UC Berkeley

Global dependence on fossil fuels as energy sources and the alarming increase of greenhouse gas emissions has necessitated the development of clean renewable energy sources for the future. Solar energy is one of the most promising renewable energy sources because of the large flux of photons from the sun. While solar energy can be collected through the use of solar cells, such photovoltaic devices are limited by the intermittency of sunlight. Therefore, new methods are necessary to store solar energy for energy usage on demand. Solar water splitting is one such method, which is attractive because it can produce hydrogen as a fuel source in a carbon free process. However, no known materials exist which can satisfy all of the requirements for efficient spontaneous solar water splitting. Here, we explore vapor phase chemical vapor deposition methods to bandgap engineer InxGa1-xN nanowires and (Ga1-xZnx)(N1-xOx) nanotubes for solar water splitting.

Significant synthetic challenges remain for the epitaxial growth of high-quality InGaN across the entire compositional range. One strategy to address these challenges has been to use the nanowire geometry because of its strain relieving properties. Here, we demonstrate the heteroepitaxial growth of InxGa1-xN nanowire arrays (0.06 ≤ x ≤ 0.43) on c-plane sapphire (Al2O3(001)) using a halide chemical vapor deposition (HCVD) technique. Scanning electron microscopy (SEM) and x-ray diffraction (XRD) characterization confirmed the long range order and epitaxy of vertically oriented nanowires. Structural characterization by transmission electron microscopy (TEM) showed that single crystalline nanowires were grown in the <002> direction. Optical properties of InGaN nanowire arrays were investigated by absorption and photoluminescence (PL) measurements. These measurements show the tunable direct band gap properties of InGaN nanowires into the yellow-orange region of the visible spectrum. To demonstrate the utility of our HCVD method for implementation into devices, LEDs and photoanodes were fabricated from InxGa1-xN nanowires epitaxially grown on GaN(001). Although initial LED and photoanode testing were successful, devices were limited by the poor internal quantum efficiency (QE) of InGaN nanowires. Current-voltage measurements of photoanodes showed that the H2O oxidation activity of InGaN nanowires is limited by recombination.

An annealing procedure in ammonia ambient is investigated to improve the QE of InxGa1-xN nanowires (0.07 ≤ x ≤ 0.42) grown on c-Al2O3. Morphological studies using scanning electron microscopy confirm that the nanowire morphology is retained after annealing in ammonia at temperatures up to 800 °C. However, significant indium etching and composition inhomogeneities are observed for higher indium composition nanowires (x = 0.28, 0.42), as measured by energy-dispersive x-ray spectroscopy and Z-contrast scanning transmission electron microscopy. Structural analysis, using x-ray diffraction and high-resolution transmission electron microscopy, indicate that this is a result of the greater thermal instability of higher indium composition nanowires. The effect of these structural changes on the optical quality of InGaN nanowires is examined using steady-state and time-resolved photoluminescence measurements. Annealing in ammonia enhances the integrated photoluminescence intensity of InxGa1-xN nanowires by up to a factor of 4.11±0.03 (x = 0.42) by increasing the rate of radiative recombination. Fitting of photoluminescence decay curves to a Kohlrausch stretched exponential indicates that this increase is directly related to a larger distribution of recombination rates from composition inhomogeneities caused by annealing. The results demonstrate the role of thermal instability on the improved QE of InGaN nanowires annealed in ammonia. Increases in the rates of recombination indicate that such a treatment will be more useful for carrier recombination devices such as LEDs, rather than carrier extraction devices such as photoanodes. For photoanodes, further work will be necessary to improve the QE of InGaN nanowires by decreasing the rate of nonradiative recombination.

Recently, (Ga1-xZnx)(N1-xOx) has gained widespread attention as a comparatively high efficiency photocatalyst for visible-light-driven overall water splitting. Despite significant gains in efficiency over the past several years, a majority of the photogenerated carriers recombine within bulk powders. To improve the photocatalytic activity, we used an epitaxial casting method to synthesize single-crystalline, high surface area (Ga1-xZnx)(N1-xOx) nanotubes with ZnO compositions up to x = 0.10. Individual nanotubes showed improved homogeneity over powder samples due to a well-defined epitaxial interface for ZnO diffusion into GaN. Absorption measurements showed that the ZnO incorporation shifts the absorption into the visible region with a tail out to 500 nm. Gas chromatography (GC) was used to compare the solar water splitting activity of (Ga1-xZnx)(N1-xOx) nanotubes (x = 0.05-0.10) with similar composition powders. Co-catalyst decorated samples were dispersed in aqueous solutions of CH3OH and AgO2CCH3 to monitor the H+ reduction and H2O oxidation half reactions, respectively. The nanotubes were found to have approximately 1.5-2 times higher photocatalytic activity than similar composition powders for the rate limiting H+ reduction half reaction. These results demonstrate that improvements in homogeneity and surface area using the nanotube geometry can enhance the photocatalytic activity of GaN:ZnO for solar water splitting.

Vapor phase growth methods were successfully developed to bandgap engineer 1-D InxGa1-xN and (Ga1-xZnx)(N1-xOx) alloys. For both alloys, results indicate that 1-D nanostructured geometries provide benefits for the formation of a more homogeneous alloy. Photoelectrochemical and photocatalytic studies demonstrate that bandgap engineering is a viable strategy towards discovering new material compositions for solar water splitting.