Early detection of the thermal and explosion hazards of chemicals is vital for safe handling and manufacturing. While thermal analysis methods using sealed-cell differential scanning calorimetry are commonly used to screen for shock sensitivity via empirical Yoshida-type equations, the reliability of the underlying thermal data and the robustness of this predictive approach require systematic validation. This study investigates the Yoshida approach from thermal analysis and energetic materials perspectives. We compiled a data set of more than 100 reactive compounds, including common reagents and energetic materials, to critically evaluate the variability of decomposition enthalpy and characteristic onset by DSC. Our results reveal significant scatter, up to several times, in reported decomposition heat values, attributable to factors such as gas leakage, calibration inconsistencies, interactions with crucible, and thermal runaway. We establish a correlation between decomposition enthalpy and the thermodynamically estimated maximal heat of explosion for the compound. Furthermore, by comparing original Yoshida and Pfizer-modified equations against experimental mechanical sensitivity data, we demonstrate that the conservative Pfizer adaptation achieves a superior recall (true positive rate) for identifying hazardous compounds. This work underscores the thermodynamic (energy content) and kinetic (onset temperature) foundations of the Yoshida-type equations, highlights the critical need for standardized, high-quality thermal data, and provides practical guidance for enhancing the reliability of early-stage hazard screening tools in chemical research and development.
Muravyev et al. (Thu,) studied this question.