PNNL

Abstract

A reliable molecular-level treatment of nucleation requires (i) accurate descriptions of the interaction potentials, (ii) a consistent theoretical approach that connects the interaction potential to the nucleation rate, and (iii) appropriate statistical mechanics to evaluate quantities of interest. In a recent PRL Nadykto et al. [1] “attack the fundamental problem of the sign preference by employing a higher level of theory such as quantum mechanics”, where quantum mechanics in this context refers to electronic structure calculations of interaction energies. They concluded that “The strong effect of the chemical nature of the core ion on the conversion of vapor molecules to clusters is essentially quantum in nature, and, thus, systematic accounting for the actual core species is impossible without taking into account the actual electronic structures of the core ions.” The purpose of this Comment is to point out conceptual errors concerning the critical importance of items (ii) and (iii) above and to clarify misrepresentations of our previous work on this topic [2]. Any consistent theory of nucleation must describe the relevant regions of configuration space that govern the kinetics and thermodynamics of cluster formation through evaporation and condensation processes – e.g., Dynamical Nucleation Theory (DNT) [3].