A new strategy to simulate nonadiabatic dynamics of molecules on metal surfaces

The team proposed a new approach to accurately describe the nonadiabatic dynamics of molecular electron transfer at metal surfaces. Their work was published in Physical Review Letters.

Numerous experimental phenomena have demonstrated that nonadiabatic energy transfer is widespread in various interfacial processes, and therefore studying it is of great importance for understanding interfacial processes such as chemisorption, electrochemistry, and plasmonic catalysis.

However, in molecule-metal surface interactions, molecular vibrations, rotations, and translations are coupled with surface phonons and electrons, resulting in extremely complex energy transfer processes. Traditional models based on electronic friction have provided some insight, but are inadequate to capture the complex energy transfer observed in experimental studies.

To address this issue, Professor Jiang Bin from the University of Science and Technology of China, Chinese Academy of Sciences, and his team developed a simulation strategy and applied it to the energy transfer dynamics of CO molecules scattering from an AU(111) surface. The strategy starts by using constrained density functional theory (CDFT) to calculate the charge transfer states for different configurations of CO molecules on the metal surface.

Then, we utilized an embedded atomic neural network (EANN) to learn the CDFT energies and generate high-dimensional nonadiabatic potential energy surfaces (PESs).Finally, we applied the independent electron surface hopping (IESH) method to simulate the energy transfer process.

The results showed that the simulations closely matched the experimental data for the vibrational final state distribution of highly vibrationally excited CO (vi=17) after scattering. The vibrational relaxation probability, average translational energy, and scattering angle distribution of low vibrationally excited CO (vi=2) were also accurately reproduced by the simulations.

Specifically, the simulation results reveal distinct energy transfer pathways for different initial vibrational states. For the high initial vibrational state, the molecular vibrational energy is mainly transferred to the surface electrons and the molecular movement. In contrast, for the low initial vibrational state, the molecular vibrational energy is only transferred to the surface electrons.

“This work represents a major advance in understanding energy transfer in molecular-surface systems. By providing a robust and accurate framework for modeling nonadiabatic dynamics, this strategy can be extended to study other nonadiabatic dynamics at surfaces and may lead to future developments in catalysis, materials science and nanotechnology.”