IR and Raman (parallel- and perpendicular-polarized) spectra in the OH stretch region for liquid water were measured some years ago, but their interpretation is still controversial. In part, this is because theoretical calculation of such spectra for a neat liquid presents a formidable challenge due to the coupling between vibrational chromophores and the effects of motional narrowing. Recently we proposed an electronic structure/molecular dynamics method for calculating spectra of dilute HOD in liquid D(2)O, which relied on ab initio calculations on clusters to provide a map from nuclear coordinates of the molecules in the liquid to OH stretch frequencies, transition dipoles, and polarizabilities. Here we extend this approach to the calculation of couplings between chromophores. From the trajectories of the fluctuating local-mode frequencies, transition moments, and couplings, we use our recently developed time-averaging approximation to calculate the line shapes. Our results are in good agreement with experiment for the IR and Raman line shapes, and capture the significant differences among them. Our analysis shows that while the coupling between chromophores is relatively modest, it nevertheless produces delocalization of the vibrational eigenstates over up to 12 chromophores, which has a profound effect on the spectroscopy. In particular, our results demonstrate that the peak in the parallel-polarized Raman spectrum at about 3250 wavenumbers is collective in nature.

We present improvements on our previous approaches for calculating vibrational spectroscopy observables for the OH stretch region of dilute HOD in liquid D2O. These revised approaches are implemented to calculate IR and isotropic Raman spectra, using the SPC/E simulation model, and the results are in good agreement with experiment. We also calculate observables associated with three-pulse IR echoes: the peak shift and 2D-IR spectrum. The agreement with experiment for the former is improved over our previous calculations, but discrepancies between theory and experiment still exist. Using our proposed definition for hydrogen bonding in liquid water, we decompose the distribution of frequencies in the OH stretch region in terms of subensembles of HOD molecules with different local hydrogen-bonding environments. Such a decomposition allows us to make the connection with experiments and calculations on water clusters and more generally to understand the extent of the relationship between transition frequency and local structure in the liquid.vibrational spectroscopy ͉ water W ater is ubiquitous in science and nature (1), so it is natural that a tremendous amount of effort has been expended trying to describe and understand the structure and dynamics of its liquid state. Vibrational spectroscopy, both IR and Raman, provides an excellent probe of the local structure in water, because a local mode's vibrational frequency is exquisitely sensitive to the local mode's molecular environment. Actually, the cleanest information about local structure in water comes from the vibrational spectroscopy not of neat water, but rather of dilute HOD in either H 2 O or D 2 O, because in these cases, respectively, the OD or OH local-mode stretch is almost completely decoupled from the other stretches in the liquid, thus functioning well as a local chromophore. IR and Raman spectra on these systems have been measured by many (2-9).Valuable information about local dynamics in liquid water can also be obtained from vibrational spectroscopy experiments, in this case of the subpicosecond time-domain variety. On this time scale a local mode's vibrational frequency is continually changing because of molecular dynamics. The resulting dynamic frequency fluctuations, also known as spectral diffusion, can be measured by transient vibrational hole-burning and three-pulse echoes (5,6,(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22). In particular, these experiments provide information about both the short-time (stretching) and longtime (making and breaking) aspects of intermolecular hydrogen bonds (23-35).We and others have developed methods for the theoretical calculation of steady-state and ultrafast vibrational spectroscopy observables (7,(23)(24)(25)(26)(27)(36)(37)(38)(39)(40). In our approach, the single vibrational mode of interest, for example, the OH stretch of HOD (when it is immersed in D 2 O), is treated quantum mechanically, whereas all other degrees of freedom (the bath) are treated classically. Thus we are making the adiabatic app...

No abstract

We present theoretical calculations of the vibrational sum-frequency susceptibility for the water liquid/vapor interface. Our approach builds on previous calculations by us and others, using the time-averaging approximation within the mixed quantum/classical formulation for coupled vibrational chromophores, and electric-field maps for transition frequencies, dipoles, polarizabilities, and intramolecular vibrational couplings. We compare our results for the imaginary part of the susceptibility to those from recent experiments, and comment about the effects of intermolecular vibrational coupling and the assignment of features in the spectrum.

We study theoretically the steady-state and ultrafast vibrational spectroscopy, in the OD-stretch region, of dilute HOD in aqueous solutions of sodium bromide. Based on electronic-structure calculations on clusters containing salt ions and water, we develop new spectroscopic maps that enable us to undertake this study. We calculate OD-stretch absorption line shapes as a function of salt concentration, finding good agreement with experiment. We provide molecular-level understandings of the monotonic (as a function of concentration) blueshift, and nonmonotonic line width. We also calculate the frequency time-correlation function, as measured by spectral diffusion experiments. Here again we obtain good agreement with experiment, finding that at the highest salt concentration spectral diffusion slows down by a factor of 3 or 4 (compared to pure water). For longer times than can be accessed experimentally, we find that spectral diffusion is very complicated, with processes occurring on multiple time scales. We argue that from 6 to 40 ps, relaxation involves anionic solvation shell rearrangements. Finally, we consider our findings within the general context of the Hofmeister series, concluding that this series must reflect only local ordering of water molecules.