Ehrenfest dynamics accelerated with SPEED

Area of Science:

• Computational Chemistry
• Quantum Dynamics
• Theoretical Chemistry

Background:

• Mixed quantum-classical methods are vital for molecular processes with multiple electronic states.
• Standard methods like surface hopping and Ehrenfest dynamics are computationally intensive due to trajectory propagation.

Purpose of the Study:

• To develop a computationally efficient variant of Ehrenfest dynamics.
• To reduce the computational cost of simulating molecular dynamics involving multiple electronic states.

Main Methods:

• Single Potential Evaluation Ehrenfest Dynamics (SPEED) was developed.
• SPEED propagates trajectories using a single local quadratic effective potential in the diabatic representation.
• This method replaces multiple Hessian evaluations with a single one per time step.

Main Results:

• SPEED was shown to be equivalent to standard Ehrenfest dynamics for systems with quadratic potentials.
• The method was successfully combined with ALMO(MSDFT2) for ab initio calculations.
• SPEED qualitatively reproduced temperature-dependent hole transfer rates and predicted charge distribution in furan systems.

Conclusions:

• SPEED offers a significant efficiency gain for on-the-fly ab initio simulations.
• The method accurately describes charge transfer processes but is limited for highly anharmonic systems like retinal photoisomerization.