Four different substrates - three grown by hydride vapor phase epitaxy and the fourth by ammonothermal technique were obtained from multiple vendors. All had nominally similar as-received specifications (resistivity, thickness, off-cut angle, bow, surface finish). The substrates were evaluated with a variety of techniques including Raman spectroscopy, photoluminescence, white light interferometry, and Nomarski imaging, enabling to detection of different concentrations of grain boundaries, sample impurities, point defects, v-shaped pits, polishing defects, crystal stress damage, and non-uniform insulating and conducive regions. Comparing these results to those from homoepitaxial growth on the same wafers, the effects are both subtle and overt. Macroscopic surface morphology, which has shown a direct correlation to leakage current, copies and exaggerates that of the underlying substrate. Donor incorporation in the epitaxy is relatively uniform, regardless of the substrate doping. However, homoepitaxy on substrates with inhomogeneous impurity incorporation did show some near band edge broadening by PL. Carbon incorporation varied with substrate offcut angle, which is supported by DFT calculations. All interfaces between epitaxy and substrate have a large Si peak, which requires further study. Following optical characterization, vertical Schottky diodes were fabricated to quickly evaluate device performance. While most of the films showed the ability to withstand high electric fields, those with more uniform properties in the substrate also showed more uniform electrical properties. These results show that the bulk substrates enable the path to high voltage vertical devices, but also show the significant influence that substrates can play in device performance.
