Composition-dependent glass transition temperature in mixtures: Evaluation of configurational entropy models

The composition dependence of the glass transition temperature (T-g) in mixtures remains an important unsolved problem. Here, it is revisited using three model systems: a series of oligomeric and polymeric cyanurates, blends of oligomeric and polymeric alpha-methyl styrene, and molecular mixtures of itraconazole and posaconazole. We evaluate several entropy-based models to determine the theoretical T-g as a function of molecular composition and compare the results against the experimental data. The assumption that the configurational entropy is invariant at the T-g is tested, where the change in configurational entropy is assumed to be given by the integral of Delta C(p)dlnT, where Delta C-p is the temperature-dependent change in the heat capacity at T-g. We find that, although the temperature-dependent heat capacities in both liquid and glassy states are nearly independent of composition for several of the systems studied (i.e., they are nearly ideal mixtures), the composition dependence of T-g is not well described by simply adding the changes in the mass-weighted configurational entropy of the components on going from the T-g in the pure state to that of the blend. The implication is that either configurational entropy is not invariant at T-g or that it cannot be obtained from the integral of Delta C(p)dlnT.