Abstract:

To deeply reveal the thermal decomposition mechanism of NH₄NO₃, two kinetic models applicable to the thermal decomposition reaction of NH₄NO₃, namely the 3-1 model (three species and one reaction) and the 4-4 model (four species and four reactions), are constructed based on thermogravimetric (TG) experimental data and CRNN algorithm. The model can predict the TG curves and the material evolution during the thermal decomposition process of NH₄NO₃, and infer the reaction pathway and intermediate products. The results show that in the 3-1 model, the calculated activation energy is 98.49 kJ/mol, which is in good agreement with the experimental value. Based on the reaction network constructed using the 4-4 model, two initial reaction paths are identified, with corresponding activation energies of 98.02 kJ/mol and 139.43 kJ/mol, respectively. The entire thermal decomposition process involves the breaking of ionic bonds between NH₄⁺ and NO₃^-, hydrogen transfer reactions, autocatalytic reactions, and interactions between products. NO_2, as a key intermediate product, participates in multiple subsequent reaction paths, ultimately generating substances such as NO₂, N₂O, NO, N₂, O₂, and H₂O.

Key words: NH₄NO₃, thermal decomposition, reacting kinetic, chemical reaction neural network (CRNN), reaction mechanism