Reply to "Comment on 'Infrared-reflectance spectra of heat-treated, sol-gel-derived silica' "

Abstract

Previously reported infrared spectra of amorphous Si${\mathrm{O}}_{2}$ have been discussed in order to clarify the origin of the shoulder at \ensuremath{\sim} 1200 ${\mathrm{cm}}^{\ensuremath{-}1}$ in the transmission spectra and of the high-frequency band at \ensuremath{\sim} 1260 ${\mathrm{cm}}^{\ensuremath{-}1}$ in the reflection spectra. Infrared measurements at oblique incidence with $s$- and $p$-polarized light show that the high-frequency response of silica (between \ensuremath{\sim} 1000 and 1300 ${\mathrm{cm}}^{\ensuremath{-}1}$) originates from two coupled modes A${\mathrm{S}}_{1}$ and A${\mathrm{S}}_{2}$. The polarization characteristics indicate that the features at \ensuremath{\sim} 1200 ${\mathrm{cm}}^{\ensuremath{-}1}$ and \ensuremath{\sim} 1260 ${\mathrm{cm}}^{\ensuremath{-}1}$ have different origin and that it is appropriate to attribute to them TO (A${\mathrm{S}}_{2}$) and LO (A${\mathrm{S}}_{1}$) character, respectively. This assignment was found to be consistent with the infrared-reflectance measurements at oblique incidence reported by Almeida [Phys. Rev. B 45, 161 (1992)] and with those at nearly normal incidence reported by Kamitsos et al. and analyzed by Kramers-Kronig transformation [Phys. Rev. B 48, 12 499 (1993)]. The two-mode interpretation of the high-frequency response of silica was found to be consistent with the trends of the effective transverse-optic and longitudinal-optic frequencies, which were calculated using the second-moment expressions of Noh and Sievers [Phys. Rev. Lett. 17, 1800 (1989)].