Nucleation of colloids and macromolecules: does the nucleation pathway matter?

TL;DR

This paper uses a fluctuating hydrodynamics framework to compare classical and non-classical nucleation pathways, showing that classical pathways are highly unlikely for large nuclei in protein solutions.

Contribution

It introduces a method to evaluate the likelihood of various nucleation pathways, including non-classical ones, using a formalism based on fluctuating hydrodynamics.

Findings

Classical nucleation pathways are extremely unlikely for large nuclei.

The probability of non-classical pathways decreases exponentially with nucleus size.

The relative probability scales as e^{-N_c/3} for the studied system.

Abstract

A recent description of diffusion-limited nucleation based on fluctuating hydrodynamics that extends classical nucleation theory predicts a very non-classical two-step scenario whereby nucleation is most likely to occur in spatially-extended, low-amplitude density fluctuations. In this paper, it is shown how the formalism can be used to determine the maximum probability of observing \emph{any} proposed nucleation pathway, thus allowing one to address the question as to their relative likelihood, including of the newly proposed pathway compared to classical scenarios. Calculations are presented for the nucleation of high-concentration bubbles in a low-concentration solution of globular proteins and it is found that the relative probabilities (new theory compared to classical result) for reaching a critical nucleus containing $N_c$ molecules scales as $e^{-N_c/3}$ thus indicating that for all but the smallest nuclei, the classical scenario is extremely unlikely.