Study on the Structure-Property Relationship between Novel Ring Sequence Index and the Performance of Azotriazole Compounds

Abstract Azotriazole is a compound composed of two five-membered triazole rings bridged by azo bonds. It offers advantages such as high energy density, high explosive energy, and safe insensitivity, making it a promising new energetic material. To investigate the quantitative structure-property relationship (QSPR) between the thermodynamic and detonation properties of 3,3′-azobis-1,2,4-triazole energetic compounds and their molecular structures, a novel ring sequence index was developed based on the atomic properties of these molecules and the positions of group connections on the triazole ring. Additionally, molecular connectivity indices for 25 molecules were calculated, and 2X, 4X, 5X and 4Xpc were optimized and selected as molecular structure descriptors. By combining these indices with the ring sequence index Ir, regression analyses were performed on the thermodynamic and detonation properties of azotriazole energetic compounds. Using these structural parameters as input layer nodes of the neural network, a model for predicting the thermodynamic and detonation properties of energetic azotriazole compounds was constructed employing a 5-2-1 neural network structure. The total correlation coefficient of each model exceeded 0.95, with average relative errors between the predicted and literature values of detonation velocity, detonation pressure, and detonation heat of 1.25%, 3.70%, and 2.05%, respectively. When the index was applied to modeling oxtriazole energetic materials, the average relative errors between the predicted and literature values of density (ρ), detonation velocity (D) and detonation pressure (P) were 1.10, 1.75, and 3.58%, respectively, indicating strong predictive performance. These results demonstrate a robust nonlinear relationship between the ring sequence index, molecular connectivity indices, and the detonation performance of triazole energetic materials.