An automated method for graph-based chemical space exploration and transition state finding.

Area of Science:

• Computational Chemistry
• Chemical Reaction Networks
• Algorithm Development

Background:

• Automated chemical space exploration is hindered by high computational costs and numerous reaction pathways.
• Existing algorithms struggle with complex chemical systems due to limitations in quantum calculations.

Purpose of the Study:

• To develop a novel algorithm for efficient chemical space exploration.
• To generate comprehensive reaction networks using heuristics and chemical theory.
• To overcome computational limitations in exploring large chemical systems.

Main Methods:

• Employing molecular graph transformations based on functional groups to generate reaction networks.
• Considering only transformations that break two bonds and form two new ones for performance enhancement.
• Estimating energy barriers using quantum chemical calculations and the growing string method.
• Identifying non-octet species and refining the reaction network through energy barrier calculations.

Main Results:

• The algorithm successfully generated reaction networks for five different chemical reactions.
• Key reaction pathways were identified in all tested chemical systems.
• The method demonstrated significant performance enhancement compared to previous algorithms.

Conclusions:

• The proposed algorithm offers an efficient and effective solution for exploring chemical reaction networks.
• This approach overcomes previous computational limitations in automated chemical space exploration.
• The method accurately identifies critical reaction pathways in complex chemical systems.