The interfacial properties of water/oil mixtures is a topic of significant interest for the oil and gas and chemical industry, as they are required for performing process calculations. However, the reported data at high temperatures and pressures are scarce. The present study focuses on simulating the interfacial tension (IFT) of mixtures of water with i) toluene (water/toluene), ii) n-dodecane (water/n-dodecane) and iii) a 50:50 % wt toluene:n-dodecane mixture (water/toluene/n -dodecane), at 1.83 MPa and temperatures ranging from 383.15 to 443.15 K. Molecular Dynamics (MD) simulations with atomistic molecular models, developed primarily for biomolecular systems were employed. In the simulations performed, the effects of the water model and the scaling of interatomic interactions by introducing a binary interaction parameter, kijkij, were assessed for the accurate reproduction of experimental data. The combination of the TIP4P/2005, SPC/E and TIP3P water models with the GAFF and Lipid14 force fields for toluene and n -dodecane respectively, coupled with appropriate binary interaction parameters, kijkij, for the interaction of carbon with oxygen atoms were found to yield accurate results in the case of binary mixtures. For the water/toluene/n -dodecane mixture, all force field combinations tested resulted in overestimated IFT values for the whole range of state points examined. Our simulations show that these widely used force fields, originating from the world of biomolecular simulations, are suitable candidates in the study of binary water/oil mixtures. Nevertheless, the introduction of the kijkij scaling parameter is not sufficient to allow the accurate reproduction of experimental IFT data for ternary water/oil/oil mixtures.@en

The world is facing today a number of challenges related to energy efficiency and security, climate change, water management and supply, and others. Basic and applied research can provide the necessary tools to address effectively these challenges. In most cases, a better understanding and accurate modeling at different levels of the underlying system is necessary. The unprecedented increase of computing power at relatively low price in recent years, the development of efficient computational algorithms and accurate models, and the design of sophisticated experimental techniques allow us today to calculate, model, and measure phenomena at the microscopic (molecular) level and predict the macroscopic properties@en