Memory effects in nonadiabatic molecular dynamics at metal surfaces

Area of Science:

• Surface science
• Theoretical chemistry
• Computational physics

Background:

• Nonadiabatic dynamics at metal surfaces are complex.
• Langevin equations are used to model these dynamics.
• Temporal correlations can influence the accuracy of these models.

Purpose of the Study:

• Investigate the impact of temporal correlations in Langevin equations for nonadiabatic surface dynamics.
• Compare Langevin dynamics with master equation approaches.
• Extend the Langevin method to anharmonic potentials and real surface systems.

Main Methods:

• Developed and analyzed a Langevin equation with temporal correlations.
• Performed comparative studies with perturbative master equation approaches.
• Utilized density functional theory (DFT) to calculate Langevin trajectories for N(2) desorption from Ru(0001).

Main Results:

• Memory effects are essential for conserving the ground state energy of a harmonic oscillator in Langevin dynamics.
• Langevin dynamics provide a superior description compared to master equations at high temperatures and friction.
• Calculated Langevin trajectories reveal that memory effects reduce energy dissipation during N(2) associative desorption.
• Proposed an ab initio scheme for calculating temporal correlation functions and dynamical friction within DFT.

Conclusions:

• Temporal correlations and memory effects are vital for accurate nonadiabatic dynamics modeling using Langevin equations.
• The Langevin approach offers advantages over master equations, especially in nonperturbative regimes and for anharmonic systems.
• This work introduces a pathway for ab initio calculations of friction and correlation functions in surface dynamics.