Nonclassical nucleation theory of condensation

Abstract

A longstanding question in the study of nucleation of liquids from the vapor phase is the validity of the classical nucleation theory (CNT) of Becker, Doring, and Zeldovich[l]. Experiments have been designed [2,3]to probe the range of validity of the classical theory, and have concluded that that theory is remarkably successful, although careful measurements are capable of showing small deviations. The introduction by Lothe and Pound [4] of large correction factors to classical nucleation rates has excited a great deal of controversy, but has led to much poorer agreement of theory with experiment. More recent theories [5] are closer to the classical prediction, but still do not agree with experiment as well as CNT. There are two ways to go beyond the CNT and the capillarity approximation that it implies. The first is to attempt direct statistical mechanical evaluation of the partition functions of small clusters of molecules, a procedure that is difficult for clusters as large as those important in nucleation processes. The second is the approach we follow [6]: the use of density functional theory in statistical mechanics to calculate the free energies of non-uniform fluids. Our approach includes, in a mean field sense, nonclassical effects such as curvature-dependent surface tensions and finite interfacial widths, and goes over naturally to the classical theory in the limit of large droplets.