Structural instability and relaxation in liquid and glassy phases near the fragile liquid limit

Abstract

Listen

Translate article

We consider two sorts of structural instability of glassy substances, each of which may be released by a relaxation process with its own characteristic relaxation time, and specific kinetic features. The first of these is the instability against relaxation out of the amorphous state into the crystalline state, while the second is the instability against relaxation within the amorphous state itself. The latter may often involve relaxation out of a homogeneous amorphous phase into a two-phase amorphous structure, but we will not specifically consider this liquid-liquid phase separation process here. In most glasses, the former (which is no more than the characteristic nucleation time) is much longer than the latter time. However, there are important classes of glasses, for instance the metallic glasses, in which the former is in fact the shorter time, a fact which is responsible for the inability to observe the glass transition phenomenon in such substances. In this paper we will be considering the relation between these two times and the specific kinetics of each. The nucleation time has been the subject of theoretical developments over a number of decades, and details will be omitted in order to concentrate on experimental studies of this phenomenon. We will described briefly the recently developed DSC techniques for determining the classical time-temperature-transformation curves for a variety of supercooled liquids, and the relation of these to the nucleation curves. The relaxation process within the amorphous state, which can be observed for cases where the nucleation time is relatively long, has a number of features which currently lack a complete explanation. In most cases the relaxation process is non-Arrhenius in its temperature dependence, nonexponential in its time dependence, and nonlinear in its structural state dependence. Some examples taken from glasses at the “fragile” edge of the deduced viscosity-temperature pattern for glassforming liquids are dealt with in detail, and the distinction between shear stress relaxation and thermodynamic stress relaxation is made. The possibility that near T g the latter relaxation time remains Vogel-Fulcher in form with T 0 ≡ T K (the Kauzmann temperature), in contrast with the common observations for (the decoupled) shear relaxation, is raised. Strong support for this notion is found in the current “specific heat spectroscopy” results of Nagel and co-workers. Microscopic relaxation processes, as observed using spectroscopic probes and neutron scattering techniques, are reviewed, and the difference in non-exponentiality from macroscopic relaxation are examined in the light of current theories. Finally, secondary relaxations in ionic and molecular glasses, and their relation to the fastest of all glassy state relaxation processes, the tunnelling modes (TLS), are briefly considered.