Analysis of AFM Images of Self-Structured Surface Textures by Directional Fractal Signature Method

Abstract

A new method, called augmented blanket with rotating grid (ABRG), has been proposed in our recent work on characterizing roughness and directionality of self-structured surface textures. This is the first method that calculates fractal dimensions (FDs) at individual scales and directions for the entire surface image data and does not require the data to be Brownian fractal. However, before the ABRG method can be used in real applications, effects of atomic force microscope (AFM) imaging conditions on FDs need to be evaluated first. In this paper, computer-generated AFM images with three different resolutions, 48 combinations of tip radii and cone angles, and 15 noise levels were used in the tests. The images represent isotropic self-structured surface textures with small, medium and large motif sizes, and anisotropic surfaces exhibiting two dominating directions. For isotropic surfaces, the ABRG method is not significantly affected (i.e. FDs changes <5 %) by image resolution, tip size (for surfaces with large motifs) and noise (except the level above 8 %). For anisotropic surfaces, the method exhibits large changes in FDs (up to −34 %). The results obtained show that the ABRG method can be effective in analysing the AFM images of self-structured surface textures. However, some precautions should be taken with anisotropic and isotropic surfaces with small motifs.