Treatment of defects and defect levels in ternary alloys using special quasirandom structures: InGaAs*

While progress has been made toward obtaining demonstrably accurate defect properties in crystalline semiconductors (e.g. GaAs or InAs) using state-of-the-art density functional theory, ternary alloys, such as InGaAs, are equally important technologically and pose significant new conceptual and computational challenges to modelling. With a random composition of In and Ga, lack of periodic crystallinity requires that the conventional supercell approaches be generalized, and local variability in the random alloy requires a statistical approach to assessing defect properties. I describe how one models intrinsic defects in In(0.5)Ga(0.5)As using a special quasirandom structure (SQS), including demonstrating the validity and adequacy of the model (convergence to bulk limits, consistency of chemical potentials), and statistical analysis of defect formation and level energies for vacancies, divacancies, antisites, and some interstitials.