Complex reaction processes in combustion unraveled by neural network-based molecular dynamics simulation

Area of Science:

• Chemical Engineering
• Computational Chemistry
• Combustion Science

Background:

• Combustion processes involve complex chemical reactions generating numerous species and radicals.
• Simulating these reactions at an atomic level is crucial for understanding combustion phenomena.
• Traditional methods face challenges in capturing the intricate details of these systems.

Purpose of the Study:

• To apply neural network-based molecular dynamics (MD) simulations to a benchmark methane combustion system.
• To reveal and characterize detailed reaction pathways, including intermediate radicals and products.
• To explore the potential of this method for discovering new chemical reaction pathways.

Main Methods:

• Utilized neural network-based molecular dynamics (MD) simulations.
• Focused on simulating the benchmark combustion of methane.
• Analyzed the generated data to identify and record chemical reactions.

Main Results:

• Successfully simulated methane combustion, revealing detailed reaction processes.
• Recorded a total of 798 different chemical reactions.
• Discovered new chemical reaction pathways previously unknown.

Conclusions:

• Neural network-based reactive MD simulations offer a powerful tool for studying complex chemical systems.
• This methodology provides unprecedented atomic-level understanding of chemical reactions.
• The approach promises a new era for simulating combustion and discovering novel reaction pathways.