Effect of structural relaxation on crystal nucleation in a soda‐lime‐silica glass

Abstract

The influence of structural relaxation on crystal nucleation has been underexplored and remains elusive. This article discusses its possible effect on the nucleation process using a stoichiometric soda-lime-silica (2Na 2 O•CaO•3SiO 2 ) glass as a model system. We show that the relaxation effect is powerful at low temperatures, close and below the glass transition, T g , and leads to a continuous increase in the nucleation rate. At any given temperature, the nucleation rate eventually reaches its ultimate steadystate corresponding to the fully relaxed supercooled liquid (SCL). However, the time to reach the steady-state is two to three orders of magnitude longer than the average relaxation time estimated by the Maxwell relation (shear viscosity / shear modulus).The proposed nucleation mechanism and model, which take relaxation into account, and related experimental results also explain the alleged "breakdown" of CNT at low temperatures reported for various glasses. It confirms a few recent papers that this apparent flaw is merely because most researchers did not prolong nucleation treatments enough to complete the relaxation process to achieve a steady state. Another remarkable result is that the actual maximum nucleation temperature, T max , is significantly lower than the previously reported values. Finally, a comparative analysis of the kinetic coefficient using viscosity versus growth velocity favors the last. These results for this soda-lime-silica glass extend and validate recent findings for lithium disilicate on the significant (but often neglected) effect of relaxation on crystal nucleation.

Theory of crystal nucleation of glass‐forming liquids: Some new developments

“…In particular, a detailed analysis of the experimental data on crystal nucleation given in Refs. 31–34 shows in agreement with a study performed in Ref. 158 that for a quantitative correct description, at least, two structural order parameters are required.…”

“…In line with the analysis performed in Refs. 29–34, we demonstrate here that (ii) at temperatures near to the glass transition range and below it, crystal nucleation is considerably affected by the glass transition. It may result in significant deviations of both thermodynamic driving force and surface tension as compared with the situation when the liquid has approached already the metastable equilibrium state prior to nucleation.…”

“…Some of such approaches are described in Refs. 31–34. In the present paper, we employ Equations () and () treating the parameters $$ \Omega _{\Delta g}$$ and $$\Omega _\sigma$$ as fit parameters.…”

“…Direct experimental data in support of the point of view that nucleation may proceed concomitantly with relaxation, its theoretical description, and the conditions for its occurrence (even more, the necessity of its occurrence for temperatures $$T\lessapprox T_{\text{max}}\cong T_g$$) are presented in Refs. 31–34. They are summarized briefly below as the basis for the analysis of their effects in thermal treatments, that is, cooling and heating processes and their dependence on their rates.…”

“…These ideas are applied to the interpretation of experimental data in Refs. 32–34. They lead to a comprehensive solution of the problem of “breakdown” of CNT at temperatures in the glass transition interval and below it.…”