Temperature-dependence measurements have been made on the chemical shift of the proton of a water molecule in the liquid state and in the gas state at varying pressure. The problem of relating these experimental data to the intermolecular forces leading to cohesion and to hydrogen-bond formation between water molecules is considered in detail. It is shown that a consistent treatment of the chemical shift, thermal, and dielectric data for water can be given based on a two-state model involving an equilibrium between a hydrogen-bonded ``icelike'' fraction and a ``monomer'' fraction whose interaction with the lattice arises entirely from London dispersion forces. Using semiempirically derived values of the chemical shift and energy associated with the condensation of water vapor to ``monomer,'' the magnitude of the shift associated with the transformation to ice is calculated. It is then shown that, on the assumption that the hydrogen bond is electrostatic in character, the ``polar'' contribution to this shift can be related through the appropriate shielding equations to the dipole moment of the water molecule in ice. The magnitude of the dipole moment derived from these relationships is found to be in excellent agreement with values derived from dielectric data. The possibility that the shielding changes may in part be due to processes other than the breaking of hydrogen bonds is considered. It is shown that the model leads to the conclusion that the chemical shift in the transformation of ice to water at 0°C could be entirely accounted for either by a stretching of the hydrogen bonds or a small amount of bending of the bonds. It is noted that if some bond breaking does occur, as required by the fact that water is a liquid, then the amount of stretching and/or bending will be limited.

No abstract

Experimental data on the effect of varying concentrations of salts of univalent diamagnetic ions on the proton resonance of water are given. A reference scale for the division of the infinite dilution shifts for the salts into contributions for the separate ions is discussed. Factors entering into the separate ion contributions are considered and a model is developed from which an attempt has been made to evaluate in a quantitative manner the magnitude of these factors. ``Effective'' hydration numbers are calculated from this model, which treats as ion-water complex as a molecular species. These hydration numbers for cations are very similar to estimates of ``primary'' hydration obtained from other sources. A decrease in effective hydration number with increasing cation radius is found. It is suggested from th data that of the halide ions, only the fluoride ion forms a hydrate in the chemical sense, the larger halide ions acting primarily to break down the water structure, the effect increasing with increasing anionic radius. These structural effects are expressed in terms of equivalent numbers of hydrogen-bonds made or broken. A structure-making effect is suggested for the lithium ion. Factors to be considered in the extension of the equations to the interpretation of concentration effects are discussed and evidence presented for different kinds of environmental changes occurringe in solutions of different salts.

An experimental study of the proton spin-lattice relaxation time T1 has been made for H2O over the temperature interval from −16 to 145°C. It is found that after correction for spin-rotational interaction, the experimental T1 behavior can be represented by the double exponential form of the rate expression used to treat the relaxation for the quadrupolar nuclei 2H and 17O in water. The oxygen-17 data are used to calculate the intramolecular contribution to the proton T1. The activation energies E1=9.4± 0.8 kcal/mole and E2=3.6± 0.1 kcal/mole for the two contributions to the intermolecular relaxation are in sufficiently good agreement with those for the intramolecular relaxation to indicate that the relaxation mechanism is the same in both cases. This mechanism involves two processes. The data indicate the process dominant at high temperature can be described as a rotational diffusion where the amplitude of angular motion increases with increasing temperature.

Experimental measurements of T1 for oxygen-17 water in n-heptane emissions were made in the supercooled region down to −31°C. The data were fit to the double exponential form of the rate equation with activation energies for the two processes of E1=11.85±0.57 and E2=3.54±0.08 kcal mole−1, respectively. Using transition state rate theory, activation entropies ΔS1*=35.3±2.3 and ΔS2*=4.63±0.25 cal deg−1 · mole−1 were calculated. It is proposed that both the low and high temperature reactions are kinetic processes. It is suggested that the large entropy term for the low temperature reaction indicates that the relaxation process involves the ``cooperative'' dissolution of a small cluster of hydrogen bonded molecules. An examination of the available data for viscosity and diffusion indicates that the high temperature relaxation process is the same as that for the rotational (T1) relaxation. It is suggested that this process involves the breaking of a single hydrogen bond. The question of whether the low temperature processes for viscous flow, diffusion, and rotational relaxation are related is discussed.