A variational approach to nucleation simulation.

Area of Science:

• Thermodynamics
• Computational Physics
• Materials Science

Background:

• Nucleation, the process of phase transition from vapor to liquid, is crucial in many physical phenomena.
• Direct simulation of nucleation is challenging due to large free energy barriers and inaccessible timescales.
• Classical nucleation theory provides a physical model but has limitations in direct simulation.

Purpose of the Study:

• To investigate the nucleation of a supersaturated Lennard-Jones vapor into the liquid phase using computer simulations.
• To apply and evaluate a recently developed enhanced sampling method for overcoming simulation time scale limitations.
• To integrate the physical insights from classical nucleation theory into the enhanced sampling approach.

Main Methods:

• Utilizing computer simulations to model the nucleation process.
• Employing an enhanced sampling method based on the variational determination of a bias potential.
• Constructing the bias potential informed by the physical model of classical nucleation theory.

Main Results:

• Successfully applied an enhanced sampling method to study vapor-to-liquid nucleation.
• Demonstrated the feasibility of integrating classical nucleation theory with advanced simulation techniques.
• Identified and examined technical challenges associated with the applied methodology.

Conclusions:

• The developed approach is highly satisfactory for studying nucleation phenomena.
• This method paves the way for accurate calculation of nucleation rates in diverse systems.
• Offers a promising computational strategy for understanding phase transitions.