A novel multiscale approach for rapid prediction of phase behaviors with consideration of molecular conformations

Abstract

The conformational distribution of flexible molecules may have a significant influence on its physical properties and phase behaviors. While atomistic molecular simulations naturally include conformational changes, they are often computationally expensive and require separate simulations for each state condition. Conversely, an equation of state (EOS) provides a rapid description for a variety of fluid properties over the whole phase space; however, because of the ignorance of molecular structure, it is difficult to obtain the interaction parameters from first‐principles. Here we propose a multiscale approach for simultaneous predictions of fluid properties and molecular conformational distribution. The PR+COSMOSAC EOS is used to provide the fugacity of chemical species in a mixture. The electrostatic and dispersion interactions in the EOS are determined from quantum‐mechanical solvation calculations and molecular dynamics simulations at a few state points. The conformational transitions are considered as chemical reactions with the equilibrium constant determined from ab initio G4 calculations. With all the EOS parameters obtained from molecular simulations at different scales, subsequent phase equilibrium predictions can be achieved within milliseconds. We validated this approach using 1,2‐dichloroethane (DCE), whose dipole moment varies from 6.34 × 10−30 C·m in the gauche‐form to 0 in the trans‐form. Our results show that this approach provides not only accurate predictions for the vapor–liquid and liquid–liquid equilibrium of DCE‐containing fluids but also quantitative descriptions of conformational distribution of DCE in these systems. This novel method can be very useful for the prediction of thermodynamic properties of fluids with explicit inclusion of molecular conformations. © 2016 American Institute of Chemical Engineers AIChE J, 62: 4047–4054, 2016

Comprehensive Assessment of COSMO-SAC Models for Predictions of Fluid-Phase Equilibria

“…Third, the dispersive interactions may be obtained from molecular dynamics or Monte Carlo simulations with suitable force fields to describe the interactions. Our recent study showed that the dispersive term can be directly obtained from the van der Waals component of the solvation free energy derived from thermodynamic integration …”

“…Our recent study showed that the dispersive term can be directly obtained from the van der Waals component of the solvation free energy derived from thermodynamic integration. 41 This assessment indicates that model efficiency strongly correlates with the type of chemical family. No significant correlation between model accuracy and polarity as well as between error and molecular size asymmetry was found.…”

Prediction of pH Effect on the Octanol–Water Partition Coefficient of Ionizable Pharmaceuticals

“…While the K OW,N value varies with the conformation of the drug molecules, we found that, in most cases, the lowest-energy conformation results in better agreement with the experiment. Although more sophisticated treatment of the conformation is possible (i.e., to include all significant conformations , ), we consider the results based on the lowest-energy conformation to be sufficient for the purposes of the present study.…”

Improved Prediction of Phase Behaviors of Ionic Liquid Solutions with the Consideration of Directional Hydrogen Bonding Interactions

“…For example, COSMO-based models like COSMO-RS and COSMO-SAC are the outstanding examples of such models. These models have a sound theoretical basis and have been shown to provide reasonably accurate predictions for a wide varieties of fluid mixtures, including organic fluids, − pharmaceuticals, − gas molecules, , supercritical fluids, , electrolytes, − and ionic liquids (ILs). − …”