Adaptive accelerated reactive molecular dynamics driven by parallel collective variables overcoming dimensionality...

Area of Science:

• Computational Chemistry
• Materials Science
• Chemical Engineering

Background:

• ReaxFF reactive molecular dynamics is crucial for studying chemical reaction mechanisms.
• Existing ReaxFF methods face limitations: high temperatures, short timescales, and dimensionality issues with collective variables.
• These limitations restrict the scope and applicability of ReaxFF simulations for complex systems.

Purpose of the Study:

• To introduce Parallel Collective Variable-Driven Adaptive Accelerated Reaction Molecular Dynamics (PCVR) to overcome ReaxFF limitations.
• To enable simulations at lower temperatures, longer timescales, and larger system sizes.
• To accurately model complex chemical reactions, such as ester oxidation, under industrially relevant conditions.

Main Methods:

• Integrated metadynamics with ReaxFF to develop PCVR.
• Employed customized Collective Variable (CV) parameterization based on bond distortion for parallel acceleration.
• Confined sampling to fixed cutoff ranges for distinct bond distortions to manage CV dimensionality.

Main Results:

• Successfully simulated the oxidation of ester-based oil with over 31,000 atoms at 500 K for 64 seconds.
• Achieved 61% efficiency compared to original ReaxFF and was approximately 37 times faster than previous accelerated methods.
• Accurately revealed the role of oxygen in ester oxidation at industrial temperatures, producing polymers consistent with experimental observations of sludge formation.

Conclusions:

• PCVR significantly enhances the applicability of ReaxFF for non-strongly coupled systems with multiple reaction barriers.
• The method overcomes system size limitations and enables simulations at lower temperatures and second-level timescales.
• PCVR advances reactive molecular dynamics, offering a powerful tool for studying complex chemical processes under realistic conditions.