Modeling Short-Range and Three-Membered Ring Structures in Lithium Borosilicate Glasses using Machine Learning Potential

TL;DR

This study demonstrates that machine learning potentials can effectively model lithium borosilicate glasses, capturing microstructural features like four-coordinated boron and three-membered rings more accurately than traditional methods.

Contribution

The paper introduces a deep learning-based potential for modeling LBS glasses, overcoming limitations of AIMD and CMD in reproducing microstructures.

Findings

MLP models show better stability than functional force-fields.

MLP accurately reproduces experimental microstructures.

Three-membered rings are well captured by MLP.

Abstract

Lithium borosilicate (LBS) glass is a prototypical lithium-ion conducting oxide glasses available for an all-solid state buttery. Nevertheless, the atomistic modeling of LBS glass using $ab$ $initio$ (AIMD) and classical molecular dynamics (CMD) simulations have critical limitations due to computational cost and inaccuracy in reproducing the glass microstructures, respectively. To overcome these difficulties, a machine-learning potential (MLP) was examined in this work for modeling LBS glasses using DeepMD. The glass structures obtained by this MLP possessed fourhold-coordinated boron ($^4$B) confirmed well with the experimental data and abundance of three-membered rings. The models were energetically more stable compared with those constructed with a functional force-field even though both the models included reasonable $^4$B. The results confirmed MLP to be superior to model the boron-containing glasses and address the inherent shortcomings of the AIMD and CMD. This study also discusses some limitations of MLP for modeling glasses.