Journal of Vacuum Science & Technology B, Vol.25, No.3, 1072-1076, 2007
Kinetic Monte Carlo simulation of InAs quantum dot growth on nonlithographically patterned substrates
In this work, the authors use the kinetic Monte Carlo method to simulate the heteroepitaxial submonolayer epitaxial growth process on a prepatterned semiconductor substrate. Experimentally, the GaAs substrates are patterned prior to growth by reactive ion etching through a self-organized nanopore anodized aluminum oxide film. At the early stages of growth, the InAs nuclei appear at the edges of the pores and then continue to grow from these locations, eventually filling the pores entirely. The shape of the pores greatly influences the location of the nuclei. If the pores are smoothed during the high temperature thermal cleaning process, the selectivity of nucleation is reduced or lost. The results of the kinetic Monte Carlo simulation demonstrate that a likely theoretical explanation of the observations is that the nonuniform distribution of the substrate elastic strain energy that is brought in by the lattice mismatch in the InAs wetting layer is the origin of the selectivity in the dot nucleation process. (c) 2007 American Vacuum Society.