In our paper we wrote "It is common to fit the temperature dependence of the volume with a polynomial and units in degrees Celsius, yielding incorrect values below 273 K. 13 " The sentence should read: "Pawlus et al. 13 fitted the temperature dependence with a polynomial."We also wrote that "A similar equation was given by Ref.18, but their polynomial for the temperature dependence of the volume yields incorrect values below 273 K." The second part of the sentence should be omitted.

No abstract

We have investigated the fragility of glycerol, propylene carbonate, and metafluoroaniline at pressures up to 700 MPa using dielectric spectroscopy. All three small-molecule glass formers do not show a pressure dependence of fragility within experimental error. The ratio of glass transition temperature and Vogel-Fulcher-Tammann temperature is pressure independent. The results are discussed in the frame of the defect diffusion model.

We have investigated the dielectric properties of a 17.2 mol % mixture of chlorobenzene and decalin in the supercooled and in the glassy state under hydrostatic pressure up to 400 MPa. The dielectric loss shows the well-known primary (α) relaxation and a secondary (β) relaxation the origin of which is still being debated. The glass transition temperature Tg increases linearly with pressure. The fragility is pressure independent. With increasing pressure the relaxation strength of the β process rises at the expense of the α process. Above around 160 K, where devitrification occurs at atmospheric pressure, the activation energy of the secondary relaxation process is roughly twice that at lower temperatures. Both energies are proportional to Tg, indicating a strong correlation between α and β processes. A possible origin of the secondary relaxation is rationalized by critical clusters in the melt.

Glass forming monohydroxy alcohols show an exponential Debye-type dielectric relaxation, which proceeds slower than the structural relaxation. Dielectric high-pressure measurements of 2-ethyl-1-hexanol reveal that isochronal scaling is violated. Temperature-volume scaling was found to be valid separately for the Debye-type and for the structural relaxation, yielding a scaling exponent of γ≈1.8.