Abstract:Accessing the thermal transport properties of glasses is a major issue for the design of production strategies of glass industry, as well as for the plethora of applications and devices where glasses are employed. From the computational standpoint, the chemical and morphological complexity of glasses calls for atomistic simulations where the interatomic potentials are able to capture the variety of local environments, composition, and (dis)order that typically characterize glassy phases. Machine-learning potentials (MLPs) are emerging as a valid alternative to computationally expensive ab initio simulations, inevitably run on very small samples which cannot account for disorder at different scales, as well as to empirical force fields, fast but often reliable only in a narrow portion of the thermodynamic and composition phase diagrams. In this article, we make the point on the use of MLPs to compute the thermal conductivity of glasses, through a review of recent theoretical and computational tools and a series of numerical applications on vitreous silica and vitreous silicon, both pure and intercalated with lithium.

What problem does this paper attempt to address?

The paper discusses how to study the thermal transport properties of glass materials through machine learning-driven simulations. Understanding thermal transport characteristics is crucial for designing production strategies in the glass industry, especially when using glass in various applications and devices. Due to the chemical and morphological complexity of glass, traditional atomistic simulations require force fields that can capture the interactions of different local environments, compositions, and disorder, which are often computationally intensive. Machine learning potentials (MLPs) serve as a cost-effective alternative that can avoid the limitations of expensive first-principles simulations, which can only handle small samples and cannot account for disorder at different scales. Additionally, MLPs may be more reliable than empirical force fields within specific ranges of thermodynamics and composition phase diagrams. The paper focuses on discussing the methods for calculating glass thermal conductivity using MLPs, including a review of theoretical and computational tools, and numerically studying the thermal conductivity of pure silica and lithium-doped silica. The study emphasizes the importance of improving the thermal conductivity properties of amorphous solids in contemporary materials science and engineering, particularly in applications such as nuclear waste immobilization and solid-state batteries. The article points out that despite extensive experimental and computational work, there are still gaps in microscale heat conduction theory regarding temperature and composition dependence. The research compares two main computational methods: lattice dynamics applicable to low-temperature solids and equilibrium molecular dynamics (EMD) simulations applicable to property sampling. The limitations of these two methods, such as handling quantum effects and the influence of system size, are also discussed in the paper. Finally, the paper demonstrates the calculation of thermal conductivity in lithium-doped silicon through molecular dynamics simulations to showcase the application of this method in solid electrolyte design, while highlighting the effects of size and disorder on thermal conductivity. Through these studies, the authors aim to advance the understanding of thermal transport properties of glass materials, particularly with the aid of machine learning techniques.

Thermal transport of glasses via machine learning driven simulations

Thermal transport of glasses via machine learning driven simulations

Applications of machine‐learning interatomic potentials for modeling ceramics, glass, and electrolytes: A review

Thermal conductivity of glasses above the plateau: first-principles theory and applications

Predicting dynamic heterogeneity in glass-forming liquids by physics-inspired machine learning

Hydrodynamic finite-size scaling of the thermal conductivity in glasses

Roadmap on machine learning glassy dynamics

Machine learning atomic dynamics to unfold the origin of plasticity in metallic glasses: From thermo- to acousto-plastic flow

Prediction of Silicate Glasses' Stiffness by High-Throughput Molecular Dynamics Simulations and Machine Learning

Prediction and screening of glass properties based on high-throughput molecular dynamics simulations and machine learning

Unveiling the Glass Veil: Elucidating the Optical Properties in Glasses with Interpretable Machine Learning

Unearthing the foundational role of anharmonicity in heat transport in glasses

Monte Carlo simulations of glass-forming liquids beyond Metropolis

Machine learning for predicting thermal transport properties of solids

Predicting Pair Correlation Functions of Glasses using Machine Learning

Evaluation of GlassNet for physics-informed machine learning of glass stability and glass-forming ability

Predicting the pair correlation functions of silicate and borosilicate glasses using machine learning

Machine learning that predicts well may not learn the correct physical descriptions of glassy systems

Elaboration of a neural-network interatomic potential for silica glass and melt

Predicting the propensity for thermally activated β events in metallic glasses via interpretable machine learning

Strength of 2D Glasses Explored by Machine-Learning Force Fields