Predicting the catalytic mechanisms of CuO/PbO on energetic materials using machine learning interatomic potentials

Double-base propellants (DBPs) have attracted significant attention in propulsion systems due to their excellent energy density and stability. This work aims to develop, for the first time, a highly accurate and efficient neural network potentials (NNP) model for DBPs, to elucidate the microscopic reaction mechanisms of nitroglycerin (NG) decomposition catalyzed by metal oxides (CuO and PbO). The NNP model was rigorously validated, demonstrating consistent accuracy in predicting mechanical and chemical properties when compared to density functional theory (DFT) calculations. Molecular dynamics (MD) simulations of NG thermal decomposition on CuO and PbO surfaces indicate that the phase transition of metal oxides increases adsorption probability and reactivity of NG. The dispersed catalysts significantly improve the stability and combustion performance of DBPs by consuming reactive small molecules. This research provides a crucial theoretical foundation and high-precision reactive force field for the design and application of DBPs, advancing the understanding of propellant combustion mechanisms and thermal decomposition processes.