Monte Carlo simulation of growth and decay processes in a cluster aggregation source

Using a spatially homogeneous Monte Carlo simulation, we model the agglomeration and fragmentation processes of Cu atoms and clusters in an Ar buffer gas. We model Cu dimer formation to occur via the stabilization of metastable Cu*(2) complexes by Cu or Ar atom collisions. In cluster growth and fragmentation, the heating and cooling effects caused by the binding and recoil energies are taken into account. In this scenario, we study the influence of the dwell time in the cluster source, the gas temperature, and the Cu and Ar densities on the cluster distribution. We find the cluster size distribution to follow a log-normal distribution. Both the average cluster size and the bound atom fraction increase monotonically with the dwell time in the source, until a saturation level is reached, in which cluster formation and decay balance each other. We find an optimum temperature window for the formation of large clusters: For smaller temperatures, collision processes are frozen in, while at larger temperatures, Ar collisions become inefficient for cluster cooling. It is furthermore shown that even a small initial dimer content in the clustering gas speeds up the clustering process considerably. Finally, we show by way of an example how to apply the present model to the description of cluster formation in a spatially inhomogeneous cylindrical aggregation source. (C) 1999 American Vacuum Society. [S0734-2101(99)01201-4].