Frequency-dependence of second refractivity virial coefficients of small molecules between 325 nm and 633 nm

Heavy rare-gas atomic pairs and the "double penalty" issue: Isotropic Raman lineshapes by Kr2, Xe2, and KrXe at room temperature

We report absolutely calibrated isotropic Raman lineshapes for Kr2 and Xe2 and for KrXe at 294.5 K and compare them to quantum-mechanically generated lineshapes by using state-of-the-art second-order Møller-Plesset and DFT/B3LYP data sets for the induced mean dipole polarizability ᾱ. A very good agreement between the numerical and the experimental data was observed but the large uncertainty margins and the short Raman frequency interval probed in our experiment prevented us from rating on a more refined scale the performance of the tested ᾱ models. These drawbacks are inherent in isotropic Raman spectrum measurements and amplified for dissimilar pairs because, for such systems and spectra, the unreliable operation of subtracting optical signals of comparable magnitude occurs twice per Raman frequency shift value, thus penalizing twice the quality of the measured data. In light of our findings and of previously reported evidence about related electric properties in Kr2 and Xe2 and in KrXe, we are left with no doubt as to the consistency of the induced-polarizability and interatomic-potential data used for these three systems at the reported level of accuracy.

The collision-induced polarizability of a pair of hydrogen molecules

Collision-induced light scattering, impulsive stimulated scattering, and subpicosecond-induced birefringence all depend on the transient changes Deltaalpha in molecular polarizabilities that occur when molecules collide. Ab initio results for Deltaalpha are needed to permit comparisons with accurate experimental results for these spectra and for refractive index virial coefficients and dielectric virial coefficients. In this work, we provide results for Deltaalpha for a pair of hydrogen molecules, treated at CCSD(T) level, with an aug-cc-pV5Z (spdf) basis set. Our values replace the best previous ab initio results for the variation of Deltaalpha with intermolecular separation, the self-consistent-field results obtained by Bounds [Mol. Phys. 38, 2099 (1979)] with a relatively small (3s2p) basis set for H2. For the six geometrical configurations studied by Bounds, the inclusion of correlation and improvements in the basis tend to increase both the trace Deltaalpha(0)0 and the anisotropy Deltaalpha2m of the pair polarizability. The change in the anisotropy is relatively small, but our values for the trace differ by factors of 2 or more from Bounds' results. For use in computing experimental line shapes, intensities, and virial coefficients, we have calculated Deltaalpha for 18 different relative orientations of a pair of H2 molecules, with the intermolecular separation R ranging from 2 a.u. (3 a.u. for a linear pair) to 10 a.u. The H2 bond length is fixed at the vibrationally averaged internuclear separation in the ground state r=1.449 a.u. Our results agree well with the CCSD(T) results for Deltaalpha obtained by Maroulis [J. Phys. Chem. A 104, 4772 (2000)] for two pair configurations of H2...H2 (linear and T-shaped) at a fixed internuclear distance of R=6.5 a.u. in a [6s4p1d] basis. As the intermolecular distance increases (for R>or=8 a.u.), the spherical-tensor components of Deltaalpha converge to the results from a long-range model that includes dipole-induced-dipole (DID) interactions, higher-multipole induction, nonuniformity of the local field, hyperpolarization, and van der Waals dispersion. Deviations from the first-order DID model are still evident for R between 8 and 10 a.u. in most orientations of the pair. At shorter range, overlap damping, exchange, and orbital distortion reduce both Deltaalpha0(0) and Deltaalpha(2)0 below their long-range limiting forms.