Abstract:The structural complexity and vast composition space of glassy materials can make design and optimization challenging. Of particular importance for glass processing and design is a material's glass‐forming ability (GFA). However, there remain many open questions regarding the underlying mechanisms of glass formation. Here, we explore the possibilities of physics‐informed machine learning for the prediction of a GFA surrogate property, glass stability, to guide our understanding of the mechanisms, as well as to inform glass processing and design. Glassy materials form the basis of many modern applications, including nuclear waste immobilization, touch‐screen displays, and optical fibers, and also hold great potential for future medical and environmental applications. However, their structural complexity and large composition space make design and optimization challenging for certain applications. Of particular importance for glass processing and design is an estimate of a given composition's glass‐forming ability (GFA). However, there remain many open questions regarding the underlying physical mechanisms of glass formation, especially in oxide glasses. It is apparent that a proxy for GFA would be highly useful in glass processing and design, but identifying such a surrogate property has proven itself to be difficult. While glass stability (GS) parameters have historically been used as a GFA surrogate, recent research has demonstrated that most of these parameters are not accurate predictors of the GFA of oxide glasses. Here, we explore the application of an open‐source pre‐trained neural network model, GlassNet, that can predict the characteristic temperatures necessary to compute GS with reasonable performance and assess the feasibility of using these physics‐informed machine learning (PIML)‐predicted GS parameters to estimate GFA. In doing so, we track the uncertainties at each step of the computation—from the original ML prediction errors to the compounding of errors during GS estimation, and finally to the final estimation of GFA. While GlassNet exhibits reasonable accuracy on all individual properties, we observe a large compounding of error in the combination of these individual predictions for the PIML prediction of GS, finding that random forest models offer similar accuracy to GlassNet. We also break down the performance of GlassNet on different glass families and find that the error in GS prediction is correlated with the error in crystallization peak temperature prediction. Lastly, we utilize this finding to assess the relationship between top‐performing GS parameters and GFA for two ternary glass systems: sodium borosilicate and sodium iron phosphate glasses. We conclude that to obtain true ML predictive capability of GFA, significantly more data needs to be collected.

Enhancing the Prediction of Glass Dynamics by Incorporating the Direction of Deviation from Equilibrium Positions

Rotation-equivariant Graph Neural Networks for Learning Glassy Liquids Representations

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

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

Roadmap on machine learning glassy dynamics

Towards interpreting the thermally activated $β$ dynamics in metallic glass with the structural constraint neural network

Toward Interpreting the Thermally Activated β Dynamics in Metallic Glass Using the Structural Constraint Neural Network

Predicting Glass Structure by Physics-Informed Machine Learning

Learning molecular dynamics: predicting the dynamics of glasses by a machine learning simulator

A geometry-enhanced graph neural network for learning the smoothness of glassy dynamics from static structure

Dynamic heterogeneity at the experimental glass transition predicted by transferable machine learning

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

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

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

Deciphering a Structural Signature of Glass Dynamics by Machine Learning

Predicting orientation-dependent plastic susceptibility from static structure in amorphous solids via deep learning

Selecting Relevant Structural Features for Glassy Dynamics by Information Imbalance

Machine-learning Inspired Density-Fluctuation Model of Local Structural Instability in Metallic Glasses

What Do Deep Neural Networks Find in Disordered Structures of Glasses?

Machine Learning Approach for Prediction and Understanding of Glass-Forming Ability.

Ensemble learning predicts glass-forming ability under imbalanced datasets