Relaxational and vibrational dynamics in the glass-transition range of a strong glass formerB2O3

Abstract

Carolyn BRODIN, BÖRJESSON, ENGBERG, TORELL, AND SOKOLOV RELAXATIONAL AND VIBRATIONAL DYNAMICS IN... The structural relaxation behavior of a strong glass former ${\mathrm{B}}_2$ ${\mathrm{O}}_3$ has been investigated over broad temperature (300–1275 K) and frequency (0.5 GHz–10 THz) ranges using depolarized light scattering. The spectra clearly show nonmonotonic temperature behavior with some dynamical crossover at about ${\mathit{T}}_{\mathit{c}}$≊800–900 K. Above ${\mathit{T}}_{\mathit{c}}$ the spectra develop qualitatively according to the general scenario predicted by the mode-coupling theory (MCT), including a fast β process and a much slower α process in addition to a vibrational contribution. However, there is disagreement between the observed functional form of the fast relaxational dynamics and that predicted by MCT. The disagreement seems to be related to the influence of low-lying vibrational contributions, the so-called boson peak, which generally seems to be more pronounced in strong glass formers. Below ${\mathit{T}}_{\mathit{c}}$ the spectra do not follow MCT predictions, not even qualitatively; the main signature is a decrease of the level of the fast relaxation spectrum. Analysis in terms of an alternative phenomenological approach, in which the fast relaxation contribution is related to the damping of the vibrational modes (giving rise to the boson peak), reveals some crossover of the damping rate at about the same temperature ${\mathit{T}}_{\mathit{c}}$ as the crossover of the fast relaxation dynamics itself, and with similar temperature dependence as that recently reported for the Brillouin linewidth. We suggest that these variations are related to the temperature dependence of the relative strength of the fast relaxation. We show that apart from differences in the vibrational contribution, strong and fragile glass formers differ concerning the temperature range of transition (between ${\mathit{T}}_{\mathit{c}}$ and ${\mathit{T}}_{\mathit{g}}$), being narrow for fragile systems (${\mathit{T}}_{\mathit{c}}$/${\mathit{T}}_{\mathit{g}}$≊1.2) and broad for stronger ones (${\mathit{T}}_{\mathit{c}}$/${\mathit{T}}_{\mathit{g}}$≊1.6 for ${\mathrm{B}}_2$ ${\mathrm{O}}_3$). © 1996 The American Physical Society. DOI: https://doi.org/10.1103/PhysRevB.53.11511 ©1996 American Physical Society