Low frequency Δν̄=0–350 cm−1, Raman intensity data were obtained from liquid water between 3.5 and 89.3 °C using holographic grating double and triple monochromators. The spectra were Bose–Einstein (BE) corrected, I/(1+n), and the total integrated (absolute) contour intensities were treated by an elaboration of the Young–Westerdahl (YW) thermodynamic method, assuming conservation of hydrogen-bonded (HB) and nonhydrogen-bonded (NHB=bent and/or stretched, O–H O) nearest-neighbor O–O pairs. A ΔH°1 value of 2.6±0.1 kcal/mol O–H ⋅⋅⋅ O or 5.2±0.2 kcal/mol H2O (11 kJ/mol O–H ⋅⋅⋅ O, or 22 kJ/mol H2O) resulted for the HB→NHB process. This intermolecular value agrees quantitatively with Raman and infrared ΔH° values from the one- and two-phonon OH-stretching regions, and from molecular dynamics, depolarized light scattering, neutron scattering, and ultrasonic absorption, thus indicating a common process. A population involving partial covalency of, i.e., charge transfer into, the H ⋅⋅⋅ O units of linear and/or weakly bent hydrogen bonds, O–H ⋅⋅⋅ O; is transformed into a second high energy population involving bent, e.g., 150° or less, and/or stretched, e.g., 3.2 Å, but otherwise strongly cohesive O–H O interactions. All difference spectra from the fundamental OH-stretching contours cross at the X(Z,X+Z)Y isobestic frequency of 3425 cm−1. Also, total integrated Raman intensity decreases occurring below 3425 cm−1 with temperature rise were found to be proportional to the total integrated intensity increases above 3425 cm−1, indicating conservation among the HB and NHB OH-stretching classes. From the enthalpy of vaporization of water at 0 °C, and the ΔH°1 of 2.6 kcal/mol O–H ⋅⋅⋅ O, the additional enthalpy, ΔH°2, needed for the complete separation of the NHB O–O nearest neighbors is ∼3.2 kcal/mol O–H ⋅⋅⋅ O or ∼6.4 kcal/mol H2O (13 kJ/mol O–H ⋅⋅⋅ O or 27 kJ/mol H2O). The NHB O–O nearest neighbors are held by forces other than those involving H ⋅⋅⋅ O partial covalency, i.e., electrostatic (multipole), induction, and dispersion forces. The NHB O–O pairs do not appear to produce significant intermolecular Raman intensity because they lack H ⋅⋅⋅O bond polarizability, but the corresponding NHB OH oscillators do contribute weakened Raman intensity above 3425 cm−1. An ideal solution thermodynamic treatment employing ΔH°1 =2.6 kcal/mol O–H ⋅⋅⋅ O, the HB mole fraction, and the vapor heat capacity, yielded a very satisfactory specific heat value of 1.1 cal deg−1 g−1 H2O at 0 °C. The NHB mole fraction, fu, from the YW treatment is negligibly small, 0.06 or less, for t<−50 °C. However, fu increases to 0.16 at 0 °C, and fu≊1 at 1437 °C, where recent shock-wave Raman measurements indicate loss of all partially covalent, charge transfer hydrogen bonding.

Raman frequencies and polarizations and infrared frequencies of water and heavy water have been obtained, and the intermolecular vibrations of water have been related to a five-molecule hydrogen-bonded C2v model consistent with x-ray data. Observed variations of the integrated Raman intensity of the 175-cm—1 hydrogen-bond-stretching vibration, with variations of temperature, have been interpreted in terms of the five-molecule model. That interpretation leads to reasonable values for the enthalpy of hydrogen-bond formation. Effects of electrolyte addition on the intensity of the 175-cm—1 band are also described.

Integrated Raman intensities of the spectral contour arising from the intermolecular librational motions of pure water have been obtained in the temperature range of ∼10°—95°C. In addition, integrated intensities of nearly symmetric librational components centered near ∼475 and ∼710 cm−1 were obtained from manual contour analysis according to two components. However, contour analysis was also accomplished by means of a special-purpose analog computer, and three Gaussian librational components having average frequencies of 439, 538, and 717 cm−1 were thus revealed. The total contour intensity, the manually determined component intensities, and the Gaussian component intensities were found to have the same temperature dependence, and that dependence was found to be in excellent quantitative agreement with the previously reported temperature dependence of the hydrogen-bond-stretching intensity [J. Chem. Phys. 44, 1546 (1966)]. Integrated Raman intensities of pure water were also obtained in the temperature range of 10°—90°C for the intramolecular valence and deformation contours in the spectral region of ∼2800–3900 cm−1, and near 1645 cm−1, respectively. The integrated intensity of the deformation contour was found to be nearly independent of temperature, but the total integrated intensity of the intramolecular valence contour was found to decrease with increasing temperature. However, heights of the high-frequency portion of the intramolecular valence contour were observed to increase, whereas heights of the low-frequency portion were observed to decrease at nearly the same rate, with increasing temperature. An isosbestic point was also found at approximately 3460 cm−1. Further, computer analysis revealed the existence of four Gaussian components having opposite temperature dependences in pairs—two intense valence components at ∼3247 and ∼3435 cm−1 were found to decrease in intensity with increasing temperature, and two weak components at ∼3535 and ∼3622 cm−1 were found to increase in intensity. Computer analysis of infrared absorbance spectra also revealed four Gaussian components at approximately 3240, 3435, 3540, and 3620 cm−1. The quantitative agreements involving temperature dependences of the intermolecular hydrogen-bond-stretching and librational intensities, as well as the intramolecular valence data, would appear to preclude models of water structure involving consecutive hydrogen-bond breakage. Continuum models of water structure are also precluded by the inter- and intramolecular intensity dependences, and particularly by the isosbestic point in the intramolecular valence region, but a model involving an equilibrium between two forms of water is consistent with all of the data. The two forms refer to water molecules which have or have not surmounted a barrier arising from a partially covalent hydrogen-bond potential of C2v symmetry, and they may be described as nonhydrogen-bonded monomeric water, and as lattice water, respectively. Polarized argon-ion-laser—Raman spectra were also obtained in the intermolecular frequency region of the water spectrum, and the depolarization ratios of the intermolecular Raman bands were found to be in complete agreement with predictions from intermolecular C2v symmetry. Studies of the intramolecular valence region were also made with polarized mercury excitation, and the spectra were analyzed by the analog method. Short-lived CS intramolecular perturbations were indicated by the observed depolarization ratios of the four Gaussian valence components. Accordingly, CS intramolecular valence perturbations occur in the lattice water, as well as in the nonhydrogen-bonded water, but the perturbations are of little importance on the intermolecular time scale.

We present the complete intermolecular dynamical spectrum of liquid water, by merging the data sets from femtosecond nonlinear-optical polarization spectroscopy with the depolarized, Bose–Einstein corrected Raman spectrum to cover the frequency range from 0–1200 cm−1. The impulse response function for liquid water at room temperature is calculated, including all of the intermolecular motions.

Precise isosbestic points occur in the Raman OH-stretching spectra from liquid water between 3 and 85 °C if cell alignment is accomplished with Newton’s rings. Isosbestic frequencies measured for the orientations X(Y,X+Z)Y=6β2, X(ZX)Y=3β2, X(Y+Z,X+Z)Y=45α2+13β2, X(Z,X+Z)Y=45α2+7β2, and X(ZZ)Y=45α2+4β2 are 3524, 3522 (note β2 agreement), 3468, 3425, and 3403 cm−1, respectively. Isosbestic points from two different measurements calculated by the relations, X(ZZ)Y-(4/3)X(ZX)Y and X(Z,X+Z)Y-(7/6)X(Y,X+Z)Y agree exactly for 45α2, 3370 cm−1. (α and β2 correspond to the mean polarizability and square of the anisotropy.) The pure α2 isosbestic frequency, 3370 cm−1, coincides with the peak of the highest frequency hydrogen-bonded (HB) Gaussian OH-stretching component. The pure β2 isosbestic point, 3522–3524 cm−1, coincides with the peak of the nonhydrogen-bonded (NHB) Gaussian OH-stretching component, next above in frequency. The α2 and β2 isosbestic points are thus thought to provide an experimental distinction between, and a clear definition of, the HB and NHB OH-oscillator classes for water. Moreover, the various OH-stretching combinations of α2 and β2 simply provide different measures of the HB→NHB equilibrium—no special information concerning the temperature dependence of this equilibrium results from use of any one linear polarizability combination over any other, including pure α2 or pure β2. The present results agree with mercury-excited data [Walrafen, J. Chem. Phys. 47, 114 (1967)] for X(Y+Z,X+Z)Y and with the corrected α2 data of d’Arrigo et al. [J. Chem. Phys. 75, 4264 (1981)]. Furthermore, the new data are in accord with the spectroscopic mixture model, but the continuum model conflicts with the observation of exact points. The isosbestic frequencies are also found to be strongly nonlinear in the amount of α2 or β2 involved in the spectra.

No abstract

Structural correlation lengths (SCL) obtained from radial distribution functions were found to be linear in the integrated Raman intensity ratios of the correlated-to uncorrelated-proton OH stretching bands from -163 "C, low-density amorphous (LDA) ice, through supercooled and ambient water, and up to 401 "C and 1730 bar, supercritical water. This linearity, from -163 to =400 "C, arises because the proton motions are correlated over hydrogen-bonded distances which are determined by the SCL. Moreover, the longest acoustic phonon wavelength, namely, that of the TA or shear mode, which occurs near ~5 . 5 cm-',l.* is also completely limited by the SCL. Hence, the proton (OH stretching) and center-of-mass (TA phonon) correlations both arise from, and are restricted in distance by, the size of the hydrogen-bonded aggregates or patches. The inclusion of LDA ice in the linear relationship also suggests a possible continuity of states from amorphous ice to supercooled water, whereas the high-temperature, supercritical limit is attained when most of the hydrogen bonding is disrupted.

Hydration forces are thought to result from the energetic cost of water rearrangement near macromolecular surfaces. Raman spectra, collected on the same collagen samples on which these forces were measured, reveal a continuous change in water hydrogen-bonding structure as a function of separation between collagen triple helices. The varying spectral parameters track the force-distance curve. The energetic cost of water ''restructuring,'' estimated from the spectra, is consistent with the measured energy of intermolecular interaction. These correlations support the idea that the change in water structure underlies the exponentially varying forces seen in this system at least over the 13-18-Å range of interaxial separations.The structure of water surrounding collagen has been studied by a wide variety of techniques and is considered a paradigm for protein hydration. It is commonly accepted that at least some water in collagen fibers differs from bulk, ambient water. Within the most simple scheme, this interstitial water is separated into two distinct classes-tightly bound and ''free'' or bulk-like (1, 2). More elaborate models imagine three or even more fractions (3-8). Tightly bound waters are believed to stabilize the triple helix by participating in the H-bond backbone (see ref. 9 for a review). In a larger context, however, the specific role of water in collagen assembly, structure and stability is still an open question.New insights into the structure and the role of water near collagen come from the high-resolution x-ray structure of a collagen-like peptide (10, 11). Malleable clusters and chains of H-bonded waters form a complex network connected to polar side chains and to backbone carbonyls on the peptide triple helices. The uncharged helices do not show any direct contacts; water directly mediates the interaction between them.This interpretation of the x-ray results correlates well with the picture that has emerged from direct measurement of force-distance curves between triple helices. It has been argued that the measured short-range repulsion, which prevents collagen molecules from coming too close to each other, originates from the energetic cost of removal of the interstitial water (12-14). The longer-range attraction, which prevents the molecules from coming too far apart and which induces self-assembly of collagen fibers from solution, appears to be associated with formation of hydrogen bonded water bridges at specific recognition sites (13). The arguments, however, have been based so far on the process of elimination of other explanations for the observed forces, rather than on direct observation of changes in water structure.The present study establishes more explicit connection between forces and water structure by observing Raman spectra from the same samples on which osmotic stress force measurements have been made. We analyze the changes in the OOH and NOH vibrational bands measured in the 3,100-3,800-cm Ϫ1 range as a function of separation between collagen triple helices and correlat...