Study of catalytic reaction kinetics using strongly typed genetic programming

Abstract

Kinetic modeling of some generic and real catalytic reactions systems were performed using strongly typed genetic programming. The reaction rate laws were created using the data generated for generic reactions and experimental data reported for the real reactions in the literature, and the structure of these equations were used to predict the possible reaction mechanisms as an aid to the traditional kinetic methods. It was found that the strongly typed genetic programming techniques are quite successful predicting the simple model arising from a simple mechanisms or complex mechanisms with simplifying assumptions. This is especially true for the surface reaction controlled systems which constitute about the 75% of the heterogeneous catalytic reactions. However, this method is not effective in the presence of complexities such as too many terms in the denominator of the rate equation, square term arising from the control of the surface reaction step in dual site reactions, square root terms in the dominator due to the dissociative nature of the reaction. Then it was concluded that the method may not be sufficient to understand the catalytic mechanisms by itself but it may help to reduce the work load required in the traditional Langmuir-Hinshelwood Hougen Watson approach that is commonly utilized for the kinetic studies of the catalytic reactions..