Effects of composition and oxidation states on the structures ofchromium-containing sodium silicate glasses: Molecular dynamicssimulations using machine learning interatomic potentials

Chromium represents a significant challenge for the vitrification of high-level nuclear waste into aluminosilicate and borosilicate glasses due to its low solubility and variable oxidation states, which can limit the waste loading due to promotion of crystallization or phase separation during processing. In this study, we modeled chromium containing silicate glasses using molecular dynamics simulations with three machine learning interatomic potentials (MLIPs): Multi-Atomic Cluster Expansion (MACE), Crystal Hamiltonian Graph Neural Network (CHGNet), and Preferred Potential (PFP). These were employed to gain insights on glass composition and oxidation states in the structures of these glasses. The goal is to evaluate the ability of these MLIPs to accurately represent the general structure of silicate glasses and chromium local environments as a function of chromium oxidation states. Density Functional Theory (DFT) based calculations and experimental data such as neutron diffraction structure factors are used to validate the structural models. The foundation models of the three MLIPs are able to reproduce general structural features of the sodium silicate glass structure consistent with experimental and DFT data, but only CHGNet and PFP accurately capture the oxidation states and local environment of Cr3+ and Cr6+. Furthermore, PFP is used to investigate the effects of the Cr3+/Cr6+ ratio and total chromium content on the structural properties of the glass network. Our results indicate that Cr6+coordinates with four oxygen atoms, with a Cr–O bond distance of 1.66 Å, whereas Cr³⁺ exhibits coordination with five to six oxygen atoms, with a Cr–O bond distance of 1.95 Å. Cr3+ plays the role of network modifier while Cr6+ influences network polymerization by reducing the concentration of non-bridging oxygens (NBO), since Na⁺ ions are consumed for charge compensation during the formation of chromate species (CrO₄²⁻). System size effects on the structural characteristics and chromium environments were also tested. This work highlights the importance of careful validation on the precision, transferability, and potential of MLIPs for modeling glasses containing transition metal elements that can exist in multiple oxidation states.