Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Chemical and Biochemical Engineering


Hugo I. de Lasa


Propane oxidative dehydrogenation (PODH) was studied using VOx/γAl2O3 and VOx/ZrO2-γAl2O3 (1:1 wt.%) catalysts, as well as consecutive propane injections under oxygen-free conditions. These catalysts were synthesized with 2.5, 5 and 7.5 wt.% vanadium loadings, and prepared using a wet saturation impregnation technique. Different characterization techniques were used to establish catalyst properties including NH3-TPD, pyridine FTIR and NH3-TPD kinetics. As well, PODH runs in the CREC Riser Simulator were developed under oxygen free atmospheres at 500-550°C, close to 1 atm., 10-20 s and 44.0 catalyst/propane weight ratio (g/g). Propylene selectivity obtained were up to 94%, at 25% propane conversion.

Using this data, a “parallel-series” model was established based on a Langmuir-Hinshelwood rate equation. Adsorption constants were defined independently, with this leading to a 6-independent intrinsic kinetic parameter model. These parameters were calculated via numerical regression with reduced spans, for the 95% confidence interval and low cross-correlation coefficients. A larger 2.82×10-5 mol.gcat-1s-1 frequency factor for propylene formation versus the 1.65×10-6 mol.gcat-1s-1 frequency factor for propane combustion was obtained. The calculated energies of activation (55.7 kJ/mole for propylene formation and 33.3 kJ/mole for propane combustion) appeared to moderate this effect, with the influence of frequency factors prevailing. Furthermore, propylene conversion in COx oxidation appeared as a non-favored reaction step, given the 98.5 kJ/mole activation energy and 4.80×10-6 mol.gcat-1s-1 frequency factor.

This kinetic model was considered for the development of a scaled-up twin fluidized bed reactor configuration. For this, a hybrid computational particle-fluid dynamic (CPFD) model featuring either “Particle Clusters” or “Single Particles” was employed. Results obtained in a 20-m length downer unit showing a 28% propane total conversion and a 93% propylene selectivity using the “Single Particle” model. However, and once the more rigorous particle cluster flow was accounted for, propane conversion was limited to 20%, with propylene selectivity staying at 94% level. Thus, the obtained results show that a PODH simulation using CPFD requires one to account for “Particle Clusters”. This type of comprehensive model is needed to establish unambiguously the PODH downer reactor performance, being of critical value for the development of down-flow reactors for other catalytic processes.