A Computational Investigation of Hydrocarbon Cracking: Gas Phase and Heterogeneous Catalytic Reactions on Zeolites

Abstract

For many years, researchers have been developing theoretical methods of estimating reaction rates and energetics when experimental measurements are not available. Recent advances have led to composite energy methods with near chemical accuracy. The performance of these new methods for predicting activation energies and rate constants have not been evaluated for large hydrocarbon cracking reactions.In this work, we investigate the suitability of using composite energy methods for estimating activation energies for the cracking reactions of many hydrocarbon species including ethyl, n-propyl, iso-propyl, n-butyl, sec-butyl, iso-butyl, neo-pentyl radicals in the gas phase. Further work using Canonical Transition State Theory (CTST) and Rice-Ramsperger-Marcus (RRKM) theory is done to estimate the rate constants for these reactions. A comparison of our theoretical methods shows that activation energies normally are predicted within 4 kcal/mol of experimental values for G3 and Complete Basis Set (CBS) composite energy methods, and reaction rate constants can be accurately.Also, in this work, quantum chemical methods have been used to predict catalytic conversion reactions of light alkanes including methane, ethane, propane, and iso-butane on zeolite surface. A silicon free cluster model and an aluminosilicate cluster model containing three tetrahedral (Si, Al) atoms (T3 cluster) was applied to investigation reaction pathways and energetics. The activation energies were obtained and compared with available experimental data. We find that the activation energy is a strong function of zeolite acidity and the relationships of the activation energy as a function of acid strength were also investigated by changing the terminal hydrogen bond length.This work not only allows for a more thorough understanding of the hydrocarbon reactions which is of high importance of petroleum and combustion industry, but also offers a reliable tools to guide the engineering reactor design which sometime cannot be achieved through direct experimental studies.