Automated Discovery of Elementary Chemical Reaction Steps Using Freezing String and Berny Optimization Methods

Area of Science:

• Computational Chemistry
• Chemical Kinetics
• Reaction Mechanism Elucidation

Background:

• Discovering elementary chemical reaction steps is crucial for understanding complex chemical processes.
• Manual identification of reaction pathways can be time-consuming and may miss novel mechanisms.

Purpose of the Study:

• To develop a fully automated protocol for the discovery of elementary chemical reaction steps.
• To investigate the reactivity of single-molecule systems relevant to combustion and atmospheric chemistry.

Main Methods:

• Utilized a combination of double-ended (freezing string) and single-ended (Berny optimization) transition-state optimization algorithms.
• Implemented a protocol for fully automated elementary reaction step discovery.

Main Results:

• Successfully detected known reaction pathways in combustion and atmospheric chemistry systems.
• Identified previously unknown reaction pathways involving significant atom rearrangements without human intervention.
• Demonstrated the utility of the automated approach in exploring chemical reactivity.

Conclusions:

• The proposed automated protocol significantly accelerates the discovery of new chemical reactions.
• This systematic approach leads to more accurate computer simulations of chemical processes.
• Automated reaction path finding is a powerful tool for advancing chemical understanding.