Water-Anion Hydrogen Bond Interactions and Dynamics from the Bulk to Silica Mesopores

The study of hydrogen bond interactions and dynamics in water is an active and exciting area of research in the Fayer group. We have recently identified a simple anionic vibrational probe, selenocyanate (SeCN), which can serve as a reliable probe of hydrogen bond dynamics in water in 2D IR vibrational echo experiments. Recent work has shown that SeCN is sensitive to local structural fluctuations in its hydrogen bonding configurations with water, which occur on a timescale of ~ 600 fs (~400 fs if water is directly probed). This timescale is obtained from the frequency-frequency correlation function (FFCF) of the CN stretch of SeCN dissolved in D2O in the dilute limit. Ultrafast thermal equilibrium fluctuations in the surrounding water push and pull at the potential energy surface, resulting in a time-dependent vibrational transition frequency, ω(t) = ω + δω(t), where the time-dependence is separated into the instantaneous fluctuation, δω(t), from the time-averaged frequency, ω. The FFCF, given by δω(0)δω(t)〉, follows the timescales over which the ensemble of SeCN oscillators lose memory of their frequencies at time zero due to fluctuations in the structure of the surrounding water. The FFCF of SeCN in D2O has a final 1.4 ps component, which is in quantitative agreement with the corresponding component in the FFCF of water. This component has been extensively studied in molecular dynamics (MD) simulations of water and arises from collective solvent fluctuations that randomize the hydrogen bond network. Therefore, SeCN is a reliable probe, both qualitatively and quantitatively, of the surrounding ultrafast motions of the water hydrogen bond network.

Figure 1. (a) The DFT-calculated fundamental CN stretching frequencies (top), and the absolute value of the transition dipole derivative (bottom) are plotted versus the electric field along the SeCN molecule evaluated at the C atom (filled black circles). In all cases, the empirical map fits are also shown (dashed red lines).

In collaboration with Ward Thompson at the University of Kansas, we have developed the first empirical frequency map for SeCN in D2O (Figure 1). An empirical frequency map provides a means to calculate vibrational frequencies and transition dipole moments from a quantity easily obtained from a classical MD trajectory: in this case an electric field component exerted on SeCN by the solvent. Consequently, an empirical frequency map drastically reduces the resources required to calculate spectroscopic observables compared to full quantum mechanical (QM) simulations. This is especially important for the future simulation and study of nanoscale systems, which are predicted to display hydrogen bond evolution orders of magnitude slower than observed in bulk. Figure 2(b) compares the experimentally measured linear absorption spectrum of SeCN in D2O to that calculated from MD simulations using the empirical frequency map in Figure 1. In Figure 3(b), the CLS (a quantity proportional to the FFCF) is shown for both the 2D IR experiments and the MD simulations with the empirical map approach. The agreement is quite close except for the presence of an additional timescale not observed in the experiments. Improvements to the model are actively being investigated with the goal of accurately simulating/predicting aqueous phase dynamics in spatially complex systems.

Figure 2. (a) Background-subtracted linear IR spectrum of the CN stretch of SeCN in D2O (black curve). A Voigt fit (red curve) to the blue side and peak of the spectrum is extended across the full frequency range, showing that the absorption spectrum is not symmetric. The spectrum has a red side wing that is shown to arise from the non-Condon effect. (b) Simulated IR spectrum (solid blue curve) of the CN stretch of SeCN in D2O compared to the measured spectrum (black curve), the vibrational frequency distribution (dashed red curve), and the spectral density (dashed violet curve).

In the context of Fayer lab research on water, an important, yet difficult, class of experiments is to study how perturbation of the bulk network by the introduction of interfacial structures affects the motions of hydrogen bonds on ultrafast timescales. One model system of interest is water confined in highly ordered porous silica with a narrow unimodal distribution of pores with an average diameter of ~ 2.4 nm. One of the major difficulties preventing the study of these systems is the high degree of scattered light they generate. Scatter signals, which can easily overwhelm the much weaker 3rd order response of the confined liquid, must be eliminated. This is a major experimental challenge that the Fayer group is currently addressing with new phase cycling and data acquisition procedures.

Figure 3. (a) Parallel, isotropic, and perpendicular CLS decays (spectral diffusion) for the CN stretch of SeCN in D2O. The solid curves are biexponential fits to the data. The top and bottom insets display the residuals obtained with single and biexponential fits to the data, respectively. (b) Simulated CLS decay (red solid curve) compared to the measured CLS decay (blue circles) for the CN stretch of SeCN in D2O. The fits are shown as solid curves of the same color.

Relevant Publications

465. "Water-Anion Hydrogen Bonding Dynamics: Ultrafast IR Experiments and Simulations," Steven A. Yamada, Ward H. Thompson, and Michael D. Fayer J. Chem. Phys. 146, 234501 (2017).