Clustering analysis to identify structural and density heterogeneity in Mg2SiO4 glass

Spatial distribution of HD (blue) and LD (gray) regions in Mg 2 SiO 4 glass (A) and the largest 10 clusters within the HD region (B). In this study, advanced data analysis techniques combined with molecular dynamics simulations are employed to investigate the structural characteristics of Mg 2 SiO 4 glass. Key structural features, including short-range order, intermediate-range order, and ring statistics, are analyzed to elucidate the distribution of Mg 2+ ions and their influence on the glass network. The DBSCAN clustering algorithm, integrated with structural analysis methods, reveals micro-phase separation with distinct Mg-rich regions. Additionally, the results highlight significant structural and density heterogeneities. To further clarify this phenomenon, a two-phase model is proposed, and the structural characteristics of each phase are characterized. The findings reveal the presence of both a high-density phase and a low-density phase, along with distinct differences in intermediate-range order between these phases. Furthermore, void distribution analysis provides deeper insights into the structural heterogeneity of Mg 2 SiO 4 glass. Advanced 3D visualization techniques offer an in-depth perspective on the atomic-scale complexity of this multi-component oxide glass. These results not only advance the fundamental understanding of amorphous Mg 2 SiO 4 but also provide valuable insights for tailoring material properties in advanced technological applications.