Identification of important normal modes in nonadiabatic dynamics simulations by coherence, correlation, and...

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Quantum Chemistry

Background:

• Simulations of nonadiabatic molecular dynamics with many nuclear degrees of freedom are becoming more accessible.
• Extracting essential nuclear motion modes is crucial for gaining insights and building simplified models.
• Traditional methods focus on statistical variance, which may not reflect influence on electronic states.

Purpose of the Study:

• To develop and present novel analysis techniques for nonadiabatic dynamics simulations.
• To identify vibrational modes most affected by electronic excitation and state interactions.
• To extract key nuclear motion modes from surface hopping simulations.

Main Methods:

• Developed three distinct analysis techniques for surface hopping nonadiabatic dynamics simulations.
• Technique 1: Identifies coherent motion using variance ratios.
• Technique 2: Employs linear regression to link normal modes to electronic properties (excitation energies, gaps, overlaps).
• Technique 3: Uses time-frequency analysis to detect frequency changes in normal modes during dynamics.

Main Results:

• Applied the three techniques to surface hopping trajectories of [Re(CO)3(Im)(Phen)]+.
• Demonstrated that nonadiabatic dynamics for this complex are governed by a few specific modes.
• Identified dominant modes as carbonyl and phenanthroline in-plane stretches.

Conclusions:

• The developed techniques effectively extract influential vibrational modes from complex simulations.
• These methods offer a pathway to simplify understanding of nonadiabatic processes.
• The approach is applicable to various nonadiabatic dynamics methods beyond surface hopping.