Energy-Based Interface Detection for Phase Change Processes of Monatomic Fluids in Nanoconfinements

An energy-based liquid–vapor interface detection method is presented using molecular dynamics simulations of liquid menisci confined between two parallel plates under equilibrium and evaporation/condensation conditions. This method defines the liquid–vapor interface at the location where the average kinetic energy of atoms first exceeds the average potential energy imposed by all neighboring molecules. This definition naturally adapts to the location of the menisci relative to the walls and can properly model the behavior of the liquid adsorbed layers. Unlike the density cutoff methods frequently used in the literature that suffer from density layering effects, this new method gives smooth and continuous liquid–vapor interfaces in nanoconfinements. Surface tension values calculated from the equilibrium MD simulations match the Young–Laplace equation better when using the radius of curvatures calculated from this method. Overall, this energy-based liquid–vapor interface detection method can be used in studies of nanoscale phase change processes and other relevant applications.