Project Details
Projekt Print View

Non-equilibrium theory of nucleation at first order phase transitions.

Subject Area Statistical Physics, Nonlinear Dynamics, Complex Systems, Soft and Fluid Matter, Biological Physics
Term from 2019 to 2024
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 430195928
 
Final Report Year 2024

Final Report Abstract

Nucleation at first order phase transitions is a non-equilibrium process. It is also a collective process, which involves a very large number of microscopic degrees of freedom. Hence to predict nucleation rates, a coarse-grained, non-equilibrium description is required. Classical nucleation theory, which builds on a diffusion process in a free energy landscape, claims to give such an description. However, classical nucleation theory requires some assumptions (such as Markovianity) that seem questionable. In this project we used projection-operator techniques to derive exact coarsegrained equations of motion for systems out of thermal equilibrium. We showed how to obtain an effective equation of motion for the nucleation process that resembles classical nucleation theory in that it contains a derivative of a thermodynamic potential (or “potential of mean force”) as a driving force. However, the resulting equations are not Markovian. Further, even if the memory functions decayed rapidly, the resulting structure would still not resemble a simple diffusion process in a free-energy landscape. A numerical analysis of molecular dynamics simulation data of colloidal hard spheres showed that the temporal extent of the memory is of the same order as the duration of the nucleation process itself. Hence, we do not easily obtain a Markovian description in this situation. We also tackled the problem of the mismatch between experimentally measured nucleation rates and their theoretical predictions in colloidal hard spheres. This mismatch had been regarded as a shortcoming of the theory. In large scale molecular dynamics simulations of hard-sphere systems, however, we demonstrated that the mismatch is explained by a reinterpretation of the experimental data. If one allows for poly-crystalline nuclei due to twinning, the discrepancy between theory and experiment is resolved. Other results obtained in this research project comprise a method to generate new trajectories from the reduced non-equilibrium description yielding a route to rigorous coarse graining, an exact reduced formalism to describe driven systems with a formulation involving equilibrium memory functions, and an analysis of tracer dynamics within a membrane.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung