Unveiling nucleation dynamics by seeded and spontaneous crystallization in supercooled liquids

L Separdar and JP Rino and ED Zanotto, COMPUTATIONAL MATERIALS SCIENCE, 199, 110802 (2021).

DOI: 10.1016/j.commatsci.2021.110802

Crystallization of supercooled liquids and glasses is of gargantuan scientific and industrial importance. However, investigating its early stages is experimentally very challenging because of the nanometric size and very short lifetime of the critical nuclei; therefore, computer simulations can substantially assist in unveiling the nucleation mechanism and dynamics. In this work, we determined the critical nucleus sizes, the transport coefficient at the nucleus/liquid interface, and the interfacial free energy employing the seeding approach at shallow supercoolings combined with spontaneous homogeneous nucleation rates (Jss) at deep supercoolings in barium sulfide liquid. The nucleation rates were calculated using the MD results for the relevant physical properties and theoretical values of the driving force without any adjustable parameter, and agree quite well with the Jss obtained directly by MD simulations. The results confirm the ability of the seeding method to estimate nucleation parameters. They not only corroborate recent findings for mW water, hard spheres, Lennard-Jones, and zinc selenide models, but also show the best simulations-CNT calculations agreement ever. As this set of relevant outcomes is dispersed for other substances; here we combined and discussed them together to reinforce and generalize the soundness of the Classical Nucleation Theory in describing and predicting crystal nucleation rates for various supercooled liquids.

Return to Publications page