Resonant Brillouin Scattering Studies of the a - in Cadmium Sulphide.

High resolution resonant Brillouin scattering was performed in the vicinity of the A-exciton in cadmium sulphide. The experimental arrangement, which consisted of a single mode dye laser, triple-pass Fabry-Perot interferometer and a grating spectrometer, enabled Brillouin shift, line -width and intensity measurements to be made on multi-Brillouin components. These components result from the scattering between acoustic phonons and propagating exciton-polariton modes in the crystalline medium. In an attempt to identify the correct additional boundary condition (ABC) which is needed in describing the simultaneous propagation of two exciton-polariton modes, a theoretical external differential scattering cross section calculation was developed via a factorization procedure. This calculation allowed a quantitative comparison between experiment and theory. The important results of this research are now given. (1) The observation of the inner polariton branch participating in the one phonon scattering processes {both interbranch (1-2'), (2-1') and intrabranch (1-1')} between exciton-polaritons and acoustic phonons. These additionally observed Brillouin components have resulted in revised values for the A-exciton parameters. (2) Broadening of both the Stokes LA (2-2') and TA (2-2') line-widths has been observed as a function of incident laser frequency. These measurements, however, suggest that the exciton damping parameter, (GAMMA), may not be a constant but instead a monotonically increasing function of incident frequency (omega)(,I) for (omega)(,I) > (omega)(,T). (3) Comparisons between the experimental intensity measurements and the theoretical external scattering cross sections for both Stokes LA (2-2') and TA (2-2') components suggest that the additional boundary condition formulated by Ting et al results in the best agreement between experiment and theory. (4) The intensity ratio between the Stokes LA (1-1') component and the Stokes LA (2-2') component is 100 times smaller than theoretically predicted even though the general trend of the LA (1-1') intensity as a function of incident frequency follows these same theoretical predictions for any of the three ABCs. (5) Least square fits between theoretical and experimental reflectivity expressions were also made. Each ABC results in a reasonably good fit with experiment when a dead layer is included in the analysis.