Glasses and time

Abstract

A glass is essentially a liquid which has failed to crystallize. The formation of a glass is therefore a kinetic problem. The processes of nucleation and growth of crystals in glass-forming liquids are reviewed in simple terms, and a discussion of sub-liquidus metastable liquid-liquid immiscibility is included. If nucleation and crystallization are avoided during cooling of a liquid and the viscosity rises above 1013 P, a second rate process is encountered when the atomic configuration cannot adjust itself quickly to a sudden change of temperature. The properties of a glass then change with time at constant temperature. After sufficient time the properties reach a constant equilibrium value and the glass is said to be stabilized. Even stabilized glass shows time-dependent elasticity but the behaviour is linear; by an application of Boltzmann's superposition principle it can be demonstrated that the equilibrium configuration will probably take many hundreds of years to be attained when the viscosity is about 1020 P. However at lower temperatures it is possible to observe the early stages of this approach to equilibrium. At room temperature contributions to time-dependent properties come from the beginnings of stabilization, from delayed elasticity and, in addition, from the motion of alkali ions.