Abstract
Recent attempts to resolve the disagreement between the present theories of vapor condensation and experiment has centered around suggesting models for the surface structure of embryonic droplets which differ from the flat‐plane surface structure for the case of hydrogen‐bonded liquids. In this paper we extend these studies. First, a brief review of the bulk and interface structure of liquid water is presented. We then extend the discussion to other liquids studied in nucleation experiments and learn that certain fundamental features of their bulk surface structure enables us to categorize them in two distinct classes: (a) liquids which exhibit orientation of the polar molecules at the liquid interface and (b) liquids of quasispherical molecules which exhibit no ordering at the interface. Also, liquids of class (a) follow the Becker–Döring nucleation theory, and liquids of class (b) follow the Lothe–Pound nucleation theory. Possible surface structure models for embryonic droplets are discussed. We conclude that the failure of the Lothe–Pound theory to explain the nucleation of polar liquids is due to its failure to account for the deviation of critical nucleus surface properties from bulk liquid surface properties. The apparent success of the Becker–Döring theory to predict the nucleation of polar liquids is believed to be fortuitous.