Vibrational spectroscopy of water at interfaces.

Understanding liquid water's behavior at the molecular level is essential to progress in fields as disparate as biology and atmospheric sciences. Moreover, the properties of water in bulk and water at interfaces can be very different, making the study of the hydrogen-bonding networks therein very important. With recent experimental advances in vibrational spectroscopy, such as ultrafast pulses and heterodyne detection, it is now possible to probe the structure and dynamics of bulk and interfacial water in unprecedented detail. We consider here three aqueous interfaces: the water liquid-vapor interface, the interface between water and the surfactant headgroups of reverse micelles, and the interface between water and the lipid headgroups of aligned multi-bilayers. In the first case, sum-frequency spectroscopy is used to probe the interface. In the second and third cases, the confined water pools are sufficiently small that techniques of bulk spectroscopy (such as FTIR, pump-probe, two-dimensional IR, and the like) can be used to probe the interfacial water. In this Account, we discuss our attempts to model these three systems and interpret the existing experiments. For the water liquid-vapor interface, we find that three-body interactions are essential for reproducing the experimental sum-frequency spectrum, and presumably for the structure of the interface as well. The observed spectrum is interpreted as arising from overlapping and canceling positive and negative contributions from molecules in different hydrogen-bonding environments. For the reverse micelles, our theoretical models confirm that the experimentally observed blue shift of the water OD stretch (for dilute HOD in H(2)O) arises from weaker hydrogen bonding to sulfonate oxygens. We interpret the observed slow-down in water rotational dynamics as arising from curvature-induced frustration. For the water confined between lipid bilayers, our theoretical models confirm that the experimentally observed red shift of the water OD stretch arises from stronger hydrogen bonding to phosphate oxygens. We develop a model for heterogeneous vibrational lifetime distributions, and we implement the model to calculate isotropic and anisotropic pump-probe decays. We then compare these results with experimental data. Clearly, recent experimental advances in vibrational spectroscopy have led to beautiful new results, providing information about the structure and dynamics of water at interfaces. These experimental and concomitant theoretical advances (particularly the unified theoretical framework of non-linear response functions) have greatly contributed to our understanding of this unique and important substance.