Length distributions of nanowires: Effects of surface diffusion versus nucleation delay

Abstract

It is often thought that the ensembles of semiconductor nanowires are uniform in length due to the initial organization of the growth seeds such as lithographically defined droplets or holes in the substrate. However, several recent works have already demonstrated that most nanowire length distributions are broader than Poissonian. Herein, we consider theoretically the length distributions of non-interacting nanowires that grow by the material collection from the entire length of their sidewalls and with a delay of nucleation of the very first nanowire monolayer. The obtained analytic length distribution is controlled by two parameters that describe the strength of surface diffusion and the nanowire nucleation rate. We show how the distribution changes from the symmetrical Polya shape without the nucleation delay to a much broader and asymmetrical one for longer delays. In the continuum limit (for tall enough nanowires), the length distribution is given by a power law times an incomplete gamma-function. We discuss interesting scaling properties of this solution and give a recipe for analyzing and tailoring the experimental length histograms of nanowires which should work for a wide range of material systems and growth conditions.