Progress in modeling of glass properties using topological constraint theory

Abstract

Prediction of glass properties as a function of composition has been a long-standing problem. Topological constraint theory has built the quantitative relationships between composition and macroscopic properties, which can be used to predict compositional dependence of glass properties. In this review, we will summarize the recent important progress for the development of the theory. The topological model for glass transition temperature has been proposed from temperature-dependent constraint theory. We will also discuss the effect of water on topological constraint, how to account for mixed alkali effect and mixed alkaline earth effect, and the application of the model to new glass family besides inorganic glasses. Topological models for hardness and thermal expansion have also been developed. Topological models for predicting elasticity have been recently proposed. We also show some example on the application of the theory in designing new glasses and we suggest some future directions of the theory.

Predicting glass structure by physics-informed machine learning

“…With this approach, the number of glasses within the prediction range of the model expands exponentially as the training data expands linearly; however, the assumption of direct correlations between reactions in binary and ternary glasses has been found to be insufficient in some systems. That is, the model can show a systematic error when calculating the structures of ternary glasses from the inputs obtained from their binary counterparts 25 .…”

“…This is partly due to the fact that glass properties often exhibit a non-linear dependence on composition 8 , which makes it challenging for models to capture such non-linearity and yield robust extrapolations. In contrast, the relationships between local structure (e.g., as captured from the glass connectivity) and properties are often fairly linear 25 . The linear nature of structureproperty relationships makes it significantly easier for models to generalize well outside their training sets, when extrapolated toward unexplored compositional spaces 19,26,27 .…”

Predicting spectroscopic properties of quaternary phosphate laser glasses

“…However, previous exploration of new glass and its properties mainly relied on the empirical “trial and error” method with long experimental cycles and low efficiency, which greatly restricted the rapid development of related fields. To overcome this limitation, researchers have tried to predict the properties of glass theoretically through empirical formulas or models 4–9 . Despite this, existing physical and empirical models, such as addition methods, statistical methods, and machine learning, have little effect on predicting C–S–P relationships.…”

Quantitatively predicting the optical and spectroscopic properties of Nd3+‐doped laser glasses

“…However, this model has been hardly used in the calculation of spectroscopic properties. Modern methods, such as molecular dynamic simulations and topological constraint theory, provide a promising alternative pathway to simulate the microstructure of glass, but they still cannot describe the macroscopic structure and are mostly used to predict physical properties of glasses 14–16 . This means that the CSP internal correlation and quantitative prediction of glass spectroscopic properties remain obscure in glass research.…”