Collisional redistribution and saturation of near-resonance scattered light

Abstract

This paper reports on our studies of near-resonant scattering of laser light in a collisional environment. A pulsed dye laser with a peak power of 55 MW/${\mathrm{cm}}^{2}$ was tuned near the 460.73-nm resonance line of strontium and the side emission was observed from an oven containing both strontium vapor and argon buffer gas. The emission was composed of three spectral components: Rayleigh scattering at the frequency of the laser ${\ensuremath{\omega}}_{L}$, fluorescence at the resonance frequency ${\ensuremath{\omega}}_{0}$ of strontium, and a third component at $2{\ensuremath{\omega}}_{L}\ensuremath{-}{\ensuremath{\omega}}_{0}$. These three components have been studied as a function of the frequency and intensity of the laser and also as a function of the argon buffer gas pressure. While the Rayleigh emission was found to vary as ${\ensuremath{\Delta}}^{\ensuremath{-}2}$ ($\ensuremath{\Delta}={\ensuremath{\omega}}_{0}\ensuremath{-}{\ensuremath{\omega}}_{L}$), the fluorescence component, which was produced by Sr-Ar collisions, was found to be asymmetric with the sign of $\ensuremath{\Delta}$ as predicted by line-broadening theory. By measuring the ratio of the intensities of the fluorescence and Rayleigh components, we were able to measure directly the collisional redistribution function, important in the study of radiative transfer in stellar and planetary atmospheres. At high laser intensities all three components were found to saturate. The results were compared with the theoretical predictions of Mollow's steady-state theory. Theoretical fits for the high-intensity results were obtained when the collisional cross sections were taken to be considerably smaller than in our low-intensity measurements. We believe the discrepancy lies in the use of a steady-state theory for a transient experiment. Effects of radiative trapping and spatial averaging are also discussed.