Applicability of the adiabatic nucleation theory to glasses

Woelk and Strey recently criticized the temperature dependency of nucleation rates of the isothermal Classical Nucleation Theory, of water in its vapor, and showed that the apparent excellent agreement of the theory with experimental data, in a narrow temperature range, is in fact a "crossover" of the theory with experimental data. Similarly, Zanotto and Fokin criticized the isothermal Classical Nucleation Theory for nucleation (devitrification) in silicate glasses, claiming that "the theory fails to describe nucleation rates in glasses quantitatively." Finally, Schmelzer, Boltachev and Baidakov stated that: "...deviations between theoretical and experimental data for the steady-state nucleation rate may reach the order of 10 20 -10 30 for condensation of one component vapors and even 10 100 or more for phase formation in glass-forming melts." This situation shows the need for new proposals to explain the phenomenon. We demonstrate that a simple Adiabatic Nucleation Theory predicts nucleation rates in "fragile" oxide glasses (T oo /T g >0.6, where Too is the Fulcher temperature and T g is the glass transition temperature) with an average deviation of only nine orders of magnitude (too high) and shows (within this uncertainty) correctly which glasses show devitrification and which ones do not, in experiments during "laboratory times." The theory could not yet be extended to "strong" glasses because of the very low Fulcher temperatures of these glasses (T oo /T g <0.6), in which the Stokes-Einstein relation loses its validity completely.