Separation of collision-induced from intrinsic molecular depolarized Rayleigh scattering optical anisotropy of the C—Cl bond

Abstract

A procedure is demonstrated for separating the intrinsic molecular component of the depolarized Rayleigh scattering (DRS) from the transient, collision-induced contribution. It depends on the use of two interference filters of differing breadths (18 and 53 cm^–1, respectively) to distinguish the narrow reorientational component of <3 cm^–1 half-width from the much broader collisional scattering. The procedure is applied to CCl₄ and four other spherically symmetric molecules in the liquid state. All exhibit small residual scattering (Rayleigh ratios R_HV≈ 0.01–0.02 × 10^–6 cm^–1) at low frequencies ω from the central line, in addition to copious induced scattering over a broader range of ω. Approximately half of the scattering by neat liquid CHCl₃, (CH₃)₂CHCl, (CH₃)₃CCl and CH₃CCl₃ is collision induced. Measurements on CCl₄ solutions of each of these anisotropic molecules, corrected for collisional scattering and extrapolated to infinite dilution, yield molecular anisotropies γ²= 2.6, 1.6, 2.2 and 2.1 Å⁶, respectively. These results give Γ_CCl=Δα_CCl–Δα_CH= 1.6, 1.5, 2.0, 2.0 ų, respectively, where Δα_CCl and Δα_CH are the indicated bond anisotropies. They are in good agreement with values obtained from electric birefringence and absolute Raman intensities. Discordant values of γ², and hence of Γ_CCl, found by Bothorel and co-workers are in error due to their failure to eliminate collisional scattering. Conformational averages 〈γ²〉 for ClCH₂CH₂Cl and for meso and racemic CH₃CHClCH₂CHClCH₃ calculated by rotational isomeric analysis using Γ_CCl= 1.5 ų, as found for (CH₃)₂CHCl, are in good agreement with values measured in CCl₄, viz., 5.5, 2.2 and 0.5 Å⁶, respectively.