Co-Oriented Fluid Functional Equation for Electrostatic interactions (COFFEE) for Mixtures: Phase Equilibria

Abstract

Predicting the phase equilibria of mixtures with thermodynamic equations of state (EOS) based on the perturbation theory approach is often challenging. In particular, when combining pure components with different polar properties, state-dependent parameters are often required to achieve a satisfactory description of the phase behavior. A likely reason for this is that EOS typically do not take the changes in the intermolecular fluid structure that result from the polar interactions into account. Instead it is assumed that the fluid retains the structure of a simpler reference fluid, e.g. a hard sphere fluid. The Co-Oriented Fluid Functional Equation for Electrostatic interactions (COFFEE) was developed to address this challenge. Within the framework of COFFEE the dipolar contribution to the free energy is separated into a near field and a far field contribution. Within the near field, which corresponds roughly to the first coordination shell around a central particle, the free energy is expanded around the fully perturbed state which makes it possible to consider the preferential orientations between neighboring particles directly. For this purpose, the free energy is formulated as a functional of the orientation distribution function (ODF), which describes mutual orientations in a statistical way. The far field is handled as in classic perturbation theories. COFFEE is based on the Stockmayer (ST) model fluid which consists of a Lennard Jones (LJ) fluid with a superimposed point dipole. COFFEE has been used to describe and predict the ODF in pure ST fluids and mixtures containing ST fluids with both central and decentral dipoles. Furthermore, COFFEE can accurately describe the phase behavior of these fluids as well as simple real fluids like hydrogen chloride while showing improvements over comparable approaches. In this contribution, the far field contribution of COFFEE is newly parametrized. This is necessary because the near field contribution was adjusted in previous work. COFFEE is then employed to predict the vapor-liquid equilibria (VLE) of mixtures containing ST, shifted ST (sST, decentral dipole) and LJ fluids. Results are compared to molecular simulation data and show improvements over a comparable EOS in several cases. Finally, the VLE for the three binary mixtures of acetone, acetonitrile, and methanol is calculated and compared to experimental results and results obtained with PCP-SAFT. COFFEE predicts the VLE of mixtures containing methanol more accurately than PCP-SAFT while being slightly less accurate for the acetone-acetonitrile mixture.