Structural cluster analysis of chemical reactions in solution

Area of Science:

• Chemical Physics
• Computational Chemistry
• Molecular Dynamics

Background:

• Analyzing atomic trajectories from chemical reactions in solution is complex.
• Identifying distinct structural states (reactants, products, transition states) is challenging, especially with solvent molecules and atom exchanges.

Purpose of the Study:

• To develop a simple and general approach for clustering structures from atomic trajectories.
• To enable automatic resolution of distinct structural clusters in complex chemical systems.
• To facilitate the analysis of large molecular dynamics trajectories.

Main Methods:

• Utilizing distance metrics invariant under permutation of identical atoms or molecules.
• Applying the method to atomic trajectories of chemical reactions in solution.
• Demonstrating robustness in the presence of atom-exchanges and numerous solvent molecules.

Main Results:

• Successfully resolved distinct structural clusters corresponding to reactants, products, and transition states.
• Showcased the ability to handle complex scenarios with atom-exchanges and hundreds of solvent molecules.
• Significantly simplified the analysis of large atomic trajectories.

Conclusions:

• The developed approach offers a powerful tool for analyzing chemical reaction dynamics in solution.
• It paves the way for constructing kinetic network models of activated processes.
• The method integrates well with existing computational schemes for conformational ensembles.