New models for tracer diffusion coefficients of hard sphere and real systems: Application to gases, liquids and supercritical fluids

Do Electrostatics Control the Diffusive Dynamics of Solitary Water? NMR and MD Studies of Water Translation and Rotation in Dipolar and Ionic Solvents

NMR-based measurements of the diffusion coefficients and rotation times of solitary water and benzene at 300 K are reported in a diverse collection of 13 conventional organic solvents and 10 imidazolium ionic liquids. Proton chemical shifts of water are found to be correlated to water OH-stretching frequencies, confirming the importance of electrostatic interactions in these shifts. However, the influence of magnetic interactions in aromatic solvents renders chemical shifts a less reliable indicator of electrostatics. Diffusion coefficients (DB) and rotational correlation times (τB) of benzene in the solvents examined are accurately described as functions of viscosity (η) by DB ∝ η-0.81 and τB ∝ η0.64. Literature values of DB and τB in alkane and normal alcohols, which were not included among the solvents studied here, are systematically faster than predicted by these correlations, indicating that factors beyond solvent viscosity play a role in determining the friction on benzene. In contrast to benzene, water diffusion and rotation are poorly described in terms of viscosity alone, even in the dipolar and ionic solvents measured here. The present data and the substantial literature data already available on dilute water diffusion show a systematic dependence of DW on solvent polarity among isoviscous solvents. The aspect of solvent polarity most relevant to water dynamics is the ability of a solvent to accept hydrogen bonds from water, as conveniently quantified by the frequency of water's OH stretching band, ΔνOH. The friction on translation, ζtr = kBT/DW, and rotation, ζrot = kBTτW, are both well correlated by functions of the form ζ(η, ΔνOH) = a1ηa2 exp (a3ΔνOH), where the ai are adjustable parameters. Molecular dynamics simulations reveal a strong coupling between electrostatic and nonelectrostatic water-solvent interactions, which makes it impossible to dissect the friction on water into additive dielectric and hydrodynamic components. Simulations also provide a tentative explanation for the unusual form of the correlating function ζ(η, ΔνOH), at least in the case of ζrot.

Modeling Tracer Diffusion Coefficients of Any Type of Solutes in Polar and Non-Polar Dense Solvents

In this work, a simple two-parameters correlation based on the Rice and Gray, Lennard-Jones, and Stockmayer theories was devised for the calculation of binary diffusion coefficients (D12) of any type of solutes at infinite dilution in polar and non-polar solvents. This equation can be relevant for systems with polar solvents, since most models in the literature fail when strong intermolecular forces predominate in solution. The new correlation embodies the Stockmayer potential without requiring the dipole moments of any component, which significantly enlarges its application. It was validated with the largest D12 database of polar and non-polar dense systems, with 8812 data points (NDP) spanning 553 systems, of which 133 have water as solvent (NDP = 1266), 89 contain polar solvents excluding water (NDP = 1405), 177 have supercritical carbon dioxide (SC-CO2) as solvent (NDP = 5028), and 154 have non-polar or weakly polar solvents excluding SC-CO2 (NDP = 1113). Overall, the model achieved an average deviation of only 3.43%, with accurate and unbiased behavior even for polar systems.

Universal model for accurate calculation of tracer diffusion coefficients in gas, liquid and supercritical systems

In this work it is presented a new model for accurate calculation of binary diffusivities (D12) of solutes infinitely diluted in gas, liquid and supercritical solvents. It is based on a Lennard-Jones (LJ) model, and contains two parameters: the molecular diameter of the solvent and a diffusion activation energy. The model is universal since it is applicable to polar, weakly polar, and non-polar solutes and/or solvents, over wide ranges of temperature and density. Its validation was accomplished with the largest database ever compiled, namely 487 systems with 8293 points totally, covering polar (180 systems/2335 points) and non-polar or weakly polar (307 systems/5958 points) mixtures, for which the average errors were 2.65% and 2.97%, respectively. With regard to the physical states of the systems, the average deviations achieved were 1.56% for gaseous (73 systems/1036 points), 2.90% for supercritical (173 systems/4398 points), and 2.92% for liquid (241 systems/2859 points). Furthermore, the model exhibited excellent prediction ability. Ten expressions from the literature were adopted for comparison, but provided worse results or were not applicable to polar systems. A spreadsheet for D12 calculation is provided online for users in Supplementary Data.