GaN Nucleation Landscape on Patterned Sapphire Shaped by the Growth Temperature of Directly Grown Boron-Compound Masks
Yunjin Heo, Donghoi Kim, Hyeonoh Jo, Aelim Ha, Soohyung Park, Jaewu Choi, Chinkyo KimAbstract
The growth temperature of directly grown boron-compound masks on patterned sapphire is associated with changes in the inferred effective accessibility of the underlying sapphire surface and in the subsequent nucleation behavior of GaN. In this work, we investigate how ammonia-borane-derived boron-compound masks grown at different temperatures shape the GaN nucleation landscape within circular SiO2 openings during the initial stage of epitaxial lateral overgrowth. The preferential nucleation position of GaN changes systematically with mask growth temperature: masks grown at 700−750 °C produce pronounced edge-biased distributions, whereas higher-temperature masks lead to more inward-shifted and spatially sparse GaN domains. Quantitative analysis of the GaN areal fraction, the number of visibly isolated domains, and the radial distribution of domain centers show that the mask growth temperature affects both the amount of GaN coverage and the spatial arrangement of GaN domains within each opening. The nonmonotonic change in the number of visibly isolated domains is interpreted as a consequence of competition between reduced lateral merging and reduced effective substrate accessibility rather than as a direct measure of the number of active nucleation sites. Coarse-grained kMC-type simulations provide a qualitative consistency test by varying the effective density and radial distribution of substrate-accessible sites. These results suggest that the growth temperature of directly formed boron-compound masks can serve as a pre-growth process parameter for tuning the early-stage intra-opening GaN nucleation landscape. This behavior is described in terms of effective substrate accessibility inferred from the GaN nucleation morphology, while the microscopic origin of the accessible regions remains to be resolved by direct structural characterization.