Improved algorithm for the direct dynamics variational multi-configurational Gaussian method

Area of Science:

• Quantum Chemistry
• Theoretical Chemistry
• Computational Chemistry

Background:

• The Direct Dynamics variational Multi-Configurational Gaussian (DD-vMCG) method simulates nuclear time evolution using quantum mechanics.
• Previous studies highlighted its potential for accurate non-adiabatic excited-state molecular dynamics.
• Further algorithmic improvements are needed for broader applicability.

Purpose of the Study:

• To enhance the accuracy and efficiency of the DD-vMCG algorithm.
• To develop new methods for managing quantum chemistry calculations within the simulation.
• To enable accurate simulations of complex chemical systems.

Main Methods:

• Developed an efficient parallel algorithm for the DD-vMCG database management.
• Improved the Shepard interpolation scheme for potential energy surfaces.
• Introduced symmetry considerations and a new phase convention for propagation diabatization.

Main Results:

• The enhanced DD-vMCG algorithm demonstrates improved accuracy and efficiency.
• Benchmark calculations on the allene radical cation confirmed a consistent non-adiabatic coupling vector field.
• The new methods effectively handle all degrees of freedom in the system.

Conclusions:

• The improved DD-vMCG method provides a robust framework for quantum dynamics simulations.
• This advancement facilitates the accurate treatment of complex chemical systems.
• Opens new possibilities for studying non-adiabatic processes in excited-state molecular dynamics.