Glass transition kinetics measured by fast scanning calorimetry

Abstract

Enthalpy relaxation data obtained for amorphous selenium by means of fast scanning calorimetry were described in terms of Tool-Narayanaswamy-Moynihan model. The effect of thermal gradients formed within the measured sample was found to explain the marked difference between the FSC and conventional DSC data obtained for a-Se. If the influence of thermal gradients is accounted for, the similar set of TNM parameters as obtained for a-Se from conventional DSC can be used to describe the FSC data. Temperature profiles of the thermal gradients as well as sources of additional uncertainties occurring within the introduced conception are discussed in detail. In addition, a normalization procedure accounting for the difference between the slopes of heat capacity in the glass and undercooled liquid regions was introduced. Correct data preparation was found to have a critical influence on the applied enthalpy relaxation evaluation procedures. Due to the unresolved issue of the prediction of manifesting thermal gradients found in case of the recent study on amorphous selenium, the FSC technique appears, for the time being, to have limited usage for enthalpy relaxation studies.

Fast differential scanning calorimetry: new solutions in data treatment and applications to molecular glass-formers

“…The origin of the broadening can be attributed to the activation enthalpy variation with temperature and to the change of distribution of the relaxation times, together with the occurrence of internal thermal lag to the sample. The latter, has been experimentally studied by Toda [3] and theoretically modeled for FDSC by Svoboda [14].…”

“…In literature there are some examples reporting the eect of dynamical thermal gradients [14] and sample mass [24] but there are no studies about the eects of the temperature of the socalled sensor support temperature T ss , which is the temperature of the reference cold nger: the value of T ss , referring to the calorimeter setup, represents also the minimum temperature available for the measurement, and, in our setup, it can be varied between 173 and 313 K. The behavior of the c p -overshoot peaks at high scanning rate (above 500 K/s) in gs.10, 11, 12 is compatible with the eect of the dynamical thermal gradient, which increases by increasing the scanning rate:…”