Plasma-assisted molecular beam epitaxy (PAMBE) growth of gallium nitride (GaN) has evolved over the past two decades due to progress in growth science and in the active nitrogen plasma source hardware. The transition from electron cyclotron resonance (ECR) microwave plasma sources to radio frequency (RF) plasma sources has enabled higher growth rates, reduced ion damage and improved operation at higher growth chamber pressures. Even with further improvements in RF plasma sources, PAMBE has remained primarily a research tool partially due to limitations in material growth rates.

This dissertation presents results based upon two modifications of a commercially available nitrogen plasma source. These modifications have resulted in record active nitrogen fluxes, and therefore record growth rates of more than 7.6 μm/h. For optimized growth conditions in the standard metal-rich growth regime, the surfaces displayed a clear step-terrace structure with an average RMS roughness (3 µm×3 µm) on the order of 1 nm. Secondary ion mass spectroscopy (SIMS) impurity analysis demonstrates unintentional oxygen incorporation of ~1×1016, comparable to the metal organic chemical vapor deposition (MOCVD) grown template layer. Additionally, a revised universal growth diagram is proposed allowing the rapid determination of the metal flux needed to grow in a specific growth regime for any and all active nitrogen fluxes available.

High temperature nitrogen rich PAMBE growth of GaN has been previously demonstrated as a viable alternative to the challenges presented in maintaining the Ga bilayer required by metal rich growth of GaN. This dissertation also present results demonstrating PAMBE growth of GaN at a substrate temperature more than 150 °C greater than our standard Ga rich GaN growth regime and ~100 °C greater than any previously reported PAMBE growth of GaN. Finally, a revised growth diagram is proposed highlighting a large growth window available at high temperatures.