Response to comment on “An experimental critique on the existence of fragile-to-strong transition in glass-forming liquids”

Abstract

In a recent study parallel plate and capillary rheometry techniques were employed by us to measure the viscosity of selenium (up to 673 K) and two GeSe liquids (up to 823 K) over a temperature range sufficient to suggest that oscillation viscometry may systematically underestimate the viscosity of liquids at high temperatures. Consequently, doubt was cast on the validity of the existence of a fragile-to-strong transition (FST) in glass-forming liquids postulated in the literature on the basis of oscillation viscometry data. The Comment by Lucas questions the accuracy of the viscosity data for GeSe liquids in our study on the basis of the observation of significant mass loss due to evaporation of Se from supercooled GeSe3 liquid when heated in a TGA at temperatures ≥823 K. Lucas also reasserts the accuracy of viscosity measured by oscillation viscometry as well as the existence of FST by showing consistency between the viscosity and the configurational entropy of Ge15Te85 liquid, within the framework of the Adam-Gibbs model. We argue that Lucas' understanding of the boundary conditions and the experimental protocols for our viscosity measurements is incorrect and present experimental evidence that indicate the lack of any detectable change in the chemical composition of the GeSe liquids during viscosity measurements. We also point out that the recently reported viscosity data for the Ge15Te85 liquid is in fact in disagreement with Lucas' entropy-based calculations. Furthermore, these calculations not only ignore the electronic contribution to specific heat for a semiconductor-to-metal transition in the Ge15Te85 liquid, but more importantly the fact that the Adam-Gibbs model may not be applicable at extremely low viscosity, where the liquid dynamics crosses over from an activated viscous regime to a non-activated fluid regime.