Assessing the performance of trajectory surface hopping methods: Ultrafast internal conversion in pyrazine

Area of Science:

• Quantum Chemistry
• Chemical Physics
• Computational Chemistry

Background:

• Trajectory surface hopping (TSH) methods are crucial for studying photoinduced nonadiabatic processes.
• Accurate simulation of internal conversion dynamics in molecules like pyrazine is essential for understanding photochemical reactions.

Purpose of the Study:

• To critically evaluate the accuracy of Tully's fewest switches surface hopping (FSSH) and Landau-Zener-type TSH (LZSH) algorithms.
• To compare these TSH methods against exact quantum dynamics for pyrazine's internal conversion.
• To validate a reduced-dimensional model for full-dimensional nonadiabatic dynamics simulations.

Main Methods:

• Nonadiabatic dynamics simulations using FSSH and LZSH algorithms.
• Comparison with exact quantum dynamics calculations on a three-state, nine-mode model of pyrazine.
• Full-dimensional on-the-fly nonadiabatic dynamics simulations interfaced with second-order algebraic diagrammatic construction (ADC(2)) ab initio electronic structure method.

Main Results:

• Both FSSH and LZSH methods show good agreement with exact quantum dynamics for diabatic populations.
• Fast population oscillations between B3u(nπ*) and A1u(nπ*) states are qualitatively reproduced by both TSH methods.
• Full-dimensional LZSH and FSSH simulations exhibit excellent agreement with each other and with the reduced-dimensional model, validating the model for 200 fs.

Conclusions:

• The LZSH algorithm provides accurate simulations of pyrazine's internal conversion dynamics.
• The three-state nine-mode model effectively captures the essential dynamics of pyrazine over 200 fs.
• The standard FSSH method propagating the electronic wave function in the adiabatic representation yields less accurate results compared to LZSH and FSSH with local diabatization.