State Tracking in Nonadiabatic Molecular Dynamics Using Only Forces and Energies

Area of Science:

• Computational chemistry
• Quantum dynamics

Background:

• Nonadiabatic molecular dynamics (NAMD) is crucial for understanding chemical reactions.
• Accurate identification of avoided crossings is essential for reliable NAMD simulations.
• Existing methods often rely on computationally expensive wave function information.

Purpose of the Study:

• To develop a novel algorithm for identifying trivial crossings in NAMD.
• To introduce a phase consistency correction for nonadiabatic couplings.
• To assess the performance of these algorithms on model systems.

Main Methods:

• An algorithm using only state energies and gradients to detect trivial crossings.
• A phase consistency correction for time-derivative nonadiabatic couplings.
• Validation using various state crossing models.

Main Results:

• The new algorithm effectively identifies trivial crossings without wave functions or overlaps.
• The phase correction algorithm improves population dynamics accuracy.
• Performance is optimal for localized nonadiabatic coupling regions.

Conclusions:

• The developed algorithms offer efficient alternatives for NAMD calculations.
• State tracking alone is insufficient for accurate population dynamics.
• Nonadiabatic coupling phase correction is vital for precise simulations.