CARNOT: a Fragment-Based Direct Molecular Dynamics and Virtual-Reality Simulation Package for Reactive Systems.

Area of Science:

• Computational Chemistry
• Chemical Kinetics
• Molecular Dynamics

Background:

• Traditional methods for studying complex reaction mechanisms are time-consuming and may miss crucial processes.
• Analyzing individual elementary reactions limits comprehensive understanding of large systems.

Purpose of the Study:

• To present a novel direct molecular dynamics (DMD) approach for simulating complex chemical reactions.
• To introduce the CARNOT virtual-reality simulation program for simultaneous reaction analysis.
• To enable efficient study of reaction mechanisms under finite temperature and pressure.

Main Methods:

• Developed a direct molecular dynamics (DMD) method partitioning reactive systems into adjustable molecular fragments.
• Extended explicit polarization theory to reactive events (ReX-Pol).
• Incorporated a graph-data structure algorithm in CARNOT for reaction network analysis and mechanism reduction.

Main Results:

• Successfully simulated a large combustion reaction (28,650 atoms) using ab initio DMD.
• Identified mechanistic and dynamical events through reaction network analysis.
• Demonstrated the applicability of DMD to systems with varying numbers of open and closed-shell fragments.

Conclusions:

• The proposed ab initio DMD method and CARNOT program offer an efficient approach to studying complex reaction systems.
• This method overcomes limitations of traditional techniques by simulating multiple reactions simultaneously.
• The approach is versatile and applicable to various complex chemical reactions beyond combustion.