Summary Retrograde gas condensation is an intriguing and seemingly counterintuitive phase behavior phenomenon in which some gas mixtures condense as the pressure decreases at certain regimes above the critical temperature. This work aims to provide an explanation that bridges the gap between phenomenological observations and microscopic mechanisms using thermodynamic modeling and molecular dynamics (MD) simulation. Phase envelopes for a selection of binary hydrocarbon mixtures have been generated and compared with experimental results using the Peng-Robinson (PR) equation of state (EOS). We also used MD simulation in the NVT ensemble to simulate the constant composition expansion of a ternary methane (C1), normal butane (nC4), and normal decane (nC10) mixture; the intermolecular energies and other structural and transport properties were computed from the MD simulations to elucidate the retrograde condensation behavior. The optimized potentials for liquid simulations-all atom (OPLS-AA) force field was used along with the leap-frog algorithm to integrate Newton’s equations of motion with the velocity rescaling thermostat to maintain the temperature. EOS modeling was in good agreement with experimental data for the studied binary mixtures. Additionally, EOS modeling was used to showcase the differences between the critical temperature and cricondentherm of different binary mixtures; this showcases the extent of the negative slope region in the phase envelope, which is associated with retrograde condensation. MD simulations revealed that the intermolecular interaction energies (IEs) among heavier hydrocarbon components with low vapor pressures increase as pressure decreases at certain regions above the critical temperature and below the cricondentherm. Gas condensate systems typically exhibit methane mole fractions of 0.6–0.8; however, the liquid dropouts were predominantly composed of heavier hydrocarbon compounds. In the MD simulation, pressures corresponding to aggregation of nC10 also showed a slow decay in the radial distribution function (RDF) between terminal and secondary carbon atoms in the nC10 with nC4, suggesting the assemblage of the heavier fractions in liquid-like molecular clusters. Furthermore, variation in the ratio of self-diffusion of the heavier components compared with the lighter components was observed, which can be attributed to the increased intermolecular interactions of the heavier components at the liquid dropout region. Despite their extensive application in studying complex fluids, reports of MD simulations in retrograde condensation are almost missing from the literature. To our knowledge, this study represents one of the first applications of MD simulations to retrograde condensation, providing direct molecular-scale evidence of how clustering, intermolecular energies, and diffusion behavior evolve during constant composition expansion.
Salem Alshammari (Sun,) studied this question.