(121a) Prediction of Liquid-Liquid Equilibria from Gibbs Ensemble Monte Carlo Simulations

While physics-based computer simulations such as molecule dynamics or Monte Carlo are widely used for the prediction of physical properties and microscopic structure, there remain numerous systems and state points where the nature of interactions between molecules can cause naïve approaches to become trapped in meta-stable states, leading to incorrect predictions. In some cases, the results are obviously incorrect, while in others, sampling limitations may be more subtle, i.e. the results look reasonable, but are instead the result of sampling an incorrect distribution of states.

In this talk, we discuss an extension of the Molecular Exchange Monte Carlo (MEMC) algorithm to improve its effectiveness when used to exchange molecules between two dense phases [1,2]. To highlight the effectiveness of the new MEMC algorithms, Gibbs ensemble Monte Carlo simulations are used to predict liquid-liquid equilibria for alkane-perfluoroalkane mixtures (for molecules with 1-8 carbon atoms) and dimethyl ether-water. The process of developing and validating configurational-bias moves is discussed. These algorithms are implemented and available to the research community in the development branch of GOMC (https://github.com/GOMC-WSU/GOMC).

References:

[1] Mohammad Soroush Barhaghi, Korosh Torabi, Younes Nejahi, Loren Schwiebert and Jeffrey J. Potoff, J. Chem. Phys, 2018, 149, 072318.

[2] Mohammad Soroush Barhaghi and Jeffrey J. Potoff, Fluid Phase Equil, 2019, 486, 106.