Physics-Based Modeling of Chemical Hazards in a Regulatory Framework: Comparison with Quantitative Structure-Property Relationship (QSPR) Methods for Impact Sensitivities

A semiempirical model based on simple physical assumptions is rigorously compared to a combination of two recent state-of-the-art quantitative structure property relationship (QSPR) methods for impact sensitivities of nitroaliphatic and nitramine compounds. For most datasets considered, it yields slightly better predictions than QSPR schemes, which is noteworthy, considering the fact that it relies only on three adjustable parameters and an equation developed independently of the data. Further advantages of this approach include physical interpretability, enhanced robustness, and wider applicability. Therefore, there is no doubt that such physics-based models provide valuable alternatives to the purely empirical relationships usually employed in regulatory contexts, especially in situations where experimental data are scarce.