Composition-Dependent Density Modeling of Chalcogenide Glasses: Empirical Evidence for Influential van der Waals Interactions and the Central Role of Lone-Pair Electrons

Abstract

We present an expanded empirical framework for predicting the density of chalcogenide glasses (ChGs) and use it to examine atomic bonding interpretations across a broad compositional landscape. Building on our earlier methodology and dataset work, we compile 468 measured densities spanning 60 elemental families and 19 constituent elements. Using an iterative nonlinear optimization to determine element-specific high- and low-concentration empirical coefficients, combined with a Sigmoid interpolation for concentration dependence, the model predicts density directly from composition with high accuracy (majority of predictions within ±2.5% of measured values; nearly all within ±5%). Beyond predictive performance, we analyze systematic trends in the fitted coefficients and show that the results may possibly be explained by a network landscape noticeably influenced by weak, non-shared-electron interactions that are well described by van der Waals radii and volumetric packing considerations, rather than exclusively by covalent electron sharing, as conventional theory promotes as untested fact. To account for the nonstructural anomalies of low-temperature aging, we introduce a conceptual framework rooted in electromagnetic interactions, establishing a testable hypothesis to be fully formalized in a forthcoming companion work. The primary contribution is a reproducible, data-driven density model that both serves as a practical design tool and motivates a testable physical interpretation of bonding and polarizability in ChGs.   Received: 13 April 2026 | Revised: 25 August 2026 | Accepted: 18 September 2026   Conflicts of Interest The authors declare that they have no conflicts of interest to this work.   Data Availability Statement The data that support the findings of this study are openly available in the Supplementary Materials (Table S1). The legend in Figure 1 reflects these data and provides a simple means for cross-referencing to Table S1 in the Supplementary Materials.   Author Contribution Statement Thomas J. Loretz: Conceptualization, Validation, Investigation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project Administration. Richard A. Loretz: Conceptualization, Methodology, Software, Validation, Formal Analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing.