Reflections of thin film surface roughness on graphs of specklegram: a novel approach

Abstract

The advancement in thin-film technology necessitated the development of more reliable, sensitive, and non-destructive methods for the quality analysis of thin films. Graph theory, the mathematical and analytical tool, has gained significant importance in analysing complex signals and images. The paper introduces a novel surrogate method based on graph theory to analyse the specklegrams of thin films for quality analysis. The method is deciphered through the complex network analysis of the electronic specklegrams of the molybdenum oxide thin films sputtered at different Argon pressures. For this, graph features are determined from the graphs constructed using the seven equally spaced columns of the specklegram data of a film. The heat map of the specklegram displays the morphological modifications in the film surface. The study reveals a decrease in the centrality measures and multifractal dimension for the film samples. The decrease of root mean square surface roughness of the films calculated from the atomic force microscopic images suggests its relation with centrality measures indicating the potential of centrality measures of the specklegrams as a surrogate method for thin-film quality analysis.