Experimental MethodsRaman Spectral Measurements: Raman spectra were obtained using a home-built, micro-Raman system similar to that used in previous studies (1). The system used in the present studies includes an Ar-Ion laser source (514.5nm, ~ 50 mW power at the sample) and a thermoelectrically cooled CCD detector (Princeton Instruments Inc., Pixis 400, 1340x400 pixel) mounted to a 300 mm focal length imaging spectrograph (SpectraPro300i, Acton Research Inc.), with a 300 g/mm grating, such that the dispersion is approximately 5 cm -1 per CCD pixel. Liquid samples were analyzed in spectroscopic 1 cm glass cuvettes contained within a thermoelectric, temperature-controlled, translating cell holder (Quantum Northwest). The solution temperature was controlled to within ± 0.01 °C over a 0°C to 60°C temperature range. Such temperature control was required in order to avoid spurious spectral features in the solute-correlated (SC) spectra resulting from small differences in temperature between the solution and pure water samples. A helium lamp was placed behind the sample so that two He lines at 587.562 nm and 667.815 nm were visible in each Raman spectrum (on either side of the OH stretch band). The He lines were used to correct for small wavelength drifts (resulting from small changes in ambient temperature and pressure) which, if uncorrected, would also produce spurious features in the SC spectra within the OH stretch band. The frequency shifts were corrected by introducing a sub-pixel shift the wavelength axis of the solution spectra so as to precisely match the He peak positions in the solution and pure water spectra. The latter shifts were performed using IgorPro (Wavemetrics Inc.) which facilitates duplication of waves with sub-pixel shifts, using a command such as wave1=wave0(p+d), where p is the pixel number (variable) d is a real constant whose magnitude is typically less than 0.1, and wave1 and wave0 are the shifted and un-shifted solution spectra, respectively.Benzene: Spectrophotometry grade benzene (99.93 % assay, EMD Chemicals Inc., Germany) was used without further purification. Water was ultra-purified (Milli-Q UF plus, Millipore Inc.) to an electrical resistance of 18.2 MΩ •cm. Saturated benzene in water was prepared by thoroughly stirring a sample consisting of water with a small excess benzene for 5 min. The solution was then allowed to equilibrate at the desired temperature for at least one day (in equilibrium with the excess benzene phase). Raman spectra were collected from the aqueous phase of the resulting solution (with an excess benzene layer at the top) as well as from a pure water sample (at the same temperature). All samples were equilibrated for at least 5 min in the temperature controlled sample cell holder, and measurements were made in the same cell position. Total (un-polarized) Raman scattering spectra of each solution of benzene in water

No abstract

Raman spectroscopy and multivariate curve resolution (Raman-MCR) are combined to yield a powerful spectroscopic method for identifying solute-induced perturbations of solvent molecules. The principles and applications of the resulting solvation-shell spectroscopy are described and illustrated using both numerical model spectra and experimental Raman spectra, including water in acetone and aqueous OH(-), as well as of both neutral and ionic acetic acid solutions. The results illustrate the quantitative capabilities of Raman-MCR as a solvation-shell spectroscopy, including fundamental limitations arising from "intensity" and "rotational" ambiguities.

Raman spectroscopic measurements are combined with various multivariate curve resolution (Raman-MCR) strategies, to characterize the aggregation of t-butyl alcohol (TBA) in aqueous solutions. The resulting TBA solute-correlated (SC) spectra reveal perturbed water OH features arising from the hydration-shell of TBA as well as shifts in the TBA CH vibrational frequency arising from TBA-TBA interactions. Our results indicate that at low concentrations (below approximately 0.5 M), there is virtually no TBA aggregation. The first aggregates formed above 0.5 M remain highly hydrated, while those formed above approximately 2 M are significantly less hydrated. Comparisons with predictions pertaining to a randomly mixed (non-aggregating) solution indicate that below approximately 1 M there are fewer TBA-TBA contacts than would be present in a random mixture, thus implying that the thermodynamic stability of the first hydration-shell of TBA suppresses the formation of direct contact aggregates at low TBA concentrations. Our results further suggest that microheterogeneous domains containing many water-separated TBA-TBA contacts form near a TBA concentration of approximately 1 M, while at higher concentrations the TBA-rich domain size distribution may resemble that in a non-aggregating random mixture.

Interactions between halide ions (fluoride and iodide) and t-butyl alcohol (TBA) dissolved in water are probed using a recently developed hydration-shell spectroscopic technique and theoretical cluster and liquid calculations. High ignal-to-noise Raman spectroscopic measurements are combined with multivariate curve resolution (Raman-MCR) to reveal that while there is little interaction between aqueous fluoride ions and TBA, iodide ions break down the tetrahedral hydration-shell structure of TBA and produce a red-shift in its CH stretch frequency, in good agreement with the theoretical effective fragment potential (EFP) molecular dynamics simulations and hybrid quantum/EFP frequency calculations. The results imply that there is a significantly larger probability of finding iodide than fluoride in the first hydration shell of TBA, although the local iodide concentration is apparently not as high as in the surrounding bulk aqueous NaI solution.