Molecular modeling of phase behavior and microstructure of associating fluids, oil reservoir fluids and chemical warfare agents

Prediction of phase equilibria and thermophysical properties of polar fluids and their multicomponent mixtures is a major challenge in the field of molecular simulation. The accuracy of thermophysical properties predictions by molecular simulation depends on the parameters of molecular model used. Molecular models, based on a Lennard-Jones plus point charge functional form, for polar fluids like carboxylic acids, mixtures involving acetone+chloroform, oil reservoir fluids and their mixtures, organophosphates like DMMP and its toxic variant sarin are proposed. Point charges for the different pseudo atoms are determined from ab initio calculations using Gaussian03. We have developed transferable force fields for all these compounds where transferable implies that the force field parameters (LJ well depth and size) for a given interaction site is transferable between different molecules, transferable to different state points and properties. Transferability is demonstrated by performing simulations on homologous groups, carboxylate esters and toxic variants of DMMP like sarin. The Lennard-Jones well depth and size parameters for the different interaction sites were determined by fitting to single-component vapor-liquid equilibrium data, while the remaining Lennard-Jones parameters were taken from the parameters of functionally similar interaction sites used in TraPPE force field (transferable potentials for phase equilibria). Histogram-reweighting Monte Carlo simulations in the grand canonical ensemble were used to determine the vapor-liquid coexistence curves, vapor pressures and critical points as predicted by the new force field for all the chemical compounds mentioned above. The saturated liquid densities, critical properties predictions using the developed force fields for all the fluids are within 1.5% of the experimental values. For chemical warfare agents like sarin where scarce experimental data is available, the boiling point predictions are within 1% of the experimental value.; A major advantage of molecular simulations over equation of state predictions is that it is possible to gain microscopic insights into the polar fluids mentioned above and most of which (carboxylic acids, acetone/chloroform mixtures) like to associate through hydrogen bonding. These simulations help us in predicting structural properties like pair distribution functions, quantitatively describing the aggregation in associating fluids, which are difficult to determine experimentally.