(731f) The Influence of Detailed Stereochemistry during Kinetic Model Construction – a Case Study for the Oxidation of Cyclohexane

In the last decade, significant progress has been made in the construction of reliable kinetic models for important processes based on radical chemistry, e.g., pyrolysis and oxidation. Such models describe the chemistry of hundreds of species with several thousands of associated reactions. To avoid tedious, error-prone and often incomplete model development, automatic kinetic model generation codes have been developed to construct models systematically, e.g., the in-house developed code Genesys. The present work demonstrates the importance of accounting for detailed stereochemistry during model construction to accurately describe the oxidation of cyclohexane. Genesys has already been extended with the feature to include detailed stereochemistry. Algorithms are implemented which detect stereocenters, generate the possible enantiomers and assign different thermodynamic properties to diastereomers. In addition, both stereoselective and -specific reactions can be defined, recognized and assigned different kinetic parameters. This requires many additional thermodynamic and kinetic parameters obtained from quantum chemical calculations which makes use of this feature labor-intensive. Thermochemical parameters of all diastereomers for cyclohexane-derivatives are calculated at the CBS-QB3 level of theory. Less computationally demanding methods, such as group additivity and rate rules, are used for the remaining species and reactions. In this work, it is shown that large differences exist between the thermodynamic properties of stereoisomers (c.f. Table 1) and their associated reactivity (e.g., factor 10 difference for an axially positioned versus an equatorially positioned reactive moiety). Potential energy surfaces are constructed which emphasize the impact of stereochemistry on the favorable reaction pathways. The final model is validated with experimental data available from literature for a variety of reactor configurations and operating conditions. The kinetic model already captures the experimental trends well, see Figure 1, without fitting any of the rate coefficients. In addition, model predictions are qualitatively better compared to other models which do not take stereochemistry into account.