Characterizing In and N impurities in GaAs from ab initio computer simulation of (110) cross-sectional STM images

Computer simulations of cross-sectional scanning-tunneling microscopy images of GaAs(110) are performed by means of density-functional theory, with the aim of characterizing In-Ga and N-As substitutional impurities. In some cases the comparison between an experimental image taken at a single negative bias with the results of our simulations is sufficient to discriminate among different types and locations of impurities. When ambiguities arise, a combined positive/negative scan of the same region will help resolve them. We find that In-Ga-N-As neighboring pairs are energetically favored relative to the isolated impurities. The main features of the images of these pairs are determined by the location of the nitrogen atoms, but details due to neighboring indium atoms are also meaningful and recognizable in the experimental images.