Connection between quasielastic Raman scattering and free volume in polymeric glasses and supercooled liquids

Abstract

Quasielastic light scattering (QLS) in the frequency interval 100–1000 GHz is measured in some polymers: polycarbonate, polybutadiene, polystyrene, and poly(methyl methacrylate). To describe the spectra, a model of the fast picosecond relaxation processes responsible for the QLS, which is based on the damping of the boson peak vibrations by the dynamic hole volume fluctuations, is used. Within the frame of the model, the intensity of the fast relaxation process is proportional to the fractional dynamic hole volume (which above the glass transition temperature Tg is known as the fractional free volume). The hole volumes can be measured using the positron annihilation lifetime spectroscopy (PALS). The comparison of the literature PALS data in the four polymers with the QLS shows an apparent correlation between the relaxation strength and the fractional dynamic hole volume in good agreement with the predictions of the model.

Intermolecular vibrations and fast relaxations in supercooled ionic liquids

“…41,85 In low-frequency Raman spectroscopy, it is known that the temperature dependence of the QES intensity, I QES , in the supercooled liquid range is steeper than below T g . [27][28][29][30][31][32][33][34][57][58][59][60] The change of slope of the I QES (T) curve at T g is at first sight a remarkable finding, since QES is related to processes taking place at picosecond time scale while the glass transition is due to the structural α-relaxation time, τ α , reaching 10 2 -10 3 s. In fact, the relationship between I QES and α-relaxation is another example of these correlations between fast and slow dynamics of glassforming liquids.…”

“…[57][58][59][60] It has been already shown that both the models give good fit to 31 but these models need assumptions concerning the functions to be used in fits. Thus, we follow here instead a less model dependent procedure, and we considered I QES as the integral up to the boson peak, say 4 < ω < 20 cm −1 , of the experimental spectrum normalized by intensity of high frequency bands.…”

“…It has been further assumed that the relaxations that promoted damping of the boson peak are due to free volume, so that a relationship between δ 2 (T) and V f has been proposed. 58,59 The data of u 2 for [bmIm]Cl are shown in Fig. 8 in a different 101 i.e., u 2 not purely vibrational but including relaxation, in order to reproduce the temperature dependence of the effective activation energy.…”

“…spectra of glass-forming liquids. [57][58][59][98][99][100] In the coupling model, [57][58][59] the pronounced rise of I QES above T g is due to the increase of the coupling parameter, δ 2 (T), between fast relaxations and the boson peak vibrations. The parameter δ 2 (T) is the fraction of the relaxational to the vibrational plus relaxational contributions, so that δ 2 (T) can be used as a measurement of I QES (T).…”

“…Such correlation strengths the proposition of an indirect mechanism for the QES contribution, 56 for instance, as in the coupling model, in which the QES intensity increases with temperature due to damping of the boson peak vibrations. 34,55,[57][58][59] Rather than a superposition model of two independent functions for QES and boson peak, the Raman spectrum is calculated in the coupling model by averaging a purely vibrational contribution, usually assumed as the Raman spectrum recorded at very low temperature, with a function giving the relaxations from all of the processes damping the vibrations. Both the superposition and the coupling models can give good fit to experimental spectra, for instance, see the studies of the glassformer GeSBr 2 31 Therefore, the coupling model is favored on the basis of arguments as mentioned above, but not just from better agreement to experiment.…”