Most theories of homogeneous nucleation are based on a Fokker-Planck-like description of the behavior of the mass of clusters. Here we will show that these approaches are incomplete for a large class of nucleating systems, as they assume the effective dynamics of the clusters to be Markovian, i.e., memoryless. We present a nucleation theory which describes these non-Markovian dynamics and show how this influences the dynamics of clusters during nucleation.
To validate our theoretical framework, we have developed a novel method for the computation of nucleation times. This method is based on sampling small pieces of trajectories in different phase space domains and piecing them together in a smart way using equilibrium properties.
The method does not have any systematic error and is very efficient; the computational effort to calculate the first passage time across a free energy barrier does not increase with the height of the barrier.
Back to Rare Events in High-Dimensional Systems