Multistate Coupled Diabatic Neural Network potential for the quantum non-adiabatic Photofragmentation of CH + 2

Key Points

• The aim is to develop a method for tracking non-adiabatic transitions in photodissociation of CH2+ accurately using neural networks.
• Introduced an automated diabatization method to generate diabatic potential energy matrices (PEMs).
• Focused on far-ultraviolet photodissociation and complex non-adiabatic transitions across excited states.
• Utilized a multistate coupled neural network framework.
• Successfully produced highly accurate PEMs for CH2+, improving modeling of its photodissociation.
• Demonstrated the model's efficiency in capturing the dynamics of non-adiabatic transitions.

Abstract