Reliable estimation of heats of formation for energetic metal-organic materials: A structure-descriptor approach for defence applications

This study presents a predictive model for condensed-phase heats of formation of metal-containing energetic complexes (MCECs) and energetic metal-organic frameworks (EMOFs), leveraging a dataset of 148 compounds. Using elemental composition, triazole rings, and metal presence, the model achieves high accuracy ( R 2 > 0.94, mean absolute error ( MAE ) ≈ 390 kJ/mol) for screening high-energy materials. It outperforms prior methods, particularly for polycyclic systems, offering a practical tool for safer design and risk assessment in defense and industrial applications. • A robust and transparent model is developed for predicting condensed phase heats of formation of MCECs and EMOFs. • The model leverages elemental composition, triazole ring content, and key metal atoms as structural descriptors. • Validation against 148 diverse experimental compounds demonstrates high predictive accuracy and generalizability. • The approach significantly improves predictions for complex and polycyclic energetic materials compared to previous models. • Reliable predictions support safer process design, risk assessment, and rational screening of hazardous energetic materials.