Structure and Density Heterogeneities of Silica Glass: Insight from Datamining Techniques

Abstract

In this study, we used computer simulations of molecular dynamics to investigate structure and density heterogeneities of silica glass at glass transition temperature, 1475 K. We employed machine learning and data analysis techniques to gain a detailed understanding of the atomic-level structure of the silica glass. The investigation found that the network structure of the silica glass is composed of clusters with varying structures and densities. This structural heterogeneity with clusters is similar to the grains seen in polycrystalline materials. This observation indicates that silica glass can exist in multiple amorphous structural forms, analogous to polymorphism in crystalline materials. Additionally, we applied detailed 3D visualizations to effectively represent the structural characteristics of the silica glass. These visualizations provided valuable insights into the complex and heterogeneous nature of the silica network.