Analysis of the size-distribution function of metallic nanoclusters in a hydrogenated amorphous carbon host

A method for mathematical processing of TEM images of nanocomposites with a pronounced noise background is proposed. This method is based on the spline filtering of an image by using the values of average background intensity and is applied to TEM images of amorphous Cu-doped carbon. A size distribution function obtained for copper clusters is in good agreement with previous data. The dependence of the cluster-to-cluster distance on cluster size is determined. This dependence validates the hypothesis of the fluctuation nature of cluster generation at a lower metal content and of the coalescence of clusters at higher metal contents.