CONTRIBUTIONS TO THE THEORY OF GLASS FORMATION AND THE GLASSY STATE*

Abstract

ABSTRACTAn attempt has been made to assimilate the most important results of recent glass research into a theory of glass formation and the glassy state and to discuss it in relation to present knowledge of physics, chemistry, and colloidal chemistry.The formation of glass from the liquid state, in contrast to crystallization, is a continuous path from the liquid, through the viscous (supercooled) to the brittle state. The transformation point (Tjp) represents the boundary temperature between viscous and brittle glass, at which boundary the property‐temperature coefficients more or less suddenly change. Property values are influenced by thermal history which can not be explained by mechanical strains. Therefore, while a melt is cooled, an inner change of state, dynamic aggregation, is assumed which takes more and more time as the transformation point is approached and is completely checked at lower temperatures. The question of the thermodynamic stability of the glassy state thus becomes a problem. The difference between “unchecked” (lasting) and “checked” displacements of state is explained with respect to a series of properties.From measurements of the viscosity and of the electric conductivity at temperatures from 1300 down to 300°C, it is shown that the liquid or viscous state follows a simple hyperbolic law, according to which these properties can only continue down to a lower limiting temperature (7V) above the absolute zcro point. The brittle state begins at the transformation point and follows a new law, knowledge of which is made more difficult by sooner or later “checking” the displacements of state. From the previous two different laws, it follows that brittle glass is something other than a supercooled liquid.The transformation point can not be thermodynamically explained. Molecular kinetic considerations make it appear to be that temperature at which densest packing with a cessation of free molecular kinetic motion of the particles is reached. Changes in the brittle state are explained from an atom‐kinetic point of view (shrinking of particle volumes) provided no changes in the state of aggregation occur. Dynamic aggregation takes place by the formation of primary and secondary particles. The structure of secondary particles is compared to a roll of coins and this picture is used to explain negative expansion under pressure, peculiarities in expansion curves, and the beginning of brittleness.According to colloidal chemical considerations, the glassy or viscous states represent a solution of secondary particles as the disperse phase in a dispersing medium of primary particles and individual molecules, and the transformation point represents the boundary temperature between viscous and brittle glass paralleled by the gelatinizing temperature characteristic of the change from a sol to a gel. Support for the colloidal‐chemical conception is obtained from observations on rapidly cooled glass melts and from reaction kinetic. Changes of property in the brittle and viscous states are proportional to the logarithm of the time, z, a law which is also found in elastic after‐working, in isothermal dissociation of carbonates, and in some other instances. This law can he deduced from v. Smoluchowski's theory of coagulation. The disappearance of, double refraction of chilled glasses due to annealing (cooling) can also be linked to the log Z law.