Prediction of physico-chemical properties for REACH based on QSPR models

Guillaume Fayet (1) , Patricia Rotureau (1) , Vinca Prana (1, 2) , Carlo Adamo (2)

1 INERIS - Institut National de l'Environnement Industriel et des Risques
2 LECIME - Laboratoire d'Electrochimie, Chimie des Interfaces et Modélisation pour l'Energie (LECIME - UMR 7575)

Quantitative Structure Property Relationship models have been developed for the prediction of flash points of two families of organic compounds selected in the PREDIMOL French Project: amines and organic peroxides. If the model dedicated to amines respected all OECD validation principles with excellent performances in predictivity, the one dedicated to organic peroxides was not validated on an external validation set, due to the low number of available data, but already presented high performances in fitting and robustness. This work highlighted the need of gathering experimental data, as in progress in the PREDIMOL project, to achieve validated reliable models that could be used in a regulatory framework, like REACH. Such models are expected to be submitted to the European Joint Research Comity (JRC) and to existing tools (like the OECD ECHA QSAR Toolbox) to be available for use by industrials and regulatory instances.

Domaines: Sciences de l'ingénieur [physics]

Guillaume Fayet, Patricia Rotureau, Vinca Prana, Carlo Adamo. Prediction of physico-chemical properties for REACH based on QSPR models. 14. International Symposium on Loss Prevention and Safety Promotion in the Process Industry, May 2013, Florence, Italy. pp.925-930, ⟨10.3303/CET1331155⟩. ⟨ineris-00976244⟩