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

In a recent article, Zhu et al. suggest that a discrepancy exist between viscosity measurements performed by oscillation viscometry compared to other techniques such as capillary or parallel plate methods. They argue that oscillation viscometry data increasingly deviate with temperature leading to a systematic mismatch at high temperature. They conclude that this discrepancy discredit the existence of most fragile-to-strong transitions (FST) reported recently that are based on this type of measurement. Here we show that when all the data available are taken into account, this argument loses its coherency. Indeed, if FST were an artefact of instrumental error then all liquids measured using oscillation viscometry over the same temperature and viscosity range should exhibit an apparent FST. Instead, data available at high temperature for Se shows no FST but in fact show an excellent match with model prediction over the full temperature range, while data for GeSe ₃ measured over the same range show a severe kink revealing an FST which therefore cannot be assigned to a mismatch between measurement techniques. Furthermore we show that oscillation viscometry data can be independently corroborated using heat capacity data via the Adam Gibbs equation. Finally we show that large vapor pressure losses occur in Ge–Se glasses at high temperature and are likely responsible for the reported deviation in viscosity measurement. This is not the case of oscillation viscometry measurements that are performed in sealed ampoules. Overall these results confirm the validity of oscillation viscometry as well as the existence of fragile-to-strong transitions.