Answers are grounded in this paper and cite their sources.
The infrared-absorption spectrum of a-SiO₂ is analyzed in terms of its transverse-optic (TO) and longitudinal-optic (LO) vibrational modes. It is shown that the independent-oscillator model for the a-SiO₂ dielectric function fails to yield a consistent value of mode strength for the optically active oxygen asymmetric stretch (AS₁) TO mode at 1076 cm^-1 (in-phase motion of adjacent oxygen atoms) when different but equivalent methods of measurement and analysis are used. This inconsistency is resolved by introducing disorder-induced mechanical coupling between the AS₁ mode and the relatively optically inactive oxygen asymmetric stretch (AS₂) mode (out-of-phase motion of adjacent oxygen atoms) into the oscillator model. Coupled AS₁- and AS₂-mode LO-TO frequency pairs are experimentally observed as peaks at approximately 1256--1076 cm^-1 and 1160--1200 cm^-1, respectively, in oblique-incidence p-polarized absorption spectra of thin a-SiO₂ films grown thermally on c-Si wafers.Additionally, two other LO-TO-mode pairs are observed in these spectra as absorption peaks at approximately 820--810 cm^-1 and 507--457 cm^-1. The simplest form of the coupled-mode model consistent with experiment is found to be one in which the AS₁-mode LO-TO frequency splitting is due to the AS₁ transverse effective charge and the AS₂-mode LO-TO splitting is due to the mechanical coupling between these two modes and not to the AS₂ transverse effective charge, which is negligibly small. The AS₂ TO and the AS₁ LO modes found at approximately 1200 and 1256 cm^-1, respectively, are shown to be consistent with experimental mode strengths and with the analytic requirements that all LO and TO modes be interspersed and that, as a result of lying between the AS₁-mode LO-TO pair frequencies, the AS₂-mode LO-TO frequency splitting be inverted. Comparison of these experimental LO-TO-mode pair frequencies with the vibrational density of states (VDOS) spectrum of a-SiO₂ shows that the TO absorption peaks correspond quite well with the VDOS spectral peaks, whereas the LO absorption peaks do not appear to exhibit any such correspondence.
No takes yet. Share an insight, caveat, or question.
Charles Kirk (1988) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: