Summary: The Art and Science of glass formation. Ideal glassformers as opposed
to "ideal glass"es.
C. Austen Angell,
Department of Chemistry,
Arizona State University, Tempe, AZ 85287
There is a common belief in glass science that, if a liquid of known melting point
supercools readily, its glass transition temperature, Tg, can be predicted from the "2/3
rule", viz., Tg = 2/3Tm. We start by arguing that this "law" is a tautology. Accordingly,
we look for extreme violations to the high end, seeking "ideal" glassformers defined by
the condition Tm<< Tg . Of course, Tm cannot be determined if Tm<< 1.5Tg so the search
must be conducted indirectly, by extrapolation of systematic results, with no final
guarantee that the "ideal" glassformer has been found. Our principal strategy is
systematic study of melting properties of closely related systems of simple character,
which we conduct using molecular dynamics simulations with "potential tuning". We
commence with known crystals of a system, perturb their interaction potentials
systematically, and determine the temperatures at which they "melt" ( or amorphize).
Having found systems in which there appears to be no crystal with a positive heat of
fusion, (hence no stable crystal phase), we then consider the effects of second
components that mix non-ideally, on the liquidus relations. The conclusion is that "ideal"
glassformers can, and do, exist.