Exploring solvent properties of high pressure carbon dioxide via computer simulation

There has been significant interest in the use of liquid and supercritical carbon dioxide (CO2) as an environmentally responsible solvent. Although CO2 has properties that make it a suitable solvent, many organic solutes are only sparingly soluble in CO2. We use Monte Carlo and molecular dynamics computer simulations of CO2 to explore solvation. We characterize the structure of CO2 using cavity size distributions. These distributions are compared to water. The distributions show that lowering density increases average cavity size, but lowering temperature has minimal affect. Water and CO2 at their triple point have similar cavity sizes. We calculate the infinite dilution Henry’s Law solubility of 31 small organic molecules using the Expanded Ensemble method. The solubility is measured at 12 liquid, near-critical, and supercritical conditions. Results show that increasing the number of fluorine atoms or methyl group substitutions for hydrogen atoms on a methane molecule increases solubility, despite the larger molecular size. The solubility data correlate well to a functional form with mean field attractive terms and scaled particle theory repulsive terms. This correlation is based on the temperature and density of CO2 and the intrinsic properties of the solute molecule. The simulations provide a solubility ranking for the molecules, and allow for calculation of partition coefficients between CO2 and water. We explore changes in freely jointed Lennard-Jones polymer chain conformation with changes in CO2 density and temperature. We find that higher temperature or solvent density increases chain dimensions, although there is a limit to chain extension from increasing density. We find that increasing density depresses the temperature where the chain transforms from a coil to a globule. We also examine the effects of changing the model parameters of the polymer.