In Silico Prediction of the Temperature-Dependent Decomposition Rate Coefficients of Symmetrical Azo Compounds

The decomposition rate coefficient (k(d)) is a crucial kinetic parameter for the decomposition of azo compounds. In this work, we develop quantitative structure-property relationship (QSPR) models for the first time to predict the decomposition kinetic parameters at a wide range of temperature for symmetrical azo compounds, including azonitriles, alkyl azo, and aryl azo. The models are developed based on norm index (NI) and quantum chemical (QC) descriptors, so-called ln k(d) (T, NI, QC) models. Based on the collected data, solvents have an apparent influence on the k(d) values for aryl azo, and the ln k(d) (T, NI, QC, sol)(aryl) (azo) model with better performance than ln k(d) (T, NI, QC)(aryl) (azo) model is thus developed. Meanwhile, the as-developed models are endowed with excellent accuracy in calculating the k(d) values of initiators, which can be further applied to new initiators beyond the data set used for modeling. Interestingly, both Arrhenius parameters and their temperatures at which the half-lives (t(1/2)) are 0.1, 1, and 10 h of azo initiators can be extrapolated from the established models. These data offer an insightful understanding of decomposition kinetics and facilitate the selection of appropriate initiators in radical polymerization. In the long term, it is expected that QSPR models with higher predictive accuracy based on the data collected from a wider range of experimental conditions can be developed, enabling the design and screening of new azo initiators.