S. Wauters

Ab initio calculations are presented on radical reactions that occur during the formation of coke in a thermal cracking unit. Kinetic parameter, for the addition reaction of the ethylbenzene radical to ethene and the subsequent cyclization of the butylbenzene radical are calculated by means of Transition State Theory and Density Functional Theory. Special care is taken to correctly treat the internal rotations to predict accurate values of the preexponential factor. The influence of the local structure of the coke matrix on the kinetic parameters is tested by calculating kinetic parameters of clusters consisting of more than one benzene ring. (C) 2002 Wiley Periodicals, Inc. | Conference: 9th International Conference on Application of the Density Functional Theory to Chemistry and Physics Location: MADRID, SPAIN Date: SEP 10-14, 2001

Ab initio density functional theory calculations have been carried out on a model reaction involved in coke formation during the thermal cracking of hydrocarbons, namely, the addition of the ethylbenzene radical to ethene. This study enables one to get more microscopic insight into the mechanistic and kinetic aspects of the reaction. A profound ab initio conformational analysis of the formed products, reactants, and transition states is made. The impact of internal rotations on the two kinetic parameters deduced from transition state theory (TST), the activation energy and the preexponential factor, has been studied in detail. Furthermore, we report on the various components that govern the kinetic parameters. Preexponential factors are very sensitive to the accuracy of constructing the microscopic partition functions. Internal rotations play a dominant role in the reaction mechanism, and their impact on the preexponential factor is large. Hence, a very accurate handling of internal rotations is of crucial importance. We present a new algorithm to extract exactly on a quantum mechanical basis the partition functions of the internal rotations. The calculations as presented here are especially important for complex reaction schemes, for which experimental data are not always available.