When vapor nucleates into liquid, the critical droplets break a key assumption of classical nucleation theory (CNT). CNT assumes that critical droplets have the same density as the final liquid—an assumption that leads to prediction on nucleation rate that can be wrong by orders of magnitude. Using state-of-the-art computer simulation strategies designed to capture rare events, we directly observe that the critical droplets are less dense than ordinary liquid. This reveals a nonclassical nucleation pathway in which droplets grow and densify simultaneously. Guided by this insight, we develop a simple theoretical model that explicitly includes droplet density as a variable. Without adjustable parameters, the model accurately predicts nucleation rates, nucleation pathways, and critical droplet properties observed in simulations. Our results identify droplet density as a key ingredient governing nucleation and provide a practical framework for improving nucleation theories.
This work has been done in collaboration with Julien Lam (UMET) and was supported by the ANR TITANS project. The article has been published in Physical Review Letters : Y. Wu, T. Philippe, A. Graini, and J. Lam, “Nonclassical nucleation pathways in liquid condensation revealed by simulation and theory,” Phys. Rev. Lett. 136, 017101 (2026).
