The dynamics of water are examined using ultrafast IR stimulated vibrational echo correlation spectroscopy. The OD hydroxyl stretch of HOD in H2O is probed with 45-fs pulses that have sufficient bandwidth (>400 cm-1) to span the entire broad spectrum. High-quality 2D correlation spectra are obtained having the correct phase relations across the broad hydroxyl band. The correlation spectra are found to evolve on multiple time scales. The time evolution of the vibrational echo correlation spectrum reflects the structural evolution of the hydrogen bond networks. The extended vibrational lifetime of the OD hydroxyl stretch of HOD in H2O facilitates the measurement of hydrogen bond dynamics for longer times than possible in previous studies of the OH stretch. Molecular dynamics simulations/electronic structure calculations are used to obtain the time correlation functions (TCF) for two water models, TIP4P and SPC/E. The TCFs are inputs to full time-dependent diagrammatic perturbation theory calculations, which yield theoretical correlation spectra. Quantitative comparison with the data demonstrates that the two water models somewhat overemphasize the fast fluctuations in water and do not contain a slow enough component to account for the slowest fluctuations. Fits to the data using a phenomenological triexponential TCF yield a slowest component of ∼2 ps, and TIP4P and SPC/E have slowest components of <1 ps. The TCF obtained from the water models and the triexponential TCF reproduce the linear absorption line shape equally well, but all miss to some extent the asymmetric “wing” on the low-energy side of the line. Therefore, the time dependence of the vibrational echo correlation spectra provides a good test for the TCF, but the absorption spectrum does not.

No abstract

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.

X-ray and neutron diffractions, vibrational spectroscopy, and x-ray Raman scattering and absorption experiments on water are often interpreted in terms of hydrogen bonding. To this end a number of geometric definitions of hydrogen bonding in water have been developed. While all definitions of hydrogen bonding are to some extent arbitrary, those involving one distance and one angle for a given water dimer are unnecessarily so. In this paper the authors develop a systematic procedure based on two-dimensional potentials of mean force for defining cutoffs for a given pair of distance and angular coordinates. They also develop an electronic structure-based definition of hydrogen bonding in liquid water, related to the electronic occupancy of the antibonding OH orbitals. This definition turns out to be reasonably compatible with one of the distance-angle geometric definitions. These two definitions lead to an estimate of the number of hydrogen bonds per molecule in liquid simple point charge/extended (SPC/E) water of between 3.2 and 3.4. They also used these and other hydrogen-bond definitions to examine the dynamics of local hydrogen-bond number fluctuations, finding an approximate long-time decay constant for SPC/E water of between 0.8 and 0.9 ps, which corresponds to the time scale for local structural relaxation.

Vibrational echo correlation spectroscopy experiments on the OD stretch of dilute HOD in H(2)O are used to probe the structural dynamics of water. A method is demonstrated for combining correlation spectra taken with different infrared pulse bandwidths (pulse durations), making it possible to use data collected from many experiments in which the laser pulse properties are not identical. Accurate measurements of the OD stretch anharmonicity (162 cm(-1)) are presented and used in the data analysis. In addition, the recent accurate determination of the OD vibrational lifetime (1.45 ps) and the time scale for the production of vibrational relaxation induced broken hydrogen bond "photoproducts" ( approximately 2 ps) aid in the data analysis. The data are analyzed using time dependent diagrammatic perturbation theory to obtain the frequency time correlation function (FTCF). The results are an improved FTCF compared to that obtained previously with vibrational echo correlation spectroscopy. The experimental data and the experimentally determined FTCF are compared to calculations that employ a polarizable water model (SPC-FQ) to calculate the FTCF. The SPC-FQ derived FTCF is much closer to the experimental results than previously tested nonpolarizable water models which are also presented for comparison.

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...

We present a new approach that combines electronic structure methods and molecular dynamics simulations to investigate the infrared spectroscopy of condensed phase systems. This approach is applied to the OH stretch band of dilute HOD in liquid D2O and the OD stretch band of dilute HOD in liquid H2O for two commonly employed models of water, TIP4P and SPC/E. Ab initio OH and OD anharmonic transition frequencies are calculated for 100 HOD x (D2O)n and HOD x(H2O)n (n = 4-9) clusters randomly selected from liquid water simulations. A linear empirical relationship between the ab initio frequencies and the component of the electric field from the solvent along the bond of interest is developed. This relationship is used in a molecular dynamics simulation to compute frequency fluctuation time-correlation functions and infrared absorption line shapes. The normalized frequency fluctuation time-correlation functions are in good agreement with the results of previous theoretical approaches. Their long-time decay times are 0.5 ps for the TIP4P model and 0.9 ps for the SPC/E model, both of which appear to be somewhat too fast compared to recent experiments. The calculated line shapes are in good agreement with experiment, and improve upon the results of previous theoretical approaches. The methods presented are simple, and transferable to more complicated systems.

In the context of vibrational spectroscopy in liquids, non-Condon effects refer to the dependence of the vibrational transition dipole moment of a particular molecule on the rotational and translational coordinates of all the molecules in the liquid. For strongly hydrogen-bonded systems, such as liquid water, non-Condon effects are large. That is, the bond dipole derivative of an OH stretch depends strongly on its hydrogen-bonding environment. Previous calculations of nonlinear vibrational spectroscopy in liquids have not included these non-Condon effects. We find that for water, inclusion of these effects is important for an accurate calculation of, for example, homodyned and heterodyned three-pulse echoes. Such echo experiments have been "inverted" to obtain the OH stretch frequency time-correlation function, but by necessity the Condon and other approximations are made in this inversion procedure. Our conclusion is that for water, primarily because of strong non-Condon effects, this inversion may not lead to the correct frequency time-correlation function. Nevertheless, one can still make comparison between theory and experiment by calculating the experimental echo observables themselves.