Coke burnoff in a typical FCC particle analyzed by an SEM mapped 2-D network pore structure

The pore morphology of a selected (congruent to 70 mum) fluid catalytic cracking (FCC) catalyst particle, as viewed on a SEM image of sectioned surface of low melting point alloy (LMPA) impregnated FCC catalyst sample, was mapped onto an approximately 'equivalent' 2-D 30 x 30 pseudo-random pore network to investigate the consequences for burnoff characteristics in the particle. The mapped 2-D structure was constructed from a set of pores having a psd close to that given by BJH adsorption results by preferentially assigning larger macropores to corresponding apparent locations in the SEM image while the rest of the meso- and micro-pores were then randomly assigned across the rest of the network. Major parameters such as pore structural configuration, residual coke, and spatial oxygen profile were tracked as a function of time. The model results were compared to experimental coke burnoff data, carried out in a micro-scale reaction unit. The model prediction fits fairly well within the experimental fluctuations. The results also showed that over 90% coke is burnt off within the first 50% of reaction time. For comparative purposes, the model predictions were also applied to a 'shuffled' structure, produced by random permutation of the same radii of the mapped structure as well as an 'optimally sparsed' structure. Burnoff in the shuffled structure exhibited the worst performance. This was attributed to the decrease in the proportion of larger pores on the network skin in contrast to the mapped and optimally sparsed structures. Since Visual inspections of SEM images of most particles tend to reveal a randomly oriented pore configuration (whereas the mapped structure was for a selected particle), the predicted aggregate performance of the FCC catalyst sample studied would be expected to lie somewhere in between the curves of the shuffled and mapped structures. The support pore architectural design of those particles could, therefore, still be improved for better performance in terms of accessibility, reactivity and selectivity. (C) 2001 Elsevier Science Ltd. All rights reserved.