A molecular perspective on Tully models for nonadiabatic dynamics.

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Quantum Chemistry

Background:

• Simulating molecular dynamics beyond the Born-Oppenheimer approximation is crucial.
• Tully's one-dimensional models (1990) became standard for testing nonadiabatic dynamics methods.
• These models test approximations for processes like nonadiabatic (re)crossings.

Purpose of the Study:

• To bridge the gap between low-dimensional Tully models and high-dimensional molecular excited-state dynamics.
• To propose three specific molecules as "molecular Tully models" for standardized testing.
• To compare different nonadiabatic dynamics simulation methods using these molecular models.

Main Methods:

• Developed a molecular perspective of Tully's one-dimensional models.
• Identified three exemplary molecules to serve as high-dimensional test systems.
• Compared trajectory surface hopping (TSH) and ab initio multiple spawning (AIMS) methods.
• Provided comprehensive simulation parameters for reproducibility.

Main Results:

• Established a correspondence between Tully models and molecular excited-state dynamics.
• Demonstrated that proposed molecules capture key features of original Tully models in a high-dimensional space.
• Highlighted method-specific differences between TSH and AIMS in simulating these molecular models.
• Generated a standardized dataset for future validation of nonadiabatic dynamics methods.

Conclusions:

• Molecular Tully models offer a realistic, high-dimensional platform for testing nonadiabatic dynamics.
• These models facilitate a unified and standardized approach to validating computational methods.
• The study provides a crucial link between simplified models and complex molecular systems.