Analysis of quasielastic light scattering inLiNbO3nearTC

We have performed Raman studies of lithium niobate from 293 to 1224 K with special emphasis upon quantitative analysis of the quasielastic scattering from 0–50 ${\mathrm{cm}}^{\mathrm{-}1}$. We find that the complete spectrum, including both the quasielastic ‘‘wing’’ and the two lowest-frequency $A_1$-symmetry optical phonons, can be fitted to the spectral distribution function predicted for a system with a relaxing self-energy; when this is done all the parameters fitted vary slowly and monotonically with temperature. The coupling parameter ${\delta }^2$(T) varies approximately as ($T_C$-T${)}^{\mathrm{-}1.4}$ in accord with the predictions of Halperin and Varma [Phys. Rev. B 14, 4030 (1976)]. This is only the second central-mode study to satisfy that defect-theory prediction. Both congruent and stoichiometric specimens were examined. The congruent sample exhibits an inverse relaxation time of 540±30 GHz that is independent of temperature; it is apparently limited by defects arising from the lack of stoichiometry. The stoichiometric sample exhibits a temperature-dependent relaxation time of the form ${\tau }^{\mathrm{-}1}$(T)=${\tau }_0^{\mathrm{-}1}$[($T_C$-T) /$T_C$] with ${\tau }_0^{\mathrm{-}1}$=1740±430 GHz (i.e., ${\tau }_0^{\mathrm{-}1}$=58±14 ${\mathrm{cm}}^{\mathrm{-}1}$); this is the expected mean-field dependence for a second-order phase transition. The data show why inelastic neutron scattering studies [e.g., M. R. Chowdhury, G. E. Peckham, and D. H. Saunderson, J. Phys. C 11, 1671 (1978)] may have been unable to detect the softening of the lowest $A_1$-symmetry optical phonon and thereby help resolve a long-standing controversy concerning the qualitative disagreement of Raman and neutron scattering data.