Assessing the ability of force-fields to predict liquid–liquid equilibria of ternary systems of light alcohols+water+dodecane by Monte Carlo simulation

A Monte Carlo simulation study of the liquid–liquid equilibria for binary dodecane/ethanol and ternary dodecane/ethanol/water mixtures

“…In the current work, volume fluctuations (through exchanges with the pressure bath) are only sampled for the two simulation boxes containing the liquid phases, whereas the volume for the third box is fixed and the non-bonded interactions are set to zero for this phase, i.e., representing an ideal gas. In the simulations by Lasich et al, [30] a two-box GEMC set-up was used.…”

“…To improve the statistics, eight independent simulations are carried out for each state point. For comparison, the equilibration periods used by Lasich et al [30] were about one order of magnitude shorter and only a single trajectory was used for each state point.…”

“…For example, these authors observed two liquid phases for the n-dodecane/ethanol mixture at T = 400 K and p = 0.6 MPa, [30] whereas experimental measurements yield an upper critical solution temperature of T = 285.6 ± 0.5 K at p = 0.1 MPa. [35] Somewhat surprisingly, Lasich et al [30] found larger disagreement for higher alcohol carbon numbers despite the fact that the TraPPE force field represents the alkyl region of the alcohols with the same Lennard-Jones parameters as used for the alkanes. Here, we re-investigate the n-dodecane/ethanol mixture to assess the extent to which this disagreement is caused by deficiencies of the force fields and by sampling problems that may be encountered in simulations for LLE.…”

“…[26,27,28,29] In addition to the sampling difficultities encountered in simulations of LLE, the relatively small transfer free energies governing mutual solubilities in these systems pose a severe test for the accuracy of the force fields used to describe the intermolecular interactions. In a recent study, Lasich et al [30] investigated water/n-dodecane/light-alcohol mixtures using Gibbs M a n u s c r i p t ensemble Monte Carlo (GEMC) simulations [31,32] and the transferable potentials for phase equilibria (TraPPE) force field [33,34] and reported severe over-predictions of the miscibility gap. For example, these authors observed two liquid phases for the n-dodecane/ethanol mixture at T = 400 K and p = 0.6 MPa, [30] whereas experimental measurements yield an upper critical solution temperature of T = 285.6 ± 0.5 K at p = 0.1 MPa.…”

“…In a recent study, Lasich et al [30] investigated water/n-dodecane/light-alcohol mixtures using Gibbs M a n u s c r i p t ensemble Monte Carlo (GEMC) simulations [31,32] and the transferable potentials for phase equilibria (TraPPE) force field [33,34] and reported severe over-predictions of the miscibility gap. For example, these authors observed two liquid phases for the n-dodecane/ethanol mixture at T = 400 K and p = 0.6 MPa, [30] whereas experimental measurements yield an upper critical solution temperature of T = 285.6 ± 0.5 K at p = 0.1 MPa. [35] Somewhat surprisingly, Lasich et al [30] found larger disagreement for higher alcohol carbon numbers despite the fact that the TraPPE force field represents the alkyl region of the alcohols with the same Lennard-Jones parameters as used for the alkanes.…”