Generating random surfaces with desired autocorrelation length

A versatile surface processing method based on electrostatic deposition of particles and subsequent dry etching is shown to be able to tailor the autocorrelation length of a random surface by varying particle size and coverage. An explicit relation between final autocorrelation length, surface coverage of the particles, particle size, and etch depth is built. The autocorrelation length of the final surface closely follows a power law decay with particle coverage, the most significant processing parameter. Experimental results on silicon substrates agree reasonably well with model predictions.