First-order transitions in glasses and melts induced by solid superclusters nucleated and melted by homogeneous nucleation instead of surface melting

TL;DR

This paper investigates first-order phase transitions in supercooled liquids and glasses induced by homogeneous nucleation of solid superclusters, challenging surface melting theories and providing a new predictive model validated by multiple materials.

Contribution

It introduces a revised classical nucleation model incorporating additional enthalpy, predicting new homogeneous nucleation temperatures and phase transition behaviors in supercooled liquids and glasses.

Findings

Solid superclusters exist above Tm and melt via homogeneous nucleation.

The revised model accurately predicts nucleation temperatures for various materials.

Critical supercooling rate of 0.198 matches experimental data on Sn droplets.

Abstract

Supercooled liquids give rise, by homogeneous nucleation, to solid superclusters acting as building blocks of glass, ultrastable glass, and glacial glass phases before being crystallized. Liquid-to-liquid phase transitions begin to be observed above the melting temperature Tm as well as critical undercooling depending on critical overheating (Tm-T)/Tm. Solid nuclei exist above Tm and melt by homogeneous nucleation of liquid instead of surface melting. The Gibbs free energy change predicted by the classical nucleation equation is completed by an additional enthalpy which stabilize these solid entities during undercooling. A two-liquid model, using this renewed equation, predicts the new homogeneous nucleation temperatures inducing first-order transitions, and the enthalpy and entropy of new liquid and glass phases. These calculations are successfully applied to ethylbenzene, triphenyl phosphite, d-mannitol, n-butanol, Zr41.2Ti13.8Cu12.5Ni10Be22.5, Ti34Zr11Cu47Ni8, and Co81.5B18.5. A critical supercooling and overheating rate (Tm-T)/Tm = 0.198 of liquid elements is predicted in agreement with experiments on Sn droplets.