Prediction of thermodynamic properties of organic mixtures: Combining molecular simulations with classical thermodynamics

Ashwin Ravichandran, Hla Tun, Rajesh Khare, Chau Chyun Chen

Research output: Contribution to journalArticlepeer-review

Abstract

The binary interaction parameters of the nonrandom two liquid (NRTL) thermodynamic model are predicted for several organic mixtures using molecular simulations. Based on the theoretical framework of the two-fluid theory, the binary interaction parameters are expressed in terms of the interaction energies, size of the molecules, and size of the local molecular domains; these quantities are calculated from molecular simulations. We show that our technique is robust in terms of its predictions involving organic mixtures with compatible chemical characteristics while we propose possible modifications in the case of mixtures involving incompatible chemical components or significant size disparity, where there is a notable difference between the interaction parameters calculated from simulations and those obtained from experimental data regression. We further demonstrate that the binary interaction parameters calculated from data regression are not unique and that molecular simulations can guide the parameter selection process by identifying physically relevant binary interaction parameters. Requiring only the local molecular structure information from molecular simulations, the method offers fast and reliable prediction of phase equilibrium properties, especially in cases where limited experimental data are available.

Original languageEnglish
Article number112759
JournalFluid Phase Equilibria
Volume523
DOIs
StatePublished - Nov 15 2020

Keywords

  • Fluid phase equilibria
  • Local composition models
  • Molecular simulations
  • NRTL activity coefficient model
  • Two-fluid theory

Fingerprint Dive into the research topics of 'Prediction of thermodynamic properties of organic mixtures: Combining molecular simulations with classical thermodynamics'. Together they form a unique fingerprint.

Cite this