Absence of thermodynamic phase transition in a model glass former.

Nature · 2000

basic_science · Level V

Where this comes from

Abstract

The glass transition can be viewed simply as the point at which the viscosity of a structurally disordered liquid reaches a universal threshold value. But this is an operational definition that circumvents fundamental issues, such as whether the glass transition is a purely dynamical phenomenon. If so, ergodicity gets broken (the system becomes confined to some part of its phase space), but the thermodynamic properties of the liquid remain unchanged across the transition, provided they are determined as thermodynamic equilibrium averages over the whole phase space. The opposite view claims that an underlying thermodynamic phase transition is responsible for the pronounced slow-down in the dynamics at the liquid-glass boundary. Such a phase transition would trigger the dynamic standstill, and then be masked by it. Here we perform Monte Carlo simulations of a two-dimensional system of polydisperse hard disks far within its glassy phase. The approach allows for non-local moves in a way that preserves micro-reversibility. We find no evidence for a thermodynamic phase transition up to very high densities; the glass is thus indistinguishable from the liquid on purely thermodynamic grounds.