(574a) Predicting Vapor-Liquid Equilibria with Augmented Ab Initio Potentials

Understanding the vapor-liquid equilibria of fluids is important for many key chemical engineering processes [1]. The topic is also challenging from a scientific perspective because it involves simultaneously predicting two diverse phases of matter. Historically, the prediction of vapor-liquid equilibrium has been addressed by proposing a suitable equation of state [2]. However, as the governing mechanism for phase separation is directly related to intermolecular interactions, the use of molecular simulations [3] that involve a suitable intermolecular potential can be particularly insightful. Simple intermolecular potentials such as the Lennard-Jones model have been used widely, but considerable progress has been [4,5] achieved in developing accurate potentials from first principles. Although these ab initio potentials are currently restricted to simple atomic systems [4] or small polyatomic molecules such as water [6], greater use of such models is likely in the future. In this work, we examine the accuracy of some recent models to predict vapor-liquid equilibria. An advantage of this approach is that it can be systematically refined to improve the quality of predictions. We illustrate this approach by combining two-body ab intio potentials with a computationally simple model of three-body interactions. Comparison with experimental data indicates that the quality of predictions is improved considerably.

1. R.J. Sadus, High Pressure Phase Behaviour of Multicomponent Fluid Mixtures, Elsevier, Amsterdam, 1992.
2. Y.S. Wei and R. J. Sadus, AIChE J. 46, 169 (2000).
3. R.J. Sadus, Molecular Simulation of Fluids: Theory, Algorithms and Object-Orientation, Elsevier, Amsterdam, 1999.
4. B. Jaeger, R. Hellmann and E. Vogel, J. Chem. Phys. 144, 114304 (2016).
5. M. Vlasiuk, F. Frascoli and R.J. Sadus, J. Chem. Phys. 145, 104501 (2016).
6. R. Bukowski, K. Szalewicz, G. C. Groenenbnoom and A. van der Avoird, Science 315, 1249 (2007).