Skip to Main content Skip to Navigation
New interface
Journal articles

First QSPR models to predict the thermal stability of potential self-reactive substances

Abstract : Self-reactive substances are unstable chemical substances which can easily decompose and may lead to explosion in transport, storage, or process situations. For this reason, their thermal stability properties are required to assess possible process safety issues and for classification purpose. In this study, the first quantitative structure–property relationships (QSPR) dedicated to this class of compounds were developed to predict the heat of decomposition of possible self-reactive substances from their molecular structures. The database used to develop and validate the models was issued from a dedicated experimental campaign on 50 samples using differential scanning calorimetry in homogeneous experimental conditions. QSPR models were derived using the GA-MLR methods (using a genetic algorithm and multi-linear regressions) using molecular descriptors calculated by Dragon software based on two types of inputs: 3D molecular structures determined using the density functional theory (DFT), allowing access to three-dimensional descriptors, and from SMILES codes, favoring the access to simpler models, requiring no preliminary quantum chemical calculations. All models respected the OECD validation guidelines for regulatory acceptability of QSPR models. They were tested by internal and external validation tests and their applicability domains were defined and analyzed.
Document type :
Journal articles
Complete list of metadata
Contributor : Gestionnaire Civs Connect in order to contact the contributor
Submitted on : Tuesday, November 8, 2022 - 3:02:22 PM
Last modification on : Tuesday, November 15, 2022 - 8:48:44 AM


2022-079 post-print.pdf
Files produced by the author(s)




Guillaume Fayet, Annett Knorr, Patricia Rotureau. First QSPR models to predict the thermal stability of potential self-reactive substances. Process Safety and Environmental Protection, 2022, 163, pp.191-199. ⟨10.1016/j.psep.2022.05.017⟩. ⟨ineris-03830691⟩



Record views


Files downloads