Atomistic modeling of surface nucleation in anorthite-based glasses

Surface crystallization is one of the most challenging issues in glass research due to the complexity of understanding small length scale phenomena that occur over long time periods. Hence, experimental studies of surface crystallization are difficult and time-consuming. In this report, the surface nucleation rates of a set of glasses in the anorthite (CaO-Al₂O₃-SiO₂) compositional family are predicted using molecular dynamics to simulate the glass/crystal interface. In addition, the Toy Landscape Model is utilized to predict the temperature dependence of nucleation. Furthermore, sodium was added to the base anorthite composition to test the validity of the methods presented here for more complex compositions. Results are verified from both experimental and computational perspectives. The approach proposed in this work can be used to guide commercial glass development more efficiently.