Vibrational energy transfer of free OH groups at the air/water interface via neural network-based molecular dynamics simulations.

By integrating neural network-based molecular dynamics simulations with the mixed quantum/classical method, we investigate the vibrational energy transfer pathways of interfacial free OH groups. Systematic analysis reveals distinct time scales associated with different energy transfer mechanisms. The results indicate that the intramolecular channel dominates overall energy relaxation and acts as the primary pathway, with the reorientation mechanism serving as a secondary contributor. In comparison, intermolecular transfer and Fermi resonance govern considerably slower dynamics processes. Consequently, the calculated total time is in good agreement with experimental observations, offering fundamental insights into intrinsic energy flow dynamics at the air/water interface.