Abstract:
Azeotropes are special mixtures in which the components have same composition both in liquid and vapor phases at a specific temperature and pressure. Because of this property, the two components act as a single liquid and boil together, making separation difficult. This study is important for improving separation process in physical chemistry. However, the prediction of azeotropic shift remains challenging due to competing molecu- lar factors. For the study of this behaviour, molecular dynamics simulations are performed using GROMACS. The vapor-liquid equilibrium of binary Lennard-Jones (LJ) mixtures is studied to understand the effects of temperature, composition, and asymmetry in interac- tion or size on phase behavior. The simulations were conducted in the range of compo- sitions (χA = 0.1 to 0.9) and temperatures (60 K, 80 K, 100 K, and 120 K) for symmetric mixture (single component system) and asymmetric mixtures that have differences in the strength of interaction (ϵ) and molecular size (σ). Number density profiles, radial distribu- tion functions (RDF), mole fraction vs. composition analysis, and interfacial analysis were used to study the structural and compositional properties of binary mixtures. The profiles of the number density showed that there was a clear separation between the vapor and the liquid phase, with a decrease in the density variations with an increase in the tem- perature. The RDF analysis indicated the existence of short-range ordering in the liquid phase, which decreased with high temperature. The compositional analysis showed the change in the concentration of each component with the overall mole fraction, whereas the interfacial analysis revealed the nature of the transition region between the vapor and liquid phases. This study shows that asymmetric interactions play a major role in the shift of the azeotropic point. These findings provide better insight into the molecular interactions that regulate the vapor-liquid equilibrium and provide information about azeotrope formation. The study was extended to a real system by investigating the vapor–liquid equilibrium of the benzene–methanol azeotrope using the AMOEBA force field; however, reliable re- sults could not be obtained due to difficulties in achieving stable pressure equilibration. In future work, more advanced machine learning potentials will be explored to improve the accuracy and efficiency of the molecular simulations.
