The current understanding of intrinsic point-defect equilibria was first reviewed. With regard to defect equilibria in the bulk, the amphoteric native defect model was applied to the Ga vacancy equilibrium concentration. The model was then extended so as to include surface Fermi-level pinning, in order to predict the profile of the equilibrium concentration of Ga vacancies in the sub-surface layers. This extension generalized the formation kinetics of point defects, and naturally predicted a surface bottle-neck effect for point-defect formation. The surfaces did not always act as perfect sinks or sources for intrinsic point-defect formation via Schottky-type defect formation. Dislocations induced a bulk bottle-neck effect in which they did not always act as perfect sinks and sources. Diffusivity data on plasma-induced intrinsic point defects at low temperatures were compared with high-temperature data on vacancies. They were found to be of acceptor type and were highly mobile, even at room temperature. The point defects were suggested to be interstitials rather than vacancies.

K.Wada: Applied Surface Science, 1995, 85, 246-52