing industry.



A-4


Research Status

The highest direct, unbiased solar photoelectrolysis efficiency was achieved using a tandem III-V semiconductor: a p-GaInP2 PEC junction integrated with a p/n-GaAs photovoltaic converted illumination to hydrogen and oxygen at 12.4% solar to hydrogen (STH) efficiency2. This measurement was made under 12x light intensity showing the viability of moderate concentration in these devices. After 20 hours, the short-circuit current density dropped from 120mA/cm2 to 105 mA/cm2 (10.8% STH) and although the entire surface continued to evolve gas bubbles, areas of localized damage were observed.

Recent results indicate that an improved efficiency can be obtained by using a more active oxygen evolution counter electrode. When RuO2 was used instead of platinum black, the two-terminal J-V curve was shifted towards higher efficiencies. An efficiency under real-solar (outdoor) conditions of 16.3% was measured, but this was under a moderate bias13. The most recent iteration of the tandem cells did not match the performance of those used to establish the 12.4% unbiased efficiency benchmark due to current difficulties in achieving reproducible synthesis, a problem that must be addressed prior to scale-up. It is likely that the short-circuit efficiency of the original materials would have been greater had RuO2 been used.

Two efforts to stabilize III-V’s using nitrides have had moderate success. The inclusion of the nitride in the bulk of GaP demonstrated an ability to reduce the corrosion on the III-V surface 14,15, however, the nitride also led to a significant loss in photoconversion efficiency. A nitrogen ion implantation treatment on p-GaInP2 surfaces (without a buried p/n GaAs junction) has been extremely successful in nearly eliminating corrosion by generating a surface nitride layer. One nitride treated sample exhibited no detectable damage after 115-hours of passing a constant -10mA/cm2 photocurrent, the equivalent to 12.3% solar-to-hydrogen conversion16. The nitride treatment did lead to an approximate 5% relative loss in (light-limited reverse bias) photoconversion ability, but this minimal loss could be tolerated by a 12% STH cell and still exceed the 10% benchmark. Approaches Research on III-Vs for PEC applications focuses primarily on utilizing proven high-efficiency configurations and engineering a stabilized surface. Protection of the surface can be accomplished through inclusion of a stabilizing agent throughout the bulk during synthesis, such as the incorporation of a dilute nitride. Other areas of research aim at surface or near-surface treatments that can chemically protect the interface while maintaining a low interface defect density and fast charge transfer to the surface attached electrocatalyst. Some surface treatments are in the form of coatings (oxides via atomic layer deposition, polymers, metal overcoats, catalysts). Other treatments are aimed at changing the near surface chemistry through application of a plasma, via electrochemical means, or by ion-implantation.

Research on III-V synthesis, not necessarily for only PEC applications, has focused on two key areas; reusing the expensive substrate or eliminating it entirely. Synthesis routes that maximize the substrate utility are epitaxial lift-off17,18 (ELO)



A-5


and inverted metamorphic multijunction19 (IMM). Both of these approaches are currently being exploited by industry (Alta Devices, RFMD). Another approach, close spaced vapor transport (CSVT), seeks to eliminate the substrate and the costly metal organic precursors. Dopant diffusion at the high temperatures required for CSVT could compromise abrupt junctions in certain cases and prevent CSVT from providing complex multilayer configurations. However, high efficiency dual-photoelectrode systems have been demonstrated that use simple combinations of a p-type photocathode and n-type photoanode5. The fast growth rate of high-quality, single conductivity type materials makes CSVT an attractive route to high volume production of dual absorber systems20, especially if growth on low-cost substrates can be accomplished.

Another approach is based on modification of the wide band gap (3.4eV) semiconductor GaN, a stable III-V material that has band edges that encompass the water splitting half-reaction potentials. This route uses the indium content in a pure nitride alloy InxGa1-xN as a lever to tune the band gap and band edge positions. InN and GaN have a lattice mismatch of close to 12% making synthesis high indium content films challenging via conventional (high temperature) methods due to strain induced effects21. Low temperature epitaxial routes that use energetic atoms in lieu of plasma have overcome this miscibility gap22,23. GaN has better corrosion resistance than other metal nitrides (e.g., Ta3N5) and phosphides (e.g., GaP), but the long term stability is still not comparable to metal oxides24–26. Stabilizing InxGa1-xN photoelectrodes is a key issue that remains to be solved.

Research on III-V materials also includes synthesizing alloys of GaN using small amounts of antimony (<5 at%) to obtain alloys with band gaps below 2.4 eV27. These alloys with dilute antimony compositions have been predicted to straddle hydrogen and oxygen evolution reactions. References

1. Green, M. A., Emery, K., Hishikawa, Y., Warta, W. & Dunlop, E. D. Solar cell

efficiency tables (version 39). Progress in Photovoltaics: Research and

Applications 20, 12–20 (2012).

2. Khaselev, O. & Turner, J. A monolithic photovoltaic-photoelectrochemical device

for hydrogen production via water splitting. Science (New York, N.Y.) 280, 425–7

(1998).

3. Shockley, W. & Queisser, H. J. Detailed Balance Limit of Efficiency of p-n

Junction Solar Cells. Journal of Applied Physics 32, 510–519 (1961).

4. Kocha, S. S., Turner, J. A. & Nozik, A. J. Study of the Schottky barrier and

determination of the energetic positions of band edges at the n- and p-type gallium

indium phosphide electrode | electrolyte interface. Journal of Electroanalytical

Chemistry 367, 27–30 (1994).



A-6


5. Kainthla, R. C., Zelenay, B. & Bockris, J. O. M. Significant Efficiency Increase in

Self-Driven Photoelectrochemical Cell for Water Photoelectrolysis. Journal of The

Electrochemical Society 134, 841 (1987).

6. Boettcher, S. W. et al. Energy-conversion properties of vapor-liquid-solid-grown

silicon wire-array photocathodes. Science (New York, N.Y.) 327, 185–7 (2010).

7. Khaselev, O. & Turner, J. A. Electrochemical Stability of p-GaInP2 in Aqueous

Electrolytes Toward Photoelectrochemical Water Splitting. Journal of The

Electrochemical Society 145, 3335–3339 (1998).

8. James, B. D., Baum, G. N., Perez, J. & Baum, K. N. Technoeconomic Analysis of

Photoelectrochemical (PEC) Hydrogen Production. 1–127 (2009).at

<http://www1.eere.energy.gov/hydrogenandfuelcells/pdfs/pec_technoeconomic_an

alysis.pdf>

9. Chuang, L. C. et al. Critical diameter for III-V nanowires grown on lattice-

mismatched substrates. Applied Physics Letters 90, 043115 (2007).

10. Tomioka, K., Kobayashi, Y., Motohisa, J., Hara, S. & Fukui, T. Selective-area

growth of vertically aligned GaAs and GaAs/AlGaAs core–shell nanowires on

Si(111) substrate. Nanotechnology 20, 145302 (2009).

11. Pendyala, C. et al. Nanowires as semi-rigid substrates for growth of thick,

In(x)Ga(1-x)N (x > 0.4) epi-layers without phase segregation for

photoelectrochemical water splitting. Nanoscale 4, 6269–75 (2012).

12. Madaria, A. R. et al. Toward optimized light utilization in nanowire arrays using

scalable nanosphere lithography and selected area growth. Nano letters 12, 2839–

45 (2012).

13. Turner, J. A. & Deutsch, T. G. Semiconductor Materials for Photoelectrolysis.

Proceedings of the 2011 U.S. Department of Energy Hydrogen and Fuel Cells

Program and Vehicle Technologies Program Annual Merit Review and Peer

Evaluation Meeting (2011).

14. Deutsch, T. G., Head, J. L. & Turner, J. A. Photoelectrochemical Characterization

and Durability Analysis of GaInPN Epilayers. Journal of The Electrochemical

Society 155, B903 (2008).

15. Deutsch, T. G., Koval, C. A. & Turner, J. A. III-V nitride epilayers for

photoelectrochemical water splitting: GaPN and GaAsPN. The Journal of Physical

Chemistry B 110, 25297–307 (2006).



A-7


16. Deutsch, T. G. & Turner, J. A. Semiconductor Materials for Photoelectrolysis.

Proceedings of the 2012 U.S. Department of Energy Hydrogen and Fuel Cells

Program and Vehicle Technologies Program Annual Merit Review and Peer

Evaluation Meeting (2012).

17. Yoon, J. et al. GaAs photovoltaics and optoelectronics using releasable multilayer

epitaxial assemblies. Nature 465, 329–33 (2010).

18. Schermer, J. J. et al. Epitaxial Lift-Off for large area thin film III/V devices.

Physica Status Solidi (a) 202, 501–508 (2005).

19. Geisz, J. F. et al. 40.8% Efficient Inverted Triple-Junction Solar Cell With Two

Independently Metamorphic Junctions. Applied Physics Letters 93, 123505 (2008).

20. Ritenour, A. J., Cramer, R. C., Levinrad, S. & Boettcher, S. W. Efficient n-GaAs

Photoelectrodes Grown by Close-Spaced Vapor Transport from a Solid Source.

ACS Applied Materials & Interfaces 4, 69–73 (2012).

21. Moses, P. G. & Van de Walle, C. G. Band bowing and band alignment in InGaN

alloys. Applied Physics Letters 96, 021908 (2010).

22. Williamson, T. L., Williams, J. J., Hubbard, J. C. D. & Hoffbauer, M. A. High In

content In[sub x]Ga[sub 1−x]N grown by energetic neutral atom beam lithography

and epitaxy under slightly N-rich conditions. Journal of Vacuum Science &

Technology B: Microelectronics and Nanometer Structures 29, 03C132 (2011).

23. Williamson, T. L., Salazar, A. L., Williams, J. J. & Hoffbauer, M. a.

Improvements in the compositional uniformity of In-rich InxGa1-xN films grown

at low temperatures by ENABLE. Physica Status Solidi (C) 8, 2098–2100 (2011).

24. Li, J., Lin, J. Y. & Jiang, H. X. Direct hydrogen gas generation by using InGaN

epilayers as working electrodes. Applied Physics Letters 93, 162107 (2008).

25. Ryu, S.-W., Zhang, Y., Leung, B., Yerino, C. & Han, J. Improved

photoelectrochemical water splitting efficiency of nanoporous GaN photoanode.

Semiconductor Science and Technology 27, 015014 (2012).

26. Waki, I. et al. Direct water photoelectrolysis with patterned n-GaN. Applied

Physics Letters 91, 093519 (2007).

27. Sheetz, R. M. et al. Visible light absorption and large band gap bowing in dilute

alloys of gallium nitride with antimony. Physical Review Letters 84, 075304

(2011).



A-8

PEC White Papers: I-III-VI2 for PEC Hydrogen Production

I-III-VI2 Copper Chalcopyrites Primary Authors & Affiliations:

Jess Kaneshiro (HNEI), Todd Deutsch (NREL), Nicolas Gaillard (HNEI), Zhebo Chen (Stanford U.), Alan

Kleiman-Shwarsctein (Solopower), Feng Zhu (MVSystems), Michael Weir (UNLV)

Introduction:

The I-III-VI2 alloyed semiconductor class, championed in the photovoltaic (PV)

scientific field by Cu(InxGa1-x)Se2 (often abbreviated “CIGS”)1, incorporates a wide

range of materials that are useful in solar energy conversion. Within this material class,

bandgaps between 1.0eV and 2.43eV can be obtained by varying the alloy ratios in each

elemental group2. Of most interest for photoelectrochemical (PEC) water-splitting is the

higher-bandgap members of this class, such as CuGaSe2 with a bandgap of 1.65eV which

has served generally as the baseline material for this application.

The variable bandgap of the copper chalcopyrite compounds has been studied

extensively, most often for PV applications, making this material class particularly

attractive for the development of PEC materials and systems by utilizing the very rich

existing knowledge base1,2

. The very high Cu(InGa)Se2 PV conversion efficiency of

20.3% (Nov. 2011) with lower-bandgap material is made possible by the strong optical

absorption due to a direct bandgap, exceptional carrier transport properties, and

compositional tunability enabling highly beneficial bandgap grading3. These traits are

just as important for a PEC device and are, for the most part, maintained in the higher-

bandgap materials like CuGaSe2. Despite the existing knowledge base, significant work

is needed to tune the material properties (e.g. bandgap, band edge alignment, and

corrosion resistance; see White Paper Appendix) for PEC applications. Previous work has

demonstrated photocurrents as high as 20mA/cm2 in a PEC application, stability in very

highly acidic electrolytes, and durability up to 420 hours4. It is worth mentioning that any

CIGSe-based PEC technologies would rely on existing fabrication techniques already

implemented for CIGSe PV technologies, allowing for the rapid deployment of PEC

devices when a suitable material is developed.

Because materials viable for PEC devices have high bandgaps, much of the visible

solar spectrum is not absorbed. This offers the opportunity to harvest the unabsorbed

photons with underlying PV cells of smaller bandgaps resulting in multi-junction devices

that can more effectively utilize incident light4. Multi-junction devices are electrically

connected in series, resulting in a summation of voltages to counteract typically

insufficient band edge alignment and kinetic overpotentials of the water-splitting

reactions. Multi-junction absorbers have been utilized to achieve world record

efficiencies in both PV5 and PEC

6 devices utilizing the III-V material class, and are the

optimal configuration in the development of most planar PEC water-splitting devices.

This is particularly convenient for alloyed semiconductors like the I-III-VI2 material class

because the ability to engineer the bandgaps by alloy compositions permit tuning to

optimize absorption via multi-layers within the same material class.

Technical Challenges:

The research philosophy in this material class has so far aimed to take an optimized

material (Cu(InGa)Se2), proven to have excellent optoelectronic properties in the PV



A-9


field, and repurpose it for the requirements of water-splitting. The first requirement is a

high bandgap of at least 1.23eV, and preferably much higher in the 1.6-2.1eV range4,7

.

By removing In from optimized PV devices resulting in CuGaSe2, the bandgap is raised

to a modest 1.65eV, which can, in an appropriate configuration, be used to split water.

CuGaSe2 thus serves as the baseline material for this material class. Its chemical

simplicity, ease of fabrication, and close relation to commercial PV Cu(InGa)Se2 has

resulted in very robust fabrication of high-quality cells, and performance and durability

have proven to be exceptional4,8,9,10.

With the increase of Ga, however, the bandgap is expanded by a rise in the

conduction band while the valence band remains misaligned with respect to the oxygen

evolution potential4. This occurs because the valence band energy is dominated by the

Cu-3d to Se-4p orbital bond, which is unaffected by the group-III alloy content (In,Ga)11

.

Furthermore, CuGaSe2 is subjected to the “doping pinning rule”, common in highly-

doped semiconductors, wherein the Fermi level is pinned at a certain level interrupting

the band bending vital to photovoltage production12

. This doping pinning precludes the

formation of a surface inversion layer that is very highly relied upon in Cu(InGa)Se2 PV

devices, resulting in a sub-proportional increase in open circuit voltages as bandgaps are

raised with an increase in the Ga/(In+Ga) ratio11,13

. This effect is also evident in PEC

devices where CuGaSe2 exhibits very low photovoltage than would be expected for its

bandgap, and therefore requires a very high voltage bias (presumably provided by light-

harvesting PV cells) to split water.

Therefore (with the design philosophy of using CuGaSe2 as a baseline material), the

largest barrier in the I-III-VI2 material class is overcoming the excessive voltage

requirements originating from the misaligned valence band edge. Because the valence

band energy is dominated by the I-VI bond (Cu-Se in the base case), it is the Cu and/or

Se content that must be modified by, for example, replacing some Cu with Ag or Se with

S. While both of which will raise the bandgap, the Se/S substitution may not significantly

change the valence band maximum required to improve the band edge alignment14

.

Research Status:

Band-edge misalignment in copper chalcopyrite films analyzed up until now makes

unassisted water splitting impossible, and necessitates an external applied bias. While

other device structures have been utilized, a monolithic stack of PEC and PV cells

creating a hybrid photoelectrode (HPE) is the ultimate goal. At 1.65eV, typical 1μm-thick

CuGaSe2 cells are at the low end of optimum bandgap range (1.6-2.1eV), and do not pass

sufficient light to underlying PV cells to provide enough voltage biasing. Recall that the

band-edge misalignment in CuGaSe2 results in a very high required voltage bias.

For this reason, alternative device designs were explored to utilize the very high

photocurrents of optimized CuGaSe2 of as high as 20mA/cm2 by leveraging it towards

photovoltage production. The resultant co-planar hybrid PV/PEC device employs 3 PV

devices in a planar configuration next to a CuGaSe2 PEC device. Because the devices are

side-by-side, the total device area is the summation of the constituent devices. This is

how the high current density of CuGaSe2 is sacrificed to liberate real estate for PV

devices to produce the voltage bias. Load-line analysis was used to determine the optimal

ratio of PV/PEC areas and the resulting standalone device was tested outdoors producing

3.53mA/cm2 at AM1.5G illumination, equivalent to a 4.35% solar-to-hydrogen (STH)



A-10


conversion efficiency4. Low-cost a-Si PV devices provided by MVSystems were utilized

in this scheme, representing a relatively cheap solution to water-splitting. With 4

junctions, however, this device serves more as a proof-of-concept, described more in the

following section, and would not by itself be a very practical device to split water

because it operates at a relatively high voltage around 1.6V (2-terminal); a result of the

unfavorably aligned band edges4.

CuGaSe2, along with possessing high photocurrents, also benefits from excellent

durability in very highly concentrated electrolytes. Higher ionic concentrations in the

electrolyte provide better ionic conductivity, resulting in lower series resistance losses

improving the “fill factor”, analogous to the term applied to J-V curves in PV technology.

In 1M H2SO4 under AM1.5G illumination and biased to produce 4mA/cm2 (the current

density indicative of 5%STH efficiency), CuGaSe2 has thus far demonstrated 420 hours

of continuous operation before significant degradation of performance.

Approaches:

Having four coplanar junctions is not very practical as the high current is split across

a larger absorption area, resulting in a lower overall device current density. A coplanar

configuration also does not benefit from spectrum splitting, which more efficiently uses

the potential of incident photons resulting in better voltage characteristics. This

demonstration was valuable, however, as a starting point for device design advancement.

Although CuGaSe2 is currently unsuited for monolithic device integration, the coplanar

device offers a proof-of-concept that I-III-VI2 materials can be used in a PEC device that

only uses light to split water. By showing that this device can achieve 4.35% STH with

economical a-Si PV cells, extended simulation shows that burying just one PV cell can

result in efficiencies upwards of 5% STH by reducing the area-division and increasing

useable current density. As material advancements are achieved in the future, they can be

inserted into this device design pathway to grasp both the full potential and the material

performance goals that must be met to proceed further towards a monolithic device where

all PV cells are buried.

The importance of material advancements is specifically highlighted by the need to

address the valence band misalignment dominated by the I-VI bonds. It has already been

theorized and demonstrated that the substitution of all or some of the selenium with sulfur

(VI-elements) can increase the bandgap in copper chalcopyrites, and furthermore,

sulfurization can lower the valence band placing it closer to the water oxidation

potential2,15,16. Material engineering can also be accomplished by manipulating the group

I elements; partial replacement of Cu with Ag can slightly raise the bandgap as well as

decrease the intrinsic p-type doping density. This may allow the formation of the surface

inversion layer, aiding photovoltage development17

.

Another potential research area is the use of Cu2ZnSnS4 (CZTS) and related materials

for PEC. This material is technically a I2-II-IV-VI4 and effectively exchanges the group

III atom (In, Ga) for 50% group II (Zn) and 50% group IV (Sn). CZTS is increasing in

interest to the PV community because of the use of more abundant and sustainable

elements18

. CZTS is natively a p-type semiconductor with a bandgap of ~1.5 eV.

However, the same elemental substitution approach from CIGS (e.g. Ag for Cu, Se for S)

applies to CZTS and has been used to improve the solar conversion efficiency19

. Only a



A-11


few reports of CZTS for PEC applications are available, leaving open the possibility of

adapting this material to water splitting20-24

.

Surface catalysis is another method of device enhancement being investigated to

improve reaction kinetics. Progress with Pt and Ru nanoparticle treatments have shown

increased performance, but lower cost solutions would be preferred.

1 I. Repins, M.A. Contreras, B. Egaas, C. DeHart, J. Scharf, C.L. Perkins, B. To, R.

Noufi, Prog. Photovolt. Res. Appl. 16 (2008) 235. 2 M. Bär, W. Bohne, J. Röhrich, E. Strub, S. Lindner, M.C. Lux-Steiner, Ch.-H. Fischer,

J. Appl. Phys. 96 (2004) 3857. 3 P. Jackson, D. Hariskos, E. Lotter, S. Paetel, R. Wuerz, R. Menner, W. Wischmann, M.

Powalla, Prog. Photovolt. Res. Appl. 19 (2011) 894. 4 J. Kaneshiro, Dissertation Thesis, University of Hawai`i at Mānoa, 2012. 5 S. Wojtczuk, P. Chiu, X. Zhang, D. Derkacs, C. Harris, D. Pulver and M. Timmons,

Photovoltaics Specialists Conference (PVSC) 35th IEEE (2010) Honolulu, HI, p. 001259. 6 O. Khaselev, A. Bansal, J.A. Turner, Int. J. Hydrogen Energ. 26 (2001) 127. 7 A. J. Bard and M. A. Fox, Acc. Chem. Res. 28 (1995) 141. 8 B. Marsen, S. Dorn, B. Cole, R. E. Rocheleau, E. L. Miller, Mater. Res. Soc. Symp.

Proc. 974E, Warrendale, PA, 2007, 0974-CC09-05. 9 J. Leisch, J. Abushama, and J. A. Turner, ECS Meeting Abstracts 502 (2006) 821-821.

10 B. Marsen, B. Cole, E. L. Miller, Sol. Energy Mater. Sol. Cells,

doi:10.1016/j.solmat.2008.03.009 (2008). 11 M. Turcu and U. Rau, J. Phys. Chem. Solids 64 (2003) 1591. 12 S. B. Zhang, S.-H. Wei, A. Zunger, J. Appl. Phys. 83 (1998) 3192. 13 D. Schmid, M. Ruckh, F. Grunwald, H. W. Schock, J. Appl. Phys. 73 (1993) 2902. 14 M. Bär, S. Nishiwaki, L. Weinhardt, S. Pookpanratana, O. Fuchs, M. Blum, W. Yang, J.D.

Denlinger, W.N. Shafarman, C. Heske, Appl. Phys. Lett. 93 (2008) 244103. 15

S.-H. Wei and A. Zunger, J. of Appl. Phys. 78 (1995) 3846-3856. 16

L. Weinhardt, Dissertation Thesis, University Wurzburg, 2005. 17 J. Kaneshiro, A. Deangelis, N. Gaillard, Y. Chang, J. Kowalczyk, E.L. Miller, Photovoltaic

Specialists Conference (PVSC) 35th IEEE (2010) Honolulu, HI, p.002448. 18 H. Wang, Int. J. Photoenergy (2011) 801292. 19 T. K. Todorov, K. B. Reuter, D. B. Mitzi Adv. Mater. 22 (2010) E156 20 D. Yokoyama, T. Minegishi, K. Jimbo, T. Hisatomi, G. Ma, M. Katayama, J. Kubota, H.

Katagiri, K. Domen, Appl. Phys. Express 3 (2010) 101202. 21 G. Ma, T. Minegishi, D. Yokoyama, J. Kubota, K. Domen, Chem. Phys. Lett. 501 (2011)

619-622. 22 M. Miyauchi, T. Hanayama, D. Atarashi, E. Sakai, J. Phys. Chem. C 116 (2012) 23945-

23950. 23 T. Kameyama, T. Osaki, K. Okazaki, T. Sibayama, A. Kudo, S. Kuwabata, T. Torimoto J.

Mater. Chem. 20 (2010) 5319-5324. 24 S. C. Riha, S. J. Fredrick, J. B. Sambur, Y. Liu, A. L. Prieto, B. A. Parkinson ACS Appl.

Mater. Interfaces 3 (2011) 58-66



A-12

PEC White Papers: WO3 for PEC Hydrogen Production

The viability for using tungsten oxide based compounds as a photoelectrode for the solar production of hydrogen.

Nicolas Gaillarda, Yat Li

b , Heli Wang

c

a. Hawaii Natural Energy Institute

b. University of California, Santa Cruz

c. National Renewable Energy Laboratory

Synopsis:

Tungsten trioxide (WO3) is an n-type semiconductor that has been chosen as a

model material to validate the concept of photoelectrochemical (PEC) water splitting.

WO3 has been an ideal photoanode material in the study of PEC water-splitting systems

because it inherently has good photon absorption generating decent amount of

photocurrent, good electron transport properties, and stability against (photo)corrosion. In

practice, WO3 photoanodes have been implemented in prototype multijunction PEC

systems including hybrid photoelectrodei and dual photoelectrode cell approach

ii. As the

basis for a thin-film photoanode in an efficient water-splitting device, pure WO3 falls

short on two principal fronts. The first barrier is a bandgap (between 2.5-2.8 eV) that is

too high to absorb an adequate portion of the solar spectrum. As experimentally

validated, the high bandgap limits achievable photocurrents, resulting in devices with

solar to hydrogen (STH) conversion efficiencies which do not exceed 3 percentiii

. In

perspective, optimized material systems will be needed to meet the DOE’s 2020

benchmark of 20% STH conversion efficiency. A second barrier for pure WO3 is the non-

optimal band-edge alignment of the conduction band. PEC experiments on WO3 films in

acidic aqueous media have indicated a conduction band minimum that is lower than the

hydrogen evolution reaction reduction potential. In the photoanode configuration, this

non-favorable band-edge alignment results in the need for a supplemental voltage bias,

complicating the design of practical water-splitting devices. As we continue forward in

the R&D of tungsten-based compounds, the primary thrust of our research plan is focus

on the bandgap and band-edge alignment issues. Based on comprehensive theoretical and

experimental feedback efforts, new multi-component tungsten-based compounds are

being developed with raised valence band maximum and conduction band minimum,

ideally resulting in a sufficiently low bandgap for high photocurrents and favorable band-

edge alignment to minimize or eliminate voltage bias requirements. Materials processing

and stability issues for promising new compounds are also be addressed.

Technology Barriers:

Although pure WO3 is not sufficient to meet DOE’s long-term PEC hydrogen

generation via water-splitting performance targets, tungsten-based compounds show great

promise with respect to the three major materials-related barriers listed in the US DOE

EERE Hydrogen program plan. This section highlights the challenges and strengths of



A-13


tungsten-based compounds in this context, and outlines research solution pathways

specific to this material class.

Addressing the Challengesiv:

Y. Materials Efficiency. Different approaches are currently being evaluated to

address WO3 bandgap reduction, band-edge alignment with water redox potential

and optimized surface activity. Incorporation of ions into WO3 bulk film can be

performed to enhance absorption properties (i.e. bandgap reduction), as predicted

by density functional theory calculation.

Z. Materials Durability. Long-term stability in acidic media needs to be evaluated on

tungsten oxide-based materials.

AB. Bulk Materials Synthesis. Several techniques including sol-gel, CVD and reactive

sputtering are ideally suited to this material. Other methods such as Atomic Layer

Deposition (ALD) should be evaluated.

AC. Device Configuration Designs. For initial studies, the multijunction device

structure can be validated via mechanical stacks (already validated with WO3).

With identification of an ideal top PEC material, the engineering of the complete

multi-junction device will be concluded. Work on fully integrated (monolithic)

devices will require close collaboration with our industrial partner (MVSystem).

Research Status:

a. Improving WO3 optical absorption

The reduction of optical bandgap remains the most important task for this material

class. Based on theoretical calculationv, a series of foreign elements have been

incorporated into tungsten trioxide bulk, including sulfur and nitrogen. In both cases, no

clear evidence of bandgap reduction was observed, although absorption characteristics

were impacted by these treatments. A closer look at UV-visible spectra revealed possible

free carrier absorption for WO3:N and WO3:S systems which could originate from defect

points in the newly formed systems. It is worth mentioning that structural

characterization (X-ray analysis) pointed out a dramatic phase modification after foreign

element incorporation, from monoclinic/orthorhombic (known structures for WO3) to

cubic-like system. The later could emerge only if a high concentration of defects (oxygen

vacancies) is present in the material. Subsequent PEC characterization pointed out poor

performances when compared to pure WO3 material; mostly related to weak carrier

transport due to grain structure degradationvi

. It is worth mentioning that in the nitrogen

case, XPS analyses performed at UNLV at newly formed thin film surface has not

revealed so far the presence of any nitrogen species, indicating weak bonds between

involved species. Tungsten oxide alloying with foreign elements has been also reported

by Pr. Augustynski (Warsaw University) and Dr. Braun (EMPA). Their study shown that

Si, Ru, Li, and Mo had a pronounced impact on the morphology of the resulting WO3



A-14


alloys (synthesized via sol-gel method), improving electrochromic characteristicsvii

.

However, water-splitting experiments indicated a net decrease of saturated photocurrent

density under AM1.5G illumination.

b. Near-surface engendering with bilayer approach

Encouraging results have been obtained when Mo is incorporated at the WO3

surface only (bilayer structure). A 15% enhancement of the generated photocurrent at

1.6V vs SCE as well as a 200 mV onset potential reduction (tested under simulated

AM1.5G illumination in 0.33M H3PO4 electrolyte) were observedviii

. It is believed that

improved bilayer performances may come from a beneficial effect of the WO3 bottom

layer on Mo:WO3 grain growth. In fact, bilayer structure shows highly crystalline grains

when compared to those of bulk Mo:WO3 materials (Mo incorporation in the whole film).

High-resolution TEM characterizations performed at NREL indeed validated this point.

Further electron spectroscopy analyses (UPS and IPES) performed at UNLV pointed out

that Mo incorporation into WO3 leads to a conduction band-edge position increase of 210

meVix

when compared to pure tungsten trioxidex. In a bilayer configuration, this

difference in Fermi level position leads to the formation of built in potential which

direction and strength promotes photogenerated holes diffusion toward the PEC

material/electrolyte interface.

c. Surface catalysis of WO3

First evaluation of RuO2 particle deposition on WO3 films has been evaluated by

UCSB for catalytic purposes. RuO2 particles have been either deposited using an

electrochemical process on HNEI reactively sputtered WO3 or deposited using spray

pyrolysis on UCSB electrodeposited WO3. No major improvement in either saturated

photocurrent or the onset potential has been observed after RuO2 surface treatment. In a

second phase, HNEI developed at RuO2 nanoparticle sputtering process using a pure Ru

target in oxidizing environment. By adjusting the deposition duration (basically less that

30 seconds), RuO2 nanoparticles (5-10 nm mean size) were successfully deposited on

tungsten trioxide thin films. Subsequent PEC characterization indicated a 20% increase in

photocurrent density a low potential while the onset potential and the saturated

photocurrent remained unchanged. This is a clear indication that surface catalysis was

effectively addressed with RuO2 treatmentiii

. In addition, full sheet RuO2 films have been

fabricated at HNEI using identical process. Tests in a two electrode configuration using

p-type material (CGSe or a-SiC) as working electrode and an optimal RuO2 film as

counter electrode have already shown improved results, i.e. a 500 mV onset potential

reduction when compared to Pt counter electrodeiii

.

d. Assessment of WO3 stability in acidic media

Durability tests were performed by HNEI on reactively sputtered WO3 thin films

in 0.33M H3PO4 electrolyte under a constant potential of 1.6V vs. SCE. Several light

sources were evaluated as an alternative to costly (and short lifespan) Xe-arc bulbs

usually used to mimic AM1.5G illumination. Tungsten bulbs were ruled out as they do



A-15


not emit enough UV. Decision was made to use UV-white LED light bulbs instead.

Under these conditions, a fairly steady photocurrent density of 1.5 mA/cm2 was recorded

on five WO3 samples over 1,200 hours. As this current density represents only half of

that usually measured under simulated AM1.5G illumination, the time of operation

reported was therefore divided by a factor of two, i.e. 600 hours, to be consistent with the

amount of coulombs that would pass through a WO3 hybrid system generating hydrogen

at a STH of 3.2% (benchmark).

e. Integration of WO3 into a monolithic device

Since hydrogen production using a standalone device is our main goal, efforts

have been made to combine both a-Si tandem solar cells and tungsten trioxide PEC

electrode in one monolithically integrated device. The main barrier in this task lies in the

process temperature incompatibility between the engine (a-Si) and the PEC material

(WO3). HNEI demonstrated via temperature cycling that WO3 couldn’t be integrated in

traditional hybrid structure where the PEC material is deposited on top of the PV system,

as a-Si solar cells degrades rapidly when exposed to 300C for 3 hours (WO3 sputtering

deposition parameters). The solution resides in a bifacial integration, where the tungsten

oxide film is deposited first on the front side of the transparent conductive oxide (TCO)

substrate (performed at HNEI) followed by the deposition of the solar cell on the

backside (performed at MVS). First monolithic devices were successfully fabricated in

2011 and tested under out-door conditions. Although, the overall STH efficiency (1.5%)

was lower than that obtained with mechanical stacks (3%), this demonstration proved that

the concept was feasible. It is worth mentioning that the limiting factor in this integration

scheme is no longer the a-Si material but the TCO substrate which can withstand

temperatures up to 550C. This makes the bifacial approach compatible with numerous

PEC materials, including I-III-VI2 (e.g. CuGaSe2).

Future of tungsten trioxide in PEC research:

With an optical bandgap of 2.6 eV, tungsten trioxide theoretical STH cannot

exceed 6% and will not reach the target fixed by DOE. Efforts have been done in the past

to decrease WO3 bandgap using foreign elements incorporation but effective reduction

has not been reported yet. Although not optimum as core solar absorber, it still attracts

lots of interests, mainly due to its good transport properties and resistance to corrosion.

Tungsten trioxide has been the workhorse of PEC since the early ages of photochemistry.

Lots of new concepts have been developed on this model system before being

implemented to other PEC materials. Examples include the development of monolithic

bifacial hybrid device recently reported by HNEI and MVS as well as the introduction of

RuO2 material as a counter electrode in photocathode-based systems. Several papers have

been also published on the use of WO3 in dye sensitized solar cellsxi

, PEC dual photo-

electrode systemsii and for methanol oxidation

xii. In the context of pure water splitting,

tungsten trioxide could serve as an excellent capping layer sitting on top of a more

efficient (yet non-resistant to corrosion) photoanode. The later is supported by the fact

that WO3 is an excellent oxygen evolution catalyst.



A-16


References

i E. Miller, R. Rocheleau, X. Deng, “Design Considerations for a Hybrid Amorphous Silicon /

Photoelectrochemical Multijunction Cell for Hydrogen Production”, Int. J. Hydrogen Energy, 2003, 28(6),

615-623. ii Wang, H.; Deutsch, T.; Turner, J. A. (2008). "Direct Water Splitting Under Visible Light with a

Nanostructured Hematite and WO3 Photoanodes and a GaInP2 Photocathode". J. Electrochem. Soc., 155,

F91 (2008). iii

N. Gaillard, Y. Chang, J. Kaneshiro, A. Deangelis and E. L. Miller, “Status of Research on Tungsten

Oxide-based Photoelectrochemical Devices at the University of Hawai’i”, Proc. SPIE, Vol. 7770, 77700V;

doi:10.1117/12.860970 (2010). iv

Fuel Cell Technologies Program Multi-Year Research, Development and Demonstration Plan, updated

September 2011, US DOE, Office of Energy Efficiency and Renewable Energy, Page 3.1 – 27,

http://www1.eere.energy.gov/hydrogenandfuelcells/mypp/index.html v M.N. Huda, Y. Yan, C-Y. Moon, S-H. Wei, and M.M. Al-Jassim: Density-functional theory study of the

effects of atomic impurity on the band edges of monoclinic WO3. Phys. Rev. B 77, 195102 (2008). vi B. Cole, B. Marsen, E.L. Miller, Y. Yan, B. To, K. Jones, and M.M. Al-Jassim: Evaluation of nitrogen

doping of tungsten oxide for photoelectrochemical water splitting. J. Phys. Chem. C 112, 5213 (2008). vii R. Solarska, B.D. Alexander, A. Braun, R. Jurczakowski, G. Fortunato,

M. Stiefel, T. Graule, J. Augustynski, “Tailoring the morphology of WO3 films with substitutional cation

doping: Effect on the photoelectrochemical properties”, Volume 55, Issue 26, 1 November 2010, Pages

7780–7787 viii

Gaillard, N.; Cole, B.; Marsen, B.; Kaneshiro, J.; Miller, E. L.; Weinhardt, L.; Bar, M.; Heske, C.

“Improved current collection in WO3:Mo/WO3 bilayer

photoelectrodes”, J. Mater. Res. 2010, 25, 1–7. ix

M. Bär, L. Weinhardt, B. Marsen, B. Cole, N. Gaillard, E. Miller, and C. Heske, “Mo incorporation in

WO3 thin film photoanodes: Tailoring the electronic structure for photoelectrochemical hydrogen

production”, Appl. Phys. Lett. 96, 032107 (2010). x L. Weinhardt, M. Blum, M. Bar, C. Heske, B. Cole, B. Marsen, and E.L. Miller: Electronic surface level

positions of WO3 thin films for photoelectrochemical hydrogen production. J. Phys. Chem. C 112, 3078

(2008). xi

Satyen K. Deb, “Opportunities and challenges in science and technology of WO3 for electrochromic and

related applications”, Solar Energy Materials & Solar Cells 92 (2008) 245–258. xii

D. V. Esposito, J. G. Chen, R. W. Birkmire, Y. Chang, N. Gaillard, “Hydrogen production from photo-

driven electrolysis of biomass-derived oxygenates: A case study on methanol using Pt-modified WO3 thin

film electrodes“, Int. J. Hydrogen Energ. 36, 9632 (2011).



A-17

http://www1.eere.energy.gov/hydrogenandfuelcells/mypp/index.html

http://www.sciencedirect.com/science/journal/00134686/55/26

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Bar%2C+M.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Weinhardt%2C+L.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Marsen%2C+B.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Cole%2C+B.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Gaillard%2C+N.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Miller%2C+E.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://scitation.aip.org/vsearch/servlet/VerityServlet?KEY=FREESR&possible1=Heske%2C+C.&possible1zone=author&maxdisp=25&smode=strresults&aqs=true

http://link.aip.org/link/applab/v96/i3/p032107/s1&KEY=FREESR



PEC White Papers: MoS2 for PEC hydrogen production

Molybdenum Disulfide for Photoelectrochemical Water Splitting Z. Chen, J.D. Benck, and T.F. Jaramillo, Department of Chemical Engineering, Stanford University

Introduction

Photoelectrochemical (PEC) water splitting provides a promising method for producing hydrogen from

water using the energy from sunlight. However, there currently exists no material system that can perform

this process efficiently and economically.

Molybdenum disulfide (MoS2) is a promising candidate photocathode material for PEC water splitting.

MoS2 is a dichalcogenide semiconductor with a bulk band gap of ~1.2 eV that occurs naturally in n-type

and p-type forms.1 The crystal structure of MoS2 consists of parallel S-Mo-S planes bound loosely

through van der Waals forces.1 This material is advantageous because it is composed of abundant

elements, can be synthesized using inexpensive processes, and is stable to corrosion at reductive

potentials.1, 2

Significant foundational work assessing the viability of MoS2 as a photoelectrode was performed decades

ago by leading researchers in the field of photoelectrochemistry. Seminal studies by Tributsch and

Bennett in 1977 showcased the photoactivity of n-MoS2 as well as p-MoS2.3 Studies by Kautek and

Gerischer,4 Gobrecht et al.,5 as well as Schneemeyer and Wrighton6 further characterized various

properties of MoS2, including its flat-band potential and band structure.

Several groups have also investigated the photocorrosion of MoS2. Fujishima performed studies of MoS2

mounted in a rotating ring disk electrode geometry and showed that photogenerated holes in n-MoS2

kinetically favor the corrosion of the surface into SO42- and Mo6+ species rather than water oxidation,

while photogenerated electrons drive H+ reduction into H2.7 However, the addition of a I-/I2 redox couple

could stabilize the surface and also shift the position of the flat-band potential. Similarly, Kubiak et al.

showed that the addition of Cl- and Br- in solution at high concentrations could limit corrosion.8

Further work focused on characterizing the performance of micro- or nanostructured MoS2

photoelectrodes. Kiesewetter et al. attempted to incorporate microcrystals of MoS2 as the absorber in a

dye sensitized solar cell with a TiO2 support and I-/I3- redox couple.9 However, this approach yielded

relatively low activity compared to bulk MoS2. Since nanostructured materials often have a higher

density of surface defect sites, this result was consistent with previous studies by Kline et al., who

demonstrated that highly structured MoS2 with surface steps produced lower photoactivity.10 The cause

was found to be due to a high rate of recombination at these surface steps, as elucidated by Furtak and

Parkinson in studies of other layered chalcogenide materials such as WSe2.11, 12

Technical Challenges

To successfully incorporate MoS2 into a functional and economical water splitting photocathode, several

materials requirements established for a single-band gap PEC device must be met.13-16 Designing

nanostructures to tune the properties of MoS2 to meet these requirements presents the following technical

challenges:



A-18


1. Catalysis

MoS2 must efficiently catalyze the hydrogen evolution reaction to minimize efficiency losses due to

kinetic overpotential. The bulk form of MoS2 consists primarily of extended flat basal planes of S-

Mo-S layers, which serve as poor catalysts for hydrogen evolution1 because proton adsorption is

energetically unfavorable. In contrast, the edge sites of the basal planes are highly active for hydrogen

evolution.17, 18 Therefore, the design of nanostructures that maximizes the exposure of active edge

sites is an important technical challenge for improving the performance of MoS2.

2. Band Gap

The band gap of MoS2 must be sufficiently large to overcome the thermodynamic and kinetic energy

requirements for splitting water. The band gap of bulk MoS2 is approximately 1.2 eV, 19, 20 which is

too small to split water without an additional potential bias. An internal (rather than external) bias

could potentially be provided by utilizing a vertically stacked tandem device structure with a separate

absorber. Nanostructuring is another approach that can be used to increase the band gap of MoS2

through quantum confinement. Further work is required to apply this strategy to appropriate MoS2

morphologies and to incorporate quantum confined MoS2 structures into a PEC water splitting

device.20, 21 In the absence of a sufficient band gap to drive unassisted water splitting, MoS2 may still

be incorporated into a multi-absorber configuration with a matching photoanode. In either case

(single absorber or multi-absorber), voltage losses from the band gap due to factors such as

insufficient light trapping and non-radiative recombination need to be minimized. The former

challenge arises from the highly reflective nature of a flat MoS2 surface, whereas the latter

recombination arises from the indirect band gap nature of bulk MoS2.

3. Charge Carrier Transport and Conversion Efficiency

Excited charge carriers must reach the semiconductor-electrolyte interface prior to recombination.

Due to its layered structure, mobility in MoS2 is highly anisotropic. While charge transport along S-

Mo-S layers is rapid (~200 cm2 V-1 s-1), mobility perpendicular to the S-Mo-S layers is more than

2000 times lower.1 Therefore, slow transport of excited charge carriers can limit device performance

in bulk MoS2. Nanostructured MoS2 can overcome this limitation by reducing the distance charge

carriers must travel before reaching the semiconductor surface, minimizing resistive as well as bulk

recombination losses. However, nanostructuring may also lead to increased surface recombination,

which can decrease carrier collection efficiency.22 Strategies such as surface passivation may be

required to avoid this potential pitfall while maintaining the benefits of decreased carrier collection

path lengths in nanostructured MoS2 photoelectrodes.

4. Device Configuration

While nanostructuring MoS2 provides many benefits for catalysis, band structure, and charge

transport, it also imposes some additional challenges for integrating this material into a functional

PEC water splitting electrode. Due to their small size, some MoS2 nanostructures have absorption

path lengths that are too short to capture a large fraction of incident sunlight, as is necessary to

achieve high energy conversion efficiency. Therefore, the development of device configurations that

maximize light absorption while maintaining the benefits of nanostructuring is crucial. In particular, it

may be necessary to combine MoS2 nanostructures with high aspect ratio support architectures,

incorporate light trapping structures, or utilize plasmonic effects to enhance light absorption.23-25 This

may also necessitate the development of novel synthesis techniques for MoS2. In the past, the

synthesis of MoS2 typically required treatment in H2S gas at high temperatures (>400°C).18, 26-28 The

presence of H2S sulfidizing agent at these temperatures is extremely corrosive, which prevents the

integration of MoS2 into many potential support structures.



A-19


5. Stability

MoS2 is not stable in alkaline (base) electrolytes. However, it is quite stable in acidic electrolytes,

and is commonly characterized in concentrated sulfuric acid.29, 30 Furthermore, while MoS2 corrodes

at oxidative potentials,1, 31 it is highly stable at the potentials required for hydrogen evolution, i.e.

more negative than 0 V vs RHE, and does not anodically corrode until potentials more positive than

0.5 V vs. RHE.29, 30 For some potential device configurations, it may be necessary to develop

strategies for stabilizing or protecting the MoS2 surface to prevent corrosion at oxidative potentials.

Research Status

Although the charge transport properties of MoS2 are not ideal in bulk form due to the anisotropic

conductivity of its layered structure,1 core-shell MoO3-MoS2 nanowires demonstrate that thin conformal

MoS2 films of only a few nanometers in thickness exhibit negligible ohmic resistance due to the small

length scale for carrier transport. Additionally, the conformal ultrathin MoS2 shell completely protects the

high aspect ratio MoO3 core architecture, which by itself is unstable and rapidly corrodes in strong acids

and at cathodic potentials. This structure remained stable in strong acid for more than 10,000 simulated

diurnal cycles.29 The stability exhibited by core-shell nanostructures opens the opportunity to study thin

conformal layers of MoS2 as a simultaneous stabilizing agent and highly efficient electrocatalyst for

photocathodes.32

Significant advances have been made in developing novel synthetic procedures for MoS2 that broaden the

parameter space for its integration into device morphologies. Recent MoS2 nanostructures are synthesized

at atmospheric pressure at considerably lower (200°C) temperatures,29 enabling the use of oxide

substrates such as fluorine-doped tin oxide (FTO). Highly active electrocatalysts of amorphous MoSx

have also been synthesized at room temperature under atmospheric pressure or low vacuum without

employing any H2S, which may enable the facile deposition of this material onto many types of

substrates.33-35

Approaches

While some properties of bulk MoS2, including its band gap and catalytic activity, are insufficient to

enable unassisted water splitting, nanostructuring provides a means to enhance this material’s favorable

properties while simultaneously mitigating its deficiencies. Nanostructured MoS2 has the potential to

satisfy all the requirements necessary to create a practical PEC photocathode.

The recent development of novel MoS2 nanostructures with unique electrocatalytic, electronic, and optical

properties has highlighted the potential of MoS2 for integration into PEC water splitting devices. Various

MoS2 nanostructures have demonstrated excellent activity for hydrogen evolution electrocatalysis. These

structures include core-shell nanowires of MoO3-MoS2,29 mesoporous MoS2 thin films, and

nanoparticulate MoS2 thin films. All of these nanostructures increase the density of active edge sites in

contact with the electrolyte, and therefore act as highly efficient catalysts for hydrogen evolution,

surpassing the activity of MoS2 nanoclusters studied previously18 and achieving solar-to-hydrogen-

relevant current densities of 10 mA/cm2 at ~200 mV overpotential.

Nanostructuring MoS2 also serves as a route to engineer novel semiconducting properties. MoS2

nanoparticles in the regime of < 10 nm exhibit blueshifts in the onset of optical absorption corresponding

to a band gap enlargement characteristic of quantum confinement.20, 36, 37 In 2 – 5 nm diameter

nanoparticles, the band gap of MoS2 can be increased to more than 2 eV,2, 20, 37 which may be sufficient to

split water without an external bias. Recently, single monolayers of exfoliated MoS2 crystals have

exhibited direct band gap photoluminescence which is not observed in the bulk form.38-40

These single

monolayer crystallites further exhibit high charge mobilities of >200 cm2 V-1 s-1 along the (0001) plane,41,



A-20


42 renewing interest in MoS2 as a material for electronic applications.43, 44 These studies show that

nanostructuring provides an effective means for tuning the band structure of MoS2.

Although some of the technical challenges associated with the development of nanostructured MoS2

photocathodes have been thoroughly addressed in recent studies, more progress is required to make this

material viable for PEC water splitting. In particular, the limitations of charge transport, conversion

efficiency, and device configuration remain major challenges.

The precise mechanisms that lead to charge carrier recombination in nanostructured MoS2 are not well

understood. Crystallographic defects such as grain boundaries, screw dislocations, atomic vacancies,

interstitial atoms, or under-coordinated surface sites could create mid-band gap energy states that act as

recombination centers. A recent study suggested that the edge sites responsible for the high catalytic

activity of MoS2 may also serve as recombination centers.45 Further work using techniques such as time-

resolved photoluminescence46, 47 will be necessary to elucidate the nature of these various factors and

enable improved material design.

References 1. Tributsch, H.; Bennett, J. C. J. Electroanal. Chem. 1977, 81, (1), 97-111.

2. Wilcoxon, J. P.; Samara, G. A. Phys. Rev. B 1995, 51, (11), 7299-7302.

3. Tributsch, H.; Bennett, J. C. J. Electroanal. Chem. 1977, 81, (1), 97-111.

4. Kautek, W.; Gerischer, H. Ber. Bunsen-Ges. Phys. Chem. Chem. Phys. 1980, 84, (7), 645-653.

5. Gobrecht, J.; Tributsch, H.; Gerischer, H. J. Electrochem. Soc. 1978, 125, (12), 2085-2086.

6. Schneemeyer, L. F.; Wrighton, M. S. J. Am. Chem. Soc. 1979, 101, (22), 6496-6500.

7. Fujishima, A.; Noguchi, Y.; Honda, K.; Loo, B. H. Bull. Chem. Soc. Jpn. 1982, 55, (1), 17-22.

8. Kubiak, C. P.; Schneemeyer, L. F.; Wrighton, M. S. J. Am. Chem. Soc. 1980, 102, (22), 6898-

6900.

9. Kiesewetter, T.; Tomm, Y.; Turrion, M.; Tributsch, H. Sol. Energy Mater. Sol. Cells 1999, 59,

(4), 309-323.

10. Kline, G.; Kam, K. K.; Ziegler, R.; Parkinson, B. A. Solar Energy Materials 1982, 6, (3), 337-

350.

11. Parkinson, B. A.; Furtak, T. E.; Canfield, D.; Kam, K. K.; Kline, G. Faraday Discussions 1980,

70, 233-245.

12. Furtak, T. E.; Canfield, D. C.; Parkinson, B. A. J. Appl. Phys. 1980, 51, (11), 6018-6021.

13. Bak, T.; Nowotny, J.; Rekas, M.; Sorrell, C. C. Int. J. Hydrog. Energy 2002, 27, (10), 991-1022.

14. Aroutiounian, V. M.; Arakelyan, V. M.; Shahnazaryan, G. E. Sol. Energy 2005, 78, (5), 581-592.

15. Nowotny, J.; Sorrell, C. C.; Sheppard, L. R.; Bak, T. Int. J. Hydrog. Energy 2005, 30, (5), 521-

544.

16. Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q. X.; Santori, E. A.; Lewis,

N. S. Chem. Rev. 2010, 110, (11), 6446-6473.

17. Hinnemann, B.; Moses, P. G.; Bonde, J.; Jorgensen, K. P.; Nielsen, J. H.; Horch, S.;

Chorkendorff, I.; Norskov, J. K. J. Am. Chem. Soc. 2005, 127, (15), 5308-5309.

18. Jaramillo, T. F.; Jorgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I. Science

2007, 317, (5834), 100-102.

19. Coehoorn, R.; Haas, C.; Dijkstra, J.; Flipse, C. J. F.; Degroot, R. A.; Wold, A. Physical Review B

1987, 35, (12), 6195-6202.

20. Wilcoxon, J. P.; Newcomer, P. P.; Samara, G. A. J. Appl. Phys. 1997, 81, (12), 7934-7944.

21. Chen, Z.; Kibsgaard, J.; Jaramillo, T. F. In Nanostructuring MoS2 for photoelectrochemical

water splitting, 2010; Hicham, I.; Heli, W., Eds. SPIE: 2010; p 77700K.

22. Osterloh, F. E. Chem. Soc. Rev. 2013.

23. Zhu, J.; Hsu, C. M.; Yu, Z. F.; Fan, S. H.; Cui, Y. Nano Lett. 2010, 10, (6), 1979-1984.



A-21


24. Zhu, J.; Yu, Z. F.; Burkhard, G. F.; Hsu, C. M.; Connor, S. T.; Xu, Y. Q.; Wang, Q.; McGehee,

M.; Fan, S. H.; Cui, Y. Nano Lett. 2009, 9, (1), 279-282.

25. Thomann, I.; Pinaud, B. A.; Chen, Z.; Clemens, B. M.; Jaramillo, T. F.; Brongersma, M. L. Nano

Letters 2011, 11, (8), 3440-3446.

26. Margulis, L.; Salitra, G.; Tenne, R.; Talianker, M. Nature 1993, 365, (6442), 113-114.

27. Zak, A.; Feldman, Y.; Alperovich, V.; Rosentsveig, R.; Tenne, R. J. Am. Chem. Soc. 2000, 122,

(45), 11108-11116.

28. Bonde, J.; Moses, P. G.; Jaramillo, T. F.; Norskov, J. K.; Chorkendorff, I. Faraday Discuss.

2008, 140, 219-231.

29. Chen, Z.; Cummins, D.; Reinecke, B. N.; Clark, E.; Sunkara, M. K.; Jaramillo, T. F. Nano Lett.

2011, 11, (10), 4168-4175.

30. Li, Y. G.; Wang, H. L.; Xie, L. M.; Liang, Y. Y.; Hong, G. S.; Dai, H. J. J. Am. Chem. Soc. 2011,

133, (19), 7296-7299.

31. Kautek, W.; Gerischer, H. Surf. Sci. 1982, 119, (1), 46-60.

32. Zong, X.; Yan, H.; Wu, G.; Ma, G.; Wen, F.; Wang, L.; Li, C. J. Am. Chem. Soc. 2008, 130, (23),

7176-7177.

33. Merki, D.; Fierro, S.; Vrubel, H.; Hu, X. L. Chem. Sci. 2011, 2, (7), 1262-1267.

34. Merki, D.; Hu, X. Energy & Environmental Science 2011, 4, (10), 3878-3888.

35. Benck, J. D.; Chen, Z.; Kuritzky, L. Y.; Forman, A. J.; Jaramillo, T. F. ACS Catalysis 2012,

1916-1923.

36. Wilcoxon, J. P.; Samara, G. A. Phys. Rev. B, Condens. Matter 1995, 51, (11), 7299-7302.

37. Thurston, T. R.; Wilcoxon, J. P. J. Phys. Chem. B 1999, 103, (1), 11-17.

38. Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Phys. Rev. Lett. 2010, 105, (13).

39. Splendiani, A.; Sun, L.; Zhang, Y.; Li, T.; Kim, J.; Chim, C.-Y.; Galli, G.; Wang, F. Nano Lett.

2010, 10, (4), 1271-1275.

40. Eda, G.; Yamaguchi, H.; Voiry, D.; Fujita, T.; Chen, M.; Chhowalla, M. Nano Lett. 2011, 11,

(12), 5111-5116.

41. Fivaz, R.; Mooser, E. Physical Review 1967, 163, (3), 743-755.

42. Mansfield, R.; Salam, S. A. Proceedings of the Physical Society of London Section B 1953, 66,

(401), 377-385.

43. Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A. Nature Nanotech. 2011, 6,

(3), 147-150.

44. Yoon, Y.; Ganapathi, K.; Salahuddin, S. Nano Lett. 2011, 11, (9), 3768-3773.

45. Chen, Z.; Forman, A. J.; Jaramillo, T. F. The Journal of Physical Chemistry C 2013, 117, (19),

9713-9722.

46. Doolen, R.; Laitinen, R.; Parsapour, F.; Kelley, D. F. J. Phys. Chem. B 1998, 102, (20), 3906-

3911.

47. Parsapour, F.; Kelley, D. F.; Craft, S.; Wilcoxon, J. P. Journal of Chemical Physics 1996, 104,

(13), 4978-4987.



A-22

PEC White Papers: Engineered Oxide Minerals

Engineered ternary and quaternary oxide minerals with optimal absorption characteristics for solar-assisted low-cost hydrogen production.

Nicolas Gaillarda, Muhammad N. Huda

b

a. Hawaii Natural Energy Institute

b. The University of Texas in Arlington

Synopsis:

Four decades after the demonstration of photoelectrochemical (PEC) water splitting by

Fujishima and Honda with TiO2i, intensive research is still ongoing to identify a suitable

semiconductor to be integrated in an efficient, cost effective, and reliable PEC system. Among

all candidates, binary transition metal oxides are still drawing lots of attention as most of them

offer good resistance to corrosion and are inexpensive to produce. However, no system having

both appropriate optical absorption and good transport properties have been discovered yet. In

the case of wide bandgap materials (i.e. TiO2 and WO3), numerous attempts have been made to

narrow their bandgaps, mainly via incorporation of foreign element such as nitrogenii.

Unfortunately, this method usually leads to an increase in structural defects and poor PEC

performancesiii. In contrast, hematite (-Fe2O3) is a material class that already has the right

absorption characteristics for solar-powered water splitting. However, hematite falls short on

electrical performances, with hole-diffusion length in the order of 20 nmiv

, though significant

progress in Fe2O3 nano-structuring have been made to resolve this issuev,vi

. From this point of

view, it appears that the one feasible research strategy might be to focus on metal oxides owning

appropriate absorption properties and improve, if necessary, their transport and/or catalytic

properties.

Nowadays, the optical and electronic properties of all existing binary systems have been

studied and are well documented. To the best of our knowledge, only two binary metal oxides

can fulfill the requirements in term of band gap for PEC applications: Fe2O3 (2.0 eV) and Cu2O

(1.95 eV). In the latter case, recent work by Pr. Grätzel (EPFL) shows that cuprous oxide can

generate a photocurrent up to 7 mA/cm2 at saturation, making this material the best binary oxide

ever synthesized for solar conversion applicationsvii

. The only issue with such system is its

thermodynamic stability. Indeed, it is possible that, over time and under operation, cuprous oxide

(Cu2O) turns into cupric oxide (CuO), which bandgap (1.3-1.6 eV) is far from ideal for PEC

applications. In the light of these examples, it appears that one need to search beyond binary

oxides to find the ideal material for low-cost solar-assisted hydrogen production.

With the increased number of elements in ternary and quaternary compounds comes the

difficulty to design a unique system capable of satisfying optical absorption, electronic

conductivity and resistance to corrosion. One approach resides in pure combinatorial analysis,

where sets of selected elements are automatically blended to form series of compounds. Though

rather challenging to perform with vacuum-based processes (e.g. sputtering), recent progress in

solution-based (ink) have greatly improved combinatorial-based PEC research and led to the



A-23


emergence of interesting new PEC systems[Parkinson]. However, one should bear in mind that

the success of such approach should not rely only on opportune material discovery. Similar

combinatorial approach has been employed in the past in the field of superconductors to

investigate compounds containing multiple cations metal-oxides. While material discoveries in

this field used to be quite “serendipitous”viii

, modern density functional theory calculation has

since permitted to predict trends to design new superconducting material classes.

The ability to design new ternary and quaternary oxide minerals for PEC

applications guided by density functional theory predictions defines the objective of this

subtask.

Research Status:

a. Copper tungstate

With an electronic band-gap of 2.2 eV and optimum surface energetics for water

splittingix,x

, copper tungstate (CuWO4) is a promising material-class that merits further

investigation. Although fairly new in the field of PEC, metal tungstates have been investigated

for more than three decades for their scintillation characteristics. A large number of papers have

been published on the microstructure and optical characteristics of this material class and metal

tungstate-based scintillators are widely used in the filed of detection. As an example, calcium

tungstate currently equipped the Dark Matter Cryogenic detector at CRESST in Gran Sasso in

Italy. Cadmium tungstate is widely used for positron emission tomography in medical diagnosis.

Finally, the Electromagnetic Calorimeter detector at the Large Hadron Collider (CERN) is made

of approximately 100,000 lead tungstate crystals. In this material class, copper tungstate is the

only material with a bandgap (2.2 eV) that is appropriate for PEC applications. However, less

than ten papers have been published specifically on the photoelectrochemical properties of

CuWO4 since Benko’s (Brock University) first report in 1982.xi

Mott-Schottky analyses reported

in 1990 by Arora (Sardar Patel University) on single crystals evidenced that CuWO4 valence

band maximum (VBM) and conduction band minimum (CBM) straddled the oxygen (OER) and

hydrogen evolution reaction (HER) potentials, respectively, an ideal situation for solar-assisted

water splittingxv

. The first thin film CuWO4 photoelectrode fabrication was reported in 2005 by

Pandey (Nagpur University) using spray-deposition methodxii

. In an attempt to reduce CuWO4

optical band gap, Chen (NREL) developed a low-temperature co-sputtering process to synthesize

amorphous CuxW(1-x)O4 thin filmsxiii

. Although copper tungstate absorption characteristics were

tunable, amorphous films presented limited photo-response to visible light and were un-stable in

aqueous solution. In 2011, Yourey (Michigan) published encouraging results on electrodeposited

CuWO4xiv

and CuWO4-WO3xv

composite systems. The authors demonstrated copper tungstate’s

stability in a 0.1M Na2SO4 solution containing 10% methanol over 12 hours at 0.5 V vs.

Ag/AgCl. Finally, Chang (HNEI) reported recently on the effect of thermal treatment on the

crystallographic, surface energetics and PEC properties of reactively co-sputtered CuWO4ix

. It

was observed that thin films fabricated at temperature below 300C had p-type conductivity and

exhibited no significant photoresponse when exposed to Air Mass 1.5Global (AM1.5G) irradiation.

However, a major improvement was observed after a post-annealing at 500C in argon for 8

hours, exhibiting a photocurrent density of approx. 400A/cm2 at 1.6V vs. SCE. More



A-24


importantly, a surface band diagram drawn from electrochemical and optical measurements

confirmed Arora’s results on the ideal positions of CBM and VBM for solar-assisted water

splitting. Electrochemical impedance study also indicated that CuWO4 transport properties must

be improved in order to achieve better performing photoanodes.

b. Down-selection of new oxide minerals via DFT calculation

It has been mentioned earlier that the introduction of impurities for doping would create

unwanted defect-states in the band gap, which, in turn, would be detrimental to the crystallinity

and the transport properties of the host materials. In addition, the presence of highly localized

orbitals at the band edges (such as at the valence band maxima or the conduction band minima of

Mott or charge transfer-type insulators), results in high effective mass charge-carriers, and hence

high electron-hole recombination rate. On the other hand, a thermodynamically stable alloy

structure with desirable materials compositions can provide better crystallinity and, hence

improved charge transport properties.

Predicting novel alloys of multi-cation metal oxides satisfying all the electronic criteria of

an efficient photocatalyst is a challenge. Nonetheless, recent advances in computers and

theoretical methods made it possible to investigate and predict materials at a fundamental level,

which were not possible before. In addition, an efficiently combined experiment-theory approach

can expedite the prediction and characterization processes further. With our new theoretical

approach for materials design by searching mineral data-base, we were able to predict several

new tungstate-based multi-cation metal-oxides. For example, from a combined mineral data-

based search and DFT calculations, we were able to predict so far unknown crystal structure of

AgBiW2O8, and the corresponding electronic properties such as band structure, density of states,

optical absorptions, etc. These results were then compared with the experimental findings. We

recently have predicted the crystal structure of a new alloy, namely CuBiW2O8, which has a

lower band gap and a higher valence band maximum than AgBiW2O8. In contrast to WO3, both

of these two materials straddle the hydrogen and oxygen reduction potentials. The theoretical

XRD plot and other electronic and optical properties will be calculated for the newly predicted

material to guide the experimental findings. For example, a predicted XRD plot for CuBiW2O8 is

shown in the following:

References:

Figure: Calculated XRD plot for

CuBiW2O8, a new tungstate based oxide

predicted by searching mineral database

followed by DFT calculations.



A-25


i Fujishima, A., and Honda, K., Nature, 238, 37–38 (1972).

ii M. Huda, Y. Yan, C.-Y. Moon, S.-H. Wei, M. Al-Jassim, Physical Review B 77, 1 (2008). iii B. Cole, B. Marsen, E. Miller, Y. Yan, B. To, K. Jones, M. Al-Jassim, J. Phys. Chem. C 112, 5213

(2008). iv

Dare-edwards, M. P.; Goodenough, J. B.; Hamnett, A.; Trevellick, P. R. J. Chem. Soc.,

Faraday Trans. 1983, 79, 2027-2041. v Sivula, K.; Formal, F. L.; Gratzel, M. ChemSusChem. 2011, 18, 432-449.

vi Vayssieres, L.; Sathe, C.; Butorin, S.; Shuh, D.; Nordgren, J.; Guo, J. Adv. Mater. 2005, 17,

2320-2323. vii

A. Paracchino; V. Laporte; K. Sivula; M. Grätzel; E. Thimsen. Nature Materials 10, 456–461

(2011). viii

Basic Research Needs for Supraconductivity, Report of the Basic Energy Sciences, 2006. ix

Chang, Y.; Braun, A.; Deangelis, A.; Kaneshiro, J.; Gaillard, N. J. Phys. Chem. C, 2011, 115,

25490–25495. x Arora, S. K.; Mathew T.; Batra, N. M. J. Phys. D: Appl. Phys. 1990, 23, 460-464.

xi Benko, F. A.; MacLaurin, C. L.; Koffyberg, F. P. Mat. Res. Bull. 1982, 17, 133-136.

xii Pandey, P. K.;Bhave, N. S.;Kharat, R. B. Mater. Lett. 2005, 59, 3149-3155.

xiii Chen, L.; Shet, S.; Tang, H.; Ahn, K.; Wang, H.;Yan, Y.; Turner, J.; Jassim, M. A. J. Appl.

Phys. 2010, 108, 043502 1-5. xiv

Yourey J. E.; Bartlett, B. M. J. Mater. Chem. 2011, 21, 7651-7660. xv

Joseph E. Yourey, Joshua B. Kurtz, and Bart M. Bartlett, J. Phys. Chem. C, 2012, 116, 3200–

3205.



A-26

http://www.ncbi.nlm.nih.gov/pubmed/21416621

PEC White Papers: BiVO4 for PEC hydrogen production

BiVO4 as a Photoanode for Photoelectrochemical Water Splitting

Kyoung-Shin Choi, Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA

Roel van de Krol, Institute for Solar Fuels, Helmholtz-Zentrum Berlin für Materialien und Energie,

Berlin, Germany

Introduction

Bismuth vanadate (BiVO4) has recently emerged as a promising material for use as a photoanode in

water-splitting photoelectrochemical cells.[1,2] It has a bandgap of 2.4 eV with the valence band (VB)

edge located at ca. 2.4 V vs. RHE (reversible hydrogen electrode), providing sufficient overpotential for

holes to photooxidize water while the conduction band (CB) edge is located just short of the

thermodynamic level for H2 production.[3]

Its bandgap is slightly larger than is desired for a photoanode

(ca. 2.0 eV) but its very negative CB position may compensate for this disadvantage as not many n-type

semiconductors that can utilize visible light have a CB edge position that is as negative as that of

BiVO4.[4,5] In addition, it is composed of only non-precious elements ensuring commercial viability for

practical use in large quantities.

In general, oxide semiconductors have wide bandgaps because the VB has mainly the O 2p character and

is located at a very positive potential.[6,7] Therefore, the production of oxide semiconductors that can

absorb a significant portion of visible–light requires decreasing their bandgap energies by raising the VB

edge. One viable strategy to achieve this is to introduce cations with occupied low binding energy s

orbitals such as Bi3+ and Sn2+.[6] BiVO4 is an exemplary case for this approach in that the hybridization of

the filled Bi3+ 6s2 state and O 2p states at the top of the VB effectively shifts the VB edge to the negative

direction, reducing the bandgap energy while its CB edge remains at a relatively negative position

compared to those of other oxides having comparable bandgaps (e.g. Fe2O3, WO3).

The use of a BiVO4 photocatalyst for solar oxidation was first reported by Kudo et al. in 1998.[2] Early

work on BiVO4 mainly focused on suspension-type photocatalysts for water oxidation or

photodegradation of organic compounds. Since the CB level does not allow for water reduction, Ag+ was

typically used as a sacrificial electron acceptor for these studies.

More recent studies on BiVO4 have focused on the preparation of BiVO4 as electrode-type materials for

use as photoanodes for photoelectrochemical cells. However, the typical efficiencies of unmodified

BiVO4 photoanodes for water oxidation were not impressive as they suffer from excessive electron-hole

recombination, poor charge transport properties and poor water oxidation kinetics. Therefore, various

strategies such as morphology control, construction of composite structures, doping, and pairing with

oxygen evolution catalysts have been developed recently to alleviate one or more of these limitations.

Further advancement in the construction of efficient BiVO4-based photoanode systems is expected once

the limitations of BiVO4 are better understood and new approaches are developed to effectively overcome

them.

Technical Challenges

Major technical challenges in constructing highly efficient BiVO4 photoanodes are summarized below.

1. Light Management

As mentioned earlier, the bandgap of BiVO4 is slightly larger than is desired for a photoanode.

Therefore, theoretical and experimental studies on composition tuning (i.e. formation of solid

solutions), which aim to reduce the bandgap (ca. 2.0 eV or smaller), are highly desired. To date, no



A-27


studies on composition tuning have resulted in a decrease in bandgap energy. Ideally, the bandgap

reduction should be achieved by elevating the VB edge position without lowering the CB edge

position. Engineering the morphology of BiVO4 electrodes to enhance light absorption is also an

important issue.

2. Recombination losses and poor electron transport properties

The separation yield of photogenerated electron and hole pairs in the BiVO4 system is assessed to be

below 30%,[8,9] making it the main limiting factor for the performance of BiVO4. Therefore, there is

an urgent need for the development of more effective morphologies, compositions (i.e. doping), and

cell structures to reduce recombination losses. Undoped BiVO4 generally shows poor electron

transport properties, which appears to be one of the main reasons for the low electron-hole separation

yield of BiVO4. A few doping studies show a significant increase in the photocurrent of BiVO4.

3. Slow hole transfer kinetics for water oxidation

The surface of BiVO4 is not particularly catalytic for water oxidation and, therefore, bare BiVO4

electrodes that are not coupled with oxygen evolution catalysts (OECs) do not show impressive

performances for photo-oxidation of water. Coupling of efficient OECs with BiVO4 without creating

undesirable interface states at the BiVO4/OEC junction is necessary.

4. Chemical and photoelectrochemical stabilities

BiVO4 is chemically stable in neutral and slightly basic media but it is not stable in strong acidic or

basic media.[10] Under illumination, it is not stable for photo-oxidation of water if it is not coupled

with OECs.[1] It appears that the slow interfacial hole transfer kinetics for water oxidation and the

resulting hole accumulation at the BiVO4/electrolyte interface results in anodic photocorrosion of

BiVO4. However, when BiVO4 is covered with appropriate OECs or hole acceptors with fast

oxidation kinetics are introduced to the electrolyte, anodic photocorrosion of BiVO4 can be

effectively suppressed.[11,12]

5. Complete water splitting to H2 and O2 by BiVO4

Since the CB edge of BiVO4 is located just short of the thermodynamic level for H2, BiVO4 alone

cannot produce H2. Therefore, application of an external bias (e.g., coupling with a PV unit) or

formation of a photoelectrochemical diode by combining an n-type BiVO4 photoanode with a proper

p-type semiconductor (photocathode) is necessary. Since the CB edge of BiVO4 is already very close

to the reduction potential of H2, composition tuning of BiVO4 to shift the CB edge position to the

negative direction may also be possible.

Research Status

Significant advancement in the construction and understanding of efficient BiVO4-based photoanode

systems has been made within a short period of time owing to various newly developed ideas and

approaches. The most frequently used synthesis methods to prepare BiVO4 photoanodes include metal

organic decomposition, chemical bath deposition, urea-precipitation method, hydrothermal synthesis,

spray deposition, and electrochemical synthesis.[1] During synthesis, various efforts to control

morphology, formation of composite structures or heterojunctions, doping or composition tuning, and

coupling with oxygen evolution catalysts (OECs) have been made. As a result, BiVO4 currently shows

the most promising performance for photo-oxidation of water in the low-bias region (< 0.6 V vs. RHE)

among all oxide-based photoanodes studied to date.[1]

Several theoretical studies have been performed to understand the optical and charge transport properties

of BiVO4 and also to identify the role of the dopants incorporated into the BiVO4 structure. However,

some controversies exist in these studies and the conduction mechanisms in BiVO4 and doped BiVO4

have not been thoroughly elucidated.



A-28


Approaches

Formation of Composite Structures Composite photoelectrodes where the main photon absorber is

combined with an additional semiconductor or a conductor are often constructed to enhance the overall

photon absorption, electron-hole separation, or charge transport processes. BiVO4 has been most

frequently coupled with WO3 to form heterojunction structures in order to improve electron-hole

separation. Rapid electron injection from the CB of BiVO4 to the CB of WO3, which is located at a more

positive potential, can physically separate electrons from holes in the VB of BiVO4 and may effectively

reduce their recombination.[13-17] Another interesting composite structure is a BiVO4/SnO2 electrode

where a thin SnO2 layer (ca. 10 nm) is placed between BiVO4 and the FTO substrate.[9,18] A significant

enhancement in IPCE was observed because the SnO2 layer with a very positive VB edge acts as a hole

mirror, meaning holes are “reflected” at the SnO2/BiVO4 junction, presumably preventing electron-hole

recombination at the FTO–related defect states. Without the SnO2 layer, FTO–related defect states

formed at the FTO/BiVO4 interface act as recombination centers. More recently, BiVO4/SnO2/WO3

multi-composite electrodes were also prepared,[16] which demonstrated that the photocurrent observed

from BiVO4 was increased by adding a WO3 layer under the BiVO4 film. The photocurrent was further

increased when a very thin SnO2 layer was inserted between BiVO4 and WO3. To achieve complete

water splitting with BiVO4 in a cost-effective manner, monolithic composite tandem structures with an

integrated PV cell should be explored. The BiVO4 would effectively protect the underlying PV junction

against photocorrosion, which is a problem for devices that are entirely based on III-V materials.[19]

Alternatively, BiVO4 can be deposited on top of a smaller-bandgap n-type semiconductor to form a n-n

heterojunction, in a similar manner as recently demonstrated for Fe2O3.[20]

Doping Studies The most effective dopants identified to date, which increase the carrier

concentration and photoelectrochemical performance of BiVO4, are Mo6+ and W6+ ions that

substitutionally replace V5+ ions.[21-24] No changes in bandgap by Mo or W doping were observed while

the increase in carrier density was confirmed by the decrease in the slope of Mott-Schottky plots. The

results of the first-principles density-functional theory (DFT) calculations, which assumed substitutional

replacement of V by Mo and W, showed that W and Mo serve as shallow donors and can effectively

increase the carrier density in BiVO4.[23] Non-metal element, P, was also doped into BiVO4 by the urea-

precipitation method where PO43- oxoanions were added as a P precursor to replace a small fraction of

VO43- oxoanions in the precursor solution.[25] The EIS measurements suggested that the presence of P

lowered the charge transfer resistance of BiVO4 remarkably.

Coupling with Oxygen Evolution Catalysts In order to improve slow hole transfer kinetics of

BiVO4 for water oxidation, BiVO4 have been coupled with various OECs (e.g. IrOx, Co3O4, Co-

Pi, and Pt).[1]

The most commonly used OEC to improve water oxidation kinetics of BiVO4 or

doped BiVO4 photoanodes has been Co-Pi. Some studies compared electrodeposition and

photodeposition (or photoassisted electrodeposition) of Co-Pi OEC onto the BiVO4 photoanodes

and found that BiVO4/Co-Pi OEC prepared by photodeposition showed a superior performance

for water oxidation.[26-28]

Recently, FeOOH was also identified as an efficient OEC that can

work very well with BiVO4.[11,12]

A few studies noted that the performance of a given OEC

varies significantly depending on the type and the synthesis method used to prepare the

photoanode.[1]

This suggests that the overall performance of a photoanode/OEC is significantly

governed by the photoanode/OEC interface and, therefore, better understanding of this interface

is necessary for further optimization of photoanode/OEC junctions. Morphology Control When a semiconductor material is produced as a polycrystalline electrode,

morphological details of the electrode such as size, shape and connectivity of the particles have a

significant impact on the interfacial energetics, kinetics and charge transport properties.[29] Therefore,

understanding and controlling the morphological aspects of polycrystalline BiVO4 electrodes provides an



A-29


effective way to enhance its photoelectrochemical properties. Atomic plane-dependent

photoelectrochemical properties of powder type-BiVO4 photocatalysts have been reported,[1,30] suggesting

that precise morphology control of BiVO4 electrodes may also be advantageous. Also, construction of

high surface area BiVO4 electrodes that will increase the volume of space-charge region and enhance

electron-hole separation is another effective way to increase the efficiencies of BiVO4 photoanodes.

References

1. Y. Park, K. J. McDonald and K.-S. Choi, Chem. Soc. Rev., 2013, 42, 2321.

2. A. Kudo, K. Ueda, H. Kato and I. Mikami, Catal. Lett., 1998, 53, 229.

3. B. Xie, H. Zhang, P. Cai, R. Qiu and Y. Xiong, Chemosphere, 2006, 63, 956.

4. M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi, E. A. Santori and N. S.

Lewis, Chem. Rev., 2010, 110, 6446.

5. T. Bak, J. Nowotny, M. Rekas and C. C. Sorrell, Int. J. Hydrogen Energy, 2002, 27, 991.

6. A. Walsh, Y. Yan, M. N. Huda, M. M. Al-Jassim and S.-H. Wei, Chem. Mater., 2009, 21,

547.

7. A. Kudo, K. Omori and H. Kato, J. Am. Chem. Soc., 1999, 121, 11459.

8. D. K. Zhong, S. Choi and D. R. Gamelin, J. Am. Chem. Soc., 2011, 133, 18370.

9. F. F. Abdi and R. van de Krol, J. Phys. Chem. C, 2012, 116, 9398.

10. F. F. Abdi, N. Firet, A. Dabirian and R. van de Krol. Mater. Res. Soc. Symp. Proc. 2012,

1446, DOI: 10.1557/opl.2012.811.

11. J. A. Seabold and K.-S. Choi, J. Am. Chem. Soc., 2012, 134, 2186.

12. K. J. McDonald and Choi, K.-S., Energy Environ. Sci., 2012, 5, 8553.

13. J. Su, L. Guo, N. Bao and C. A. Grimes, Nano lett., 2011, 11, 1928.

14. P. Chatchai, Y. Murakami, S.-y. Kishioka, A. Y. Nosaka and Y. Nosaka, Electrochim.

Acta, 2009, 54, 1147.

15. S. J. Hong, S. Lee, J. S. Jang and J. S. Lee, Energy Environ. Sci., 2011, 4, 1781.

16. R. Saito, Y. Miseki and K. Sayama, Chem. Commun., 2012, 48, 3833.

17. K. Zhang, X.-J. Shi, J. K. Kim and J. H. Park, Phys. Chem. Chem. Phys., 2012, 14, 11119.

18. Y. Liang, T. Tsubota, L. P. A. Mooij and R. van de Krol, J. Phys. Chem. C, 2011, 115,

17594.

19. O. Khaselev and J. Turner, Science, 1998, 280, 425

20. M.T. Mayer, C. Du, and D. Wang, J. Am. Chem. Soc., 2012, 134, 12406.

21. H. Ye, J. Lee, J. S. Jang and A. J. Bard, J. Phys. Chem. C, 2010, 114, 13322.

22. W. Luo, Z. Yang, Z. Li, J. Zhang, J. Liu, Z. Zhao, Z. Wang, S. Yan, T. Yu and Z. Zou,

Energy Environ. Sci., 2011, 4, 4046.

23. H. S. Park, K. E. Kweon, H. Ye, E. Paek, G. S. Hwang and A. J. Bard, J. Phys. Chem. C,

2011, 115, 17870.

24. S. P. Berglund, A. J. E. Rettie, S. Hoang and C. B. Mullins, Phys. Chem. Chem. Phys.,

2012, 14, 7065.

25. W. J. Jo, J.-W. Jang, K.-j. Kong, H. J. Kang, J. Y. Kim, H. Jun, K. P. S. Parmar and J. S.

Lee, Angew. Chem. Int. Ed., 2012, 51, 3147.

26. S. K. Pilli, T. E. Furtak, L. D. Brown, T. G. Deutsch, J. A. Turner and A. M. Herring,

Energy Environ. Sci., 2011, 4, 5028.

27. S. K. Pilli, T. G. Deutsch, T. E. Furtak, J. A. Turner, L. D. Brown and A. M. Herring,

Phys. Chem. Chem. Phys., 2012, 14, 7032.



A-30


28. T. H. Jeon, W. Choi and H. Park, Phys. Chem. Chem. Phys., 2011, 13, 21392.

29. K.-S. Choi, J. Phys. Chem. Lett., 2010, 1, 2244.

30. G. Xi and J. Ye, Chem. Commun., 2010, 46, 1893.



A-31

PEC White Papers: Hematite for PEC hydrogen production

Hematite (α-Fe2O3) as a Photoelectrode for Photoelectrochemical Hydrogen Production

Heli Wang, National Renewable Energy Laboratory

Isabell Thomann, Rice University

Arnold J. Forman, Stanford University

Yat Li, University of California at Santa Cruz

Mahendra Sunkara, University of Louisville

Moreno de Respinis, Delft University of Technology

1. Introduction

The photoelectrochemical (PEC) production of hydrogen and oxygen, via water splitting

reactions with photo-generated electrons and holes, was first demonstrated by Fujishima and

Honda at a chemically biased TiO2 with ultraviolet (UV) light.1 Hematite (α-Fe2O3) has several

advantages over other semiconductor materials for this purpose. With a band gap around 2 eV,2,3

it could utilize 40% of the incident solar spectrum. It has an excellent chemical stability in a

broad pH range3 and its valence band (VB) is appropriate for the oxygen evolution reaction

(OER).2-4

Moreover, it is abundant on earth, low in cost, and non-toxic making hematite an

attractive candidate for PEC water splitting. Thus, it was investigated extensively in the late 70’s

and early 80’s,4-9

mostly in the form of bulk, non-textured electrodes.

2. Technical Challenges and Research Status

Despite being one of the few promising materials for large-scale solar hydrogen production,

hematite presents a number of challenges for application in PEC water splitting.

2.1 Energetic mis-match

First, like many n-type oxides, the conduction band (CB) of iron oxide lies 0.2 - 0.4 eV positive

of the hydrogen evolution reaction (HER),3 meaning that an external bias is needed to drive this

reaction, which reduces the overall efficiency. One way to provide this bias is via a multijunction

device,10

in which the bottom photovoltaic (PV) cell/layer provides the necessary bias for the

top PEC layer to drive the water splitting reactions.10-13

Such a bias can also be provided by

combining two photoelectrodes in series, one n-type and one p-type.14-17

This dual-electrode

configuration separates OER to the n- and HER to the p-type photoelectrodes, respectively,

increasing the number of candidate semiconductors that could be used to build a stand-alone

solar water splitting device.16

Alternatively, monolithic devices can be based on a photoanode

biased with an integrated p-n junction.18

A promising approach consists of an n-n heterojunction

PEC device in which a photoanode is deposited onto an n-type semiconductor that boosts the

energy of the electrons.18-21

An alternate strategy to improve the band edge alignment is via

modifications of the semiconductor surface, either with a pH-insensitive group, producing a

surface dipole that is independent of pH, or by introducing a desired surface dipole/charge (see

e.g. [Ref Lewis doi 10.1021/cr1002326]). Another possible way to shift the band edge up would

be via quantum confinement.22

A 0.3-0.6 eV CB shift due to the quantum confinement effect

could potentially locate the conduction band of hematite above the HER energy.



A-32


2.2 Poor photo-response

It is generally accepted that the following are reasons for the poor photo-response of hematite:

(1) Very short minority carrier (hole) diffusion length;4

(2) High resistivity;

(3) Low electron mobility and high recombination rate of photo-generated charge carriers.4,6

(4) Low visible light absorption coefficient

Since these effects are all inter-related, but difficult to probe quantitatively or remediate

synthetically, the mechanisms and processes responsible for the low photo-response of hematite

are far from being clearly understood. A deeper understanding of the semiconducting properties

of the material is certainly needed for improvement. Different approaches, including various

doping and alloying schemes3,7-9

have been attempted to increase the photo-response of iron

oxide. As a result, significant progress in photo-response has been made recently by the research

community.

2.2.1 Short hole diffusion length

For hematite, the minority carrier (hole) diffusion length is <10 nm, resulting in very inefficient

carrier extraction. The recent development of high surface-area nanomaterials opens up new

opportunities in overcoming this limitation from a structural design perspective. Nanostructured

hematite photoanodes provide a much shorter diffusion path for minority carriers to reach the

hematite/electrolyte interface, which can result in increased carrier collection efficiencies as has

been observed in nanostructured hematite particle thin films.23

Thin films based on oriented

nanorod arrays24-27

also enhance the photogenerated carrier transport along the nanorods due to

fewer grain boundaries (possible recombination centers) and a directed electron movement

toward the back contact. The diameter of bundled ultrafine hematite nanorods is typically 4-5

nm,22,24-26

comparable to the hole diffusion length in hematite. Recently, several research groups

have significantly improved the photo-response of hematite where a few milliamps/cm2 of

photocurrent density were obtained at 1.23 VRHE with nanostructured thin films.21,28-36

2.2.2 High resistivity

Poor electrical conductivity is another major hurdle for hematite. Elemental doping has been

demonstrated to be a promising method to potentially address this limitation by significantly

enhancing hematite’s donor density.28,29,37,38

The development of performance-enhancing doping

strategies has recently attracted a lot of attention.3,34,38-41

For example, Si-doping has been

developed by spray pyrolysis and atmospheric pressure chemical vapor deposition

techniques.21,28,29,37

The Si-doped hematite nanostructures achieved a photocurrent of 2.3

mA/cm2 at 1.23 VRHE, without the aid of a cocatalyst.

29 Sn-doped hematite, which can be

achieved through Sn diffusion from an FTO substrate induced by thermal treatment or

intentionally mixing the iron with Sn precursor during hydrothermal growth, has recently shown

substantially improved PEC performance.30,31,35,42

Moreover, Ti-doping can enhance the donor



A-33


density of mesoporous hematite film by nearly two orders of magnitude, resulting in a two-fold

enhancement of photocurrent as compared to undoped films.43

To further improve hematite electrodes, general strategies to understand the limiting factors for

conductivity and photo-activity which are dominated by processes operating on the nanoscale

(both in the bulk and at surfaces) are still required. Furthermore synthetic methods for forming

suitable thin films with controllable composition, doping and proper structures, to optimize

photocurrent and to cathodically shift the photocurrent onset potentials are required.

2.2.3 Recombination

A high charge carrier recombination rate is considered to be due to inefficient carrier diffusion

and slow water oxidation kinetics at the hematite/electrolyte interface. One method addressing

the recombination found in poor crystalline quality materials is to apply an ultra-thin single

crystal hematite film on highly conductive TiSi2 nanonets.36

The thin layers facilitate charge

transport to the back contact, minimize recombination at crystal defects and grain boundaries and

reduce diffusion distances for charge carriers. It was reported that the absorbed photon

conversion efficiency is almost the same as the incident-photon-to-electron conversion efficiency

(IPCE), indicating excellent efficiency of charge separation and collection (quantum yield).

While nanostructuring is likely needed to reduce carrier path lengths to catalytic surfaces, it will

be necessary to optimize carrier dynamics in these nanostructured materials for reduced

recombination.44-46

This will require strategic modification of the hematite surface to minimize

catalytic overpotentials (addition of co-catalysts) and to eliminate surface traps/defects.47-49

For

example, an IrO2 co-catalyst substantially shifted the photocurrent onset potential of Si-doped

hematite to a more negative potential. Other oxygen evolution catalysts such as cobalt ions and

cobalt phosphate have also been developed for PEC water oxidation at hematite.29,50

2.2.4 Low absorption coefficient

If accompanied by high visible light absorption, the short minority carrier diffusion length would

not be a major challenge. Unfortunately, being an indirect band-gap semiconductor, hematite is a

relatively weak absorber of near band-gap photons (∼0.1-1 μm absorption length in the 500-600

nm range),51

resulting in challenges for the optical design of the photoelectrode structures. In

particular, novel strategies are required to increase the IPCE near the band-edge, and to improve

light management and concentration in thin hematite films. One promising strategy may be the

application of metallic plasmonic nanostructures or nanoparticles interspersed within hematite

thin films to boost the effective near band-edge absorption.51

Such a strategy can help to alleviate

the large mismatch between the length scales of absorption and carrier extraction. This could be

done by confining incident sunlight close to the semiconductor/liquid interface, where the space

charge layer can promptly separate the photocarriers, and thereby reducing recombination.52

Plasmon-enhanced photocatalytic activity of hematite has already proven successful with Au, Ag

and Cu nanoparticles.51-53

2.3 (Photo)Corrosion under extended operation conditions



A-34


In terms of the practical application of hematite photoelectrodes for scalable solar hydrogen

production, an improved understanding of extended operational stability will be required. It may

be necessary to develop protective, corrosion resistant coatings which also permit charge transfer

to solid/electrolyte interface.

Moreover, facile and large-scale production methods will be needed for hematite films27,38,54

and

nanophotonic enhancement structures.55

Substrate materials may be optimized to achieve a good

ohmic contact for majority carrier extraction. If metal nanoparticles are to be used for photon

management in large-scale applications, non-precious metals group (n-PMG) materials will be

needed, calling for novel strategies to improve their corrosion stability.

3. Approaches

While different strategies have been explored to overcome the limitations inherent to hematite

photoanodes and encouraging progress has been made, further research is clearly needed to

better understand the fundamental properties and processes of nanostructured hematite and

limiting factors. The complex nature of the system requires combined experimental and

theoretical studies at an advanced level. Moreover, feasible approaches should focus on:

o Fundamental properties of both pristine and doped hematite; Nanostructured arrays and

extremely thin film absorbers (ETA);29

Limiting factors for conductivity and photo-

activity;

o Surface chemistry at the nano-scale, including defects and/or oxygen vacancies; Charge

generation, carrier extraction and electrochemical reactions;

o Suitable catalysts for fast water oxidation kinetics at the nanostructured

hematite/electrolyte interface and reduced recombination;

o Optimizing the electromagnetic properties of photoelectrode structures for improved light

management and concentration; modifying optical absorption in thin absorber structures

(iterative computational-experimental strategies);

o Synthetic methods for forming suitable thin films, with controllable composition, doping

and proper structures; Proper substrate – film interface for large scale fabrication;

o Modeling of photoelectrochemical process at nanostructured hematite; The influence of

(nanoscale) structure on carrier generation, extraction and chemical reaction kinetics

(modeling and experiments, e.g. time-resolved optical spectroscopies).

References

1. A. Fujishima and K. Honda, Nature, 238, 37 (1972).

2. M. Grätzel, Nature, 414, 338 (2001).

3. Y-S. Hu, A. Kleiman-Shwarsctein, A. J. Forman, D. Hazen, J-N Park and E. W. McFarland,

Chem. Mater., 20, 3803 (2008).

4. J. H. Kennedy and K. W. Frese, Jr., J. Electrochem. Soc., 125, 709 (1978).

5. R. A. Fredlein and A. J. Bard, J. Electrochem. Soc., 126, 1892 (1979).

6. M. P. Dare-Edwards, J. B. Goodenough, A. Hamnett and P. R. Trevellick, J. Chem. Soc.

Faraday Trans. 1, 79, 2027(1983).



A-35


7. J. H. Kennedy and M. Anderman, J. Electrochem. Soc., 130, 848 (1983).

8. C. Leygraf, M. Hendewerk and G. Somorjai, J. Solid State Chem., 48, 357 (1983).

9. J. E. Turner, M. Hendewerk, J. Parmeter, D. Neiman, G. A. Somorjai, J. Electrochem. Soc.,

131, 1777 (1984).

10. E. L. Miller, R. E. Rocheleau, X. M. Deng, Int. J. Hydrogen Energy, 28, 615 (2003).

11. E. L. Miller, D. Paluselli, B. Marsen, R. E. Rocheleau, Solar Energy Mat. Solar Cells, 88,

131 (2005).

12. O. Khaselev and J. A. Turner, Science, 280, 425 (1998).

13. A. J. Bard and M. A. Fox, Acc. Chem. Res., 28, 141 (1995).

14. A. J. Nozik, Appl. Phys. Lett., 30, 567 (1977).

15. R. Abe, T. Takata, H. Sugihara and K. Domen, Chem. Commun. 3829 (2005).

16. H. Wang, T. Deutsch and J. A. Turner, J. Electrochem. Soc., 155, F91 (2008).

17. H. Wang and J. A. Turner, J. Electrochem. Soc., 157, F173 (2010).

18. Van de Krol R., Gratzel M.; “Photoelectrochemical hydrogen production”, Springer (2012)

19. Saito R., Miseki Y., Sayama K.; Chem. Commun., 48, 3833 (2012).

20. Su J., Guo L., Bao N., Grimes C.A.; Nano Lett. 11, 1928 (2011).

21. Sivula K., Le Formal F., and Gratzel M.; Chem. Mater. 21, 2862 (2009).

22. Vayssieres, L.; Sathe, C.; Butorin, S. M.; Shuh, D. K.; Nordgren, J.; Guo, J. H., Adv. Mater.

2005, 17, 2320-2323.

23. U. Björkstén, J. Moser and M. Grätzel, Chem. Mater., 6, 858 (1994).

24. N. Beermann, L. Vayssieres, S.-E. Lindquist and A. Hagfeldt, J. Electrochem. Soc., 147,

2456 (2000).

25. L. Vayssieres, N. Beermann, S.-E. Lindquist and A. Hagfledt, Chem. Mater., 13, 233 (2001).

26. T. Lindgren, H. Wang, N. Beermann, L. Vayssieres, A. Hagfeldt, S.-E. Lindquist, Sol.

Energy Mater. Sol. Cells, 71, 231 (2002).

27. B. Chernomordik, H. B. Russell, U. Cvelbar, J. B. Jasinski, V. Kumar, T. Deutsch, and M. K.

Sunkara, Nanotechnology, 23, 194009 (2012).

28. I. Cesar, A. Kay, J. Matinez and M. Grätzel, J. Am. Chem. Soc., 128, 4582 (2006).

29. A. Kay, I. Cesar and M. Grätzel, J. Am. Chem. Soc., 128, 15714 (2006).

30. K. Sivula, R. Zboril, F. Le Formal, R. Robert, A. Weidenkaff, J. Tucek, J. Frydrych and M.

Grätzel, J. Am. Chem. Soc., 132, 7436 (2010).

31. J. Brillet, M.Grätzel and K. Sivula, Nano Lett. 10, 4155 (2010).

32. K. Sivula, F. Le Formal, and M.Grätzel, ChemSusChem, 4, 432 (2011).

33. C. J. Sartoretti, M. Ulmann, B. D. Alexander, J. Augustynski, A. Weidenkaff, Chem. Phys.

Lett., 376, 194 (2003).

34. C. J. Sartoretti, B. D. Alexander, R. Solarsko, I. A. Rutkowska, J. Augustynski, R. Cerny, J.

Phys. Chem. B, 109, 13685 (2005).

35. Y. Ling, G. Wang, D. A. Wheeler, J. Zhang and Y. Li, Nano Lett. 11, 2119 (2011).

36. Y Lin, S. Zhou, S. Sheehan and D. Wang, J. Am. Chem. Soc., 133, 2398 (2011).

37. I Cesar, K. Sivula, A. Kay, R. Zboril and M. Graetzel, J. Phys. Chem. C, 113, 772(2009).

38. G. M. Wang, Y. C. Ling, D. A. Wheeler, K. E. N. George, K. Horsley, C. Heske, J. Z. Zhang

and Y. Li, Nano Lett., 11, 3503 (2011).

39. Hu, Y.-S.; Kleiman-Shwarsctein, A.; Stucky, G. D.; McFarland, E. W. Chem. Commun. 2652

(2009).

40. Jang, J. S.; Lee, J.; Ye, H.; Fan, F.-R. F.; Bard, A. J. J. Phys. Chem. C 113, 6719 (2009).



A-36


41. H. Tang, M.A. Matin, H. Wang, T. Deutsch, M. Al-Jassim, J. Turner and Y. Yan, J. Appl.

Phys. 110, 123511 (2011).

42. M. A. Lukowski and S. Jin, J. Phys. Chem. C, 115, 12388 (2011).

43. G M. Wang, Y. C. Ling, D. A. Wheeler, K. E. N. George, K. Horsley, C. Heske, J. Z. Zhang

and Y. Li, Nano Lett., 11, 3503 (2011).

44. Pendlebury, S. R.; Barroso, M.; Cowan, A. J.; Sivula, K.; Tang, J.; Gratzel, M.; Klug, D.;

Durrant, J. R. Chem. Commun., 47, 716 (2011).

45. Cowan, A. J.; Barnett, C. J.; Pendlebury, S. R.; Barroso, M.; Sivula, K.; Gratzel, M.; Durrant,

J. R.; Klug, D. R. J. Am. Chem. Soc., 133, 10134 (2011).

46. Peter, L. M.; Wijayantha, K. G. U.; Tahir, A. A. Faraday Discussions 155, 309 (2012).

47. Klahr, B.; Gimenez, S.; Fabregat-Santiago, F.; Hamann, T.; Bisquert, J. J. Am. Chem. Soc.,

134, 4294 (2012).

48. Le Formal, F.; Tetreault, N.; Cornuz, M.; Moehl, T.; Gratzel, M.; Sivula, K. Chem. Sci. 2,

737 (2011).

49. Spray, R. L.; McDonald, K. J.; Choi, K.-S. J. Phys. Chem. C, 115, 3497 (2011).

50. D.K. Zhong, M. Cornuz, K. Sivula, M. Graetzel and D. R. Gamelin, Energy Environ. Sci., 4,

1759 (2011).

51. Thomann, I.; Pinaud, B. A.; Chen, Z.; Clemens, B. M.; Jaramillo, T. F.; Brongersma, M. L.

Nano Lett. 11, 3440 (2011).

52. Warren S.C., Thimsen E.; Energy Environ. Sci., 5, 5133 (2012).

53. Gao H., Liu C., Jeong H.E., Yang P.; ACS Nano, 6, 234 (2012).

54. U. Cvelbar, Z. Chen, M.K. Sunkara, M. Mozetic, Small , 4, 1610 (2008).

55. Polman, A.; Atwater, H. A. Nat. Mater. 11, 174 (2012).



A-37

2012 PEC White Paper: Photovoltage of Hematite

Photovoltage of -Fe2O3

Shannon W. Boettcher, University of Oregon


Muhammad N. Huda, University of Texas at Arlington


In order to split water without an external bias the photovoltage generated by a complete

PEC system must exceed 1.23 V, the thermodynamic minimum.1 In general, the photovoltage

generated by a semiconductor under illumination is proportional to the logarithm of the ratio of

photocurrent to the recombination current.2 For high quality semiconductors such as Si or GaAs

the attainable photovoltage under one sun illumination is typically about 0.4 V less than the

semiconductor bandgap (though both cannot do the water splitting task due to un-suitable band

gaps and issues of oxidation/corrosion). For α-Fe2O3, there are at least two important

recombination pathways that limit the photovoltage output to, in optimized cases, near 0.6 V,

roughly 1.5 V less than the 2.1 eV bandgap.3 In fact, the derivation of the dependency of

photovoltage in reference 2 assumed Boltzman-type particle distribution, whereas in case of the

highly correlated electrons (which results in high localizations) in the conduction band of -

Fe2O3 can make the situation even worse. The two recombination pathways can be broadly

classified as (i) interface, and (ii) bulk recombinations.. However, the fundamental limitations of

-Fe2O3 can be traced to the lack of bulk photoelectron available for conduction.

The first recombination pathway is majority electron transfer across the electrostatic

barrier at the electrode/electrolyte interface. This recombination mechanism could in principle be

mitigated by shifting the band-edge positions of an α-Fe2O3 electrode by surface modification to

increase the equilibrium band bending and hence retard recombination via forward majority

carrier transport.

Bulk and depeletion region recombination within the semiconductor are also critical. It is

well-known that the low mobility of carriers in α-Fe2O3 leads to short collection lengths. Hamann

has recently shown in controlled thin-film α-Fe2O3 model systems that minority holes are

collected via drift over a ~ 6 nm portion of the depletion region where a large electric field

exists,4 leading to large bulk recombination currents and low internal quantum efficiencies (<

0.5) even in the best devices. These high bulk recombination rates are fundamentally related to

the localized electronic structure of α-Fe2O3 which can be described as a charge-transfer

insulator.5 In this type of insulators, the bulk electron effective mass is extremely high, and hence

the immobility of electrons are manifested by large electron-hole recombinations after

photoexcitations.

In general, there are two ways to improve transport of the bulk photocurrent of -Fe2O3.

As it is well known that the main conduction mechanism in -Fe2O3 is by small polaronic effect,

reduction of volume would somewhat improve the photocurrent.6 However, volume reduction by

doping is minimal, and can increase the photocurrent minutely.5 The second way involves the

significant modification of the conduction band. The conduction band minimum (CBM),

composed of Fe 3d orbitals, is dispersionless. Modification of the conduction band by selective

doping is a possibility.7 For example, it has been shown that Ti doping in -Fe2O3 can modify

the CBM and lower the band gap, though not much dispersion was obtained to lower the electron

effective mass. However, the dopant states may simply add additional bands below or in the



A-38


CBM of -Fe2O3, rather than modifying it; for example, Sc doping in -Fe2O3.7 An alternate

approach would be to alloy the dopant materials with -Fe2O3, in which case an overall crystal

phase transition is a possibility. It remains a largely undiscovered area regarding the nature of

these new phase-spaces and the possible electronic properties of these alloys. With selective

alloy strategies, the local magnetic moments can also be destroyed or compensated, which is

another manifestation of the electron localization in -Fe2O3. This will effectively take -Fe2O3

(with new alloy crystal phase) out of “charge transfer” insulator regime. Though very

challenging, these new alloys signatures can be predicted by DFT and be synthesized in the lab.

An important area of basic α-Fe2O3 research will be to determine what the photovoltage

limits are in appropriately modified systems with inhibited forward electron transfer and hence

recombination currents dominated by bulk/depletion region processes. Strategies for minimizing

forward electron recombination current should focus on increasing band-bending and selectively

inhibiting the kinetics of this “back reaction” transfer to solution species which, for water

splitting, occurs primarily in the form of oxygen reduction. The latter may be realized through

hole-selective (electron blocking) contact layers,8 and/or addition of an appropriately selective

catalyst.9 However, given the inherent materials limitations of the pure α-Fe2O3 electronic

structure, it is a challenge to improve the photovoltages sufficiently for application as a single

bandgap photoelectrode for photoelectrochemical water splitting with unbiased efficiencies in

excess of 10%. It may still find a role in tandem or two-bandgap devices coupled to

photoelectrodes which can provide the remainder of the photovoltage but its photocurrent must

still be improved so as not to limit the overall device current.

References

1. Walter, M.; Warren, E.; McKone, J.; Boettcher, S. W.; Qixi, M.; Santori, L.; Lewis, N.

S., Solar Water Splitting Cells. Chem. Rev. 2010, 110 (10), 6446-6473.

2. Lewis, N. S., A Quantitative Investigation of the Open-Circuit Photovoltage of the

Semiconductor Liquid Interface. J. Electrochem. Soc. 1984, 131 (11), 2496-2503.

3. Klahr, B. M.; Hamann, T. W., Current and Voltage Limiting Processes in Thin Film

Hematite Electrodes. Journal of Physical Chemistry C 2011, 115 (16), 8393-8399.

4. Klahr, B. M.; Martinson, A. B. F.; Hamann, T. W., Photoelectrochemical Investigation of

Ultrathin Film Iron Oxide Solar Cells Prepared by Atomic Layer Deposition. Langmuir 2011, 27

(1), 461-468.

5. Huda, M. N.; Al-Jassim, M. M.; Turner, J. A., Mott insulators: An early selection

criterion for materials for photoelectrochemical H-2 production. J. Renew. Sustain. Energy 2011,

3 (5).

6. Kleiman-Shwarsctein, A.; Huda, M. N.; Walsh, A.; Yan, Y. F.; Stucky, G. D.; Hu, Y. S.;

Al-Jassim, M. M.; McFarland, E. W., Electrodeposited Aluminum-Doped alpha-Fe2O3

Photoelectrodes: Experiment and Theory. Chemistry of Materials 2010, 22 (2), 510-517.

7. Huda, M. N.; Walsh, A.; Yan, Y. F.; Wei, S. H.; Al-Jassim, M. M., Electronic, structural,

and magnetic effects of 3d transition metals in hematite. Journal of Applied Physics 2010, 107

(12), 123712.

8. Bisquert J.; Cahen D.; Hodes G.; Ruhle S.; Zaban A., Physical Chemical Principles of

Photovoltaic Conversion with Nanoparticulate, Mesoporous Dye-Sensitized Solar Cells. J. Phys.

Chem. B, 2004, 108 (24), pp 8106–8118



A-39


9. Gorlin Y.: Jaramillo T.; A Bifunctional Nonprecious Metal Catalyst for Oxygen

Reduction and Water Oxidation. J. Am. Chem. Soc., 2010, 132 (39), pp 13612–13614.



A-40

The viability of using amorphous silicon carbide (a-SiC) as a photoelectrode for PEC hydrogen production

Jian Hua, Feng Zhua, Ilvydas Matulionisa Nicolas Gaillardb Todd Deutsch c, Heli Wang c a MVSystems, Inc. b Hawaii Natural Energy Institute c National Renewable Energy Laboratory Date: October 1, 2012 Revised: November 30, 2012 Revised: August 12, 2013 Overview Water splitting using photoelectrochemical (PEC) devices based on hydrogenated amorphous silicon (a-Si:H or a-Si in short) thin films and its alloys has the potential for low-cost and efficient hydrogen production. In 1998, Rocheleau et al reported use of the a-Si triple junction solar cell in a photoelectrode cell which was similar to an integrated photovoltaic (PV)/electrolyzer system (no wires or cell interconnections), and a solar-to-hydrogen (STH) efficiency of 7.8% was achieved in basic electrolyte [1]. More recently, Reece et al reported the STH efficiencies of 4.7% for a wired configuration and 2.5% for a wireless configuration in such a photoelectrode system which can be operated in near-neutral pH conditions [2]. The photoelectrode system comprising the a-Si solar cell interfaced to hydrogen and oxygen-evolving catalysts is considered a scheme mimic the photosynthetic process within a leaf that converts the energy of sunlight into chemical energy by splitting water to produce O2 and H2 [2]. The main drawback of using a-Si thin films is its poor corrosion resistance in electrolyte. In the above scheme, in order to avoid the a-Si solar cells (and ITO layer) exposed to electrolyte directly and thus enhance the durability of the entire PEC cell, the HER or OER catalyst was needed to separate the a-Si solar cells and ITO layer from electrolyte. Alternatively, one could use a “hybrid” photoelectrode scheme [3]. In this configuration, the a-Si photovoltaic junction is protected by a semiconductor that makes a photoelectrochemical junction which is more durable in electrolyte than a-Si. Because this layer is connected in series, it contributes additional voltage allowing for a tandem (instead of triple) a-Si solar cell to be used leading to possibly increased photocurrent depending on current-matching of the solid-state devices. In addition, fabrication of a-Si solar cells is simplified. Using hydrogenated amorphous silicon carbide (a-SiC:H or a-SiC in short) as the photoelectrode adds extra advantages over conventional metal oxides. First, a-SiC is of a bandgap of 1.9~2.3 eV which can be readily tuned with alternation of C source gas during growth [4], allowing more photocurrent and hence increased STH efficiency. For a-SiC with a bandgap of ~2 eV, the maximum available photocurrent approaches ~15 mA/cm2, leading potentially to an STH efficiency of ~18%. Secondly, incorporation of the C in the film should lead to an increase in the corrosion resistance compared to the use of conventional a-Si films [5]. Thirdly, a-SiC has been extensively studied over the last three decades due to its importance in thin film Si solar cells, for instance, it has been used as a p-type window layer in a-Si solar cells for over 20 years [5,6], or as the absorbing layer in a-SiC p-i-n solar cells [7-10]. The state-of-art device performance exhibited a conversion efficiency of ~7%, with Jsc ~13 mA/cm2 and Voc >0.9V [10]. This vast knowledge base accumulated in PV filed over last three decades should be very useful in



A-41

ericm

Typewritten Text

PEC White Paper: a-SiC for PEC Hydrogen Production

development of a-SiC photoelectrode for PEC hydrogen production. Finally, fabrication of a-SiC films is routinely prepared by the Plasma Enhanced Chemical Vapor Deposition (PECVD) technique at a low temperature of ≤ 200 ºC, in contrast with its crystalline counterpart which usually requires a high growth temperatures (1000-2000 ºC) [11,12]. Note that growth of a-SiC is identical to that for a-Si since both use the PECVD technique and can be deposited in a same system. This is a very attractive for mass-production of the hybrid PV/a-SiC photoelectrode devices in a cost-effective fashion. Research Status The fabrication of the a-SiC photoelectrode consisting of p-type and intrinsic SiC layers has been previously reported [13-17]. The a-SiC photoelectrode exhibits a photocurrent density of ~8 mA/cm2 @-1.5V (V vs. SCE). To eliminate the external biases, hybrid PV/PEC device comprising a-SiC photoelectrode and an a-Si tandem solar cell has been developed, as shown in Fig.1(a) [18-21]. This type of hybrid PEC cell exhibited maximum photocurrent density of ~5 mA/cm2 at zero bias (2-electrode configuration using RuO2 counter electrode as OER catalyst). In addition, the hybrid device of a-Si/a-SiC configuration exhibits durability of >500 hours when tested at a constant current of ~ -4 mA/cm2 in a 0.25M H2SO4 electrolyte, as tested so far. Hydrogen production was observed in short-circuit condition.

a-SiC (i)

SnO2Glass

a-SiC(p)

a-Sitop cell a-SiC(p)

a-Si(n)

a-SiC(p)

a-Si(n)

a-Si (i)

a-SiC cell

a-Si (i)

lightlight

(a)

(b)

a-Sibottom cell

a-SiC(n)

Fig.1. (a) Schematic diagram of the hybrid PV/a-SiC device. (b) Progress in improvement of the performance of the hybrid PV/a-SiC device1. Fig.1(b) shows yearly progress in improvement of the photocurrent current and the STH efficiency. We see that, with improved energetics and catalytic activity at the a-SiC/electrolyte interface using Ru nanoparticles treatment and enhanced performance of the photovoltaic solar cell, the short circuit photocurrent density increases to ~5 mA/cm2, or equivalent to the STH efficiency of ~6.1%. The performance of the hybrid PEC cells could be improved significantly once the PV cells and surface modification are further optimized. A-SiC as a photoelectrode makes possible an entirely a-SiC based hybrid PEC device, which could be produced in large quantities with lower cost using a cluster tool PECVD/Sputter system designed and fabricated at MVSystems, Inc. Addressing the Challenges and Possible Solutions 1. Materials Efficiency The poor interface between a-SiC and electrolyte was the main technical barrier, which resulted in a large over-potential loss and difficulty to extract charges at the interface. In order to minimize the surface barrier (i.e., partially caused by SiOx native oxide) and increase STH efficiency, surface modification on

1 For proprietary reason, the high performance PV device is termed as “X cell”, whose details are not disclosed here.



A-42

ericm

Typewritten Text


a-SiC was necessary and has proven to be crucial to enhance the device performance. Over the past several years, several methods have been evaluated, including the HF etch, H2 plasma treatment, thin a-SiNx and thin C-rich layer coverage by PECVD technique, CH3-termination (methylation) or metal nanoparticles. The most promising technique so far is the surface treatment by low work function (WF) nanoparticles, which effectively reduces the onset potential and increases photocurrent from <1 mA/cm2 to ~3 mA/cm2 at zero potential measured at 2-electrode setup. (1) Surface treatment by metal nanoparticles. There are two key factors contributing to performance improvement of the hybrid device: a low Schottky barrier at a-SiC/metal interface, and catalytic property of the metal nanoparticles. For high work function metals (Pt, Pd and Au), the calculated barrier height is high, i.e., >1 eV; whereas for Ti, Ru and W with low WF, the barrier height is much lower. This difference in barrier height could partly explain why all low WF metal particles lowered overpotentials whereas all high WF metals increased them. In addition, although Ti exhibits a lower barrier height (0.39 eV) than that of Ru (0.77 eV), the latter usually leads to much higher photocurrent than the former at zero potential. One possibility is that Ru is of a better catalytic activity in hydrogen evolution than Ti. Yamada et al showed that, with Ru nanoparticles, the electron transfer from the photogenerated QuPh•_NA to Ru nanoparticles results in hydrogen evolution even under basic conditions (pH10). In addition, the size of the Ru nanoparticles has effect on the catalytic reactivity for hydrogen evolution [22]. Similar behaviors such as size effect and hydrogen evolution in pH10 electrolyte were also observed for a-SiC [23]. In summary, with improved energetics and catalytic activity at the a-SiC/electrolyte interface using low WF Ru nanoparticles treatment, the overpotential loss is reduced to ~1.6V (@~3mA/cm2). This overpotential value is very close to ideal water splitting condition, i.e. ~1.5V. Further reduction in overpotential is expected by refining the Ru naoparticles treatment, for instance, enhancing catalytic activity by applying Pt particles on Ru coated a-SiC photoelectordes and hybrid devices. Meanwhile, cheaper low work function metals (e.g. W) were evaluated but more work is needed to establish its effectiveness. (2) Modification of the a-SiC material properties. It should be noted that C incorporation results in increase in disorder in a-SiC due to different bond lengths in C-C (1.54 Å) and Si-Si (2.24 Å) bonds [24], and increase of the density of states in the mid-gap region [25]. Deposition conditions such as substrate temperature, doping, type of source gases, RF power, pressure, etc, all affect a-SiC film properties. Generally the electronic quality of a-SiC is best at lower values of Eg (less than ~2.1eV) since the film is less defective. N-type and p-type doping can be achieved with good control of the Fermi level. So far, bulk properties of a-SiC have not been altered. It is important to know that, like a-Si, a-SiC is by nature a weakly n-type semiconductor. The conductivity activation energy of intrinsic a-SiC, as measured at MVSystems, is about 0.9 eV. By adding small amount of boron into the film, the Fermi energy (EF) of intrinsic a-SiC will shift towards mid-gap, leading to a change in surface energy band structure. If assuming the surface EF shifts downwards as in the bulk, the band alignment between the Fermi level and the O2/H2O redox potential would be improved. However, since the surface EF is also critically affected by the surface states, which tends to pin the EF, how much effect from doping in the intrinsic a-SiC is not known yet, and this will be determined by more experiments. 2. Materials Durability.



A-43

ericm

Typewritten Text


In the past, durability on hybrid PV/a-SiC device has been successfully demonstrated for up to 310 hours at a constant current density of -1mA/cm2 [26]. Recently, we have further achieved the durability of >500 hours on hybrid devices, under constant bias leading to a photocurrent density of ~4 mA/cm2 in 0.25M H2SO4 electrolyte.2 This improvement in durability was due to contributions from several factors. The enhanced performance of the hybrid device, i.e., surface modification by the low work function metal nanoparticles (W nanoparticles were used in this test) allowed the device under test to sustain higher current biases. In addition, optimized process conditions for growing solid-state devices helped minimize defects. In fact, after about 700-hour test, no pinhole was found on the tested sample. Also, corrected testing procedure helped avoid interferences from handling and epoxy sealing. As a result, the yield and consistency of the durability test has been improved considerably. This result demonstrates that the a-SiC based hybrid device can be of good durability of >500 hours. 3. Device Configuration Designs. Monolithically integrating a-SiC photoelectrode with Si solar cells is fairly straight forward, because both are fabricated using the same technique, PECVD, and at a similar low temperature of ≤ 200 ºC. The key question here is how to maximize the performance (photocurrent and voltage) of the hybrid device through current match between the solar cell and photoelectrode. Table 1 shows the calculated performance of the hybrid devices where the solar cells and a-SiC layer is assumed to be in a good current match. Note that in case (1), where the hybrid PV/a-SiC device is of a-Si pin/pin/a-SiC pi configuration, the limiting factor is the a-Si tandem solar cell with a filtered photocurrent density of ~6 mA/cm2 3. Hence, the STH efficiency is limited as ~7.6%. To reach this goal, in addition to improving the surface energetics as described previously (see “Materials Efficiency”), the possible approaches include:

(1) Current match by optimizing thickness of each junction; (2) Using pin type configuration for the a-SiC photoelectrode; (3) To employ high performance solar cells.

In order to achieve current match, thickness of each intrinsic layer in the device was altered and tests were done using ITO as the top contact. As a result, the Jsc of ~5.2 mA/cm2 was achieved due to better current matching. At 1.5V (practical water splitting bias), the photocurrent reaches ~4.1 mA/cm2. When inserting an amorphous n+ layer (~10nm thick) underneath ITO, the barrier at a-SiC/ITO interface is eliminated. Thus a much better FF (0.65) was obtained. Another noticeable change in device performance was the increase of Voc (~2.5V), which is caused by a higher built-in electric field in a p-i-n configuration (compared with the p-i configuration). Further increase in Jsc is expected with optimized n+ layer properties. In order to further enhance the STH efficiency beyond 10%, a more powerful solar cell engine must be used. One possible choice is outlined in Table 1.

2 As of this writing the device under test has 700 hours, and is still operational. The test continues. 3 The a-SiC layer (100 nm thick) could generate maximum 8.8 mA/cm2 photocurrent density, and the realistic operational current density (Jph) is estimated as ~7.6 mA/cm2, assuming FF=0.75 and the current density at the maximum power point is estimated by

JscFF * .



A-44

ericm

Typewritten Text


Table 1. Calculated performance of the hybrid PV/a-SiC devices. A conceptual configuration of a hybrid X cell/a-SiC device could be the one shown in Fig.2, where a-SiC photoelectrode (of pin structure) is integrated monolithically with the X cell. The combination of the novel X cell and the a-SiC pin cell could generate Jsc in a range of 14~17 mA/cm2, or possible operational photocurrent density of 12.4~15 mA/cm2 (assuming FF=0.8) equivalent to a STH efficiency of 15~18%.

Fig.2. Schematic diagram for the hybrid X cell/a-SiC pin photoelectrode. It should be noted that the triple junction solar cells as used in a photoelectrode system comprising the a-Si solar cell only [1-2], in spite of its high Voc (~2.2V), could only provide a limit operation current density, i.e., ~8.3 mA/cm2 [27]. These solar cells thus are not suitable for achieving a STH efficiency of 11%. Appendix: Current Matching in Multi-junction Solar Cells (A Review) References [1] R. E. Rocheleau, E. L. Miller, and A. Misra, Energy & Fuels 1998, 12, 3-10 [2] S. Y. Reece, J. A. Hamel, K. S. Thomas, D. Jarvi, A. J. Esswein, J. J. H. Pijpers, D. G. Nocera, Science 334, 645 (2011). References therein. [3] E. L. Miller; R. E. Rocheleau, S. Khan, International Journal of Hydrogen Energy 29 (2004) 907 – 914 [4] Madan and Shaw, The Physics and Applications of Amorphous Semiconductors, Academic Press, San Diego (1989), p.155.

a-SiC Photoelectrode PV devices Hybrid

Configuration Eg (eV)

Jph (mA/cm2) Available

Voc (V) Configuration

Jph (mA/cm2) Filtered

Voc (V)

STH (%)

possible

(1) p-i 2 7.66 (100nm) 0.5-0.6 a-Si/a-Si

(620nm/132nm) 6.15 1.8-1.9 7.56

(2) p-i-n 2 12.4 (500nm) >1 X cell

>25 >0.7 >12 lig

ht

a-SiC pin

HIT

cell

Metal Metal nanoparticles

4H++O2

2H2O

2H2

4H+

OERcatalyst

X cell



A-45

ericm

Typewritten Text


Appendix: Current Matching in Multi-junction Solar Cells –A Review Table of Contents:

1. High efficiency in multi-junction solar cells. 2. Current matching in two-terminal multi-junction solar cells. 3. Effect of sub-cell bandgaps on current-matching. 4. Effect of top-cell thickness on current matching. 5. Current matching in a-Si based multi-junction solar cells. 6. Current matching in PEC devices.

1. High efficiency in multi-junction solar cells. Multijunction solar cells provide a simple and straightforward way of overcoming the fundamental conversion efficiency limitation of the single junction solar cell [1]. The fundamental concept underlying multijunction solar cells is “spectrum splitting.” In this configuration, the top junction which is of the highest bandgap “filters” the sunlight to the bottom junction. Thus, only high-energy photons are absorbed in the top cell whereas photons with energy less than the top junction bandgap pass through towards the inner cells (or the bottom cell). This concept is illustrated in Fig.1, showing the solar spectrum (upper graph) divided up into two regions for conversion by a two-junction cell (lower graph). A larger open-circuit voltage (Voc) across the top junction than that across the bottom cell is achieved due to a larger bandgap in the top cell, leading to more incident power conversion. This is the “spectrum-splitting” effect.

Top junction(Eg~1.8 eV)Bottom junction(Eg~1.4 eV)Substrate

Fig.1. AM1.5 Global Spectrum and schematic of a two-junction solar cell with different bandgaps, (1.4 eV and 1.8 eV for the bottom and top cell), showing the spectral regions converted by each junction. The thermodynamic efficiency limit of the multi-junction stack (with infinite number of cells) can be calculated based on the detailed recombination-generation balance equation as originally proposed by Shockley and Queisser, given by [2]



A-46

ericm

Typewritten Text

PEC White Paper Appendix on Current Matching

where ηmc(ε) is the monochromatic cell efficiency, and i(ε,V) is the current in a monochromatic cell. σSB (= 5.67 × 10-8 Wm-2 K-4) is the Stefan–Boltzmann constant. For Ts = 6000 K and Ta = 300 K, the sun and ambient temperature, respectively, one finds an efficiency of 68.2% for 1 sun illumination intensity and 86.8% for 45900 suns intensity [2], the highest efficiency limit of known ideal photovoltaic converters and much higher than that of a single junction (31% for unconcentrated and 40.7% for concentrated solar cells [3]). In practice, the efficiency of a multi-junction solar cell is much lower than the above theoretical value. So far, the highest efficiency achieved in a concentrated multi-junction solar cell was 44% (roughly half of the theoretical limit), as reported by Solar Junction (USA), using a monolithic two-terminal lattice matched triple junction cell of GaInP/GaAs/GaInNAs operating at a concentration factor of 947 suns [4]. In unconcentrated multi-junction solar cells, Sharp (Japan) recently reported the highest efficiency of 37.7% using an GaInP/GaAs/GaInAs triple junction configuration [5]. The difference between the actual cell efficiency and theoretical limit indicates much more efforts would be needed to reduce various losses and hence improve the performance of the multi-junction solar cells. These losses include reflection loss at the cell surface and various sub-cell interfaces, resistive loss at the sub-cell interfaces, current matching among sub-cells, etc. In the following, issues mainly related to the current-matching will be reviewed. 2. Current matching in two-terminal multi-junction solar cells. A multi-junction solar cell can be constructed in different ways, depending on connection of power leads to the multijunction stack. For instance, for a two-subcell stack, there could be of a two-, three-, or four-terminal configuration. Among these, the two-terminal series-connected configuration provides truly monolithic, fewest possibilities for interconnection of the devices, requiring a simpler cell structure and processing. However, this configuration requires that the photocurrents of the subcells be closely matched, since in this series connection the subcell with the least photocurrent dictates the current generated by the entire device. The current-voltage characteristics of the two-terminal series-connected m junction devices can be described by [6]

where Vi(J) is an individual J –V curve the ith device. In order for each individual subcell to operate at its own maximum-power point, it is apparent that the maximum current density must be the same for all the subcells, i.e. Jmp,1 = Jmp,2 = . . . = Jmp,m. If this is the case, then the maximum power output of the combined multijunction device is the sum of the maximum power outputs of the subcells. Otherwise, if the subcells do not all have the same Jmp,i, some of the subcells will operate away from their maximum power points. As a result, the output power from the entire device will be reduced. 3. Effect of sub-cell bandgaps on current-matching.



A-47

ericm

Typewritten Text


For a cell with n junctions numbered from top to bottom as 1/2/. . ./n and with corresponding bandgaps Eg1/Eg2/ . . . /Egn, the short-circuit current density of the mth subcell is given by

where λm = hc/Egm is the wavelength corresponding to the bandgap of the mth subcell, and Φm(λ) the incident spectrum seen by the mth subcell. This equation shows that JSC depends on the bandgaps of the various junctions. For instance, in the simple case of a two-junction cell, because the bottom junction is filtered by the top junction, the bottom-junction current density JSC,2 depends on both Eg1 and Eg2, whereas JSC,1 depends only on Eg1. Fig.2 shows calculated Jsc,1 and Jsc,2 as a function of top cell bandgap for Eg2 = 1.42 eV for the AM1.5 global spectrum [6].

Fig.2 JSC1 and JSC2 as a function of top-subcell bandgap Eg1 for a bottom-subcell bandgap Eg2 = 1.42 eV. The figure shows that as Eg1 decreases, JSC1 increases and JSC2 decreases, becoming less than JSC1 for Eg1< 1.95 eV. The JSC for the entire solar cell will be the lesser of JSC1 and JSC2. This quantity is a maximum at the current-matched bandgap Eg2 = 1.95 eV, and falls off rapidly as Eg1 is decreased below 1.95 eV. The current-matching constraint puts relatively tight constraints on the selection of bandgaps for the various junctions in this structure. The highest efficiency as mentioned previously, 68.2% (under 1 sun illumination) and 86.8% (under concentrated illumination), is obtained with an infinite number of solar cells, each one biased at its own voltage and illuminated with monochromatic radiation [2]. In the case of finite number of cells, many authors have calculated the optimal bandgap combination [7-11]. For a two-junction series-connected cell under 1 sun illumination, a maximum efficiency of 44.3% [8], 42% [2], 40.7% [9] and 38% [10] or 35% [11] was deduced for the optimum bandgap pairs of 1.0/1.8 eV, 1.0/1.9, 0.97/1.65 eV and 1.13/1.75 eV, respectively. For the three-cell configuration, the above authors also reported the calculated 1-Sun efficiency in a range of 35 ~ 50% using various bandgap combinations: 35.6% with 1.1/1.55/2.5 eV [8], 50.3% with 1.0/1.6/2.2 eV [10] and 49% with 0.8/1.4/2.3 eV [2]. 4. Effect of top-cell thickness on current matching. Alternation of the top-cell thickness is a commonly used method for achieving current matching in fabrication of two-junction solar cells [12]. This is because the absorption coefficient for solar-cell materials is finite, and a cell of finite thickness will not absorb all the incident above-bandgap light. The thinner the cell, the greater the transmission (this is particularly true for photons near the bandgap where the absorption is very small). If, before thinning, JSC1 (top cell) < JSC2 (bottom cell) then the top subcell can be thinned to make JSC1 = JSC2. Although the actual current at the maximum power point is not same



A-48

ericm

Typewritten Text


as Jsc, this should be a very good approximation for high-quality, high-bandgap cells. A similar approach can be also applied to other multi-junction solar cells such as the triple junction stack, in order to achieve current matching. In this case, thickness of both the top sub-cell and the middle sub-cell is alternated for optimization. 5. Current matching in a-Si based multi-junction solar cells. Amorphous silicon (a-Si in short) multijunction solar cells fabricated in a stacked structure is particularly successful both because there is no need for lattice matching, as is required for crystalline heterojunctions, and also because the bandgap is readily adjusted by alloying with Ge, C or by forming nanocrystallin silicon (nc-Si). Another added advantage is the improved stability in a-Si tandem solar cells compared with the a-Si single junction devices, since the former can utilize thinner intrinsic a-Si layers leading to reduced light-induced degradation. Finally, since a multijunction cell delivers its power at a higher operating voltage and lower operating current than a single-junction cell, the lower current reduces resistive losses. Higher solar conversion efficiency than single-junction cells has been achieved in multijunction, a-Si-based solar cells. Fig.3 shows schematic diagrams of a few a-Si multijunction configurations presently used in most commercial modules.

(a) (b) (c) (d)

Fig.3 Schematic diagrams of some typical a-Si multijunction solar cells.

The a-Si/a-Si tandem solar cell (Fig.3(a)) was very first kind of a-Si based multi-junction devices developed. The highest efficiency so far reported in the a-Si/a-Si tandem solar cell is ~12% [13]. Compared with the a-Si single junction device, it exhibits not only increased efficiency, i.e., ~12% vs. ~10%, but also improved stability under illumination, i.e., degradation of ≤5% vs. >10% in a first 100 hours test [14]. However, the performance of this type of tandem solar cell is limited by its bandgap since both junctions are of an identical bandgap (~1.75eV). Replacing the a-Si bottom sub-cell with a-SiGe (Eg=1.4~1.6 eV) helps increase of the Jsc due to the spectrum splitting effect, leading to an improved efficiency of ~14% [15]. Compared with a-Si, a-SiGe is more defective. This limits the even higher performance of a-Si/a-SiGe tandem solar cells. A more promising approach is the “micromorph” tandem solar cell where a more stable nc-Si (Fig.3(b)) is used as the bottom sub-cell. Besides, since nc-Si is of a bandgap of ~1.1 eV, the a-Si/nc-Si pair is closer to optimum bandgap combination for achieving a high efficiency. The reported highest efficiency of a-Si/nc-Si tandem solar cells has reached 14.7% [16]. One of key techniques used in a-Si/nc-Si tandem devices is the introduction of a highly transparent tunneling junction between the a-Si top and nc-Si bottom sub-cells, which not only minimizes the series resistance, but also helps reflect more short-wavelength photons back into the a-Si layer, resulting in increased Jsc in the a-Si sub-cell and achieve better current matching with the nc-Si bottom sub-cell [17].

a-SiGe



A-49

ericm

Typewritten Text


Further improvement in efficiency has been achieved using the triple junction configuration as shown in Figs.3(c) and (d). These triple junction devices, a-Si/a-SiGe/a-SiGe and a-Si/a-SiGe/nc-Si devices, fully make use of the spectrum splitting effect by utilizing three absorbing layers of different bandgaps (1.8/1.6/1.4 eV and 1.8/1.5/1.1 eV respectively), lead to an improved efficiency of 15.2% [18] and 16.3%, respectively [19].

6. Current matching in PEC devices. Water splitting using photoelectrochemical (PEC) devices based on a-Si thin films and its alloys offers a potential low-cost and efficient hydrogen production approach. Currently, two different PEC schemes utilizing a-Si multi-junction solar cells have been explored. One is the use of the a-Si triple junction solar cell in a photoelectrode cell which is similar to an integrated photovoltaic (PV)/electrolyzer system (no wires or cell interconnections), and a solar-to-hydrogen (STH) efficiency of 7.8% was demonstrated in basic electrolyte [20]. An alternative scheme is a “hybrid” photoelectrode [21], where the a-Si photovoltaic junction is integrated with a semiconductor photoelectrode that makes a photoelectrochemical junction which is more durable in electrolyte than a-Si. Because this layer is connected in series with an a-Si multi-junction device, current-matching between the PV cell and photoelectrode is necessary in order to achieve a high STH efficiency (details of the hybrid device consisting of a-Si tandem cell/a-SiC photoelectrode is described in the main text). Compared with the case of solid-state devices operating in the air, operation of the PEC device in electrolyte is more complicated, involving not only photovoltaic and optical, but also electrochemical phenomena. In order to analyze these behaviors, Miller et al [21] have developed integrated models comprising both a lumped-circuit model for a photocell (shown in the left-hand side of Fig.4(a)) and an electrochemical model for the current-dependent load (shown in the right-hand side of Fig. 4(a)), whose kinetics is determined by the Butler–Volmer equation. The current-dependent overpotential loss due to charge transfer at the electrode surface and additional potential drop due to ions transport through the electrolyte are included in this analysis. This one-dimensional model is able to capture the key physical and chemical nature of the PEC system.

Fig.4 (a) Integrated models of triple-junction photoelectrode; (b) load-line analysis to determine photoelectrochemical operating point [21]. This analysis shows that the operating point for the PEC device is no longer depends only on the current-matching among individual sub-cells of the PV device, and will be determined by both the light J-V curve of the PV cell and the electrochemical load curve, or the intersection of these two curves as shown in Fig.4(b).



A-50

ericm

Typewritten Text


Reference [1] W. Shockley and H. Queisser, J. Appl. Phys. 32, 510–519 (1961) [2] A. De Vos and H. Pauwels, Appl. Phys. 25, 119–125 (1981) [3] A. Luque and A. Mart, “Theoretical Limits of Photovoltaic Conversion and New-generation Solar Cells” (Chapter 4), Handbook of Photovoltaic Science and Engineering, 2nd Ed. Edited by A. Luque and S. Hegedus, John Wiley & Sons, 2011. Reference therein. [4] Solar Junction (USA) news release, http://www.sj-solar.com/about_us/latest-news.php, October 15, 2012 [5] Sharp (Japan) news release, http://www.sharp-world.com/corporate/news/121205.html, December 5, 2012 [6] D. J. Friedman, J. M. Olson and S. Kurtz, “High-efficiency III–V Multijunction Solar Cells” (Chapter 8), Handbook of Photovoltaic Science and Engineering, 2nd Ed. Edited by A. Luque and S. Hegedus, John Wiley & Sons, 2011. Reference therein. [7] J. J. Loferski, Proc. 12th IEEE Photovoltaic Specialists Conference, 1976, p.957-961. [8] A. Bennett and L. C. Olsen, Proc. 13th IEEE Photovoltaic Specialists Conference, 1978, p. 868-873. [9] I. Tobias, A. Luque, Progress in Photovoltaics: Research and Applications 2002, 10, 323–329 [10] M. E. Nell and A. M. Barnett, IEEE Trans. Electron Devices ED-34 257 (1987). [11] F. Meillaud_, A. Shah, C. Droz, E. Vallat-Sauvain, and C. Miazza, Solar Energy Materials & Solar Cells 90 (2006) 2952–2959 [12] S. R. Kurtz, P. Faine, J. M. Olson, Journal of Applied Physics 1990, 68, 1890 [13] J. Yang, K. Lord, S. Guha and S.R. Ovshinsky, Mat. Res. Soc. Symp. Proc. Vol. 609, 2000 [14] T.Yoshlda, K.Maruyama, O.Nabeta, Y.Ichikawa, H.sakai and Y.Uchida, Proc. 19th IEEE Photovoltaic Specialists Conference, p.1095, 1987 [15] X. Deng, Proc. 31st IEEE Photovoltaic Specialists Conference, p. 1365-1370, 2005 [16] K.Yamamoto, A. Nakajima, M.Yoshimi, T. Sawada, S. Fukuda, T. Suezaki, M. Ichikawa, Y. Koi, M. Goto, T. Meguro, T. Matsuda, M. Kondo, T. Sasaki, Y. Tawada, Solar Energy, 77, 939(2004) [17] D. Fischer, S. Dubail, J. A. Anna S&an, N. Pellaton Vaucher, R. PI&z, Ch. Hof, U. Kroll, J. M&r, P. Tomes, H. Keppner, N. Wyrsch, M. Go&, A. Shah, and K-D.Ufert, Proc. 25th IEEE Photovoltaic Specialists Conference, 1996, p. 1053 [18] J. Yang, A. Banerjee, K. Lord, S. Guha, 2nd World Conf. On Photovoltaic Energy Conversion, 387 (1998) [19] B. Yan, G. Yue, L. Sivec, J. Yang, and S. Guha, Appl. Phys. Lett. 99, 113512 (2011) [20] R. E. Rocheleau, E. L. Miller, and A. Misra, Energy & Fuels 1998, 12, 3-10 [21] E. L. Miller; R. E. Rocheleau, S. Khan, International Journal of Hydrogen Energy 29 (2004) 907 – 914



A-51

ericm

Typewritten Text


PEC White Paper Appendix on Energetic Mismatch

Energetic Mismatch



Moreno de Respinis, Delft University of Technology

Nicolas Gaillard, Hawaii Natural Energy Institute

Shannon Boettcher, University of Oregon

Energetic mismatch is a major challenge for most semiconductors studied for PEC water

splitting. For water splitting to occur using a single semiconductor photoelectrode, the

semiconductor conduction band minima (CBM) must be more negative of the hydrogen

evolution reaction (HER) and valence band maxima (VBM) more positive of the oxygen

evolution reaction (OER), respectively.1-3

If these criteria are not met, the equilibrium band

bending will not be sufficient for the semiconductor to generate a quasi-Fermi level splitting

greater than 1.23 V, the thermodynamic minimum needed for photoelectrochemical water

splitting.3b

Although some wide band-gap materials satisfy this criterion (e.g. SrTiO3), this is at the expense

of charge carrier generation, as only a small fraction of the solar spectrum is absorbed. This leads

to very low water splitting efficiencies.

Figure 1. a) Band diagram of Ta oxide-oxynitride-nitride (left figure), and b) their optical absorbance (right

figure)26a, 26b

One method to tackle this issue is via bandgap engineering. One example is the tantalum oxide

system. Ta2O5 has Ta5d orbitals (which make up the conduction band) and O2p orbitals (which

make up the valence band) which straddle the redox potential of hydrogen and oxygen evolution,

respectively. However, its bandgap of 3.9 eV preclude visible light absorption. The nitride form

Ta3N5, has N2p orbitals substituting the O2p orbitals leading to a decreased bandgap of 2.1 eV

and near-ideal band-edge positions. The poor stability of Ta3N5 in aqueous environment,

however, limits use as a photoanode for water oxidation. The poor electronic properties further

limit the quantum efficiency for carrier collection and the quasi-Fermi level splitting (and hence

photovoltage generated by a photoelectrode). Possible routes to overcome these shortcomings are

by improving the surface catalytic activity of Ta3N525

, or by forming the oxynitride phase TaON.



A-52


TaON combines light absorption in the visible range with photochemical stability (see Figures 1a

and 1b). O2p and N2p orbitals hybridize to form the valence band.

β-TaON has shown photocatalytic activity with quantum efficiency up to 76% (for 400 nm

photons) and current density greater than 3.5 mA/cm2 at 0.6 V vs. Ag/AgCl

26c. Despite these

promising results, the synthesis of β-TaON is non-trivial, and performance-limiting bulk and

surface defects are yet to be identified and addressed. The photovoltage obtained from

photoelectrodes is not sufficient for overall water splitting, and overall water splitting required an

additional applied bias of 0.6 to 1 V, despite the large band-gap.

Another method to circumvent the energetic mismatch issue is to apply a bias in addition to the

photovoltage generated by the photoelectrode. A solar cell (or multi-junction device) placed

underneath the photocatalytic material can provide such bias, allowing hydrogen production

from renewable solar energy only.6-8

The bias, in essence, “makes up” the energetic mismatch

and thus enables the counter electrode to drive the reaction that does not match with the band

edge level. In other words, for semiconductors whose band edge positions limit the attainable

photovoltage to less than 1.23 V, additional voltage can be applied externally so that the total

voltage exceeds 1.23 V. For example, n-TiO2 has a conduction band edge that is a few hundred

mV below the hydrogen evolution potential. The maximum possible photovoltage such a

photoanode can generate is the energy difference between the photochemical redox reaction

driven at the surface (for n-TiO2 this is the oxygen evolution reaction) and the flat band position

(which is near the conduction band edge for moderately n-doped samples). If this difference is

less 1.23 eV, that photoelectrode is incapable of splitting water alone, regardless of the band gap.

The monolithic p-GaInP2/n/p-GaAs photoelectrochemical-photovoltaic tandem cell device is a

classic example of a photoelectrode biased by an underlying photovoltaic device.8 The device,

consisting of a top p-GaInP2 layer connected in series to an n/p GaAs bottom cell on a GaAs

substrate, showed a 12.4% solar-to-hydrogen (STH) conversion efficiency with 24 h lifetime.

Alternatively, monolithic devices can be based on a photoanode biased with an integrated p-n

junction.9 Another approach consists of an n-n heterojunction PEC device in which a photoanode

is deposited onto an n-type semiconductor that boosts the energy of the electrons.9-12

Dye-sensitized nanocrystalline-nanoporous solar cell (Grätzel cell) could act as similar function.

A light-to-hydrogen conversion efficiency of 4.5% was reported from the configuration

combined with a nanostructured WO3 photoanode.1

The energy for spontaneous water splitting can also be provided by combining two

photoelectrodes in series, one n-type and one p-type. Nozik in his earlier work used single

crystals of n-TiO2 and p-GaP as photoelectrodes and bonded them together through the ohmic

contacts.13,14

NREL recently reported the performance of individual photoelectrodes connected

together in the outside circuit.15

This combination relaxes the criteria governing the band edge

positions. The dual-electrode configuration separates OER to the n- and HER to the p-type

photoelectrodes, respectively, increasing the number of candidate semiconductors that could be

used to build a stand-alone solar water splitting device.15,16

The key concept is that each

photoelectrode generates a photovoltage given (in the ideal case where bulk recombination is

minimized) by the difference from the flat band potential and the redox couple of interest. The

sum of the two photovoltage must exceed 1.23 V in order for water splitting to occur. Therefore



A-53


to maximize the photovoltage generated by a n-type photoanode, the band edges should raised

(i.e. flat band very negative on the electrochemical scale) as high as possible. To maximize the

photovoltage from a p-type photoelectrode the band edges should be lowered (flat-band potential

very positive on the electrochemical scale.

In order to achieve the ideal photovoltages dictated by the band edge positions, however,

requires high-quality semiconductor materials with sufficient diffusion lengths to minimize the

degradation of the photovoltage due to bulk recombination. The ideal photovoltage for any

semiconductor is roughly 300-400 mV less than the band-gap, but is rarely achieved due to non-

ideal band edge alignment, bulk recombination, or both.

In a conceptually related approach, Domen and co-workers developed Z-scheme systems with

two photocatalysts and redox couple (mediator).17-19

The key challenge with the redox Z-scheme

is preventing internal shunting through the conducting redox mediator. Lewis and co-workers

have developed p- doped Si microwires20,21

and are working to incorporate them with n-doped

photoanodes in a tandem cell ion-conducting-membrane-supported tandem photoelectrode.22

The

common point in these multiple photo-absorber approaches is that two or more photons are

required to generate one electron in the external circuit. Such approaches would allow the use of

lower energy photons that are unused in single semiconductor configurations.

Other strategies strategy to improve the band edge alignment is via modifications of the

semiconductor surface, either with a pH-insensitive group, producing a surface dipole that is

independent of pH, or by introducing a desired surface dipole/charge.23

Another possible way to

shift the band edge up would be via quantum confinement, although this also increases the band-

gap.24

A 0.3-0.6 eV CB shift due to the quantum confinement effect could potentially locate the

CB of hematite above the HER level, although bulk recombination issues would still remain that

would likely limit the photovoltage output.

References

1. M. Grätzel, Nature, 414, 338 (2001).

2. T. Bak, J. Nowotny, M. Rekas, C.C. Sorell, Int. J. Hydrogen Energy 27, 991 (2002).

3. R. van de Krol, Y. Liang and J. Schoonman, J. Mater. Chem. 18, 2311 (2008).

3b. Kumar, A.; Santangelo, P. G.; Lewis, N. S. J. Phys. Chem. 1992, 96, 834-842.

4. A. Kudo and Y. Miseki, Chem. Soc. Rev. 38, 253 (2009)

5. A. J. Bard and M. A. Fox, Acc. Chem. Res., 28, 141 (1995).

6. E. L. Miller, R. E. Rocheleau, X. M. Deng, Int. J. Hydrogen Energy, 28, 615 (2003).

7. E. L. Miller, D. Paluselli, B. Marsen, R. E. Rocheleau, Solar Energy Mat. Solar Cells, 88, 131

(2005).

8. O. Khaselev and J. A. Turner, Science, 280, 425 (1998).

9. R. van de Krol, M. Grätzel; “Photoelectrochemical hydrogen production”, Springer (2012)

10. R. Saito, Y. Miseki, K. Sayama; Chem. Commun., 48, 3833 (2012).

11. J. Su, L. Guo, N. Bao, C.A. Grimes; Nano Lett. 11, 1928 (2011).

12. K. Sivula, F. Le Formal and M. Grätzel, Chem. Mater. 21, 2862 (2009).

13. A. J. Nozik, Appl. Phys. Lett., 30, 567 (1977).



A-54


14. A. J. Nozik, Phil. Trans. R. Soc. Lond. A, 295, 453 (1980).

15. H. Wang, T. Deutsch and J. A. Turner, J. Electrochem. Soc., 155, F91 (2008).

16. H. Wang and J. A. Turner, J. Electrochem. Soc., 157, F173 (2010).

17. R. Abe, T. Takata, H. Sugihara and K. Domen, Chem. Commun. 3829 (2005).

18. K. Maeda and K. Domen, J. Phys. Chem. Lett., 1, 2655 (2010)

19. K. Maeda, M. Higashi, D. Liu, R. Abe and K. Domen, J. Am. Chem. Soc., 132, 5858 (2010)

20. J.R. Maiolo III, B.M. Kayes, M.A.Filler, M.C. Putnam, M.D. Kelzenberg, H.A. Atwater,

N.S. Lewis, J. Am. Chem. Soc. 129, 12346 (2007)

21. S.W. Boettcher, J.M. Spurgeon, M.C. Putnam, E.L.Warren, D.B. Turner-Evans,

M.D.Kelzenberg, J. R. Maiolo, H.A. Atwater, N.S. Lewis, Science 327, 185(2010)

22. E.L. Warren, S. W.Boettcher, J.R. McKone, N.S. Lewis, Proceedings of SPIE Volume 7770:

Solar Hydrogen and Nanotechnology V, H. Idriss and H. Wang eds. 77701F, (2010).

23. M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. X. Mi, E. A. Santori and N.

S. Lewis, Chem. Rev., 110, 6446 (2010).

24. L. Vayssieres, C. Sathe, S.M. Butorin, D.K. Shuh, J. Nordgren, J.H. Guo, J. H., Adv. Mater.

17, 2320 (2005).

25. Y. Li , T. Takata , D. Cha , K. Takanabe , T. Minegishi, J. Kubota, K. Domen; Adv. Mater.

2012

26a, 26b, 26c. Domen et al., J. Phys. Chem. B, 107 (2003) 1798; Catal. Today 78 (2003) 555; J.

Am. Chem. Soc., 2010, 132 (34)



A-55

PEC White Paper Appendix: Theoretical Studies

Design and Characterization of

Photoelectrodes from First Principles

Team Affiliation Task

Tadashi Ogitsu

Brandon Wood

Wooni Choi

Lawrence Livermore National

Laboratory

Photocatalysts-electrolyte

interface

Muhammad N. Huda University of Texas at Arlington Novel photo-catalytic

materials design

Su-Huai Wei National Renewable Energy

Laboratory

Corrosion mitigation strategy

Although significant performance improvements have been realized since the

first demonstration of sunlight-driven water splitting in 1972, mainstream

adoption of photoelectrochemical (PEC) cells remains limited by an absence

of cost-effective electrodes that show simultaneously high conversion

efficiency and good durability. Here we outline current and future efforts to

use advanced theoretical techniques to guide the development of a durable,

high-performance PEC electrode material. Working in close collaboration

with experimental synthesis and characterization teams, we use a twofold

approach focusing on: 1) rational design of novel high-performance electrode

materials by methods beyond traditional band-engineering; and 2)

characterization and optimization of the electrode-electrolyte interface.

Introduction A photoelectrochemical (PEC) hydrogen production device uses sunlight and water to

generate hydrogen gas with no adverse emissions, and as such is considered an ideal sustainable

energy solution. Since the first successful demonstration of hydrogen production from sunlight

and water using a TiO2 photoelectrode in 1972,2 steady improvement on solar-to-hydrogen

(STH) efficiency has been made. For instance, in 1997-1998, a silicon triple-junction solar

harvester combined with cobalt-based co-catalyst resulted in a STH efficiency of about 8%.3, 4

In

1998, the current record STH efficiency of 12.4% was established at NREL using a

GaInP2/GaAs tandem cell with a Pt co-catalyst.5

Despite these breakthroughs, the PEC research community has faced great challenges in

achieving high STH and durability simultaneously, which has impeded commercial use of PEC

technology. Since no known material currently satisfies established U. S. Department of Energy

target windows that would lead to widespread market adoption,6-12

it is highly desirable to devise

a targeted, rational approach for developing entirely new photoelectrode materials and/or surface

modifications. This has proven challenging, in part because it is not fundamentally understood

how the interplay between the various electrode materials properties impacts the overall device



A-56


performance. In this regard, advanced

theoretical techniques are ideally suited for

offering a detailed description of the

underlying properties, and as such could be an

extremely valuable tool for intelligently

guiding future design efforts.

In practice, many of the materials

properties that directly impact solar harvesting

efficiency, corrosion resistance, and catalytic

activity appear to be correlated. As a result,

optimizing one property (e.g., STH efficiency)

frequently compromises another (e.g.,

durability). One of the challenges has been to

understand the nature of this correlation in the

bulk electrode materials, and whether key

properties can be simultaneously optimized in

real devices. Accordingly, our first approach

focuses on the use of first-principles density

functional theory (DFT) calculations to

investigate mechanisms for independently

improving intrinsic semiconductor properties

to achieve improved performance. The understanding provided by these studies has led to

consistent modification of PEC material design strategies over the last decade (see Fig. 1).1, 13-33

A second challenging aspect of this problem is the lack of information on the microscopic

properties of electrode-electrolyte interface. When the electrode is immersed in electrolyte (even

without illumination), the surface becomes contaminated by foreign chemical species. When the

electrode is illuminated, the situation becomes

even more complex. In addition to hydrogen

and oxygen evolution, various additional types

of chemical reactions, including photocorrosion,

can be driven at the electrode-electrolyte

interface by photogenerated carriers. Without

detailed information on the microscopic

structure and chemistry of the interface,

formulating a consistent strategy to optimize

interfacial properties becomes extremely

difficult, if not impossible. In order to better

understand the active interfacial processes, we

have begun performing large-scale first-

principles molecular dynamics simulations to

examine the structure and reactivity of realistic

electrode-electrolyte interfaces (see Fig 2).29-33

Such simulations have become possible only in

recent years, thanks to significant advances in supercomputer performance and novel software

algorithms

Figure 1: Cartoon images describing one possible strategy for

improving TiO2 electrode properties.1 From left to right:

Schematic electronic structure of TiO2;6) H2/O2 water redox

potentials; schematic band structure of an ideal PEC electrode;

plots on levels of atomic orbitals (from ref [23]). TiO2 has a

conduction band (CB) that is too low with respect to the water

redox potentials and possesses localized character owing to the Ti

3d states; similarly, the O 2p-derived valence band (VB) is too

low and localized. An ideal material should have CB and VB

edges straddling the H2/O2 redox potentials so as to overcome

any electron-transfer reaction barriers with minimal energy loss,

and possessing delocalized character to ensure good carrier

transport. The right-hand plot shows that one can rationally

substitute atoms in order to improve the band alignment and the

transport, for instance by substituting Ti with Ta (or W) and O

with N.1

Figure 2: It was demonstrated that the local bonding topologies

of an oxidized III-V surface are a descriptor of chemical

activity (left). This local model approach was successfully

applied to rationalize the chemical activities observed in the

direct first-principles molecular dynamics simulations of the

water/III-V interface (right). From ECS presentation, Boston

(2011) by Wood, Schwegler, and Ogitsu.



A-57


To summarize, our theory effort focuses on two areas: electrode material design and

characterization of interfacial properties. Dr. Huda at UTA will refine and deploy the materials

design strategies by innovative mineral database search algorithm and calculate the materials’

electronic and optical properties at both crystalline and nano-crystalline phases. The LLNL team

will continue their theoretical studies on the semiconductor-electrolyte interface using first-

principles molecular dynamics simulations. Their focus will be on investigating surface

stabilization and activation mechanisms based on modification of the interfacial structure and

chemistry. Dr. Wei at NREL will develop corrosion mitigation strategy based on modification of

activation potentials of oxidation and reduction reactions. In all cases, information obtained from

these activities will be shared with the collaborators on the materials synthesis and

characterization teams, as well as the wider PEC research community, in order to accelerate the

development of a viable PEC electrode material.

Task I: Rational Design of Electrode Materials (Task lead by Huda at UTA) Focus: Theoretically/computationally design of novel crystalline and nano-crystalline

photocatalyst materials and their simultaneous optimization of electronic, optical, and transport

properties.

A general scheme to predictively tune materials band structure properties are following: (1) By

isovalent doping, (2) by passivated co-doping, (3) By predicting novel alloys, and (4) By

modifying the shape and size of the materials, e.g., by forming nanostructures.

Our proposed research topics mainly targeting number 3 and 4 of the above scheme:

(i) Novel photocatalysts prediction by mineral database searching: The discovery of

efficient photo-catalysts is one of the grand challenges for energy conversion and storage. As

naturally occurring materials do not fulfill all the required electronic criteria, these electronic

requirements in materials are usually achieved by a band-engineering approach, where the

electronic structures of the materials are engineered by selective doping. However, the

introduction of impurities generally creates unwanted defect states in band gap, which are

detrimental to the transport properties due to poor crystallinity34

. To-date shifting the optical

absorption spectrum of a material to the visible region by doping-only process has not been

successful in improving the photoconversion efficiency significantly.

Instead of following a simple band-engineering-only approach, we plan to follow a

“natural selection” process. As the minerals were already formed in earth over millions of years,

they clearly possess thermodynamic stability, and represent ideal candidates to design materials

for energy-related applications. Though these naturally occurring minerals by themselves may

not be directly suitable for energy conversion, the knowledge of chemical compositions of these

minerals may lead to a proper photo-conversion material. Their compositions and properties can

easily be determined by theory calculations, and, instead of doping, a new alloy based on the

selected mineral will be predicted. Once predicted by theory, these existing but untested

structures then can be evaluated experimentally for photocatalysis. In this proposed research, we

will also develop an efficient selection algorithm for mineral database search.

(ii) Nano-crystalline photocatalysts: The current understanding at the “nano” level of

oxides is not very clear, and leads to misleading assumptions to the photocatalytic potential of



A-58


nano-crystals. Thus understanding of size- and shape-induced effects is crucial for efficient

design of metal-oxide nano-crystalline photocatalysts. First principle theories, such as DFT and

time-dependent-DFT (TDDFT), are state of the art theoretical methods to shed light in these

aspects. Several key issues remain challenging in metal-oxide nano-crystals so far; some of them

are: (i) identification of the fundamental gap and the actual optical gap, (ii) the nature of energy

levels (“band”) in the nano-crystals, (iii) the effects of surface and surface passivation, (iv) the

difference between the flat band potential (surface) and the band position in the “bulk” of nano-

crystals, (v) transport of charge carriers after photo-excitation, and (vi) extraction or injection of

electrons from or to the nano-crystals’ surface.

We have recently shown that a unique set of self-passivated and charge compensated

delafossite nanocrystals can be highly stable with some interesting optical properties. These

behavior needs to be explored in other metal oxides nanocrystals with various sizes and shaps as

well.

Task II: Characterization and Optimization of the Electrode-

electrolyte Interface (Task lead by Ogitsu at LLNL)

Focus: Gain atomistic insight into the properties of electrode-electrolyte interface in order to

develop a corrosion mitigation strategy and to improve STH conversion efficiency.

LLNL’s plan initially involves investigation of the GaInP2 (001) surface both with and

without an electrolyte. This will be done in order to understand which microscopic interfacial

properties are necessary to achieve high STH, and how these are affected by the presence of an

electrolyte. We will then simulate and compare the in-situ XAS/XES data for our model systems

with experimental results from the UNLV surface characterization team, which will allow us to

verify which of our proposed atomistic mechanisms are indeed related to hydrogen evolution and

corrosion. Simulation results will be compared with data from NREL’s recent attempts at surface

stabilization via nitrogen treatment. In doing so, we aim to extract a mechanism for atomistic

surface stabilization and to develop an improved corrosion mitigation strategy for GaInP2. It is

expected that these efforts will interface closely with and provide feedback to the rational

materials design effort (Task I), particularly given that Dr. Wei is recognized as one of the

leading experts on GaInP2.35-38

i) Simulation of the GaInP2(001)-water interface: The first subtask involves the

application of previously established static and dynamic methods for GaP(001) and InP(001) to

the alloy material GaInP2(001). In addition to extracting structural motifs that can be used in

subsequent models, we will assess the role of alloy structure on the chemical properties of the

interface. Simulations will be done with and without additional surface contaminants.

ii) Investigation of subsitutional impurities: In FY12, T. Deutsch et al. (NREL)

discovered that a specific type of nitrogen incorporation to GaInP2 electrode may improve the

durability significantly with an acceptably minimal impact on STH efficiency. For our second

subtask, we intend to investigate in detail the effect of nitrogen doping on surface corrosion

resistance in order to gain microscopic insight on the atomistic surface stabilization mechanism.



A-59


iii) Simulation of XAS/XES spectra: For our third subtask, we plan to simulate X-ray

absorption (XAS) and emission (XES) spectra of reference model surfaces. These will be done

on clean surfaces, as well as oxygen- and hydroxyl-contaminated surfaces, which are better

approximations to the electrode structure when immersed in electrolyte. The goal of this stage is

to establish and validate accurate models for the precursory interfacial structure present prior to

illumination. For this part of research activity, the LLNL team is collaborating with Dr.

Prendergast at LBNL (Computational Spectroscopy Group of The Molecular Foundry). Notably,

optimized computational procedures for simulating the P-L2,3 edge XAS/XES of GaInP2 have

already been established and tested.

iv) Surface oxide modeling: Experimental characterization of actual GaInP2 electrodes by

the UNLV team found a >1nm native oxide on the surface. If resources permit, we will develop

and investigate indium/gallium oxide-water interface models, including the generation of

corresponding computational spectra to compare with the UNLV/NREL results. These could be

particularly useful for detailing the atomistic corrosion mechanism, as well as its relationship

with the hydrogen evolution mechanism. An initial comparison of the experimental data with our

current models should allow us to better assess the necessity of these additional spectra. If

necessary, we will revise our model structures until we come to a satisfactory agreement between

experiment and theory.

Task III: Corrosion Mitigation Strategy (Task lead by Wei at NREL)

Focus: Investigate on energy barrier of oxidation reaction, and develop corrosion mitigation

strategy

The photocorrosion in semiconductor is mainly due to photo generated holes which oxide

the semiconductor or photo generated electrons that reduce the semiconductor. In order to

improve the durability we have to either migrate the photo generated holes or electrons to other

materials, which are stable for water splitting, or find semiconductors with chemical potentials of

all semiconductor oxidation reactions below the water oxidation potential and chemical

potentials of all semiconductor reduction reactions above the water reduction potential, so the

photo generated holes and electrons will relax to water oxidation and reduction levels, thus

driving the water splitting reactions instead of driving the semiconductor reactions. For example,

we calculated the electron potential of GaP reduction reaction and hole potential of GaP

oxidation reaction as shown in Fig. 7. The GaP reduction potential is found to be above the water

reduction potential and thus photo generated electron will most likely drive the H2O/H2 reaction.

The GaP oxidation potential is, however, above the water oxidation potential and thus hole will

most likely drive the GaP oxidation instead of O2/H2O reaction. If we can engineer the GaP

surface to create an energy barrier larger than the energy difference between GaP oxidation and

water oxidation potentials to block the GaP oxidation reaction, then we may be able to

significantly improve the durability of GaP as water splitting photocatalysts.



A-60


Figure 3: Band alignment of GaP with respective to water redox potentials as well as the electron potential of GaP

reduction reaction and hole potential of GaP oxidation reaction.

Summary In summary, we have outlined a twofold approach that uses advanced theoretical

techniques to develop design and optimization strategies for efficient, durable PEC

photoelectrodes. First, we use first-principles density functional theory calculations to

successfully predict novel photocatalyst materials of new composition by an intelligent search of

the mineral database and asses their electronic, optical, and transport properties at both

crystalline and nano-crystalline phase for further screening. Second, we use first-principles

molecular dynamics simulations and model free-energy reaction barrier calculations to examine

the structure and chemistry of the surface, with the goal of improving device stability and

performance. Third, based on the free-energy reaction barrier calculations, we will develop

corrosion mitigation strategy. Together with the efforts of other members of the DOE/EERE

Photoelectrochemical Hydrogen Production Working Group, the theory team’s input should

provide much-needed insight into how specific photoelectrode materials properties should be

combined so as to engineer and optimize devices that meet the DOE market adoption targets.

References 1

W.-J. Yin, H. Tang, S.-H. Wei, M. M. Al-Jassim, J. Turner, and Y. Yan, Phys Rev B 82

(2010). 2

A. Fujishima and K. Honda, Nature 238, 37 (1972).



A-61


3 R. E. Rocheleau and E. L. Miller, Int J Hydrogen Energ 22, 771 (1997).

4 R. E. Rocheleau, E. L. Miller, and A. Misra, Energy & Fuels 12, 3 (1998).

5 O. Khaselev and J. A. Turner, Science 280, 425 (1998).

6 "Hydrogen and Fuel Cells Program Plan", (Department of Energy, Washington D.C.,

2011) 7

"Fuel Cell Technologies Program Multi-Year Research, Development and Demonstration

Plan: 3.1 Hydrogen Production, 2011 Interim Update", (Department of Energy,

Washington D.C., 2011) 8

"Basic Research Needs for Solar Energy Etilization", (Department of Energy, Washington

D.C., 2005) 9

"Hydrogen Posture Plan", (Department of Energy, Washington D.C., 2006) 10

"Basic Research Needs for the Hydrogen Economy", (Department of Energy, Washington

D.C., 2004) 11

"National Hydrogen Energy Roadmap", (Department of Energy, Washington D.C., 2002) 12

"A National Vision of Ameriaca's Transition to A Hydrogen Economy -- To 2030 and

Beyond", (Department of Energy, Washington D.C., 2002) 13

A. Walsh, J. L. F. Da Silva, Y. Yan, M. M. Al-Jassim, and S.-H. Wei, Phys Rev B 79

(2009). 14

A. Walsh, et al., Energ Environ Sci 2, 774 (2009). 15

A. Walsh, Y. Yan, M. M. Al-Jassim, and S.-H. Wei, Journal of Physical Chemistry C 112,

12044 (2008). 16

A. Walsh, Y. Yan, M. N. Huda, M. M. Al-Jassim, and S.-H. Wei, Chem Mater 21, 547

(2009). 17

M. N. Huda, Y. F. Yan, S. H. Wei, and M. M. Al-Jassim, Phys Rev B 78 (2008). 18

M. N. Huda, Y. Yan, A. Walsh, S.-H. Wei, and M. M. Al-Jassim, Appl Phys Lett 94 (2009). 19

M. N. Huda, Y. Yan, A. Walsh, S.-H. Wei, and M. M. Al-Jassim, Phys Rev B 80 (2009). 20

M. N. Huda, Y. Yan, C.-Y. Moon, S.-H. Wei, and M. M. Al-Jassim, Phys Rev B 77 (2008). 21

W.-J. Yin, S.-H. Wei, M. M. Al-Jassim, and Y. Yan, Phys Rev Lett 106 (2011). 22

W.-J. Yin, S.-H. Wei, M. M. Al-Jassim, and Y. Yan, Appl Phys Lett 99 (2011). 23

W.-J. Yin, S.-H. Wei, M. M. Al-Jassim, J. Turner, and Y. Yan, Phys Rev B 83 (2011). 24

M. N. Huda, Y. Yan, A. Walsh, S.-H. Wei, J. A. Turner, and M. M. Al-Jassim, in Solar

Hydrogen and Nanotechnology V, edited by H. W. H. Idriss, 2010), Vol. 7770. 25

M. N. Huda, A. Walsh, Y. Yan, S.-H. Wei, and M. M. Al-Jassim, J Appl Phys 107 (2010). 26

C. Feng, W.-J. Yin, J. Nie, X. Zu, M. N. Huda, S.-H. Wei, M. M. Al-Jassim, J. A. Turner,

and Y. Yan, Appl Phys Lett 100 (2012). 27

Y. Gai, J. Li, S.-S. Li, J.-B. Xia, and S.-H. Wei, Phys Rev Lett 102, 036402 (2009). 28

W.-J. Yin, S. Chen, J.-H. Yang, X.-G. Gong, Y. Yan, and S.-H. Wei, Appl Phys Lett 96,

221901 (2010). 29

"Characterization and Optimization of PhotoelectrodeSurfaces for Solar-to-Chemical Fuel

Conversion", (Department of Energy, Washington D.C., 2010) 30

B. Wood, T. Ogitsu, and E. Schwegler, in SPIE Optics + Photonics 2010 (SPIE, San Diego,

2010). 31

B. C. Wood, T. Ogitsu, and E. Schwegler, Journal of Photonics for Energy 1, 016002

(2011). 32

B. C. Wood, T. Ogitsu, and E. Schwegler, The Journal of Chemical Physics 136, 064705

(2012).



A-62


33 "Characterization and Optimization of PhotoelectrodeSurfaces for Solar-to-Chemical Fuel

Conversion", (Department of Energy, 2011) 34

K.-S. Ahn, Y. Yan, and M. Al-Jassim, Journal of Vacuum Science & Technology B:

Microelectronics and Nanometer Structures 25, L23 (2007). 35

S. H. Wei and A. Zunger, Appl Phys Lett 56, 662 (1990). 36

S.-H. Wei and A. Zunger, Phys Rev B 57, 8983 (1998). 37

A. Franceschetti, S.-H. Wei, and A. Zunger, Phys Rev B 52, 13992 (1995). 38

S.-H. Wei, D. B. Laks, and A. Zunger, Appl Phys Lett 62, 1937 (1993).



A-63

Appendix B

Technoeconomic Analysis of Photoelectrochemical (PEC) Hydrogen

Production

Acknowledgements ________________________________________________________________________ B-3

Table of Contents ___________________________________________________________________________ B-4

Table of Figures ____________________________________________________________________________ B-7

Executive Summary _____________________________________________________________________ B-10

B-1

Technoeconomic Analysis of

Photoelectrochemical (PEC)

Hydrogen Production

Final Report December 2009

Prepared by: Brian D. James

George N. Baum Julie Perez

Kevin N. Baum

One Virginia Square 3601 Wilson Boulevard, Suite 650

Arlington, Virginia 22201 (703) 243-3383

DOE Contract Number: GS-10F-009J DOE Technical Monitor: David Peterson

Deliverable Task 5.1: Draft Project Final Report

B-2

Acknowledgements

The authors of this report would like to acknowledge the contributions of the US Department of Energy (DOE) Photoelectrochemical (PEC) Hydrogen Working Group, whose members provided invaluable technical and programmatic guidance throughout the analysis project.

While the working group consists of approximately 31 members from academia, industry, and the DOE, a core group of six individuals worked closely with DTI to assist in the technical direction of the analysis. These members merit particular recognition:

Dr. Eric Miller, University of Hawaii, PEC H2Ms. Roxanne Garland, Technology Development Manager, US DOE, EER&E

Working Group Leader

Dr. Eric McFarland, University of California, Santa Barbara Dr. Thomas Jaramillo, Stanford University Dr. John Turner, National Renewable Energy Laboratory (NREL) Mr. Robert Perret, Nevada Technical Services, LLC.

The work described in this report was performed under contract to the US Department of Energy, Office of Fuel Cell Technologies, DOE Office of Energy Efficiency and Renewable Energy.

B-3

http://www1.eere.energy.gov/hydrogenandfuelcells/pdfs/pec_technoeconomic_analysis.pdf

Table of Contents 1. Executive Summary ................................................................................................................ 92. Introduction ........................................................................................................................... 223. PEC Operating Principles ..................................................................................................... 22

3.1 PEC Electrolysis ............................................................................................................. 223.2 PEC Efficiency ............................................................................................................... 233.3 PEC Reactor Types ........................................................................................................ 243.4 PEC Optical Windows ................................................................................................... 24

3.4.1 PEC Reactor Window Refraction/Reflection Effects ............................................. 243.4.2 Type 1 and 2 PEC Cell Window Transmittance ..................................................... 263.4.3 Type 3 and 4 PEC Window Transmittance ............................................................. 273.4.4 Window Chemical Properties ................................................................................. 293.4.5 Type 1 and Type 2 Covered Pond Water Vaporization .......................................... 29

3.5 Solar Insolation .............................................................................................................. 303.5.1 Type 1 and Type 2 System Insolation ..................................................................... 303.5.2 Type 3 Tilted Planar Array System Insolation ....................................................... 343.5.3 Type 4 Tracking Concentrator System Insolation .................................................. 37

3.6 Solar Shadowing ............................................................................................................ 413.6.1 Type 3 Panel Separation Distance for Minimal Shadowing ................................... 413.6.2 Type 4 Panel Separation Distance for Low Shadowing ......................................... 41

4. Photoelectrolysis Reactor Engineering Designs and Costs .................................................. 424.1 Type 1 Single Bed Colloidal Suspension Reactor ......................................................... 44

4.1.1 Photoelectrode Reactor Bed Particles ..................................................................... 444.1.2 Nanoparticle Fabrication and Cost .......................................................................... 464.1.3 Type 1 Solar-to-Hydrogen Conversion Efficiency ................................................. 474.1.4 Type 1 Reactor Bed ................................................................................................ 484.1.5 Bed Trough System ................................................................................................. 494.1.6 Continuous Bag (Baggie) System ........................................................................... 494.1.7 Plastic Films ............................................................................................................ 504.1.8 Ports ........................................................................................................................ 514.1.9 Laminating and Sealing Machines .......................................................................... 514.1.10 Bed Headspace Considerations ............................................................................... 524.1.11 Capital Costs ........................................................................................................... 524.1.12 Type 1 Reactor Cost Summary ............................................................................... 53

4.2 Type 2 Dual Bed Colloidal Suspension Reactor ............................................................ 544.2.1 Photoelectrode Reactor Bed Particles ..................................................................... 544.2.2 Nanoparticle Cost for Type 2 System ..................................................................... 544.2.3 Type 2 Solar-to-Hydrogen Conversion Efficiency ................................................. 554.2.4 Type 2 Reactor Bed ................................................................................................ 554.2.5 Dual Bed Reactor Assembly ................................................................................... 564.2.6 Type 2 Reactor Cost Summary ............................................................................... 574.2.7 Type 1 and Type 2 Reactor Bed Technology Summary ......................................... 58

4.3 Type 3 and 4 PEC System PV Cell Properties, Fabrication, and Cost .......................... 584.3.1 Photocell PEC Operation ........................................................................................ 584.3.2 PEC Photocell Cost Factors .................................................................................... 62

B-4


4.3.3 PEC Photocell Cost Prediction ............................................................................... 644.4 Type 3 Planar Array System .......................................................................................... 68

4.4.1 PEC Planar Array Design ....................................................................................... 694.5 Type 4 Solar Concentrator System ................................................................................. 71

4.5.1 Concentrator PEC Cell Technology ........................................................................ 714.5.2 Solar/Thermal Concentrators .................................................................................. 734.5.3 PEC Solar Concentrator Design for this Study ....................................................... 744.5.4 Solar Collector Costs .............................................................................................. 76

4.6 Summary of All PEC Reactor Systems .......................................................................... 785. Gas Processing Subassembly ................................................................................................ 79

5.1 H2-O2 Gas Mixture Safety ............................................................................................. 795.2 Compressors ................................................................................................................... 805.3 Gas Cooling and Water Vapor Removal ........................................................................ 815.4 Hydrogen Separation from Contaminants ...................................................................... 82

5.4.1 Pressure Swing Adsorption (PSA) .......................................................................... 825.4.2 Other Separation Methods ...................................................................................... 84

5.5 Piping ............................................................................................................................. 865.6 Capital Costs of Gas Processing Components ............................................................... 87

6. Control System Subassembly ............................................................................................... 886.1 Components .................................................................................................................... 886.2 Wiring ............................................................................................................................. 906.3 Capital Costs .................................................................................................................. 92

7. General Cost Assumptions and Calculations ........................................................................ 937.1 Default H2A Parameters ................................................................................................ 937.2 System Common Parameters ......................................................................................... 94

7.2.1 Operating Capacity Factor ...................................................................................... 957.2.2 Reference Year Dollars ........................................................................................... 957.2.3 Site Preparation Parameter ...................................................................................... 957.2.4 Engineering & Design Parameter ........................................................................... 957.2.5 Process Contingency Parameter .............................................................................. 957.2.6 Project Contingency Parameter ............................................................................... 957.2.7 Up-Front Permitting Costs ...................................................................................... 96

7.3 System Specific Parameters ........................................................................................... 967.3.1 Baseline Uninstalled Costs ..................................................................................... 967.3.2 Installation Cost Factor ........................................................................................... 96

8. Specific System Capital Costs .............................................................................................. 968.1 Reactor Costs .................................................................................................................. 96

8.1.1 Type 1 and Type 2 Reactor Nanoparticle Costs ..................................................... 968.1.2 Baggie Sizing .......................................................................................................... 968.1.3 Quantity of Baggies ................................................................................................ 978.1.4 Excavation of Land for Reactor Bed Placement ..................................................... 988.1.5 Type 1 and Type 2 Reactor Costs ........................................................................... 998.1.6 Type 3 and Type 4 Reactor Costs ......................................................................... 100

8.2 Piping Costs ................................................................................................................. 1018.3 Pump Costs ................................................................................................................... 1028.4 Compressor, Heat Exchangers, and PSA ..................................................................... 103

B-5


8.5 Land Required .............................................................................................................. 1038.6 Capital Cost Summary ................................................................................................. 104

9. Specific System Operating Costs ........................................................................................ 1059.1 Electricity Consumption ............................................................................................... 1059.2 Utility Usage ................................................................................................................ 1069.3 Yearly Replacement Costs ........................................................................................... 1069.4 Yearly Maintenance Costs ........................................................................................... 1079.5 Production Facility Plant Staff ..................................................................................... 107

10. PEC System and Hydrogen Production Cost Results ...................................................... 10810.1 Type 1 Single Bed Colloidal Suspension System .................................................... 10810.2 Type 2 Dual Bed Colloidal Suspension System ....................................................... 11210.3 Type 3 PEC Planar Array System ............................................................................ 11610.4 Type 4 PEC Tracking Solar Concentrator System ................................................... 120

11. Summary of Results and Conclusions for Levelized Hydrogen Costs ............................ 12411.1 PEC Hydrogen Production Systems ......................................................................... 12411.2 Hydrogen Production Cost Comparison ................................................................... 12411.3 H2 Cost Sensitivity to System Parameters ............................................................... 12511.4 Discussion of Results ................................................................................................ 126

11.4.1 Particle Bed PEC ................................................................................................... 12611.4.2 Photocell PEC ....................................................................................................... 127

11.5 PEC System Development Recommendations ......................................................... 127

B-6


Table of Figures Figure 1-1: Type 1 End View ....................................................................................................... 11Figure 1-2: Type 2 Multi-Baggie Assembly End View ................................................................ 12Figure 1-3: Schematic of a Generic PEC Photocell ...................................................................... 13Figure 1-4: Type 3 PEC Panel Layout .......................................................................................... 14Figure 1-5: Type 4 Concentrator PEC Design .............................................................................. 15Figure 1-6: PEC System Capital Cost Summary .......................................................................... 17Figure 1-7: Comparison of PEC System Levelized H2 Cost ........................................................ 18Figure 1-8: Hydrogen Cost Sensitivity Analysis Results ............................................................. 18Figure 3-1: Bandgap Energy for PV Materials ............................................................................. 23Figure 3-2: PEC Window Refraction ............................................................................................ 25Figure 3-3: PEC Window Refraction ............................................................................................ 25Figure 3-4: PEC Window Reflectance using Fresnel Reflection Equations ................................. 26Figure 3-5: Plexiglas Specification Window Reflectance ............................................................ 26Figure 3-6: Spectral Transmission - Plexiglas Acrylic ................................................................. 27Figure 3-7: Spectral Transmission - Lexan Polycarbonate ........................................................... 28Figure 3-8: Solar Spectrum ........................................................................................................... 28Figure 3-9: Plexiglas Chemical Resistance ................................................................................... 29Figure 3-10: Yearly Average Solar Irradiance On Horizontal Surfaces ....................................... 31Figure 3-11: Monthly Variation of Daily Radiation on Horizontal Surface ................................. 32Figure 3-12: Hourly Irradiance on a Horizontal Surface .............................................................. 32Figure 3-13: Variation of Daily Refracted Radiation on Horizontal Surface over a Year ........... 33Figure 3-14: Hourly Refracted Input Energy on a Horizontal Surface ......................................... 33Figure 3-15: Yearly Average Solar Irradiance on Surface Inclined at Local Latitude ................. 34Figure 3-16: Monthly Variation of Daily Radiation on Surface Inclined at 35o Latitude Angle . 35Figure 3-17: Hourly Irradiance on a 35o Inclined Surface ............................................................ 36Figure 3-18: Variation of Daily Refracted Radiation on 35o Inclined Surface ............................. 36Figure 3-19: Hourly Refracted Input Energy on 35o Tilted Panel ................................................ 37Figure 3-20: Yearly Average Solar Direct Normal Irradiance ..................................................... 38Figure 3-21: Monthly Variation of Daily Radiation on Tracking Concentrator (No Shading) .... 38Figure 3-22: Daily Variation of Radiation Incident on Tracking Concentrator (No Shading) ..... 39Figure 3-23: Average Day Hourly Irradiance with Inter-array Shading and Window Loss ........ 40Figure 3-24: Average Month’s Daily Radiation on Tracking Concentrator (With Shading) ....... 40Figure 3-25: Type 3 Shadowing North-South Separation Limit .................................................. 41Figure 3-26: Type 4 Shadowing East-West Separation Limit ...................................................... 42Figure 3-27: Type 4 Shadowing North-South Separation Limit .................................................. 42Figure 4-1: Type 1 PEC Nanoparticle Structure ........................................................................... 45Figure 4-2: PEC Nanoparticle Micrograph ................................................................................... 45Figure 4-3: Particle Coating Major Assumptions ........................................................................ 46Figure 4-4: PEC Nanoparticle Production Cost Breakdown ....................................................... 47Figure 4-5: Nanoparticle Cost vs. Annual Production Rate ........................................................ 47Figure 4-6: Type 1 System STH Bounds ...................................................................................... 48Figure 4-7: Type 1 Colloidal Suspension Reactor ........................................................................ 48Figure 4-8: End View of Type 1 Baggie Configuration ............................................................... 49Figure 4-9: Top view of Type 1 Baggie Configuration ................................................................ 50

B-7


Figure 4-10: Port Specs and Installation Costs ............................................................................. 51Figure 4-11: Production Specifications for Laminating and End Sealing Machines .................... 52Figure 4-12: Headspace Daily Vertical Rise in 40 ft Wide Bed (to scale)– June 21 .................... 52Figure 4-13: Capital Recovery Factor for Baggie Production (1 tonne/day H2 ) ......................... 53Figure 4-14: Type 1 Reactor Capital Costs for 1TPD .................................................................. 53Figure 4-15: Type 2 PV Nanoparticle Structures ......................................................................... 54Figure 4-16: Type 2 System STH Bounds .................................................................................... 55Figure 4-17: Type 2 Dual Bed Colloidal Suspension Reactor ...................................................... 56Figure 4-18: Type 2 End View of a Dual Bed Reactor Assembly ................................................ 56Figure 4-19: Top View of Two Type 2 Reactor Bed Assemblies ................................................ 57Figure 4-20: Type 2 Reactor Capital Costs for 1TPD .................................................................. 57Figure 4-21: PEC Multilayer Cell Configuration ......................................................................... 59Figure 4-22: PEC Configuration from Gibson Patent .................................................................. 59Figure 4-23: PEC Configuration from Gratzel Patent ................................................................. 60Figure 4-24: PEC Configuration from McNulty Patent Application ........................................... 60Figure 4-25: PEC Cell Research Descriptions and Performance .................................................. 61Figure 4-26: PV Cell Efficiencies ................................................................................................. 62Figure 4-27: Nanosolar Roll Printing ........................................................................................... 63Figure 4-28: Layout of the Nanosolar PV cell .............................................................................. 63Figure 4-29: Baseline PEC Cell Cost Model based on NREL Solar Cell Cost Study ................. 66Figure 4-30: Nanosolar PV Cell Diagram ................................................................................... 67Figure 4-31: Future Projected PEC Cell Cost Model based on DFMA Analysis ........................ 68Figure 4-32: Layout of Type 3 PEC Panel .................................................................................... 69Figure 4-33: Analogous PV Array Structure ................................................................................ 70Figure 4-34: Type 3 Planar Array Field Layout ........................................................................... 70Figure 4-35: Variation of PV Efficiency with Concentration ...................................................... 72Figure 4-36: Entech PV Refractor System 22:1 concentration ratio ............................................ 73Figure 4-37: Junction Tracking Solar Thermal Trough Array ...................................................... 73Figure 4-38: Offset Parabolic Cylinder Reflector PEC ................................................................ 74Figure 4-39: Receiver Solar Input Diagram .................................................................................. 75Figure 4-40: Receiver Details ....................................................................................................... 75Figure 4-41: Type 4 Baseline 1 TPD System Layout ................................................................... 76Figure 4-42: Type 4 Solar Collector Cost Estimate ...................................................................... 77Figure 4-43: Type 4 Cost Reduction with Increased Concentration Ratio ................................... 77Figure 4-44: Summary of Reactor Parameters for 1TPD PEC Systems ....................................... 78Figure 5-1: Gas Processing Components ...................................................................................... 79Figure 5-2: Explosion Limit Pressure/Mixture Dependence for H2/O2 Mix ............................... 80Figure 5-3: Water Vapor Fractions ............................................................................................... 82Figure 5-4: PSA Sizing for Absorption of Oxygen Contaminant Gas .......................................... 84Figure 5-5: Nano-porous Membranes ........................................................................................... 86Figure 5-6: Gas Heat Exchanger and Compressor Properties ..................................................... 87Figure 5-7: Capital Costs of Gas Processing Components (without piping) ................................ 88Figure 6-1: Control System Components ..................................................................................... 89Figure 6-2: Sensors and Controllers ............................................................................................. 90Figure 6-3: Typical Subassembly Design Showing Control System - Type 1 ............................. 91Figure 6-4: Control System Wiring and Conduit Quantities ........................................................ 91

B-8


Figure 6-5: Capital Costs of Control System Components ........................................................... 92Figure 7-1: H2A Default Values used for all PEC Systems ......................................................... 93Figure 7-2: Parameters Common to All Systems ........................................................................ 94Figure 7-3: System Specific Parameters ...................................................................................... 96Figure 8-1: Davis-Bacon Hourly Wage Rates ............................................................................. 98Figure 8-2: Excavation Cost Estimate for Type 1 and Type 2 Systems using Arizona Costs ...... 99Figure 8-3: Type 1 and Type 2 Baseline Reactor Costs .............................................................. 99Figure 8-4: Type 3 and Type 4 Baseline Reactor Costs ............................................................ 100Figure 8-5: Summary of Reactor Parameters for PEC Systems (1 Tonne H2/day Module) ...... 101Figure 8-6: Piping Sizes and Unit Costs for PEC Systems ......................................................... 102Figure 8-7: Cooling water needs for Heat Exchangers .............................................................. 103Figure 8-8. Gas Processing Major Component Cost .................................................................. 103Figure 8-9: Land Requirements ................................................................................................. 104Figure 8-10: Capital Cost Summary .......................................................................................... 104Figure 9-1: Electrical Power Consumption (average power over year) ..................................... 105Figure 9-2: Utilities Usage .......................................................................................................... 106Figure 9-3: Replacement Costs ................................................................................................... 106Figure 9-4: Plant Staff Requirements for Baseline Plants .......................................................... 107Figure 10-1: End View of Three Type 1 Single Bed Baggie Reactors ....................................... 108Figure 10-2: Bill of Materials for Installed Type 1 Baseline 1TPD System .............................. 109Figure 10-3: Type 1 Baseline 1 TPD System Direct Capital Components ................................. 110Figure 10-4: Type 1 Baseline 10 TPD H2 Production Cost Elements - $1.63kg ....................... 110Figure 10-5: Type 1 Sensitivity Analysis Parameters ............................................................... 111Figure 10-6: Type 1 H2 Cost Sensitivity ($/kgH2) ..................................................................... 111Figure 10-7: Type 2 Reactor Unit ............................................................................................... 112Figure 10-8: Type 2 Total System Layout .................................................................................. 112Figure 10-9: Bill of Materials for Installed Type 2 Baseline 1TPD System .............................. 113Figure 10-10: Type 2 Baseline 1 TPD System Direct Capital Components ............................... 114Figure 10-11: Type 2 Baseline 10 TPD H2 production cost elements – $3.19/kg ..................... 114Figure 10-12: Type 2 Sensitivity Analysis Parameters .............................................................. 115Figure 10-13: Type 2 H2 cost sensitivity ($/kgH2)..................................................................... 115Figure 10-14: Type 3 Baseline System for 1TPD ....................................................................... 116Figure 10-15: Bill of Materials for Installed Type 3 Baseline 1TPD System ............................ 117Figure 10-16: Type 3 Baseline 1 TPD System Direct Capital Components ............................... 118Figure 10-17: Type 3 Baseline 10 TPD H2 Production Cost Elements – $10.36/kg H2 ............ 118Figure 10-18: Type 3 System Sensitivity Analysis Parameters ................................................. 119Figure 10-19: Type 3 Cost Sensitivities ($/kgH2) ...................................................................... 119Figure 10-20: Type 4 System Layout ......................................................................................... 120Figure 10-21: Bill of Materials for Installed Type 4 Baseline 1TPD System ............................ 121Figure 10-22: Type 4 Baseline 1 TPD System Direct Capital Components ............................... 122Figure 10-23: Type 4 Baseline 10 TPD H2 Production Cost Elements – $4.05/kg H2 .............. 122Figure 10-24: Type 4 Sensitivity Analysis Parameters .............................................................. 123Figure 10-25: Overall Type 4 Cost Sensitivities ($/kgH2) ......................................................... 123Figure 11-1: Levelized costs for H2 Produced by Baseline PEC Systems ................................. 124Figure 11-2: Sensitivity Analysis Parameters ............................................................................. 125Figure 11-3: Sensitivity Analysis Results ................................................................................... 125

B-9


1. Executive Summary

Photoelectrochemical (PEC) production of hydrogen is a promising renewable energy technology for generation of hydrogen for use in the future hydrogen economy. PEC systems use solar photons to generate a voltage in an electrolysis cell sufficient to electrolyze water, producing H2 and O2 gases. A major advantage of PEC systems is that they involve relatively simple processes steps as compared to many other H2 production systems. Additionally, they possess a wide operating temperature ranges, with no intrinsic upper temperature limit and a lower temperature of slightly below 0oC without a warm-up period, and well below 0oC with a warm-up period dependent on outside temperature. The primary challenges for PEC are to develop materials with sufficient photovoltage to electrolyze water, to minimize internal resistance losses, to have long lifetime (particularly corrosion life), to maximize photon utilization efficiencies, and to reduce plant capital cost.

Under contract to the US Department of Energy, Directed Technologies Inc. (DTI) has conducted a techno-economic evaluation of conceptual PEC hydrogen production systems. Four basic system configurations are chosen by DOE’s PEC Working Group to encompass the technology spread of potential future PEC production systems. Overall system designs and parameters, costs of implementation, and costs of the output hydrogen were determined for each of the four conceptual systems. Each system consisted of a PEC reactor that generates H2 and O2, a gas processing system that compresses and purifies the output gas stream, and ancillary equipment.

The first two of the four system configurations examined utilize aqueous reactor beds containing colloidal suspensions of PV-active nanoparticles, each nanoparticle being composed of the appropriate layered PV materials to achieve sufficient bandgap voltage to carry out the electrolysis reaction. The third and fourth system configurations use multi-layer planar PV cells in electrical contact with a small electrolyte reservoir and produce oxygen gas on the anode face and hydrogen gas on the cathode face. They are positioned in fixed or steered arrays facing the sun.

The four specific system types conceptually designed and evaluated in the report are: Type-1: A single electrolyte -filled reactor bed containing a colloidal suspension of PEC nanoparticles which produce a mixture of H2 and O2 product gases. Type-2: Dual electrolyte-filled reactor beds containing colloidal suspensions of PEC nanoparticles, with one bed carrying out the H2O => ½ O2 + 2 H+ half-reaction, the other bed carrying out the 2H+ => H2 half-reaction, and including a mechanism for circulating the ions between beds. Type-3: A fixed PEC planar array tilted toward the sun at local latitude angle, using multi-junction PV/PEC cells immersed in an electrolyte reservoir. Type-4: A PEC solar concentrator system, using reflectors to focus the solar flux at a 10:1 intensity ratio onto multi-junction PV/PEC cell receivers immersed in an electrolyte reservoir and pressurized to 300 psi.

B-10


The first step in determining hydrogen production was to evaluate the useable solar insolation levels for the four system types. For this report, we have assumed the location of the PEC reactors to be at Daggett, CA, at 35

Solar Irradiation

o North latitude near Barstow, CA. This is a high insolationNREL solar measurement site and is near the solar-thermal power field at Kramer Junction, CA. For solar inputs we used the solar radiation tables for ground radiation compiled by NREL in the Solar Radiation Data Manual1 and the SOLPOS2 program to calculate hourly solar variation over a year.

For Type 1 and Type 2 reactor beds, the solar input consists of the component of direct radiation incident on the horizontal bed plus the diffuse radiation from the sky. For these, the yearly mean of the average daily radiation energy input (after window refraction loss) is 5.55 kW-hr/m2 per 24 hour period. One issue for these horizontal bed PEC systems is that the H2 output variation between summer and winter can vary by a factor of 3.2 for a clear environment and by a greater factor in the event of extensive winter cloud cover. Since this study didn’t include a monthly hydrogen demand profile, the beds were sized for an average yearly production (averaging 1,000 kg H2/day over a year) without regard for potential seasional demand varations. The number of beds will need to be increased if the winter H2 demand is greater than 31% of the summer demand.

The Type 3 system fixed planar PEC cell panels are inclined toward the equator at an angle equal to local latitude. This inclination allows the array, in general, to maximize overall capture of direct solar flux throughout the year and results in a much more leveled output between summer and winter. The system captures the solar direct component determined by panel tilt angle and the solar zenith and azimuth angles, and also captures much of the diffuse radiation component. The yearly mean of the average daily radiation energy input (after window refraction and inter-array shading losses) is 6.19 kW-hr/m2 per 24 hour period.

The Type 4 system reactor consists of arrays that track solar direct radiation and focus the energy onto PEC receivers. While it captures the maximal direct solar radiation, it receives only a very small amount of diffuse radiation, since the concentrating mirrors have a narrow Field of View. The yearly mean of the average daily radiation energy input (after window refraction and inter-array shading losses) is 6.55 kW-hr/m2 per 24 hour period.

The Type 1 and 2 reactors are shallow horizontal pools or beds, filled with water, nanoparticles, and a KOH electrolyte, and having a flexible clear plastic thin-film envelope, or baggie, to contain the slurry and capture the gas produced while simultaneously allowing light to penetrate to the particles. With no gas production, the thin-film plastic cover will float on the water, however, as gas is produced, the cover will lift to accumulate output gas. The cover is sized to allow it to rise and fall over a 24 hour day and thus average out the gas flow to the gas handling

Type 1 and Type 2 Particle Bed Systems

1 Solar Radiation Data Manual for Flat Plate and Concentrating Collectors, 1961-1990, NREL Report, W. Merion, S. Wilcox. 2 NREL MIDS SOLPOS (Solar Position and Intensity) model, Distributed by the NREL , Center for Renewable Energy Resources, Renewable Resource Data Center t http://rredc.nrel.gov/solar/codesandalgorithms/solpos/ .

B-11


http://rredc.nrel.gov/solar/codesandalgorithms/solpos/�

subassembly. This gas handling subassembly is therefore sized for the average daily gas output rate over the highest production day (June 21) rather than the peak hourly output.

The PEC nanoparticles are modeled as 40 nm conductive substrate particles onto which ~5nm thick anodic and cathodic photo-active coatings are deposited. This results in a multi-layer PV/PEC unit with a multi-photon response to achieve the requisite electrolysis voltage either from single above-threshold photons or from multiple below-threshold-energy photons. Details of the PEC nanoparticle material system are not yet well defined through experimental data, so reasonable extrapolations have been made from the current level of knowledge. Thus, in consultation with the PEC Working Group, we have modeled the nanoparticles as 40nm diameter Fe

Type 1 System Reactor

2O3

particles coated with an additional photoactive layer. For the Type 1 system particles, both hydrogen and oxygen are evolved from the surface of the nanoparticle. For current experimental particles, lab tests measuring conversion of absorbed photons to electrons have demonstrated an Incident-Photon-to electron Conversion Efficiency (IPCE) peak value of 2.5% for 360nm (3.4 eV) photons, and values to 10% have been predicted.

An end view of three baggie/bed structures is shown in Figure 1-1. A single baggie/bed is 1060 ft long x 40 ft wide. The assumed baseline Solar-to-Hydrogen (STH) conversion efficiency3 is 10%. The system for 1 tonne per day (TPD) H2

yearly average production consists of 18 baggies. This Type 1 reactor is the simplest PEC embodiment and has the lowest capital cost.

Figure 1-1: Type 1 End View

DrivewayBaggies

Transparent Film

Type 2 is the second type of colloidal suspension reactor and employs separate beds for the OType 2 System Reactor

2 gas production reaction and the H2 gas production reaction. The O2 and H2

beds are linked together with diffusion bridges to allow the transport of ions but prevent gas and particle mixing. A 0.1M KOH electrolyte is common to both beds and facilitates transport of ionic species. These beds also contain an intermediary reactant denoted “A”, which participates in the reactions, but is not consumed. “A” can be iodine, bromine, iron or other elements. A typical set of equations describing the nanoparticle photoreactions is:

Bed I (O2 evolution bed): 4 photons + 4 A + 2 H2O => 1 O2 + 4H+ + 4 ABed II (H

- 2 evolution bed): 4 photons + 4H+ + 4 A- => 2 H2 + 4 A

The nanoparticles are similar in structure to nanoparticles in the Type 1 system; however, the anodic particles would differ somewhat from the cathodic particles. We have modeled the Type 2 nanoparticles as Fe2O3

3 Solar-to-hydrogen (STH) conversion efficiency is the energy ratio of the H2 produced (lower heating value) by a reactor divided by the total solar energy incident on the reactor.

substrate particles onto which an additional photoactive layer is deposited.

B-12


As with the Type 1 single bed system, a baggie system is utilized, but with the addition of a continuous feed-through bridge passage between bed pairs (for ion diffusion between beds) and a slurry mixing system. The slurry mixer, to facilitate mixing within the bags and diffusion across the bridges, consists of perforated pipes through which the slurry is continuously pumped and circulated from the bed center to the bed edges The multi-baggie/bed assembly, shown in Figure 1-2, consists of one half-baggie (H2), one full size baggie (O2), a second full size baggie (H2), and a second half-baggie (O2

). Dimensions of the baggie/bed assembly shown in the figure are 200 ft long x 20 ft wide. The width of the bed is reduced compared to Type 1 baggies to reduce the diffusion distances.

Figure 1-2: Type 2 Multi-Baggie Assembly End View FRONT VIE W

Driveway

Transparent Film

BridgeDriveway Perforated Pipes

Baggies

The Type 2 system requires approximately twice the solar absorption area as Type 1 because of the separation of the complete reaction into dual beds. Thus the STH efficiency is 5% and the system for 1 TPD H2

average production consists of 347 such assemblies. The costs for the Type 2 reactor are 4.2 times higher than the Type 1 because of the near- doubling of reactor area and amount of nanoparticles, the added porous membranes, the added slurry circulation system, the additional number of ports, and the added manufacturing complexity

The Type 1 and 2 systems are innovative and promising approaches to PEC hydrogen generation, but are relatively immature compared with the standard PEC planar cell approach. Consequently they have greater uncertainty in prediction of performance and costs, such as:

• Detailed definition of Nanoparticle PV materials and fabrication • Production costing of particle fabrication • Effective photo-reactive area (photon capture area) on a given nanoparticle • For the Type 2 reactor: uncertainty in diffusion times across diffusion bridge • For the Type 2 reactor: uncertainty in whether there is 100% exclusive generation of O2

in Bed I and H2

in Bed II.

The Type 3 and Type 4 Systems are extrapolated from current experimental PEC systems, using planar PV cells. Most PEC research to date has dealt with this type of cell. The photocell PEC system utilizes a PV cell generating sufficient voltage to electrolyze water, with modifications to allow it to survive in an electrolyte. The cell generates electrons from incident photons and has either integral electrodes immersed in the electrolyte as shown in

Type 3 and Type 4 Photocell Systems

Figure 1-3, or electrically connected spaced apart electrodes immersed in the electrolyte. For the PEC cell, the PV materials absorb photons to generate electrons for electrolysis at a total voltage on the order of 1.6-2.0 volts and conduct the electrons between the oxygen gas generating anode and the

B-13


hydrogen gas generating cathode. In experimental systems, the required voltage is higher due to losses in ion and electron transport and other losses. The electrolysis gases are separated by their physically separate reaction sites to create separate outlets for pure H2 and pure O2

. In a common embodiment, the cell front face illuminated by solar radiation is a conductive window that functions as the electrolysis anode. Multi-junction PV active layers are used to use multiple sub-threshold photons to reach desired overall voltage and increase solar spectrum utilization.

Figure 1-3: Schematic of a Generic PEC Photocell

There are multiple PEC cell configurations which can be used, some using membrane separation of the gases and others relying solely on buoyancy separation. For this costing study, we have based our cell design on the simplest generic design assuming an open electrolyte compartment and buoyant separation of gases. Costing of the PV/PEC active components relies heavily on the cost estimates, projections, and achievements in the solar cell industry. To estimate cell cost, we have assumed the PV cell advances of:

• Minimized thickness of individual PV layers • Use of low cost printing techniques for material deposition • Use of lower cost PV materials, when possible • Low cost conductive coatings to protect against corrosion

To predict PEC cell costs, a cell component cost analysis was carried out, consistent with an NREL solar cell costing study4

and consistent with recent solar cell predictions of $1/W for thin film solar cells.

Planar PEC arrays are similar to planar solar cell PV arrays, except that the cell electrodes are in direct contact with the PEC electrolyte and output is H

Type 3 PEC Panel

2 and O2

4 "Thin film PV manufacturing: materials costs and their optimization", Zweibel, K., Solar Energy Materials & Solar Cells, 2000, Elsevier.

gas rather than an external electric current. Each panel is made up of multiple cells, with the cell area being as large as can be readily manufactured. The arrays are fixed in place and inclined toward the sun at a tilt angle from horizontal equal to the local latitude.

B-14


Figure 1-4: Type 3 PEC Panel Layout

Each individual panel is 1 m wide and 2 m in length, and has a baseline STH efficiency of 10%. The system for 1 TPD H2 average production consists of 26,923 such panels.

The Type 4 system uses a solar concentrator reflector to focus solar direct radiation onto the PEC cell. A solar tracking system is used to maximize direct radiation capture. Solar concentrators, which can use reflectors or lenses to focus the solar energy, substantially reduce the cost impact of the PV component of the system, but add the costs of the concentrators and steering systems. Therefore, the PV components comprise a smaller fraction of the Type 4 total system cost, and more costly cell materials (i.e., GaAs/GaInAs) with higher efficiencies, are cost effective. A PEC concentrator system can potentially use a concentration ratio of 10-50 suns; however, we limited our system to 10:1, which has been demonstrated in lab tests.

Type 4 Solar Concentrator PEC

For the concentrator PEC system, the water reservoir and the H2 and O2 collected are pressurized by the inlet water pump at relatively low added cost. Pressurization to 300 psi obviates the need for a separate compressor, minimizes water vapor loss by the reactor, and reduces O2 gas bubble size, which minimizes potential bubble scattering of incident photons at the anode face.

The concentrator PEC design for this analysis uses an offset parabolic cylinder array to focus radiation on a linear PEC receiver, as shown in Figure 1-5. The offset parabolic array has advantages of reduced structural weight, no aperture blockage, and location of the active receiver components, water feed, and hydrogen collection piping in the reflector base assembly. The

electrolyte, H2

SS cathode

PV-2

PV-2

TCO

electrolyte, O2

Front View

O2 manifold H2 manifold

Anode

Face

Water

Input

B-15


PEC receiver component is a linear array of PEC cells at the parabolic reflector focal point. The array has 2-axis steering to track the solar direct radiation.

Figure 1-5: Type 4 Concentrator PEC Design

Each individual concentrator array is 6 m wide and 3 m in height with a baseline STH efficiency of 15%. The system for 1 TPD H2

average production consists of 1,885 such reactors. In estimating reflector/collector costs for the Type 4 system, we based our costs on an NREL study of parabolic trough solar thermal power systems.

The Baseline Receiver uses a concentration ratio of 10:1. Increasing concentration ratio to 20:1 with the same PEC cell reduces the Plexiglas window span, with a thinner/lower cost window, and also reduces the PV surface area/cost and cell encapsulation area/cost. It is estimated that a doubling of concentration ratio to 20:1 could reduce the basic reactor cost by 17%.

The gas processing subassembly collects, compresses, purifies and delivers the product hydrogen to the production facility limits. The outlet pressure of hydrogen at the plant gate is 300 psi (20.4 atm., 20.7 bar) to provide a system comparable to other DOE H2A-modeled production plants.

Gas Processing

In all of the PEC systems, oxygen and hydrogen are produced, which raises combustibility issues. In the Type 2, 3, and 4 systems, the H2 and O2

are inherently separated in the PEC reactor so combustibility problems don’t arise. However, in the Type 1 system, the product gas within the headspace of the reactor bed (and subsequently fed to the gas processing systems) is a stoichiometric mixture of oxygen and hydrogen with a small amount of water vapor. As these gases are a combustible mixture, special precautions must be taken to ensure safety. However, in numerous industrial processes, compression of flammable mixtures is routinely accomplished. Consequently, hydrogen/oxygen mixtures are deemed a design concern rather than a problem.

For compressing the gas mixture, a compressor with intercooling is used in Type 1, 2, and 3 systems. For Type 1 systems, the compressor compresses an H2/O2 mix to 305psi prior to input into the H2 separation unit to allow for a 5 psi pressure loss in the separation. For Type 2 and Type 3 systems, the compressor compresses nominally pure H2 to 300psi for delivery to the

B-16


plant gate. For Type 4 systems, no compressor is needed as the pure H2 gas is already at 300psi coming out of the reactor.

A pressure swing adsorption (PSA) system is selected as the best H2 separation system for this application. PSA operates by flowing a pressurized stream of gases (e.g., at 305 psi) across a multi-component adsorbent bed (commonly layers of activated carbon, zeolite, silica gel, etc.) to preferentially capture an undesired gaseous species on the surface of the adsorbent. In the process, there is loss of hydrogen that is contained in the absorption bed at the beginning of the vent cycle. As the bed is depressurized, this hydrogen is expelled and lost out the vent. A second H2 loss occurs during the purge cycle, as pure hydrogen that is used to actively vent the system of impurities. Because of these PSA losses, the Type 1 system must produce about 11% more hydrogen from its reactor than the other systems.

Plant control systems serve many functions including local and remote monitoring, alarming and controlling of plant equipment and functions. We have assumed a level of control sophistication consistent with full functionality and safe operation.

Plant Control System

Several baseline assumptions were made to obtain the estimated yearly capital costs and operating costs for each of the four systems. On assessing yearly costs for each system’s capital equipment investment, an appropriate return on investment (ROI) was used. In order to evaluate the ROI, a discounted cash flow (DCF) analysis was performed using the H2A Production Model, Version 2.0. The H2A Costing Model provides a structured format to enter parameters which impact cash inflows and outflows associated with the construction and operation of a Hydrogen Production Plant. There are numerous plant-specific parameters which must be entered. Additionally, there are H2A Default values for many of the parameters which can be modified to meet specific circumstances. Once all parameters have been entered, the H2A model computes the levelized cost of hydrogen in $/kgH

General Cost Assumptions

2. For this study, we have not taken a costcredit for the byproduct O2 generated by the reactors.

PEC reactor sizes and system costs are summarized and compared in Specific Capital Costs for baseline systems

Figure 1-6.

B-17


Figure 1-6: PEC System Capital Cost Summary

The control system makes up a substantial part of total capital cost. Much of this cost comes from the hydrogen/oxygen sensors which monitor the gas stream for any leaks or contamination.

The total cost of produced hydrogen assumed a 10 tonne per day (TPD) plant consisting of ten of the baseline 1TPD modules described above. The hydrogen production cost results calculated from the baseline system designs and the H2A model are:

Overall Hydrogen Production Costs of Baseline Systems

• Type 1: $ 1.63/kg H• Type 2: $ 3.19/kg H

2

• Type 3: $10.36/kg H2

• Type 4: $ 4.05/kg H2

2

Figure 1-7 shows a breakdown of these costs into: capital costs, decommissioning costs, fixed O&M, and variable costs. Note that these are the H2 production costs for producing 300 psi hydrogen at the plant gate, and do not include delivery or dispensing costs.

Type 1 Type 2 Type 3 Type 4Single Bed Colloidal

Suspension

Dual Bed Colloidal

SuspensionFixed Flat

PanelTracking

ConcentratorGross Production (kgH2/day) 1,111 1,000 1,000 1,000Net Production (kgH2/day) 1,000 1,000 1,000 1,000Mean Solar Input (kWh/m2/day) 5.25 5.25 6.19 6.55Baseline Efficiency (STH) 10% 5% 10% 15%

1060’x40’x0.3’ 200’x20’x1.2’ 2m x 1m 6m x 3mbed bed panel reflector

Number of Reactors 18 347 26,923 1,885Photon Capture Area (m2) 70,540 126,969 53,845 33,924Reactor total cost $212,257 $892,934 $8,343,345 $3,135,209 Gas Processing Cost $684,283 $356,654 $917,338 $33,771 Controls Cost $173,944 $440,826 $319,862 $279,774 Hardware total cost $1,070,484 $1,690,414 $9,580,545 $3,448,754 Land Cost $11,330 $20,393 $27,076 $27,537 Total capital cost $1,081,814 $1,710,807 $9,607,621 $3,476,291

Dimensions of reactor

B-18


Figure 1-7: Comparison of PEC System Levelized H2 Cost

For the Type 1 and Type 2 systems, the levelized cost is quite low, but there is a large amount of development work and uncertainty in producing an operating system having these baseline performance parameters. The Type 3 system is the most mature of the concepts, with multiple small scale examples fabricated, but the substantial capital costs dominate H2 production cost. The Type 4 system has been implemented at lab scale with good efficiency. For the Type 4 production system, costs are moderately low and are dominated by the solar collector structure.

An HHydrogen Cost Sensitivity

2 production cost sensitivity analysis assessed the variation in H2

Figure 1-8

cost as a function of STH efficiency, PEC cell component cost, and PEC cell lifetime. The range of evaluation parameters for the sensitivity analyses and the results for the four systems are shown in .

Figure 1-8: Hydrogen Cost Sensitivity Analysis Results

Efficiency Particle Cost Particle Lifetime Efficiency Particle Cost Particle Lifetime

5% 0.1x 1 Year 2.5% 0.1x 1 Year

10% 1x 5 Year 5% 1x 5 Year15% 20x 10 Year 7.5% 20x 10 Year

Efficiency PEC Cell Cost PEC Cell Lifetime Efficiency PEC Cell Cost PEC Cell Lifetime

5% $80/m2 5 year 10% $200/m2 5 year

10% $153/m2 10 year 15% $316/m2 10 year

20% $200/m220 year 25% $450/m2

20 year

Type 1 Sensitivity Analysis Parameters Type 2 Sensitivity Analysis Parameters

Type 3 Sensitivity Analysis Parameters Type 4 Sensitivity Analysis Parameters

$0.00

$2.00

$4.00

$6.00

$8.00

$10.00

$12.00

Type 1 Type 2 Type 3 Type 4

Cost

($/k

g H

2)

System

Other Variable Costs (including utilities)

Fixed O&M

Decommissioning Costs

Capital Costs (Direct Capital + Indirect Capital + Land )

B-19


This study has shown that, within the cost assumptions used, production of HDiscussion of Results

2 by PEC systems can be economically viable in several configurations, upon successful resolution of the research challenges. Each system is discussed below.

Type 1 and 2 particle bed systems are innovative and cost effective PEC approaches, but they are immature and unproven compared with the standard PEC cell approach. Given the study assumptions as to efficiency and nanoparticle effectiveness, the Type 1 and 2 systems yield the lowest cost hydrogen. A unique advantage of these systems vs. the Type 3 panel array is that the gas collection bags are capable of storing the product gas output over a day’s production to average out the demands on the gas processing system rather than requiring the processors to handle the peak gas output (as is the case for the Type 3 system).

Key Unique Type 1 characteristics include: 1. Lowest predicted H22. Product gas in this system is a stoichiometric mixture of H

costs, given study efficiency assumptions 2 and O2

3. December output is 31% of June output, so the system would need to be enlarged ifDecember output were a driving requirement rather than just the yearly average.

raising safety concern

Key Unique Type 2 characteristics include: 1. Low predicted H22. Performance results hinge on minimal losses due to ion transport

costs, given study efficiency assumptions

3. Nanoparticles separately tailored for O2 production and H24. December output is 31% of June output, so the system would need to be enlarged if

December output were a driving requirement rather than just the yearly average.

production

$1.61

$1.61

$1.49

$1.71

$1.96

$2.29

$- $5.00 $10.00 $15.00 $20.00

LifetimeYrs: 1/5/10

Particle CostMultiplier: 0.1/1/20

Efficiency%:5,10,15

Production Costs ($/kgH2)

Type 1

$8.64

$6.90

$6.14

$14.19

$12.59

$18.75

$- $5.00 $10.00 $15.00 $20.00

LifetimeYrs: 5/10/20

PEC Cell Cost$/m2: 80/153/200

Efficiency%: 5/10/20


Type 3

$3.85

$3.70

$2.85

$4.49

$4.45

$5.55

$- $5.00 $10.00 $15.00 $20.00

LifetimeYrs: 5/10/20

PEC Cell Cost$/m2: 200/316/450

Efficiency%:10/15/25


Type 4

$3.17

$3.13

$2.53

$3.33

$4.49

$5.23

$- $5.00 $10.00 $15.00 $20.00

LifetimeYrs: 1/5/10

Particle CostMultiplier: 0.1/1/20

Efficiency%: 2.5/5/7.5


Type 2

B-20


The greater uncertainties in the Type 1 and 2 systems include:

• Incomplete definition and demonstration of the optimal nanoparticle PEC materials, including effective photon voltages, resistance losses, corrosion effects, lifetime

• Incomplete definition of fabrication techniques and production costing of the particles • Fraction of effective photo-reactive area (photon capture area) on a given base

nanoparticle • Annual production quantity of the photoactive nanoparticles. (This study assumed an

annual production quantity sufficient for supplying a 500 tonne H2 capacity each year, yielding a particle cost of $304/kg. Nanoparticle cost would increase significantly if annual production corresponded to that required to produce only 10 tonnes H2

each year.)

Type 3 and 4 Photocell Systems

have benefited extensively from the current high development activity in the solar cell area, particularly in efforts to drive down the costs of thin film solar cells. Solar cells can be used to generate solar electricity to separately electrolyze water. However, with sufficient development, the PEC cell concept has the potential to be more efficient than separate solar cell/electrolyzer systems, since it eliminates the materials and fabrication costs of the solar cell current carrier conductor grid. The PEC cell can also be used under pressure to eliminate the need for a separate compression stage. Relative to the particle bed systems, variation in output between summer and winter for the photocell systems is significantly less.

Key unique Type 3 characteristics include: 1. Highest H22. Benefits from and relies on development of low cost thin film PV materials

production costs, due to large areas of PEC cell component

3. For PEC cells, the cell packaging costs are significantly higher than the PV material costs 4. Highest gas compression cost, because compressor is sized for peak hourly production 5. Tilt angle, nominally the latitude angle, can be optimized to achieve the most level H2

gas output over the year of all the options over the full year including environmental variations.

Key unique Type 4 characteristics include:

1. Moderately low H22. In-cell compression of gas eliminates need for separate compressor

costs, near the Type 2 estimate

3. Increased efficiency possible with PEC development, and high temperature operation 4. Decreased H25. Offset reflector array for the concentrator reduces structural and piping costs.

cost with higher concentration ratio – to potentially below $3/kg

6. December output is 53% of June output, so the system would need to be enlarged if December output were the driving requirement rather than just the yearly average.

The body of the main report is divided into several sections for ease of use. Initially, the basic science aspects are discussed. Next, the engineering designs for hydrogen production,

Organization of Report

B-21


purification, and compression are laid out. Finally, the resulting system capital costs and corresponding hydrogen production costs are determined.

B-22


APPENDIX C

PHOTOELECTROCHEMICAL WATER SPLITTING: STANDARDS.

EXPERIMENTAL METHODS AND PROTOCOLS

http://download.springer.com/static/pdf/533/bfm%253A978-1-4614-8298-7%252F1.pdf?auth66=1383411637_68d655b0ab1d4378420e4ab389ba58fa&ext=.pdf

Springer: Photoelectrochemical Water Splitting: Front Material C-1

SpringerBriefs in Energy

For further volumes:http://www.springer.com/series/8903

http://download.springer.com/static/pdf/533/bfm%253A978-1-4614-8298-7%252F1.pdf? auth66=1383411637_68d655b0ab1d4378420e4ab389ba58fa&ext=.pdf


http://www.springer.com/series/8903

Zhebo Chen • Huyen N. DinhEric Miller

Photoelectrochemical WaterSplitting

Standards, Experimental Methods,and Protocols

123



Zhebo ChenDepartment of Chemical EngineeringStanford UniversityStanford, CAUSA

Huyen N. DinhNational Renewable Energy LaboratoryHydrogen Technologies

and Systems CenterGolden, COUSA

Eric MillerFuel Cell TechnologiesU.S. Department of EnergyWashington, DCUSA

ISSN 2191-5520 ISSN 2191-5539 (electronic)ISBN 978-1-4614-8297-0 ISBN 978-1-4614-8298-7 (eBook)DOI 10.1007/978-1-4614-8298-7Springer New York Heidelberg Dordrecht London

Library of Congress Control Number: 2013943566

� The Author(s) 2013This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part ofthe material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission orinformation storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodology now known or hereafter developed. Exempted from this legal reservation are briefexcerpts in connection with reviews or scholarly analysis or material supplied specifically for thepurpose of being entered and executed on a computer system, for exclusive use by the purchaser of thework. Duplication of this publication or parts thereof is permitted only under the provisions ofthe Copyright Law of the Publisher’s location, in its current version, and permission for use mustalways be obtained from Springer. Permissions for use may be obtained through RightsLink at theCopyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law.The use of general descriptive names, registered names, trademarks, service marks, etc. in thispublication does not imply, even in the absence of a specific statement, that such names are exemptfrom the relevant protective laws and regulations and therefore free for general use.While the advice and information in this book are believed to be true and accurate at the date ofpublication, neither the authors nor the editors nor the publisher can accept any legal responsibility forany errors or omissions that may be made. The publisher makes no warranty, express or implied, withrespect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)



Preface

The methods and definitions presented herein are the product of an effort supportedby the U.S. Department of Energy (DOE) to form a consensus among a number ofexperienced researchers in the area of photoelectrochemical (PEC) hydrogenproduction from various DOE-supported laboratories (including national labs andacademic institutions) and other international partners. An early result of this effortwas the production of a consolidated version of the guide presented here, pub-lished in the form of a review paper in the Journal of Materials Research in 2010,[1] and is reprinted in part with permission in this book. The extended guidance inthe present work aims to accelerate materials development by establishing stan-dards for methods, definitions, and reporting protocols that will enable directcross-comparison of materials’ properties and performance metrics. The intent isto facilitate knowledge transfer on a global scale.

The authors who have contributed to the writing of this book are members ofthe PEC Standards Working Group, assembled by the Energy Efficiency andRenewable Energy’s Fuel Cell Technologies Office at the U.S. DOE. Theseauthors include:

• Zhebo Chen (Department of Chemical Engineering, Stanford University,Stanford, CA 94305-5025)

• Todd G. Deutsch (Hydrogen Technologies and Systems Center, NationalRenewable Energy Laboratory, Golden, Colorado 80401)

• Huyen N. Dinh (Hydrogen Technologies and Systems Center, NationalRenewable Energy Laboratory, Golden, Colorado 80401)

• Kazunari Domen (Department of Chemical System Engineering, University ofTokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan)

• Keith Emery (National Center for Photovoltaics, National Renewable EnergyLaboratory, Golden, Colorado 80401)

• Arnold J. Forman (Department of Chemistry and Biochemistry, University ofCalifornia—Santa Barbara, Santa Barbara, CA 93106-5080)

• Nicolas Gaillard (Hawaii Natural Energy Institute, University of Hawaii atManoa, Honolulu, HI 96822)

• Roxanne Garland (Fuel Cell Technologies Office, U.S. Department of Energy,Washington, D.C. 20585)

v



• Clemens Heske (Department of Chemistry, University of Nevada—Las Vegas,Las Vegas, NV 89154-4003)

• Thomas F. Jaramillo (Department of Chemical Engineering, Stanford Univer-sity, Stanford CA 94305-5025)

• Alan Kleiman-Shwarsctein (Department of Chemical Engineering, University ofCalifornia—Santa Barbara, Santa Barbara, CA 93106-5080)

• Eric Miller (Fuel Cell Technologies Office, U.S. Department of Energy,Washington, D.C. 20585)

• Kazuhiro Takanabe (Division of Physical Sciences and Engineering, KingAbdullah University of Science and Technology (KAUST), Thuwall, Kingdomof Saudi Arabia)

• John Turner (Hydrogen Technologies and Systems Center, National RenewableEnergy Laboratory, Golden, Colorado 80401)

The chapters written herein represent many years of collaborative discussionand review, and we hope that their content can enable new researchers in the fieldof PEC water splitting to rapidly gain traction in their own laboratories towards thedevelopment of high efficiency materials.

A number of international researchers participated in providing excellentfeedback on the text written in this book, including many affiliated with theInternational Energy Agency’s Hydrogen Implementing Agreement Task 26. In noparticular order, we thank Grant Mathieson, Bruce Parkinson, Jennifer Leisch,Theanne Schiros, David Peterson, Ib Chorkendorff, Peter Vesborg, Kendra Kuhl,Blaise Pinaud, Julie Tuttle, Sarah Havig, Berc Kalanyan, Billie Abrams, CandaceChan, Nelson Kelly, Shiwei Lin, Nianqiang Wu, Shane Ardo, Nick Strandwitz,Lorna Jeffery Mingu, Daniel Schaadt, Marie Mayer, Ke Sun, Nikolaos Vlacho-poulos, Qiang Huang, Juan Hodelin, Jian Jin, Anna Goldstein, Kevin Sivula,Kazuhiro Sayama, Isabell Thomann, Yue Tak Lai, Ilwhan Oh, and Sonia JulianaCalero.

Reference

1. Z. Chen, T.F. Jaramillo, T.G. Deutsch, A. Kleiman-Shwarsctein, A.J. Forman, N. Gaillard,R. Garland, K. Takanabe, C. Heske, M. Sunkara, E.W. McFarland, K. Domen, E.L. Miller,J.A. Turner, H.N. Dinh, J. Mater. Res. 25(1), 3–16 (2010)

vi Preface



Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Efficiency Definitions in the Field of PEC . . . . . . . . . . . . . . . . . . . 72.1 Overview of Efficiency Definitions . . . . . . . . . . . . . . . . . . . . . 72.2 Efficiency Definition for Benchmarking . . . . . . . . . . . . . . . . . . 82.3 Diagnostic Efficiencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2.3.1 Applied Bias Photon-to-Current Efficiency. . . . . . . . . . . 102.3.2 Incident Photon-to-Current Efficiency (IPCE)/External

Quantum Efficiency (EQE) . . . . . . . . . . . . . . . . . . . . . 112.3.3 Absorbed Photon-to-Current Efficiency

(APCE)/Internal Quantum Efficiency (IQE) . . . . . . . . . . 132.4 Half-cell Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142.5 Summary of Efficiency Definitions . . . . . . . . . . . . . . . . . . . . . 15References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3 Experimental Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173.1 Electrode Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.1.1 Electrode Preparation Considerations. . . . . . . . . . . . . . . 173.1.2 Photoactive Semiconductor Material . . . . . . . . . . . . . . . 173.1.3 Substrate Considerations . . . . . . . . . . . . . . . . . . . . . . . 193.1.4 PEC Electrode Connections . . . . . . . . . . . . . . . . . . . . . 213.1.5 Electrode Surface Area Determination . . . . . . . . . . . . . . 24

3.2 Cell Setup and Connections for Three-and Two-Electrode Configurations. . . . . . . . . . . . . . . . . . . . . . 263.2.1 Basic Photoelectrochemical Test Setup . . . . . . . . . . . . . 263.2.2 Selecting the Counter Electrode . . . . . . . . . . . . . . . . . . 283.2.3 Selecting the Reference Electrode . . . . . . . . . . . . . . . . . 283.2.4 Choosing the Electrolyte . . . . . . . . . . . . . . . . . . . . . . . 293.2.5 Connecting the Electrodes to the Potentiostat . . . . . . . . . 313.2.6 Device Testing Approaches . . . . . . . . . . . . . . . . . . . . . 31

3.3 Catalyst Surface Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . 323.3.1 Principle of Surface Catalysis . . . . . . . . . . . . . . . . . . . . 33

vii



http://dx.doi.org/10.1007/978-1-4614-8298-7_1

http://dx.doi.org/10.1007/978-1-4614-8298-7_1

http://dx.doi.org/10.1007/978-1-4614-8298-7_1#Bib1

http://dx.doi.org/10.1007/978-1-4614-8298-7_2

http://dx.doi.org/10.1007/978-1-4614-8298-7_2

http://dx.doi.org/10.1007/978-1-4614-8298-7_2#Sec1



















http://dx.doi.org/10.1007/978-1-4614-8298-7_3

http://dx.doi.org/10.1007/978-1-4614-8298-7_3
































3.3.2 Selecting the Catalyst . . . . . . . . . . . . . . . . . . . . . . . . . 333.3.3 Morphological Considerations . . . . . . . . . . . . . . . . . . . 343.3.4 Deposition Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 343.3.5 Electrical Characterization . . . . . . . . . . . . . . . . . . . . . . 37

3.4 Spectral Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373.4.1 The AM 1.5 G Reference Spectrum . . . . . . . . . . . . . . . 373.4.2 Reference Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4 PEC Characterization Flowchart . . . . . . . . . . . . . . . . . . . . . . . . . 45Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5 UV-Vis Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495.1 Knowledge Gained from UV-Vis Spectroscopy . . . . . . . . . . . . . 495.2 Limitations of UV-Vis Spectroscopy . . . . . . . . . . . . . . . . . . . . 505.3 Method for Performing UV-Vis Spectroscopic Measurements . . . 52

5.3.1 Experimental Parameters . . . . . . . . . . . . . . . . . . . . . . . 535.3.2 Transmission UV-Vis . . . . . . . . . . . . . . . . . . . . . . . . . 535.3.3 Diffuse Reflectance UV-Vis . . . . . . . . . . . . . . . . . . . . . 545.3.4 Absorption UV-Vis . . . . . . . . . . . . . . . . . . . . . . . . . . . 555.3.5 Required Equipment for UV-Vis Measurements . . . . . . . 56

5.4 Analysis of Band Gap Energies from UV-Vis Spectra . . . . . . . . 57References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

6 Flat-Band Potential Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . 636.1 Illuminated Open-Circuit Potential (OCP). . . . . . . . . . . . . . . . . 63

6.1.1 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . 636.1.2 Limits of Experiments . . . . . . . . . . . . . . . . . . . . . . . . . 646.1.3 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656.1.4 Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676.1.5 Open-Circuit Potential and pH . . . . . . . . . . . . . . . . . . . 67

6.2 Mott–Schottky. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686.2.1 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . 686.2.2 Limits of Experiment . . . . . . . . . . . . . . . . . . . . . . . . . 696.2.3 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706.2.4 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

6.3 Three-Electrode j–V and Photocurrent Onset. . . . . . . . . . . . . . . 736.3.1 Potential Range of Photocurrent Generation . . . . . . . . . . 736.3.2 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . 756.3.3 Limits of Experiment . . . . . . . . . . . . . . . . . . . . . . . . . 776.3.4 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 796.3.5 Time Required for Preparation, Experiment,

and Data Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . 81

viii Contents


















http://dx.doi.org/10.1007/978-1-4614-8298-7_4

http://dx.doi.org/10.1007/978-1-4614-8298-7_4


http://dx.doi.org/10.1007/978-1-4614-8298-7_5

http://dx.doi.org/10.1007/978-1-4614-8298-7_5




















http://dx.doi.org/10.1007/978-1-4614-8298-7_6

http://dx.doi.org/10.1007/978-1-4614-8298-7_6




































6.3.6 Data Analysis and Expected Results . . . . . . . . . . . . . . . 826.3.7 Flat-Band Potential From Photocurrent Onset . . . . . . . . . 82

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

7 Incident Photon-to-Current Efficiency and PhotocurrentSpectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877.1 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877.2 Limitations of the Technique . . . . . . . . . . . . . . . . . . . . . . . . . 88

7.2.1 Confidence in Results . . . . . . . . . . . . . . . . . . . . . . . . . 887.2.2 Corrosion Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . 897.2.3 Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

7.3 Pitfalls of the Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907.4 Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

7.4.1 Preparation Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917.4.2 Calibration of Lamp and Sample Measurement. . . . . . . . 917.4.3 Applied Bias IPCE Experiment. . . . . . . . . . . . . . . . . . . 927.4.4 White Light Bias IPCE Experiment. . . . . . . . . . . . . . . . 92

7.5 Measurement and Analysis Time. . . . . . . . . . . . . . . . . . . . . . . 937.5.1 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

7.6 Photocurrent Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . 95References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

8 2-Electrode Short Circuit and j–V . . . . . . . . . . . . . . . . . . . . . . . . 998.1 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 998.2 Limitations of Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

8.2.1 Credibility of Results. . . . . . . . . . . . . . . . . . . . . . . . . . 1008.2.2 Corrosion Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . 1008.2.3 Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

8.3 Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1008.3.1 Experimental Setup and Procedure . . . . . . . . . . . . . . . . 1008.3.2 2-Electrode j–V Measurement . . . . . . . . . . . . . . . . . . . 1018.3.3 Preparation and Measurement Time . . . . . . . . . . . . . . . 1018.3.4 Data Analysis and Expected Results . . . . . . . . . . . . . . . 102

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

9 Hydrogen and Oxygen Detection from Photoelectrodes . . . . . . . . . 1059.1 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059.2 Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059.3 Experimental Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

9.3.1 Batch Reactor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089.3.2 Flow Reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1099.3.3 Recirculating Reactor (in Vacuum) . . . . . . . . . . . . . . . . 109

Contents ix








http://dx.doi.org/10.1007/978-1-4614-8298-7_7

http://dx.doi.org/10.1007/978-1-4614-8298-7_7

http://dx.doi.org/10.1007/978-1-4614-8298-7_7






























http://dx.doi.org/10.1007/978-1-4614-8298-7_8

http://dx.doi.org/10.1007/978-1-4614-8298-7_8






















http://dx.doi.org/10.1007/978-1-4614-8298-7_9

http://dx.doi.org/10.1007/978-1-4614-8298-7_9













9.4 Limitations of Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1109.4.1 Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

9.5 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1119.6 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

10 Stability Testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11510.1 Knowledge Gained . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11510.2 Limits of Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11510.3 Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11610.4 Data Analysis and Expected Result . . . . . . . . . . . . . . . . . . . . . 116Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

x Contents












http://dx.doi.org/10.1007/978-1-4614-8298-7_10

http://dx.doi.org/10.1007/978-1-4614-8298-7_10










Wiley Publication found at: http://www.wiley.com/WileyCDA/WileyTitle/productCd-0470823976.html

APPENDIX D

On Solar Hydrogen and Nanotechnology

Lionel Vayssieres

ISBN: 978-0-470-82397-2

704 pages

March 2010

Description

More energy from the sun strikes Earth in an hour than is consumed by humans in an entire year. Efficiently

harnessing solar power for sustainable generation of hydrogen requires low-cost, purpose-built, functional materials

combined with inexpensive large-scale manufacturing methods. These issues are comprehensively addressed in On

Solar Hydrogen & Nanotechnology – an authoritative, interdisciplinary source of fundamental and applied knowledge

in all areas related to solar hydrogen. Written by leading experts, the book emphasizes state-of-the-art materials and

characterization techniques as well as the impact of nanotechnology on this cutting edge field.

Addresses the current status and prospects of solar hydrogen, including major achievements, performancebenchmarks, technological limitations, and crucial remaining challenges

Covers the latest advances in fundamental understanding and development in photocatalytic reactions,semiconductor nanostructures and heterostructures, quantum confinement effects, device fabrication,modeling, simulation, and characterization techniques as they pertain to solar generation of hydrogen

Assesses and establishes the present and future role of solar hydrogen in the hydrogen economy Contains numerous graphics to illustrate concepts, techniques, and research results

On Solar Hydrogen & Nanotechnology is an essential reference for materials scientists, physical and inorganic chemists, electrochemists, physicists, and engineers carrying out research on solar energy,

photocatalysis, or semiconducting nanomaterials, both in academia and industry. It is also an invaluable resource for graduate students and postdoctoral researchers as well as business professionals and consultants

with an interest in renewable energy.

D-1

http://www.wiley.com/WileyCDA/Section/id-302475.html?query=Lionel+Vayssieres

http://www.wiley.com/WileyCDA/WileyTitle/productCd-0470823976.html

Contents

List of Contributors xvii

Preface xix

Editor Biography xxiii

PART ONE—FUNDAMENTALS, MODELING, AND EXPERIMENTALINVESTIGATION OF PHOTOCATALYTIC REACTIONS FOR DIRECT

SOLAR HYDROGEN GENERATION

1 Solar Hydrogen Production by Photoelectrochemical

Water Splitting: The Promise and Challenge 3

Eric L. Miller

1.1 Introduction 3

1.2 Hydrogen or Hype? 4

1.3 Solar Pathways to Hydrogen 5

1.3.1 The Solar Resource 5

1.3.2 Converting Sunlight 6

1.3.3 Solar-Thermal Conversion 7

1.3.4 Solar-Potential Conversion 8

1.3.5 Pathways to Hydrogen 9

1.4 Photoelectrochemical Water-Splitting 10

1.4.1 Photoelectrochemistry 10

1.4.2 PEC Water-Splitting Reactions 10

1.4.3 Solar-to-Hydrogen Conversion Efficiency 13

1.4.4 Fundamental Process Steps 14

1.5 The Semiconductor/Electrolyte Interface 14

1.5.1 Rectifying Junctions 14

1.5.2 A Solid-State Analogy: The npþ Junction 15

1.5.3 PEC Junction Formation 17

1.5.4 Illuminated Characteristics 19

1.5.5 Fundamental Process Steps 20

COPYRIG

HTED M

ATERIAL

http://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-2

1.6 Photoelectrode Implementations 23

1.6.1 Single-Junction Performance Limits 23

1.6.2 Multijunction Performance Limits 24

1.6.3 A Shining Example 27

1.7 The PEC Challenge 28

1.7.1 What’s Needed, Really? 28

1.7.2 Tradeoffs and Compromises 29

1.7.3 The Race with PV-Electrolysis 29

1.8 Facing the Challenge: Current PEC Materials Research 29

Acknowledgments 32

References 32

2 Modeling and Simulation of Photocatalytic Reactions at TiO2 Surfaces 37Hideyuki Kamisaka and Koichi Yamashita

2.1 Importance of Theoretical Studies on TiO2 Systems 37

2.2 Doped TiO2 Systems: Carbon and Niobium Doping 39

2.2.1 First-Principle Calculations on TiO2 39

2.2.2 C-Doped TiO2 41

2.2.3 Nb-Doped TiO2 45

2.3 Surface Hydroxyl Groups and the Photoinduced Hydrophilicity of TiO2 51

2.3.1 Speculated Active Species on TiO2 – Superoxide Anion (O2�)

and the Hydroxyl Radical (OH.) 51

2.3.2 Theoretical Calculations of TiO2 Surfaces and Adsorbents 51

2.3.3 Surface Hydroxyl Groups and Photoinduced Hydrophilic

Conversion 53

2.4 Dye-Sensitized Solar Cells 58

2.4.1 Conventional Sensitizers: Ruthenium Compounds and Organic Dyes 58

2.4.2 Multiexciton Generation in Quantum Dots: A Novel Sensitizer

for a DSSC 59

2.4.3 Theoretical Estimation of the Decoherence Time between

the Electronic States in PbSe QDs 60

2.5 Future Directions: Ab Initio Simulations and the Local

Excited States on TiO2 63

2.5.1 Improvement of the DFT Functional 64

2.5.2 Molecular Mechanics and Ab Initio Molecular Dynamics 65

2.5.3 Description of Local Excited States 66

2.5.4 Nonadiabatic Behavior of a System and Interfacial

Electron Transfer 67

Acknowledgments 68

References 68

3 Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces 77G.I.N. Waterhouse and H. Idriss

3.1 TiO2 Single-Crystal Surfaces 78

3.2 Photoreactions Over Semiconductor Surfaces 80

vi Contentshttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-3

3.3 Ethanol Reactions Over TiO2(110) Surface 81

3.4 Photocatalysis and Structure Sensitivity 83

3.5 Hydrogen Production from Ethanol Over Au/TiO2 Catalysts 84

3.6 Conclusions 87

References 87

4 Fundamental Reactions on Rutile TiO2(110) Model

Photocatalysts Studied by High-Resolution Scanning

Tunneling Microscopy 91

Stefan Wendt, Ronnie T. Vang, and Flemming Besenbacher

4.1 Introduction 91

4.2 Geometric Structure and Defects of the Rutile

TiO2 (110) Surface 93

4.3 Reactions of Water with Oxygen Vacancies 96

4.4 Splitting of Paired H Adatoms and Other Reactions Observed on Partly

Water Covered TiO2(110) 98

4.5 O2 Dissociation and the Role of Ti Interstitials 101

4.6 Intermediate Steps of the Reaction Between O2 and H Adatoms

and the Role of Coadsorbed Water 106

4.7 Bonding of Gold Nanoparticles on TiO2(110)

in Different Oxidation States 112

4.8 Summary and Outlook 115

References 117

PART TWO—ELECTRONIC STRUCTURE, ENERGETICS,

AND TRANSPORT DYNAMICS OF PHOTOCATALYST

NANOSTRUCTURES

5 Electronic Structure Study of Nanostructured Transition

Metal Oxides Using Soft X-Ray Spectroscopy 125

Jinghua Guo, Per-Anders Glans, Yi-Sheng Liu,

and Chinglin Chang

5.1 Introduction 125

5.2 Soft X-Ray Spectroscopy 126

5.2.1 Soft X-Ray Absorption and Emission Spectroscopy 126

5.2.2 Resonantly Excited Soft X-Ray Emission

Spectroscopy 127

5.3 Experiment Set-Up 127

5.3.1 Beamline 128

5.3.2 Spectrometer and Endstation 129

5.3.3 Sample Arrangements 131

5.4 Results and Discussion 132

Acknowledgments 139

References 139

Contents viihttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-4

6 X-ray and Electron Spectroscopy Studies of Oxide Semiconductors

for Photoelectrochemical Hydrogen Production 143

Clemens Heske, Lothar Weinhardt, and Marcus B€ar


6.2 Soft X-Ray and Electron Spectroscopies 145

6.3 Electronic Surface-Level Positions of WO3 Thin Films 147

6.3.1 Introduction 147

6.3.2 Sample Handling and the Influence of X-Rays, UV-Light

and Low-Energy Electrons on the Properties of the WO3 Surface 147

6.3.3 Surface Band Edge Positions in Vacuum – Determination

with UPS/IPES 149

6.3.4 Estimated Surface Band-Edge Positions in Electrolyte 151

6.3.5 Conclusions 153

6.4 Soft X-Ray Spectroscopy of ZnO:Zn3N2 Thin Films 154


6.4.2 The O K XES Spectrum of ZnO:N Thin Films – Determination

of the Valence Band Maximum 154

6.4.3 The Impact of Air Exposure on the Chemical Structure

of ZnO:N Thin Films 155

6.4.4 Conclusions 157

6.5 In Situ Soft X-Ray Spectroscopy: A Brief Outlook 158

6.6 Summary 158

Acknowledgments 159

References 159

7 Applications of X-Ray Transient Absorption Spectroscopy

in Photocatalysis for Hydrogen Generation 163Lin X. Chen


7.2 X-Ray Transient Absorption Spectroscopy (XTA) 165

7.3 Tracking Electronic and Nuclear Configurations in Photoexcited

Metalloporphyrins 171

7.4 Tracking Metal-Center Oxidation States in the MLCT State

of Metal Complexes 176

7.5 Tracking Transient Metal Oxidation States During Hydrogen Generation 178

7.6 Prospects and Challenges in Future Studies 180

Acknowledgments 181

References 181

8 Fourier-Transform Infrared and Raman Spectroscopy of Pure

and Doped TiO2 Photocatalysts 189

Lars Osterlund


8.2 Vibrational Spectroscopy on TiO2 Photocatalysts: Experimental

Considerations 191

viii Contentshttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-5

8.3 Raman Spectroscopy of Pure and Doped TiO2 Nanoparticles 195

8.4 Gas–Solid Photocatalytic Reactions Probed by FTIR Spectroscopy 199

8.5 Model Gas–Solid Reactions on Pure and Doped TiO2

Nanoparticles Studied by FTIR Spectroscopy 205

8.5.1 Reactions with Formic Acid 205

8.5.2 Reactions with Acetone 221

8.6 Summary and Concluding Remarks 229

Acknowledgments 230

References 230

9 Interfacial Electron Transfer Reactions in CdS QuantumDot Sensitized TiO2 Nanocrystalline Electrodes 239

Yasuhiro Tachibana


9.2 Nanomaterials 240

9.2.1 Semiconductor Quantum Dots 240

9.2.2 Metal Oxide Nanocrystalline Semiconductor Films 241

9.2.3 QD Sensitized Metal Oxide Semiconductor Films 242

9.3 Transient Absorption Spectroscopy 245

9.3.1 Principle 245

9.3.2 Calculation of Absorption Difference 245

9.3.3 System Arrangement 246

9.4 Controlling Interfacial Electron Transfer Reactions

by Nanomaterial Design 247

9.4.1 QD/Metal-Oxide Interface 248

9.4.2 QD/Electrolyte Interface 250

9.4.3 Conducting Glass/Electrolyte Interface 252

9.5 Application of QD-Sensitized Metal-Oxide Semiconductors to Solar

Hydrogen Production 258

9.6 Conclusion 260

Acknowledgments 260

References 260

PART THREE—DEVELOPMENT OF ADVANCED NANOSTRUCTURES

FOR EFFICIENT SOLAR HYDROGEN PRODUCTION FROM CLASSICALLARGE BANDGAP SEMICONDUCTORS

10 Ordered Titanium Dioxide Nanotubular Arrays as Photoanodes

for Hydrogen Generation 267

M. Misra and K.S. Raja


10.2 Crystal Structure of TiO2 268

10.2.1 Electronic and Defect Structure of TiO2 269

10.2.2 Preparation of TiO2 Nanotubes 272

10.2.3 Energetics of Photodecomposition of Water on TiO2 279

References 288

Contents ixhttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-6

11 Electrodeposition of Nanostructured ZnO Films and Their

Photoelectrochemical Properties 291

Torsten Oekermann


11.2 Fundamentals of Electrochemical Deposition 292

11.3 Electrodeposition of Metal Oxides and Other Compounds 294

11.4 Electrodeposition of Zinc Oxide 295

11.4.1 Electrodeposition of Pure ZnO 295

11.4.2 Electrodeposition of Doped ZnO 297

11.4.3 P-n-Junctions Based on Electrodeposited ZnO 298

11.5 Electrodeposition of One- and Two-Dimensional ZnO Nanostructures 298

11.5.1 ZnO Nanorods 298

11.5.2 ZnO Nanotubes 301

11.5.3 Two-Dimensional ZnO Nanostructures 302

11.6 Use of Additives in ZnO Electrodeposition 303

11.6.1 Dye Molecules as Structure-Directing Additives 303

11.6.2 ZnO Electrodeposition with Surfactants 307

11.6.3 Other Additives 311

11.7 Photoelectrochemical and Photovoltaic Properties 312

11.7.1 Dye-Sensitized Solar Cells (DSSCs) 312

11.7.2 Photoelectrochemical Investigation of the Electron Transport

in Porous ZnO Films 316

11.7.3 Performance of Nanoporous Electrodeposited

ZnO Films in DSSCs 320

11.7.4 Use of ZnO Nanorods in Photovoltaics 321

11.8 Photocatalytic Properties 322

11.9 Outlook 323

References 323

12 Nanostructured Thin-Film WO3 Photoanodes for Solar Water

and Sea-Water Splitting 333

Bruce D. Alexander and Jan Augustynski

12.1 Historical Context 333

12.2 Macrocrystalline WO3 Films 334

12.3 Limitations of Macroscopic WO3 336

12.4 Nanostructured Films 336

12.5 Tailoring WO3 Films Through a Modified Chimie Douce Synthetic Route 339

12.6 Surface Reactions at Nanocrystalline WO3 Electrodes 342

12.7 Conclusions and Outlook 345

References 346

13 Nanostructured a-Fe2O3 in PEC Generation of Hydrogen 349Vibha R. Satsangi, Sahab Dass, and Rohit Shrivastav


13.2 a-Fe2O3 350

x Contentshttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-7

13.2.1 Structural and Electrical/Electronic Properties 350

13.2.2 a-Fe2O3 in PEC Splitting of Water 351

13.3 Nanostructured a-Fe2O3 Photoelectrodes 352

13.3.1 Preparation Techniques and Photoelectrochemical Response 353

13.3.2 Flatband Potential and Donor Density 365

13.4 Strategies to Enhance Photoresponse 368

13.4.1 Doping 368

13.4.2 Choice of Electrolytes 373

13.4.3 Dye Sensitizers 374

13.4.4 Porosity 375

13.4.5 Forward/Backward Illumination 375

13.4.6 Loading of Metal/Metal Oxide 377

13.4.7 Layered Structures 377

13.4.8 Deposition of Zn Islands 380

13.4.9 Swift Heavy Ion (SHI) Irradiation 382

13.4.10 p/n Assemblies 385

13.5 Efficiency and Hydrogen Production 386

13.6 Concluding Remarks 388

Acknowledgments 393

References 393

PART FOUR—NEW DESIGN AND APPROACHES TO BANDGAP

PROFILING AND VISIBLE-LIGHT-ACTIVE NANOSTRUCTURES

14 Photoelectrocatalyst Discovery Using High-Throughput Methods

and Combinatorial Chemistry 401Alan Kleiman-Shwarsctein, Peng Zhang, Yongsheng Hu,

and Eric W. McFarland


14.2 The Use of High-Throughput and Combinatorial Methods for the

Discovery and Optimization of Photoelectrocatalyst Material Systems 402

14.2.1 The Use of High-Throughput and Combinatorial Methods

in Materials Science 402

14.2.2 HTE Applications to PEC Discovery 405

14.2.3 Absorbers 408

14.2.4 Bulk Carrier Transport 411

14.2.5 Electrocatalysts 412

14.2.6 Morphology and Material System 412

14.2.7 Library Format, Data Management and Analysis 414

14.3 Practical Methods of High-Throughput Synthesis of Photoelectrocatalysts 415

14.3.1 Vapor Deposition 416

14.3.2 Liquid Phase Synthesis 417

14.3.3 Electrochemical Synthesis 419

14.3.4 Spray Pyrolysis 422

Contents xihttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-8

14.4 Photocatalyst Screening and Characterization 423

14.4.1 High-Throughput Screening 424

14.4.2 Secondary Screening and Quantitative Characterization 432

14.5 Specific Examples of High-Throughput Methodology

Applied to Photoelectrocatalysts 437

14.5.1 Solar Absorbers 437

14.5.2 Improving Charge-Transfer Efficiency 443

14.5.3 Improved PEC Electrocatalysts 448

14.5.4 Design and Assembly of a Complete Nanostructured

Photocatalytic Unit 451

14.6 Summary and Outlook 453

References 454

15 Multidimensional Nanostructures for Solar Water Splitting:

Synthesis, Properties, and Applications 459

Abraham Wolcott and Jin Z. Zhang

15.1 Motivation for Developing Metal-Oxide Nanostructures 459


15.1.2 PEC Water Splitting for Hydrogen Production 460

15.1.3 Metal-Oxide PEC Cells 460

15.1.4 Dye and QD Sensitization 462

15.1.5 Deposition Techniques for Metal Oxides 462

15.2 Colloidal Methods for 0D Metal-Oxide Nanoparticle Synthesis 463

15.2.1 Colloidal Nanoparticles 463

15.2.2 TiO2 Sol-Gel Synthesis 464

15.2.3 TiO2 Hydrothermal Synthesis 465

15.2.4 TiO2 Solvothermal and Sonochemical Synthesis 466

15.2.5 TiO2 Template-Driven Synthesis 468

15.2.6 Sol-Gel WO3 Colloidal Synthesis 470

15.2.7 WO3 Hydrothermal Synthesis 470

15.2.8 WO3 Solvothermal and Sonochemical Synthesis 470

15.2.9 WO3 Template Driven Synthesis 471

15.2.10 ZnO Sol-Gel Nanoparticle Synthesis 473

15.2.11 ZnO Hydrothermal Synthesis 474

15.2.12 ZnO Solvothermal and Sonochemical Synthesis 475

15.2.13 ZnO Template-Driven Synthesis 479

15.3 1D Metal-Oxide Nanostructures 481

15.3.1 Colloidal Synthesis and Fabrication 481

15.3.2 Synthesis and Fabrication of 1D TiO2 Nanostructures 481

15.3.3 Colloidal Synthesis and Fabrication of 1D WO3 Nanostructures 486

15.3.4 Colloidal Synthesis and Fabrication of 1D ZnO Nanostructures 487

15.4 2D Metal-Oxide Nanostructures 488

15.4.1 Colloidal Synthesis of 2D TiO2 Nanostructures 488

15.4.2 Colloidal Synthesis of 2D WO3 Nanostructures 490

15.4.3 Colloidal Synthesis of 2D ZnO Nanostructures 491

xii Contentshttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-9

15.5 Conclusion 492

Acknowledgments 493

References 493

16 Nanoparticle-Assembled Catalysts for Photochemical

Water Splitting 507

Frank E. Osterloh


16.2 Two-Component Catalysts 509

16.2.1 Synthetic and Structural Aspects 509

16.2.2 Photocatalytic Hydrogen Evolution 511

16.2.3 Peroxide Formation 513

16.2.4 Water Electrolysis 515

16.3 CdSe Nanoribbons as a Quantum-Confined

Water-Splitting Catalyst 516

16.4 Conclusion and Outlook 518

Acknowledgment 519

References 519

17 Quantum-Confined Visible-Light-Active Metal-Oxide

Nanostructures for Direct Solar-to-Hydrogen

Generation 523

Lionel Vayssieres


17.2 Design of Advanced Semiconductor Nanostructures

by Cost-Effective Technique 524

17.2.1 Concepts and Experimental Set-Up of Aqueous

Chemical Growth 524

17.2.2 Achievements in Aqueous Design of Highly Oriented

Metal-Oxide Arrays 528

17.3 Quantum Confinement Effects for Photovoltaics

and Solar Hydrogen Generation 529

17.3.1 Multiple Exciton Generation 530

17.3.2 Quantum-Well Structures 531

17.3.3 Intermediate Band Materials 531

17.4 Novel Cost-Effective Visible-Light-Active (Hetero)Nanostructures

for Solar Hydrogen Generation 533

17.4.1 Iron-Oxide Quantum-Rod Arrays 533

17.4.2 Doped Iron-Oxide Quantum-Rod Arrays 541

17.4.3 Quantum-Dot–Quantum-Rod Iron-Oxide

Heteronanostructure Arrays 545

17.4.4 Iron Oxide Oriented Porous Nanostructures 546

17.5 Conclusion and Perspectives 548

References 548

Contents xiiihttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-10

18 Effects of Metal-Ion Doping, Removal and Exchange on Photocatalytic

Activity of Metal Oxides and Nitrides for Overall Water Splitting 559

Yasunobu Inoue


18.2 Experimental Procedures 561

18.3 Effects of Metal Ion Doping 561

18.3.1 Sr2þ Ion-Doped CeO2 561

18.3.2 Metal-Ion Doped GaN 564

18.4 Effects of Metal-Ion Removal 569

18.5 Effects of Metal-Ion Exchange on Photocatalysis 573

18.5.1 YxIn2�xO3 573

18.5.2 ScxIn2�xO3 580

18.5.3 YxIn2�xGe2O7 582

18.6 Effects of Zn Addition to Indate and Stannate 583

18.6.1 Li1.6Zn1.6Sn2.8O8 584

18.6.2 Ba3Zn5In2O11 584

18.7 Conclusions 585

Acknowledgments 586

References 586

19 Supramolecular Complexes as Photoinitiated Electron Collectors:

Applications in Solar Hydrogen Production 589

Shamindri M. Arachchige and Karen J. Brewer


19.1.1 Solar Water Splitting 589

19.1.2 Supramolecular Complexes and Photochemical

Molecular Devices 590

19.1.3 Polyazine Light Absorbers 591

19.1.4 Polyazine Bridging Ligands to Construct Photochemical

Molecular Devices 594

19.1.5 Multi-Component System for Visible Light Reduction of Water 595

19.1.6 Photoinitiated Charge Separation 596

19.2 Supramolecular Complexes for Photoinitiated

Electron Collection 598

19.2.1 Photoinitiated Electron Collection on a Bridging Ligand 598

19.2.2 Ruthenium Polyazine Light Absorbers Coupled Through

an Aromatic Bridging Ligand 600

19.2.3 Photoinitiated Electron Collection on a Platinum Metal 602

19.2.4 Two-Electron Mixed-Valence Complexes for Multielectron

Photochemistry 604

19.2.5 Rhodium-Centered Electron Collectors 605

19.2.6 Mixed-Metal Systems for Solar Hydrogen Production 613


List of Abbreviations 616

Acknowledgments 616

References 617

xiv Contentshttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-11

PART FIVE—NEW DEVICES FOR SOLAR THERMAL

HYDROGEN GENERATION

20 Novel Monolithic Reactors for Solar Thermochemical Water Splitting 623

Athanasios G. Konstandopoulos and Souzana Lorentzou


20.1.1 Energy Production and Nanotechnology 623

20.1.2 Application of Solar Technologies 624

20.2 Solar Hydrogen Production 624

20.2.1 Solar Hydrogen Production: Thermochemical Processes 625

20.2.2 Solar Chemical Reactors 626

20.3 HYDROSOL Reactor 627

20.3.1 The Idea 627

20.3.2 Redox Materials 627

20.3.3 Water Splitting: Laboratory Tests 629

20.3.4 HYDROSOL Reactors 630

20.3.5 Solar Testing 631

20.3.6 Simulation 633

20.3.7 Future Developments 636

20.4 HYDROSOL Process 636


Acknowledgments 638

References 638

21 Solar Thermal and Efficient Solar Thermal/Electrochemical

Photo Hydrogen Generation 641Stuart Licht

21.1 Comparison of Solar Hydrogen Processes 641

21.2 STEP (Solar Thermal Electrochemical Photo) Generation of H2 646

21.3 STEP Theory 648

21.4 STEP Experiment: Efficient Solar Water Splitting 653

21.5 NonHybrid Solar Thermal Processes 657

21.5.1 Direct Solar Thermal Hydrogen Generation 657

21.5.2 Indirect (Multistep) Solar Thermal H2 Generation 659


References 661

Index 665

Contents xvhttp://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-12

http://media.wiley.com/product_data/excerpt/76/04708239/0470823976-2.pdf

D-13

E-1

Appendix E

Photoelectrochemical Hydrogen Production

Edited by Roel van de Krol & Michael Gratzel

Preface _______________________________________________________________________________________ E-6

Table of Contents ___________________________________________________________________________ E-8

Electronic Materials: Science & Technology

Series Editor: Harry L. TullerProfessor of Materials Science and EngineeringMassachusetts Institute of TechnologyCambridge, [email protected]

For further volumes:http://www.springer.com/series/5915

http://www.springer.com/engineering/energy+technology/book/978-1-4614-1379-0

Springer: Photoelectrochemical Hydrogen Production: Front Matter E-2

Roel van de Krol l Michael Gratzel

Editors

PhotoelectrochemicalHydrogen Production



EditorsRoel van de KrolDepartment of Chemical Engineering/Materials for Energy Conversionand StorageFaculty of Applied SciencesDelft University of TechnologyP.O. Box 5045, 2600 GA DelftThe [email protected]

Michael GratzelLaboratory for Photonics and InterfacesEcole Polytechnique Federale de LausanneCH-1015 Lausanne, [email protected]

ISSN 1386-3290ISBN 978-1-4614-1379-0 e-ISBN 978-1-4614-1380-6DOI 10.1007/978-1-4614-1380-6Springer New York Dordrecht Heidelberg London

Library of Congress Control Number: 2011939087

# Springer Science+Business Media, LLC 2012All rights reserved. This work may not be translated or copied in whole or in part without the writtenpermission 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 inconnection 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 theyare not identified as such, is not to be taken as an expression of opinion as to whether or not they aresubject to proprietary rights.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)



Preface

Hydrogen is a highly versatile fuel that may become one of the key pillars to

support our future energy infrastructure. It can be efficiently converted into elec-

tricity using a fuel cell, or it can directly drive an internal combustion engine. Using

hydrogen is clean; the only reaction product upon oxidation is pure water, with little

or no exhaust of greenhouse gases. It can even be converted into more convenient

form of fuel, a liquid hydrocarbon, using excess CO2 and well-established Fischer–

Tropsch technology. However, hydrogen does not occur freely in nature, and

producing hydrogen in a clean, sustainable, and economic way is a major challenge.

This book is about tackling that challenge with semiconductors, using water and

sunlight as the only ingredients. The ultimate aim is to make a monolithic photo-

electrode that evolves hydrogen and oxygen at opposite sides of the electrode, so

that they can be easily separated. Finding semiconductors that can do this efficiently,

at low cost, and without suffering from corrosion is far from trivial. The emphasis in

this book is on transition metal oxides, a low-cost and generally very stable

class of semiconductors. There is a darker side to these materials, though. The

bandgap of metal oxide semiconductors is often a bit too large, and the optical

absorption coefficient is usually small. In addition, the catalytic activity for

water oxidation or reduction at the surface is generally poor, and the electronic

charge transport properties can be downright horrible. This issues have thwarted

many earlier efforts in the late 1970s and early 1980s to reach the “Holy Grail”

of solar water splitting. In the past few years, however, exciting breakthroughs

in nanotechnology have stimulated a huge amount of renewed interest in this

field. This book attempts to summarize both the basic principles and some of the

important recent developments in photoelectrochemical water splitting. While

we cannot even hope to approach completeness in a single volume, we never-

theless hope that both experts and newcomers in this field find something useful

here that helps their research.

The book is organized into four parts. The first part covers basic principles and

is specifically aimed at undergraduate and graduate students, as well as colleagues

who are new to the field. Chapter 1 provides a brief motivation for our interest in

solar hydrogen production. The properties of semiconductors, the semiconductor/

v



electrolyte interface, and basic PEC device operation are covered in Chap. 2, while

an overview of photoelectrochemical measurement techniques is given in Chap. 3.

The second part of the book is on materials properties and synthesis. In Chap. 4,

Kevin Sivula discusses the intrinsic properties of a-Fe2O3 (hematite) that limit its

performance as a photoanode, and how these limitations can be overcome by

nanostructuring. Kazuhiro Sayama outlines the properties of ternary and mixed

metal oxide photoelectrodes in Chap. 5, showing recent results on BiVO4 and a

high-throughput screening method. In Chap. 6, Bruce Parkinson takes the high-

throughput concept to the next level by discussing combinatorial approaches to

discover new candidate materials and to screen thousands of compositions in a

quick and systematic fashion. The third part of the book is on devices and device

characterization. This part consists of a single, extensive chapter by Eric Miller,

Alex DeAngelis, and Stewart Mallory on multijunction approaches and devices for

solar water splitting (Chap. 7). They analyze the merits of various tandem config-

urations and materials combinations, and give an overview of key aspects to be

considered in future research efforts. The fourth and final part of the book gives an

overview of some of the future perspectives for photoelectrochemical water

splitting. In Chap. 8, Julian Keable and Brian Holcroft take a closer look at the

economic and business perspectives, and set the device performance targets that

need to be met in order to commercialize the technology. In the final chapter, Scott

Warren describes how some of the recent developments in nanotechnology and

nanophotonics can be leveraged in solar water splitting materials, offering an

exciting glimpse at future performance breakthroughs (Chap. 9).

Putting together a volume like this is a big undertaking in which many people are

involved. First and foremost, the editors express their sincere thanks to all the

contributors. We hope they are pleased with the fruits of our collective labor, and

greatly appreciate their patience during the lengthy course of this project. We thank

the people of Springer for their encouragement and support throughout the project:

Elaine Tham, Lauren Danahy, Merry Stuber, and especially Michael Luby. A final

and special thanks goes to the series editor, Prof. Harry Tuller, for inviting us to edit

a volume on the exciting subject of solar water splitting.

Delft, The Netherlands Roel van de Krol

Lausanne, Switzerland Michael Gratzel

vi Prefacehttp://www.springer.com/engineering/energy+technology/book/978-1-4614-1379-0


Contents

Part I Basic Principles

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Roel van de Krol and Michael Gratzel

2 Principles of Photoelectrochemical Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Roel van de Krol

3 Photoelectrochemical Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Roel van de Krol

Part II Materials Properties and Synthesis

4 Nanostructured a-Fe2O3 Photoanodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Kevin Sivula

5 Mixed Metal Oxide Photoelectrodes and Photocatalysts . . . . . . . . . . . . . 157

Kazuhiro Sayama

6 Combinatorial Identification and Optimization of New

Oxide Semiconductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Bruce A. Parkinson

Part III Devices and Device Characterization

7 Multijunction Approaches to Photoelectrochemical

Water Splitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Eric L. Miller, Alex DeAngelis, and Stewart Mallory

vii



Part IV Future Perspectives

8 Economic and Business Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277

Julian Keable and Brian Holcroft

9 Emerging Trends in Water Photoelectrolysis . . . . . . . . . . . . . . . . . . . . . . . . . 293

Scott C. Warren

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

viii Contentshttp://www.springer.com/engineering/energy+technology/book/978-1-4614-1379-0


ISBN 978-0-9815041-6-2E-10

iea hydrogen implementing agreement task 26: advanced ......a2 database development c2 advanced...

Documents