Impact of the size and degree of branching of alkanes on the rate rules approach: the case of isomerizations
Résumé
The accurate and high-throughput generation of kinetic data for combustion detailed chemical kinetic models remains a challenge given the large size of these models. For large organic molecules, the majority of kinetic data of the kinetic models are estimated using quantitative structure activity relationships within reaction classes and reaction rate rules approaches. In this work, we question the limits of the rate rule approach for the isomerization reaction classes, using electronic structure calculations and reaction rate theories. Systematic calculations were performed to investigate the effect of the size and the degree of branching of alkyl radicals on the rate coefficients and their consequences on the rate rule approach. Our computed kinetic data show that when the size of alkyl group is increased, the rate coefficients remains close to each other. This allows the use of methyl as representative models of larger alkyl groups to investigate the influence of increasing the degree of branching. The computed rate coefficients for 1,3-and 1,4-H-atom shifts show that the increase of the branching level, with spectator methyl groups in the transition state cyclic structure can strongly increase the rate coefficients, up to several orders of magnitude. Consequently, a single rate rule is not feasible for any degree of branching of a reaction belonging to the same isomerization reaction class. As variation in rate coefficients are important, this would lead to an explosion of the number rate rules as a function of branching in 5-, 6-, and 7membered cyclic transition states. A new approach is demonstrated and proposed where model transition states are tabulated, and methyl groups are assumed as alkyl groups and all combinations of substitutions in the model TSs, for a given reaction class, are included in a table with associated ab initio rate coefficients. The automation of the construction of such tables is possible and could be an interesting high-throughput / high-accuracy alternative to on-the-fly ab initio calculations of kinetic parameters.
Origine | Fichiers produits par l'(les) auteur(s) |
---|