Study of initial stages of thin film growth by means of atomistic computer simulation and image analysis: Comparison with experimental data

Abstract

The initial stages of metal film growth were studied by computer experiment. The structures were generated by an atomistic model combining the molecular dynamics and Monte Carlo approaches and the analysis of obtained structures was performed by image processing algorithms. Film morphology was tested by several methods and it was found that especially Voronoi tessellation and Quadrat Counts methods are best suited for a description of spatial distribution of objects. The same morphological analysis was applied to experimental data – to micrographs of discontinuous films of Ag, Au and In grown on dielectric or graphite substrates. It was found that for small evaporation rates the atomistic model is in a good agreement with experiments. However, when increasing the evaporation rates the effect of secondary nucleation starts to be pronounced and a discrepancy between modelled and experimentally derived morphological results occurs. This effect was demonstrated for indium films with intensive secondary nucleation.