THEORIES OF LIQUID-GLASS TRANSITION

A number of theories have been proposed to describe atomic (molecular) movements in glassy solids; their common point is that they consider collective processes when the temperature approaches T(g), despite the fact that the heterogeneity of the microscopic structure of glassy materials has been implied by a number of theories. The structure of a glassy solid may be considered as a random distribution of local regions of spatially fluctuating density and high energy (and entropy) in an atomically disordered continuum. These local regions have often been described in terms of free volume or alternatively as quasi-point defects (qpd). Above T(g), the concentration, C(d), of qpd is temperature dependent as the system remains in metastable thermodynamic equilibrium, but when the temperature is decreased below T(g), C(d) becomes constant. The hierarchical constraints assumption yields a theory of correlated molecular movements relating to a lifetime tau(mol) characterizing those movements to the structural state, i.e. C(d), via a correlation parameter increasing with C(d). Such an approach is compared to those of other authors and can be used in order to describe the rheological behaviour of glasses near T(g). The preceding views result from physics of the solid state; it is worth comparing them with the results obtained by the mode coupling theory applied to the liquid state. In the latter case, it is deducted that at some temperature T(c) higher than T(g), there is an ergodic to non-ergodic phase transition. This phenomenon corresponds to a 'blocking' effect of the dynamic behaviour of the liquid and has sometimes been claimed to be the true liquid-glass transition. In order to reconcile these approaches, it is necessary to take into account thermally activated processes occurring at T(c) which stabilize the liquid state for a while until T(g) is reached. Then the transition at T(g) corresponds to the thermally activated and collective modes being frozen in.