GROWTH AND CHARACTERIZATION OF ULTRAWIDE-BANDGAP OXIDE SEMICONDUCTORS

Benjamin L. Dutton

doi:10.7273/000007363

Power electronics are ubiquitous in the world today. From the wall charger for a phone to the inverter in an electric vehicle, electricity is converted into a usable form by these power electronics and delivered to the device. With an estimated 30% of all electrical energy passing through power electronics for frequency switching and conversion, even small losses of energy as waste heat during each of these conversions result in an outsized impact. While silicon, with a bandgap of 1.12 eV, is the basis of a vast majority of power electronic devices, wide bandgap semiconductors (bandgap >3 eV) have the potential to push the envelope in terms of device performance. GaN (3.2 eV) and SiC (3.4 eV) have already transformed possibilities for low and high voltage applications, respectively, due to their enhanced breakdown voltages, which enable devices with improved size, weight, and power (SWaP). Substantial development of wide bandgap semiconductors as optoelectronic materials has occurred concurrently given the abundant possibilities for sub-bandgap defect introduction and bandgap modulation with alloying. A significant drawback in using GaN and SiC technologies is their lack of a native melt-grown substrate, leading to substantial cost increases. β-Ga2O3 is an emerging ultrawide bandgap (4.85 eV) semiconductor which can be grown directly from the melt, unlike GaN, SiC, and other ultrawide bandgap materials currently under development (AlN, diamond, c-BN). This dissertation seeks to further β-Ga2O3 based technology with a vertically integrated approach, from fundamental understanding of how defects incorporate and alter optoelectronic properties to optimization of crystal growth and processing by leveraging these defects diagnostically. Mn is explored as an alternative acceptor dopant to the incumbents (Fe, Mg) typically used in insulating substrates. Unique polarization and orientation dependent optical absorptions, or pleochroism, of the Mn:β-Ga2O3 highlighted the anisotropy of its monoclinic structure, which was also manifest in other optically active dopants (Cr, Zn, Cu). Alloying with In and Gd was attempted in order to modulate the size of the optical bandgap and improve overall photoresponse; however, the large ions severely degraded crystal quality and introduced secondary phases. Spatial photoluminescence (PL) and Raman spectroscopy measurements are presented in conjunction with mass spectrometry techniques as tools for understanding impurity segregation across different melt-growth methods (Czochralski, edge-defined film-fed growth, floating zone, and vertical gradient freeze). Barium titanate served as a model material for crucible selection, in order to better understand oxide-metal crucible interactions which may then be applied to eliminate expensive iridium from β-Ga2O3 growth. Lastly, attempts at growing metastable rutile structured oxides (GeO2, SiO2) are presented as alternatives to β-Ga2O3, given their potential as p-type doped ultrawide bandgap materials.

GROWTH AND CHARACTERIZATION OF ULTRAWIDE-BANDGAP OXIDE SEMICONDUCTORS

Files and links (1)

Abstract

Metrics

Details