Abstract
<jats:p>Despite being a century old, classical nucleation theory (CNT) remains in wide use to this day for qualitatively describing the phase change kinetics spanning a wide variety of materials relevant to both natural and commercial processes. Unfortunately, the absolute nucleation rates predicted by CNT often disagree with experimental measurements by many orders of magnitude. This discrepancy is evident even in the simplest of model systems -- so-called hard sphere (HS) colloids -- where the constituents are assumed to be non-deformable, share a common size, and not to interact with each other. In this work, a simple thermodynamic correction, which introduces a new phase reflecting the densified nucleated fluid fraction formed out-of-equilibrium, is used to overcome the key source of error in determining the activation energy barrier of nucleation in HS systems. Applying it is found to benefit CNT rate predictions by increasing the activation energy by an average of 54% across the four different system supersaturations investigated falling within the coexistence region. Separately, two alternative thermodynamic corrections are proposed for broader application (to also include self-interacting atoms, molecules, or colloids) by leveraging the CNT energy profile directly to identify the smallest thermodynamically stable cluster, and in so doing provide increases in the predicted activation energy barrier for steady-state nucleation of 33% and 50%. This range has particular significance in the context of previous simulations of HS systems that have indicated the (uncorrected) energy barrier predicted by CNT is 30% to 50% too low. </jats:p>