Fast Dynamics in Glass Forming Systems: Vibrations vs Relaxation
- PDF / 808,214 Bytes
- 12 Pages / 414.72 x 648 pts Page_size
- 25 Downloads / 195 Views
ABSTRACT Two contributions specific for the spectra of the fast dynamics in glass forming systems, a broad quasielastic scattering and the boson peak, are analyzed. It is shown that the vibrational contribution (the boson peak) decreases strongly in fragile systems. Some speculations about dependence of the degree of fragility (a la Angell) on peculiarity of the spectrum of fast dynamics are presented. The existence of some intrinsic relation between the broad quasielastic contribution and the boson peak is demonstrated from analysis of the recent neutron and Raman scattering data. It is shown that this relation can be explained in framework of the model of damped oscillator. The model ascribes the quasielastic contribution to the scattering of light or neutrons on the vibrations around the boson peak, which are damped by some relaxation channel and have a quasielastic part in their response function. It is demonstrated that the model can explain many peculiar properties of the fast dynamics in the Raman, neutron and far-infrared absorption spectra. INTRODUCTION It is well-known that the main difference between dynamic spectra of crystals and glasses (or other disordered systems) appears in the mesoscopic frequency range (v-l0-1000 GHz). Ordered crystals in this frequency range usually have acoustic vibrations with Debye-like density of states g(v)ctv 2 only. However, in all glass forming systems a strong deviations from the Debye-like behavior was found [1]: at very low frequencies (v>ra, here V is a sound velocity and ra is an interatomic (intermolecular) distance. In the frequency range around the boson peak the value V/Q is still significantly large than ra and we assume that the hydrodynamic approach for M(co,Q) still can be applied. There are some experimental results supporting this assumption: the soundlike behavior expected in the hydrodynamic limit, f-uVq and Focf22 , has been found up to
75
-400GHz for LA modes in SiO 2 and some other glasses [28]; the recent measurements of inelastic X-ray scattering also demonstrated the same behavior for LA mode in SiO 2, glycerol and LiCl-water solution even at frequencies above the boson peak [29]. So, this assumption looks reasonable although still needs a theoretical justification. Now one can make some qualitative predictions. Let's analyze the "relaxation-like" contribution in the case of a weak damping IM(Q) I
Data Loading...
