Investigating the role of network former interactions on charge carrier diffusivity in glasses

Abstract

Ionic transport is a critical property for the glass industry, since emerging applications such as sensors, batteries, and electric melting are based on the phenomenon. Short-range interactions (anion-charge carrier) have not been able to explain the total activation barrier observed experimentally, and, as such, it is critical to understand the larger role of all ions in a glass, not just the carrier and the ‘site’ ions. This research focuses on the role of network formers and their impact on diffusion in glasses, something that current models lack an explicit explanation of. Atomistic simulations with randomly generated parameters for the cation potentials and classical simulations were used to determine the diffusion coefficients and activation energies for synthetic network formers. Using this database, explainable machine learning algorithms were employed to explore network former interactions and determine which parameters are the most influential for ion diffusion. Results suggest that the bond length of the cations changes the geometry of the structure contributing the greatest to cation-modifier interactions.