GaN layer growth optimization for high power devices

In this work, a novel method for GaN layer optimization — statistical multi-parameter design of experiments (DOE) — is presented. According to the statistical model obtained, increasing the buffer layer V/III ratio is beneficial for minimizing the full width at half maximum (FWHM) of the X-ray diffraction rocking curve for the (002) reflection. Statistical models were also obtained for background electron concentration and room temperature Hall mobility, but further data analysis and electrical measurements lead us to the conclusion that those models are disturbed by the presence of a highly conductive layer near the GaN/sapphire interface. Inclusion of an AlxGa(1−x)N isolation layer results in a reduction of two orders of magnitude in the measured background concentration, as well as a significant increase in Hall mobility, without degradation of the crystalline quality.