A practical solar water splitting device that can produce H₂ at a cost less than PV + electrolysis is difficult to envision without it possessing a simple construction that employs widely available materials and inexpensive processing techniques. A tandem cell consisting of 2 solution-processed photoelectrodes (photoanode/photcathode) can reasonably reach this goal. While n-type oxide semconductors have been established as viable photoanode materials, cheap and stable p-type photocathodes are less developed. In this presentation our gorups&’ progress in the development of solution-processed stable photocathodes is presented and their application toward overall photoelectrochemical water splitting tandem cells is demonstrated. Three classes of promising materials are discussed: copper-based oxides in the delafossite phase, ternary and quaternary chalcogenides, and 2D transition metal chalcogenides. In each case the challenges to maximize photogenerated charge collection in thin films prepared by solution-based deposition approaches are highlighted and the use of bottom-up nanostructuring to gain insight into routes for improvement and overcome these challenges are established. The optimization of these materials for solar water reduction using overlayers and catalysts, and the stability under reasonable operation conditions are addressed. Routes to improve the unassisted and overall solar water splitting performance in tandem cells with commonly used oxide photoanodes are finally outlined.

This talk will present an overview of our research strategy to develop a semiconductor-based device capable of over 20% solar-to-hydrogen efficiency with several thousand hours of stability under operating conditions. Our goal is to improve solar-to-hydrogen (STH) efficiency from 12.4% to over 20% by developing novel tandem semiconductor materials and configurations. Our approach focuses on classes of materials that have either demonstrated exceptionally high efficiency or theoretically can produce highly efficient materials. The four tandem material sets in our research portfolio are III-Vs on GaAs, InGaN on Si, III-V-N on Si, and dual photoelectrode pairings. We are evaluating catalytic nitride, oxide, and sulfide-based semiconductor surface modifications to extend durability from a few hundred hours to nearly 1000 hours, with a stretch target of 3500 hours. Cost reductions in the synthesis of high-efficiency III-V photoelectrochemical devices could be realized from emerging epitaxial synthesis technologies such as spalling, epitaxial lift-off, or close-space vapor transport to yield economical hydrogen production. Our group addresses synthesis cost, in a limited capacity, by investigating novel material configurations.
Ultimately, we plan to demonstrate a prototype photoreactor that produces 3 L of standard hydrogen within an 8-hour period under moderate solar concentration (10x), which can be accomplished using only 6 cm² of a 20% STH material.

Layered transition metal dichalcogenides of the general formula, MX₂ (where M= Mo, W and X=S, Se), are of significant interest as photoelectrode materials in solar water-splitting due to their favorable bandgap, high absorption coefficients and good photoelectrochemical stability. However, the surfaces of these materials are highly heterogeneous exhibiting macroscopic terraces and step edges; along with other micro/nano-scale defects. It has been suggested previously that the collection of photogenerated carriers at the semiconductor/electrolyte interface occurs efficiently on smooth flat terraces; whereas the step edges and other defects act as carrier recombination sites. The efficiency of the overall photoconversion process in such materials, hence, would be crucially determined by the interfacial charge transfer occurring at these surface motifs and their surface densities. Thus, understanding the charge transfer process at the microscopic level on the surface of these materials is foundational to their application in solar water-splitting.
To this end, we have carried out spatially resolved local photo-current measurements on layered chalcogenide photoelectrodes using the technique of laser beam induced current microscopy. Importantly, the studies showed that, more than the step edges, the presence of terraces with low photoactivities primarily accounted for the efficiency losses in these materials. In an effort to expose the underlying reasons for the existence of low-performing terraces, local topographic and spectral response measurements were subsequently carried out. It was generally established that terraces, at the micro/nano-scale, are not uniform but textured and the texturing differs from one terrace to another. The results also showed that the local electronic structures of terraces differ from one another. Furthermore, compared to high photoactivity terraces, the low-performing terraces clearly indicated the presence of sub-bandgap surface states that could be acting as carrier traps. In summary, our study, employing spatially resolved techniques, has yielded significant phenomenological insights into the photoconversion process on the class of layered chalcogenide photoelectrode materials for solar water-splitting.

Photoelectrochemistry (PEC) is one of the most efficient methods to produce alternative fuels, although the efficiency, cost, and durability of lab-scale systems are currently not at the level required to make this technology economically feasible. The chalcopyrite material class, typically identified by its most popular alloy Cu(In,Ga)Se₂, provides exceptionally good candidates to meet the requirements identified for cheap, sustainable solar fuels production. As we recently reported[i], co-evaporated 1.7 eV CuGaSe₂ offers very high-saturated photocurrent densities (15 mA.cm^-shy;2 in 0.5M H₂SO₄ under AM1.5G illumination), long durability (up to 400 hours), and high Faradaic efficiency (>85% for non-catalyzed systems). Although CuGaSe₂ has the highest bandgap of the copper chalcopyrite class, its optical characteristics are still too close to that of amorphous silicon (a-Si), a low-cost material our research team has identified as an ideal photovoltaic driver in a monolithic hybrid photoelectrode device. Nevertheless, a solar-to-hydrogen efficiency of 3.7% was achieved using a co-planar integration scheme. In order to improve the water-splitting efficiency further, novel chalcopyrite alloys with bandgap greater than 1.7 eV must be developed.
In the present communication, we report on our effort to synthesize wide bandgap (1.8 eVG<2.2 eV) band-gap chalcopyrite materials for un-biased PEC water splitting using PEC/PV hybrid devices. Specifically, we investigate the impact of sulfur on the optical and photoelectrochemical characteristics of the copper chalcopyrite material class. Using co-evaporated 1 mu;m-thick CuGaSe₂ as baseline system, we demonstrate that selenium can be substituted by sulfur using a simple annealing step. With this protocol, a dramatic change in optical properties was observed, with a bandgap increase from 1.7 eV (CuGaSe₂) to 2.4 eV (CuGaS₂), in good agreement with theoretical predictions[ii]. Then, by simply adjusting the indium content in the film during the initial growth process, red 2.0 eV CuIn_0.3Ga_0.7S₂ was obtained. Mott-Schottky analysis indicated a 200 mV anodic shift of the flatband potential with increasing bandgap (300 meV), suggesting that the bandgap modification in sulfurized films primarily stems from a downward shift of the valence band, an ideal situation for p-type PEC systems. Linear sweep voltammetry performed in 0.5M H₂SO₄ under AM1.5_G simulated illumination revealed the excellent photo-conversion properties of 2.0 eV CuInGaS₂, with photocurrent densities of 3.5 and 6.0 mA/cm² at 0 V_RHE and -0.4 V_RHE, respectively, with negligible dark current from flatband potential (+0.5 V_RHE) to photocurrent saturation (-0.5 V_RHE). Preliminary results obtained on PV/CuInGaS₂ hybrid structures will be also presented.
[i] N. Gaillard, D. Prasher, J. Kaneshiro, S. Mallory, and M. Chong, MRS Spring Meeting, Z2.07 (2013).
[ii] M. Bär, W. Bohne, J. Rohrich, E. Strub et al., Appl. Phys. Lett. 96, 3857 (2004).

With the emergence of highly efficient perovskite materials, there is a need to understand the excited state behavior and charge separation events in these new materials[1,2]. The scientific issue related to the utilization of organic metal halide perovskite materials in solar fuels generation remains a scientific challenge because its susceptibility to degradation. The interaction of CH₃NH₃PbI₃ with water vapor (90% humidity exposure) has allowed us to identify the transformations leading to the deterioration of the perovskite solar cells. The reaction with H₂O at the surface of the perovskite film forms shallow traps in the band structure so that the portion of the CH₃NH₃PbI₃ crystal which is pristine remains largely unaffected [3]. In order to circumvent direct exposure of the perovskite film to water electrolysis, we have employed a tandem water splitting assembly composed of a BiVO₄ photoanode and a single-junction CH₃NH₃PbI₃ hybrid perovskite solar cell [4]. This unique configuration allows efficient solar photon management, with the metal oxide photoanode selectively harvesting high energy visible photons and the underlying perovskite solar cell capturing lower energy visible-near IR wavelengths in a single-pass excitation. The sustained photovoltage and photoconversion efficiency of the single-junction CH₃NH₃PbI₃ PV even under low energy excitation enables exceptional performance in tandem light-harvesting assemblies.
References
Manser, J. S.; Kamat, P. V., Band Filling with Charge Carriers in Organometal Halide Perovskites. Nature Photonics 2014, 8, 737-743.
Stamplecoskie, K. G.; Manser, J. S.; Kamat, P. V. Dual Nature of the Excited State in Organic-Inorganic Lead Halide Perovskites Energy Environ. Sci. 2015, 8, 208 - 215
Christians, J. A.; Miranda Herrera, P. A.; Kamat, P. V., Transformation of the Excited State and Photovoltaic Efficiency of CH3NH3PbI3 Perovskite upon Controlled Exposure to Humidified Air. J. Am. Chem. Soc. 2015, DOI: 10.1021/ja511132a
Chen, Y.-S.; Manser, J. S.; Kamat, P. V., All Solution-Processed Lead Halide Perovskite-BiVO4 Tandem Assembly for Photolytic Solar Fuels Production. J. Am. Chem. Soc., 2015, 137, 974-981.

For the first time, double-deck WO₃/BiVO₄ inverse opal photoanodes (DDIO-WO₃/BiVO₄) were prepared by swelling-shrinking mediated polystyrene template synthetic routes, and the use of the photoanodes in photoelectrochemical cells under simulated solar light was investigated. The double-deck photoanodes represented the compact interface between WO₃ and BiVO₄, inheriting the periodically ordered macroporous nanostructure. More significantly, the DDIO-WO₃/BiVO₄ inverse opal photoanodes prepared from the optimized fabrication condition demonstrated a photocurrent that was about 40 times higher than that of the pure inverse opal WO₃ photoanodes at a bias of 1.23V vs. RHE. Even without an added catalyst, they produce an outstanding photocurrent density of about 3.3 mA/cm² at a bias of 1.23V vs. RHE, which profits from improving the poor charge carrier mobility of BiVO₄ by combining it with a WO₃ skeleton and a shrouded bilayer inverse opal structure with a large surface area and good contact with the electrolyte.

Solar-to-fuel efficiency in photoelectrochemical cells is often limited by the recombination rate of photo-excited charge carriers and the kinetic overpotential at the semiconductor/electrolyte interface. In small-polaron semiconductors such as BiVO₄, carrier transport and electrocatalysis are thermally activated and benefit from higher operating temperatures. In this work, we studied the effect of heating in nanostructured BiVO₄ in aqueous electrolyte. Our results have shown that while the open-circuit potential decreased with temperature at a rate of 2.1 mV/K, the photocurrent increased dramatically. By elevating the temperature from 9 #8451; to 42 #8451;, the saturation current density increased by 52%, from 2.1 to 3.2 mA/cm². The stability under different temperatures was also explored. This work demonstrates that heating is a promising route to improve the photoelectrochemical activity of BiVO₄, paving the way for further studies in solid-state cells.

Photoactive materials used in the oxygen evolution reaction for clean hydrogen water splitting technologies have been researched since the discovery of TiO₂ as a photocatalytic material in 1972.^[1] Our current research is based on constructing a photoelectrocatalytic cell comprised of thin film photo-electrodes in a dual photo-system configuration which can thermodynamically attain the highest efficiency of any other cell architecture in a photocatalytic cell.^[2] This requires the photoelectrodes to be photo active, durable, low cost, scalable, possess varying band gaps, and requires one to be transparent. Electrodes are fabricated via sol-gel synthesis followed with thin film deposition by dip-coating; a technique which is low cost, easily scalable, and utilizes low temperature procedures. Our research is focused on the low temperature fabrication of transparent photoanodes made of WO₃ and/or BiVO₄. Tungsten trioxide, with its large internal quantum efficiency, band gap of 2.7eV, and its stability in acidic conditions make it a good candidate for water splitting technologies.^[3] However, for a system working in neutral conditions, bismuth vanadate with its band gap of 2.4eV is more stable at pH 7 and absorbs more visible light. These properties may allow BiVO₄ to be used as a protective and n-tandem layer to coat WO₃ and may as well increase charge carrier separation of the photoelectrode. Several approaches will be discussed in order to increase the efficiency of these materials; most notably: doping for increased conductivity and charge separation, nanostructuration for increased specific surface, and addition of catalysts to facilitate the water oxidation kinetics.^[4] WO₃ and BiVO₄, characterized by scanning electron and transmission electron microscopes, x-ray diffraction, UV-visible absorption and transmission, x-ray photoelectron spectroscopy, and photoelectrochemistry, are promising candidates for thin film photoelectrodes in a dual photo-system photoelectrocatalytic cell.
[1] A. Fujishima and K. Honda, Nature 1972, 238, 37-38.
[2] J. R. Bolton, S. J. Strickler and J. S. Connolly, Nature 1985, 3116, 495-500.
[3] B. D. Alexander, P. J. Kulesza, I. Rutkowska, R. Solarska and J. Augustynski, Journal of Materials Chemistry 2008, 18, 2298-2303.
[4] M. W. Kanan and D. G. Nocera, Science 2008, 321, 1072-1075.

Metal oxides have emerged as attractive candidates for photoelectrochemical water splitting, mainly due to their good stability in aqueous solutions, easy synthesis, and low cost. One of the most promising metal oxide photoanodes is bismuth vanadate (BiVO₄)¹. This material has been shown to evolve oxygen under illumination with visible light, and is stable in aqueous solutions with pH between 3 and 11. In addition, its conduction band edge is close to the reversible hydrogen electrode potential, which means that only little extra voltage is needed for evolution of hydrogen at the counter electrode in a BiVO₄-based water splitting device. One of the challenges for BiVO₄ is the efficient separation of electrons and holes. Recent work has shown that nanostructuring and doping are effective solutions for this problem^2-4. Another issue that appears to limit the efficiency is slow oxygen evolution kinetics. The deposition of a co-catalyst, such as cobalt phosphate (CoPi), has been shown to greatly enhance the photocurrent and appears to address this issue^1,5. However, the exact mechanism for this improvement is not yet fully clear.
In this study, we use Intensity Modulated Photocurrent Spectroscopy (IMPS) to examine the photocurrent kinetics of spray deposited BiVO₄ photoanodes. An LED is used to illuminate the sample with a modulated intensity, and the real and imaginary parts of the opto-electrical impedance (i.e., the ratio between photocurrent and light intensity) are recorded. To interpret the resulting spectra, we used a model developed by Peter et al. that allows one to distinguish the rate constants for surface recombination and charge injection into the electrolyte.⁶ A comparison of bare and CoPi-catalyzed BiVO₄ reveals that at modest applied potentials, the CoPi reduces the surface recombination rate by ~2 orders of magnitude. A similar surface passivation effect has been reported for CoO_x-catalyzed hematite.⁶ More surprisingly, the CoPi seems to reduce the rate constant for charge injection into the electrolyte by a factor of ~10. Although the surface passivation effect still outweighs the reduction in charge transfer kinetics, resulting in higher photocurrents for CoPi-catalyzed BiVO₄, the latter effect seems to contradict its function as an electrocatalyst. At more positive applied potentials (> 1.2 V vs RHE) the rate constants for surface recombination and charge transfer no longer depend on the illumination intensity, which implies that bulk recombination dominates in this potential regime. The implications of these intriguing observations will be discussed.
References
[1] F. F. Abdi, R. van de Krol, J. Phys. Chem. C, 2012, 116, 9398.
[2] F. F. Abdi et al., Nat. Commun. 4:2195 (2013) 1.
[3] T. W. Kim et al., Science 343 (2014) 990.
[4] X. Shi et al., Nat. Commun. 5:4775 (2014) 1.
[5] D. Wang et al., J Phys. Chem. C 116 (2012) 5082.
[6] L. M. Peter, J. Solid State Electrochem., 2013, 17, 315.

A complete control on morphology and microstructure of photoelectrodes is needed to establish precise correlations between structure and photocatalytic properties. Solution-based colloidal methods for preparing oxide semiconductors allow for the use of inexpensive and scalable processing techniques, like spin-coating or doctor-blades, and can give mesoporous thin films with controlled morphology at different lengtscales. The prototypical example of this approach is the mesoporous TiO₂ used in dye-sensitized solar cells, where porosity, crystallite size and shape can be tuned independently. Recently, Sivula et al. have revealed the potentiality of using colloidal-based ink for hematite films to better understand the impact of the hematite structural properties on its performance as a photoanode for water oxidation.[1] As for hematite, morphology seems to be key in the performance optimization of m-BiVO₄ photoanodes. In fact, the state-of-art photocurrent reported so far for m-BiVO₄ has been measured in films with nanoporous morphology.[2] However, low temperature routes to nanocrystal-based ink for the B-V-O family have not been fully developed yet.
Herein, our recent work on colloidal nanocrystal-based inks for BiVO₄ and Sb doped-BiVO₄ photoanodes will be discussed.[3, 4]The nanocrystal composition and morphology was controlled through a one-pot seeded growth approach in which bismuth or bismuth antimony nanocrystals were reacted with the vanadium precursor in properly chosen conditions. After the synthesis, optically clear and colloidally stable of nanocrystals were used as ink to obtained monoclinic BiVO₄ and Sb-BiVO₄ upon annealing. In-situ XRD studies gave insight into the phase transformations occurring at different temperatures. Finally, charge carrier dynamics data and preliminary photoelectrochemical measuraments highlight the potential of these novel ink-based routes to the Bi-V-O photoanode family to build meaningful correlations between film structure and practical solar water oxidation functions.
1. (a) Brillet, J.; Gratzel, M.; Sivula, K Nano Lett. 2010, 10, 4155; (b) Sivula, K.; Le Formal, F.; Gratzel, M. ChemSusChem, 2011, 4, 432.
2. Kim, T. W.; Choi, K.-S. Science 2014, 343, 990.
3. Loiudice, A.; Cooper, J. K.; Mattox, T.; Sharp, I. D.; Buonsanti, R. "Colloidal Bi₂O_2.7/VO_x nanocrystal heterodimers for ink-based BiVO₄ films" submitted.
4. Loiudice, A.; Cooper, J. K.; Mattox, T.; Thao, T.; Drisdell, W.; Ma, J.; Wang, L-W; Yano, J.; Sharp, I. D.; Buonsanti, R. "Ink-based mesoporous Sb-BiVO₄ films" in preparation.

The US Department of Energy&’s (DOE) Fuel Cell Technologies Office has made significant progress in hydrogen and fuel cell technology advancement and cost reduction. With the expected rollouts of fuel-cell vehicles by major automotive manufacturers over the next several years, enabling technologies for the widespread production of affordable renewable hydrogen becomes increasingly important. Solar water-splitting processes, including photo-assisted electrolysis as well as the direct photoelectrochemical and thermochemical routes, can play a significant role. However, fundamental materials challenges remain which limit conversion efficiency and durability of these pathways, therefore preventing large-scale technoeconomic viability. Innovations are needed in the development of functional materials and interfaces which address the thermodynamic and kinetic limitations to performance and lifetime at the macro-, meso- and nano-scales. Researchers are increasingly adopting Clean Energy Materials Genome Initiative methodologies integrating advanced computation, experimentation and informatics to accelerate the materials discovery and development process. Examples of this approach specifically applied in the R&D of photoelectrochemical and thermochemical processes are discussed, and beneficial cross-cutting implications for all solar hydrogen pathways are highlighted.

Although fully integrated devices for solar fuels are in the early stages, it is possible to consider how a future technology might look from what we know today. Recent studies by the Joint Center for Artificial Photosynthesis that assess the net energy balance for both a solar hydrogen device and a full facility find that the value is positive as long as sufficient efficiency and durability are achieved. JCAP&’s scientific and engineering programs are focused on improvements in materials systems and device designs to meet these two criteria. In this talk I will highlight our recent work on discovery and characterization of catalysts, light absorbers, corrosion barriers, membranes and architectures for hydrogen generation, and discuss the sensitivity of sustainability considerations to them.

The photoelectrochemical (PEC) process is an innovative approach for hydrogen production in which energy collection (solar absorber) and water electrolysis (catalysis) are mated into a single device. However, the bulk of the research over the last 40+ years has focused on metal oxides were the efficiency of PEC solar hydrogen generation is low due to the combined challenges of poor solar absorption and poor electronic properties. More suitable semiconductors for high efficiency solar photoconversion such as the III-Vs suffer from material instability and slow interfacial kinetics towards the water redox reactions. Integration of electrocatalysts to the semiconductor&’s surface is a necessary approach to both stabilize the interface and increase catalysis, thus enhancing the overall device performance. Stable catalysts for surface modification are particularly beneficial if they are also potentially low-cost, scalable, and transparent. Work on hydrogen evolution catalysts has been a very active area of research where numerous molecular, nanomaterial, and bulk catalysts have been developed. Noble metals, particularly platinum, are most commonly applied as they are the most active for the water redox reactions. These metals are neither earth abundant nor low-cost, so identifying catalytic systems that can match the activity and stability of noble metals but are based on earth abundant materials are clearly a high-priority area of research.
We will report on the immobilization of a cobaltoxime hydrogen evolution catalyst on a p-GaInP₂ surface modified with a thin TiO₂ layer via atomic layer deposition. Under illumination in pH 13 aqueous solution, the GaInP₂-TiO₂-cobaltoxime photocathode exhibited remarkable stability during 24 hours of continuous operation at 0 V vs. RHE. We observed a turnover number of 303,000 with a turnover frequency of 210 min^-1 during that time. A high IPCE, up to ~70% through the visible range (<690nm), was measured for this structure, clarifying the advantage of a visible-light transparent hydrogen evolution catalyst.

Hydrogen is the simplest solar fuel to produce and in this presentation we shall give a short overview of the pros and cons of various tandem devices [1]. The large band gap semiconductor needs to be in front, but apart from that we can chose to have either the anode in front or back using either acid or alkaline conditions. Since most relevant semiconductors are very prone to corrosion the advantage of using buried junctions and using protection layers offering shall be discussed [2-4]. Next we shall discuss the availability of various catalysts for being coupled to these protections layers and how their stability may be evaluated [5, 6]. Examples of half-cell reaction using protection layers for both cathode and anode will be discussed though some of recent examples under both alkaline and acidic conditions. Si is a very good low band gap semiconductor and by using TiO₂ as a protection layer we can stabilize it for both H₂ and O₂ evolution [7, 8, 9, 10]. Notably NiO_x promoted by iron is a material that is transparent, providing protection, and is a good catalyst for O₂ evolution. We have also recently started searching for large band gap semicondutors like III-V based or pervoskite materials and follow the same strategy by using protection layers and catalysts [11].
References
[1] B. Seger et al. Energy & Environment Science 7 2397 (2014)
[2] B. Seger, et al. Angew. Chem. Int. Ed., 51 9128 (2012)
[3] B. Seger, et al., JACS 135 1057 (2013)
[4] B. Seger, et al., J. Mater. Chem. A, 1 (47) 15089 (2013)
[5] R. Frydendal, et al. Accepted Chem.Elec.Chem (2014).
[6] E. A. Paoli, et al. Chemical Science, (2014), DOI: 10.1039/C4SC02685C
[7] A. B. Laursen et al., Energy & Environment Science 5 5577 (2012)
[8] A. B. Laursen, et al. Chem. Com. 49 4965 (2013)
[9] A. B. Laursen, et al.,Phys. Chem. Chem. Phys., 15 20000 (2013)
[10] B. Mei, et al. J. Phys. Chem. Lett. 5 1948 (2014)
[11] M. Malizia, et al. J. Mater. Chem. A DOI: 10.1039/C4TA00752B (2014)

Photoelectrochemical hydrogen production using sensitized photoelectrode as photoanode is a viable choice to store solar energy into chemical fuels. The systems that employ organic dyes or quantum dots as photosensitizers often require external bias or sacrificial reagents to overcome the charge transfer limitations at photoanode. We now report a bias-free photoelectrochemical hydrogen production system using glutathione-capped gold nanoclusters (Au_x-GSH) as photosensitizer. These clusters provide a favorable reduction and oxidation window to induce hydrogen evolution reaction (HER) and oxygen evolution reaction (OER). A two electrode photoelectrochemical cell (PEC) consisted of Au_x-GSH sensitized TiO₂ photoanode and Pt counter electrode was fabricated. When the photoanode was illuminated with visible light (> 420 nm), we observe a significant photocurrent activity under short circuit condition in an aqueous buffer (pH = 7). Over an hour long period of illumination, we detected 2 µmoles of hydrogen with a Faradaic efficiency of 65 %. By introducing EDTA as a sacrificial reagent in the electrolyte, around three times of hydrogen yield was observed. The capability of this newly discovered photosensitizer to conduct photoelectrochemical hydrogen production without applying external bias is superior to other photosensitizers. Our study paves the way to explore other nanoclusters with different physical and chemical properties as photosensitizers for hydrogen production from sunlight.

Electrochemical energy conversion devices that can directly capture and store solar energy in the form of fuels are a promising alternative to increase the share of renewables in our current energy landscape. Significant research efforts have been devoted towards the development of photoelectrochemical components (i.e. light absorbers and water splitting catalysts) that are able to spontaneously split water in the presence of solar irradiation. These efforts have led to major advances in the solar-fuels field, while only limited attention has been given to understanding the factors that drive economically viable solar-fuel generators. The study presented here introduces a system agnostic approach that allows for the understanding of economic factors behind the design of solar-hydrogen generators. Using this framework, we evaluate the underpinning effects of the material selection for the light absorption and water splitting components on the cost of the generated fuel ($/Kg of H₂). Furthermore, designs for solar water-splitting devices are optimized for a given material&’s system in order to minimize the cost of hydrogen production. The results presented here provide insights into engineering aspects related to the sizing of devices and the use of light concentration components that, when optimized, can lead to costs below $2.90 per kilogram of hydrogen after compression and distribution. In order to achieve these low hydrogen production costs, optimized devices will need to be composed by water splitting components that are significantly smaller in area than light absorption units (by a factor 10-100&’s). Moreover, the analysis demonstrates that in optimized devices the cost of hydrogen is primarily driven by the light-absorbing component while the material selection for the electrolysis components has minor effects. The findings discussed here can help direct research and development efforts towards the fabrication of scalable solar-hydrogen generators that are cost competitive with commercial energy sources.

Up to the present, the best way to achieve the highest solar-to-hydrogen (STH) energy conversion efficiency is the combination of high-efficiency photovoltaic (PV) cells and electrochemical cells (ECs) for water splitting with low overpotential. Such a combination liberate us from a constraint for materials and structures: generating high electrochemical potential from photons while keeping tolerance against corrosion in water. In this work, we connected concentrator photovoltaic (CPV) modules with a sun tracker to commercial polymer ECs and obtained a STH efficiency of 15.3% under a direct normal irradiance (DNI) of approximately 770 W/m² in Miyazaki, Japan. The sunlight incident on the surface of a CPV module (20×20 cm² in area) was concentrated to a triple-junction PV cell with a ratio of 800 and the conversion efficiency to electricity was 24%. The output of the module was connected to ECs by copper wires without any circuit. The operation point of the system was determined in an autonomous manner as the cross point between two current-voltage characteristics for the CPV module and the EC. Such direct connection eliminates characteristic loss associated with electrical circuits for DC-AC conversion and power-maximum-point tracking. On the other hand, the direct connection imposed us the optimization of series/parallel connection among PV modules and ECs for obtaining the maximum STH efficiency. The number ratio of PV modules and ECs in series determined the operation voltage which should be close to the power maximum voltage of the PV module. On the other hand, an increase in the number of parallel connections for ECs reduced the series resistance, which is probably due to the polymer membrane, and enhanced STH efficiency by reducing the voltage loss from the terminals of the EC to the free energy of H₂ (1.23 V).
The strategy for maximizing STH efficiency will be discussed. Briefly, STH efficiency is the product among 1) an efficiency of a PV module, 2) the ratio of the operation point voltage over the power maximum voltage of the PV module and, 3) the ratio of redox potential of H₂ evolution over the operation point voltage, and 4) Faraday efficiency. Using the configuration of 3PV-4EC series connection with 3 parallel lines for ECs, amounting to 12 ECs in total, each efficiencies above are 1) 0.24, 2) 0.92, 3) 0.69 and 4) 1.0, leading to 0.153 as an STH efficiency. The maximum STH efficiency of 20% can be obtained by increasing 1) to 0.27, 2) to 0.95, and 3) to 0.77. The PV module efficiency 1) of 0.27 is quite realistic even with existing CPV modules and even higher values are within our scope. The loss due to EC overpotential 3) then governs the STH efficiency. Toward the maximum efficiency of 0.83, which is the ratio ΔG/ΔH for H₂O evolution, we need both low series resistance associated with the membrane and a small overpotential at the electrode. Low current density is of course beneficial for these, but it is at a sacrifice of cost.

Separation of charge carriers and reaction products in photocatalysts or photoelectrodes is a key issue for efficient solar water splitting. However, implementation requires spatially separated oxidation and reduction parts in a single system which have not been developed. In addition, enlarging the solar absorption bandwidth and usage of abundant materials are also important. To this end, we have developed nano-photochemical cells composed of open-ended carbon nanotube arrays with photocatalysts on the outside and metal co-catalysts on the inside of each nanotube. The carbon nanotubes were synthesized within Titania nanotubes, improving the visible light response of the photocatalyst and promoting fast charge separation. The oxidation sites are on the outside of the tubes, where the Titania is located, and the reduction sites are on the inside of the tube, which was decorated with Pt nanoparticles.
We will present results on the enhanced photocatalytic performance, determined from the degradation of methylene blue, and the photoelectrochemical performance, determined from photocurrent density under visible light. The separation of the oxidation and reduction sites is verified by the photodeposition of metal and metal oxide under visible light. Ag nanoparticles were deposited on the inside of the tubes, indicating that inside of the tubes is the location of the reduction sites, while MnO_x was deposited on the outside of the tubes, implying that the outside is the location of the oxidation sites. This demonstrates the spatial separation of oxidation and reduction sites within a single nanotube with enhanced photocatalytic and photoelectrochemical activity under visible light. This work has been partially supported by the NSF through Grants DMR 0645596 and CMMI 1335417.

Atomic layer deposition (ALD) has recently emerged as a powerful tool for protection of photoelectrodes from corrosion, either by introduction of interfacial layers between catalysts and semiconductors or by direct integration of catalysts onto semiconductors. However, the composition and structure of materials deposited by ALD often differ significantly from those synthesized by other methods. In this work, we investigate the activity of cobalt oxide deposited by plasma-enhanced ALD (PE-ALD) for catalyzing - both electrochemically and, when integrated with a silicon light absorber, photoelectrochemically - the oxygen evolution reaction (OER). We find that the electrocatalytic activity is significantly improved at a lower deposition temperatures. By controlling composition and structure, we determine that this improved activity is related to the amorphous character catalyst deposited at low temperature, which provides higher concentrations of active sites for the OER reaction per geometric area. The thin and conformal cobalt oxide films display ultra-stable electrochemical performance for at least 6 days at controlled current of 10 mA cm^-2. When the low temperature PE-ALD CoO_x is integrated onto nanotextured p⁺n-Si surfaces, significantly enhanced photoelectrochemical water oxidation activity is observed. A photocurrent as high as 30 mA cm^-2 has been achieved on CoO_x/p⁺n Si at 1.23 V vs. RHE under AM 1.5 solar simulator illumination at high pH (13.6), among the highest photocurrents achieved on crystallized Si under the same conditions. This work provides insight into the role of structural disorder on catalytic activity and provides a path to higher performance photoelectrodes for solar fuels systems.

Metal oxide nanostructures with hetero-contacts and phase boundaries offer unique platform for designing materials architectures for solar harvesting applications. Besides the size and surface effects, the modulation of electronic behavior due to junction properties leads to modify surface states that promote higher efficiency. The growing possibilities of engineering nanostructures in various compositions (pure, doped, composites, heterostructures) and forms (particles, tubes, wires, films) has intensified the research on the integration of different functional material units in a single architecture to obtain new materials for solar energy harvesting application.
In this work we present the deposition and modification of semiconducting metal oxides and their multilayers (TiO₂, Fe₂O₃ and TiO₂/Fe₂O₃) for photoelectrochemical (PEC) hydrogen production. The deposition parameters for thin film creation were optimized with respect to the PEC performance of the resulting materials in both alkali solution and simulated seawater. The long-term performances of the metal oxide photoanodes were determined in alkali and seawater electrolyte, as well. The results presented that the multilayered TiO₂/Fe₂O₃ photonanode yielded higher photocurrent density (1.8 mAcm^-2 at 1.23 V) with very stable conditions even after 1-week measurement.

The ability to resolve the interfacial dynamics of a heterogeneous catalytic reaction will be an important tool for applying theoretical calculations to chemical reactions and for designing artificial photosynthetic systems that take advantage of these dynamics. Transient optical spectroscopy at the n-type SrTiO₃/water interface reveals the picosecond to nanosecond kinetics of the initial hole transfer of the water oxidation reaction and its activation barrier as a function of surface potential. On-going ultrafast infrared studies relate the disappearance of the hole in the semiconductor with the transformation of surface adsorbed species. Another study focuses on catalysts with a low over-potential for the water oxidation reaction, such as Co₃O₄ and IrO₂, where the reaction is initiated by a unique photodiode configuration that allows sensitivity to interfacial hole transfer rates. These interfacial hole transfer rates are being studied as a function of the electrolyte interface and pH at open circuit, as well as light intensity and voltage during ongoing O₂ evolution.

The solar-driven photocatalytic splitting of water into hydrogen and oxygen is a potential source of clean and renewable fuels. However, three decades of global research have proven this multi-step reaction to be highly challenging. Perhaps it can be attributed, in part, to our limited understanding of the dynamics and mechanism of the different photo-induced charge transfer processes.
Here I will describe unique quantum dot systems that enable both manipulation and probing of photo-induced charge carriers within the nanocrystal. I will demonstrate our approach for disentangling the separate rate determining steps in each particle across the very wide range of time scales, which follow the initial light absorption process. Emphasis will be given to techniques for in-situ measurements of photocatalytic reactions on the single particle level.
Finally, I will point to how this information might benefit future design of effective artificial photocatalytic systems for renewable direct solar-to-fuel energy conversion.

Three-dimensionally (3D) structured photoelectrodes offer several advantages for achieving increased efficiency for solar-driven water splitting, including higher surface area for catalysis, enhanced optical absorption, improved carrier collection, and more facile removal of product gases. However, the benefits of 3D structuring are often accompanied by detrimental influences such as reduced photovoltage and high surface recombination. Understanding the structure-property relations that give rise to these positive and negative influences of 3D structuring is thus of great importance for optimizing efficiency in photoelectrochemical (PEC) devices. Unfortunately, complex and intertwined structure-property relationships in 3D photoelectrodes often make this task very difficult. In order to systematically study these structure-property relationships, we have fabricated well-defined 3D-structured p-Si photocathodes possessing a metal-insulator-semiconductor (MIS) architecture that has shown great promise for achieving high efficiency and good stability.^1-3. By using a combination of electroanalytical, spectroscopic, modeling, and solid-state measurement techniques and tools, we demonstrate a systematic approach for deconvoluting the influences of 3D structure on PEC performance of these photoelectrodes. Particular focus is placed on in situ micro-scale analysis of electronic properties with scanning photocurrent microscopy to give local information about carrier collection and photocurrent generation. By combining these measurements with modeling, we compare the optical, catalytic, ohmic, and recombination losses 3D and planar structured MIS photocathodes. The results are summarized in current-voltage loss-analysis diagrams, highlighting the challenges and opportunities for using 3D structuring in PEC devices for water splitting.
References
[1] Munoz, A.G. et al.,. Journal of the Electrochemical Society 156 (9), D337-D342 (2009).
[2] Y. W. Chen, J. D. Prange, S. Duhnen, Y. Park, M. Gunji, C. E. D. Chidsey and P. C., Nature Materials, 2011, 10, 539-544.
[3] Esposito, D.V., Levin, I., Moffat, T.P., & Talin, A.A.,. Nature Materials 12 (6), 562-568 (2013).

Dilute III-V alloys have garnered immense interest as suitable materials for solar hydrogen generation due to their compositional tunable band gaps, high carrier mobilities and high absorption coefficients. Till date there are no III-V materials that can satisfy all the stringent criteria for photoelectrochemical water splitting. In this study, we report two new III-V materials GaSbN and GaSbP that addresses current materials bottleneck for photoelectrochemical water splitting. Applicability of the Ga(Sb_x)N_1minus;x alloys for practical realization of photoelectrochemical water splitting is investigated using first-principles density functional theory incorporating the local density approximation and generalized gradient approximation plus the Hubbard U parameter formalism. Prior results with calculations revealed that a relatively small concentration of Sb impurities is sufficient to achieve a significant narrowing of the band gap, enabling absorption of visible light.¹ Theoretical results predict that Ga(Sb_x)N_{1minus;x ,} GaSb_xP_1-x alloys with 2 eV band gaps straddle the potential window at moderate to low pH values, thus indicating that dilute Ga(Sb_x)N_1minus;x alloys could be potential candidates for splitting water under visible light irradiation. Theoretical computations with Sb composition beyond 7% change the electronic band gap from direct to indirect. In the case of GaSbP a small amount of Sb incorporation will change the indirect GaP band gap to direct band gap.
Experimental synthesis is carried out using metal organic chemical vapor deposition method. Results confirm that severe band gap reduction occurs with incorporation of antimony in to GaN as predicted by the theoretical calculations.^2.3 As predicted from DFT calculations, slight incorporation of antimony into gallium phosphide slightly reduced the band gap of gallium phosphide and changed the band gap transition from indirect to direct.⁴ This presentation will highlight our results with both synthesis and photoelectrochemical characterization of GaSbN and GaSbP alloys.
Acknowledgements: Financial support from US Department of Energy (DE-FG02-07ER46375) and NSF (DMS1125909).
1. R.M. Sheetz, E. Richter, A.N. Andriotis, C. Pendyala, M.K. Sunkara and M. Menon, “Visible light absorption and large band gap bowing in dilute alloys of gallium nitride with antimony”, Phys. Rev. B, 84, 075304 (2011)
2. S. Sunkara, V.K. Vendra, J.B. Jasinski, T. Deutsch, A.N. Andriotis, K. Rajan, M. Menon and M.K. Sunkara, "New Visible Light Absorbing Materials for Solar Fuels, Ga(Sbx)N1-x”, Adv. Mater., 26 (18), 2878-2882 (2014)
3. S. Sunkara et al., “Band Gap Engineering of Gallium Antimony Nitride Alloy Nanowires and Photoelectrochemical Properties”, To be submitted (2014).
4. H. Russell et al., “Indirect to Direct Transition for Gallium Phosphide with Alloying Antimony”, To be submitted (2014).

There are a limited number of earth-abundant materials known to satisfy the demands required of the photoabsorber for practical photoelectrochemical (PEC) water oxidation. In an effort to exapnd this palate we report the growth and characterization of an uncommon bixbyite phase of iron(III) oxide (β-Fe₂O₃) epitaxially stabilized via atomic layer deposition on an conductive, transparent, and isomorphic template (Sn-doped In₂O₃). β-Fe₂O₃ is reported to possess a bandgap (Eg) as low as 1.7 eV, suggesting the possibility that it could provide superior solar absorption as compared to α-Fe₂O₃ (Eg = 1.9-2.2 eV) and therefore be more suitable OER photoanode. This smaller bandgap makes β-Fe₂O₃ an outstanding candidate for use in tandem photoelectrochemical cells, for which recent modelling studies predict will optimally require a 1.7 eV bandgap top cell. Despite these promising signs, β-Fe₂O₃ has not previously been considered for PEC water splitting applications largely due to its synthetic elusivity. We find that as a photoanode, unoptimized β-Fe₂O₃ ultrathin films perform similarly to their ubiquitous α-Fe₂O₃ (hematite) counterpart, but reveal a more suitable bandgap (1.8 eV), a ~0.1 V improved onset potential, and longer wavelength (>600 nm) spectral response. Stable operation under basic water oxidation justifies further exploration of this atypical phase and motivates the investigation of other unexplored metastable phases as new PEC materials.

Developing photocatalysts capable of fully harvesting photons in the whole visible light range (400-700 nm) is indispensible for realizing highly efficient photocatalytic solar energy conversion. These photocatalysts typically have a red color. Narrowing the bandgap of wide-bandgap photocatalysts and exploring red materials are two aspects of obtaining red photocatalysts. In the past years, many impressive red photocatalysts have been investigated and demonstrated encouraging photocatalytic activities. This talk will mainly introduce our efforts in narrowing the bandgap of anatase TiO₂ and g-C₃N₄ for fully absorbing red light, utilizing Ta₃N₅ photoanodes, and exploring some unknown red photocatalysts. By forming a gradient B/N codoping in the around 50 nm thick shell of anatase microspheres, a red anatase TiO₂ with strong absorbance in the whole visible light range was obtained. The photoanode of the red TiO₂ has the photoelectrochemical water splitting ability with the irradiation to ca. 700 nm.
References:
[1] G. Liu, L.-C. Yin, J. Q. Wang, P. Niu, C. Zhen, Y. P. Xie, H.-M. Cheng, A red anatase TiO₂ photocatalyst for solar energy conversion, Energy & Environmental Science 5 (11), 9603, (2012).
[2] G. Liu, P. Niu, H. M. Cheng, Visible-light-active elemental photocatalysts, ChemPhysChem, 14 (5), 885, (2013)..
[3] G. Liu, L. C. Yin, P. Niu, W. Jiao, H. M. Cheng, Angewandte Chemie International Edition 52 (24), 6242, (2013).
[4] Y. Q. Yang, C. H. Sun, L. Z. Wang, Z. Liu, G. Liu, H. M. Cheng, Adv Energy Mater 2014, 4, 1400057.
[5] Y. P. Xie, Z. B. Yu, G. Liu, X. L. Ma, H. M. Cheng, Energy Environ Sci 2014, 7, 1895.
[6] P. Niu, G. Liu, L. C. Yin, H. M. Cheng, Advanced Materials 2014, In press
[7] C. Zhen, G. Liu, H. M. Cheng, Submitted.
[8] G. Liu, L. C. Yin, P. Niu, W. Jiao, H. M. Cheng, Submitted.

Nickel (II) oxide (NiO) is a p-type semiconductor with applications as an electrocatalyst for water oxidation and as a cocatalyst for water splitting photocatalysis. Here we show that NiO nanocrystals independently exhibit photocatalytic water reduction properties under band gap excitation. Photoelectrochemical and surface photovoltage scans reveal that the p-type character of the NiO nanocrystals is enhanced, likely due to surface oxidation to Ni₂O₃, and that transfer of photoexcited electrons from NiO to Pt occurs under band gap illumination.

Tantalum oxide and many tantalite-based systems have been reported to show extraordinarily high activities and quantum yields when decomposing water under ultraviolet (UV) illumination [1]. Although pure tantalum oxide shows some photocatalytic activity, loading with a nickel-based co-catalyst improve the initial H₂ production rate by 3 orders of magnitude and result in stoichiometric decomposition of pure water into H₂ and O₂ [2]. Here we investigate structure-reactivity relations in these photocatalysts to explore both activation and de-activation mechanisms.
A series of controlled Ni-NiO core-shell co-catalyst structures were prepared on a Ta₂O₅ substrate by tuning the heat treatment conditions. The photocatalytic activity of each structure was tested in pure liquid water in a photoreactor system interfaced to a gas chromatography (GC) for H₂ detection. Atomic level transmission electron microscopy (TEM) was employed to study the differences in the initial co-catalyst structures, as well as the evolution of the materials after exposure to UV light and water.
Changes in the core-shell morphologies resulted in large changes in H₂ production rates and deactivation time. Increased H₂ production was found to be related to an increase in the thickness of NiO shell due to suppression of the reverse reaction. TEM images showed that the core-shell co-catalyst structures deactivated primarily due to a loss of metallic Ni from the core structure. During deactivation, the catalyst transformed either to structures consisting of NiO nanoblocks or hollow NiO shells. The phase transformations occurring during deactivation were associated with Ni diffusion processes that are driven by light illumination. The information we have gained by correlating microstructures with reactivities is now being employed to design improved co-catalysts structures.
Acknowledgement:
The support from US Department of Energy (DE-SC0004954) and the use of TEM at John M. Cowley Center for High Resolution Microscopy at Arizona State University is gratefully acknowledged.
References:
[1] A. Kudo, Y. Miseki, Chem. Soc. Rev. 38 (2009) 253-278
[2] H. Kato, A. Kudo, Chem. Phys. Lett. 295 (1998) 487-492

Cu₂ZnSnS₄ (CZTS) compound has attracted intense attention as a light absorbing material due to not only large optical absorption coefficient of around 10⁵ cm^-1, but also earth-abundance of constituent elements. With rapid development of nanotechnology, one-dimensional nanostructures of CZTS compound have been investigated as a potential form to achieve high efficiency because the nanostructures are expected to be capable of capturing more light and enhancing charge separation and transport. The reported nanostructures of CZTS materials include single/poly crystalline nanowires and nanotube which were synthesized by using conventional anodic aluminum oxide (AAO) as a hard template. However, the previously reported nanowires and nanotube were too long (~50 mu;m) to be applied to practical one-dimensional devices with vertically-aligned geometry because long nanowire length suffers from inefficient carrier collection due to the short carrier diffusion length (~1 mu;m) of CZTS materials.
Here, we report a well-controlled fabrication route for vertically-aligned CZTS nanorod arrays via simple sol-gel process. To achieve vertically-aligned nanostructure with controlled wire length, our approach relies on an AAO template assisted growth-and-transfer method, consisted of aluminum foil iodization, nanowire growth, and array transfer to carrier collector/ substrate, followed by removal of a template. For the ink preparation, the Cu, Zn, Sn precursors and thiourea are dissolved in 2-methoxyethanol from which the AAO template was infiltrated. The structure, morphology, composition, and optical absorption properties of the resulting CZTS nanorod samples were characterized using X-ray diffraction (XRD), Raman spectroscopy, transmission electron microscopy (TEM), energy dispersive X-ray spectrometry (EDS), scanning electron microscopy (SEM) and UV-vis spectroscopy. The photoelectrochemical (PEC) properties of CZTS nanorod arrays were also characterized by three-electrode configuration using an Ag/AgCl reference electrode and a Pt wire counter electrode to demonstrate their feasibility as low cost photocathode material. The PEC performance could be improved by surface modification with Pt catalyst. The CZTS nanorod photocathode revealed higher photocurrent density than the thin film counterpart.

Today, demands for clean and sustainable energy is increasing. Among several alternative energy sources, solar energy is expected to support world energy consumption. Solar water splitting is one of the solar-based energy generation method. This process produces exciton from the material-water interface by solar light that is finally converted hydrogen gas or photocurrent. In other word, the produced energy can be stored as photoelectrochemical (PEC) cell.
There are a few considerable materials for water splitting from metal oxides to polymers. ZnO is selected as main material due to its electrical property. ZnO is n-type semiconductor material which has large (=3.3 eV) and direct band gap. Moreover, it can be easily synthesized through hydrothermal growth. Nevertheless, there is one critical issue on ZnO. Owing to its improper band gap position, it can only absorb UV light that exists only 3% of sunlight on the ground.
Therefore, several efforts are tried to reduce the limitation and to extend absorbing wavelength region. For example, deposition of noble metals such as Pt, IrO2, Au and Ag on the surface of the water splitting material by RF sputtering, LPCVD, ALD and other vacuum technique are presented for the catalyst or trigger of plasmon effect that eventually result in increased efficiency. However, those vacuum based methods are required prolonged working time, cost and high temperature.
In order to meet facile synthesis way, in this study, photoreduction process is introduced to produce metal nanoparticles for plasmon effect. In brief, silver nanoparticles are synthesized by UV irradiation for only several minutes on the ZnO nanowire arrays. Consequently, ZnO/Ag hierarchical structure is produced. For the proof of this concept, 3-electrode method is employed for measuring photocurrent. Likewise, plasmon enhanced effect is observed by various analytic tools

We report a facile method to fabricate three-dimensional RuO₂ branched Au-TiO₂ nanowire heterostructures and their application as the photostable electrode for efficient oxygen generation, due to the high catalytic properties of the RuO₂ for the water oxidation and the plasmonic absorption with efficient charge separation to TiO₂ and Au interface. The RuO₂ nanobranches are produced on the Au-TiO₂ nanowires by a thermal annealing process at relatively low temperature of ~ 180 ^oC. The real time structural evolution during the growth verified the direct transition of the amorphous phase of RuO₂ precursors to crystalline RuO₂ nanoparticles with the tetragonal crystallographic structure at 180 ^oC, without the intermediate change formation, supported by the nucleation theory. The RuO₂ nanobranches, highly crystalline, are then grown on the crystalline nanoparticles due to the surface diffusion of the Ru, followed by the rapid oxidation in O₂ ambient. The RuO₂ branched Au-TiO₂ nanowire arrays shows a remarkable enhancement in the photocurrent density by approximately 60 % and 200 %, in the UV-visible and Visible region, respectively, compared with pristine TiO₂ nanowires. Furthermore, there is no significant decrease in the device&’s photoconductance with UV-visible illumination during 1 day, making it possible to produce oxygen gas without the loss of the photoactvity.

Iron oxide (hematite) is a promising material to convert solar energy directly into chemical energy. Hematite is an abundance and non-toxic semiconductor which possesses an optimal band gap for visible light absorption. However, its electronic conductivity properties hinder its performance in photoelectrochemical cells. This work describes the influence of defects on hematite surface caused by different atmosphere (N₂ and O₂) of thermal treatment. The impacts of those defects in the conductivity, morphology and photoelectrochemical response of the hematite films were evaluated. Hematite films were deposited on a conductive substrate (FTO) by microwave-assisted hydrothermal process. The hematite films were heat treated at 750°C for 30 minutes in two different atmospheres with a gas flux (O₂ and N₂). The thickness was estimated to be 32-38 nm using the absorbance spectra analysis. The images of atomic force microscopy (AFM) and scanning electron microscopy (SEM) indicate the morphological no significant changes occurred on the films surface. A study of the wettability was carried out and a similar interaction of the both films with the electrolyte was found. The films treated in nitrogen and oxygen gas flux showed respectively a photocurrent of 0.30 and 0.15 mA cm^-2 at 1.23 V_RHE, presenting a good stability for the film treated in N₂. The carrier density were estimated from Mott-Schottky plot around 1,03.10²⁰ cm^-3 (N₂) and 1,54.10¹⁹ cm^-3 (O₂), indicating a strict relation with the improvement on the conductivity, since there is no presence of effective dopants. Finally, further investigations are necessary to better understand the changes caused by the presence of defects on hematite surface.
Acknowledgements
We gratefully acknowledge #64257;nancial support from the Brazilian agencies of FAPESP (Grants 2011/19924-2 and 2012/19926-8), CAPES, CNPq (Grants no. 473669/2012-9), Instituto Nacional em Eletrocirc;nica Orgacirc;nica (INEO), NanoBioMed Brazil Network (CAPES), and and CDMF (Grants no. 2013/07296-2).

Iron oxide is one of the most promising semiconductors for application as photoanode in photoelectrochemical cells. A simple and cheap route to prepare hematite photoelectrodes is the sol-gel method. Hematite thin films can be also prepared using the polymerized complex (PC) method that is a sol-gel derived technique. This methodology involves the formation of complex of metal ions that then undergoes polymerization. In addition, the PC method allows optimal control of stoichiometry and incorporation of impurities during the process. In this work, pure and doped hematite thin films were prepared at two different heat treatments (500°C and 800°C) using the PC method. The α-Fe₂O₃ thin films were modified using Zn²⁺ and Sn⁴⁺ ions as dopants and its photoelectrochemical properties were studied in comparison with pure hematite films. Top-view scanning electron microscopy images of the films revealed that the hematite films prepared at 500°C exhibited a smooth surface and those prepared at 800°C presented a porous surface. Thus, the higher surface area of the α-Fe₂O₃ films treated at 800°C influenced positively on its photocatalytic performance. Furthermore, the modification of α-Fe₂O₃ with Zn²⁺ and Sn⁴⁺ ions resulted in a better photoresponse and stability as showed by the linear voltammetry and chronoamperometry results.
Acknowledgements
We gratefully acknowledge #64257;nancial support from the Brazilian agencies of FAPESP (Grants 2011/19924-2 and 2012/19926-8), CAPES, CNPq (Grants no. 473669/2012-9), Instituto Nacional em Eletrocirc;nica Orgacirc;nica (INEO), NanoBioMed Brazil Network (CAPES), and and CDMF (Grants no. 2013/07296-2).

Photoelectrochemical cells (PECs) make it possible to transform solar energy directly into chemical energy via water splitting through the interaction of water and sunlight in the presence of a semiconductor. Among the suitable materials, tungsten oxide (WO₃), which has a wide range of technological applications, has been widely studied recently for utilization as working electrode in PECs. In this work, we synthesized WO₃ nanoparticles via aqueous route using two different methods and subsequently deposited the materials obtained on FTO (SnO₂:F) either by the doctor&’s blade technique or by spin deposition. The deposition step was followed by heat treatment (450 0C, 3h) for all films. The films produced by each method were analyzed by X-ray diffraction and their crystallographic arrangement was identified using the JCPDS catalog. Their electrochemical performance and stability were analyzed in a conventional electrochemical cell using the produced films as working electrode. Using this same system, it was possible to determine the number of carriers (N_D) of approximately 8,3.10¹⁹ cm^-3 and the flat band potential (V_fb) of approximately 0,15 V for each film. The good photoelectrochemical activity of the films was attributed to the tuning of their structural, electronic and catalytic properties.
Acknowledgements
We gratefully acknowledge #64257;nancial support from the Brazilian agencies of FAPESP (Grants 2011/19924-2, 2012/19926-8 and 2013/05471-1), CAPES, CNPq (Grants no. 473669/2012-9), Instituto Nacional em Eletro#770;nica Orga#770;nica (INEO), NanoBioMed Brazil Network (CAPES), CEM-UFABC and CDMF (Grants no. 2013/07296-2).

In the past few decades, most efforts have been focused on developing binary metal oxides as suitable materials for photoelectrode in solar water splitting. Nevertheless, only few candidates have shown promising photocurrents (> 2 mA/cm²), such as Fe₂O₃, WO₃ and Cu₂O. This limited number of options has generated large interest in the development of a new generation of photoelectrode materials based on multinary metal oxides (ternary or quarternary), giving a huge number of additional possibilities. BiVO₄ is an example of a ternary metal oxide which is gaining a large amount of interest in the past couple of years, and AM1.5 photocurrents exceeding 4 mA/cm² have been reported. This material, however, has a bandgap of 2.4 eV, which means that it is theoretically impossible to reach solar-to-hydrogen (STH) efficiency of more than 9% with AM1.5 illumination. Further exploration of other ternary metal oxides that do not share the same limitation is therefore necessary.
In this work, we explore the suitability of iron tungstate as a photoanode material. This material is relatively unexplored, as shown by less than 10 existing publications on this material—only one of which is related to photocatalysis [1-3]. We have developed a spray pyrolysis recipe for depositing thin films of n-type orthorhombic Fe₂WO₆ at a relatively low temperature of 500 ^oC. UV-Vis measurements reveal that the direct and indirect bandgap values for Fe₂WO₆ are 2.3 eV and 1.6 eV, respectively. This means that an AM1.5 photocurrent of ~24 mA/cm² and an STH efficiency of up to 30% can be potentially achieved. However, we observe much lower photocurrents for our films (~0.1 mA/cm²) due to extensive carrier recombination. This recombination appears to be related to the high concentration of defects within the material, as suggested by the rather high free carrier concentration of ~1.7 x 10²⁰ cm^-3. Heat treatment at 800 ^oC in air reduces the carrier concentration by one order of magnitude and significantly improves the photocurrent. Results from intensity modulated photocurrent spectroscopy (IMPS) measurements suggest that this improvement is accompanied by suppression of surface recombination. Finally, detailed photoelectrochemical analysis reveals that the main performance limitation of this material is the very low carrier separation efficiency (<10%). Future efforts should then be directed towards overcoming this limitation, through nanostructuring and/or doping strategies.
References
1. Khader et al. J. Solid State Chem. 2 (1998) 170
2. Bharati et al. J. Mater. Sci. 16 (1981) 511
3. Walczak et al. J. Mater. Sci. 27 (1992) 3680

One of the major challenges in photoelectrochemical water splitting is the development of highly efficient, stable, and low-cost photoanodes for water oxidation. The most efficient photoanodes reported up to now were based on rare elements and/or turned out to be unstable under operational conditions. While passivation of conventional photovoltaic materials like silicon seems to be a promising strategy, the development of low-cost metal oxide photoanodes still attracts a great interest. Oxide photoanods mostly investigated in recent years were typically based on stable binary oxides like Fe₂O₃ or WO₃ which suffer from disadvantageous electronic (short diffusion length, low conductivity) and photoelectrochemical (too positive positions of band edges) properties having detrimental effect on overall photoconversion efficiency. The interest has therefore turned to ternary oxides which offer much larger variations of composition, promising the possibility of better tuning of optical and electronic properties. Thus, for example, impressive photoconversion efficiencies - far exceeding those obtained at binary oxides - have been recently obtained with photoanodes based on BiVO₄, a semiconductor with strong visible light absorption [1].
We have been recently investigating nanostructured photoanodes based on another low-cost ternary oxide, CuWO₄ prepared by sol-gel technique. This material is very attractive due to following reasons [2-4]: (i) bandgap of 2.2 eV (~564 nm); (ii) excellent photochemical stability in aqueous electrolytes with optimized chemical composition [4]; (iii) high surface catalytic activity towards oxygen evolution [4]. However, the reports available on this material so far have also identified the key obstacles hampering the efficiency of CuWO₄-based photoanodes: very short hole diffusion length, and very bad electron transport properties [2,4]. Our different experimental strategies (doping, compositing with large bandgap electron collectors, etc.) to address these limitations will be presented.
References
[1] F. F. Abdi, L. Han, A. H. M. Smets, M. Zeman, B. Dam, R. van de Krol,
Nat. Commun. 4, 2195 (2013).
[2] Chang, Y.; Braun, A.; Deangelis, A.; Kaneshiro, J.; Gaillard, N. J. Phys. Chem. C 115, 25490 (2011).
[3 Yourey, J. E.; Bartlett, B. M. J. Mater. Chem. 21, 7651 (2011).
[4] Yourey, J. E.; Pyper, K. J.; Kurtz, J. B.; Bartlett, B. M. J. Phys. Chem. C 117, 8708 (2013).

Zinc Oxide (ZnO) with wurtzite crystal structures show piezoelectric properties, and have been widely studied to fabricate nanogenerators (NG). Recently, ZnO have been studied for hydrogen evolution by piezo-driven water splitting without light illumination. However, most studies have been utilized mechanical distortion of ZnO to generate temporal piezoelectric potential. Here, we reported photoelectrochemical (PEC) water splitting by using Li-doped ZnO nanowires (NW) as a PEC cell. Corona poling process on Li-doped ZnO NW induced permanent polarization in the ZnO NW, and induced favorable band bending to direction for favorable carrier transfer to electrolytes. Interestingly, PEC performance and stability were strongly depending on the direction of polarization. Furthermore, poled NW exhibited high stability and steady light response in acidic electrolyte.

Solar hydrogen generation by water splitting in photo-electrochemical cells (PEC) belongs to the holy grail of sustainable energy supply. The search for high-performing, affordable, and corrosion-resistant photo-electrode materials is an on-going quest. Hematite, a-Fe2O3 is in many respects a promising solution, particularly because it is an abundant and low-cost material with a band gap that makes it operational even for visible-light applications. But in spite of decades of extensive research, hematite still suffers from substantial deficiencies of conductivity. There is a controversy of whether this is based on the premature recombination of electron-hole pairs in the bulk or on the recombination of charge carriers at the surface. Lack of information of the electronic structure of typical PEC electrode materials for bulk and surface may be one reason why progress in the field is lacking.
Our recent combination of electrochemical and x-ray spectroscopy methods has provided new insight in the interaction of DC bias, electrolyte and photo-excitation with hematite. Electrochemical oxidation of sol-gel derived hematite photo-anodes creates a new, stable and hitherto unknown surface state in the upper Hubbard band, as evidenced by O1s near edge x-ray absorption fine structure (NEXAFS) spectra [1]. Anodization of creates at least two new pre-edge structures in the O1s NEXAFS spectra, which can be assigned to e_g spin up symmetry transitions in the O2p charge transfer band and in the Fe3d-type upper Hubbard band. Upon further analyses, we find a parallel evolution of the relative spectral weight of these transitions with the photocurrent vs. DC bias, strongly supporting that two types of electron holes are contributing to the historically suggested two types of photocurrent in hematite [2]. The fact that our experiment was carried out entirely in-situ with exposure of the electrode to visible light (or dark) and electrolyte (or dry) during x-ray spectroscopy strongly supports previous speculations that the electronic structure of the surface, and not only the short hole diffusion length in the bulk, may impede the performance of hematite photo-anodes. Our in-situ ambient-pressure and Fe-resonant XPS valence band studies on hematite during oxidation suggest that the charge transfer satellites frequently visible in Fe 2p NEXAFS spectra are indeed active in the valence band right near the Fermi energy.
[1] Bora et al., JPCC 2011, 115, 5619.
[2] Braun et al., JPCC 2012, 116, 16870.
[3] Gajda-Schrantz et al., PCCP 2013,15, 1443
[4] Braun et al., ChemPhysChem 2012, 13, 2937.

We report a near-infrared-driven photoelectrochemical water splitting using ZnO nanorod-array decorated with CdTe quantum dots and plasmon-enhanced upconversion nanoparticles (UCN). The plasmon enhanced the intensity of the upconversion emission, which improved the photocurrent and the gas evolution rate of the photoelectrochemical reaction greatly. We will demonstrate a process of utilizing near infrared (NIR) to drive the photoelectrochemical water splitting reaction. The Au-UCNs can significantly enhance the upconversion emission intensity by plasmonic effect. Thus, the enhancement of photocurrent and gas evolution was achieved. The results offer a convincing demonstration that energy can be converted from NIR to chemical fuel. We believe that our strategy is fundamental to the design of solar energy devices and should become an accepted technique for solar energy utilization at the NIR and IR regions.

There have been numerous efforts over recent years to incorporate plasmonic metal nanostructures into photoelectrochemical systems, either semiconductor suspensions or, less frequently, thin film semiconductor photoelectrodes [1]. Although significant enhancements in photon-to-current conversion efficiencies have been reported for some water splitting photoanodes [2], they generally did not translate into large net photocurrents. N-type tungsten trioxide, WO₃, is one of very few semiconductor materials that combine visible light absorption (E_g = 2.5 eV) with remarkable long-term stability as photoanode during water photoelectrolyses performed in appropriate acidic solutions [3]. Although its band energetics, with the open-circuit photopotential of ca 0.45 V vs RHE, does not allow overall water splitting, WO₃ is well suited for the application in a tandem device, combined with a photovoltaic, PV, cell capturing longer solar light wavelengths. One of main limitations to efficient operation of such a device are low visible light absorption coefficients of WO₃ related to its indirect optical transition. To address this issue we are exploring incorporation to the photoanode of various plasmonic metal nanoparticles. Recent results demonstrate the critical importance of combining the metal NPs with capping agents to limit charge carriers recombination at the semiconductor surface [4].
Acknowledgement: This work is supported by the Polish-Swiss Research Program and by the funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement SOLAROGENIX N° NMP4-SL-2012-3100333.
References:
[1]. W. Hou, S. B. Cronin, Adv. Funct. Mater. 2013, 23, 1612-1619.
[2]. S. Linic, P. Christopher, D. B. Ingram , Nature Mater.2011, 10, 911-921.
[3]. R. Solarska, R. Jurczakowski, J. Augustynski, Nanoscale2012, 4, 1553-1556.
[4]. R. Solarska, K. Bienkowski, S. Zoladek, A. Majcher, T. Stefaniuk, P. J. Kulesza and J. Augustynski, Angew. Chem. Int. Ed. 2014, 53: DOI: 10.1002/anie.201408374.

This presentation deals with the mechanism of plasmon-enhanced solar-to-fuel conversion by inorganic semiconductor. It will discuss the hot electron injection from a plasmonic metal to a semiconductor. It will introduce the plasmon-induced resonant energy transfer (PIRET) from a plasmonic metal to a semiconductor. In addition, this talk will show the effort on the development of metal-semiconductor photocatalysts and photoelectrodes for solar fuel generation.

Two photoelectrode systems were individually optimized for the process of solar-driven overall water splitting in acidic electrolytes: 1) phase-pure Pt-doped anatase TiO₂, prepared by Ion Layer Gas Reaction (ILGAR), was employed as an electrocatalytic protection layer on device-grade p-type Cu(In,Ga)Se₂. It will be shown that the gradual increase of the Pt-concentration within the TiO₂-layer passes through an efficiency- and stability-maximum of the electrode. At this maximum, Incident Photon-to-Current Efficiencies (IPCE) of 80% could be achieved over the full range of the solar spectrum (AM1.5). Photocurrent densities reached more than 12.5 mAcm^-2 at the thermodynamic potential for H₂-evolution in 0.5 M H₂SO₄. Further improvements of the heterojunction, realized by atomic layer deposition (ALD), will be discussed. 2) Nanoparticulate RuO₂ was deposited on n-type Si by electrophoretic deposition. During deposition, an ultra-thin SiO₂ film was simultaneously formed that acts as protection layer. Oxygen evolution photocurrent densities of up to 12 mAcm^-2 could be achieved at 1.23 V vs RHE in 0.5M H₂SO₄. Both photoelectrodes maintained photocurrent densities of 12 mAcm^-2 for more than 25 hours, demonstrating the promising stability offered by these novel Pt-TiO₂ and SiO₂/RuO₂ protection layers. The film and interface properties were investigated with photoelectron spectroscopy, electron microscopy and surface photovoltage measurements, offering new insights in the structure of electronic defects in the Pt-doped TiO₂ layer and at the Si/RuO₂ interface. Finally, device development in combination with a two-junction tandem solar cell will be discussed.

Hematite (iron oxide) and tungsten oxide are two oxides of interest as photoanode material for photoelectrochemical water splitting. Implementing these oxides in a heterojunction proved to be a succesfull strategy to improve charge diffusion in those materials. Nevertheless the efficiency of such photoanode is still hindered by the low conductivity of semiconducting metal oxides. To solve this limitation, light management using plasmonic or scattering nano to micro-structures has raised interest in recent years. Photonic effects can be tailored to confine light inside the film and increase the propagation length inside the material. Light absorption becomes independant of the film thickness, and thinner films with higher conductivities can be implemented.
By using a solution based processing, I was able to develop photoanodes composed of tungsten oxide microspheroids coated with a hematite ultra-thin film overlay. By tuning the spheroids dimension, in the micrometric to submicrometric range, different photonic effects were observed experimentally. Finite-difference time-domain simulations of light propagation inside the film microstructure allowed a quantitative analysis of the different photonic regimes observed experimentally. Both experimental and mathematical studies lead to a better understanding of photonic effects triggered by the spheroid-like microstructure and their influence on the oxides photoactivity.

Photo-catalytic hydrogen production from water has been pursued for decades and hundreds of catalysts have been designed and tested. Among the many catalysts that have been studied, TiO₂ is known to be the most stable semiconductor for hydrogen production. However, due to its wide band-gap energy (~3.2 eV), it absorbs light only in the UV range. A major challenge is to extend its light absorption range into the visible region and hence increase the rate of energy transfer. In this presentation, we describe a method to extend the light absorption range of TiO₂ by coupling it with Au and Cu₂O nanocrystals. We report the photo-catalytic hydrogen production from water over colloidally fabricated catalysts composed of Au-Cu₂O nanoparticle-nanocrystal composites deposited on TiO₂ (anatase) in order to exploit the plasmon resonance (Au nanoparticles) and the pn-junction (Cu₂O-TiO₂) effects. The Au-Cu₂O composite is comprised of ~8 nm Au nanoparticles grown on ~480 nm Cu₂O octahedral nanocrystals by partial galvanic replacement. The Au-Cu₂O composite was then coupled with TiO₂ nanoparticles and tested for hydrogen production from water. The rate of hydrogen production over 2wt.% Au/Cu₂O-TiO₂ was found to be slightly higher than 2wt.% Cu₂O-TiO₂, ~10 times that of 2wt.% Au-TiO₂ and ~30 times that of TiO₂ alone. Characterization of the catalyst using X-ray photoelectron spectroscopy, HR-TEM, and Raman spectroscopy reveal that in situ reduction of Cu⁺¹ into Cu⁰ occurs upon photoreaction. Remarkably, we find that after this reduction the catalyst systems becomes active under visible light and exhibits prolonged stability, with very little signs of deactivation. Our studies indicate that Cu⁰ takes part in the reaction probably by providing additional sites that act as recombination centers for hydrogen atoms to form molecular hydrogen. Our work highlights a previously unexplored photo-catalysis mechanism in Cu₂O-TiO₂ systems and the effect of metals in augmenting their efficiency at water splitting.

Fossil fuels are being depleted rapidly and solar energy conversion is a very promising path to fill the energy gap. Solar energy can either be converted to electricity by solar cells, or to solar fuels by photoelectrochemical cells. For both applications, it is highly advantageous to use one-dimensional nanostructures as the photoactive material due to their high surface-to-volume ratio. In addition, the physical properties of most semiconductors (absorption depth and charge carrier diffusion length) demand the use of one-dimensional nanostructures to improve the transport of photogenerated charge carriers to the nanowire surface while maintaining a sufficient film thickness for optimal photon absorption.
On the search for more efficient photocatalysts and photoelectrodes, we present the development and photocatalytic/photoelectrochemical characterization of several Cu₂O nanowire-based systems: (1) axially segmented p|n-Cu₂O nanowires dispersed in an aqueous solution for autonomous photocatalytic water splitting, and (2) interconnected p-Cu₂O nanowire networks for photoelectrochemical water splitting. In the former configuration, electrodeposition in polycarbonate membranes with parallel cylindrical pores is applied to combine p- and n-type Cu₂O along the nanowire length. After dispersing these wires in an electrolyte, formation of H₂ gas is demonstrated without the application of an external bias. Using this configuration, we investigated the influence of the number of p- and n-type segments on the photocatalytic efficiency. In the latter configuration, electrodeposition is performed inside polycarbonate membranes with interconnected pores fabricated by ion-track technology at the GSI linear accelerator (UNILAC) in Darmstadt to obtain three-dimensional nanowire network structures consisting of interconnected one-dimensional nanowires. These nanowire networks are used to prepare photocathodes, to subsequently investigate the influence of nanowire parameters (i.e. diameter and length) on the photocurrent.

Copper oxide (Cu₂O) has been widely studied as p-type photocathode for water splitting (Eg= 2 eV with suitable position of band edges for proton reduction)^[1]. Unfortunately, it encounters a problem of stability in solution due to photocorrosion; thus, it has to be protected. Different ways are studied like covering electrodes with NiO_x^[2], or with CuO passivation layer^[3].
Our photoelectrodes are synthetized by electroplating and anodization onto Fluorine doped Tin Oxide (FTO) coated glass substrates. First, copper metal is grown by electroplating, which resembles bubbles on the surface, then a part of the copper metal is oxidized in Cu₂O and CuO (Eg= 1.5eV) which shapes peaks of CuO on the bubbles of Cu/Cu₂O. This morphology increases the specific area at the interface electrode/electrolyte. Preliminary test in electrochemistry exhibits high photocurrents in visible range (best without any catalyst: (-)2.5 mA/cm² at 0V vs RHE, pH 7); however, photocorrosion is still observed. The strategy is to combine these two previous p-type semiconductors with n-type semiconductor (e.g. TiO₂) to ensure photo-stability over time and to prevent recombination between photo-generated electrons and holes. A layer of TiO₂, synthetized via a sol-gel process, is deposited by dip-coating technique on the top of the photocathodes. Due to an appropriate band energy structure, the photo-generated electrons from Cu₂O are easily transferred to the conduction band of CuO, then to the conduction band of TiO₂, and eventually transferred straight to the interface electrode-electrolyte.
The morphology, crystal structure and the electrical properties of the photoelectrodes are respectively characterized by (field emission gun-) scanning electron microscopy, X-Ray diffraction and current-voltage measurement. The impact of these parameters onto the electrode performances will be discussed, as well.
[1] Hara,M. et al., Chem. Commun., 1998, 357-358.
[2] Lin, C-Y. et al., Chem. Sci., 2012, 3, 3482.
[3] Wang, P. et al., Nanoscale, 2013, 5, 2952.

Coupling suitable electrocatalysts with light-absorbing semiconductor structures has lead to significant improvements in the efficiency of state-of-the-art water splitting photoelectrodes, by reducing the overpotential required for water oxidation. Typical electrocatalysts such as Pt, RuO₂ and IrO₂ can reduce the oxygen evolution reaction (OER) overpotential of metal oxide-based photoelectrodes, albeit with a considerable increase in the overall cost of the devices. Cheaper, more abundant materials, on the other hand, offer a very promising alternative to precious metals. Cobalt oxide-based catalysts have been extensively explored in combination with a large number of metal-oxide semiconductors, producing 0.2 - 0.3 V shifts of the measured photocurrent onsets. More recently, nickel-based oxides have been identified as active earth abundant electrocatalysts for OER in alkaline media [1] with overpotentials comparable to those of Pt catalysts [2]. Coupling NiFeO_x with suitable semiconductor structures, such as hematite, has been shown to produce a cathodic shift of the photocurrent onset of -0.24 V [3]. A detailed understanding of the nature of the interaction between the semiconductor and electrocatalyst in such structures, however, is still lacking. In this paper, we combine photoelectrochemical characterization techniques and in-situ spectroscopic methods to investigate the effect of NiFeO_x on doped α-Fe₂O₃. The focus is on the role of structure and composition of the semiconductor and electrocatalyst, and the effect of these parameters on the charge carrier dynamics, particularly the charge transfer processes at the solid-solid and solid-liquid interfaces.
[1] Smith R.D.L. et al., J. Am. Chem. Soc., 135 ( 2013) 11580.
[2] Gong M. et al., Nat. Commun., 5 (2014) 4695.
[3] Wang D. et al., Angew. Chem. Int. Ed., 52 (2013) 1 - 5.

By mimicking one of nature's most fundamental processes - photosynthesis - the conversion of sunlight, water and carbon dioxide (CO₂) into useful fuels is one of the key sustainable technologies to provide storable solar energy and carbon-free environment. One of the main challenges in artificial photosynthesis is the production of stable H₂ and O₂ from water splitting in an efficient way. Therefore, high efficiency and stable water splitting under abundant visible solar spectrum has been intensively studied over the last four decades but with limited success. Recently, group-III nitrides (i.e., GaN, InGaN) have attracted considerable attention, as they possess the thermodynamic and kinetic potential requirements for water splitting under deep visible and near-infrared light irradiation [1-2]. Recently, we have demonstrated that by tuning the near surface Fermi-level on p-GaN nanowires, the quantum efficiency can be enhanced by nearly two orders of magnitude [3]. However, it has remained unknown if the same concept can be extended into In-containing visible light active nitrides wherein tuning of the Fermi-level has not yet been realized. In this context, we have studied the effect of p-type Mg dopants on the water splitting activity of dual-gap GaN/InGaN nanowire arrays for efficient water splitting under visible light. We show that InGaN, a widely used semiconductor for solid-state lighting and power electronics, can be transformed to be the most active photocatalyst under visible light irradiation by precisely engineering the surface band bending through controlled Mg-dopant incorporation. The absorbed photon conversion efficiency can be enhanced by more than 30 times and can reach ~69%, the highest value ever reported for one-step overall water splitting under visible light irradiation (up to 475 nm). Furthermore, in order to extend the absorption edge towards deep-visible (i.e., red) solar spectrum, novel hexagonal disk-on-wire InGaN/GaN nanostructures have been developed using plasma-assisted molecular beam epitaxy. Such unique nanostructures can be a potential material platform to study the thermodynamic and kinetic potentials of III-nitrides for deep-visible and near-infrared solar irradiation. Moreover, we will report on a detailed study of the water oxidation mechanism of III-nitride nanowires, which is commonly known to be the sluggish and rate limiting reaction on most of the metal-oxides for overall water splitting.

Corrosion of high-quality III-V semiconductors has been a major hurdle for photoelectrochemical water splitting. Conformal layers of thick TiO₂ (4 to 120 nanometers) have been shown to prevent corrosion of GaAs and other III-V materials in aqueous media. In this work, a high-efficiency dual-junction 1.9 eV InGaP / 1.42 eV GaAs device was shown to perform unassisted water splitting with 10.5% solar to hydrogen (STH) efficiency. An amorphous, electronically defective, “leaky” TiO₂ protection layer, grown by atomic layer deposition (ALD) and coated with Ni oxygen-evolving catalysts, was used to passivate the anodic surface of the III-V structure, which was connected in series with NiMo hydrogen-evolving nanoparticles dropcast on Ni film. The device remained stable in 1.0 molar aqueous KOH for over 150 hours with a current density of 8.5 milliamperes per square centimeter and ~100% Faradaic efficiency.

Solar fuel generation by direct photoelectrolysis of water is a pathway towards energy storage and transportation demands of a sustainable energy economy. Unlike photovoltaics (PV), photoelectrochemical (PEC) devices are operated in aqueous electrolytes. While consequences on device durability are widely discussed [1], the impact of sunlight absorption in the electrolyte on solar to hydrogen (STH) conversion have been neglected until recently [2]. We discuss the loss of solar photon flux during transmission through water using a detailed balance approach to derive STH efficiency prospects, discover a PEC design principle relating electrolyte thickness and allowable overvoltage loss, and derive optimum band gap combinations for PEC tandems.
PEC performance - beyond material quality known from PV - critically depends on sufficient threshold voltage to drive the water splitting reaction, chemical stability in contact to an electrolyte, and alignment of the band edge potentials [1]. High band gap materials may split water without bias, but neglect substantial parts of the solar spectrum. Tandem absorber structures overcome this undesirable trade-off featuring promising efficiency potentials. Epitaxial GaInP₂/GaAs photoelectrodes enabled world record solar to hydrogen (STH) conversion efficiencies of up to 12.4% [3].
Our detailed balance calculations for typical PEC operating conditions [2] show much higher STH efficiency prospects for GaInP₂ top junction tandems with a lower bottom junction band gap inaccessible by lattice matched growth of classical III-Vs on GaAs. Inverted metamorphic (IMM) growth is an established technique that enabled advanced multi-junction PV [4]. We have established a preparation technique for IMM PEC devices and demonstrate unassisted water splitting. Our results show increased STH efficiencies of 15% and above, with a pathway towards 2-junction IMM PEC devices with specifically adjusted band gaps boosting the prospects to 24% and beyond.
[1] Z. Chen et al., J. Materials Research 25, 3 (2010).
[2] H. Döscher et al., Energy and Environmental Science 7, 2951 (2014).
[3] O. Khaselev and J. Turner, Science 280, 425 (1998).
[4] J. Geisz et al, Applied Physics Letters 91, 023502 (2007).

Stable molecular linkers that can couple light absorbers and molecular catalysts will be important components in building a fully functioning artificial photosynthetic assembly that utilizes CO₂ reduction and H₂ evolving molecular catalysts. In this study we examine the stability of two classes of linkers, silanes and phosphonates, attached at one end to redox couples that serve as surrogates for electrocatalysts, and at the other to a transparent conductive oxide, fluorine-doped tin oxide (FTO). This oxide is selected due to its potential application as an anti-corrosion protection coating for photocathodes. The stability of the redox couple - linker - FTO systems was investigated electrochemically in aqueous solutions at room temperature over a range of potentials and at extreme acidic, basic and neutral pH values. Evaluation of coverages and chemical composition by x-ray photoemission spectroscopy (XPS) reveals that in general the linkers exhibit several degradation pathways, and that they dissociate from the FTO surface even at neutral pH.

Hydrogen is an ideal fuel for fuel cell applications. Hydrogen has to be produced from renewable and carbon-free resources using nature energies such as sunlight if one thinks of clean energy and environmental issues. In this regard, a photoelectrochemical (PEC) cell consisting of semiconductor photoelectrodes that can harvest light and use this energy directly for splitting water is a more promising way for hydrogen generation. Abundant and inexpensive oxide semiconductor such as ZnO has been recognized as a promising photoelectrode, but the photoconversion efficiency is substantially limited by its large band gap and rapid charge recombination. Recently, doping with 4d transition metal, such as Mo, has been carried out to remarkably enhance the PEC performance of many photoanodes including TiO₂, BiVO₄, and Fe₂O₃. Nevertheless, 1-D Mo-doped ZnO nanostructures have not been reported for PEC water splitting. We report the first demonstration of cobalt phosphate (Co-Pi) assisted Zn_1-xMo_xO nanorods (NRs) as visible-light-sensitive photofunctional electrodes to fundamentally improve the performance of ZnO NRs for PEC water splitting. The maximum photoconversion efficiency could be successfully achieved as high as 1.05%, with the significant photocurrent density of 1.4 mA cm^-2. More importantly, in addition to achieve the maximum incident photon to current conversion efficiency (IPCE) value of 86%, it could be noted that the IPCE of Zn_1-xMo_xO photoanodes at the monochromatic wavelength of 450 nm is up to 12%. Our PEC performances are comparable to those of many oxide-based photoanodes in recent reports. The improvement in photoactivity of PEC water splitting may be attributed to the enhanced visible-light absorption, increased charge-carrier densities, and improved interfacial charge-transfer kinetics due to the synergistic effects of Mo incorporation and Co-Pi modification, thus contributing to photocatalysis. The new design of constructing highly photoactive Co-Pi assisted Zn_1-xMo_xO photoanodes enriches the doping community and sheds light on developing high efficiency photoelectrodes for solar-hydrogen field.
Reference:
[1] Y.G. Lin, Y.K. Hsu, Y.C. Chen, S.B. Wang, J.T. Miller, L.C. Chen, and K.H. Chen, Energy Environ. Sci., 5, 8917 (2012).
[2] Y.G. Lin, Y.K. Hsu, Y.C. Chen, L.C. Chen, S.Y. Chen, and K.H. Chen, Nanoscale, 4, 6515 (2012).
[3] Y.K. Hsu, Y.C. Chen, Y.G. Lin, L.C. Chen, and K.H. Chen, J. Mater. Chem., 22, 2733 (2012).
[4] Y.G. Lin, Y.K. Hsu, A.M. Basilio, Y.T. Chen, K.H. Chen, and L.C. Chen, Optics Express, 22, A21 (2014).
[5] Y.K. Hsu, S.Y. Fu, M.H. Chen, Y.C. Chen, and Y.G. Lin, Electrochimica Acta, 120, 1 (2014).
[6] Y.G. Lin, Y.C. Chen, J.T. Miller, L.C. Chen, K.H. Chen, and Y.K. Hsu, ChemCatChem, 6, 1684 (2014).
[7] Y.G. Lin, Y.K. Hsu, Y.C. Chen, B.W. Lee, J.S. Hwang, L.C. Chen, and K.H. Chen, ChemSusChem, 7, 2748 (2014).

As a p-type semiconductor, CuBi₂O₄ is of potential interest as a photocathode or photocatalyst for the generation of hydrogen fuel by photoelectrochemical water splitting. Here, we describe the preparation of CuBi₂O₄ nanocrystals by a hydrothermal approach. The formation of CuBi₂O₄ and its 1.7 eV band gap are confirmed by powder X-ray diffraction (XRD) and optical studies (UV-Vis spectroscopy). According to TEM the particles are irregular shaped with average particle size of 25.7 ± 4.7 nm. Surface photovoltage (SPV) measurements on CuBi₂O₄ nanoparticle films give positive photovoltage at >1.7 eV photon energy, in agreement with hole majority carrier transport. The photovoltage is enhanced on fluorine-doped tin oxide (FTO) substrate and saturates at +0.21 V for 4,350 nm thick films. Photoelectrochemical scans on nanocrystal films immersed in 0.1 M K₂SO₄ aqueous solution at pH = 7 show photocathodic current of up to 6.45 mu;A cm^-2 with an onset potential of +0.75 V vs. NHE in the presence of Na₂S₂O₈ used as a sacrificial electron acceptor. Small amount of hydrogen is evolved under visible light when illuminated in the presence of sacrificial agent such as methanol. The significance of these findings will discuss the limitations of hydrogen evolution by CuBi₂O₄ photocatalyst from water.

Bismuth based band gap engineered multimetal oxide - bismuth titanates (Bi₂Ti₂O₇) belongs to a unique class of pyrochlore family having earth abundant building blocks, with a band-gap of ~2.8 eV. Unfortunately, these eco-friendly systems suffer from extensive charge recombination upon photoillumination and affect the light driven reactions, such as dye degradation, water (H₂O) splitting and carbon dioxide (CO₂) reduction. In order to leverage the charge separation and improve its photocatalytic activity, bismuth titanate (Bi₂Ti₂O₇) nanoparticles were wrapped with RGO (N-doped) to form a composite heterostructure in a cost effective way. An electrostatic ‘self-assembly&’ approach was followed to fabricate the hybrid structure. The optical and structural features of the heterostructure were characterized using UV-Visible, XPS (X-ray photoelectron Spectroscopy), SEM (Scanning Electron Microscopy) and TEM (Transmission Electron Microscopy) analysis. The UV-visible analysis revealed a red-shift in the absorbance with the increase in the RGO loading in the composite structure. The improved charge separation of the composite was verified using EIS (Electrochemical Impedance Analysis) and PL (Photoluminiscence) analysis. Photoelectrochemical measurements indicated an enhancement of ~136% in the photocurrent. The improved visible light absorption (‘red shift&’ in the absorbance) and enhanced charge separation mechanism was believed to occur through a [Ti-O-C] bond formation (similar to ‘carbon doping&’) between the oxide phase and the conductive carbon matrix. This was observed using XPS analysis. As a case study, light driven hydrogen generation efficiency of the photocatalyst was studied from water/methanol mixture, using a slurry reactor with UV-visible light illumination. The amount of evolved hydrogen was quantified using a Gas chromatography (GC). This showed an enhancement in the hydrogen yield by ~3 times, compared to pristine BTO, till a 1 wt% of RGO loading. A further increase in the RGO loading indicated a decrease in the hydrogen yield. This can be a resultant of two factors: (1) due to the light shielding effect of the RGO, as it can block the light absorption and reduce the light intensity at surface and (2) blockage of the active sites of the catalyst. The repeated use of the photocatalyst demonstrated a stable hydrogen production with time and no structural disruption, indicating a robust BTO/RGO photocatalyst system. The study shows N-doped RGO is a cost effective substitute for expensive noble metals such as Platinum (Pt) to design and synthesize an eco-friendly composite photocatalyst systems. The method utilized to elevate the charge separation can be a general strategy to improve electron-hole life-times for various other oxides of different class such as delafossite, perovskite and sillenite.

Clean and renewable hydrogen energy is considered as one of the best future energy sources due to its high power density and no carbon dioxide emission. Utilizing visible light-driven photocatalysts and solar energy to produce large-scale H₂ by water splitting is a promising technique without waste of energy in comparison with traditional electrolysis method. In the present study, particulate photocatalyst GaN:ZnO was prepared by two steps. First, Ga₂O₃ was synthesized by plasma enhanced atomic layer deposition (PEALD) on substrate at 250 ^oC, and ZnO was then deposited to Ga₂O₃ to form alternating laminated layers of Ga₂O₃/ZnO. In the second step, Ga₂O₃/ZnO was nitridated under NH₃ atmosphere at 850 ^oC for 2 h to form GaN:ZnO. Both of as-deposited Ga₂O₃ and ZnO are amorphous. The growth rate of Ga₂O₃ is 0.23 #506;/cycle when the lengths of pulse time for the sequence triethylgallium (TEG)-N₂-Ar/O₂ plasma-N₂ of the ALD process are 0.1-5-5-5 s. The growth rate of ZnO is 0.37 #506;/cycle when the lengths of pulse time for the sequence diethylzinc (DEZ)-N₂-Ar/O₂ plasma-N₂ of the ALD process are 0.1-5-5-5 s. The further transformation of Ga₂O₃/ZnO laminated layer to form GaN:ZnO photocatalyst will be discussed.

Photoelectrochemical water splitting is a promising approach to convert the solar energy into hydrogen which can be stored and transported. III-Nitride semiconductors are promising materials as a photoelectrode because of their wide range of band gap energies and their ability to split water without external bias. We proposed an undoped GaN/AlN/n-type GaN (u-GaN/AlN/n-GaN) as a new structure of nitride photoelectrode. [1] Electrons photoexcited in the u-GaN are injected into an electrolyte because polarization charge at GaN/AlN interfaces generates large potential difference inside the structure. Photogenerated holes, on the other hand, flow in the opposite direction and are extracted by tunneling from u-GaN to n-GaN. The u-GaN/AlN/n-GaN structure, therefore, works as photocathode though it consists of only n-type semiconductors. The crystal quality of AlN is highly important in this structure because amount of polarization charge depends on it. In this work, we studied the effect of AlN thickness and growth temperature on the photoelectrochemical properties of the u-GaN/AlN/n-GaN photocathode.
The samples were grown by metal organic chemical vapor deposition (MOCVD). A GaN buffer layer was first deposited on a sapphire substrate followed by n-GaN, a few nm thick AlN and u-GaN. Three series of u-GaN/AlN/n-GaN structure were prepared with different AlN growth temperature of 800, 950 and 1130 ^oC. Each series consists of several samples with different thickness of AlN layer ranging from 3 to 6 nm. The photoelectrochemical properties were measured in a 0.5M H₂SO₄ under Xe lamp illumination.
As the polarization charge exists at AlN/GaN interface, thicker AlN layer generates larger bias inside. AlN layer needs to be thick enough to form a tunnel junction between u-GaN and n-GaN. Thick AlN layer, on the other hand, impedes tunneling and is difficult to grow without relaxation. Therefore, the sample which had AlN layer with an optimum thickness of about 4 nm showed the most positive turn on voltage of photocurrent in each series.
The samples with AlN grown at 800^oC which is the lowest growth temperature among the three series showed the best photoelectrochemical properties, though the quality of the AlN layer considered to improve as the growth temperature increases. We checked the AlN layer with STEM, AFM, and XRD analysis to find out the reason. AlN grown at 800 ^oC showed most abrupt interface while AlN grown at 950 ^oC showed the smoothest surface morphology. We concluded from these results that abruptness of AlN/GaN interfaces is the most important for the u-GaN/AlN/n-GaN photocathode.
[1] A. Nakamura, K. Fujii, M. Sugiyama, Y. Nakano, Phys. Chem. Chem. Phys. 16 (2014) 15326.

An InGaN photocatalyst was produced H₂ efficiently without extra bias and sacrificial reagent. The H₂ evolution rate was 0.9 mL/cm²/h. Its energy conversion efficiency from light to hydrogen was as high as 2.4%.
Nitride semiconductors are attractive materials not only for energy saving devices such as LEDs and LDs but also clean energy devices. Nitride photocatalyst are possible to produce “clean hydrogen” from water using light energy [1-3]. GaN photocatalyst with NiO co-catalyst produced H₂ stable and highly efficient [3]. Since a GaN has wide gap and can absorb only 1.4% of solar energy at AM 1.5. The solar spectrum has the highest irradiation intensity in the visible region. To enhance H₂ evolution rate by using sunlight, it is necessary to narrow bandgap rather than that of GaN. In order to overcome this problem, we adopt the InGaN as light absorption layer. Since the bandgap of InGaN is possible to control from 0.65 to 3.42 eV, InGaN photocatalyst can absorb light energy not only ultra violet region but also visible region. InGaN is very attractive material for light absorption of sunlight. In this paper, we will report the high efficiency of InGaN photocatalyst.
We used the InGaN absorption layer (110 nm), n-type GaN conduction layer (3 mu;m) on a c-sapphire substrate grown by MOVPE, and NiO co-catalyst was deposited on its surface. The InGaN absorption layer is graded InN molar fraction from 0 to 0.08. The InGaN photoelectrode was connected to Pt counterelectrode, and was immersed into 1 mol/L NaOH electrolyte without any sacrificial reagent. Our measurement system was without extra bias. A 300-W Xe-lamp with energy density of 100 mW/cm² was irradiated on the InGaN photoelectrode surface.
The H₂ evolution rate was 0.9 mL/cm²/h. The energy conversion efficiency from light to hydrogen was 2.4%. The InGaN surface was not etched. This result indicates that InGaN photoelectrode is very useful to produce H₂ from water. InGaN is necessary to product high efficiency photoelectrode.
References
[1] K. Fujii, T. Karasawa, and K. Ohkawa, Jpn. J. Appl. Phys.44, L543 (2005).
[2] M. Ono, K. Fujii, Y. Iwaki, A. Hirako, T. Yao, and K. Ohkawa, J. Chem. Phys. 126, 054708 (2007).
[3] K. Ohkawa, W. Ohara, D. Uchida, and M. Deura, Jpn. J. Appl. Phys. 52, 08JH04 (2013).

Hydrogenated amorphous silicon carbide (a-SiC:H) is an attractive semiconducting photoelectrode for photoelectrochemical water splitting due to its strong light absorption, its high stability, its composition made entirely from earth abundant elements and its scalable fabrication in industrial processing techniques.
Several observed limitations for a-SiC:H include the late photocurrent onset potential which is essentially due to the electron energetics dependence on the semiconductor-liquid junction, and the poor catalytic activity on the electrode surface. To address the former limitation, we have deposited a thin n-type amorphous titanium dioxide (TiO₂) as a front surface field layer onto the p-type/intrinsic a-SiC:H, forming a rectifying hetero p-i-n junction. The p-i-n structure created an internal electric field that increased the operating photovoltage, and subsequently improved the drift mechanism of photogenerated charge carriers across the space charge region. The enhancement of the photovoltage lead to a very positive photocurrent onset potential of +0.8 V vs. RHE and exhibited a photocurrent density of 8.5 mA/cm² at 0 V vs. RHE with only a 100 nm absorber layer and without employing any hydrogen evolution co-catalyst.
In addition, we have investigated the influence of the dopant level on the electronic conducting behaviour of the TiO₂, which showed a variation of photocurrent onset potentials. Using electrochemical impedance spectroscopy and Mott-Schottky analysis, we estimated the band edge positions and the electronic properties of each of the semiconducting components. Electronic band diagrams have been developed to illustrate the thermodynamic and the interfacial energetics of electron transfer at the semiconductor-liquid interface.
Finally, the a-SiC:H photocathode with a front surface field TiO₂ layer showed a high stability for 12 hours of operation under photocatalytic conditions. This high performance, very thin, and earth-abundant photocathode is very promising for integration with smaller band gap solar absorbers to form a bias-free multijunction photoelectrode.
Overall, we have been able to identify and address the possible limiting factor for a-SiC:H photocathodes, and achieved a significant advancement in understanding and constructing a highly efficient photoelectrode for solar water splitting.

We report on the optimization of thin film silicon based, integrated photovoltaic-electrochemical water-splitting devices. Silicon thin film technology represents an attractive way to realize large scale water-splitting applications because it enables scalable low-cost production, earth-abundance and versatility. The combination of amorphous hydrogenated silicon (a-Si:H) and microcrystalline silicon (µc-Si:H) thin films in tandem or triple junction devices allows for the efficient utilization of the solar spectrum and the generation of a voltage well above the thermodynamically required 1.23 V to drive the oxygen and hydrogen evolution reactions. In practice a voltage above 1.6 V is regularly required due to additional overpotential losses at the interface and in the electrolyte.
a-Si:H and µc-Si:H layers were deposited on transparent glass/SnO₂:F substrates by plasma enhanced chemical vapor deposition. The optical and electronic material properties and the interfaces were carefully adjusted to optimize the current density of tandem and triple junction solar cells while maintaining a sufficiently high voltage for practical water-splitting. We achieved a-Si:H/a-Si:H tandem and a-Si:H/µc-Si:H/µc-Si:H and a-Si:H/a-Si:H/µc-Si:H triple junction solar cells which provide open-circuit voltages V_oc above 1.8 V, 1.9 V and 2.3 V, respectively, and exhibit photovoltaic conversion efficiencies well above 10%.
In the integrated photovoltaic-electrochemical device a metallic layer at the solid/electrolyte interface between the solar cell and the electrolyte was implemented for multiple purposes: (i) it functions as a back reflector for transmitted photons and thereby increases the probability of photon absorption and charge carrier generation in the silicon and thus enhances the photo current of the device, (ii) it forms a good electronic contact to the silicon semiconductor and provides an efficient charge transfer at the solid/electrolyte interface, (iii) it protects the silicon solar cell against corrosion, and (iv) it can provide catalytic activity and hence lower the overpotential for hydrogen evolution. Considering this broad range of demands, the metal has to be chosen carefully. Here, we systematically evaluated thin layers (approx. 300 nm) of Ti, Ni, Ag, Cu, Au, Al, and Pt on glass substrates as well as attached to the photocathodes with regards to their stability and catalytic activity. Electrochemical experiments were conducted both in acidic and alkaline solutions because both the stability against corrosion and the catalytic activity depend on the pH of the used electrolyte.
Modeling the integrated photovoltaic-electrochemical system in terms of a simple series connection of a solar cell and an electrolysis cell shows good agreement with experimental results and allows for a detailed analysis of the losses and a prediction of efficiency limits for thin film silicon based water-splitting devices based on the solar cell performance.

Producing hydrogen fuel directly from photoelectrochemical (PEC) water splitting is considered as one of the most scalable and cost-effective approaches for using solar energy. Because of its earth-abundant#65292; low-cost and narrow bandgap (E_g=1.12eV) characteristics, silicon (Si) is an attractive candidate for photoelectrode of a PEC system. Lots of vertically aligned silicon nanostructures, such as nanoholes, nanowires and nanotips, have been investigated for use in solar-hydrogen production, because one-dimensional (1D) materials are beneficial for improving the transportation of carriers.
Highly ordered silicon nanopillars (SiNPs) with a granular morphology have been fabricated by the Au and Ag combined metal assisted chemical etching (MACE). The diameter and the center-to-center distance between the adjacent nanopillars have been accurately controlled by the size of a polystyrene (PS) template. The MACE conditions could precisely tailor the length and size of SiNPs. These pillars efficiently increased the photon scattering to increase the optical path length as well as the absorption probability. The sub-wavelength structure of granular SiNP arrays showed excellent antireflection property with low reflectance of ~5%, which was much less than that of conventional SiNPs, within the wavelength range of 400-1000 nm. The granular SiNPs revealed a porous morphology, which is beneficial for increasing the electrode area in hydrogen evolution. Henceforth, better electrical contact between Si cathode and electrolyte led to the reduction in overpotential required for photoelectrochemical hydrogen reaction. Compared to the conventional SiNPs, our granular SiNP sample enhanced the photocurrent density of ~10%.

One-dimensionally periodic lamellar nanopatterns derived from the self-assembly of symmetric poly(styrene-co-methylmethacrylate)(PS-PMMA) diblock copolymer are incorporated on the surface of p-type single-crystalline silicon to form photocatalytic electrodes for solar water splitting. The resulting nanostructured silicon photocathodes exhibit improved photoelectrochemical performance compared with a bare silicon without nanostructure implementation. Experimental studies on the optical properties and photoelectrochemical performance characteristics, together with optical modeling of photon absorption based on the finite-difference time-domain (FDTD) method, indicate that the combined effects of anti-reflection and increased surface area contribute to the enhancement of light absorption and facile desorption of gas bubbles, as well as reduction of photocatalytic over-potential for hydrogen evolution reaction.

Photoelectrochemical (PEC) water splitting cells for H₂ production have attracted intensive attention to convert solar energy to a clean and storable fuel [1]. Silicon is a promising material for PEC water splitting, because it is an earth-abundant and low cost material and has small bandgap (1.12eV) [2]. However, silicon forms an insulating SiO₂ or even corrodes under oxidative potentials in aqueous solutions during oxygen evolution reaction (OER) [3, 4]. Therefore a corrosion-resistant protection strategy is required in order to use silicon for efficient and durable PEC water splitting cells [3, 4].
Here, we propose a novel strategy to protect a silicon photoanode for OER during water splitting reaction. Our silicon protection strategy is to isolate vulnerable Si surface from a solution by coating chemically inert dielectric films (such as SiN_x or SiO₂). Then, since SiN_x and SiO₂ films are insulating, we locally form OER catalyst patches directly on the silicon photoanode by patterning the dielectric films to allow injection of photo-excited holes to water. Our silicon photoelectrode design provides high durability by completely burying the silicon surface under the films and OER catalysts patches. At the same time, it maximizes light absorption for high water oxidation efficiency, since SiO_x and SiN_x are well-known anti-reflection coating in silicon photovoltaic cells. Indeed, we confirmed that SiO_x/n-Si photoanode with Pt patches operated over 19 hours without degradation in a highly oxidative condition (2.98V vs RHE, 21mA/cm², 1M NaOH) under simulated 1-sun illumination. In addition, we reduced the OER overpotential more than 600 mV by forming a buried Si p/n junction under a protection layer, compared to the n-Si photoanode, while lasting more than 24 hours. In the presentation, we will present photoelectrochemical water oxidation characteristics of our new silicon electrodes in more detail.
Acknowledgement
This work was supported by the Korea CCS R&D Center (KCRC) grant funded by the Korea government (Ministry of Science, ICT & Future Planning) (NRF-2014M1A8A1049303).
Reference
1. N.S Lewis, and D.G. Nocera, Proc. Natl. Acad. Sci. USA 103, 15729 (2006).
2. J. Oh, T.G. Deutsch, H.-C. Yuan, and H.M. Branz, Energy and Environ. Sci. 4, 1690 (2011).
3. Y. W. Chen, J.D. Prange, S. Dühnen, Y. Park, M. Gunji, C.E. Chidsey, P.C. Mclntyre, Nature Mater. 10, 539 (2011).
4. S. Hue, M.R. Shaner, J.A. Beardslee, M. Lichterman, B.S. Brunschwig, N.S. Lewis, Science. 344, 1005 (2014).

Strategies for successful development of new sustainable energy technologies requires consideration of appropriate levels of life cycle assessment, theory, computation, and experiment, as well as the explicit incorporation of a manufacturability perspective throughout the product research, development, demonstration, and deployment cycles. To accelerate the discovery and development of material, in particular nanomaterials, it is critical to integrate these factors seamlessly into a comprehensive approach. However, success in this endeavor frequently depends intimately on tools to characterize nanomaterials at all stages of their development to ensure processes are scalable and desired material properties are escorted to and throughout manufacturing.
A class of versatile characterization techniques essential to the development of nanomaterials vital to a range of energy sciences research are those provided by and relying on synchrotron radiation-based approaches. In the soft x-ray regime, x-ray absorption spectroscopy (XAS), x-ray emission spectroscopy (XES), resonant inelastic x-ray scattering (RIXS), and photoelectron spectroscopy (PES) are the core methodologies employed. Similarly, there is a corresponding suite of hard x-ray methods represented by XAS, XES, RIXS, small angle x-ray scattering along with several additional x-ray scattering approaches that comprise core approaches currently being utilized. A concurrent theme in all of the synchrotron radiation studies is the focus on high-spatial and -spectral resolution coupled with the capability to conduct non-destructive measurements in-operando.
One these methods, soft x-ray spectromicroscopy with the scanning transmission x-ray microscope (STXM) at the Molecular Environmental Science (MES) Beamline 11.0.2 of the Advanced Light Source (ALS). The MES STXM is capable of imaging with a spatial resolution well below 10 nm, can directly probe the light element K-edges below 2 keV via x-ray absorption, and can be employed to probe a diverse range of materials including particulates and highly air-sensitive materials. Representative results from several particulate-based systems will be presented highlighting the utility of the STXM for a broad range of materials research. A critical discussion will follow on developmental opportunities for STXM that will include recent developments in ptychography, as well as other x-ray methods for the characterization of materials systems for energy science.

We report insights from experiments on the influence of surface modifications on structure and reaction chemistry associated with the heterogeneous oxidation of water on hematite (α-Fe₂O₃)-based photoanodes. Specifically, we apply a classical surface science approach to understand the influence of dopants, co-catalysts, and composite oxide components on structure and reactivity. We have utilized a range of techniques for surface analysis to characterize the structure and properties of Ni-doped and mixed-oxide hematite single-crystal surfaces formed by vapor deposition under controlled conditions. XPS, LEIS, LEED and STM were used to characterize the structure of Ni-modified thin films of α-Fe₂O₃ model catalysts with different morphologies and geometries. TPD, XPS, and UPS were used to study water adsorption and reaction, and characterize the influence of Ni modification on thermal and photochemical reaction mechanisms. Ni doping caused a new termination for the α-Fe₂O₃(0001) film and also induced new surface chemistry, probed by water TPD, which revealed a new, higher temperature OH recombination desorption peak that is due to more stable surface-bound OH groups, as identified by UPS. These surface-science type experiments were combined with electrochemical water oxidation measurements on photoanodes prepared by thin-film and nanomaterials synthesis to elucidate new information on the surface phases of hematite-based photoanodes and on their specific stability and reactivity toward photoelectrochemical water splitting. In addition to studying the effect of dopants and oxide composites, we have also investigated the facet-dependent activity and stability of Co₃O₄ for water oxidation using the well-defined morphologies of nanocubes and nanooctahedra synthesized using a hydrothermal method without the use of surfactants. Cobalt oxide is often used as a cocatalyst on hematite and other oxide photoelectrocatalysts to enhance water oxidation activity, and is an interesting electrocatalyst in its own right. Our studies demonstrated that the OER activity, measured in terms of overpotential and current density, showed significant dependence on the predominant facets exposed, where the (111) crystal surface vastly out-performs the (100) crystal surface.

During the last 5-10 years, many research efforts have been focused on the development of new and improved metal oxide-based semiconductors for photoelectrochemical water splitting. As a result of these efforts, there are now a handful of metal oxides that show photocurrents in excess of 4 mA/cm² under simulated sunlight, most of which have been developed in the last 4 years. Moreover, several demonstrations of stand-alone (i.e., bias-free) water splitting systems based on metal oxide photoelectrodes have been reported. Important factors that have enabled these breakthroughs are the progress in the synthesis of nanostructured materials with controlled sizes and shapes, the development of effective co-catalysts, and the increased focus on new materials (i.e., other than TiO₂, Fe₂O₃ and WO₃).
Despite these promising developments, the efficiencies of most oxides are still far away from the theoretical efficiencies. The mismatch between carrier diffusion length and optical absorption depth remains a key issue. Nanostructuring does not always solve this, since recent work by several groups has shown that surface recombination is an important loss mechanism for several important oxides. We find that this is also the case for BiVO₄. We performed an intensity-modulated photocurrent spectroscopy (IMPS) study on BiVO₄ modified with cobalt phosphate. The results suggest that the CoPi strongly reduces the surface recombination rate, but that it also suppresses the kinetics of hole injection into the electrolyte. The implications of this surprising observation will be discussed.
One possible approach to avoid the need for nanostructuring is to concentrate the optical absorption in a smaller region of the material. Towards this end, we have explored the surface modification of BiVO₄ photoanodes with Ag@SiO₂ core-shell nanoparticles. Localized surface plasmon resonances in the Ag core are found to enhance the optical absorption in the BiVO₄, mainly due to far-field effects (scattering). Intriguingly, we find that the improvement in photocurrent (x2.5) is much larger than the increase in optical absorption (x1.3). We tentatively attribute this to a plasmon-induced enhancement of the water oxidation kinetics.
We will conclude by showing some initial results for Fe₂WO₆, a ternary oxide with an indirect bandgap of 1.6 eV, and critically discuss whether or not this material merits further study.

Using sunlight to generate fuels through direct photoelectrochemical PEC processes is a promising approach to extend solar energy conversion into the transportation sector or provide storage of solar energy for electricity generation. One approach to meeting this challenge is to split H₂O into H₂ and O₂ photoelectrochemically in a tandem device configuration where two photoabsorbers in series effect the proton reduction and water oxidation half-reactions. Toward these ends, we have been exploring nanostructured semiconductors that could be incorporated into such a tandem system. Nanostructuring provides several advantages over non-porous electrodes such as anti-reflectivity, higher surface area (which lowers the charge-transfer requirements per unit area), and unique catalyst binding properties. In this presentation, we will discuss our recent work on the interfacial charge transfer processes of surface-adsorbed catalysts on (a) nanoporous Si photocathodes (a.k.a. “black Si”) and (b) nanoporous metal oxide photoanodes (e.g., BiVO₄ and related oxides).

To date, devices based on p-GaInP₂ epitaxial thin films, in a tandem configuration with p/n-GaAs, represent the “world record” material for photoelectrochemical (PEC) solar water splitting at 12.4% solar-to-hydrogen (STH) conversion efficiency¹. For further improvement of the material it is essential to understand the electronic and chemical surface structure of p-GaInP₂ thin films and the p-GaInP₂/electrolyte interface, and how it changes during prolonged PEC operation.
We have employed a suite of experimental techniques, in particular lab-based soft X-ray and UV photoelectron spectroscopy (XPS and UPS) and inverse photoemission spectroscopy (IPES) under ultra-high vacuum conditions at UNLV. This was coupled with synchrotron-based ambient pressure (tender) X-ray photoelectron spectroscopy (AP-XPS)² at the ALS. The obtained information from the lab-based techniques includes the (initial) chemical surface composition as well as the band edge energies, Fermi energy, and the electronic surface bandgap. In addition, AP-XPS provides information about the electronic and chemical structure of the p-GaInP₂ film at the solid/liquid interface during operando measurements, and thus sheds light on the influence of electrolyte and applied potential on the material properties. Based on this data, a comprehensive picture of the p-GaInP₂ electronic and chemical surface structure can be drawn, and insights into the photocathode/electrolyte interface can be gained.
¹ O. Khaselev and J.A. Turner, Science 280, 425 (1998).
² S. Axnanda, E. J. Crumlin, et al., “Using “Tender” X-ray Ambient Pressure X-ray Photoelectron Spectroscopy as A Direct Probe of Solid-Liquid Interface”, submitted.

Bismuth vanadate (BiVO₄) is a promising n-type semiconducting material for photoelectrochemical (PEC) water splitting devices. It has a suitable band-gap energy (~2.4 eV), a favorable conduction band edge position (just below the water reduction potential), and it is relatively stable in near-neutral aqueous environments. Moreover, it is made of cheap, non-toxic, earth-abundant materials and can be easily produced on a large scale with techniques such as spray pyrolysis. Consequently, to make production of solar fuels with BiVO₄ a viable solution, its limitations have to be identified and addressed accordingly.
Notably, BiVO₄ suffers from substantial surface recombination losses. This problem has been partially resolved by decoration with oxygen evolution catalysts^1-3. However, strong parasitic absorption by these catalysts can make it prohibitive to use them in tandem devices, requiring front-side illumination. In addition, the catalytic activity at low bias potentials still remains low for photoanodes modified with co-catalysts. Therefore new strategies to tackle surface recombination are necessary. Even more importantly, it is of paramount significance to understand the physical and chemical nature of the semiconductor/electrolyte interface in BiVO₄ photoanodes, specifically the role of the electronic surface states constituting the main pathway for recombination.
We have found that the PEC performance of BiVO₄ photoanodes can be dramatically improved by prolonged exposure to illumination in an open circuit configuration. Thus, a record photocurrent for bare, undoped, uncatalysed BiVO₄ photoanodes of 3.3 mAcm^-2 (at 1.23V vs RHE) has been achieved for anodes subjected to such light treatment. Moreover, photoelectrochemical tests with a sacrificial agent have yielded catalytic efficiency near unity, suggesting elimination of the major pathway for recombination. Finally, we have shown that photocharging induces a significant cathodic shift in the photocurrent onset potential of about 0.3 V.
We ascribe the observed phenomena to photo-induced passivation of surface states of BiVO₄. Such passivation results in a diminished Fermi level pinning effect and consequently, an increase in the observed photovoltage. We have been able to identify the energy level of these apparent surface states using a novel technique and propose how it correlates to the photocurrent onset potential. Ultimately, we have demonstrated a novel approach to investigate and improve the nature of the BiVO₄/electrolyte interface, which can be potentially applied to other photoanode/photocathode materials.
1. D. K. Zhong et al., JACS, 2011, 133, 18370-7.
2. C. Ding et al., PCCP, 2013, 15, 4589-95.
3. S. K. Choi et al., PCCP, 2013, 15, 6499-507.

Metal oxide semiconductors are actively investigated for use as photoanodes in photoelectrochemical water splitting systems because of their improved stability relative to traditional non-oxide semiconductors under aqueous conditions. Among these materials, monoclinic scheelite bismuth vanadate (BiVO₄) is a particularly promising material and has been the subject of significant interest due to the favorable alignment of its band edge positions for water oxidation, its moderate bandgap of 2.5 eV, and its relatively long minority hole diffusion lengths. However, little is known about the native defects in this system, their effect on photoelectrochemical performance, and possible passivation schemes. In this work, we have utilized a range of spectroscopic techniques, in conjunction with density functional calculations, to understand the nature of recombination centers in this material and develop methods for their passivation. Steady state photoluminescence measurements reveal no band edge emission, consistent with the identification of BiVO₄ as an indirect bandgap semiconductor. However, sub-bandgap luminescence is observed; temperature-dependent measurements of this radiative recombination process reveal a 620 meV deep trap that is thermally quenched by a 31 meV activated state. These observations are consistent with a donor-acceptor pair recombination mechanism with a ~100 µs lifetime. However, annealing the thin films of BiVO₄ in a H₂ atmosphere improves the photoanodic current onset potential by ~100-200 meV, increases the fill factor, and significantly reduces the sub-bandgap photoluminescence. These results on thin films, together with XRD and solid state ¹H NMR analysis on BiVO₄ powders annealed in H₂, suggest passivation of native defect states by incorporation of hydrogen into the lattice. The results highlight that detailed understanding and controlling of carrier trapping in metal oxides, which often exhibit complex native defect properties and compositional non-uniformities, provides significant opportunity for increasing photoelectrochemical water splitting performance.

We investigated the advantages of using helical nanostructures in photoelectrochemical (PEC) solar water splitting via a traditional WO₃/BiVO₄ heterojunction system. A helical WO₃ array was first fabricated as an electron transporting medium, followed by subsequent coating with BiVO₄, metal doping and surface modification to form a heterojunction with a compact connection. This novel heterojunction, based on a helical array framework, displayed outstanding performance when used in PEC solar water oxidation. A maximum photocurrent density of approximately 5.35±0.15 mA/cm² was achieved at 1.23 volts (V) versus the reversible hydrogen electrode (RHE), and related hydrogen and oxygen evolution was also observed. Theoretical simulations and analyses were performed to verify the advantages of this helical structure. The combination of effective light scattering, improved charge separation and transportation and an enlarged contact surface area with electrolytes due to the use of the BiVO₄-decorated WO₃ helical nanostructures led to the highest reported photocurrent density to date of 1.23 V vs. RHE.

There has been significant progress on using bismuth vanadate (BiVO₄) as a photoanode, which exhibits attractive stability, onset potential, and photocurrent. However, a significant portion of sunlight cannot be absorbed by the BiVO₄ due to its relatively large bandgap. Moreover, the saturated photocurrent is still far lower than the theoretical limit, likely due to short minority carrier diffusion length and the high surface recombination rate. In this work, we fabricated a Si-BiVO₄ heterostructure to utilize the solar spectrum more efficiently by taking advantage of the light absorption in Si. The turn-on voltage of such a heterostructured photoanode is substantially improved by nearly 400 mV, compared with BiVO₄ photoanodes alone. Furthermore, we demonstrate that the photocurrent is strongly activated by temperature, increasing by 250% going from room temperature to 65 ^oC. We attribute this improvement to the thermally-activated nature of reaction kinetics and carrier transport.

In response to the global concerns on the challenges of climate change and sustainable energy supply, the use of abundant solar energy is becoming an attracting approach for solar chemical and electricity generation. Aimed at developing new nanostructures for efficient photocatalytic and photo-electrochemical solar fuel generation, we have been focusing the following two aspects of semiconductors; 1) band-gap engineering of a group of layered metal oxides including titanate, tantalates and niobate-based pervoskites for use as visible light driven photocatalysts for solar hydrogen generation; and ) dual functional photoelectrochemical reactor design which can simultaneously split water to generate solar hydrogen and remove organic pollutants or waste products for value-added chemical production. These newly-developed materials and PEC systems underpin important applications of solar fuel generation and environmental remediation.

The development of catalysts for solar fuels, solar cells and photo-assisted water and air cleaning require robust, routes capable of producing advanced complex materials. These multi-functional devices should both be effective in absorbing photons for generation of holes and electrons, charge transport and catalysis of surface reactions and be stable. Quantum confinement in typically less than 3-4 nm sized oxide systems may also be utilized for improved charge separation and transport properties which puts high demands on the precision of the processing routes, although they still need to be of low cost. For such tough combination of demands solution based routes are probably the best suited for many systems, but there is a need for further development of these processes, before they can be fully exploited for industrial scale manufacture of complex tailored materials. Here is discussed two types of solution based processing routes to doped and non-doped oxide nano-particles, thin- and ultra thin films; one based on reactive alkoxides that yield pure homo and heterometallic oxides and one based on organically modified salts that yield metals and metal-in-oxide nano-composites. The oxide systems involve doped and non-doped transition metal oxides, spinels and perovskites as nano-particles, films and ultra-thin films. Emphasis is on the ultra-thin complex oxides, metals and metal-oxide nano-composites which are promising for large scale precise modification of semi-conductor structures for use as e.g. water splitting devices were OER and HER catalysts, band and transport control through quantum size effects, modification of surface charge and corrosion stability all may be controlled by such thin layers. The materials prepared were characterized with a large range of analytical techniques and their structures will be related to their properties.

Development of low cost materials with E_g = 1.8 eV and electronic properties suitable for integration in a tandem cell is a major hurdle for the development of solar water splitting. Ternary III-V semiconductor alloys like GaAs_1-xP_x are promising candidates for these applications. However, the traditional metal-organic chemical vapor deposition (MOCVD) growth technique is expensive, with costs near $10,000/m². Close-spaced vapor transport (CSVT), which uses water as a transport agent, is an alternative, plausibly inexpensive technique for the deposition of GaAs_1-xP_x recently shown to produce high quality GaAs.^1,2 GaAs_1-xP_x growth by CSVT utilizes a powder source composed of mechanically ground GaAs and GaP, which allows for control of composition and dopant density in the product. Previous studies^3-5 have shown that GaAs_1-xP_x can be grown using CSVT, but no electronic characterization has been reported.
We report the growth of n-GaAs_1-xP_x, 0.3 le; x le; 0.7, on GaAs using CSVT. Film composition and epitaxy were confirmed by XRD rocking curves. The ratio of transported phosphorous to arsenic is modulated by growth temperature, and can be controlled to produce near-unity phosphorous incorporation from the source. Initial electron microscopy studies of the GaAs_1-xP_x films show irregular surface morphology and stacking faults throughout the films; however, non-aqueous photoelectrochemical characterization shows open-circuit potentials of at least 1 V for all compositions and maximum short-circuit currents of 7.4 mA/cm² (x = 0.3), indicating reasonable electronic quality. Advances in GaAs_1-xP_x production as well as spectral response and optical reflectivity will be presented. This growth method may make a significant impact on the photoelectrochemical field by providing easy access to a tunable bandgap material for high efficiency solar water splitting.
(1) Ritenour, A. J.; Cramer, R. C.; Levinrad, S.; Boettcher, S. W. Appl. Mater. Interfaces2012, 4, 69-73.
(2) Ritenour, A. J.; Boucher, J. W.; DeLancey, R.; Greenaway, A. L.; Aloni, S.; Boettcher, S. W. Energy Environ. Sci.2014.
(3) Gottlieb, G. E. J. Electrochem. Soc.1965, 112, 192-196.
(4) Purohit, R. K. J. Mater. Sci.1968, 3, 330-332.
(5) Hoss, P.-A.; Murray, L. A.; Rivera, J. J. J. Electrochem. Soc. Solid State Sci.1968, 115, 553-556.

The III-V tandem GaAs:GaInP2 PV:PEC (photovoltaic:photoelectrochemical) cell has achieved 12.4% solar-to-hydrogen (STH) water splitting efficiency, but the instability of GaInP₂ in contact with electrolyte greatly limits device lifetime. A recent discovery in our group demonstrated that unmodified MOCVD-grown p-GaAs epilayers made surprisingly stable photocathodes that could operate for 100 hours in 3M sulfuric acid at high photocurrent density with negligible degradation. Although not fully understood, we believe this exceptional stability results from surface chemistry unique to MOCVD-grown p-GaAs epilayers. Thus, we were motivated to demonstrate a tandem cell that uses MOCVD-grown p-GaAs in place of p-GaInP₂ at the semiconductor-electrolyte junction. Furthermore, GaAs has the advantages of being the most commonly studied and perhaps least resource-constrained III-V. A single GaAs absorber (1.42 eV band gap) has a theoretical light-limited photocurrent density of 32 mA/cm² (under AM 1.5G illumination), but does not provide sufficient voltage to split water. For two GaAs junctions in tandem configuration, the voltage is doubled and the current density is halved, but the maximum photocurrent density of 16 mA/cm² is still quite high (nearly 20% theoretical STH efficiency). Unlike traditional epitaxial III-V tandem devices that use two different band gap absorbers (e.g. GaAs:GaInP₂ with 1.81 and 1.42 eV band gaps) to split the solar spectrum and achieve current matching, it was necessary to design and fabricate a GaAs:GaAs tandem with an optically thin top absorber. The optical thinning approach has the advantage of current matching by simply tuning the top absorber thickness, but does not benefit from the additional voltage that a higher band gap top absorber would provide. Regardless, the GaAs:GaAs tandem device provided sufficient voltage to drive water splitting at nearly 9 mA/cm² (>10% STH efficiency) at short circuit in on-sun measurements. Further results on stability testing of the GaAs:GaAs tandem will also be presented.

In the past decades, numerous semiconducting materials, especially oxide semiconductors, have been investigated as potential photoelectrodes in photoelectrochemical (PEC) system with a view to efficient light-induced water splitting for solar-hydrogen conversion. Compared to other metal oxide semiconductors, α-Fe2O3 (hematite) has the advantage such as the small band gap energy of ~2.0 eV, which enables it to absorb most of the photons of solar spectrum. Furthermore, its chemical stability, widespread availability and innocuity make it promising to be employed in a large scale. Unfortunately, the ultrafast recombination of the photogenerated carriers and the poor minority charge carrier mobility lead to a short hole diffusion length in α-Fe2O3, which severely limits the overall photocurrents produced by solar light. We reported that doping with metal ions displayed positive effects on the efficiency of α-Fe2O3 photoanodes, which was mostly attributed to the improved charge conductivity.
In this study, a facile solution-based method was developed to fabricate hematite nanorods coated with ultrathin overlayer of TiO2 or AgxFe2-xO3. The core/shell nanorod structures of α-Fe2O3/TiO2 and α-Fe2O3/AgxFe2-xO3 were obtained by annealing solution-fabricated β-FeOOH nanorod arrays which were first ultrasonicated in TiO2 sol and Ag+ aqueous solution, respectively. In the α-Fe2O3/TiO2 nanorod structure, TiO2 overlayer could extract photogenerated holes from α-Fe2O3 core via the quantum-mechanical tunneling process, resulting in promoted charge carrier separation, and hence greatly improved photoelectrochemical performance for water splitting, with IPCE increased by a factor of 4.5 from ~2.0% to 9.0% at 400 nm. In the α-Fe2O3/AgxFe2-xO3 nanorod structure, the surface doping of Ag+ ions gave rise to increased electron donor density, which also led to the enhancement in photoelectrochemical performance with IPCE at 400 nm increased from ~2.0% to ~7.5%.

In an inversed two-photon driven (tandem) device, Si can be used as the small band gap photoanode material to drive the oxygen evolution reaction (OER). However, Si-based photoanodes are inherently unstable in the highly oxidizing environment during OER. Therefore, suitable strategies have to be developed to protect the Si substrate. In this contribution the applicability of Si-based small band gap photoanodes modified by sputter deposited NiO thin films will be presented. It is shown that NiO enables long term stability of Si-based photoanodes in highly alkaline conditions by combined electrochemical measurements with quartz crystal microbalance and ICP-MS studies.^[1] Furthermore, treatment of the as-deposited NiO thin films in Fe-containing electrolyte significantly enhances the photoelectrochemical performance of the photoanode assembly while maintaining its stability. Thus, current density of 10 mA/cm² (requirement for >10% efficient devices) could be obtained at 1.15 V vs. RHE (Reversible Hydrogen Electrode) under red-light (38.6 mW/cm²) irradiation. These results places the Fe-treated NiO protected Si photoanodes among the best performing Si-based photoanodes in alkaline media. Finally the applicability of Ir/IrO_x thin films for the protection against corrosion in acidic electrolytes will be discussed.^[2]
[1] B. Mei, A. A. Permyakova, R. Frydendal, D. Bae, T. Pedersen, P. Malacrida, O. Hansen, I. E. L. Stephens, P. C. K. Vesborg, B. Seger, I. Chorkendorff, JPC Letters 5, 3456 (2014). [2] B. Mei, B. Seger, T. Pedersen, M. Malizia, O. Hansen, I. Chorkendorff, P. C. K. Vesborg, JPC Letters 5, 1948 (2014).

The advances in our understanding of nature largely rely on the development of measurement techniques. At microscopic scale in which many basic chemical and physical processes take place, devices composed of single nanowire represent a unique platform allowing measurement possibly beyond the limit of ensemble average. This is in particular helpful in the field of artificial photosynthesis. Semiconductor nanowires can be a good candidate for solar-to-fuel conversion, and moreover is an ideal platform to look into the photoelectrochemcial (PEC) process at microscopic level. In this work we demonstrate such a well-controlled single nanowire device, with silicon (Si) as the model semiconductor light-absorber and platinum (Pt) as the hydrogen evolution reaction (HER) catalyst. The device served as a robust platform, whose photovoltage could be reproducibly modulated by the surface doping profile at the interface. Moreover as a model system we can quantify at single nanowire surface that the photogenerated electron flux was much reduced to about 10 electrons/ nm².s at 0 V vs. reversible hydrogen electrode (RHE), illustrating the benefits of nanowire geometry in terms of reduced electrochemical overpotential. This report extends the single-nanowire measurement technique into solar-to-fuel processes in solution phase, and invites further fundamental discoveries at such a microscopic level.

Achieving stable operation of high efficient photoanodes used as components of solar water splitting devices is critical to realizing the promise of this renewable energy technology. Here, we show that a p-type transparent conducting oxide (p-TCO) can function as a selective hole conduct and corrosion protection layer on photoanodes used for light-driven water oxidation. NiCo₂O₄ with inverse spinel structure was used as the p-TCO, and we show that this material has the requisite electronic structure, stability, transparency, and hole conductivity to achieve sustained and efficient solar water oxidation.
NiCo₂O₄ was deposited on n-Si (100) and np-Si (100) by reactive sputtering at substrate temperature of 70-100 °C. No additional post annealing was needed. he film was not crystalline based on XRD and Raman measurement. P-type conductivity was confirmed by Seebeck. The conductivity was 50-60 S/cm. Light transparency of >70% (wavelength of 400 nm and beyond) was achieved with a NiCo₂O₄ thickness of 40 nm. Photoelectrochemical (PEC) evaluation was performed in aqueous 1M NaOH (pH 14) with simulated AM1.5 illumination; these conditions would rapidly corrode the n-Si photoanodes in the absence of a protection layer.
As expected, p-NiCo₂O₄ forms a rectifying contact to n-Si. PEC performance of the np-Si/p-NiCo₂O₄ structures is excellent, particularly when a thin NiFe oxygen evolving catalyst is applied. An onset potential of 0.95 V vs. RHE is observed, which is one of the lowest reported for a Si-based photoanode. The current density at the reversible potential for water oxidation (1.23 V vs. RHE) is >25 mA cm^-2, and the current rises to a limiting value of 30 mA cm^-2 at more anodic potentials, which demonstrates the attractive combination of transparency and low-resistance hole conductivity in the NiCo₂O₄. Long-term testing indicates multi-day stability with minimal decrease in performance or observable corrosion of the Si photoanode. This work demonstrates that p-TCOs are promising as corrosion protection layers for stable water oxidation photoanodes. In depth characterization of both the solid-solid interfaces between NiCo₂O₄ and light absorber/catalyst and of the solid/electrolyte interface will be discussed
This material is based upon work performed by the Joint Center for Artificial Photosynthesis, a DOE Energy Innovation Hub, supported through the Office of Science of the U.S. Department of Energy under Award Number DE-SC0004993.

Photoelectrochemical (PEC) water splitting could provide a sustainable means of hydrogen fuel production.¹ Recent efforts in PEC water splitting have focused on developing materials suitable for application in a dual-absorber device configuration due to the high solar-to-hydrogen efficiencies this architecture could enable.^{2, 3}
Silicon is a promising candidate small band gap absorber material for application in a tandem PEC device due to its abundance, relatively low cost, excellent charger carrier transport, and near-ideal band gap. A number of recent studies have reported promising silicon photocathode structures.^4-6 However, several challenges remain to be addressed to develop silicon photocathodes that are as efficient and stable as necessary for economical water splitting. The silicon must be combined with an active catalyst to reduce the kinetic overpotential necessary to drive the hydrogen evolution reaction (HER) at the electrode surface.^4-6 The silicon must also be protected to prevent corrosion or oxidation, which can destroy device performance.⁶
We discuss our recent progress in designing active and stable silicon photocathodes. First, we show that molybdenum sulfide nanomaterials can provide both corrosion protection and catalytic activity when integrated into silicon photocathodes.⁷ Using a thin MoS₂ surface layer, we create molybdenum sulfide/silicon electrodes that show no loss in performance after 100 hours of operation. Transmission electron microscopy measurements reveal the atomic scale features of this electrode design that result in its excellent performance. To further improve this cathode&’s efficiency, we incorporate a second molybdenum sulfide nanomaterial, highly catalytically active [Mo₃S₁₃]^2- clusters,⁸ resulting in precious metal-free devices with photocurrent onset potentials within ~150 mV of the best reported Pt/Si photocathodes. We discuss further improvements in the design of silicon photocathodes to maximize the photocurrent onset potential and device durability including the incorporation of new, highly active HER electrocatalysts. Based on our findings, we propose strategies for further improving the performance of water splitting photocathodes.
1. M. G. Walter, et al., Chemical Reviews, 110, 6446 (2010).
2. M. F. Weber, et al., Journal of The Electrochemical Society, 131, 1258 (1984).
3. L. C. Seitz, et al., ChemSusChem, 7, 1372 (2014).
4. J. R. McKone, et al., Energy & Environmental Science, 4, 3573 (2011).
5. S. W. Boettcher, et al., Journal of the American Chemical Society, 133, 1216 (2011).
6. B. Seger, et al., Journal of the American Chemical Society, 135, 1057 (2013).
7. J. D. Benck, et al., Advanced Energy Materials (2014).
8. J. Kibsgaard, et al., Nat Chem, 6, 248 (2014).

Most moderate-band gap semiconductors are prone to photocorrosion in aqueous solution, and developing methods to stabilize them as photoelectrodes is an important step towards efficient integrated solar fuels devices. Amorphous TiO₂ films deposited with atomic layer deposition on various n-type semiconductors (Si, CdTe, and GaP) have recently been shown by our group to form rectifying junctions and to achieve efficient photo-driven water oxidation for >100 hours in highly alkaline electrolytes (1, 2). In these previous studies, the TiO₂ was postulated to be conductive to anodic current because of mid-gap defect energy states arising from either its amorphous structure or extrinsic impurities. To improve the stability and junction characteristics of TiO₂ on n-type semiconductors, it is necessary to develop a detailed understanding of the properties of TiO₂ that give rise to good performance in n-type semiconductor/TiO₂ photoanodes. Here, we show that in addition to amorphous TiO₂, crystalline TiO₂ thin films (both anatase and rutile) deposited with various methods (ALD and sputtering) can also form high-performance rectifying junctions with n-Si; the structure of these films and the Si/TiO₂ interface is characterized with transmission electron microscopy, Raman spectroscopy, and x-ray diffraction. The photovoltage generated by Si/TiO₂/Ni photoelectrodes in aqueous solution is shown to be controlled by annealing the devices in different environments (reducing vs oxidizing); such behavior is attributed to the shifting of the TiO₂ Fermi level and the introduction of additional defect states within the TiO₂ band gap. X-ray photoemission valence measurements show that the distribution of defect states within the band gap of the various TiO₂ materials is different, yet they conduct in similar ways. Finally, all deposition methods that produce conformal films stabilize Si during photo-driven water oxidation. Overall, these results show that the properties of TiO₂ that allow for conduction of carriers from the valence band of n-Si through the TiO₂ are more general than previously thought, and we also show that a variety of easily-fabricated TiO₂ materials can be fabricated and utilized for semiconductor protection in efficient solar fuels generators.
1. S. Hu, et al., Science 2014, 344 (6187), 1005-1009.
2. M. F. Lichterman et al., Energy & Environmental Science 2014, 7, 3334-3337.

Hydrogenation of titania introducing surface disorder and oxygen vacancy on it had an impact on the development of a next level metal oxide photocatalyst [1]. It was believed that increased charge conductivity and appeared visible/infrared light absorbance by the hydrogenation lead a significant improvement of photoconversion ability to generate a sustainable energy such as hydrogen, methane and methanol. Li and co-workers found, however, the hydrogenated titania (H-TiO₂) exhibited extremely low photoconversion efficiency in visible light region (below 1% IPCE) whereas that in ultraviolet light region increased five-fold compared to an untreated titania [2]. This indicates the color change of the H-TiO₂ from white to black was caused by a creation of mid-gap electronic state due to Ti-H bond, not by an up-shifting of valance band maximum (VBM) of a titania [3]. Hence the proper band-gap engineering enabling the H-TiO₂ to utilize visible light could provide an opportunity to overcome photoconversion limits of the wide band-gap oxide photocatalysts.
In this presentation, we report simultaneous hydrogenation and nitrogenation of titania (HN-TiO₂) by using mixed gas plasma treatment to increase visible light photoactivity as well as charge conductivity. The color of a titania was changed to dark yellow through the controlled hydrogen and nitrogen plasma treatment, whereas the colors of H-TiO₂ and N-TiO₂ were dark gray and yellow, respectively. The incident photon-to-current efficiency (IPCE) result of the HN-TiO₂ has proved broader visible light photoactivity over the 500 nm range, where a shape of the IPCE graph was similar to the trend of absorbance spectrum. The electronic states of titanium, oxygen and nitrogen elements in the HN-TiO₂ have indicated Ti-H and Ti-N bond were created on the surface of titania, and thereby its VBM level was up-shifted about 2 eV compared to the untreated titania. In addition, it was found a lot of small dots of about 5 nm size were created on the surface of titania nanopaticles by a contol of the plamsa treatment, and these dots could affect to facilitate charge separation during the photoconversion reaction. As a result, the HN-TiO₂ has exhibited increased hydrogen and methanol evolution rarte more than seven-fold compared to the untreated titania, under air mass 1.5G radiation.
[1] Xiabo Chen et al., Science 331, 746, 2011
[2] Gongming Wang et al., Nano Lett. 3026, 11, 2011
[3] Zhou Wang et al., Adv. Funct. Mater. 5444, 23, 2013

Our investigations with silane-modified TiO₂ have revealed a beneficial effect of functionalization on the photoelectrochemical performance on spin-coated electrodes. However, in order to produce large area photoelectrodes, a more scalable manufacturing technology is required. Inkjet printing can fulfil this role and furthermore allow a finer control over coating morphologies.
In this work, a series of inkjet-printed photoelectrodes were prepared with silane-functionalized TiO₂ nanoparticles, and investigated as electrodes for photoactivated water splitting. The catalyst layers, having thickness between 300 and 800 nm, were printed on FTO-coated glass supports, from cellulose stabilized dispersions. For comparison, electrodes of similar thicknesses were also prepared by spin-coating. After removing the stabilizer at 300 °C under air atmosphere, the electrodes were characterized in photoelectrochemical cells containing 0.5 M H₂SO₄ as electrolyte and a platinum ring as counter electrode.
Under simulated sunlight, the best photocurrent densities for the oxygen evolution reaction were obtained for the inkjet-printed electrodes prepared with functionalized particles (up to 0.25 mA cm^-2 at 1 V(SHE), compared to 0.16 mA cm^-2 for pristine TiO₂). Microscopy of the printed electrodes shows structurally homogenous coatings with evenly distributed roughness. The electrodes were further characterized by infrared, ultraviolet and Raman spectroscopies, and by electrochemical impedance spectroscopy. Under continuous illumination at 0.7 V(SHE), the electrodes showed no significant drop in photocurrent within five hours.

In this work, an innovative approach to the fabrication of large-area photoelectrodes, namely cold gas spraying of metal oxide particles, is presented. Conventional large area coating techniques usually employ wet chemical methods with subsequent calcination steps to obtain enhanced bonding between the catalyst particles and the substrate. In the cold gas spraying process, particles are accelerated to high velocities (up to 1200 m/s) by a pressurized gas, which is preheated and then expanded in a de Laval type nozzle. On impact with the substrate, the kinetic energy of the particles is converted to a short heat spike by deformation, fracture and fusion processes, and an intimate interface to the back contact is formed.
For a first demonstration, TiO₂ aggregates (5 - 20 µm) were evaluated for the preparation of TiO₂ photoelectrodes (3 x 3 cm²). These TiO₂ aggregates were supplied by EVONIK and were prepared via spray drying of P25 particles. In XRD measurements they reveal 90 % anatase and 10 % rutile. In photoelectrochemical experiments, these cold gas sprayed TiO₂ photoelectrodes show seven times higher photocurrents in the photooxidation of water (at 1.23V(SHE) [1]) than reference electrodes prepared by the established doctor blade technique. This is explained by a beneficial bonding and interaction between the TiO₂ particles and the titanium substrate, established due to the impact.
In this contribution, we investigate the influence of the carrier gas employed in the cold gas spray process on the photoelectrochemical properties of the resulting TiO₂ films. In the application of different gases the temperature and the velocity of the particles can be controlled in the spray process. The TiO₂ aggregates have a higher impact velocity and thus higher kinetic energy if helium is used as carrier gas as compared to nitrogen or argon. Nevertheless the highest IPCE values are observed for TiO₂ films sprayed with nitrogen as carrier gas among the tested gas carriers. Detailed physico-chemical and electrochemical characterization indicate the preservation of the crystalline structure but also the formation of interband states at the TiO₂ surface. In the XPS analysis of N₂-TiO₂ films, covalent bonded nitrogen was found on the surface.
Furthermore, first results on the preparation of cold sprayed electrodes based on other metal oxides (e.g. WO₃) are presented.
Reference:
1. I. Herrmann-Geppert, P. Bogdanoff, H. Gutzmann, T. Dittrich, T. Emmler, R. Just, M. Schieda, F. Gärtner and T. Klassen, ECS Transactions, 58 (30) 21-30 (2014)

Hydrogen fuel is a clean alternative power source for applications that span all energy consumption sectors: industrial, commercial, residential, and transportation. Electrochemical production of H₂ can also serve as an energy storage mechanism for excess electricity production from renewables. Classical electrolysis systems tend to be large in scale, posing challenges for their implementation, especially for applications that require portability. In this work, we present novel microfluidic techniques that can be used to fabricate electrolyzers at length scales which are attractive across the spectrum of applications. Power generation for portable electronic devices can be enabled by microfluidic water-splitting devices, and large scale fuel production can be achieved by panels with parallelized microsystems. Moreover, microscale devices have several advantages over macroscale systems, as transport processes in the devices can be easily controlled and tuned to improve device performance.
Here, we report two microfluidics platforms for water splitting. The first, a membrane-less microelectrolyzer that uses laminar flow principles in microchannels to separate the product gases without the need for an ion conductive membrane. The inertial lift force due to the hyperbolic shape of electrolyte&’s velocity profile, keeps the two gas streams separated until they reach different collection outlets. These membrane-less devices achieve current densities as high as 147 mA/cm² at an efficiency of 47%, and can be operated under any water-based electrolyte; including acidic, basic or near-neutral solutions. The crossover of gasses in the device is low, allowing it to continuously generate nearly-pure H₂ gas streams with O₂ concentrations below the 4% flammability limit. The second device presented is an air-based microelectrolyzer which absorbs water from ambient air humidity in a Nafion® thin-film where water electrolysis takes place. The evolved gases later diffuse through the polyelectrolyte film to be collected in separate microchannels. The large surface to volume ratio of the device provides efficient humidity absorption form air, and current densities above 3 mA/cm² can be reached under stable operational conditions. Effect of various factors such as Nafion film&’s thickness, air convection speed, and its water content are studied, providing guidelines for the development of high-current density vapor-fed water-splitting devices. The development of photoelectrodes for the fabrication of optofluidic microchips capable of spontaneously producing pure Hydrogen from sunlight and water will also be discussed.

Titanium oxide photoelectrodes have been used for water splitting for a few decades since titania is stable and low-cost, but have low solar-to-hydrogen efficiencies. Perovskite halides (e.g., CH₃NH₃PbX₃) have recently emerged as an efficient light absorber system, leading to solar cell efficiencies above 17%. The combination of the two materials offers a new opportunity to achieve higher efficiency for hydrogen production. Photoelectrodes of titania films and of composite films of titania and perovskite halite, respectively, are prepared onto the substrate of ITO glass. The perovskite films are produced by spin-coating of solution-processable organic-nonorganic chemical reaction of methylammonium iodide with lead iodide. The TiO₂ thin films are deposited on ITO glass by RF magnetron sputtering of TiO₂ ceramic target or by spin coating. Since the perovskite films are sensitive and reactive with water, composite films of titania and perovskite are designed to protect the perovskite from water contact. The films are characterized by X-ray diffraction, Scanning electron microscopy, and Ultraviolet-Visible spectroscopy. The two types of photoelectodes are investigated for water-splitting hydrogen production under UV or solar light irradiation by photoelectron chemical (PEC) reaction. The hydrogen production rate using anatase TiO₂ electrode is 2556 mu;mol h^-1m^-2 under UV light, but there is nearly no reaction using Xe lamp as irradiation source. Since perovskite films of (CH₃NH₃)PbI₃ are favorable light harvesters under UV and visible light irradiation, and lambda; is between 350nm and 800 nm, the composite films of titania and perovskite would achieve efficient water splitting using Xe lamp as irradiation source.

Protective TiO₂ coatings grown by atomic layer deposition have been shown to stabilize group IV, III-V, and II-VI semiconductors for photoelectrochemical water oxidation, yet relatively little is understood about the mechanisms of conduction and protection. Initial work on the conduction mechanism for ultrathin TiO₂ protection layers on Si showed temperature-independent conduction suggesting direct tunneling through the SiO₂/TiO₂ bilayer [1]. Investigations of thicker TiO₂ films conversely showed temperature-dependent conduction and revealed a linear relation between overpotential and ALD-TiO₂ thickness. Polaronic hole-hopping via traps in the TiO₂ was proposed as a conduction mechanism that matched the empirical data and further suggested that if the traps could be manipulated, the conduction could be tunable as well [2]. This same conduction pathway, utilizing traps approximately 1eV below the TiO₂ conduction band edge, may also be responsible for the highly conductive TiO₂ films over 100 nm in thickness now being used to protect a variety of substrates for water oxidation [3]. Overlooked thus far, however, has been the role of interfacial oxides expected to behave as tunnel junctions in series with the “electrically leaky” TiO₂. This work shows that the interfacial SiO₂ in metal/SiO₂/Si and metal/TiO₂/SiO₂/Si water splitting anodes does behave as a tunnel oxide and, therefore, the overall device efficiency is particularly sensitive to the properties of this layer. ALD-Al₂O₃ is also studied using bilayer Al₂O₃/SiO₂ as an analog to the leaky and asymmetric TiO₂/SiO₂ system of greatest interest, to further elucidate the differing conduction mechanisms. If the TiO₂ conductivity can be made sufficiently high, as shown in recent reports, overall device efficiency will depend most significantly on small changes in the interfacial oxide, meaning studying this layer is of great importance both for understanding the conduction mechanisms and for realizing reliable devices of optimal efficiency [4].
[1] Y.W. Chen, J.D. Prange, et al. Nature Materials,10, 539-44 (2011).
[2] A.G. Scheuermann, et al. Energy Environmental Science, 6, 2487 (2013).
[3] S.Hu et al. Science, 344, 1005 (2014).
[4] A.G. Scheuermann, et al. ECS Transactions, 64, 265 (2014).

The conversion of solar energy into chemical energy in the form of hydrogen via photocatalytic water splitting is an abiding challenge for production of clean, storable and sustainable energy carriers. Z-scheme systems, which employ two different semiconductors for H₂ and O₂ evolution, are capable of utilizing visible light more efficiently than a single-component photocatalyst because the energy required for driving each photocatalyst can be reduced. A major challenge in developing high-performing Z-scheme water splitting systems lies in ensuring efficient transfer of electrons between H₂ and O₂ evolution photocatalysts without deteriorating their intrinsic photocatalytic properties.
Herein, we report a Z-scheme system consisting of H₂ evolution photocatalyst (HEP)/metal layer (M)/O₂ evolution photocatalyst (OEP), which utilizes the metal layer for electron transport, taking SrTiO₃:La,Rh/Au/BiVO₄ as a prototype. SrTiO₃:La,Rh/Au/BiVO₄ (HEP/M/OEP) plates were prepared by a particle transfer method. SrTiO₃:La,Rh/Au/BiVO₄ systems exhibit photocatalytic activities for overall water splitting that are 6 and 20 times higher than powder suspensions and SrTiO₃:La,Rh/BiVO₄ systems without metal layers, respectively. The SrTiO₃:La,Rh/Au/BiVO₄ systems achieve an apparent quantum yield of 5.9% under 418-nm monochromatic light irradiation and a solar-to-hydrogen conversion efficiency of 0.2%. The high performance of this system is attributable to the presence of the Au layer, which effectively transferred photogenerated electrons from BiVO₄ to SrTiO₃:La,Rh. Because the conductive metal layer anchors the respective particulate photocatalysts firmly, the activity of SrTiO₃:La,Rh/Au/BiVO₄ is almost independent of the pH of reaction solutions. This observation is distinct from the conventional Z-scheme water splitting systems that necessitate the aggregation of particles by electrostatic attractive force for the interparticle electron transfer.
The concept of HEP/M/OEP devices can not only produce enhanced photocatalytic activity in Z-scheme water splitting but also overcome the limitations of conventional Z-scheme systems, which necessitate the addition of redox couples, aggregation of photocatalyst particles, or formation of composites for electron relay, thus extending the kinds of photocatalytic materials applicable to the reaction.

The Ni/NiO core/shell structure is one of the most efficient co-catalysts for solar water splitting when coupled with suitable semiconducting oxides. It has been shown that pretreated Ni/NiO core/shell structures are more active than pure Ni metal, pure NiO or mixed dispersion of Ni metal and NiO nanoparticles.^{1, 2} However, Ni/NiO core/shell structures on TiO₂ are only able to generate H₂ but not O₂ in aqueous water. Here we investigate nanoscale structure-reactivity relations and the de-activation and photocorrosion process taking place under light illumination in pure liquid water.
Ni/NiO core/shells on TiO₂ were prepared and tested for photocatalytic/photochemical reactivity. H₂ evolution was detected but no O₂ was produced. Inductively coupled plasma mass spectrometry (ICP-MS) was utilized to determine Ni²⁺ concentration in the solution before and after reactions. An increase of Ni²⁺ in the water was detected and matched the amount of H₂ produced. It was found that the core/shell structure plays an important role for H₂ generation but the system undergoes deactivation due to a loss of metallic Ni. The H₂ evolution was generated by a photochemical reaction which involved photocorrosion of Ni metal. The nature of the hydrogen evolution reaction in these systems was investigated by correlating photochemical H₂ production with atomic resolution structure determined with aberration corrected transmission electron microscopy. During the H₂ evolution reaction, the metal core initially formed partial voids which grew and eventually all the Ni diffused out of the core-shell into solution leaving an inactive hollow NiO void structure. This photocorrosion occurred either due to direct contact with the water through cracks or a Kirkendall effect where Ni diffused along grain boundaries in the NiO shell onto the particle surface where dissolution took place. The morphological changes only took place in the presence of light suggesting a room temperature diffusion process driven entirely by light illumination. The detailed morphology and mechanism will be discussed.
References:
1.Domen, K. , Kudo A. and Onishi, T. J. Catal.,1986, 102, 92-98.
2.Townsend, T. K.; Browning N. D. and Osterloh, F. E. Energy Environ. Sci., 2012, 5, 9543
3.The support from US Department of Energy (DE-SC0004954) and the use of ETEM at John M. Cowley Center for HR Microscopy at Arizona State University is gratefully acknowledged.

Highly ordered niobium oxynitride microcones with nano-wire like structure were formed by two step synthesis: anodization and then nitridation in ammonia atmosphere for one hour. The morphological, structural and optical properties were studied for the oxynitride and the oxide forms. XPS analysis confirmed the formation of oxynitride microcones. Nitridation shifts the absorption edge from 450 nm for the oxide form to 550 nm for the oxynitride form. The microcones were utilized in solar-spectrum water photoelectrolysis, showing 52% increase in the photocurrent compared to that of the oxide counterpart. This enhancement that is found to be the highest reported for niobium oxide photocalaysts is believed to be due to the highly ordered structure and the nitrogen incorporation. Theoretical calculations for the absorption and the photoconversion efficiency performed using Comsol Multiphysics software confirmed the effect of the ordered structure of microcones on enhancing the properties of the niobium oxide photocatalysts.

The hematite has come up as promising candidate to provide a clean, unlimited, sustainable and renewable energy free from carbon produced by photoelectrochemical (PEC) cells. Among the various routes for obtaining hematite, one that has drawn much attention is the hydrothermal route, which is a versatile and controlled way to produce one-dimensional (1D) nanostructure. It's worth to mention that the hematite prepared from this method is preceded by the akaganeite formation. For instance, the akaganeite crystalline structure has the ability to trap ions, such as chloride ions from reaction medium. This means that during the thermal treatment (to obtain hematite phase) chloride ions are eliminated as gas changing the crystallographic arrangement and it fundamental properties. This work describes the influence of trapped ion change (Cl^-, OH^-, Br^- and F^-) on the akaganeite structure and its effect on hematite properties obtained after the thermal treatment. The trapped ion change on the akaganeite structure was made with an ionic salt solution at 55 °C for 2 hours keeping constant stirring. After the trapped ions change, akaganeite was thermal treated at 750 °C/30min to obtain hematite phase. It was found by using several techniques that the akaganeite prepared with different ions has strong influence on the final properties of nanostructured hematite films. From X-ray diffraction was identified the hematite phase with highly conductive (001) basal plane orientation, which is perpendicular to the substrate for all films. As consequence of this modification the photocurrent response increase 15% with Br^- compared to as prepared (Cl^- containing), while with F^- or OH^- the photocurrent reduces. Furthermore, the Br^- ion change increases the electrochemical stability illustrated on the chronoamperometry measurements. Finally, a film prepared in presence of Br- ions was observed an enhancement of the donor density (N_D) and reduction of contact angle (high wettability). These results suggest that the akaganeite structure with different ion trapped can improve the hematite/liquid interface.
Acknowledgements
We gratefully acknowledge #64257;nancial support from the Brazilian agencies of FAPESP (Grants 2011/19924-2, 2012/19926-8 and 2013/05471-1), CAPES, CNPq (Grants no. 473669/2012-9), Instituto Nacional em Eletro#770;nica Orga#770;nica (INEO), NanoBioMed Brazil Network (CAPES), CEM- UFABC and CDMF (Grants no. 2013/07296-2).

A pyrochlore based bismuth titanate (BTO) photocatalyst with Fe as an earth-abundant additive (Fe-BTO) has been examined in the presence of UV-vis illumination of water-methanol (sacrificial agent) system. The performance of the Fe-BTO was investigated by examining the effects of catalyst loading, light intensity, methanol concentration, and catalyst stability upon repeated use. Among the parameters evaluated, the hydrogen yield of 37 mL g^-1 using 150 mg of catalyst was determined effective with a methanol concentration of 20M. In additional insights, all of the time resolved experiments indicated the existence of a methanol concentration dependent 2-zone region: zone 1- slow photoactivity and zone 2 - accelerated photoactivity. The presence of the 2-zone region is attributed to the intermediates formed during the methanol oxidation. The formation of the intermediate formic acid, which is thermodynamically favored, is identified as one of the key stages in the reaction.. Further, repeated use of the photocatalyst leads to over a 70% loss in the Fe-BTO photoactivity due to the formation of surface-functional groups. The surface functional groups can be removed by a simple oxidative surface treatment (generally a thermal - based approach) to recover the activity of the photocatalyst without impacting the surface or physical properties of Fe-BTO.

2D layered transition metal dichalcogenide (TMD) materials such as MoS₂ and WSe₂ have received much attention as alternatives to Pt for hydrogen evolution reaction (HER) catalysts. It is well known that catalytic performance of MoS₂ is correlated to the exposed surface area of catalytically active edge-sites. However, correlation of gradual electronic transition states of TMD hybrids and catalytic behavior has been rarely studied. Here, we synthesized MoS₂/rGO hybrid catalysts with enhanced charge transfer and conductivity of electrocatalyst. We synthesized the various types of MoS₂/rGO with tunable size and morphology by controlling amount of graphene oxide (GO) during solvothermal synthesis of MoS₂. Furthermore, MoS₂/rGO hybrids shows gradual changes of electronic structures from 2H to 1T depending on the amount of GO during the synthesis. We studied catalytic performance of the MoS₂/rGO hybrids depending on the morphology and electronic structures, and found that the appropriate amount of GO can transform MoS₂ to 1T phase with large exposed edge-sites. The best MoS₂/rGO hybrids shows a Tafel slope of 45mV/dec which was 6mV/dec higher than Pt catalyst in our test.

The reaction cycles of water-splitting based on redox reactions for sodium Na are composed of four reactions, which are hydrogen generation by solid-liquid reaction (1), metal separation by thermolysis (2), oxygen generation by hydrolysis (3), and phase transition of Na metal (4).
2NaOH(s) + 2Na(l) → 2Na₂O(s) + H₂(g) (1)
2Na₂O(s) → Na₂O₂(s) + 2Na(g) (2)
Na₂O₂(s) + H₂O(l) → 2NaOH(s) + 1/2O₂(g) (3)
Na(g) → Na(l) (4)
The Na cycle theoretically requires more than 1000 °C in thermodynamic equilibrium condition, however, the operating temperature is reduced to below 500 °C by nonminus;equilibrium techniques using phase transition of metal Na vapor. The thermodynamic analyses are performed by using the parameters such as operating temperature and partial pressures of the products obtained by the experiments to determine that the sodium redox cycles are potential hydrogen production technique as thermochemical energy storage.
References:
[1] H. Miyaoka et al, Int. J. Hydrogen Energy, 2012, 37, 17709-17714
[2] H. Miyaoka et al, Energy Procedia, 2014, 49, 927-934

We have grown GaN layers alloyed with GaAs, GaSb and GaBi compounds using plasma-assisted molecular beam epitaxy (PA-MBE) and extensively characterized their structural, optical and electrical properties. Electronic band structures of these so-called highly mismatched alloys (HMAs) are described by the band anticrossing (BAC) model which predicts that the alloys should exhibit a wide range of direct energy gaps. We have shown previously that the energy gap of GaN_1-xAs_x alloys varies from 0.7eV to 3.4eV. An even larger modification of the band structures is anticipated for more extremely mismatched GaN_1-xSb_x and GaN_1-xBi_x alloys. The large band gap range and controllable conduction and valence band edge positions makes the HMAs promising materials for efficient solar energy conversion devices. These HMAs may be suitable for solar water splitting applications for hydrogen production. As efficient photoelectrodes, the bandgap of the semiconductor must be >2 eV to induce electrochemical decomposition of water but still small enough to absorb a significant portion of the solar spectrum. In addition the band edges must also straddle the H₂O redox potentials.
At dilute doping levels, substitutions of As, Sb and Bi into the N sublattice results in formation of localized energy levels above the valence band in GaN. Our measurements on GaN doped with As and Sb have demonstrated that the As and the Sb impurity levels lie at about 0.7eV and 1.2eV above the valence band edge of GaN, respectively.
The BAC model predicts that at a higher concentration of the group V elements the interaction of the impurity levels with the extended states of the valence band leads to formation of an impurity-derived, fully occupied narrow band that plays a role of the new valence band edge. This results in an abrupt upward shift of the valence band edge and a reduction of the optical gap of the HMAs. We have achieved the enhanced incorporation of As, Sb and Bi by growing the layers at extremely low temperatures (down to about 100^oC). Although the layers become amorphous for high As, Sb and Bi content, the measured composition dependence of the optical absorption edge are consistent with the predictions of the BAC model, indicating that the amorphous HMAs samples have a short-range order resembling random crystalline alloys. The large band gap range and controllable positions of the conduction and valence bands make these HMAs promising materials for solar water splitting applications.

Photoelectrochemical (PEC) cell is a reliable device for harnessing and converting abundant solar energy into hydrogen by making use of green path [1]. Its been a challenge to identify a perfect semiconducting material, possessing all desirable properties of a photoanode for a PEC cell. Cadmium sulphide is one of the most suitable and studied PEC material. This is due to its capability to absorb the visible light photons (E~1.5eV - 3eV), and its well suited band-edge positions to facilitate the water-splitting reaction (H₂O→H₂+O₂). However, the dissolution of CdS in electrolyte during the photo-illumination (termed as photocorrosion) is a major drawback [2]. One of the practical methods to inhibit such photocorrosion over the semiconductor surface is loading of the water oxidation co-catalyst (WOC). The role of WOC&’s lies in increasing the water oxidation kinetics; which then releases the excess photogenerated holes on the semiconductor surface. In past, inert metal (Pt, RuO₂, Ru) modified CdS surface had been used to limit the unwanted photocorrosion [3]. Commercially, usage of such materials is un-economic option for any technological usage. Recently, transition metal-hydroxides and metal-oxides (M-OH; M=Co, Ni, Cu, Mn) are emerging as cost-effective, abundant and robust WOCs. In present work, we have loaded nano-sized metal-hydroxides on the surface of CdS by chemical impregnation method. Such modified electrodes show an enhanced photocurrent, as well as stability that improved to several hours. The work shows promise of WOC&’s in achieving an improvement in the performance and stability of the PEC cell for desirable technological applications.
References:
[1] A. Fujishima and K. Honda, Nature, 1972, 238, 37.
[2] D. Meissner, R. Memming and B. Kastening, Chem Phys. Lett., 1986, 127, 5
[3] K. Kalayanasundaram, E.B. Borgarello, D. Duonghong an dM. Graetzel, Angew. Chem. Int. Ed. Engl., 1981, 20, 987.

Control of the amounts of exposed edge sites of molybdenum disulfide (MoS₂) is the key factor to develop the efficient catalyst for hydrogen evolution reaction (HER). Here, we describe, for the first time, a chemistry for growing MoS₂ by atomic layer deposition (ALD) and how to control the amounts of edge sites of a nanoporous MoS₂ film. We ascribed that the non-ideal mode of ALD growth on planar surfaces could be used in controlling the relative fractions of active-edge sites of MoS₂. Moreover, we investigated how the amounts of edge sites affect to the on-set potentials and the current densities in the water splitting performance. We achieved the exchange current densities for HER up to 20 mA cm^-2 at -0.3 V versus reversible hydrogen electrode, indicative of our ALD approach as a novel strategy in designing HER materials.

Water and sunlight playing in harmony in presence of semiconductor, such as iron oxide (α-Fe₂O₃), can provide a clean, unlimited, sustainable, and renewable energy free from carbon produced by photoelectrochemical (PEC) cells. Although α-Fe₂O₃ meets many requirements, the efficiency reported until now is far from theoretical prediction. In the last years, our research group has developed and applied hematite vertical nanorod to split water into molecular hydrogen and oxygen. This talk will highlight our recent results focusing on the use of electrochemical impedance, photoelectrohemical and electronic transmission microscopy to investigate the interfaces solid/liquid and solid/solid (iron oxide film/substrate adherence) and their impacts on hydrogen generation from water splitting. Vertically oriented iron oxide photoelectrodes were synthesized in a single step under hydrothermal conditions. The enhancement on photoeletrocatalytic activity and concomitant reducing of required overpotential was achieved by effective control-treatment of morphology and surface of vertically aligned hematite nanorods. For instance, undoped and tin oxide-doped hematite electrodes showed a significant improvement on the photocurrent values from 1.0 to 1.6 and from 2.2 to 2.6 mA/cm2 at 1.23 V against H2 reference electrode. These findings demonstrated that the temperature plays an important role on the iron oxide (structural, morphological, and catalytic) properties and that many influences must work in great harmony in order to produce a promising hematite photoanode
Acknowledgements
We gratefully acknowledge #64257;nancial support from the Brazilian agencies of FAPESP (Grants 2011/19924-2 and 2012/19926-8), CAPES, CNPq (Grants no. 473669/2012-9), Instituto Nacional em Eletrocirc;nica Orgacirc;nica (INEO), NanoBioMed Brazil Network (CAPES), and CMDF (2013/07296-2).

Photocatalytic and photoelectrochemical water splitting under sunlight has been studied as a means of large-scale production of renewable hydrogen. It is necessary to harvest long-wavelength visible light to achieve sufficient solar-to-hydrogen conversion efficiencies with reasonable quantum efficiencies. In this talk, recent progress in the water splitting reaction on various forms of semiconducting materials is presented.
A semiconductor photocatalyst can generate both hydrogen and oxygen on the surface when the band gap straddles the reduction and oxidation potentials of water. Some (oxy)nitride semiconductors have band gap the potential of which is suitable for overall water splitting under visible light. (Ga_1-xZn_x)(N_1-xO_x) and ZrO₂-modified TaON are representative examples. However, the potential of those materials for solar energy conversion is limited because their absorption edge wavelengths are shorter than 500 nm. Recently, it was found that LaMg_1/3Ta_2/3O₂N could be activated in the overall water splitting reaction under visible light irradiation up to approximately 600 nm by appropriate surface modifications to suppress the self-oxidation.
Photocatalytic materials are applicable to photoelectrochemical water splitting when being immobilized on a conductive substrate. In photoelectrochemical reactions, an external voltage can be applied to compensate for the potential deficiency to drive redox reactions on a counter electrode. Alternatively, a photoanode and a photocathode can be connected in series. In such a tandem configuration, the working photocurrent and the potential are determined by the intersection of the steady current-potential curves of the respective photoelectrodes. Therefore, it is important to develop photoelectrodes that generate high photocurrent with a small applied voltage.
The author&’s group developed the particle transfer method for fabrication of electrodes of particulate photocatalysts. It was found that photoanodes of particulate LaTiO₂N fabricated by particle transfer generated an order of magnitude higher photocurrent than those prepared by the conventional electrophoretic deposition. The enhanced performance is considered to result from a smaller barrier at the interface between the conductor and the semiconductor particles. It should be noted that the particle transfer method is applicable to immobilization of various particulate semiconductors onto metal layers. It is possible to fabricate a photocatalyst plate consisting of a hydrogen evolution photocatalyst and an oxygen evolution catalyst immobilized onto a conductive metal thin layer, in which the metal layer allows for efficient electron transfer between the two kinds of photocatalysts. The photocatalyst plate was found to show significantly higher activity than a conventional powder suspension system in overall water splitting.

Photocatalytic water splitting with semiconductor materials has been investigated as a clean and renewable process for directly converting sunlight into chemical energy. In particular, multiferroics MFs have recently been used for applications in both photocatalysis (PC) and photovoltaics (PV) due to their ferroelectric properties and narrow band gaps, allowing them to harness the majority of solar radiation in the visible range. As typical MFs, BiFeO3 (BFO) and Bi2FeCrO6 (BFCO) have been recognized as potential materials for PV and visible-light PC applications owing to their suitable band gap (1.4-2.8 eV) and good chemical stability. However, the investigations on such materials for photocatalytic water splitting are still limited and efforts have to be undertaken to demonstrate their full potential. An efficient PV system is at the basis on an effective PC process. Thus, the control of PV properties of MFs is a critical issue for achieving highly efficient photocatalytic system. Here we will present, the controlled growth and characterization of BFCO and BFO thin films and nanostructures via pulsed laser and hydrothermal techniques. The PV properties of such systems and their photocatalytic activity will be also discussed.

We have recently demonstrated visible light-induced selective reduction of CO₂ to HCOOH using a p-type semiconductor particles (N-doped Ta₂O₅) linked with a metal-complex electrocatalyst (ruthenium-complex), which enabled noble reductive photoreaction over the metal-complex electrocatalyst [1#8210;3]. To realize an up-hill reaction for energy storage, electrical coupling between powdered photocatalysts for CO₂ reduction and water oxidation reactions is necessary to facilitate electron transfer in a system. Reduced graphene oxide (RGO) was reported to be an attractive solid electron mediator for constructing the Z-scheme watersplitting in a powdered system [4]. Hence we combined RGO with the metal-complex/semiconductor photocatalyst. [Ru(dpbpy)(bpy)(CO)₂]²⁺ (dpbpy :4,4&’-diphosponate-2,2&’-bipyridine, bpy: 2,2&’-bipyridine) (RuC) linked with SrTiO₃:Rh (RuC/SrTiO₃:Rh), RGO and BiVO₄, were employed as a H⁺ reduction photocatalyst, a solid electron mediator, and a water oxidation photocatalyst, respectively.
Photocatalytic H₂ evolution reaction from a mixture of BiVO₄ and SrTiO₃:Rh (Rh:4 at.%) in the presence and absence of RGO or RuC were performed in test tubes containing an Ar-bubbled acidified water (pH3.5) under visible-light irradiation (lambda;>390 nm). Photoactivity retained during 18 h irradiation, and H₂ evolution rate was enhanced by 5.6 and 1.7 times by modification of RuC onto SrTiO₃:Rh and addition of RGO, respectively. These results indicate that RuC operates as a hydrogen-generation site, while RGO functions as a solid-state electron mediator.
In addition, we evaluated overall watersplitting under visible-light irradiation (lambda;>420 nm) for the (RuC/SrTiO₃:Rh)-(RGO/BiVO₄) system in a closed gas-circulation system. H₂ and O₂ evolved stoichiometrically and the turnover number for RuC was calculated to be 244 after 16 h irradiation, which indicates that this Z-scheme system split water photocatalytically. It was also found that the reaction rate decreased with time because RuC gradually eliminated from semiconductor surface.
In summary, we have successfully constructed a powdered Z-scheme system for photocatalytic water splitting operating under visible light irradiation utilizing a combination of metal-complex, RGO, and semiconductors.
References
[1] S. Sato, et al., Angew. Chem. Int. Ed., 49, 5101 (2010), [2] T. M. Suzuki, et al., Chem. Commun., 47, 8673 (2011), [3] T. M. Suzuki, et al., J. Mater. Chem., 22, 24584 (2012)., [4] A. Iwase, et al., J. Am. Chem. Soc., 133, 11054 (2011)

Oxynitrides and nitrides have recently attracted much interest in the field of solar water splitting. Among them, tantalum oxynitride (β-TaON) and tantalum nitride (Ta₃N₅) have gained considerable attention. They have favorable bandgaps (2.4 and 2.1 eV, respectively), and due to their favorable band positions they are thermodynamically able to split water without any external bias. Due to the delocalized N-2p orbitals, the materials are expected to show excellent hole transport properties. However, the controlled synthesis of high-quality thin-films of β-TaON and Ta₃N₅ is very challenging. In this study, we propose a novel in-situ annealing/UV-vis monitoring technique for the controlled and reproducible synthesis of thin films of any Ta-O-N phase called ‘oxynitrogenography&’. The use of oxynitrogenography can in principle be extended to the synthesis of other oxynitrides as well.
Sputtered tantalum thin films are nitridated at high temperatures (>500°C) in a tube furnace under controlled flows of ammonia, water and hydrogen, while the optical transmission of the film is monitored in-situ. The optical absorption changes of the films can be directly correlated to the presence of different phases (Ta₂O₅, TaO_xN_y, Ta₃N₅), due to their different absorption edges. As a result, the thermodynamic equilibrium conditions to obtain these various phases are determined, and a phase diagram is constructed. The physical characteristics, measured by the in-situ UV-vis and XRD, confirm the high control over the phase transformations.
The electronic properties of the Ta-O-N phases are measured by time resolved microwave conductivity (TRMC). As expected, nitrogen incorporation significantly improves the mobility and lifetime of the photo-generated carriers compared to Ta₂O₅. While the carrier mobility of TaON and Ta₃N₅ is comparable to that of BiVO₄ (10^-2-10^-3 cm²/Vs), the lifetime is in the order of tens of milliseconds and comparable to that of crystalline silicon. To the best of our knowledge, this is higher than any previously reported carrier lifetime for any metal oxide-based semiconductors (BiVO₄, Fe₂O₃, WO₃, Cu₂O). This shows the superior semiconducting properties of oxynitrides compared to their parent oxides, and hints towards the excellent quality of the films.
For the photoelectrochemical (PEC) measurements, various TaO_xN_y thin films are deposited and measured under AM 1.5 irradiation. The photocurrent onset potential is found to be at 0.1 V vs. RHE, close to the reversible potential for hydrogen evolution, as desired. Despite the excellent electronic properties of the films, the photocurrent is not exceptional, being 0.8 mAcm^-2 and 0.4 mAcm^-2 at 1.43 V vs RHE for Ta₃N₅ and TaON, respectively. The limited photocurrent can be explained by charge recombination at the non-optimized semiconductor/electrolyte and semiconductor/back contact interfaces. We will discuss possible strategies to mitigate interfacial recombination and share some initial results.

Satisfied with a suitable band gap (2.1 eV) and band position straddling water redox potentials absolutely required for desirable photocatalysts, a tantalum nitride (Ta₃N₅) is expected one of promising semiconductor materials for visible-light-responsive water splitting. Based on such the background, Ta₃N₅ photoanodes has been developed for high photoelectrochemical (PEC) performance by introducing electrode preparations with different structures such as nano-rod and thin film, and by modifying the surface of electrode with loading various oxygen reduction reaction (OER) catalysts. The photocurrent densities of pure Ta₃N₅ photoanodes at 1.23 V_RHE for OER significantly increased from those attempts, whereas their photocurrent onset potentials were still very positive close to 0.8~0.9 V_RHE.^[1] The positive onset indicating a low photoactivity exposes the limitation of pure Ta₃N₅ as a photoanode, even if it has the band position theoretically possible to spit water. Furthermore, there is no doubt that the photoanode applicable to tandem PEC cells should suggest more negative onset potential for efficient solar water splitting.
Herein we approached for lower onset potential of Ta₃N₅ photoanode toward OER by flat band potential engineering, i.e. shifting the band position of bulk Ta₃N₅ itself. Although a-element-doped Ta₃N₅ has been reported for PEC performance to date, it was not enough to propose the clear discussion for effect of doping on the photoactivity or to show high PEC performance for water splitting.^[2] Co-doping with Mg and Zr into Ta₃N₅ structure was thus attempted to determine the influence of doping on the PEC performance over the Ta₃N₅ photoanode, especially on the photocurrent onset potential. The Mg-Zr co-doped Ta₃N₅ was successfully prepared by the flux-assisted nitridation at 1173 K for 20 h. Subsequently, the photoanodes loading the synthesized doped-Ta₃N₅ powder were fabricated by a particle transfer method.^[3] According to the XRD patterns and UV-visible DRS spectra, the doped Ta₃N₅ exhibited the identical crystal structure and absorption edge (~600 nm) to those of pure Ta₃N₅. Nevertheless, the PEC activity of the co-doped Ta₃N₅ photoanode under AM 1.5G-simulated sunlight, which CoO_x as an OER catalyst was electrodeposited on the surface of electrode, revealed a significant difference from pure Ta₃N₅. The anodic photocurrent corresponding to oxygen evolution for the doped photoanode was up-risen largely at the beginning of 0.5~0.6 V_RHE, indicating a negative shift of onset-potential when compared with that of pure Ta₃N₅ above. This obvious shift of onset potential should be attributed to the differentiated band structure of the doped Ta₃N₅. More detailed characteristics and discussion for the Mg-Zr co-doped Ta₃N₅ photoanode will be reported in the presentation.
REFERENCES
[1] Y. Li et al., Adv. Mater., 2013, 25, 125.
[2] Y. Kado et al., Chem. Commun., 2012, 48, 8685.
[3] T. Minegishi et al., Chem. Sci., 2013, 4, 1120.

Tantalum nitride (Ta₃N₅) is an attractive visible light absorber for solar water splitting due to its favorable optical band gap of 2.1 eV with conduction and valence bands that straddle the redox potentials for water reduction/oxidation, thermodynamically enabling the semiconductor to achieve unassisted water splitting under solar illumination.¹ As synthesized, Ta₃N₅ is naturally n-type and is commonly used as a photoanode. If 100% utilization of photons is realized, then Ta₃N₅ could produce a photocurrent density of 12.5 mA/cm² which, if coupled with an appropriate photocathode in a tandem photoeletrochemical (PEC) device, could perform unassisted water splitting at a solar-to-hydrogen efficiency of ~15%.² However, the record Ta₃N₅ device produces only 7 mA/cm².³ Therefore, understanding the materials limitations of Tashy;₃Nshy;₅ by studying its opto-electronic properties is important for designing and engineering a device architecture that maximizes Ta₃N₅ performance as a photoanode for PEC water splitting. One important consideration for efficient light absorption is the direct or indirect nature of the optical bandgap—a matter that has conflicting reports in the literature. Our goal was to definitively assign direct or indirect interband transitions to two prominent features in Ta₃N₅ optical absorptance spectra occurring at 2.1 and 2.5 eV⁴ using theoretical and experimental methods. Sophisticated Density Functional Theory (DFT) calculations were performed to model the electronic band structure and dielectric function of Ta₃N₅. Spectroscopic ellipsometry was conducted to elucidate the dielectric function of Ta₃N₅ experimentally. Theoretical and experimental results were in good agreement and enabled the identification of both of the transitions, 2.1 and 2.5 eV, as direct interband transitions.⁶ A direct optical band gap semiconductor is advantageous because a thin absorber layer can be used while still maintaining efficient light absorption. DFT calculations also showed high effective masses of electrons and holes in multiple directions indicative of low charge carrier mobilities, which is consistent with experimental findings that photocurrent density scales with increasing Ta₃N₅ surface area.^5,6
References:
(1) Chun et al, J. Phys. Chem. B 2003, 107 , 1798-1803.
(2) Seitz et al, ChemSusChem2014, 7, 1372-1385.
(3) Li et al, Nat. Commun. 2013, 4, 2566.
(4) Pinaud et al, Chem. Mater. 2014, 26, 1576-1572.
(5) Pinaud et al, J. Phys. Chem. C 2012, 116(3), 15918-15924.
(6) Morbec et al, Phys. Rev. B. 2014, Accepted.

One of the biggest challenges for solar water splitting is the lack of both efficient and stable semiconductor anodes for water oxidation reaction. Despite the significant progress made recently in III-V compound semiconductors (e.g. GaP, InGaAs) and metal oxides (e.g. Fe₂O₃, TiO₂, BiVO₄) for this application, enhanced operational stability of the former and better charge transport of the latter are still needed. In this context, metal nitrides are interesting because of their oxide-like stability and semidoncutor-like charge transport properties [1]. This presentation will focus on optoelectronic properties of Sn₃N₄ semiconductor with spinel crystal structure as studied by thin-film experiments and first-principles calculations.
Sn₃N₄ with spinel crystal structure in polycrystalline thin film form was synthesized by the previously reported methods [2] of reactive sputtering of metal target in atomic nitrogen atmosphere. The optical absorption onset in the 1-2 eV range is in good agreement with 1.5-1.6 eV calculated band gap from GW theory. For the synthesized samples, the electron concentration was 10¹⁸ cm^-3 with mobility of ~1 cm²/Vs, showing non-degenerate semiconductor-like temperature dependence. Sn₃N₄ band gap was determined to straddle the water oxidation and reduction potentials, according to experiments (both in air and in the solution) and theory (surface ionization potential calculations with correction for water interaction).
Unfortunately, in part due to the polycrystalline character of the current Sn₃N₄ thin films, the measured photocurrent was limited by the 50-100 nm minority carrier (holes) diffusion length, which is on the order of the grain sizes. This observation calls for improvement in thin film growth process to achieve larger grains and hence larger photocurrent. Another possible contribution to low photocurrent is the calculated relatively large hole effective mass (12.9m_e), which contrasts with very small electron effective mass (0.18m_e), and is quite surprising as for a nitride semiconductor. Comparing to other structurally- and chemically- related materials (e.g. SnO₂, Cu₃N, GaN etc) suggest that both of these factors play a role in the larger hole effective masses of Sn₃N₄. As a potential solution, theoretical calculations suggest alloying of Sn₃N₄ with other group-IV nitrides, which can improve hole effective masses by both band engineering and polymorphism of this materials system.
In summary, even though Sn₃N₄ still has its challenges, overall novel metal nitrides [3] constitute an interesting materials family for solar water splitting semiconductor anode applications.
[1] J. Phys. Chem. Lett., 5, 1117 (2014)
[2] Mater. Horiz., 1, 424 (2014)
[3] Chem. Mater., 26, 4970 (2014)

Hematite (α-Fe₂O₃) nanostructures have been extensively studied as photoanodes for photoelectrochemical (PEC) water splitting. However, the photoactivity of pristine hematite nanostructures is limited by a number of factors, including poor electrical conductivity and slow oxygen evolution reaction kinetics. Previous studies have shown that the incorporation of extrinsic dopants and/or intrinsic defects can substantially enhance the photoactivity of hematite photoanodes by modifying their optical and electrical properties. In this talk, I will highlight our recent progress in exploring chemically modified hematite photoanodes for solar water splitting.

The development of improved materials is key to advancing technologies for solar water-splitting. Among the many technological challenges, several include: catalyst development for the hydrogen evolution reaction (HER), catalyst development for the oxygen evolution reaction (OER), the development of semiconductor absorbers for use as photoanodes and photocathodes, and engineering the interfaces between semiconductors and catalyst materials for corrosion protection and unimpeded charge transport. This talk will aim to describe our recent advances in these areas.
More specifically, in this talk we will: (1) Explore the current state of materials development for catalyzing the HER and the OER, both in acid and in base; (2) describe targets for catalysis based on recent simulations of practical, realizable solar-to-hydrogen (STH) efficiencies based on the current state of materials research; (3) Discuss challenges and opportunities in developing photocathodes based on PV-grade materials and developing photoanodes based on metal oxides and metal nitrides. Some of the key underlying themes include: the importance of turnover frequency in assessing catalysis, the role of nanostructured and microstructured morphologies, and the importance of interfaces and interfacial engineering.

Direct solar to fuel conversions offer an attractive solution to the replacement of fossil fuels with a sustainable energy source. However, design of efficient, stable, and cost effective photoelectrochemical cells is hindered by materials development. One dimensional nanostructures such as semiconductor nanowires have attractive properties that could address these materials challenges. However, the synthesis of one-dimensional nanostructures with desired properties is often complicated by the difficulty in changing the composition without sacrificing control over the morphology and material quality. Here, we present a simple method we have developed based on solid state diffusion utilizing atomic layer deposition to controllably alter the composition of metal oxide nanowires.1 This compositional control allows for modification of the optical, electronic, and electrochemical properties of the semiconductor nanowires. Specifically, we demonstrate the doping process resulted in enhancement in the performance for water oxidation, demonstrating that this simple and general method can be used to control the properties of one-dimensional nanostructures for use in a variety of applications including solar-to-fuel generation.
1. Resasco, J. ; Dasgupta, N. ; Roque-Rosell, J. ; Guo,J. ; Yang, P. J. Am. Chem. Soc. 2014, 136, 10521.

Due to their abundance, non-toxicity, optical and physicochemical performance TiO₂ and α-Fe₂O₃ are currently most studied semiconductors for photoelectrochemical (PEC) water splitting. However, despite large attempts to improve their PEC performance the reported efficiencies for TiO₂ or α-Fe₂O₃ photoanodes remained limited because of low conductivity and high recombination rate observed for hematite, while poor PEC performance in for TiO₂-based photoanodes originates from its wide band gap and thus limited absorption in visible range.
In this contribution we demonstrate the effect of coupling of both materials on a nanoscale, by preparing photoanodes based on their nanocomposites. Thus we aimed to improve the charge carrier separation and reduce surface trapping of α-Fe₂O₃ by forming its intimate heterojunction with TiO₂ matrix. In addition transition metal doping of TiO₂ was applied to improve its conductivity and visible light absorption.
Nanocomposites combining doped TiO₂ and α-Fe₂O₃ constituents are prepared using non-aqueous sol-gel synthesis following two different paths: (i) by mixing of the colloidal dispersions of as-synthesized doped TiO₂ nanoparticles and Fe₂O₃ precursor nanoparticles and (ii) by partial coating of the as-synthesized doped-TiO₂ core nanoparticles surfaces with hematite shell. Spin-coating or doctor-blading of the as-prepared nanocomposites on FTO glass substrate, upon annealing at 680 °C under oxygen flow leads to a transparent, porous film. For doped TiO₂-Fe₂O₃ photoanodes the photocurrent of 2.25 mA/cm² at 1.23 V vs. RHE is obtained under simulated solar AM 1.5 illumination (100 mW/cm²) in 1M NaOH electrolyte solution, which is higher than reported photocurrent obtained for single component TiO₂ or α-Fe₂O₃ photoanode.
By means of XRD, XPS, UVVis and FESEM analysis of nanocomposites, we aimed to give a broader view on the role of the size, morphology and electronic structure of the nanocomposite constituents, the formation of heterojunction and composite nanostructuring on PEC performance.

Rhodium doped titanium dioxide and strontium titanate powders and single crystals have been codoped with antimony, tantalum or niobium in order to increase their range of light absorption to include visible light. Doping with noble metals is known to improve the viability of these titanates as photoanodes for use in solar water splitting.1 The use of single crystals has allowed for more comprehensive investigation of these materials compared to previous literature by allowing for depth profiling.
Rhodium doping is known to reduce the n-type properties of TiO2 since it preferentially forms Rh4+ in the TiO2 lattice, pinning the Fermi level deeper in the band gap. Charge compensation by Sb5+ mitigates this, encouraging Rh3+ incorporation.2 Strontium titanate doped with rhodium and antimony powder has been shown to be an effective photocatalyst for both oxygen and hydrogen evolution.3
Powder doping has been achieved by simple solid-state reactions of appropriate mole amounts of dopant oxide and titanate powders. Single crystal doping was achieved by high temperature solid-state diffusion of dopant from the doped powder to the crystal.
X-ray diffraction data confirms the absence of dopant oxides in the synthesised powders and high resolution X-ray Photoelectron Spectroscopy (XPS) was used to probe the surface of the crystals to determine the concentration and oxidation state of the dopant elements. In addition valence band spectra were obtained in order to elucidate the electronic structure of the resultant materials. XPS data for single crystal titanium dioxide corroborates published reports for TiO2 powder, whilst XPS depth profiling has provided further valuable insight into the distribution of dopants below the surface of the material.
Strontium titanate is of particular interest as, when doped with Rh and Sb, it exhibits p-type nature at the surface, with n-type characteristics in the bulk. This is due to surface segregation of mixed oxidation state Sb ions, with Sb3+ existing in high concentrations at the surface, while Sb5+ exists in the bulk.4 Consequently, investigation of this material to determine the presence of an intrinsic p-n junction has been undertaken to shed new light on its effectiveness as a photocatalyst for water splitting.
1. H. Hideki; A. Kudo, J. Phys. Chem. B 106 (19), 2002, p.5029
2. F. E. Oropeza; R.E. Egdell, Chem. Phys. Lett. 515, 2011, p.249
3. K. Furuhashi; Q Jia; A. Kudo; et al. J. Phys. Chem. C 117 (37), 2013 p.19101
4. P.A. Cox; R.G. Egdell; C. Harding; W.R. Patterson; P.J. Tavener, Surf. Sci. 123, 1982, p.179

Recently, protective and hole-conductive coatings of ALD-grown, amorphous “leaky” TiO₂ have proven to be instrumental for stabilizing various technologically-important semiconductor photoanodes (e.g., n-type Si, GaAs, GaP and CdTe) for water oxidation. Here, we will show that conformal coatings of ALD-TiO₂ protects structured photoanodes like Si microwires, GaAs nanowires and GaAs nanowire/planar Si tandem for efficient, quantitative and stable oxidation. Wire-array photoanodes reduce > 90% of material usage to achieve comparable photocurrents, while exhibiting enhanced aqueous stability. For example, “leaky” TiO₂ protected Si wire-arrays can stably oxidize water for > 2000 hours in 1 M KOH passing the same amount of charge as of >1 year outdoor operation.
To further improve performance of TiO₂-coated semiconductor photoanodes and to expand selection of materials as “leaky” protective coatings, we extensively investigated n-Si/TiO₂ interfaces as a model system. Solid-state current-voltage measurements at variable temperatures and impedance measurements showed that electrical behavior of n-Si/TiO₂ interfaces are rectifying with a barrier height of ~0.8 V, and this is also supported by x-ray photoelectron spectroscopy study of incremental ALD cycles grown on Si. Finally, the distinctive electrical behavior of n-Si/”leaky”TiO₂ interfaces will be compared with Schottky and semiconductor/liquid junctions, and strategies for improving the performance of TiO₂-coated photoelectrodes will be discussed.

Nanostructured TiO₂/WO₃ bi-layered thin films were investigated as photoelectrode in photoelectrochemical (PEC) cell for generation of hydrogen by splitting of water. In this study bi-layered junction was fabricated for best performance by tailoring thickness of the TiO₂ over WO₃ thin films. Thin films were synthesized by sol-gel spin coating method and were subjected to XRD (for phase and particle size analysis), UV-spectroscopy (for optical characterization), AFM (for surface topography), SEM (for surface morphology) and Mott-Schottky analysis (for Flat band potential). Wide absorption spectrum in visible region along with a sharp shoulder in UV region was observed. AFM and SEM picture of bi-layered thin film reveal uniform and well defined distribution of nanoparticles on the substrate i.e. growth of crystallites perpendicular to the substrate. XRD analysis revealed exhaustive evolution of anatase TiO₂ and monoclinic WO₃ in all the samples, particle size varied between (45-60 nm). A maximum photocurrent density of 0.93 mA/cm² at 0.75 V/SCE and solar to hydrogen conversion efficiency of 0.94% was recorded for 930 nm thick modified bi-layered photoelectrode in 1 M NaOH as electrolyte. A significant rise in photocurrent was recorded in comparison to Pristine TiO₂ and WO₃ thin film samples. As the thickness of TiO₂ varied over WO₃, current density increased. Results are encouraging and show that bi-layered films yielded significant gain in photocurrent, compared to mono-layered pristine samples.

Molecular hydrogen (H₂) is one of the world's most important chemicals with a global production rate of approximately 50 billion kg per year, mainly used for petroleum refining and for synthesizing ammonia (NH₃)-based fertilizers. As hydrogen is mainly produced from fossil fuels, developing an alternative, renewable pathway to produce H₂ in a cost-competitive manner would have a significant impact in reducing fossil fuel usage and CO₂ emissions. One attractive pathway for clean hydrogen production is through electrochemical processes, such as solar photoelectrochemical (PEC) water splitting or electrolysis coupled to renewable energy sources such as wind or solar.
The hydrogen evolution reaction (HER, 2H⁺ + 2e^minus; → H₂) constitutes half of the water splitting reaction. To increase process efficiency, active catalysts for the HER are needed. Currently platinum is the best known HER catalyst as only small overpotentials are required to drive high reaction rates, but the scarcity and high cost of Pt may limit its widespread technological use. Non-noble metal alternatives include nickel and nickel alloy catalysts but they are typically not stable in acidic solutions preventing use in proton exchange membrane-based (PEM) electrolysis, which have significant advantages over more conventional alkaline electrolyzers.
Very few earth-abundant catalysts have shown efficacy for the HER in strong acids. Recently transition metal phosphides have emerged as highly active and acid-stable HER catalysts. This development was partly inspired by phosphides efficacy for hydrodesulfurization (HDS) catalysis. The commonalities among catalysts for HDS reactions and the HER are not surprising since both reactions rely on a catalyst's ability to bind an important reactive intermediate, hydrogen, in a close to thermo-neutral manner.
Here we investigate a well-known HDS catalyst, molybdenum phosphide (MoP), and find that engineering the surface of MoP to more closely resemble its state during HDS operating conditions leads to significant enhancements in both catalytic activity and stability for the HER. Reports in the HDS literature indicate that sulfur plays an important role in the HDS over transition metal phosphides. For MoP the most active sites are thought to originate from a surface phosphosulfide generated during HDS reactions. To explore if a surface phosphosulfide is also beneficial for HER, we introduced sulfur in the surface region of the synthesized MoP which led to remarkable improvements in catalyst performance. MoP with a phosphosulfide surface (MoP|S) exhibits HER activity among the highest of any non-noble metal electrocatalyst, while remaining nearly perfectly stable in acid. The extraordinarily high activity and stability of this catalyst opens up avenues to replace Pt in technologies relevant to renewable energy such as PEM electrolyzers and solar PEC water-splitting cells.

Interfacing semiconducting Nanowires (NWs) with potent electrocatalysts composed of earth-abundant elements will significantly improve their performance as photoelectrodes in Z-scheme artificial photosynthesis devices. Hydrogen evolution reaction (HER) whereby protons and electrons are combined into molecular hydrogen, a clean and high density energy carrier, is catalyzed most effectively by the Pt group metals, which are both expensive and scarce, prompting widespread efforts to discover cost-effective and earth abundant materials for their replacement. Transition metal sulfur centers, which naturally exist as active sites for hydrogen evolving enzymes nitrogenase and hydrogenase, and their solid state analogues have sparked intense research interest due to their promising HER activity. Our recent result has shown that the incomplete cubane-like clusters [Mo₃S₄] is an efficient co-catalyst for the evolution of hydrogen when coupled to GaP NWs, with significant increase in H₂ production rate compared to bare GaP NW. The structure of the molecular catalyst was not stable and went through structural change during the hydrogen evolution reaction into solid state amorphous nanoparticles without deteriorating the catalytic activity. Combined characterization with Surface Enhanced Raman Spectroscopy, Aberration Corrected High Resolution TEM and X-ray Adsorption Spectroscopy studies have been conducted to interrogate the change in local structure observed at the catalyst/semiconductor interface. Such measurements are crucial for identifying active catalytic species, understanding the reaction mechanism and developing more efficient and stable transition metal sulfide catalyst.

Hydrogen production via photoelectrochemical (PEC) water splitting with high solar-to-hydrogen (STH) efficiency and high durability has been a significant challenge for decades. It requires that semiconductor materials&’ band gap, band edge, optoelectronic efficiency, and stability must be satisfied simultaneously. Metal oxides may be stable, STH efficiencies have been limited by issues related to the wide band gap, absorption, charge mobility, recombination, interfacial kinetics, etc. Crystalline III-V materials offer an alternative pathway to efficient STH conversion, with stability remains an issue.
Besides the semiconductor materials, the challenge also requires effective usage of solar spectrum. To address the band-edges mismatch, structures and configurations have been developed for unbiased PEC water splitting. So, semiconductor materials and structures as well as STH efficiencies will be discussed in the presentation.

Organolead halide peroskite quantum dots (PQDs) have recently received significant attention due to their unique optical and electronic properties for solar energy conversion and other potential application. In particular, PQDs are better suited for device fabrication in large area and on flexible substrates, like plastics, with low cost processing such as screen printing. One major issue with PQDs is instability that is likely associated with bandgap trap states due to surface defects. We have synthesized CH₃NH₃PbX_nY_3-n PQDs with tunable emission wavelength from green to red. The optical properties of PQDs result from the variation in their size and compositions (X and Y=Cl, Br, I; n=0, 1, 2). Furthermore, we developed and evaluated several strategies to passivate surface states and thereby enhance the stability of the PQDs. The structural and optical properties of the PQDs have been characterized using a combination of spectroscopy and microscopy techniques. In addition, exciton dynamics of the PQDs have been investigated using ultrafast laser techniques. Together, new insight is obtained about the electronic structure of the PQDs including bandgap states and their relation to instability, which has important implications in solar energy conversion applications.

In this presentation, we will discuss how to utilize the general strategies for rational design of engineering metal oxides to simultaneously meet the criteria required for spontaneous, efficient water-splitting by sunlight. We will show that to achieve high optical absorption of visible light and high carrier mobility, the incorporated donor-acceptor combination must reach a threshold concentration and that the bandgap reduction depends critically on the donor-acceptor concentration. We will use ZnO and Fe₂O₃ as examples to show how the photoelectrochemical (PEC) performance of metal oxides may be improved by using the co-ally concept. Furthermore, based DFT band-structure calculations, we find that the PEC performance of anatase TiO₂ can also be greatly improved by co-alloy approach. We predict that (Ta, N) and (Nb, N) pairs are the optimal donor-acceptor combinations in the high-concentration regime, and (Mo, 2N) and (W, 2N) combinations are good candidates in the low-concentration regime for engineering TiO₂ to meet bandgap, optical absorption, band edge, and mobility criteria. Our design principles are general and can be applicable for searching for other new PEC water-splitting semiconducting materials.

Photochemical charge generation, separation, and transport at nanocrystal interfaces are central to photoelectrochemical water splitting, a pathway to hydrogen from solar energy. Here we use surface photovoltage spectroscopy (SPS) to probe these processes in nanocrystal films of photoanode and photocathode materials, incl. WO₃, NiO, C₃N₄, M:SrTiO₃, HCa₂Nb₃O₁₀, CdSe, p-Si, and BiVO₄. Charge injection between nanoparticles can be observed, as well as redox reactions at nanocrystal-liquid interfaces. The ability to monitor these voltage-generating processes with SPS provides new insight into nanoscale charge separation and promotes the development of nanocrystal applications in photoelectrochemical cells, photovoltaics, and as photocatalysts.

A solar-driven water-splitting cell is generally comprised of light absorbers, electrocatalyts, membrane separators and an electrolyte solution in a specific system geometry. The overall solar-to-hydrogen conversion efficiency of such a system depends on the performance and materials properties of all the individual components as well as the design of the system. Significant advancements in modeling/simulation and prototyping development of solar-hydrogen devices have been made at Joint Center for Artificial Photosynthesis (JCAP). In this talk, I will present a comprehensive multi-physics model for a solar-hydrogen cell that accounts for the performance of photoabsorbers and electrocatalyts and the transport properties of electrolytes and membrane separators. The whole cell model was employed to optimize geometries of prototype designs, to define operational conditions and constraints for various system designs, to provide target materials properties and to evaluate the viability of new design concepts. Specifically, the optimal band-gap combination of a tandem photoabsorbers, the performance limits in near-neutral pH electrolytes and the target transport properties of the membrane separators will be discussed in detail. A few novel cell designs, including a vapor feed device and a solar concentrator coupled solar-hydrogen cell will also be discussed. In the past years, guided by the modeling and simulation activity, a robust prototype, the louvered design, has been developed. A solar-to-hydrogen conversion efficiency that exceeds 10% has been achieved in an integrated device with product gas separations.

One of the possible solutions to the world&’s rapidly increasing energy demand is the development of
new photoelectrochemical cells with improved light absorption. This requires development of
semiconductor materials which have appropriate bandgaps to absorb a large part of the solar spectrum
at the same time as being stable in aqueous environments. I will demonstrate an efficient,
computational screening of relevant oxide and oxynitride materials based on electronic structure
calculations resulting in the reduction of a vast space of 5400 different materials to only 15 promising
candidates. The screening is based on an efficient and reliable way of calculating semiconductor band
gaps. The outcome of the screening includes all already known successful materials of the types
investigated plus some new ones which warrant further experimental investigation.
Reference : I. E. Castelli, T. Olsen, S. Datta, D. D. Landis, S. Dahl, K. S. Thygesen and K. W. Jacobsen, Energy Environ. Sci. 5, 5814 (2012)

Au-nanocontacts on p-type silicon with (100) and (111) surface orientation were prepared by electron beam deposition of 5 nm thick Au-films and annealing below the Si-Au eutectic temperature (300°C) for 10 min. The Si/Au heterojunctions were subsequently chemically (in HF, 50%) and electrochemically (in NH₄F, 40%) conditioned in order i) to remove a highly resistive interlayer, caused by gold-silicon inter-diffusion and SiO₂ formation, and ii) to transform the resulting electrodes, which were initially not responding to illumination, to photoelectrodes. Thereby, photocurrent densities of 25 mAcm^-2 at the thermodynamic potential, E₀, for evolution of hydrogen could be realized at pH0. Non-photoactive electrodes proved almost overpotential-free evolution of hydrogen but a more sluggish behavior in comparison to Pt-metal sheets. Incomplete transformation to photoactive electrodes, in turn, resulted in a “mixed” behavior with photocurrent densities at E₀ in the mA-range and moderate overpotentials for the same current densities in the dark. Investigation by Scanning Electron Microscopy shows that morphology and distribution of the nanocontacts adapt to the surface symmetry of the substrates: on (100), the Au-nanocontacts show a near-cubic symmetry in both individual shape and distribution; on (111), thin Au-nanosheets are formed with threefold symmetry, visible by hexagonal holes and the orientation of the nanosheet boundaries. Comparative model experiments with chemically prepared Si-Au interfaces, not exposed to elevated temperatures, were carried out. It will be shown that the applied (electro)chemical steps are capable to (reversibly) switch between complete Fermi-level pinning and unpinning. Thereby, non-photoresponsive electrodes become photoactive and vice versa. Transmission Electron Microscopy result suggest that this effect has to be attributed to the formation (or removal) of an interfacial SiO₂ layer between the metal and the semiconductor which is able to suppress the formation of metal-induced gap states. The characterization of these Au-Si systems will be finally completed by results from photoelectron spectroscopy and reflection spectroscopy near Brewster&’s angle.

Photo-induced charge transfer at the interface of two materials is a
fundamental process in (i) photovoltaic and (ii) photocatalytic
applications. The photo-induced time-dependent electron dynamics are
computed for different interfaces by a combination of ab initio
electronic structure and time-dependent density matrix methodology. A
dissipative equation of motion for the reduced density matrix for
electronic degrees of freedom is used to study the phonon-induced
relaxation of hot electrons in the simulated systems. Non-adiabatic
couplings between electronic orbitals are computed on-the-fly along
nuclear trajectories. Equations are solved in a basis set of orbitals
generated ab initio from a density functional.[1] For an application
to photovoltaic effect, one explores light-induced electric current in
a model of a simplified photovoltaic cell composed of a Si
nano-crystal co-doped with p-and n- type doping, interfacing with Au
electrodes. Charge carrier dynamics induced by selected
photo-excitations show that hole relaxation in energy and in space is
much faster than electron relaxation. Use of the continuity equation
for electric current allows to identify substantial local currents at
the Si/Au interfaces and small overall net charge transfer across the
slab. [2] For an application to photocatalytic water splitting, charge
transfer dynamics is explored at the interface of supported metal
nanocluster and liquid water. The metal cluster introduces new states
into the band gap of semiconductor TiO2 surface, narrows the band gap
of TiO2, and enhances the absorption strength. The H2O adsorption
significantly enhances the intensity of photon absorption, which is
due to the formation of metalminus;oxygen (water) coordination bonds at the
interfaces. The metal cluster promotes the dissociation of water,
facilitates charge transfer, and increases the relaxation rates of
holes and electrons. [3] Reported results help in understanding basic
photophysical and protochemical processes contributing to harvesting
solar energy by photovoltaics and photoelectrochemical water
splitting.
1. Huang, S.; Kilin, D. S., Charge Transfer, Luminescence, and Phonon
Bottleneck in TiO2 Nanowires Computed by Eigenvectors of Liouville
Superoperator. J. Chem. Theor. Computation 2014, 10 (9), 3996-4005.
2. Han, Y.; Micha, D.; Kilin, D., Ab initio study of the photocurrent
at the Au/Si metal semiconductor nano-interface. Mol. Phys. 2014, in
print, DOI: 10.1080/00268976.2014.944598.
3. Huang, S.; Inerbaev, T. M.; Kilin, D. S., Excited state dynamics of
Ru10 cluster interfacing anatase TiO2(101) surface and liquid water.
J. Phys. Chem. Lett. 2014, 5, 2823-2829

Recently, GaN has been attracting interest in photoelectrochemistry and photocatalysis due to the very favorable energy position of its band edges with respect to many redox levels [1,2]. Additionally, the flexibility in alloying and doping of III-nitrides is expected to provide an unprecedented control over the electronic properties of the surface. Moreover, very important systems investigated in heterogeneous photocatalysis are hybrid systems comprising a semiconductor with a co-catalyst, such as Pt, located on its surface [3]. Here, GaN is also advantageous because of the suitable alignment between the work function of many co-catalysts and its energy band edges, which may be useful e.g. for water splitting [4]. The transport and recombination of photo-generated charge carriers is particularly important in photocatalysis. Here, we use GaN with different deposited Pt structures as a controllable model system to investigate the kinetics of photo-generated charge carriers in hybrid photocatalysis. For this purpose we vary the thickness and coverage of Pt on intrinsically n-type (0001) GaN layers. Using contact potential difference (CPD) and photoconductivity (PC) measurements we study systematically the influence of Pt on the processes involved in the capture and decay of photo-generated charge carriers. We find that for Pt layers with thickness <1nm the magnitude of photo-induced changes in CPD (PC), which directly monitor the capture of photo-generated holes (electrons) at the surface (conduction band states), are very similar to those observed for Pt-free GaN. However, the time dependence of the charge capture becomes considerably faster in these GaN-Pt systems. For thicker Pt layers (>1nm) the capture of photo-generated charges becomes negligible. In experiments where we changed the structure of Pt on GaN via annealing in hydrogen, where a close thick Pt layer (>1nm) is transformed into separated Pt islands, we observe that after annealing the photo-induced CPD and PC increase dramatically to values close to those observed for Pt-free GaN, indicating that the spatial distribution of Pt rather than its average amount determines the photo-generated charge capture. Our experimental data and results are discussed within the known electronic structure of GaN (including surface band bending and bulk defect states) and the excitation and recombination processes taking place upon illumination, as well as when this excitation is ceased.
[1] J. Howgate, S. J. Schoell, M. Hoeb, W. Steins, B. Baur, S. Hertrich, B. Nickel, I. D. Sharp, M. Stutzmann, M. Eickhoff, Adv. Mater. 22, 2632 (2010)
[2] S. Schäfer, S. A. Wyrzgol, R. Caterino, A. Jentys, S. J. Schoell, M. Hävecker, A. Knop-Gericke, J. A. Lercher, I. D. Sharp, M. Stutzmann, J. Am. Chem. Soc. 134, 12528 (2012)
[3] Z. Zhang and J. T. Yates, Chem. Rev. 112, 5520 (2012)
[4] D. Wang, A. Pierre, M. G. Kibria, K. Cui, X. Han, K. H. Bevan, H. Guo, S. Paradis, A.-R. Hakima, Z. Mi, Nano Lett. 11, 2353 (2011)

In this contribution, we will discuss our work on predicting the thermodynamic ability of materials to act as a “photocatalyst” and drive the splitting of water or the reduction of CO₂ to hydrocarbons [1-2]. Many semiconducting materials are known experimentally to either reduce protons or oxidise water under illumination, whilst relatively very few are found to be active for the overall splitting of water. This raises the question of whether this experimental lack of overall watersplitting activity is thermodynamic in nature or arises from reaction kinetics being slow compared with the rate of electron-hole recombination. This question is rather pertinent, as the two scenarios require different approaches to convert a material into a practical photocatalyst; use in a Z-scheme or in a photoelectrochemical cell for the former case and as a part of a heterojunction in the latter case [3]. We thus developed a computational approach to predict the chemical potentials of free electrons in the conduction band, free holes in the valence band and excitons (electron-hole pairs) of potential photocatalysts, relative to the potentials associated with the proton reduction, CO₂ reduction and water oxidation reactions [1].
We have so far applied our computational approach [1,2] to photocatalysts based on organic polymers (poly(p-phenylene) [4]) and carbon nitride materials [5], and will review our results for these applications in our contribution. Specifically, we will show how our calculated potentials rationalise the experimentally measured (lack of) photocatalytic activity and how our calculated spectra can help identify the structure of the carbon nitride samples with the highest activity. We will also discuss how one can use this type of calculation to rapidly screen potential novel photocatalyst materials and to provide insight into the nature of the heterojunction formed between different materials and as to why the formation of such a junction may result in a mixture of both materials becoming an overall watersplitting photocatalyst.
[1] Guiglion, P.; Butchosa, C; Zwijnenburg, M.A., J. Mater. Chem. A. 2014, 2, 11996.
[2] Butchosa, C.; Guiglion, P; Zwijnenburg, M.A., J. Phys. Chem. C. 2014 DOI: 10.1021/jp507372n.
[3] Jang, J.S., Kim, H.G. and Lee, J.S., Cat. Today. 2012, 185, 270.
[4] Yanagida, S.; Kabumoto, A.; Mizomoto, K.; Pac, C.; Yoshino, K., J. Chem. Soc. Chem. Commun. 1985, 474.
[5] Wang, X.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin, G.; Carlsson, J.M.; Domen, K.; Antonietti, M. Nature Mater. 2009, 8, 76.