| A computational chemistry study was done to investigate several aspects of catalytic cracking models and mechanisms. The C--C bond properties of n-alkanes of varying lengths are investigated, with relevance to initiation steps. Two catalytic cycles were investigated to study the cracking of hexane to yield propane and propene, using Bronsted-acid and Lewis-acid catalysts and including hypothesized initiation steps. Zeolites and ionic liquids are cracking catalysts that are difficult to model, and hence some aspects of modelling zeolites and ionic liquids are computationally considered, including probable species present in one important class of ionic liquids, and effects of differing zeolite fragment models upon proton affinity.; Examination of radical formation suggests that the C2H 5 is formed preferentially over other alkyl radicals. Proton affinity results for C--C bonds show that proton affinity is not constant and rises with increasing carbon chain length and progression towards the center bonds of an alkane. The Bronsted catalytic cycle investigation demonstrates that the existence of particular intermediates depends primarily on the relative difference in proton affinities of the catalyst and the reacting alkane. The Lewis catalytic cycle, using Al2Cl7- as catalyst, has a large activation barrier, which is partially due to the generation of two incompletely solvated anions. The thermodynamics of the x Pyr + y HCl + z AlCl3 results suggest PyrH+, Al2Cl7- and AlCl4 - are the predominant species in these mixtures, with extra HCl units complexed to the anions. The proton affinities of several zeolite fragment models indicate that they are mild acids and suggest that additional modelling still needs to be done. |
