How can we understand heteroepitaxial growth modes with a computer simulation? (A perspective on the article, “heteroepitaxial growth modes with dislocations in a two-dimensional elastic lattice model”)

Abstract

Please cite this article in press as: S. Tsun In many textbooks, heteroepitaxial growth of crystals is categorized into three typical modes: layer-by-layer growth (Franck–van der Merwe growth), layer-by-layer growth followed by island formation (Stranski–Krastanow growth) and three-dimensional island growth (Volmer–Weber growth). Understanding which one of the growth modes appears for a certain substrate-adsorbate combination has been a longstanding problem from both fundamental and application viewpoints [1]. So far a number of theoretical models and computational methods have been proposed on this problem ranging from macroscopic ones to atomistic ones, although they often contain somewhat ad hoc parameters or otherwise suffer from too much simplification to describe all of the three modes. Difficulty in this problem comes from the interplay of atomic diffusion, formation of three-dimensional islands, and formation of dislocations that relax the strain caused by lattice mismatch between substrate and adsorbate. The simplest solid-on-solid model or lattice model, in which adsorbed atoms are allowed to occupy the same lattice positions as the substrate atoms, could take into account the atomic diffusion and island formation [2], while it is insufficient to consider lattice mismatch and misfit dislocation. The misfit dislocation can be described if atomic positions are treated in real (continuum) space just like in molecular dynamics,