Assessment of trajectory surface hopping methods in long-time nonadiabatic dynamics

Area of Science:

• Quantum Dynamics
• Computational Chemistry
• Theoretical Physics

Background:

• Modeling long-time population dynamics is crucial for understanding quantum systems.
• Trajectory surface hopping (TSH) is a common method, but its accuracy depends on various parameters and corrections.

Purpose of the Study:

• To comprehensively assess various TSH methodologies for the spin-boson model.
• To identify optimal parameter choices and rank the performance of different TSH approaches.
• To investigate the impact of decoherence and initial condition sampling on population dynamics.

Main Methods:

• Evaluated combinations of TSH approaches with decoherence correction methods.
• Incorporated different decoherence-time computation schemes and initial-condition sampling methods.
• Assessed the Shenvi, Subotnik, and Yang (SSY) phase correction approach.

Main Results:

• The SSY correction was found to generally over-accelerate population relaxation.
• Wigner sampling led to overestimated transfer rates, necessitating decoherence corrections.
• Boltzmann sampling demonstrated strong performance with bare TSH, potentially due to error cancellation.

Conclusions:

• Methodological choices, particularly initial condition sampling, significantly impact TSH accuracy.
• Bare TSH with Boltzmann sampling offers a surprisingly effective approach for specific scenarios.
• Further improvements in TSH methods are suggested based on observed trends.