Heteroepitaxial growth on silicon surface : a Monte Carlo study

The purpose of this thesis is to investigate the initial stages of the growth of heteroepitaxial films on Si substrates. Two prototype systems were chosen for this research: first is Ge/Si(001), where the two species have similar chemical properties; second is CaF2/Si(111), in which the ionic epitaxial film and substrate have similar crystal structures. Both are strained heteroepitaxial films because of their lattice mismatch. These systems have attracted much attention largely due to various promising applications in micro-electronics and fundamental interest in the basic studies of heteroepitaxy.;The Metropolis Monte Carlo method is used for this research. For Ge/Si, because of short range interaction forces between atoms, the CELL method is developed and applied to this research. Results have shown that the such a method is fast, efficient and is easily adapted to study all other systems with short range interaction forces; the other method, called the BIG JUMP, is also developed. The results have shown that the BIG JUMP method is particularly useful in generating equilibrium or metastable configurations.;The initial stage of MBE growth of Ge on Si(001) was studied using MC method combined with CELL method, or both CELL and BIG JUMP methods. It was found that at least an (8 x 8 x 8) computational cell with six layers that were allowed to move was needed for the simulations. 2% acceptance of MC moves was found to lead a quicker energy minimisation process. This result implies that the energy minimisation process involves big jumps of atoms, corresponding to atom diffusion on a real surface. The energy map of a Ge atom on Si(001)(2 x 1) was calculated and compared with ab initio calculation. An exchange mechanism of a Ge adatom with a Si atom of the substrate was found. This mechanism can be used in understanding the ordered structure observed during the initial stage of MBE growth of Ge on a Si(001) substrate surface.