Published online by Cambridge University Press: 15 February 2011
Recently we developed a method that allows to derive a height-height autocorrelation function (ACF) for a randomly rough surface of a thin – film growing by the 3D island mechanism. In this paper, the ACF and roughness spectrum for several versions of the island growth model deduced by this method are presented and discussed. In the model dimension d=1+1 the ACFs are deduced in exact analytical form, and for d=2+1 the ACF' s are calculated numerically. In both model dimensions for the rms roughness the universal analytical expression connecting its with physical deposition parameters is derived. The impact of the growth hillock shape and the hillock space distribution are studied. It is shown that the ACF form is determined by the hillock space distribution and in the cases of an ideally random Poisson distribution and slightly ordered distribution the ACF is nearly gaussian although the surface profile height distribution is non-gaussian. The rms roughness is strongly dependent on the hillock shape while the autocorrelation length is determined by the hillock number density. The theoretical predictions are compared with published TEM and STM-data and the scaling behaviour of the rms roughness is analysed.