A theoretical study of vibrational mode coupling in H5O2+

Abstract

The vibrational mode coupling in the protonated water dimer is investigated by performing two types of quantum calculations of the vibrational levels of H5O2+ and D5O2+, utilizing the OSS3(p) potential energy surface by Ojamaë et al. [L. Ojamaë, I. Shavitt, and S. J. Singer, J. Chem. Phys. 109, 5547 (1998)]. One is four-dimensional (4D), treating only the central O⋯H(D)+⋯O moiety. Three of the four modes considered, the asymmetric stretch and the two bends, are largely the vibrations of the central proton, while the fourth mode is essentially the O⋯O stretching vibration. The vibrational levels of O⋯H(D)+⋯O are calculated rigorously, as fully coupled (FC), and also in an adiabatic (3+1)D approximation, where the proton asymmetric stretch is treated as adiabatically separated from the other three degrees of freedom. The second set of calculations, designated VCI, is full-dimensional, 15D; it is performed by the code MULTIMODE, which does configuration interaction (CI) calculations using a basis determined from a vibrational self-consistent field Hamiltonian. The FC 4D and 15D VCI calculations give very similar fundamental frequencies of the two bending modes of the central proton, as well as the O⋯O stretch. They differ substantially only for the fundamental of the proton asymmetric stretch, the VCI value being about 25% lower than the FC 4D result. This shows that the asymmetric stretch is strongly coupled to the vibrations outside the O⋯H(D)+⋯O fragment, in contrast to the two proton bending modes and the O⋯O stretching vibration. The FC 4D and 15D VCI calculations predict the same frequency ordering of the four vibrational modes of the O⋯H(D)+⋯O moiety, and are in excellent agreement with respect to the H–D shift of the fundamentals of the shared proton modes. The adiabatic (3+1)D treatment is not quantitatively accurate, yielding fundamental frequencies of the proton vibrational modes which are considerably different from the FC 4D results. Our results have potentially significant implications for the assignment of the bands associated with shared proton vibrations in the recently reported infrared multiphoton photodissociation spectrum of the protonated water dimer.

IR Spectrum of the H5O2+ Cation in the Context of Proton Disolvates L−H+−L

“…Band c has been assigned by most authors to the δ(H 2 O) bending vibration of the peripheral water molecules (Table 4), and this is consistently reproduced by theory for the bare H 5 O 2 + cation except in the calculations of Dai et al 16 and Kaledin et al 47 is the basicity of L, the stronger will be the OH‚‚‚L interaction, and the more significant will be these frequency shifts. This trend is actually seen in the condensed-phase data as L increases base strength from benzene to phosphine oxide (points [4][5][6][7][8].…”

“…Moreover, this ν as OHO stretching vibration is believed to have lower energy than the associated bending vibrations. 15,16 This contradicts a norm of IR spectroscopy where stretching vibrations always have higher frequencies than the associated bending vibrations. In addition, an unusual broadening, diminishment, and even disappearance of IR bands closely associated with the OH + O unit may occur, for example, the C-O-C stretch in the diethyl ether (Et 2 O) 2 H + disolvate.…”

“…34 cm -1 ). 16 Band c at ca. ∼1700 cm -1 is always well separated from the other bands, so these designations allow us to tabulate data in the form of Table 4, revealing consistency in the designation of the remaining band a in the range 830-1000 cm -1 .…”

Gas‐Phase Infrared Spectrum of the Protonated Water Dimer: Molecular Dynamics Simulation and Accuracy of the Potential Energy Surface

“…+ in the 500-2000 cm À1 region (in which the modes of the bridging proton are expected) by Asmis et al [1] has raised questions [2,3] about the assignment [4] of the observed bands and stimulated more experimental work. [3,5,6] Most recently, infrared predissociation (IRPD) spectroscopy has been applied to the complex of H 5 O 2 + with Ar, [5,6] and the spectrum obtained shows some differences to that obtained by IRMPD.…”

“…5) The OSS3(p)-Fit [10] of the MP2/cc-pVTZ(aug-O) PES used for the 4D and VCI calculations [2] shows frequency deviations from the direct MP2/cc-pVTZ The harmonic approximation is not adequate for strongly hydrogen bonded systems, and in the light of the approximations on the PES we comment on previous quantum calculations that take anharmonicities into account. The quality of the dynamic description depends on the number of degrees of freedom included, on the choice of coordinates and on the treatment of couplings.…”