Structure of Si/Ge nanoclusters: Kinetics and thermodynamics

The optimal structure and element distribution of SixGe1−x clusters was investigated in terms of free energy. The methods employed were computational simulations based on classical molecular dynamics. Clusters obtained in our previous work were further simulated through annealing at different temperatures. In addition, a combination of molecular dynamics and a semi-grand-canonical Monte Carlo algorithm was used to find a free-energetically favorable element configuration for the clusters. The results show that annealing at conventional temperatures improves the clusters’ sphericity only slightly, and they remain much more amorphous than clusters cut out from crystalline bulk; only at extreme annealing temperatures are the sphericity and crystallinity notably improved. Furthermore, Ge atoms are found to segregate to the surface of the clusters, which greatly reduces the free energy of the clusters.