Date of Award

Spring 1-1-2015

Document Type


Degree Name

Doctor of Philosophy (PhD)


Materials Science and Engineering

First Advisor

Steven M. George

Second Advisor

Todd G. Deutsch


A sustainable energy economy depends critically on the conversion of renewable energy resources, whose inherent variability requires a storage mechanism. Pathways for conversion of solar energy, being the most abundant, to fuel represent crucial areas of research. Hydrogen as a chemical energy carrier is storable and transportable, while being a feedstock for ammonia fertilizer that is essential to global food supply. Direct photoelectrochemical conversion of sunlight via water splitting is a prominent concept for clean, scalable, cost-effective, and locally produced hydrogen, but the technology is not yet commercially viable. Here, we address the technical challenges of realizing economical solar hydrogen production using III-V semiconductor-based devices: high conversion efficiency and extended lifetime in aqueous electrolyte.

Solar-to-hydrogen conversion efficiency is a fundamental metric for evaluating progress that will impact introduction of commercial solar water-splitting systems. Its definition is generally agreed upon, but measurement technique standards are not well defined. We demonstrate common practices, show how they can lead to significant error, and introduce methodology and cross-validation practices for improved accuracy. The advanced techniques are relevant to device configurations based on tandem absorbers, necessary for achieving maximum conversion efficiency.

We outline the development pathway for III-V tandem devices to reach maximum efficiency, demonstrate progress toward 15% enabled by a new architecture allowing lower bandgaps, and investigate alternative p-i and p-n PEC junction doping profiles that enhance photovoltage. We identify reflection as the primary loss and model anti-reflective TiO2 coatings that demonstrate improved photocurrent.

We present findings on the intrinsic stability of III-V photocathodes and the development of stabilizing surface modifications. We show that water vapor reversibly passivates p-GaInP2 surfaces and derive a model describing the behavior. Bare p-GaAs photocathodes etch ~100x slower than other III-V photocathodes due to residual surface As. Bare p-GaInP2 is unstable, but surface modification involving nitridation and/or PtRu alloy co-catalyst deposition offers corrosion resistance. We show that sputtered PtRu consistently provides better initial performance than other treatment variations, making it preferable for device development. Department of Energy progress milestones are exceeded for STH efficiency and approached for durability, while considerable reduction of device processing cost remains to be addressed.