Date of Award
Summer 8-13-2020
Degree Name
Master of Science (MS)
Degree Type
Thesis
Abstract
β-Ga2O3 is an emergent semiconductor for power electronics applications. It has a wide band gap of 4.8 eV and is transparent on the whole spectrum of visible light up to deep ultraviolet. It has a high Baliga figure of merit (BFOM) — a weighted numerical combination of the dielectric constant, charge carrier mobility, and critical breakdown field —, which is commonly used for a quantitative comparison of semiconductors for high-current operation and power switching applications. β-Ga2O3 can be grown as thin films or as large single crystals by melt growth-techniques, which is important for scalable manufacturing. However, β-Ga2O3 suffers from a lack of p-type dopants and a low thermal conductivity. Presently, all applications are based on n-type β-Ga2O3; the introduction of p-type β-Ga2O3 would enable bipolar power devices. Additionally, high-power switching results in elevated temperatures where heat retention can impede electronic performance.
The objective of this thesis is to investigate, using first-principles density-functional-theory calculations: (1) the efficiency of doping with Bi to raise the valence band of β-Ga2O3 to a level at which p-type doping is achievable, and (2) the possibility of improving the thermal properties of β-Ga2O3 by alloying with the lightest Group-13 cation, B.
We find that doping with Bi creates mid-gap states derived from the Bi 6s electrons at similar energy level to candidate acceptors through an anti-bonding hybridization of the Bi lone pair with the O 2p states. The associated states are more dispersed than the pristine valence band, as they derive from the delocalized s-states of Bi rather than the p-states of highly electronegative O, which dominate the valence band. Our calculations indicate that these intermediate states are natively filled, and at an appropriate energy level to use a co-dopant like Mg, N, or even native Ga vacancies as p-type dopants.
We explored several pathways to include B as an alloy in β-Ga2O3, including the high-entropy alloy (HEA) approach and a variety of ordered binary and ternary alloys. Despite these efforts, we do not find a stable alloy, since the small B atoms reject the octahedral and tetrahedral coordination of the cations in β-Ga2O3 in favor of a flat triangular coordination, as observed in B2O3. We conclude that B is likely insoluble in β-Ga2O3 at concentrations high enough to substantively improve its thermal conductivity.
Language
English (en)
Chair
Rohan Mishra
Committee Members
Katharine Flores Chuan Wang