ABSTRACT
Kinetic Monte Carlo (kMC) simulations along with density functional theory (DFT) calculations were used to investigate the aggregation of size-selected Nb3Oy (y = 5, 6, 7) clusters deposited onto the Au(111) surface. Recent STM experiments showed that the cluster binding sites and sizes of the cluster assemblies on the Nb3Oy/Au(111) surfaces strongly depend on the stoichiometry of the clusters, i.e., the oxygen-to-niobium ratio. To better understand the origins of these differences, kMC simulations of the nucleation and growth of cluster assemblies were performed using energy barriers for diffusion and intercluster interactions estimated from DFT calculations of cluster binding and dimerization energies, respectively. Comparisons of the kMC simulations with STM images of the as-deposited Nb3Oy/Au(111) surfaces at RT and after high temperature annealing were used to further optimize the energetics and gauge the importance of nearest neighbor interactions. The kMC simulations demonstrate that the assembly of Nb3Oy clusters on Au(111) are largely controlled by the magnitude of the barriers for diffusion and interparticle-bond formation, while changes at higher temperatures are sensitive to the binding energies between nearest neighbors. Simulations for the Nb3O5 and Nb3O6 clusters, which exhibit smaller cluster assembly sizes in STM, required larger diffusion barriers as well as different barriers for interparticle binding, which reflected differences in DFT calculated dimerization energies. The results demonstrate the effectiveness of combined DFT and kMC calculations for understanding how the stoichiometry affects the aggregation of small oxide clusters on a metal surface.
