No abstract

The vibrational Fermi resonance of two liquids, methanol (CH3OH) and dichloromethane (CH2Cl2), is investigated by measuring changes in the position and intensity of Fermi-coupled Raman bands as a function of pressure, in a diamond anvil cell. The Fermi resonance of interest occurs in the 2900 cm-1 spectral region, where coupling between the CH symmetric stretch fundamental and a CH bend overtone gives rise to two prominent bands. The methanol results reveal a pressure induced transition through exact resonance at 1.25 GPa, where the two coupled states decompose into a pair of fully mixed hybrid bands. In dichloromethane, on the other hand, the two coupled states are driven farther apart and become less mixed with increasing pressure. The Fermi resonance coupling coefficient, W, is found to be constant in each liquid up to pressures exceeding 1 GPa (W ≈ 52.6 and 22.3 cm-1 in CH3OH and CH2Cl2, respectively). The anharmonic shift of the CH bend is about 10 cm-1 in both liquids, determined by comparing the frequencies of the fundamental and Fermi resonance corrected overtone. The results are compared with those of previous Fermi resonance studies using solvent, phase, isotope, temperature, and pressure variation. In addition to yielding a robust method for quantifying Fermi resonance, pressure variation is shown to offer a powerful aid to the resolution of spectral assignment ambiguities.

Raman vibrational frequencies of HCl in argon were measured at pressures up to 110 MPa. The mean frequency of the asymmetric Q-branch is shown to accurately measure vibrational shifts through a density region where line shape changes due to motional narrowing render the peak maximum an inaccurate measure of pressure induced frequency shifts. A semiclassical, analytical expression utilizing Hutson’s HCl–Ar pair-potentials is used to determine the derivative of the HCl vibrational frequency with respect to Ar density in the limit of zero density. The predictions are in reasonable agreement with experimental results, although the experimental frequency shifts are about 20% smaller (less redshifted) than theoretical predictions, which may represent the influence of multibody interactions. Experimental HCl Raman Q-branch and S-branch linewidths and peak shifts are compared qualitatively with previous R-branch (IR absorption) results. Separation of the vibrational (Q-branch) and rotational parts of the frequency shift suggest that the rotational contribution is positive (blueshifted) for all J values and approaches zero with increasing J.

Molecular vibrational frequency shifts are modeled by treating the vibrating solute as a breathing sphere with repulsive (hard-core) and attractive (mean field) solvent−solute interactions. The solvent-induced repulsive force exerted on the normal mode of a vibrating solute molecule is obtained using the derivative of the molecular volume with respect to the normal mode coordinate in conjunction with an analytical expression for the chemical potential of a hard sphere solute immersed in a solvent of hard spheres. The volume derivative of the vibrating solute molecule is calculated by considering the solvent molecule as an assembly of interpenetrating or fused hard spheres whose individual motions are given by normal mode coordinates. The calculation of the repulsive force is simplified by equating the normal mode volume change of the multisphere solute to a volume change of a spherical solute. The anisotropy of the solute−solvent system is used to adjust the spherical solute diameter so its chemical potential matches the chemical potential of the multisphere solute. This adjustment is critical to correctly scale the spherical volume change to the anisotropy of the polyatomic solute. For the diatomic solutes, N2 and HCl, the breathing sphere model predicts repulsive forces and frequency shifts that closely agree with a near exact diatomic hard sphere model. Applied to the three different CO2 vibrations, the explicit calculation of the different excluded volume changes associated with each vibration gives predicted frequency shifts that agree with available pressure-dependent frequency data.