Nucleation and crystallization of lithium diborate glass.

Abstract

The magnitude and temperature dependence of both the nucleation and crystal growth rates in lithium diborate glass were determined in the temperature range, 490 to 520°C. Comparison of the nucleation rates predicted by Classical Nucleation Theory and those found experimentally shows that the predicted classical nucleation rates are about 95 orders of magnitude smaller than the experimentally determined values. In addition, Classical Nucleation Theory does not predict the temperature dependence found experimentally. Comparison is also made with silicate glass systems which have been shown to exhibit homogeneous nucleation. Crystal nucleation in the lithium diborate glass almost certainly proceeds by a homogeneous mechanism. Comparisons are made between experimentally obtained values of the crystal growth rate in lithium diborate glass and those computed using surface nucleated crystal growth theory. Although the temperature dependence of the experimental growth rates at large undercoolings appears to be described well by the latter model, the computed values of the growth rates are about 60 orders of magnitude too small. Using a temperature dependent surface tension (obtained from fitting crystal nucleation data) in the surface nucleated crystal growth model partially reduces the discrepancy between the experimental and calculated magnitudes of the growth rate, but produces an incorrect prediction for the temperature dependence of the growth rate.