Abstract Gordon has shown that the Landau theory of second-order phase transitions when applied to the glass transition provides a plausible explanation of the apparent coincidence of the temperature To of viscosity divergence and the T2 of the Kauzmann paradox. It can be shown however that two current statistical mechanical theories of the transition, the Gibbs-DiMarzio theory and one proposed by Goldstein which takes into account nonconfigurational contributions to the entropy, do not follow the Landau theory. We are thus presented with a dilemma: if the Landau theory is applicable to the glass transition, neither of the two statistical mechanical theories can be, and vice versa.
Read full abstract