(324b) Optimal Design of Hierarchically Structured Deactivation Resistant Catalysts

The yield of chemical reactions in nanoporous catalysts could be significantly increased by introducing a uniform distribution of broad pore channels or macropores of the same size [1-3]. Detailed optimization studies of these hierarchically structured catalysts showed that a distribution of macropore sizes does not increase the yield, when compared to catalysts with constant macropore size. For arbitrary kinetics in the nanoporous catalyst, there is an optimal macroporosity and an optimal channel size, leading to a maximum yield. For deNOx [3] and autothermal reforming, significant improvements in yield were demonstrated.

However, these studies did not include the effects of catalyst deactivation, which is of great industrial importance. Therefore, the optimal structure of hierarchically structured catalysts for reactions affected by deactivation was determined. Initially, this study has focused on determining whether a bidisperse or a bimodal catalyst is optimal, compared to a conventional monodisperse catalyst. To this end, geometry based optimizations of the broad pore channel network, employing a continuum approach are performed with the macroporosity of the broad pore channels and the diameter of the channels as the input variables. The integrated scaled yields over time from the optimum bidisperse and bimodal structures are compared with the monodisperse catalyst to decide if significant improvements have been obtained. A case study involving deactivation of hydrodemetalation catalysts is used to demonstrate this approach.

Investigations are also aimed at comparing uniform pore networks with fractal networks to see if deactivation is significantly affected by the type of macropore network. Gavrilov and Sheintuch [4] have demonstrated the superiority of fractal pore network architecture over uniform pore network architecture for positive order reactions in the Knudsen diffusion regime. Villermaux and co-workers [5] have simulated a Devil's comb in two dimensions as possible macropore architecture for hierarchically structured catalysts. They accounted for deactivation by poisoning, and their results indicate that fractal distributor pore networks may hold certain advantages over non-fractal distributor networks. Compared to a uniform channel architecture, fractal catalysts may contain a larger internal surface area (depending on the generation of the fractal object), which can still be accessed by the reactant of interest, and which will also require a longer time to be completely deactivated. Therefore, fractal distributor pore networks are interesting candidates for the optimal design of deactivation resistant catalysts.

This methodology can be applied to design better deactivation resistant catalysts, for example, in hydrodemetalation catalysts which get deactivated due to both coking and metal sulfide deposition, or in deNOx catalysis where poisoning due to impurities in the flue gas stream such as Arsenic Oxide (As2O3) may be significant [6].

References:

[1]Wang, G.; Johannessen, E.; Kleijn, C. R.; de Leeuw, S. W.; Coppens, M. O. Optimizing Transport in Nanostructured Catalysts: A Computational Study. Chem. Eng. Sci. 2007, 62, 5110-5116.

[2]Johannessen, E.; Wang, G.; Coppens, M. O. Optimal Distributor Networks in Porous Catalyst Pellets. I. Molecular Diffusion. Ind. Eng. Chem. Res. 2007, 46, 4245-4256.

[3] Wang, G.; Coppens, M. O. Calculation of the Optimal Macropore Size in Nanoporous Catalysts and Its Application to DeNOx catalysis. Ind. Eng. Chem. Res.2008, 47, 3847-3855.

[4] Gavrilov, C.; Sheintuch, M. Reaction Rates in Fractal vs. Uniform Catalysts with Linear and Nonlinear Kinetics. AIChE J. 1997, 43, 1691-1699.

[5] Mougin, P.; Pons, M.; Villermaux, J. Reaction and Diffusion at an Artificial Fractal Interface: Evidence for a new Diffusional Regime. Chem. Eng. Sci. 1996, 51, 2293-2302.

[6] Beeckman, J. W.; Hegedus, L. L. Design of Monolith Catalysts for Power Plant NOx Emission Control. Ind. Eng. Chem. Res. 1991, 30, 969-978.