Computer simulation of microclusters, surfaces, and interfaces

Abstract The structure and thermodynamics of condensed–phase microclusters, the planar liquid surface, and the liquid‐solid interface are of special interest in several fields broadly related to the general discipline of chemical physics. Until recently, we have been able to say with confidence very little about the microscopic (or molecular) properties of these systems and have resorted to the application of macroscopic thermodynamics or heuristic models of a simplistic nature. The heuristic approach requires a successful choice of a suitable model for the problem at hand which (1) reasonably well represents the system of interest and (2) provides a relatively simple formalism from which numerical quantities can be obtained. While these models provide insight into the origins of certain features of a particular microcluster or interface, they cannot serve to elucidate the molecular structure since this is assumed in one way or another. For example, in a real sense clusters of a few tens of atoms are “all surface” described by atomic reconstruction grossly different from any macroscopic crystal or liquid surface. Hence, attempting to argue that a microcluster should be quantitatively viewed as a “crystal or liquid droplet” possessing properties of the macroscopic state of matter is highly questionable and has troubled nucleation theorists for decades.