Phase and interface determination in computer simulations of liquid mixtures with high partial miscibility

Area of Science:

• Computational chemistry
• Physical chemistry
• Materials science

Background:

• Partially miscible solutions pose simulation challenges when component concentrations are similar across phases.
• Distinguishing molecules between phases in such systems is computationally difficult.

Purpose of the Study:

• To develop and apply a novel density-based clustering approach for analyzing partially miscible solutions.
• To address the challenge of phase identification in simulations with high mutual solubility.

Main Methods:

• A self-tuning, density-based clustering algorithm was developed.
• The algorithm was applied to a mixture of an ionic liquid and benzene.
• The computational efficiency was evaluated for analyzing long simulation trajectories.

Main Results:

• The proposed algorithm effectively identifies molecules belonging to different phases.
• The method demonstrates near-linear scaling, making it suitable for large datasets.
• Successful application to an ionic liquid-benzene mixture validated the approach.

Conclusions:

• The density-based clustering method offers a robust solution for phase identification in challenging partially miscible systems.
• Its computational efficiency makes it valuable for analyzing extensive molecular dynamics and Monte Carlo simulations.