Dielectric materials exhibiting very high dielectric constant (k) of more than several hundreds, low leakage current and low losses, and ideally exhibiting biocompatible properties are critical for application as gate oxide for the next generation of nanoelectronics based on nanoscale CMOS devices, for supercapacitors for energy storage, for microchip embedded energy storage capacitors for biomedical implantable microchips, and for capacitors for the next generation of non-volatile memory devices. A grand challenge of such devices is that, when scaling down to sub-μm range, some high-k materials like (Ba, Sr)TiO₃ usually suffer from serious reduction of dielectric constant and other dielectric properties. Several reports have shown that interfaces of electrodes play an important role in the dielectric properties. Our group recently demonstrated a novel approach to achieving dielectric thin films with extremely high dielectric constant using nanolaminates of TiO₂ and Al₂O₃. We demonstrated Al₂O₃/TiO₂ nanolaminate films with sub-layer thickness, which exhibit dielectric constants of up to 1000 for frequencies â?¤ 10⁴ Hz. A step-like decrease in dielectric constant to ~ 50 occurred between 10⁴-10⁵ Hz. However, these nanolaminates exhibited relatively high leakage current and losses. We now have developed new Al₂O₃/TiO₂ (TAO) nanolaminates, with greatly improved dielectric properties, consisting of alternating TiO₂ and Al₂O₃ sublayers synthesized by atomic layer deposition. Insertion of a Al₂O₃ interfacial layer at the interface between the top Pt electrode layer and the nanolaminate structure yields Al₂O₃/TiO₂ nanolaminate with still high dielectric constant, but with orders of magnitude lower losses and leakage current density than for the prior Al₂O₃/TiO₂ nanolaminates. The unusual high dielectric constant is attributed to the Maxwell-Wagner relaxation between conducting TiO₂ and insulating Al₂O₃, while the low loss and leakage current density is attributed to effective current blocking by the interfacial Al₂O₃ layer. We also report characteristic properties of the nanolaminates as a gate-oxide and a capacitor from our test-bed devices. The results shown in this presentation indicate that the substantially improved TAO nanolaminates can make a major impact in a new generation of nanoelectronic devices (nanoscale CMOS), energy storage capacitors, implantable biomedical microchips, and memory devices.

Metal/Insulator/Metal (MIM) junctions are thin-film structures with potential application as high-frequency (THz) rectifiers for opto-electronic devices. Intriguing MIM applications include (a) rectennas â?" an energy harvesting technology in which electromagnetic radiation is captured via an antenna and rectified to DC power via a rectifier, (b) infrared cameras, detectors and energy harvesting, (c) high-speed data interconnects, and (d) chemical sensors for the monitoring of electro-chemical reactions. Desired properties of MIM junctions for such applications include fast electron transport, low turn-on voltage, high nonlinearity and asymmetry, and sub-micron diode area. These stringent requirements are major obstacles that need to be addressed for successful implementation of MIM junctions. Despite active research since the 1960s, a clear understanding of the influence of material properties and interfaces on the diode performance is missing. This has been the major focus of our work on MIM structures. In this particular study, results will be presented on our efforts to perform interfacial engineering by substituting different insulator layers (Nb₂O₅, Al₂O₃, TiO₂, HfO₂) deposited by ALD onto Nb/Pt thin film template metal pairs. The fabricated MIM junctions are characterized via XPS and UPS to calculate the barrier heights at the interfaces. The barrier height information together with the current-voltage response of the MIM devices provides a direct correlation of material property/interfacial effects (in part) and device performance. TEM analysis of the structures was performed to compare the interface morphology between different MIM devices. The different insulators that were studied include Nb₂O₅, Al₂O₃, TiO₂, HfO₂. The materials were chosen based on a working hypothesis developed from our preliminary studies that creating a near-zero eV barrier height at one of the interfaces should result in devices with superior rectification properties. In addition, studies are in progress to examine the influence of electrical fatigue on the above devices. Barrier height change before and after electrical fatigue testing will also be evaluated. Based on this detailed interfacial characterization, significant advances in our current understanding of interfacial property versus device performance will be presented. Such understanding can have wide reach for applications such as MIM capacitors and MIM devices for memory devices.

In the last decade perovskite based large band gap oxide semiconductor heterostructures have attracted considerable interest, with the most prominent example being the LaAlO₃/SrTiO₃ interface where the formation of a 2-dimensional electron gas was discovered. Perovskite oxide multilayer heterostructures, with the current transport cross plane, are also of interest for energy conversion applications. In this case the transport properties across the interfaces for electrons (holes) are governed by the band discontinuities at the conduction (valence) band. Yet, in contrast to the III/V semiconductors, the band offsets at oxide semiconductor interfaces are not well characterized with the knowledge of the band offsets at perovskite oxide-perovskite oxide interfaces being even more limited. To allow for the optimization of the electronic transport properties in perovskite large band gap semiconductor heterostructures the band offsets have to be investigated and then based on this knowledge suitable materials combinations can be proposed for given applications. In the talk, we will present a study of the characterization and engineering (tuning) of the band offsets at perovskite oxide-perovskite oxide interfaces characterized via in-situ photoelectron spectroscopy in an ultra high vacuum system which couples a photoelectron spectrometer with a pulsed laser deposition chamber. SrTiO₃ â?"based perovskite thin films were stepwise deposited on niobium doped SrTiO₃ single crystals while monitoring the substrate and deposit core level lines. In conjunction with the knowledge of the valence band positions and band gaps of the materials forming the interface, the band offsets were deduced. We will show how based on these results the band offsets can be tuned by using solid solutions of perovskites as deposit on SrTiO₃ to achieve a wide range of band offsets and discuss potential applications of such perovskite oxide-perovskite oxide interfaces to solve problems in energy harvesting.

Wide band gap perovskite structure (ABO3) oxide materials, such as SrTiO3 (STO), are presently being investigated for use in a wide variety of innovative electronic devices including adaptive electronics [1] and energy harvesting elements. In such materials, defect-related properties can provide new functionality or have a marked detrimental effect on properties adversely impacting device performance. Cation nonstoichiometry is known to have a particularly strong influence on the electronic transport of perovskite structure metal titanate thin films, yet a fundamental understanding of many aspects of transport in nonstoichiometric oxide thin films remains not well understood. Device applications of interface proximity effects, such as nonvolatile memories based on resistive switching phenomena, are also hindered by a lack of understanding of transport in disordered nonstoichiometric oxide thin film systems. In the talk, we will present a study of electronic conduction in nonstoichiometric epitaxial STO films and interface proximity effects at Pt-STO heterojunctions where interface formation and Schottky barrier heights have been characterized via in-situ photoelectron spectroscopy in an ultra high vacuum system which couples a photoelectron spectrometer with a pulsed laser deposition chamber. Transport properties of donor doped and nominally undoped nonstoichiometric STO films with cation A/B ratios varying from 0.9 to 1.2 [2] have been investigated including electrical conductivity, Hall effect, and Seebeck effect measurements from 4 K to 600 K. Based on the transport data and density functional theory calculations, the influence of nonstoichiometry on carrier transport in STO will be discussed. This study helps clarify the use of defect-related properties to control electronic conduction and interface proximity effects in a beneficial manner. [1]Sieu D. Ha and Shriram Ramanathan, J. Appl. Phys. 110, 071101 (2011) [2] J. D. Baniecki, M. Ishii, H. Aso, K. Kobayashi, K. Kurihara, K. Yamanaka, A. Valionis, R. Schafranek, accepted to Applied Physics Letters, Nov 1 (2011)

Recent advances in oxide thin film synthesis have led to a surge in reports of novel properties localized to oxide interfaces. Many of these reports have focused on low temperature electronic behavior. However, growth techniques such as oxide molecular beam epitaxy (MBE) can also be used to construct superlattices with enhanced mixed ionic/electronic conductivity at elevated temperatures. In this work, we explore the high temperature electrical and structural properties of Sr-doped LaFeO3 superlattices grown by ozone-assisted MBE on singly-terminated DyScO3 (110) substrates. The (001)-oriented superlattices are grown with periodically inserted SrO planes in the LaFeO3 perovskite structure, with the period ranging from 5 nm down to 0.8 nm. The insertion of a formally non-polar plane in a polar lattice allows the systematic study of nanoscale size effects on point defect behavior and electrical conduction in the space charge region. We employ in situ synchrotron x-ray scattering at the Advanced Photon Source in parallel with DC/AC electrical measurements to examine structure-ionic/electronic conductivity relationships in the superlattices as a function of temperature and oxygen partial pressure. The results will be compared with similar measurements conducted on Sr-doped LaFeO3 alloys with equivalent composition but no Sr ordering. Work supported by the U. S. Department of Energy under Contract No. DE-AC02-06CH11357 (BES-DMSE).

Cu/ZnO-based catalysts have been extensively investigated in several industrial applications, such as methanol synthesis, CO removal, and hydrogen generation. For flow-type gas-solid reactions, micro-porous nanostructures are attractive due to the effective diffusion of reactants and heat therein. Moreover, strong metal-support interaction (SMSI) effect in these Cu/ZnO nanostructures was shown to cause microstrain in the Cu nanoparticles, to which an increase in catalytic performance was attributed [1-3]. These earlier findings led us to design and develop a novel technique to tailor the Cu/ZnO structures and interface. Recently, microwave (MW) irradiation has widely applied in synthetic organic chemistry, yielding dramatic increases in reaction rates and yields. Since the magnitude of MW absorption is related to dipole oscillation in a material, MW annealing can provide selective heating and lead to â?omolecular hot spotsâ? in the pristine materials. Here, we demonstrate MW treatment to be an effective tool to control the interface in composite catalysts for enhanced SMSI and thus significantly improve the activity in methanol reforming reactions. The pristine CuO-ZnO inverse opals were prepared by direct infiltration of Cu- and Zn-containing nitrates solution into polystyrene opals, which had been introduced and formed inside micro-channels. Afterwards, various MW treatments were performed over these CuO-ZnO frameworks. For the binary system comprising CuO/ZnO composites, only the CuO component is known to be a strong MW absorber. Therefore, hot spot formation at CuO sites decorated on the ZnO support is likely to result in pronounced microstructural rearrangement at their interface due to selective MW absorption by CuO. Using these MW treated CuO-ZnO inverse opals as catalyst precursors for methanol reforming reaction, not only higher methanol conversion efficiency (over 96%) and higher hydrogen production rate (more than 230 mmol per g-hr) can be obtained at reduced operation temperature, but also lower CO production (less than 200 ppm) and remarkable stability have been achieved. Detailed studies using XRD, Raman, XPS, XAS, and HRTEM analyses revealed the formation of defect and microstrain and suggested a strong metal support interaction at the Cu/ZnO interface. This may account for the good correlation between the increase in defects in NTs and the enhancement in catalytic performance after MW treatment.

Energy conversion and storage present enormous materials challenges. In the case where the energy carriers are chemical bonds, abundant and effective catalysts facilitating solar-to-fuel and fuel-to-electricity conversions are essential. Because most chemical reactions are thermally-activated, carrying out reactions such as hydrogen evolution (in a solar- or electricity-driven electrolyzer) and oxygen reduction (in a fuel cell) at elevated-temperatures is a promising approach for improving performance. In this talk, I will highlight fundamental investigations of surface chemistry and electrochemistry in several oxide model catalysts (operating above 500 °C) relevant to the above reactions. A variety of fabrication and novel characterization techniques were developed, including photolithography, model-based impedance spectroscopy, and in-situ ambient-pressure X-ray photoelectron spectroscopy. Insights from these fundamental studies are used to guide rational materials discovery and microstructure designs to dramatically enhance device efficiencies.

Ionic-displacement induced electric field and its interaction with the built-in electric field in a space charge region are promising for tuning charge transport behaviors of heterojunctions. Piezoelectric semiconductor materials such as wurtzite structure ZnO and GaN are substantial building blocks for modern solid state devices, whose dual attributes (the piezoelectric and the semiconducting) make it feasible to engineer semiconductor interface by strain. Through the process of photoelectrochemical (PEC) water splitting, we demonstrated an effective strategy for engineering the barrier height of a heterogeneous semiconductor interface by piezoelectric polarization. The PEC anode consisted of a piezoelectric ZnO thin film that was deposited on an ITO substrate. The PEC cell exhibited a consistent enhancement or reduction of photocurrent when tensile or compressive strains were applied to the ZnO anode, respectively. The photocurrent variation is attributed to a changed barrier height at the ZnO/ITO interface, which is a result of the remnant piezoelectric field across the interface due to a non-ideal free charge distribution in the ITO electrode. This phenomenon is different from the piezopotential-driven charge flow and is related to the material and interface properties. From our system, ~1.5 mV barrier height change per 0.1% applied strain was identified, whereas a larger barrier height change per unit strain can be expected for other semiconductor-piezoelectric material interfaces. This discovery renders a new pathway for engineering the interface barrier without altering the interface structure or chemistry, which is promising for improving the performance of many electronics, optoelectronics, and photovoltaic devices.

Yttrium-doped barium zirconate has emerged as an attractive candidate electrolyte for intermediate-temperature solid oxide fuel cells and electrolyzers because of its combination of high proton conductivity, exceeding the ionic conductivity state-of-the-art oxygen ion conductors, and excellent chemical stability. While there has been tremendous progress in the processing of barium zirconate in order to attain reproducibly high conductivities, fundamental aspects of the conduction mechanism remain unclear and hence the crystal-chemical characteristics establishing the magnitude of the conductivity are unknown. Here, we demonstrate that proton transport is governed by association of protons to the dopant species. We determined proton diffusivity in barium zirconate doped with Lu, Y and Gd, by a combined measurement of A.C. impedance spectroscopy and thermogravimetry. For all dopant species, the diffusivity displays downward curvature in the Arrhenius representation, a feature strongly indicative of proton trapping. Evaluation of the curvatures in the context of a trapping model provides proton binding energies of 28, 28 and 41 kJ/mol, for Lu, Y and Gd dopants, respectively. The experimental proton binding energies plotted with respect to the ionic radius thus suggest a volcano-like plot in which the lowest binding energy lies between Lu and Y dopants. Both electrostatic and elastic effects of dopant likely impact the proton binding energy in the oxide, resulting in an optimal intermediate value for dopant size. Tailoring the dopant ionic radius between Lu and Y, such as Er, is anticipated to improve the proton conductivity in the doped barium zirconate.

GaN/ZnO solid solutions emerge as promising photocatalysts for the overall water splitting reaction (H2O â?' H2+1/2O2) because its bandgap is considerably smaller by about 1 eV than those (3.3 eV) of GaN and ZnO semiconductors. Such a significant bandgap reduction makes it possible to prepare materials capturing a larger fraction of the incident solar spectrum than traditional photocatalysts, such as TiO2. Dimensionality is a parameter of the utmost importance in a material changes in spatiality may introduce dramatically different properties. Doping is another important factor for controlling these properties. Here, we will present our approaches to synthesize GaN/ZnO solid solution nanosheets with indium-incorporation. The thickness of these nanosheets can be as low as about 2 nm. It is found that indium plays an important role in the formation of the nanosheets, and contributes to their narrow bandgap energy of about 2.55 eV. An X-ray absorption near- edge structures study reveals that indium increases chemical disorder. X-ray excited optical luminescence measurements show that indium atoms distort the crystalâ?Ts lattice, thus reducing zinc-acceptor impurities in the GaN lattice. We will also discuss its catalytic activities.

The photoelectrochemical (PEC) cell is well-known to utilize the photoelectrolysis (solar energy-assisted water splitting H2 generation process) and is believed to be the essence of H2 economy. Photocathode is an active electrode in photoelectrolysis that supplies photo-generated electrons to react with protons in the electrolyte and form H2 gas directly. Si is very efficient in broad absorption of solar spectrum due to its small band gap and thus efficient photo-induced carrier generation, low-cost, abundant, and can be potentially a good photoelectrode candidate. However, Si is not favorable for spontaneous water splitting due to the energy band level position with respect to the water redox levels. Herein, we report the design, fabrication, and characterization of metal nanoparticles and metal oxides nanowires co-catalyzed p-type Si photocathode. Our heterogeneous photoelectrode consists of a n-type metal oxide (ZnO or TiO2) layer sandwiched between the metal nanoparticle coating and the underneath p-type Si. We demonstrate that the formed heterojunction junction can improve the band bending in Si, minimize the electron-hole recombination, improve the interfacial change transfer kinetics, and maximize the efficiency of direct carrier transfer to ions in solution. Effects such as doping and band-gap of the metal oxide coating, morphology of metal oxide nanowires, etc. on the overall PEC performance are studied and show enhanced photocurrent and shifted onset potential to the positive of water reduction level. Ni is used to replace expensive Pt as a H2 evolution catalyst and shows high efficiency (any specific numbers). This study offers an understanding of photoelectrolysis using a heterogeneous system and enables a design and optimization of the heterogeneous photoelectrode for practical PEC devices and solar H2 generation.

Naturally occurring mineral oxides and sulphides that are found abundantly in the earth offer an enticing alternative to Si as the active layer in semiconductor solar cells. Cuprite (Cu₂O) is one example of an archetype hole-doped semiconductor of such oxide minerals that possesses several interesting characteristics useful for solar cells production: low cost, non-toxicity, good mobilities, fairly high minority carrier diffusion length, and direct energy gap of approximately 2 eV. Cu₂O also lends itself to cheap and scalable fabrication techniques like thermal oxidation, electrochemical deposition, etc.. As the lack of high quality n-Cu₂O precludes the use of a homojunction, semiconductor heterojunctions (or Schottky junctions) are required to make photovoltaic (PV) devices with Cu₂O as the absorber. However, to date Cu₂O PV devices have shown poor experimental performance compared to the promising efficiencies from theoretical predictions. We have found that the performance of n-type La-doped SrTiO₃/p-type Cu₂O heterojunctions is superior to that of other heterojuctions studied till date in relevance to the high open circuit voltages in the range of 600-800 mV that we observe for this device structure. A comparative study of such devices with Cu2O obtained on the one hand through thermal oxidation of Cu sheets (polycrystalline, large-grained Cu₂O) and on the other by pulsed laser deposition, PLD (singly-oriented, small grains with low-angle grain boundaries) was conducted. The importance of grain boundaries in Cu₂O affecting the device efficiency was studied through I-V characteristics and conducting atomic force microscopy (c-AFM) measurements of our heterojunction devices in dark and under AM1.5 illumination. c-AFM of our solar cell structure reveals photo-induced shunt paths at the boundaries in the Cu₂O. This effect is pronounced in the PLD grown small grained samples, as the grain boundary cross-section of the low angle boundaries in these samples is higher than in the thermally oxidized samples. The effect of reducing the effective grain boundary cross-section in the large grained samples on, specifically, the fill factor by reducing the shunt paths in the device will be discussed. Supported by the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.

Although Zn based semiconductors ZnX (X = O, S, Se, Te) lend themselves to facile synthesis in nanowire and nonotube form, their large band gap prevent their application in optical devices, such as solar cells and LEDs. We present a way to engineer (lower) the band gap through uniaxial strain. In this ab initio work, we show that ZnX in the wurtzite crystal structure exhibits the following universal behavior: under uniaxial compressive stress along the wurtzite c-axis ([0001] direction), the band gap of the system initially increases, goes through a maximum, and then decreases abruptly, allowing for a wide range of possible uniaxial stress-induced band gaps. We consider ultrathin nanowires with diameters in the 1-3 nm range and bulk ZnX, thus bracketing the experimentally achieved diameters of ~50 nm and hence their properties. Our calculations were performed using standard density functional theory (DFT), hybrid DFT, and quasiparticle GW methods. Although DFT has the well-known deficiency of underestimating band gaps, trends in changes in the band gaps (e.g., with respect to strain), is expected to be well represented. This widely held notion is confirmed by our hybrid DFT and GW calculations. While quantum confinement always increases the nanowire band gap relative to the corresponding bulk values, uniaxial tensile and large compressive strains reduce the band gap relative to the equilibrium situation, similar to the behavior in bulk ZnX [Yadav et. al, Phys. Rev. B, 81, 144120 (2010)]. The band gap reduction under compressive strains is accompanied by the onset of a structural transformation from the wurtzite to a graphite-like coordination environment: each Zn atom becomes coplanar with the 3 X atoms below in one layer; likewise, in the adjacent layer, each X atom becomes coplanar with the underlying Zn atoms. These results have important implications for the engineering of the band gap of ZnX systems through uniaxial strain, achievable in core/shell type nanowires.

The theoretical maximum conversion efficiency of a solar cell using a single band gap is around 30% [1]. About 45% of energy loss in the conversion of solar energy is due to the so-called spectral mismatch: low energy photons are not absorbed by the solar cell while high energy photons are only partly used by the solar cell. Therefore either the solar cell can be adapted to use the solar spectrum more efficiently â?" e.g. by combining multiple semiconductor materials with different band gaps (such tandem cells can reach more than 40% efficiency [2]) â?" or the solar spectrum can be adapted to the solar cell with the use of a photon conversion layer. In the last case, either two lower energy photons can be added to obtain one higher energy photon that can be absorbed by the solar cell (upconversion) or one higher energy photon can be split to obtain two photons with a smaller energy that can then be absorbed by the cell (quantum cutting phenomenon used in downconversion (DC)) [3]. We investigate the potential of Nd doped SrTiO₃ films for the photon down shifting of ultraviolet photons to infrared ones. We first studied the structural properties of such newly developed layers grown by pulsed laser deposition (PLD). We also show that these layers allow the conversion of photons from 250-350 nm excitation to 900 nm. We implemented the optimized Nd:SrTiO₃ thin films on a conventional silicon solar cell and we show that the internal quantum efficiency is increased at around 350 nm which validates the concept of downconversion applied to solar cells. [1] W. Schokley and H.J. Queisser, J. Appl. Phys. 32, 510 (1961). [2] M.A. Green, K.Emery, Y. Hishikawa, and W. Wartta, Prog. Photovoltaics 17, 85 (2009). [3] T. Trupke, M.A. Green, and P. Wurfel, J. Appl. Phys. 92, 1668 (2002).

Tungsten oxide (WO3) is a good photoanode material for water oxidation but it is not an efficient absorber of sunlight because of its large band gap (2.6 eV). Recently, stable clathrates of WO3 with interstitial N2 molecules were synthesized [1], which are isostructural to monoclinic WO3 but have a substantially smaller bang gap, 1.8 eV.We have studied the structural, electronic, an vibrational properties of N2-WO3 clathrates using ab-initio calculations and analyzed the physical origin of their gap reduction. We found that both electronic and structural effects are responsible for such reduction: a small yet non negligible charge transfer between N2 molecules and the WO3 lattice, combined with structural modifications of the host lattice induced by the molecule. We also studied the effect of atomic dopants, in particular rare gases. In the case of Xe we observe a strong hybridization between the host atom and the lattice, leading to a reduction of the band gap of about 0.9 eV. [1] Q.Mi et al. (preprint)

High energy density capacitors are to store more than ten times as much energy per unit volume than the common capacitor (device for storing electric charge) , and the accumulated energy in dielectrics is determined by the dielectric permittivity and the breakdown strength, which varying linearly with the dielectric permittivity and quadratically with the electric filed. We have developed lead free high energy density capacitor materials, {Ba(Zr0.2Ti0.8)O3}(1-x) {(Ba0.7Ca0.3)TiO3}x [x = 0,0.10,0.20,0.30,0.40,0.50,0.60,0.70,0.80,0.90,1.0 --(BZT(1-x) BCTx ] with high dielectric constant and moderate breakdown voltage. The ceramic materials were prepared using high energy ball milling for 4 hours at 400 rpm. The ball milled powders were calcined at 1250C for 10hrs. Ceramic pellets having 13mm diameter were prepared using hydraulic press (2 ton) and sintered at 1500C for 4 hrs. X-ray diffraction studies of the sintered pellets revealed the phase pure perovskite crystal structure. The crystal structure was further confirmed by Raman spectral analysis. The SEM images revealed monolithic grain growth in samples sintered at 1500C. The preliminary data show high dielectric constant and moderate polarization (~Ps~ 15-25 µC/cm2) with saturated ferroelectric loops, high piezoelectric coefficient (d31) ~380pC/N, moderate breakdown field ~ 153kV/cm and high energy density of 7.48 J/cm3 was obtained in the sintered pellets. Details of the results will be presented at the meeting.

There have been growing interests in transparent conducting oxides (TCOs) such as ZnO, In₂O₃, and SnO₂ due to their superior optical transparency and electrical conductivity, which leads to various optoelectronic applications. Recently, studies on amorphous compounds of TCOs such as InGaZnO₄ (IGZO) have been attracting a great deal of interests due to their application as high-mobility channel layer in TFT. Even though many studies have emphasized that the spherical symmetry and the large size of s orbitals of metal ions mainly contribute to the high electronic mobility, full microscopic understanding has not been achieved yet. In particular, first-principles calculations confirm that oxygen p orbitals are also important for low-energy structure of conduction bands in TCOs. In this study, we apply a tight binding (TB) approximation to analyze the conduction bands of crystalline or amorphous TCOs. We explicitly consider oxygen s and p orbitals in addition to metal s orbitals. We identify parallel or isotropic coupling conditions inherent in the local configurations found in most TCOs. The result of TB model shows that the conduction bands of TCOs show the dispersion relation of massive Dirac particles (E= â^s(ε²+γ²k² )). As such, linear-dispersion bands appear at high energies or high carrier densities. The linear dispersion relation is a result of overlap integral between metal s and oxygen p orbitals. As an application of our model, we evaluate the electronic mobility of single crystalline ZnO by considering ionized impurity scattering and optical phonon scattering. It is found that the mobility has a minimum value at the carrier density of 10¹⁹cm^-3 which is in good agreement with experiment. From the detailed analysis, the upshift at higher carrier densities is a result of the increase in the group velocity and the suppressed electron scattering that is characteristic of the linear band.

Grain boundaries can significantly affect the functional propertis of oxide ceramics. This is particularly evident in the case of polycrystalline BaZrO3 where the excellent bulk proton conductivity is limited by high grain boundary resistivity. Such effects can qualitatively be explained by space-charge formation around a positively charged grain boundary core, but detailed understanding of the structure and defect chemistry of such interfaces are scarce. In the present work, the defect structure of a SUM3 111 grain boundary is investigated by defect calculation performed with Density Functional Theory. The concentration of defects in bulk and in the grain boundary core are calculated, and the concentration of defects in the resulting space-charge region is derrived. Variations in the defect structure and Schottky-Barrier height with external parameters such as temperature and water vapor pressure are discussed. Our calculations are consistent with the general conception that the grain boundary core is positively charged due to accumulation of oxygen vacancies and/or protons. We also perform calculation with effectively negative nitrogen acceptors dissolved into the grain boundary core and discuss whether this process can suppress space-charge formation.

The solar cells are very important technology in the growing area of renewable energy. In this context, TiO2 is the prototype metal oxide because of its catalytic activity. Pure TiO2 has a wide band gap (â^¼3.2 eV for anatase and ~3.0 eV for rutile) that favors absorption of ultraviolet light. For this reason, pure TiO2 is not good for photocatalysis. However substitutional cation and anion doping of TiO2 have been pursued as approaches to extend light absorption into the visible-light region. Doping with nitrogen is considered to be promising because it brings an acceptor level above the valence band maximum (VBM). This state leads to effective band gap narrowing. However, studies also show that partially occupied impurity bands can act as recombination centers and reduce the photo-generated current.[1] One way of overcoming this difficulty is anion and cation co-doping. Recent experiments report significant enhancement of photocatalytic activity for W/N co-doped titania.[2-4] Within density functional theory (DFT) framework, we use Heyd-Scuseria-Ernzehorf screened hybrid functional (HSE06)[5] to calculate electronic and atomic structures of W/N and W/S co-doped anatase TiO2. Mixing of semi-local PBE and non-local HF exchange energies improves description of the defect states and of the band gaps. An exact exchange contribution of 22% gives material properties in good agreement with experimental data for W/N co-doped anatase TiO2 as well as for pure rutile and anatase phases. In this context, we investigate co-doping of titania with W and S as an alternative photoabsorption system. Our results suggest W/S-TiO2 to have strong visible light response as good as W/N co-doped anatase catalysts. [1] Y. Gai, J. Li, S. S. Li, J. B. Xia and S. H. Wei, Phys. Rev. Lett. 102, 036402â?"036405 (2009) [2] B. Gao, Y. Ma, Y. Cao, W. Yang and J. Yao, J. Phys. Chem. B 110, 14391â?"14397, (2006). [3] Y. F. Shen, T.Y. Xiong, T.F. Li and K. Yang. Appl. Catal. B. Environ. 83 , 177â?"185 (2008) [4] A. Kubacka, G. Colon b, M. F. Garcia, Appl. Catal. B: Environ. 95, 238-244, (2010). [5] J. Heyd, G. E. Scuseria, M. Ernzerhof, J. Chem. Phys. 118, 8207 (2003).

Silicon, as one of the most abundant material on earth and has the most advanced manufacturing technology, is wildly used in solar panels for solar energy conversion. On the other hand, although Si for solar-powered water decomposition has been studied since the early 1970s, to the best of our knowledge, there is no solid solution to Si based photoelectrode for this particular application. Our approach towards solving this problem and reported here is to utilize material integration strategies, targeting to reduce the Si overpotential to the thermodynamic water oxidation potential and improve its long-term stability by preventing anodic passivation on the Si surface. Protection layers are deposited through RF magnetron sputtering, atomic layer deposition, or sol-gel spin coating, after which a thin layer (5 nm) of Co metal is deposited using electron beam evaporation on the protection layer and then converted to cobalt-phosphate (Co-Pi) using an electrochemical method at neutral pH reported before. Three different wide band gap materials with different intrinsic dopant type are deposited on n-Si substrates without removing native oxides, including n-type, p-type and highly conductive metal oxides, which are TiO2, NiOx and ITO, respectively. These materials themselves are fairly stable in aqueous solution of natural pH and under moderate biases. It is found that all these materials can significantly improve the durability of electrodes compared to that of the unprotected n-Si. Furthermore, thickness effect of these three types of functionalization layer is systematically studied. It is found that since TiO2 is a strong oxidation photocatalyst, increasing the thickness of TiO2, electrochemical measurement showed a behavior transition from n-Si tunneling current to recombination limited photocurrent. Thickness of ITO typically affects the resistance of the funcationlization layer and thus ohmic loss, also light intensity at the Si/ITO interface as well. Black Si with nanoscale textures on the surface is prepared using metalâ?"assisted chemical etching. Light absorption is greatly improved after 8 mins etching at room temperature. Same protection layers are deposited on the textured Si substrates. With the increasing of the depth of the porous layer on substrates, current density is generally increased and most importantly on-set potential is effectively shift to the negative potential. However, none of systems are capable to give an on-set potential negative to the water oxidation level even with the Co-Pi catalyst. Sol-gel prepared NiOx coated n-Si photoelectrode effectively moves the on-set potential to the negative of the water oxidation level gives dramatically improved photocurrent density. Our research indicated the great potential of using heterogeneous black Si photoelectrodes with proper surface functionalization for efficient and durable water splitting.

Because of their unique ability to reconcile high transparency with good electrical conductivity, transparent conductive oxides (TCOs) are key materials in many applications such as organic light-emitting diodes or photovoltaic solar cells. With a 85% transmittance and a resistivity of a few 10^-4 Ω.cm, the prevailing TCOs, indium tin oxide (ITO) and, more recently, aluminium zinc oxide (AZO) are already well optimized in terms of electrical and optical properties for transparent electrodes [1]. However, good charge collection in photovoltaic devices requires not only high conductivity, but also a proper Fermi-level alignment at the photoactive layer/TCO interface in order to avoid depletion and carrier trapping near the interface. Simulations on heterojunction solar cells [2] have shown the critical dependence of the device efficiency on the TCOâ?Ts work function (WF). It is thus of prime interest to accurately adjust this subtle physical quantity [3] without, in doing so, altering its optical and conductive properties. In this work, we investigate ways of tuning upward as well as downward the WF of DC-pulsed argon-sputtered AZO thin films and consecutively qualify their performances as a contact with hydrogenated amorphous silicon (a-Si:H) absorber layers used in thin film photovoltaic solar cells. Accurate WF measurements are performed by the Kelvin probe method using a stable calibration with aluminium and ruthenium samples. First, we find a marked increase (almost 0.5 eV) of the WF of AZO after an exposure to oxidizing plasma and a clear decrease (0.3 eV) after argon plasma exposure, whereas no stable effect of annealing could be evidenced. Second, we report successful attempts to adjust the WF by the application of a thin buffer layer of chosen metal oxides: a 30nm-thick intrinsic ZnO layer lowers the WF, while tungsten oxide increases it. Sheet resistance is monitored at each step and does not show significant increase. Partial understanding of the surface chemistry causing these important WF shifts is provided by angle-dependent X-ray photoemission spectroscopy (XPS) analysis. Finally, contact resistance between p- or n-doped a-Si:H and these TCOs with differently tuned WF is measured using the transmission line method. This approach highlights the importance of WF matching to obtain good contacts. [1] Exarhos G. J., Thin Solid Films (2007), 515 [2] Centurioni E., IEEE Electron Device Letters (2003) 24 3 [3] Cahen D., Kahn A., Advanced Materials (2003) 15 4

P-type transparent conductive oxides (TCOs) are key materials for the development and widespread use of transparent electronics. Up to date, the performance of p-type TCOs is considerably lower than those reported for n-type TCOs, thus new material with better properties are still required. In this study we report on a new, non delafossite p-type TCO, obtained by co-doping of chromium oxide (Cr2O3) with magnesium and nitrogen. We demonstrated that nitrogen enhances the optical transparency of the films in the visible range while, Mg improves the conductivity while retaining the p-type character of the material. Co-doping with both elements produces a p-type oxide with resistivity as low as of 3 Ωcm and transmission as high as 65 %. These values are higher or comparable with those reported in literature for other p-type TCOs despite the fact that the material was deposited by spray pyrolysis. This technique allows for a rapid screening of material compositions with the drawback of producing film with poor morphology in comparison with the same material deposited by more sophisticated techniques. Therefore there is room for improve the performance of Cr2O3 by using more advanced techniques. To this end, preliminary results on samples grown by PLD will be presented.

Transparent conducting materials (TCMs) are essential components of a number of opto-electronic devices such as thin film photovoltaics (PV) and light emitting diodes (LED). Materials possessing both high optical transparency and high conductivity are almost universally n-type. Thus hole conductors (p-type) of similar quality have the potential to open a new design space for optoelectronic devices and transparent electronics and therefore are of great scientific interest. We report on our efforts to develop a p-type, transparent conducting thin film composed of earth-abundant elements. Our first-principles calculations of the defect formation energy shows that, compared with ZnO, the p-type doping of ZnS is possible. ZnS can be doped heavily with Cu and it can remain p-type as the charge compensation caused by the â?ohole-killersâ? such as anion vacancies can be avoided. Thin films of Cu-doped ZnS were synthesized using pulsed laser deposition with Cu contents in the range of 5 to 15 atomic %. Thermopower and Hall effect measurements confirmed the p-type nature of our films. We find that transparency and resistivities of the films are among the highest and lowest, respectively, reported for a p-type transparent conducting material. The resistivities are as low as 19 mΩ-cm and these same films exhibited an optical transparency of greater than 70% across the visible spectrum. Transparent rectifying diodes were fabricated and the I-V characteristics of these -type Cu-doped ZnS/n-type ZnO heterostructures will be discussed. By utilizing a transparent p-type conductor, photovoltaic cells could potentially be manufactured using new types of absorber materials. In addition, the development and optimization of such a p-type TCM could open new frontiers in the development of transparent transistors, light emitting diodes, electro-chromic windows and a plethora of additional optoelectronic devices.

Transparent conducting oxides (TCOs) having n-type conductivity has been reported in several binary compounds, for example in doped In2O3, ZnO, and SnO2. On the other hand, p-type TCOs having optimum band gap, high concentration of holes, and high mobility are rare. The nature and makeup of the valence and conduction bands, presence of native defects that promote hole compensation through activation of donor defects, or local lattice distortions that cause formation of small polarons could be some of the contributing factors that makes the search for p-type TCOs difficult. Thus far, the search for candidate TCO materials have been primarily guided on the basis of the requirement for optical band gap. Since the common p-type binary oxides such as Cu2O or Ag2O do not meet the requirements in terms of band gap and hole mobility, it is necessary to survey a much larger pool of ternary compounds. In order to screen through such a search space, we studied a set of design principles that a candidate material must meet in order to qualify as a potential p-type TCO. These design principles require: (i) low formation energy and shallow nature of the hole producing defects (ii) large formation energy of the hole killing defects (iii) thermodynamic stability of the compound within a relatively large range of oxygen chemical potential (iv) large hole mobility and (v) optical band gap ~3.1 eV or greater necessary to achieve transparency. To exemplify this procedure of search, we used Cu2O and V2O5 and Ag2O and V2O5 as the baseline compounds that can combine to form the ternaries Cu3VO4 and Ag3VO4, respectively. Using first-principles methods we have studied the properties of these ternary oxovanadates in light of the above design principles. G. Trimarchi et al., Phys. Rev. B 84, 165116 (2011). Supported by the US Department of Energy, Office of Basic Energy Sciences as part of an Energy Frontier Research Center.

Transparent conducting oxides (TCOs) for thin film photovoltaics must be highly conducting and transparent in a broad spectral range. The current TCO technology is based on materials such as In₂O₃:Sn (ITO) and ZnO:Al (AZO) with resistivities in the low 10^-4Ω-cm range that are highly transparent in the wavelength range of 300 to 1000 nm. Transparency of conventional TCOs in this limited spectral range arises from the high absorption in the near infrared region due to the high free carrier concentration in these materials. This prevents using TCOs in many PV technologies including Si and high efficiency multijunction solar cells. In order to maintain low resistivity with reduced free carrier absorption, TCOs with high mobility is required. Among all TCOs, cadmium oxide (CdO) is known to have the highest intrinsic electron mobility. However, as TCOs, undoped CdO films suffer from severe UV absorption loss due to the relatively small band gap of ~2.2-2.6eV. In this work we explored the electrical and optical properties of CdO, undoped and doped with various n-type dopants, grown using pulsed laser deposition (PLD). We found that nominally undoped CdO films have Hall mobilities in the range of 50-150 cm²/Vs with carrier concentrations of 3x10²⁰-2x10¹⁹ cm^-3. Comparison of these results with the calculated electron mobility indicates a high compensation ratio of 0.7 in udoped CdO. Intentional doping of CdO with Ga and In increases the electron concentration to 5-15x10²⁰ cm^-3 with the corresponding mobilities of 300-150 cm²/Vs and extremely low resistivity of less than 5x10^-5Ω-cm. Moreover, because of the Burstein-Moss shift, the optical band gap of the doped films shifts to>3 eV. We will show with an optimal doping, the optical transmittance of CdO can be extended to λ>1500 nm, making it an ideal TCO for thin film photovoltaics especially for structures incorporating low band gap absorbers.

Despite many years of research aimed at finding more efficient oxide semiconductors, titanium dioxide (TiO2) is still one of the best materials for photocatalysis and solar energy conversion. Rutile is the thermodynamically most stable TiO2 phase, while anatase is widely considered more important for photocatalytic and photovoltaic applications. This talk will report on hybrid functional electronic structure calculations of the structure and energetics of photo-generated electrons and holes in the bulk and at the majority (101) surface of anatase TiO2. Excitons formed upon UV irradiation are found to become self-trapped, with electron and hole polarons localized at Ti3+ and Oâ?" lattice sites, respectively. At the surface, the trapping sites generally correspond to undercoordinated Ti3+5c and Oâ?"2c surface atoms, or to isolated OH species in the case of a hydroxylated surface. The polaron trapping energy is considerably larger at the surface than in the bulk, indicating that it is energetically favorable for the polarons to travel from the bulk to the surface. Implications for TiO2-based photocatalysis are discussed.

Solid-solid interfaces between different metal oxide particles provide opportunities to synergistically combine their unique photoelectronic properties upon generation of new ones. While a variety of experimental approaches leading to the coupling of two semiconductors have been successfully employed for colloidal systems [1] there are very few reports about the charge separation properties of dry particle systems that feature corresponding types of interfaces. TiO₂, SnO₂ and ZrO₂ particles were prepared by chemical vapor synthesis, transformed into colloidal dispersions and subsequently aggregated to yield mesoporous nanoparticle networks. Annealing in vacuum was used to generate particle contacts and to also enable the investigation of solid-gas interface effects. A comparison with sol-gel derived titania aerogels revealed critical new insights into the structure-property relationship.[2] Under controlled experimental conditions in terms of temperature and pressure as well as energy and flux of photons, the activation of molecular oxygen at the surface of photoexcited nanocrystals was used to compare with UV-Vis diffuse reflectance and electron paramagnetic resonance the photoactivity of nanoparticle networks with different concentrations of homo- and heterojunctions. As a major result we found that networks of intermixed nanoparticles exhibit substantially enhanced crossections for charge separation.[3] The superior performance holds also in comparison to homogeneous metal oxide nanoparticle networks that exhibit a higher charge recombination loss in comparison to powders of unconnected nanocrystals.[4] Thus, the adjustment of the composition and concentration of solid-solid interfaces inside the nanoparticle network clearly determines the branching ratio between the separation and recombination of photogenerated charge carriers. The approach of particle intermixing and aggregation cycles presents an interesting new concept for the synthetic realization of heterojunctions in highly dispersed materialsâ?T systems. [1]L. Carbone, P.D. Cozzoli, Nano Today 2010, 5, 449. [2]S. Baumann et al. Langmuir 2011, 27, 1946. [3]N. Siedl et al. J. Phys. Chem. C 2009, 113, 15792. [4]N. Siedl et al. J. Phys. Chem. C, 2009, 113, 9175.

We present here two studies on thin perovskite films grown und single crystal substrates with PLD for high temperature fuel cell and photoelectrochemical cell basic studies with synchrotron radiation. Reversible and irreversible discontinuities at around 573 K and 823 K in the electric conductivity of a strained 175 nm thin film of (La0.8Sr0.2)0.95Ni0.2Fe0.8O3-δ grown by pulsed laser deposition on SrTiO3 (110) are reflected by valence band changes as monitored in photoemission and oxygen K-edge x-ray absorption spectra (XAS). The irreversible jump at 823 K is attributed to depletion of doped electron holes concomitant with reduction of Fe3+ towards Fe2+, as evidenced by oxygen and iron core level soft XAS, and possibly of a chemical origin, whereas the reversible jump at 573 K possibly originates from structural changes. An approximately 125 nm thick pulsed laser deposited blue, nonstoichiometric WO3 film grows on TiO2 (110) in the [220] direction. Oxidative treatment at 400 C turns the film color from blue to yellow and improves the film quality considerably, as shown by improvement of the Kiessig oscillations in the X-ray reflectometry curves. Detailed analysis of resonant valence band photoemission spectra of the as-deposited nonstoichiometric blue film and oxidized yellow film suggests that a transition near the Fermi energy originates from the nonstoichiometry, i.e., oxygen deficiency, and insofar poses electronic defect states that partially can be eliminated by heat treatment in oxygen. The defects of the as-deposited blue film seem to be located throughout the film, except for the top surface due to exposure to oxygen in ambient air. Thermal after treatment under oxygen heals the defects in the bulk, whereas residual defect states appear to remain near the film substrate interface. Potential strain at the substrate film interface due to lattice mismatch may be one origin for the remanence of the defect states in the bulk. 1) A. Braun, X. Zhang, Y. Sun, U. Müller, Z. Liu, S. Erat, M. Ari, H. Grimmer, S. S. Mao, T. Graule, Correlation of high temperature X-ray photoemission valence band spectra and conductivity in strained LaSrFeNi-oxide on SrTiO3(110), Applied Physics Letters 95, 022107, 2009 2) A. Braun, S. Erat, X. Zhang, Q. Chen, T.-W. Huang, F. Aksoy, R. Löhnert, Z. Liu, S. S. Mao, T. Graule Surface and bulk oxygen vacancy defect states near the Fermi energy in 125 nm WO3-δ/TiO2 (110) film: A resonant valence band photoemission spectroscopy study, J. Phys. Chem. C 2011, 115, 16411â?"16417.

Copper oxide has received renewed attention in recent years as a non-toxic and earth-abundant solar absorber material. The most efficient Cu2O-based photovoltaics have relied on ZnO as an n-type window layer for forming a p-n junction. In this work, we employ ZnO films processed via three different solution methods: electrodeposition, zinc acetate decomposition, and diethyl zinc decomposition. The ZnO films are extensively characterized to gain insight into variations in defect density, film microstructure and morphology, as given through x-ray diffraction data, SEM, and photoluminescence measurements. Solar cells are prepared by electrodepositing Cu2O on top of the ZnO, thereby eliminating possible damage to the Cu2O during ZnO deposition. Photovoltaic performance varies substantially among the three samples. We use light and dark J-V measurements, in addition to EQE and impedance spectroscopy, to correlate changes in the electronic, structural, morphological, and interfacial properties of the ZnO layer to observed changes in device performance. From this, we draw conclusions regarding routes to higher-efficiency ZnO/Cu2O photovoltaics.

Dye sensitized solar cells (DSSCs) are electrochemical cells, where dye molecules are used as light absorbers and a titania nanoparticle network is used for electron transport. DSSCs attracted great attention due to their respectable efficiency with very low fabrication cost, good performance under diffuse light conditions and ability to be fabricated on flexible substrates. Its main efficiency limiting factor is the random hopping of electrons within the titania nanoparticle network, which causes carrier trapping and recombination. The charge transport and collection can be enhanced by employing ordered nanostructures such as nanowire or nanotube arrays. However, nanowire/nanotube based DSSCs with very high efficiencies have yet to be demonstrated, primarily due to their lower surface area for dye adsorption and surprisingly low carrier mobility similar to those of random nanoparticle networks. In this talk, we report the fabrication of DSSCs using highly crystalline free-standing titania nanotube arrays. The high crystallinity leads to high electron mobility and diffusion length, allowing thick nanotube films to be used for improving the long wavelength light absorption. This greatly enhances the photocurrent and power conversion efficiency as compared to that of nanotube DSSCs in earlier studies. By employing a thick titania nanotube layer of 70 µm, retaining highly crystalline anatase structure at high annealing temperatures and solving the contact issue between the nanotube and transparent conducting electrode, DSSCs with power conversion efficiency higher than 10% and short circuit current density of 30mA/cm2 has been achieved. At the end of the talk, I will discuss our attempts to fabricate all-inorganic radial heterojunction solar cells, which allows efficient charge carrier separation without compromising light absorption. * Supported by NSF DMR1104994 and NYSTAR.

To achieve significant improvements in the efficiency of dye sensitized solar cells (DSSCs) there is interest in developing effective sensitizing dyes that absorb in the infra-red (IR) region of the spectrum. This approach will increase the maximum achievable photocurrent. However the red-shifted absorption in new IR sensitizer dyes is often achieved at the expense of shifting down the dyeâ?Ts lowest unoccupied molecular orbital (LUMO) level. This results in a reduction in the free energy/electronic coupling available between the dye and acceptor states which also results in poor injection efficiency.1, 2 Consequently the photocurrent is reduced, negating the benefits of increased light harvesting from the red-shifted dye. We have explored the use of WO3 and SnO2 nanocrystalline films in DSSCs since they have lower conduction band energies than TiO¬2. Analysis of the spectral response of the devices was used to estimate the injection efficiency.3 The results demonstrate a correlation between the relative conduction band energy of the different substrates and the electron injection efficiency. We demonstrate that these n-type semiconductors and composites of them show promise as alternatives for sensitization with near-IR dyes due to significant increases in electron injection efficiency relative to TiO2. We also present analysis of the strengths and weaknesses of these materials in working cells. 1. S. E. Koops, B. C. O'Regan, P. R. F. Barnes and J. R. Durrant, Journal of the American Chemical Society, 2009, 131, 4808-4818. 2. A. Listorti, C. Creager, P. Sommeling, J. Kroon, E. Palomares, A. Fornelli, B. Breen, P. R. F. Barnes, J. R. Durrant, C. Law and B. O'Regan, Energy Environ. Sci., 2011. 3. P. R. F. Barnes, A. Y. Anderson, S. E. Koops, J. R. Durrant and B. C. O'Regan, Journal of Physical Chemistry C, 2009, 113, 1126-1136.

Ferroelectric materials exhibit a range of photochemical anomalies that make them an interesting alternative to the traditional materials under investigation as photocatalysts. Of particular importance is the spatially selective REDOX chemistry that arises due to the band bending imposed by the ferroelectric dipole as a result of the movement of cation in the lattice. A ferroelectric material develops an internal p-n junction that effectively separates the photoexcited electrons and holes which produces spatially selective photochemistry and a range of anomalous and interesting interactions. This means that on a ferroelectric surface there is inherent separation from regions of oxidation and reduction. Recently this anomaly has been demonstrated for the narrow band gap multiferoic material BiFeO3. In this work we show that BiFeO3 can be used to photochemically decolourise rhodamine blue (RhB) by 98% in 10 minutes under AM1.5 illumination. This compared to 30 minutes of irradiation for 97% decolourisation using nanostructured TiO2. We have developed a novel auto-combustion technique that uses a citric acid fuel to produce BiFeO3 that has a particle size centred around 20nm and exhibits and optical band gap onset at 2eV. The as produced BiFeO3 nanopowder is formed of loose agglomerates that are readily dispersed using low energy ultrasonication to produce a high surface area nanostructured catalyst. This material was then used to photodecolourise10ppm RhB(aq) using AM1.5 illumination under a standard apparatus arrangement and constant stirring. The extent of decolourisation was measured using UV-vis absorption of the dye peak. Using XPS we show that there is an inherent instability in the catalyst under an aqueous solution at band gaps around 2eV due to the location of the valance band. At higher band gaps the valence moves below a location where it is possible for the injection of holes into the BiFeO3 can occur. The associated photocorrosion can occur through oxidation of the Fe-O bond in the material if the material is treated as a solid solution or ternary compound. Techniques that are not sensitive to surface changes, such as XRD, can mask the photocorrosion and indicate that the catalyst has not been influenced. If photocorrosion is controlled then the material exhibits exceptional photocatalysis of standard commercial dye materials outperforming TiO2 significantly. The exceptional catalytic performance is explained in terms of the size of the particles, the ability of the particles to exhibit spatial photochemistry which enables the reactions to proceed more readily and control of the Stern layer around the nanoparticles. In summary through the synthesis of a nanostructured material with a measured band gap of 2eV we have produced a highly active photocatalyst. The catalyst performance is inherent to the materials properties of the catalyst due to the multiferoic nature of the catalyst.

Post-transition metal oxides (ZnO, In2O3, SnO2, TiO2, etc.) play an essential role in modern optoelectronic devices because they have many unique physical properties such as superb structural stability in solution, good catalytic activity, simultaneous high electron conductivity and optical transmission. Therefore, they are widely used in energy related optoelectronic applications such as photovoltaics and photoelectrochemical (PEC) water splitting. In this talk, I will discuss our first-principles theoretical study of the band structure, optical, and doping properties of oxides, demonstrating the complexity of the binary and multinary oxides. We find that (i) many oxides have non-equivalent fundamental and optical band gaps, arising from parity forbidden band edge transitions; (ii) The band gap renormalization in highly dope oxides arises from the nonparabolic nature of the conduction band (i.e., not a rigid shift as previous thought), and is highly sensitive to the electronic states of the dopants; (iii) The ground state crystal structure of the non-isostructural (InMO3)m(ZnO)n (M=In, Ga, Al) alloy is complex because they need to satisfy several fundamental rules such as the octahedron and the octet rules; (iv) Amorphous oxides can still have good electrical properties even if they possess many dangling bonds. We have also analyzed the doping asymmetry problem in the wide-gap oxides and proposed approaches to overcome the doping bottlenecks in oxides and tune their material properties for energy applications.

Hydrogen is a common impurity in that can be electrically active as a donor and an acceptor in a semiconductor, leading to a variety of behavior that may impact the desired performance. In one class of materials, wide-band-gap semiconducting oxides, the role of hydrogen has been strongly linked to the n-type conductivity. In these systems hydrogen has been shown to act predominantly or exclusively as a shallow donor as an interstitial or while substituting on an oxygen site. In addition to the resulting n-type conductivity, the incorporation of hydrogen or other donor dopants also leads to an increase in compensating acceptor defects. The interaction of hydrogen with cation vacancies, the dominant acceptor impurity involved in compensation, has yet to be fully explored. Using first-principles calculations employing hybrid functionals that do not suffer from the band-gap problem, we investigate the complexes formed between isolated cation vacancies and interstitial hydrogen donors. We report on the formation energies and binding energies of the hydrogenated cation vacancies in the transparent semiconducting oxides SnO₂, In₂O₃ and Ga₂O₃.^[1] We find that hydrogen can interact strongly with vacancies in both atomic and molecular form, and can have a significant impact on the electrical properties of devices employing these oxides. We also compute the associated vibrational frequencies to assist in the experimental detection of the proposed complexes, comparing to experiment when available. ^[1] J.B.Varley, H.Peelaers, A.Janotti, C.G. Van de Walle, J.Phys. Cond. Matter 23, 334212 (2011).

Tunable bandgaps, facile doping and mobile charge carriers are highly desirable properties of functional oxides for water splitting, solar cells and other optoelectronic applications. In this study such materials were searched for using Inverse Design methodology, which involves intelligent choice of the initial materials search space, followed by high-throughput theoretical screening of this space along with combinatorial thin film deposition and functional property mapping to detect the most promising candidate materials, followed by targeted theoretical calculations and bulk ceramic and single crystal synthesis and characterization, including synchrotron-based diffraction and spectroscopy. In this presentation, examples and results from all stages of the Inverse Design will be presented. As a result of this study, Cr2MnO4 was identified as a promising wide-bandgap high-mobility p-type dopable oxide. Mn-based ternary oxides with A2BO4 stoichiometry were chosen as the search for high-mobility p-type oxides, because Mn2+ ions with d5 electronic configuration in oxides lead to an anti-bonding character of the valence band that is likely to lead to low effective masses and large mobility of holes. In addition, such Mn-based materials are likely to have large exchange splitting, and hence wider band gaps, promising for transparent conductive oxide (TCO) applications. All known Mn-based A2BO4 materials were theoretically screened for electronic and optical band gaps, for hole effective masses and for native point defects. As a result of this search, the Cr2MnO4 normal spinel was identified as the most promising material. Cr2MnO4 with all Mn in d5 electronic configuration and anti-bonding character of the valence band maximum has hole effective mass of less than 10me which is smaller compared to that of Co-based p-type spinels with d6 electronic configuration and non-bonding character of the valence band maximum. Optical absorption onset of Cr2MnO4 is at 3.5 eV, well within the range for transparent conductive oxide (TCO) applications. No defect self-compensation is calculated in Cr2MnO4, but this material has no intrinsic shallow acceptor defects either, so extrinsic doping is required. Cr2MnO4 has nominally all Mn atoms on tetrahedral sites, so no hole self-trapping effects are expected. Cr2MnO4 material was measured to have 2.8 â?" 3.1 eV optical absorption onset in the technologically relevant thin film form and negligible electrical conductivity in the powder form, consistent with the theoretical predictions of the absence of shallow acceptor defects. Theoretical studies aimed to identify effective extrinsic dopants and experimental studies aimed to dope Cr2MnO4 p-type in bulk ceramic, thin film and single crystal forms will be reported. This research was funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, as part of the Energy Frontier Research Center â?oCenter for Inverse Design"

The role of nitrogen defects in a range of oxides (MgO, CaO, BaO, Al2O3, In2O3, Sc2O3, Y2O3, La2O3, TiO2, SnO2, ZrO2, BaZrO3 and SrZrO3) is investigated by First principles defect calculations. The predominating nitrogen related defects (N, NH, NH2) are evaluated by considering the Gibbs energy of various defect reactions and their temperature and partial pressure dependencies. As such, trends in the nitrogen defect chemistry of oxides are discussed. We find that nitrogen in most cases acts as a deep acceptor when substituted on an oxygen site. However, when nitrogen is bound to hydrogen, the otherwise isolated N 2p states are shifted towards the valence band maximum and thereby can lead to band gap narrowing. All the oxides except SrZrO3, exhibit a negative binding energy between N and H, i.e. nitrogen is stabilized by hydrogen. The N-H binding energy is evaluated from total energy defect calculations as well as analysis of the site projected Density of States of N and H.

Transparent conducting oxides (TCO) have the advantage of being electronically conductive and optically transparent in the spectrum range of visible light. However, its application is severely limited by the fact that the most successful TCO materials, such as ZnO, In2O3, SnO2 etc, are hard to dope p-type, although they can all be reliably doped n-type. This disparity in doping is caused by low-lying valence band maximum (VBM) consisting of O p states and conduction band minimum (CBM) mainly consisting of cation s states, i.e. most TCO materials feature large ionization energy and high electron affinity. A promising alternative approach relies on transition metal oxides in which the d orbitals form the major components of VBM states and therefore possess lower ionization energy. However, the most promising TCO of this type, Cu2O is too low in band gap and is hard to be doped n-type. The search for transition metal based TCO materials that alleviate these issues is currently focused on ternary oxides combining transition metal oxides and other oxides that have larger gap and lower ionization energy. The most common oxides of this kind are delafossites such as CuAlO2, CuGaO2 etc. In order to systematically examine how the variation of cations can alternate electronic structures and affect the doping propensity, we conducted first principles electronic structures calculations for ABO2 delafossites in which A=Cu,Ag and B=B, Al, Ga, In and Sc. The ionization energies and the electron affinities are obtained by supercell slab model calculations. The calculations are based on the density functional method (DFT), using a plane wave basis set, together with the projector augmented wave (PAW) approach of describing ionic potentials. To overcome the band gap reduction problem of local and semi-local functionals, we employed the hybrid functional in the framework of Heyd-Scuseria-Ernzerhof (HSE). We show that HSE not only remarkably improve the band gap values but also significantly change the alignment of VBM and CBM, providing us withreliable predictions for the ionization energies and electron affinities. By comparing the band alignments for delafossite as well as for Cu2O and Ag2O, we discuss the general doping trend for these oxide compounds.

The acceptor doping in wide band gap oxides is a controversially discussed issue. The properties of nitrogen in ZnO have been discussed on the basis of theoretical and experimental work. We present unambiguous evidence of single nitrogen on oxygen site being a deep acceptor in ZnO using electron paramagnetic resonance (ESR). The magnitude of the resolved hyperfine interaction of the unpaired spin with the nuclear spin of the nitrogen (I=1) already indicates predominant p-character and interaction with a single nitrogen p-orbital (localized instead of delocalized over four nearest neighbours as expected for a shallow defect). Using additional monochromatic light illumination allows to determine the photo-ionisation cross section from the defect level to the conduction. Modelling the onset and shape of the cross section provides a precise estimate of the single acceptor level at E_VB + 1.4 eV. The nitrogen-related level at 160 meV above valence band obviously involves nitrogen within a complex defect structure. The free hole carrier density in Cu₂O can be influenced by nitrogen doping and can be tuned from mid 10¹⁴ (undoped) to 10¹⁸ cm^-3 (nitrogen doped). The determination of the valence band offset between ZnO and Cu₂O measured by photoelectronspectroscopy shows that the valence band of Cu₂O is about 2 eV higher in energy compared to ZnO, and the nitrogen acceptor has a level 0.2 eV above valence band.

Highly N-type ZnO doping for high mobility transparent conducting oxides (TCOs) requires an understanding and control of electrically-active native point defects that interact strongly with extrinsic dopants. We used depth-resolved cathodoluminescence spectroscopy (DRCLS), optical absorption threshold, and Hall techniques to measure and correlate native point defects systematically with donors, acceptors, and electronic bands in degenerately doped ZnO. These techniques provide conduction band Fermi levels, gap renormalization, and depth-dependent free carrier variations in degenerately Ga-doped ZnO (GZO) vs. growth temperature and ambient gas. Previous studies showed that GZO grown by pulsed laser deposition produced such ZnO with near-record TCO carrier densities [Appl. Phys. Lett. 97, 072113 (2010)] with Hall and secondary ion mass spectrometry measurements showing the only significant acceptors and donors to be Zn vacancy-related (V_Zn â?"R) or Ga_Zn antisites, respectively. [Phys. Rev. B 84, 115202 (2011)]. DRCLS reveals that Zn vacancy-related (V_Zn-R) defects [Phys. Rev. B 81,081201R(2010)] decrease monotonically as Hall free carrier concentration and integrated conduction band emissions increase, demonstrating a process-dependent filling of Zn vacancies by Ga atoms. Zinc vacancies play a major role, acting not only as V_Zn acceptors and Ga_Zn donors, but also as compensating centers as Ef rises. At high temperatures, XPS-measured segregation of Ga to interfaces occurs that uncovers previously Ga-substituted V_Zn, raising (lowering) acceptor (donor) densities. DRCLS free carrier densities inside these sub-micron-thick GZO films can be extracted from DRCLS Ef cutoff energies. These can vary with depth on a scale of tens of nm, introducing spatial variations in Ef â?" Ec that display the same (V_Zn-R) density dependence on a nanometer scale. Comparison of optical absorption threshold, DRCLS Fermi levels and lineshapes with Hall free carrier densities yields benchmark conduction band effective masses and band curvatures. The ratio of integrated spectral intensity for V_Zn(acceptor)/conduction band (donor) increases nearly linearly with Hall-measured N_A/N_D ratio for Ar-grown GZO, but exhibits a dramatically different defect dependence vs. N_A/N_D for GZO grown in forming gas that can be attributed to self-compensation of Ga_Zn donors that form with additional V_Zn acceptors as well as temperature-dependent H passivation of V_Zn. The availability of Zn sites either during growth or subsequent annealing reflects a balance between donor and acceptor formation mediated by native point defects and highlights new approaches to control both by subsequent processing. The ability to quantitatively monitor the acceptor/donor ratio spectroscopically and in depth from the free surface throughout the bulk film provides a novel characterization tool that provides physical insight into the growth and annealing of degenerately-doped metal oxides.

Bandgap engineering of transparent conductive electrodes has potential to benefit the performance of many emerging optoelectronic devices. For example, the efficiency of CIGS solar cells can be improved by increasing the position of the conduction band minimum of ZnO-based electrodes through alloying with MgO because of better electron extraction. However, the well-known issue of electrical property deterioration (i.e. decreased conductivity Ïf, mobility μ, and carrier concentration n) with increasing Mg content in Zn1-xMgxO (ZnMgO) has thus far prevented significant consideration of this material for optoelectronic device applications. This talk highlights our recent work in addressing this bandgap/electrical properties trade-off challenge. Specifically, we investigate how the careful manipulation of intrinsic defects in high Mg (~30%) content ZnMgO can significantly enhance electron generation and transportation. Ga-doped ZnMgO thin films with 30 at. % Mg were grown using Pulsed Laser Deposition on sapphire in Ar/O2 atmosphere. To control the concentration of intrinsic point defects, we varied the Oxygen partial pressure (PO2) during deposition between 1Ã-10-7 Torr to 5Ã-10-3 Torr, keeping the total pressure constant. The fundamental absorption onset of the reported ZnMgO:Ga samples was at 3.9 eV, which is 0.4 eV larger compared to base-line ZnO:Ga sample. Relatively high electrical conductivity values (Ïf) in excess of 350 S/cm, which are record values for ZnMgO with 30 at. % Mg, were commonly observed for films deposited at all PO2 lower than 1Ã-10-4 Torr, with Hall mobilities (μ) of ~11 cm2Vâ^'1sâ^'1 and free electron densities (n) of ~2.0x1020 cm-3 . Sharp decreases of Ïf, n and μ were observed with an increase of PO2 above 1Ã-104 Torr. Sharp reductions in Ïf, n and μ with increasing PO2 were observed for films deposited with PO2 > 1Ã-10-4 Torr. This critical value of PO2 and the slope of log(n) as a function of PO2 were consistent with the result of our first-principles density functional theory (DFT) calculations. However, typical n values obtained from experiments were an order of magnitude higher than theoretical predictions. These theoretical and experimental results were rationalized by considering thermodynamic and kinetic effects during the growth. We will report on experiments which are exploring this hypothesis by annealing a set of sample originally deposited under different PO2 conditions to examine their relaxation towards an equilibrium state. In addition, results from temperature dependent Hall, XRD, AFM, and FTIR will be presented in discussing the potential mechanisms for the observed μ dependence on PO2. Overall, this study addressed the bandgap/electrical properties trade-off challenge in ZnMgO by manipulating intrinsic defects, and provides insight into the impact of intrinsic defects on electron generation and transport in a bandgap-engineered ZnMgO through direct comparison of the experimental and theoretical results.

Aluminated ZnO (ZnO:Al, AZO) is a widely used transparent conducting oxide (TCO) for thin film solar cells. Depending on the actual doping, the band gap of AZO is found between 3.35-3.45 eV. This limits the amount of UV light actually transmitted into a thin film solar cell with a TCO front contact. In particular for cells optimised for diffuse lighting conditions the parasitic absorption within the TCO is a critical factor, reducing the light available for the absorber. In order to change this absorption we demonstrate the tunability of the AZO absorption edge by introducing additional magnesium up to nominal Mg/Zn ratios of 0.2. The inclusion leads to shifts in the band gap of up to 0.2 eV, while maintaining conductivity of the film. In addition changes the refractive index and work function are lowered, making the quaternary TCO an interesting material for conductive index matching layers and buffer layers. Samples have been prepared by spray pyrolysis using organic solvents and various aluminium and magnesium precursors. We will present a throughout characterisation of the crystalline quality, electrical and optical properties as function of Mg concentration.

Metal oxide (MO) films (ZnO and CuO) were synthesized by hydrothermal methods and treated with hydrogen and oxygen plasmas. From uv-visible transmittance spectra, we have found that the optical band gaps of MO films blue-shifted with hydrogen plasma treatment, but red-shifted with oxygen plasma treatment. By alternating the treatment sequence of hydrogen and oxygen plasmas, the MO optical band gap values can be reversibly tuned with the tunable ranges as wide as 80 and 550 meV for ZnO and CuO, respectively. The infrared emissivity of MOs is also found to be tunable with plasma treatment. The mechanism for reversible tuning of optical property is proposed based on the results of optical emission, photoluminescence, X-ray diffraction, and scanning electron microscopy characterization. Compared to conventional metal ion doping and high temperature annealing methods, the use of low-temperature hydrogen and oxygen plasmas is more environmentally friendly.

The WO3 is not ideal material for unassisted photoelectrochemical hydrogen production because the large band gap. In order to improve the visible range absorption of solar spectrum two possible ways are considered; to dope the material or insert oxygen vacancies which cause the shift in the absorption coefficient to the lower energy range. Different preparation methods have different advantages in film quality and production cost. In this work, the tungsten oxide films were prepared by rf magnetron sputtering from an oxide target at room temperature. The density of the native oxygen defects were modulated by controlling total pressure in the sputtering plasma and the band gap was tailored between 2.4 to 3.2 eV. The surface morphology evolution was investigated by atomic force microscopy (AFM). The XRD spectra of WO3 films deposited at different Ar pressure show that all films are amorphous irrespective of the total pressure in the chamber. The effect of working pressure on electrical resistivity was investigated by means of four probes Van der Pauw method. The temperature dependence of electrical conductivity was measured in the 300 - 625K range. The resistivity is very high at room temperature but decrease quickly as it rises, showing semiconductor behavior. We found that the resistivity increases with the increasing of sputtering pressure. Photoluminescence spectra obtained from samples deposited with various Ar pressure show a typical luminescence behavior with two emissions, narrow UV and broad green-yellow band. The energy position of these bands depends on Ar pressure in the sputtering chamber. We propose that this effect can be due to changes in the transport of the sputtered species in the background argon gas influencing the number of oxygen vacancies. We discuss the correlation between the PL results and resistivity measurements.

Delafossite CuAlO2 (CAO) is known as a p-type transparent conducting oxide (TCO). It has a wide range of applications, including optoelectronic devices of solar cells, flat panel displays, and others [1], including thermoelectric (TE) materials [2]. There are several techniques have been used for synthesizing the oxide, such as hydrothermal method, sol-gel, solid state reaction at high temperature, DC sputtering and microwave heating [3]. Clearly, the direct microwave heating of starting raw materials for the synthesis of CAO enabled us to save energy consumption, comparing to previous reports in literature. In order to investigate the optical and TE properties of CAO, a single phase of CAO powder was produced by direct reaction of mixture of aluminum nitrate nanohydrate (Al(NO3)3.9H2O) and copper acetate (Cu(CH3COO)2) with 1:1 molar ratio of Cu:Al, by a 600 W microwave radiation for 20 min. Finally, solid product was produced. It was composed of quite distorted plates with 200-350 nm thick. Each was a CAO single crystal. Its atomic vibration presented Al-O and Cu-O bonding, in accordance with the solid phase. The optical properties were used to determine its direct and indirect band gaps, 3.8 eV and 2.1 eV, respectively. The photoluminescence (PL) at room temperature was studied, and showed two peaks centered at 585 nm (2.10 eV) and 760 nm (1.63 eV), with the first peak corresponded to the indirect band gap. According to the study of TE properties, its Seebeck coefficient was positive value, indicating that holes were the majority of charge carries. By increasing of the test temperature, both the electrical resistivity and absolute value of Seebeck coefficient were decreased, but the power factor appeared in the opposite manner, increased until reaching a maximum value at 1073 K. [1] H. Kawazoe, M. Yasukawa, H. Hyodo, M. Kurita, H. Yagagi, and H. Hosono, Nature 389, 939-942 (1997). [2] K. Park, K.Y. Ko, H.-C. Kwon, and Nahm, J. Alloy Comp. 437, 1-6 (2007). [3] L. Torkian, and M.M. Amini, Matter. Lett. 63, 587-588 (2009).

Due to the spectral limitation of popular photocatalysts (PC) TiO₂ and ZnO, there is quest for modification of the existing PC to enhance their photocatalytic performance. It has been demonstrated that the coupled nanostructured semiconductors in form of nanocomposites (NC) enhance the performance by the mutual charge transfer of carriers between each semiconductor components with compatible chemical and electrical properties. ZnO (Bg:3.37 eV) n-type semiconductor, is most important PC because of its high photosensitivity and stability. CuO (Bg: 1.7 eV), a p-type semiconductor, can be used in conjunction with ZnO to further improve its photocatalytic activity (PCA). The enhanced PCA is anticipated from the improved junction potential between ZnO-CuO, which helps in efficient e^-/p⁺ pair charge separation upon excitation. ZnO-CuO NCs were synthesized via mechanical grinding. ZnO (~51.7 nm) and CuO (~20.4nm) obtained from US Research Nanomaterials Inc. were ball-milled (CuO Wt.% = 1, 5, 10, 15, 20, 30, 40, and 50) in planetary ball mill at 400 rpm with powder to ball weight ratio 1:10 in deionized water medium for 20 hrs. The ball milled mixture was later dried at 110^oC for 3 h in air. The Uv-vis MB spectra in the presence and absence of ZnO-CuO NCs were recorded as a function of Uv exposure time. The degradation rate constant of MB was determined from the first-order kinetic equation â?"ln(Co/C)=kt, where Co is the initial concentration of MB and C is the concentration of MB after UV exposure in the presence of NCs for time â?otâ?, and â?okâ? is the rate constant. The concentration of MB was estimated from the UV-vis peak of MB at 291 nm. ZnO-CuO (wt.% 10) exhibited superior PCA among others. This enhancement in PCA results from the improved junction potential, which increases the separation efficiency of photogenerated e^-/p⁺ pairs in ZnO. The PCA of ZnO-CuO (wt.% 10, k=0.35/min) was found to be superior to that of pure ZnO (k=0.11/min). With the increase in Wt.% CuO a decrease in PCA is observed. This decrease in PCA results from increased recombination centers with the increase of CuO concentration. Furthermore, ZnO-CuO (Wt. 10%) NCs were balled milled for 12, 24, 36, and 48 hours, to investigate the particle size dependence of PC efficiency of NCs. The average particle size of ZnO calculated using Scherrerâ?Ts equation was 52.7, 50.3, 49.8, and 48.5 nm, for 12, 24, 36 and 48 hrs, respectively. A continues increase in k with the decrease in particle size was observed and is attributed to the formation of many micro p-n junctions with increased surface area of NPs upon particle size reduction. The pH dependent study of MB degradation study show that pH of 7.1 corresponds to the highest MB removal rate for ZnO-CuO (Wt.10%). In conclusion, PCA of ZnO-CuO was higher than that of either pure ZnO or CuO NPs. The reasons for the improved PCA of ZnO-CuO NCs are tentatively attributed to an effective e^-/p⁺ separation by p-n junctions formed in the ZnO-CuO NCs.

The semiconductor-insulator-semiconductor (SIS) solar cells with relatively high conversion efficiency and low cost were formed by simply depositing transparent conduction oxides (TCOs) films on silicon (Si) wafer substrate. In this work, pulsed laser deposition (PLD) was adopted to deposit the indium zinc oxide (IZO) films which serve as the TCO layer of SIS solar cell samples. By varying the In contents and substrate temperatures, amorphous IZO films with the resistivity as low as 4.5Ã-10^-3 Ω-cm and transmittance higher than 80% in the visible-light wavelength range was achieved. Such an IZO layer was then adopted on the n-type Si wafer substrate to complete the preparation of SIS devices with IZO/SiO_x/n-Si structure. Under the AM1.5 illumination condition, the solar cell sample exhibited the best performance with the open-circuit voltage(V_oc) = 0.39 V, the short-circuit current density (J_sc) = 46.2 mA/cm², the fill factor = 39.7% and the conversion efficiency = 7.43%. Current-voltage (I-V) measurement of the SIS solar cell under dark condition observed a rectifying behavior in which the current is dominated by the electrons tunneling from the Si substrate to the IZO due to its relatively high work function. The tunneling current was affected by the thickness of interfacial SiO_x layer and was a crucial factor relating to the device performance. Such a thermal oxide layer thickness is determined by substrate heating temperature during IZO deposition and an about 2.03-nm thick SiO_x layer was observed in the solar cell sample with the highest conversion efficiency by transmission electron microscopy. Moreover, the device performance improvement was found to correlate to the increase of In content of IZO layer. When the In:Zn ratio was increased from 1:3 to 1:1 (in at.%), the V_oc increased from 0.29 to 0.39 V due to the bandgap enlargement of IZO layer by the Burstein-Moss shift effect. In the meantime, the low resistivity feature of IZO film caused the increase of J_sc from to 30.9 to 46.2 mA/cm² and effectively reduced the series resistance of devices. Consequently, the amorphous IZO/SiO_x/n-Si SIS solar cell with satisfied performance could be achieved.

The feasibility of vanadium substitution into Li2FeSiO4 to achieve beyond one electron reaction to enhance capacity, and the effects of such substitution on the structural and electrochemical properties of Li2FeSiO4, have been investigated by first-principles calculation based on density functional theory (DFT) within the generalized gradient approximation adding Hubbard-like correlation (GGA+U). The evolution of the local structures and electronic structures for Li2Fe0.5V0.5SiO4 upon delithiation are analyzed in detail. It is demonstrated that vanadium substitution into Li2FeSiO4 may be thermodynamically possible to achieve more than one lithium ion extractions and, therefore, to significantly enhance the capacity of the Li2FeSiO4 cathode material, and improve electronic conductivity with lower band gap to Li2FeSiO4. Smaller cell parameter changes and more minor local structural distortions with vanadium substitution might be beneficial to stabilize the delithiated phases.

MoS2 is a very promising material for photocatalysis and it has many current applications in catalytic hydrodesulfurization. Similar to graphene, it is a layered material. Recently, it has been shown that it transitions from an 1.6 eV inidirect bandgap to a 1.9 eV direct bandgap semiconductor when reduced to a monolayer. Important for its usefulness e.g. in catalytic hydrogen splitting, is not only its bandgap but also its band alignment when deposited on different substrates. Using CVD grown MoS2 on a copper surface, we use XPS to ascertain the identity of the material and the nature of its internal bonding when on this metallic substrate. Spectroscopy also shows a metal induced reduction of the bandgap to 1.5 eV and a strong signature of n-type doping through the underlayer. Density Functional Theorey calculation corroborate this finding and provide a microscopic understanding of the bandgap and â?"alignment depending on the number of MoS2 layers and the presence of any substrate.

The conversion of solar energy to produce hydrogen by means of water splitting is a promising technique to achieve a clean and renewable source of energy for the world. In principle, this technique is economical and environmental friendly, however, current photocatalytic systems suffer from inefficient energy conversion because the oxidation and reduction reaction pathways are not yet optimized. Engineered interfaces may offer materials designers extra degrees of freedom to tune and optimize systems for photocatalytic water splitting. In this contribution, we will explore using density functional theory whether the surface chemistry and band gap of TiO2 anatase layer can be tuned by an appropriate underlayer. Specifically, we investigate the TiO2/Cu2O (anatase/cuprous oxide) heterostructure. Anatase and cuprous oxide are both very strong candidates for photocatalysis and photovoltaics. In particular, anatase has the most efficient photoactivity, a high stability, and is very suitable for industrial applications. Here we study the anatase and cuprous oxide interface and explore (1) whether chemical effects at the interface can result in the tuning of the band gap of the photocatalyst for different interface structures, and (2) how deeply the interface effects extend into the surrounding materials. From our computed band structures, densities of states, and surface energies, we study interfacial effects on the valence and conduction band energy levels of both TiO2/Cu2O. In the context of photocatalysis, this effort demonstrates that surface junctions may be useful to optimize oxidation and reduction pathways. Hence, our computational study can guide experimental efforts to understand how interface effects influence surface potentials of metal oxides.

Mixed oxide thin-film transistors (TFTs) have been extensively researched and have attracted a lot of industrial attention for large area electronics due to their high mobility and improved stability under electrical bias stress compared to amorphous silicon TFTs. However, in addition to their electrical stability, it is very important to study their mechanical, thermo-mechanical and electro-mechanical stability for advanced flexible applications, which have been reported in our study. The indium-zinc oxide TFTs under test are fabricated at low temperature on thin flexible heat stabilized polyethylene naphthalate (PEN). To study their mechanical stability, the TFTs are deformed on cylindrical surfaces for both tensile and compressive stresses by bending samples both parallel and perpendicular to channel layer, respectively and varying radii for different time spans. The thermo-mechanical stability tests are conducted by bending them on cylindrical surface in an aggressive environment of 85 °C temperature and 85 % relative humidity. For the stress parallel to channel layer, both mobility and sub-threshold swing improve with tensile stress and opposite effects are observed with compressive stress. However, within experimental error no significant changes are seen for stress perpendicular to the channel. The TFTs are found to be extremely stable for all three sets of experiments and none of the changes lead to their complete failure.

ZnO, CdO, and ZnS are group IIâ?"VI direct bandgap semiconductors with wide gaps around 3.3 eV, 2.4 eV, and 3.6 eV, respectively. Mixing ZnO with ZnS or CdO produces ternary materials ZnOxS1-x and CdxZn1-xO that are wide bandgap semiconductors with tunable electronic and optical properties, making them of interest for instance as buffer layers in thin film photovoltaics. Control over the ternary material atomic composition is necessary for the engineering of the material characteristics for applications. This can be achieved by atomic layer deposition (ALD), which is a material deposition method enabling atomic level-controlled growth, by appropriately alternating ZnO, ZnS, and CdO ALD cycles to produce the ternary ZnOxS1-x and CdxZn1-xO films. By focusing on the investigation of ZnOxS1-x film growth characteristics and material properties, ALD of ZnS, ZnO, and ZnOxS1-x films from dimethylzinc (DMZn), H2O, and H2S was performed by systematically varying ZnO/(ZnO+ZnS) ALD cycle ratios from 0 (ZnS ALD) to 1 (ZnO ALD). The deposited films were characterized by means of x-ray diffraction, x-ray photoelectron spectroscopy, and spectroscopic ellipsometry, and the electronic and optical characteristics of CdxZn1-xO films deposited from dimethylcadmium (DMCd), diethylzinc (DEZn), and H2O were determined for comparison. The ZnOxS1-x ALD growth and crystal structure resemble those of ZnS up to a 0.6 cycle ratio, at which point XPS indicates about 10% oxygen is incorporated into the film. For higher cycle ratios the film structure becomes amorphous, which is confirmed by XRD. This is also reflected in the optical constants as determined by spectroscopic ellipsometry and, in particular, the optical bandgap transforms from direct type for the (cubic) ZnS-like phase to a narrower bandgap with amorphous characteristics, causing bandgap bowing. A direct bandgap is recovered at yet higher ZnO/(ZnO+ZnS) cycle ratios, where properties converge toward ZnO ALD in terms of film growth rate, crystallinity, and composition. Similar bandgap behavior as a function of film composition was observed for the deposited CdxZn1-xO alloys due to structural disorder originating from phase transition, and the determined optical constants of the CdxZn1-xO films showed trends reflecting changes in material composition and crystal structure. Overall, a strong effect of the film atomic composition on the material, electronic, and optical properties is confirmed for the ZnOxS1-x and CdxZn1-xO thin films.

Low cost, oxides are appealing for solar power conversion in photovoltaics and photoelectrochemical cells because of material stability but are limited by wide band gaps. We have shown recently that alloying of mismatched II-VI compounds results in a large reduction of the band gap. Here we report on a method of tuning the band gap of ZnO through the substitution of a small fraction of anion atoms with an isovalent element, which leads to a dramatic restructuring of the valence band due to an anticrossing interaction between the localized states of the substitutional ions and the extended valence band states of the host semiconductor. Specifically we discuss the structural and optical properties of ZnO_1-xSe_x alloys synthesized by pulsed laser deposition and the challenges of synthesizing such materials. A dramatic red shift of the absorption edge is observed for small concentrations of incorporated Se. Through use of an electrochemical junction, we show that we can extract carriers using photons with energies significantly lower (to E<2 eV) than the ZnO band gap of 3.27 eV. In order to explain the spectral dependence of the absorption coefficients, we developed a model that considers interactions between localized states of the Se atoms and the valence band ZnO host matrix. Fitting the model to the experimental data allowed determination of the energy level of Se as 0.9 eV above the valence band and the VBAC coupling constant as 1.2 eV. Combining these results with previously determined parameters we were able to predict composition variation of the band gap as well as the valence and conduction band offsets in the whole composition range of ZnO_1-xSe_x. We will also discuss predictions of the band gap narrowing in other group II-VI highly mismatched alloys.

LaCrO3 (LCO) is a wide-gap antiferromagnetic insulator for which the electronic structure is in question. Optical spectroscopy studies for the family of LaMO3 perovskites (M = Sc to Cu) suggest a change over from Mott to charge transfer behavior at M = Cr. XPS Cr 2p core-level line shape analysis and UPS spectra for polycrystalline LCO suggest that interband excitations across the gap involve d states, which points to Mott-like behavior. Yet, ours and others electronic structure calculations reveal that the valence band spectrum measured by XPS is more accurately accounted for if U = 0, suggesting that LCO is not a strongly correlated material. Optical absorption measurements for polycrystalline LCO suggests that the charge transfer gap is ~3.3 eV. Moreover, polycrystal based studies reveal that Sr substitution for La in LCO redshifts the band gap. Our goal in this work is to investigate the role of epitaxial strain as a prelude to studying the effect of Sr doping on the LCO band gap and photoconductivity. To this end, we have grown LCO films by MBE on (001)-oriented LaAlO3, (LaAlO3)0.3(Sr2AlTaO6)0.07, MgO and GdScO3, along with quartz. In all cases, the band gap of the substrate exceeds that of LCO, allowing us to determine the latter without interference from the former. All films except those on quartz were epitaxial; films on quartz were polycrystalline and fully relaxed. This choice of substrates allowed us to systematically vary the in-plane strain from -1.2% to +1.0%. Optical absorption data reveal that unstrained LCO exhibits a gap of 3.73 eV, and that compressive and tensile strain both result in increases in the band gap of up to ~0.2 eV. Hybrid functional calculations carried out to date reveal that the band gap increase in the compressive direction can be quantitatively accounted for. The tensile strain calculations are in progress.

I will discuss opportunities and some representative results on the problem of utilizing correlated oxides that undergo Mott transition or sharp phase transitions in emerging energy efficiency, energy conversion and storage technologies. Proximity to phase transitions involving multiple functional property changes enables design of low-dimensional Mott materials as adaptive layers for active climate control. The ability to trigger rapid phase transitions in low mobility oxides creates opportunity to fabricate low latency switches for electronics. Phase coexistence in the proximity to the transition is a possible route for enhanced energy storage through divergence in dielectric constant. It is hence crucial to understand the fundamental materials science of such complex materials: how does synthesis influence the electronic-structural interactions and the resulting dynamics? The role of compositional control in enabling superior functional properties will be considered in detail. I will also discuss properties of such materials in membrane form and their stability in freestanding quasi-two dimensional structures. Examples of high temperature instrumentation that have been designed to probe such properties in-situ will be discussed briefly.

Sizable tuning of band gap in transition metal oxides without losing the functionality arising from the transition metal has long been a challenge, due mainly to their rigid electronic structure. In this presentation, we show that the band gap of a ferroelectric oxide could largely be modified by site-specific substitution with a Mott insulator. For the ferroelectric, instead of a simple perovskite, we employed the bismuth-layered perovskite Bi₄Ti₃O₁₂, which was alloyed with the LaCoO₃. La and Co were introduced into the Bi₄Ti₃O₁₂ matrix to substitute for Bi and Ti, respectively, preferentially in the vicinity of the Bi₂O₂ sub-layer. The details on the preferential substitution have been studied with high-resolution STEM-EELS and density functional calculations. This resulted in a dramatically decreased gap (as much as 1 eV) indicated by a spectroscopic ellipsometry study. Interestingly, the Bi₂Ti₃O₁₀ perovskite-like sub-layers were not as much disturbed, maintaining the strong ferroelectricity. From the density functional calculation, the lowered band gap was found to be due to a split-off state from the conduction band of Bi₄Ti₃O₁₂ triggered by the selective Co²⁺ substitution for Ti. As a result of the decreased band gap, an enhanced photoelectronic response was observed, confirming that the site-specific substitution technique could be a novel route to controlling the band gap in complex oxides, desired for emerging oxide opto-electronics and renewable energy applications. The work was supported by the U.S. Department of Energy, Basic Energy Sciences, Materials Sciences and Engineering Division and the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory.

To date, no semiconducting material has been discovered that simultaneously meets all the criteria required for economical hydrogen production via light-driven direct water splitting using a semiconductor photo-electrode. Considerable work has been directed at metal oxides due to their expected stability and low costs, unfortunately after 35 years of work little progress has been made, efficiencies for these oxides remains very low. For a viable material, semiconductors for photoelectrochemical water splitting must have the same fundamental internal quantum efficiency as the commercial high efficiency PV devices. Multi-component transition metal oxides are complex materials, making intuitive guesses impossible and a focused search very challenging. A computational approach may be the only approach that can give us the necessary insight into these mixed metal oxides and allow us to narrow the composition space leading us towards a successful material. Previously, we explored a series of ternary cobalt spinels oxides, where group 13 cations (Al, Ga, In) occupy the octahedral spinel sites. We demonstrated that while good optical absorption can be obtained, the transport properties are very poor owing to the localized Co 3d states, and hence the observed photoelectrochemical activity is minimal. While this rules out this class of oxide materials (combining 3d^n and ns^0 cations), we now suggest a number of alternative multiternary oxides, which may overcome these intrinsic limitations and provide candidate photoelectrode materials in the near-term.

The use of semiconductors to split water into H₂ fuel and O₂ gas is a promising technology for renewably producing chemical fuels that can be stored and transported. While hundreds of candidate materials have been found that can utilize ultraviolet photons to split water, there are only a handful of stable semiconductors with visible light activity. Many of these materials are oxynitrides whose synthesis and chemistry is more complex than that of traditional oxides. The perovskite oxynitride LaTiO₂N is known to be a medium band gap semiconductor (Eg ~ 2.2 eV) whose energy levels are suitably positioned to allow overall water splitting in the absence of an external potential (based on the evaluation of half-reactions carried out in sacrificial reagents), but for which direct water splitting has not yet been observed. New synthetic procedures have been developed that can produce LaTiO₂N products with substantially different band gaps (2.1 â?" 2.25 eV) and carrier concentrations (product colors are brown, orange, or red), and new insights into the average and local structures of these phases will be presented based on the results of high resolution synchrotron x-ray and neutron powder diffraction studies as well as more local PDF, XANES and scanning TEM experiments. Epitaxial thin film samples of LaTiO₂N have been prepared on top of both insulating and conductive custom single crystal substrates that we have grown, and these films have been evaluated for their optical properties and photoelectrochemical activity. Quantum efficiencies for water oxidation up to 20% in ultraviolet light and 5% with visible light have been observed in the presence of a strong external bias. The measured properties will also be discussed in comparison to results from DFT-based calculations, with the role of anion disorder included through a cluster expansion approach and the additional role thermal fluctuations considered through molecular dynamics simulations on representative structures.

In typical semiconductors, the transport of electrons and holes occurs in the conduction and valence bands, respectively. In many oxides, however, such carriers tend to self-trap, leading to a small-polaron transport mechanism, which is generally associated by very low carrier mobility. In solar energy conversion applications like solar cells or water splitting, self-trapping is particularly detrimental, because it hinders harvesting the energy of photoexcited electron-hole pairs. Thus, band-structure (band gap, optical absorption) and transport properties should be engineered simultaneously in the materials design of oxides for energy application, requiring the ability to discriminate theoretically the two (band- vs. small-polaron) conductivity types. While standard density functionals are notoriously inaccurate in the description of localized polaronic states, an accurate prediction can be achieved by employing a generalized Koopmans condition [1] that recovers the physical condition of the energy-linearity as a function of the (fractional) occupation number.

Self-trapping can be viewed as the exothermic reduction (oxidation) of individual host atoms following the capture of electrons (holes) from the conduction (valence) band. While the subsequent change of the formal oxidation number often has a small effect on the local atomic charge [2], it is associated with distinct changes of the local structure and magnetic moment. We calculated a range of different polaronic systems [3-6], for both electrons and holes and for both free and defect bound small polarons, including doped ZnO (Li_Zn, N_O, V_Zn defects), large band gap Zn_1-xMg_xO alloys, TiO₂ (rutile and anatase), Rh₂ZnO₄ and MnO. The analysis of these different cases leads to distinct trends for the materials disposition towards self-trapping depending on structure and electronic configuration, and provides insights about promising routes for band-gap engineering that avoids carrier self-trapping.

Work in collaboration with H. Raebiger, J.A. Chan, T.R. Paudel, H. Peng, and A. Zunger. Supported by the US Department of Energy, Office of Basic Energy Sciences as part of an Energy Frontier Research Center.

[1] S. Lany, Phys. Stat. Sol. (b) 248, 1052 (2011).
[2] H. Raebiger, S. Lany, A. Zunger, Nature 453, 763 (2008).
[3] S. Lany, A. Zunger, Phys. Rev. B 80, 085202 (2009).
[4] J.A. Chan, S. Lany, A. Zunger, Phys. Rev. Lett. 103, 016404 (2009).
[5] S. Lany, A. Zunger, Phys. Rev. B 81, 205209 (2010).
[6] A.R. Nagaraja, N.H. Perry, T.O. Mason, Y. Tang, M. Grayson, T.R. Paudel, S. Lany, A. Zunger, J. Am. Ceram. Soc., DOI: 10.1111/j.1551-2916.2011.04771.x

To understand the performance of future photoelectrochemical solar fuel systems, it is necessary to develop a high fidelity, predictive simulation tool that integrates light absorption, carrier transport, catalysis, ion diffusion, and fuel transport on the meso-scale. We report here on the light absorption and carrier transport characteristics of a core-shell WO₃/Si Z-scheme photoelectrochemical device. Crystalline silicon (Si), with a bandgap of 1.1eV, and tungsten oxide (WO₃), with an approximate bandgap of 2.6eV, have a combined photovoltage and an appropriate band alignment to make the photoelectrolysis of water thermodynamically possible. Existing knowledge on the synthesis of these two materials suggests that fabrication can be straightforwardly done. Growth of high aspect ratio silicon microwires with high crystallinity has already been developed.¹ A dual junction Z-scheme can be realized by using the wires as a scaffold for the deposition of the tungsten oxide. Light absorption and carrier transport were investigated by simulation using full wave electromagnetic simulations (Lumerical) and device simulations (Sentaurus) to optimize device design for maximum efficiency. We investigated silicon wires with lengths of 100µm, and diameters of 2µm, with a 7µm pitch, covered vertically with 50µm of tungsten oxide with a thickness of 1µm. For conditions of illumination with the AM 1.5 spectra and an IQE of 1, the WO₃/Si Z-scheme device results in a short circuit current of 1mA/cm², limited by the tungsten oxide absorption. Observation of the spatial absorbed light profile reveals that the silicon is largely shadowed when an opaque Z-scheme contact material is used (e.g. aluminum), and that such a contact material also absorbs a significant portion of the light (up to 50% at a thickness of 100nm, completely covering the interface between the photoelectrodes), which severely limits the silicon absorption. The aluminum absorption can be reduced by a factor of 5 by limiting contact area to the tops of the wires. These findings demonstrate that the material selection and geometric design of the material used as an electrical contact between the two photoelectrodes is of the utmost importance for the optimization of silicon light absorption. Device simulation results will be used to guide experimental WO₃/Si tandem device development. The Si/WO3 device configuration is experimentally realizable and uses two photoelectrodes in tandem to achieve sufficient voltage to split water with practical efficiencies. This design also capitalizes on the use of high aspect ratio wires for the orthogonalization of light absorption and carrier transport to attain high efficiencies with indirect bandgap materials. It also significantly increases the device surface area for catalysis and decreases material usage by at least 90% in comparison to planar models. ¹Kayes, BM; Filler, MA; Putnam, MC; Kelzenberg, MD; Lewis, NS; Atwater, HA. Appl. Phys. Lett. 91, 10 (2007).

ZnO quantum dots (Q-dots) in the size range 4-10 nm have been synthesized by thermal hydrolysis in alkaline zinc acetate solution. Ensambles of Q-dots in the solid state were prepared in form of thin transparent films for spectro-electrochemical measurements. By a precise control of the electrochemical potential and thus the Fermi level at the back-contact, it was possible to form a steady-state condition of populated states in the conduction band of the quantum dots. This leads to a depletion of the optical absorption and a shift in the absorption onset as a function of the applied potential. This is similar to the Burstein-Moss shift observed in degenerate semiconductors. From the absorption change it is possible to extract the absolute position of the valence- and conduction band edges, which previously has been found difficult to obtain for nanoparticles with traditional electrochemical methods like Mott-Schottky measurements. With this method we have been able to follow the position of the band edges as a function of particle size. We have also noticed that even as the absorption decreases when a potential greater than the band edge position is applied the absorption is actually increasing for a lower applied potential, which is a phenomena we intend to explain.

Some of the most promising stable semiconductors discovered to date for visible-light-driven solar water splitting are oxynitride materials. LaTiO2N is an oxynitride compound whose potential for driving overall water splitting has not been realized despite its suitable band gap (2.2 eV) and its ability to separately drive the half reactions for H2 and O2 production in suitable sacrificial reagents. The limitations preventing the observation of overall water splitting are not yet known. Despite belonging to the common perovskite structure family, the crystallography of this compound is quite complex due to its very low symmetry (triclinic, P-1) and the likelihood of disorder among the three anion sites in this space group. A computational methodology with several steps was therefore necessary to achieve an understanding of the role of disorder in this compound, and its influence on the optical properties. The structure of LaTiO2N was investigated by developing a first-principles cluster expansion model and then performing Monte Carlo (MC) simulations on the model. Although an anion-ordered structure is predicted at 0K, it is found that the anion sites will be disordered at typical experimental synthesis temperatures (600 â?" 1000 °C). The amount of partial order predicted by theory is in very good agreement with the values refined from powder neutron diffraction studies. To form a more detailed picture of the bonding in this material and to analyze electronic and optical properties, snapshots from a MC simulation on a modest cell (16 formula units) were further refined by direct DFT calculations, followed by molecular dynamics (MD) simulations performed over several picoseconds to capture thermal fluctuations. The theoretical model of this anion-disordered structure was determined to be truly representative of the actual material by comparing the experimentally measured neutron pair distribution functions (PDFs) on powder samples to the time-averaged PDFs obtained from MD trajectories. Agreement is excellent, comparable to fits achieved through experimental structural refinements. Turning to electronic and optical properties, many-body perturbation theory (GW/BSE) has been applied to ordered structures to assess the errors inherent to the use of DFT for calculating band gaps and other excited state energies. In the ordered low temperature phase, a strong onset for optical absorption is found to occur only 0.2 eV above the minimum direct gap. In the anion-disordered phase, the calculated band gap is about 0.2 eV larger and the onset of optical absorption edge is more gradual. Thermal fluctuations were found to have a modest impact on the optical threshold. Calculated optical responses will be discussed in the context of measured experimental data on powder and thin film samples.

Lowering the operating temperature of solid oxide fuel cells to 500-700°C could allow the use of a larger variety of cheaper materials for interconnects and insulation in SOFCs than the high-temperature formulations required today. This would lower the cost of SOFCs and increase their adoption. However, it also requires cathode and electrolyte oxides with fast oxygen conduction at low temperatures. Searching for such new oxides experimentally is time-consuming and expensive. Here we use automated high-throughput ab initio methods to computationally search for fast oxygen conductors and useful descriptors for O migration energetics. Since perovskites are a class of known fast oxygen conductors, we explore migration energetics and many possible descriptors within this important family of materials. We then compare the trends and lack of trends that we have found to existing experimental data and theoretical assumptions about perovskite diffusion to assess both the validity of our approach and that of prevalent assumptions about perovskite ionic transport.

Solar energy harvesting by photocatalytic water splitting requires a charge collection material that can efficiently absorb visible light and transport the generated charge to a liquid interface. While efficient light absorption is the most obvious parameter for optimizing photocatalyst performance, the transport of photogenerated charge to the catalyst surface is equally important. We therefore study the electronic structure and catalytic performance of epitaxial high-crystallinity Rh-doped SrTiO3 thin films. Rh:SrTiO3 is known to be an efficient catalyst for producing hydrogen in a water splitting reaction.[1] Our purpose is to use thin film samples to determine the valence state of the Rh impurity in the SrTiO3 host and the effects of the impurity valence state on the catalytic efficiency. We use x-ray photoelectron spectroscopy (XPS) for determining the Rh valence as a function of crystal growth parameters and show that stable Rh3+ and Rh4+ states can be achieved by a suitable selection of oxygen pressure during crystal growth. X-ray absorption and emission spectroscopies (XAS and XES) are used to determine the locations of empty and filled in-gap impurity levels. A mid-gap level related to the Rh4+ state and a Rh3+ level close to the valence band top have been identified. Cyclic voltammetry measurements of thin Rh:SrTiO3 films grown on Sr2RuO4 electrode layers show that Rh:SrTiO3 works as a typical p-type semiconductor, with a characteristic cathodic response to visible light illumination. The electrochemical responses of the thin film catalysts are an order of magnitude larger than those of similar Rh:SrTiO3 powder catalysts.[2] We attribute the improved efficiency on the higher crystallinity of the thin-film samples. [1] A. Kudo and Y. Miseki, Chem. Soc. Rev. 38, 253 (2009). [2] K. Iwashina and A. Kudo, J. Am. Chem. Soc. 133, 13272 (2011).

The mechanism of Ni substitution into the oxide semiconductor InTaO₄ has been studied through a combination of structural and spectroscopic techniques, providing insights into its previously reported photoactivity. Magnetic susceptibility and x-ray absorption near-edge spectroscopy (XANES) measurements demonstrate that nickel is divalent within the host lattice. The combined refinement of synchrotron and neutron powder x-ray diffraction data indicates that the product of Ni-doping has the stoichiometry of (In _1-xNi _2x/3 Ta_x/3 )TaO₄ with a solubility limit of x ~ 0.18, corresponding to 12% Ni on the In site. Single phase samples were only obtained at synthesis temperatures of 1150 °C or higher due to the sluggish reaction mechanism that is hypothesized to result from small free energy differences between (In _1-xNi _2x/3 Ta_x/3 )TaO₄ compounds with different x values. Undoped InTaO₄ is shown to have an indirect band gap of 3.96 eV, with direct optical transitions becoming allowed at photon energies in excess of 5.1 eV. Very small band gap reductions (less than 0.2 eV) result from Ni-doping, and the origin of the yellow color of (In _1-xNi _2x/3 Ta_x/3 )TaO₄ compounds instead results from a weak ³A_2gâ?'³T_1g internal dâ?'d transition not associated with the conduction or valence band that is common to oxide compounds with Ni²⁺ in an octahedral environment.

There has been an interest in exploring III-II spinels (A2BO4) for their electronic and optoelectronic applications. Spinels have a complex structure where two cations A and B occupy two inequivalent sublattice of oxygen, namely, the octahedral (Oh) sites and the tetrahedral sites (Td). Cross-substitution in a partially inverse spinel could induce â?oanti-siteâ? doping and generate charge carriers. A recent theoretical study divides spinels into four Doping Types (DT) based on the relative energy levels of anti-site defects in the band structure. This offers a possibility of rational search for spinels for targeted electrical properties, especially the carrier concentration and conductivity. Here, we synthesize a few prototype materials predicted to be of different types, and test the DT theory by relating their electrical properties to the cation site occupancies. The materials are made via different routes by solid state reaction and by pulsed laser deposition, to have a portfolio of both equilibrium and non-equilibrium samples. Next, we employ diffraction techniques including anomalous x-ray diffraction and neutron diffraction to determine Td and Oh site occupancies by the host cations and dopants. Comparison of site occupancies to experimentally determined conductivity and carrier density shows that, 1) Ga2ZnO4 and Ga2MnO4 are DT-1 as predicted by the theory. The materials remain insulating even with significant cross-substitution. Fermi level pinning makes it difficult to create large number of uncompensated donors or acceptors. 2) Co2ZnO4 and Co2NiO4 are typical DT-2 spinels. Their conductivity originates from the divalent cation (Zn2+ or Ni2+) occupying the Oh sites whereas Co3+ on the Td sites remains electrically inactive. 3) Cr2MnO4 belongs to the DT-4 family. Undoped samples are largely insulating because neither anti-defect is electrically active. External doping introduces substitutional defects into Td or Oh sites and enhances the conductivity. We have validated the Doping Type theory in spinels by experiments. This will allow theorists to screen a large pool of spinel oxides and help experimentalists to identify candidates with targeted property such as high carrier density and p-type conduction.

Topologic boundaries in complex oxides can exhibit highly anisotropic electronic responses that are different from the bulk material due to built in changes in structure and symmetry that are connected to the order parameters. We present recent results on anisotropic electronic conductance and local band structure investigations of isosymmetric phase boundaries in mixed phase Ca-doped BiFeO3 thin films. The origin and nature of the observed conductivity is probed using a combination of temperature dependent scanning probe microscopy, high resolution transmission electron microscopy, electron energy loss spectroscopy and X-ray diffraction.

Transparent conducting oxides (TCO) are particularly attractive thanks to their high optical transparency in the visible range and their good electrical conductivity. Their use is relevant for a wide ranging of applications including for instance optoelectronic devices, solar cells or sensors. By considering the rapid growth of the global photovoltaic market associated with the need of TCO for almost all thin solar cells, this will clearly emphasize the need for improving the TCO overall properties while using abundant, non-toxic and non-expensive materials. Our work focuses on a better understanding of the electro-optical properties of FTO, especially on their relationship with the structural properties. The optimization of the experimental growth conditions results in the fabrication of FTO thin films at the state of the art. We have thoroughly investigated in details the influence of the experimental growth conditions of the spray pyrolysis deposition on glass (such as substrate temperature, deposition rate, film thickness) on the physical properties of FTO thin films. It is clearly shown for instance by x-ray diffraction and Hall effect measurements that the electron mobility is directly correlated with the x-ray peak width. The presence of twin boundaries is revealed by TEM imaging and plays an important role in the electrical properties of FTO thin films. The existence and role of these twins will be described and discussed in details.

Electrochromic materials such as WOx and NiOx have been investigated intensively for enabling energy efficient technologies. In a typical electrochromic device configuration, a cathodic electrode (e.g., WOx) and an anodic electrode (e.g., NiOx) are sandwiched into a layered structure with a solid Li ion electrolyte in-between. Currently, the NiOx counter electrode suffers from relatively poor performance limiting potential large-area window applications. The overall goal of our study is to improve the performance of the NiOx-based counter electrode, via the addition of metals and/or dopants during the NiOx sputter deposition process. Specifically, Zr and Li additives were utilized to improve the transparency of the bleached state between 400 and 500 nm and optimize other electrochromic performance metrics, such as optical modulation and switching speed. In this study, we will present progress in the characterization of the multicomponent NiOx-based thin films by X-ray diffraction (XRD), transmission electron microscopy (TEM), Raman spectroscopy, UV-Vis-NIR spectroscopy, X-ray photoelectron spectroscopy (XPS) and in-situ spectroelectrochemical measurements. In addition, we will demonstrate significant enhancements in the electrochromic performance of NiOx:Li:Zr relative to the more commonly investigated NiOx:Li:W materials.

Titania (TiO₂) is an anorganic compound which features a wide variety of applications, for instance, for photovoltaics and photocatalysis. Its fundamental band gap amounts to about 3 eV, as confirmed by experimental as well as theoretical studies. However, the optical-absorption onset, marked by a pronounced peak in the spectrum, occurs only at approximately 4 eV. This has been drawn back to the vanishing of the matrix elements for optical transitions between the highest valence and the lowest conduction-band states. These transitions are dipole forbidden due to the rutile crystal structure of TiO₂. The resulting lack of absorption across the entire visible spectral range somewhat hampers photovoltaic or photocatalytic applications. We employ very recent theoretical spectroscopy methods based on quasiparticle band-structure calculations and the solution of the Bethe-Salpeter equation to investigate how alloying TiO₂ with TiS₂ affects the optical absorption in the visible spectral range. Our calculated results for the composition-dependent dielectric functions and absorption coefficients account for excitonic as well as local-field effects. From our ab-initio calculations we find that for small compositions x the band structure of TiO _2(1-x) S _2x still resembles the one of TiO₂, except for defect-induced levels within the fundamental band gap. However, the optical-transition matrix elements for the alloy indicate that the excitation of electrons from the bulk-like valence bands into the conduction bands becomes dipole-allowed, since the substitution of O atoms by S atoms breaks the rutile symmetry of the TiO₂ lattice. Hence, significant contributions to the optical absorption arise at around 3 eV for the TiO _2(1-x) S _2x system. In addition, also the defect levels within the gap create an absorption band at even lower photon energies. We show that the incorporation of S into TiO₂ leads to pronounced absorption of visible light for S concentrations as low as 1.5 %.

The concept of non-compensated n-p codoping [1] provides promising opportunities for tuning the electronic, optical, and magnetic properties of a variety of materials. In this talk, we apply this concept to several specific example systems in oxide semiconductors and carbon-based materials. First, we explore theoretically the optimal quantum efficiency of n-p codoped TiO2 as intermediate-band solar cells (IBSCs) under two different design schemes [2]. The first preserves the ideal condition that no electrical current be extracted from the intermediate band (IB), with the maximum quantum efficiency of 52.7%. In the second scheme, current is also extracted from the IB, resulting in a further enhancement in the maximum efficiency to 56.7%, while also relaxing the stringent requirement that the IB location be close to the optimum value. This finding makes it more feasible to realize IBSCs with high quantum efficiencies. As a second example, we study ferromagnetism in n-p codoped In2O3 and ZnO systems within first-principles density functional theory and experimentally [3]. It is found that n doping plays the elegant dual role of further enhancing the ferromagnetic stability of a local magnetic polaron, and more crucially, mediating the nonlocal magnetic coupling between two polarons. In the third example, we investigate the effects of n-p codoping on the absorption and magnetic properties 3d transition metal (TM) adatoms on graphene [4]. It is found that n-p codoping can enhance the magnetic moment and adsorption energy of TM-adsorbed graphene systems while preserving their unique electronic properties, thus pointing to ferromagnetic order in graphene. * Supported by NNSF of China, USNSF, and DMSE/BES of USDOE. [1] Zhu W. G., et al., Phys. Rev. Lett. 103, 226401 (2009). [2] Wu F. C., Lan H. P., Zhang Z. Y., and Cui P., submitted (2011). [3] Qi S. F., et al., Phys. Rev. B 84, 205204 (2011). [4] Qi S. F, Zhang Z. Y., and Xu X. H., to be submitted.

Bandgap narrowing of oxide semiconductors has been recognized as the main avenue for enhancing their performance in photo-electrochemical solar energy conversion. In a recent study, non-compensated n-p co-doping has been identified as an enabling concept to narrow the band gap of TiO2 with dramatically enhanced photoreactivity [W. Zhu et al., Phys. Rev. Lett. 103, 226401 (2009)]. Non-compensated n-p co-doping takes advantage of the strong tendency of the dopants with opposite charge states to bind together, which greatly enhances their thermodynamic and kinetic solubilities for dopant incorporation. Furthermore, non-compensated n-p codoping enables tuning of the impurity band location and chemical potential by choosing different combinations and concentrations of the n- and p-type dopants. In this talk, I will discuss the results of a series of x-ray spectroscopy studies carried out at synchrotron radiation facilities on N/Cr non-compensated co-doped TiO2 as a model system for addressing some of the key issues surrounding the physics of non-compensated co-doping, namely: 1) How the N and Cr dopants impact the electronic structure of TiO2, 2) The microscopic mechanisms associated with the band gap narrowing, 3) The degree of localization of the N/Cr induced states within the TiO2 band gap, and 4) The dynamics of the light induced electron transfer with chemical specificity. These results may be relevant towards a sound understanding of the microscopic mechanisms responsible for the band gap reduction and carriers mobility in TiO2, with the outcomes providing crucial feedback for the development of oxide semiconductors with dramatically enhanced photoreactivity.

The scanning tunneling microscopy (STM) provides a unique approach to understanding the physics and chemistry of different adsorbates on TiO₂ [1]. The knowledge of adsorption behaviors of molecules, like O₂, CO, CO₂, water, and methanol [2-6], has led us to a situation that we may directly investigate the chemical reactions, especially the photocatalytic reactions with atomic resolution [7]. Here, I report our recent progress on the characterization of the photocatalytic dissociation of methanol [7] and water [8], and the reaction between water and CO₂ under UV irradiation with the model system of the reduced rutile TiO2(110)-1x1 surface. We provide first direct evidence to demonstrate that adsorbed water and methanol molecules can be dissociated under UV irradiation using low temperature scanning tunneling microscopy. Such a dissociation process is quite different from the spontaneous dissociation of water [6] and methanol [9] or ethanol [10] at oxygen vacancies. It is also observed a possible reaction process between individual water and CO₂ molecules, different from the direct reduction of CO₂ by inelastic tunneling electrons [4]. The characterization of the chemical reactions with atomic resolution can be an important way for a better understanding of the underlying mechanisms of these reactions, and thus important for the design of better photocatalytic systems for efficient water splitting and reduction of CO₂.

TiO2 is a promising material for photocatalytic water splitting. In order to achieve efficient electron hole-pair generation under sunlight illumination and to adapt the redox potentials to those required for water splitting it is necessary to combine TiO2 with nanostructured semiconductors. In this respect a nanostructured TiO2/Si composite represent a promising material combination. In this work 10x(TiO2/Si) multilayer structures have been grown by sputtering with systematic variation of the TiO2 and Si layer thicknesses. For that a combinatorial materials synthesis approach with a computer controlled movable shutter system was applied. After rapid thermal annealing in inert gas and oxygen atmosphere the multilayers have been investigated by high resolution transmission electron microscopy, mirco-Raman and dynamic secondary ion mass spectrometry for their structure, stability and crystalline structure. It was found that the photocatalytically more active anatase TiO2 is stabilized up to 1100°C annealing temperature in these multilayer structures. This finding is particularly important since only at this high temperature simultaneous formation of embedded Si nanocrystalline layers can be achieved.

The performance of semiconductor photoelectrodes is often hampered by the mismatch between length scales for solar photon absorption (~ 1-100 microns) and charge transport (~ 10-100 nm). In this paper we will present a number of strategies we have employed to overcome this mismatch, including: (1) The use of metallic particles embedded at the surface or at the base of an iron oxide (Fe2O3) thin film in order to induce plasmonic resonances and multilayer interference effects that can be engineered to strongly concentrate sunlight close to the electrode/liquid interface, precisely where the relevant reactions take place. By comparison of spectral features in the enhanced photocurrent spectra to full-field electromagnetic simulations, the contribution of surface plasmon excitations is verified. (2) The development of hierarchically branched nanorod structures which serves as a model architecture for efficient photoelectrochemical devices as it simultaneously offers a large contact area with the electrolyte, excellent light-trapping characteristics, and a highly conductive pathway for charge carrier collection. Nanorod arrays of TiO2 were synthesized with nano-scaled branches that improve efficiency by means of improved charge separation and transport within the branches due to their small diameters as well increased surface area which facilitates the hole transfer at the TiO2/electrolyte interface. (3) The development of extremely high surface area transparent conducting oxide (TCO) films. We have synthesized TCO films (e.g. indium tin oxide, ITO) with roughness factors of approximately 1000, that maintain high optical transparency as well as a film conductivity similar to that of dense TCO films. With such high surface area, one can deposit an ultra-thin absorber layer (semiconductor) and achieve the overall absorption of a thick layer, but with the advantage of minimal charge-transport distance to either the semiconductor-liquid interface or that between the semiconductor and the back contact.

Anatase TiO₂ has attracted much attentions as an eco-friendly material because of its applications such as dye-sensitized solar cells, and its super hydrophilicity [1,2]. These intriguing functionalities are strongly influenced by the presence of crystal defects [3]. However the difficulty of defect management often hinders the establishment of a clear understanding of these fascinating properties for useful applications. Thus, in order to more fully control the appealing characteristics of anatase TiO₂, control of the crystalline quality are indispensable. A common strategy for defect management is to optimize the growth temperature and ambient pressure based on the defect thermodynamic equilibria. However, in the case of TiO₂, this strategy is complicated by the comparable cohesive energies of the anatase and rutile crystal structures. This narrows the window of thermodynamic variables available to obtain a high quality single anatase phase [4]. As an alternative, it is well known in the growth of compound semiconductors that the kinetic conditions, such as the adatom surface migration, adsorption and desorption dynamics strongly influence the formation of crystal defects [5]. Kinetic factors are easily controlled by tuning the growth rate, and it is applicable to many deposition techniques in general. Here, we applied this approach using pulsed laser deposition which has two primary parameters available to control the growth rate: the laser fluence and repetition rate. The defect density in the films was first characterized from the electrical resistivity. The carrier density measured at 300 K showed a minimum value around a growth rate of 5-7 Χ 10^-3 nm/sec independent of whether the growth rate was changed by the fluence or repetition rate. To identify the origin of these carriers, the activation energy ( E_a) was estimated from an Arrhenius fit to the temperature dependent electrical transport measurement. Estimated ( E_a) were similar order to the energy level of oxygen vacancy states calculated from first principles. As an independent identification of the donors, photoluminescence (PL) measurements were performed with various excitation powers. The critical excitation power ( P_crit) was measured, where the oxygen vacancy states are filled with carriers and the corresponding integrated intensity of PL spectra saturates. The P_crit showed a similar growth rate dependence to that of the carrier density. Hence, we can conclude that the formation of oxygen vacancies can be suppressed by optimizing the growth rate. [1] R. Wang et al., Nature 388, 431 (1997). [2] B. Oâ?TRegan et al., Nature 353, 737 (1991). [3] Y. Kim et al., J. Power Sources 175, 914 (2008). [4] M. Lazzeri et all., Phys. Rev. B 63, 155409 (2001). [5] S.N. Lee et al., J. Cryst. Growth 250, 256 (2003).

Transition metal oxides have received much interest lately for the possible use in both organic and inorganic electronic devices. However the deposition methods available to metal oxides are limited by their high melting point so methods such as sputtering have to be used. In this work reactive sputtering is used as a method to control the stoichiometry and hence the energy level structure of deposited molybdenum oxide films. By using a combination of inert and reactive gas (Argon and Oxygen) and varying the ratio of gas flow into the chamber it has been shown that it is possible to control the oxidation state of the deposited material. A combination of UPS, XPS, AFM and ellipsometry has been used to characterise films deposited via reactive sputtering using various processing parameters such as blend ratio, overall pressure and power of the plasma. Results indicate that as the concentration of oxygen within the chamber is increased the deposited material gradually shifts from a highly reflective metallic film to a transparent semiconductor. UPS and XPS are used to look at the energy levels in relation to the oxidation state of the film. Three different deposition regimes are observed; the first is for oxygen concentrations between 0 and 10%, in this region metallic layers are deposited that have occupied energy levels that extend to the Fermi level, this is due to the presence of +1 and +2 oxidation states. The second is for concentrations between 10 and 20%, in this region the +1 and +2 oxidation states are eliminated and the occupied states no longer extend to the Fermi level, instead a short bandgap of around 1-1.5eV is observed and a highly absorbing film is deposited. The third regime is for concentrations above 20%, in this region only +5 and +6 oxidation states are present and a large bandgap is observed of around 3-3.3eV resulting in a highly transparent semiconducting film.

The electrochemical stabilization of non-oxide semiconductor photoanodes has been a long standing goal for photoelectrochemical fuel production. In this talk we describe the investigation of manganese oxide thin films for chemical protection of silicon and gallium phosphide photoanodes during water oxidation. Manganese oxide films were deposited using atomic layer deposition with bis(ethylcyclopentandienyl)manganese and water at 150 °C. Junctions between manganese oxide and n-type silicon exhibited photovoltages of ~420 mV under simulated solar illumination and were electrochemically stable in 1 M KOH (aq) solutions. Photoactive Schottky-type barriers were formed by the manganese oxide films and not by a semiconductor/liquid junction. Thus, the photoelectrochemical behavior was independent of the electrochemical potential of the solution redox couple. Appreciable current densities (>10 mA cm-2) were observed with manganese oxide film thicknesses of >5 nm indicated that tunneling was not an active mechanism of electron transport. Reliance on tunneling transport limits metal oxide thicknesses to ~2 nm, which might also limit the stability and chemical impermeability offered by the metal oxide thin films. The bare manganese oxide films exhibit modest catalytic behavior for water oxidation and increases in current density were achieved by coupling Ir- and Ni-based electrocatalysts to the manganese oxide surfaces. Guidelines for protection strategies exploiting engineered metal oxide thin films are outlined.

The key challenge in using sunlight to drive important photoelectrochemical reactions is designing light absorbing materials and structures that simultaneously achieve a) efficient absorption of sunlight, b) efficient electron-hole separation and transport, and c) surfaces with high electrochemical reactivity, all while avoiding corrosion or oxidation of the absorbing material that would destroy one or more of the above properties. Corrosion and/or oxidation are especially significant for aqueous systems, which are most important in research that addresses large-scale energy and environmental problems. We have recently demonstrated that atomic layer deposition (ALD) can be used to protect the surface of silicon so that it can serve as a stable photoanode for water splitting. Water oxidation, which has long been recognized as a key step in fuel synthesis from sunlight, is a photoelectrochemical reaction that would normally cause Si and many other high-quality absorbers to oxidize destructively. Atomic layer deposition is a method for depositing nanoscopic and pinhole-free films, often over complex substrate topography, via a series of self-limiting chemical reactions. Well-established in research on nanoelectronic materials, it is only now being investigated for applications in energy science and technology, providing an opportunity for high-impact discoveries. This presentation will briefly review prior results on surface protection of Si photoanodes in water splitting, and show that ultrathin ALD-grown TiO2 provides a protective metal oxide passivation of nanometer thickness that stabilizes heretofore unstable semiconductor surfaces under water oxidation conditions while permitting facile electronic carrier transport by tunneling. A nanoscale Ir catalyst layer minimizes use of precious metal catalyst and avoids optical losses that occur when coating the semiconductor under a thick protective layer. In addition to functioning as an excellent oxidation catalyst, the large workfunction Ir layer acts as hole transfer mediator between the n-Si photoanode and water. This nanocomposite structure appears to avoid the Fermi level pinning typical of most metal semiconductor and electrolyte/semiconductor interfaces. The resulting combination of stability, large photovoltage, and high current densities we have observed at low overpotentials far exceeds the performance of previously reported Si photoanodes for water oxidation and may be applicable to other anode materials beyond silicon.

The facile growth of p-type transparent (or semitransparent) conducting oxides is a long standing challenge. Here, we specifically demonstrate how intentional non-equilibrium growth can be used as a design tool to qualitatively increase the hole concentration and electrical conductivity of Co₂ZnO₄ spinel, a known p-type oxide semiconductor. The resultant films have conductivity Ïf â?^ 10 S/cm and are effective hole transport layers for organic photovoltaics. More generally, we show how first principles theory can be used predictively to identify materials and classes of materials in which non-equilibrium growth is likely to yield improved material properties. Our first-principles theory shows that the antisite defects Co on the tetrahedral Zn-site (Co_Td) and Zn on octahedral Co-site (Zn_Oh) are the dominant intrinsic defects in Co₂ZnO₄. With nominal valences of 3+ for Co and 2+ for Zn in this III-II spinel, it is expected that Co³⁺_Td would be a donor state and Zn²⁺_Oh would be an acceptor site. However, the Co_Td site defect level is resonant in the valence band, making it electrically neutral and allowing the electrically active Zn_Oh acceptor to yield p-type conductivity independent of the concentration of Co_Td defects. Further, theory predicts that for equilibrium growth, Co₂ZnO₄ is p-type even though for equilibrium growth it is always Co rich with N[Co_Td] about 100 times larger than N[Zn_Oh]. This is confirmed by anomalous x-ray diffraction site occupancy measurements on bulk ceramic samples grown in air at 800 °C. Hall and Seebeck effect measurements confirm the predicted p-type conduction. Given this, non-equilibrium growth can increase the hole concentration and conductivity in two ways. First, since Co_Td is electrically neutral and Zn_Oh is an acceptor, any stoichiometrically neutral one-for-one Co-Zn site swaps on the cation sub-lattice will be net hole creators. Second, if excess Zn (relative to the Co₂ZnO₄ stoichiometry) can be incorporated through non-equilibrium growth, it will create Zn_Oh site defects with nearly unity efficiency as the Zn has no place else to go. To test this, we have used combinatorial co-sputtering from angled CoO and ZnO sources to grow Co_2-xZn_1+xO₄ thin film â?olibrariesâ? with intentional composition gradients on 2â?x2â? glass substrates. We find that while both proposed non-equilibrium growth doping mechanisms are active, the incorporation of excess Zn is by far the stronger effect. In particular, for Co_2-xZn_1+xO₄ films grown at T_S = 340 °C in pO₂ = 10^-5 atm, we find a monotonic increase in electrical conductivity from Ïf = 0.2 S/cm at x = -0.3 to Ïf = 5 S/cm at x = 0.3. Above, x = 0.3, the electrical conductivity decreases and ZnO is observed as an impurity phase in x-ray diffraction. Subsequent post-deposition annealing in air at temperatures ranging from 300 °C to 800 °C decreases the conductivity and shifts the onset composition of ZnO impurity phase formation to lower x (less Zn).

TiO₂ is one of the most widely studied oxide materials for energy applications. While there is a large body of work on producing ultra-thin smooth films for surface studies, little has been done to study the effect of non-equilibrium synthesis on the phase and morphology of TiO₂. Here we will discuss the growth of supported films of nanostructured TiO₂ with tunable porosity using non-equilibrium pulsed laser deposition and will explore the effect of surface morphology on photocatalytic performance. We examine the evolution of the phase, structure, and morphology of titania thin films as a function of the growth temperature (600-900°C), laser fluence (0.48-2.20 J/cm²), deposition rate (0.07-2.0 Ã./s), and epitaxial film-substrate lattice mismatch (from -1.89% to 4.18%). We have dramatically expanded the range of growth parameters applied to TiO₂ and have accessed non-equilibrium growth conditions that give rise to complex structures. Using an array of techniques, we have observed an epitaxial-strain driven stabilization of anatase over rutile at growth temperatures well above the bulk structural phase transition. We have probed the nature and stability of these metastable films in an attempt to probe the energetics of the epitaxial-strain stabilization by annealing films with different strain conditions and have observed the ability to change film morphology and phase composition. By dramatically varying the deposition rate and temperature, we have also observed an order of magnitude change in surface roughness, including the first ever formation of self-assembled, c-axis oriented, fully-epitaxial nanoporous anatase films. These nanoporous structures can be fine-tuned by varying the substrate [YAlO₃ (110), LaAlO₃ (001), NdGaO₃ (110), SrTiO₃ (001), and DyScO₃ (110)] and pore diameters have been found to vary from 10-140 nm and films have been found to possess upwards of 33% porosity in the as-grown state. We will discuss the energetics of this complex structural evolution with specific attention to the competition between nucleation, growth, surface and interfacial energies, and epitaxial strain in an attempt to provide a design algorithm for the development of high-surface area films with tailored surface structures, pore sizes, and densities. Finally, we will introduce ongoing work on the study of photocatalytic activity in these films with attention to increasing photoelectrochemical reactions under AM1.5 light. This includes the study of the effect of changing film morphology on optical absorption as well as light induced degradation of organics.

Transition metal oxides are explored as a treasure trove of novel device-oriented electronic functionalities ranging from interfacial electronic conductance and magnetoelectric coupling to superconductivity. Equally prominent is the role these materials play in energy related applications such as solid-oxide fuel cells (SOFC). Remarkably, materials used in physical applications such as ferroelectric titanates are electrochemically similar to the typical memristive TiO2 and SrTiO3 systems. Similarly, materials such as cobaltites and manganites are broadly explored in physical and energy applications. These similarities naturally lead to question on what is the link between physical and electrochemical responses at the surfaces and interfaces on transition metal oxides. Understanding the interplay between reversible and irreversible electrochemical processes and physical properties necessitates the development of techniques capable of probing these phenomena locally. In this presentation, summarized are the recent studies of reversible-and irreversible electrochemical processes in several prominent materials systems, including cobaltite thin films and incipient ferroelectric titanates obtained using in-situ and ex-situ combination of scanning probe microscopy and scanning transmission electron microscopy. In fully reversible systems, SPM allows probing kinetics and thermodynamics of bias-induced electrochemical transformations. In particular, in non-ferroelectric titanates we observe remament resonses and electromehcnical hysteresis loops, as well as early stages of the electroforming process (formation of conductive Magneli phase filaments). Using the extended Ginsburg-Landau formalism, we explore the coupling between polarization and chemical degrees of freedom and demonstrate that local vacancy injection can give rise to ferroelectric-like state. However, polarization dynamics is now slaved to the ionic responses, providing consistent explanation to multiple observations of relaxor-like polarization in non-ferroelectric oxides. In the irreversible systems, SPM-STEM allows probing vacancy dynamics and ordering, providing window into atomic-level solid state electrochemistry of bulk and interfacial phases. Research supported (SVK, AB) by the U.S. Department of Energy, Basic Energy Sciences, Materials Sciences and Engineering Division and partially performed at the Center for Nanophase Materials Sciences (SVK), a DOE-BES user facility.

The potential use of oxide semiconductors as energy devices, such as dye sensitized solar cells and artificial photosynthesis, utilizing the interface with other semiconducting materials has been investigated [1,2]. Since these devicesâ?T functionality is controlled by the lifetime of electrons and holes at the interface, the interfacial electronic structure is an essential parameter which dominates the device performance. According to previous studies of perovskite oxide heterostructures, the constituent electronic structures are sensitive to the interfacial termination [3,4]. These facts imply the possibility of high tunability of the performance in oxide devices. To fully exploit the fascinating properties of functional oxides, the precise investigation of the control of the electronic structure in oxides by interface engineering is indispensable. In this study, as a first step, we demonstrate a dramatic modulation of the electronic properties of anatase type titanium dioxide (TiO₂) thin films deposited on LaAlO₃ (001) substrates using atomic scale engineering. Anatase TiO₂ is a simple binary oxide semiconductor that is the one of the promising materials for energy devices. Heterostructures formed by TiO₂ thin films deposited on (001)-oriented LaAlO₃ can have two distinct interfacial stacking sequences. These we will refer to as AlO₂-terminated, where the structure is TiO₂/AlO₂-LaAlO₃, and LaO-terminated, where it is TiO₂/LaO-LaAlO₃. By changing the interface termination of the LaAlO₃ from AlO₂ to LaO, the electronic properties of the subsequent TiO₂ film are dramatically transformed from the intrinsic insulating state to a high mobility metallic state, while maintaining excellent optical transparency, and crystalline quality. Such interface engineering is conceptually quite different from other electron doping techniques, such as chemical doping, allowing the carrier density to be increased without negatively impacting the electron mobility. These results have important implications in future energy devices utilizing this interface. [1] J. Grätzel, Nature 414, 338 (2001). [2] S. Sato et al., J. Am. Chem. Soc. 133, 14250 (2011). [3] A. Ohtomo and H. Y. Hwang, Nature 427, 423 (2004). [4] M. Minohara et al., Phys. Rev. B 81, 235322 (2010).