Role of dynamic heterogeneities in crystal nucleation kinetics in an oxide supercooled liquid

Abstract

The temperature at which the classical critical nucleus size is equal to the average size of the cooperatively rearranging regions (CRR) in a supercooled liquid has been referred to as a "cross-over" temperature. We show, for the first time, using published nucleation rate, viscosity, and thermo-physical data, that the cross-over temperature for the lithium disilicate melt is significantly larger than the temperature of the kinetic spinodal and is equal or close to the temperature corresponding to the maximum in the experimentally observed nucleation rates. We suggest that the abnormal decrease in nucleation rates below the cross-over temperature is most likely because, in this regime, the CRR size controls the critical nucleus size and the nucleation rate. This finding links, for the first time, measured nucleation kinetics to the dynamic heterogeneities in a supercooled liquid.

Time-dependent nucleation rate measurements in BaO⋅2SiO2 and 5BaO⋅8SiO2 glasses

“…While sometimes explained as experimental error, the effect is far outside experimental error in these mBaO•nSiO2 studies. Some investigators have tested possible theoretical explanations such as elastic stress, changing size of structural units, spatial heterogeneity and dynamical heterogeneity to account for the effect [27,[29][30][31], but the correct explanation remains unclear due to the lack of experimental or modeling evidence.…”

The race within supercooled liquids—Relaxation versus crystallization

“…In Ref. 29, we showed that the crossover temperature for that liquid is significantly larger than the temperature of the (estimated) kinetic spinodal and is equal or close to the temperature corresponding to the maximum in the experimentally observed nucleation rates. However, in Ref.…”

“…15 In Ref. 29, we made a rough estimate of T KS and demonstrated that it should likely be located below the temperature of (experimental) the nucleation rate maximum for one silicate liquid, but we did not compare it with T K . Finally, in a recent paper, 30 the authors concluded that there is no sign of a thermodynamic or even a "pseudo" spinodal in melt crystallization.…”

“…We made a first attempt on this matter while studying the interplay between cooperatively rearranging regions (CRR) in a supercooled silicate liquid and crystal nucleation kinetics. 29 The temperature at which the classical critical nucleus size is equal to the average size of the CRR has been referred to as a "crossover" temperature. In Ref.…”

“…However, in Ref. 29, we only made a preliminary rough estimate of T KS using the experimental induction times, t ind . On that occasion, the relationships between t ind , the (intrinsic) nucleation timelag, and τ N were not known.…”