Size calibration of epitaxial islands via a two-step growth protocol: Kinetic Monte Carlo and effective-medium theory study

The kinetic Monte Carlo technique has been used to simulate irreversible growth of epitaxial islands in two consecutive steps. At the first step a small quantity of adatoms was simultaneously deposited at the surface at random positions and the atoms freely diffuse until nucleating new islands or until being caught by the earlier nucleated ones. It was found that the distribution of the Voronoi cell (VC) areas around the island centers could be accurately described by the Gaussian distribution (GD) which was narrower than the GD describing the VCs of randomly distributed nucleation centers. Thus, our simulations provide an alternative explanation of the narrowing that was observed experimentally and attributed to elastic forces. At the second step the surface was exposed to an atomic deposition flux that was chosen to be small enough for the nucleation of new islands was strongly suppressed and the growth was dominated by the aggregation of deposited atoms into existing islands. At this step the island size distributions (ISDs) obtained could be also well described by the GD only more peaked than the corresponding VC area distributions. The narrowing has been explained in the framework of an effective medium theory. In several cases the simulated VC area distributions and ISDs semi-quantitatively agreed with those observed experimentally. Furthermore, the two-step growth made the island diameter distributions much more symmetric than those obtained under the conventional irreversible growth setup. It is suggested that this technique may provide a method of controlled growth of the island ensembles with narrow and symmetric size distributions in practically any system: homo- or heteroepitaxial.