Per-Atom Identification: Structure/Dynamics of Supercritical Fluids

At the liquid-gas critical point, the boiling curve terminates and the two phases merge into a single phase of supercritical fluid (SCF). SCF shows an anomalous blend of liquid-like and gas-like traits, which not only leads to industrial applications, but also raises questions on the fundamental physics that governs its exotic physicochemical behaviors. In this work, we answer this question by identifying liquid-like and gas-like molecules coexisting in the SCF, captured by machine learning analysis on simulation data. The deltoid coexistence region encloses the Widom line, and it may therefore be termed the Widom delta. The number fractions are found to follow simplified two-state thermodynamics model, and are closely related to the magnitude of supercritical anomaly which originates from the fluctuation between microstates. Furthermore, we provide evidence that the microstate number fraction is a more fundamental descriptor of supercritical fluid compared to conventional order parameters, e.g. bulk density. These results strongly support modeling the SCF as a mixture of two interchangeable microstates, of which macroscopic properties arise from the microscopic partitioning of the space by molecules.