Catalyst deactivation pore mouth

Some studies of potential commercial significance have been made. For instance, deposition of catalyst some distance away from the pore mouth extends the catalyst s hfe when pore mouth deactivation occui s. Oxidation of CO in automobile exhausts is sensitive to the catalyst profile. For oxidation of propane the activity is eggshell > uniform > egg white. Nonuniform distributions have been found superior for hydrodemetaUation of petroleum and hydrodesulfuriza-tion with molybdenum and cobalt sulfides. Whether any commercial processes with programmed pore distribution of catalysts are actually in use is not mentioned in the recent extensive review of GavriUidis et al. (in Becker and Pereira, eds., Computer-Aided Design of Catalysts, Dekker, 1993, pp. 137-198), with the exception of monohthic automobile exhaust cleanup where the catalyst may be deposited some distance from the mouth of the pore and where perhaps a 25-percent longer life thereby may be attained. 

The two limiting cases for the distribution of deactivated catalyst sites are representative of some of the situations that can be encountered in industrial practice. The formation of coke deposits on some relatively inactive cracking catalysts would be expected to occur uniformly throughout the catalyst pore structure. In other situations the coke may deposit as a peripheral shell that thickens with time on-stream. Poisoning by trace constituents of the feed stream often falls in the pore-mouth category. 

If a fraction a of the total catalyst surface has been deactivated by poison, the pore-mouth poisoning model assumes that a cylindrical region of length (aL) nearest the pore mouth will have... 

These observations could be interpreted as a pore mouth catalysis. It was suggested that EU-1 fresh catalyst comprises two types of active sites, inner and external acid sites, the first ones which are non selective to isomerization and sensitive to deactivation, the second ones selective to isomerization but non sensitive to deactivation. Selectivity of inner acid sites could be estimated by difference between results obtained after 45 minutes and those obtained after 8 hours. These results are shown on table 3. 

Three obvious models which could describe the observed reaction rate are (a) concentration equilibrium between all parts of the intracrystalline pore structure and the exterior gas phase (reaction rate limiting), (b) equilibrium between the gas phase and the surface of the zeolite crystallites but diffusional limitations within the intracrystalline pore structure, and (c) concentration uniformity within the intracrystalline pore structure but a large difference from equilibrium at the interface between the zeolite crystal (pore mouth) and the gas phase (product desorption limitation). Combinations of the above may occur, and all models must include catalyst deactivation. 

On heating deactivated parent H-mordenite (80 to 33% cumene conversion), quantities of desorbate are so low (30.6 g/gram catalyst) that the desorbable deactivants, and hence the catalyst activity, must be at the pore mouth in the deactivated material. Non-desorbable polynuclear aromatics fill the mordenite tube. On the other hand, aluminum-deficient H-mordenite did not deactivate significantly for the same cumene treatment. Activity of this catalyst could be throughout the tube, but because of the disperse nature of the alumina sites, the high activity of parent H-mordenite, only active at its mouth, is not approached. 

A second example of metal transport by single-component gas involves the formation of metal carbonyls that are unstable in carbon monoxide. Shen et al. (72) discovered that supported nickel particles are not stable in carbon monoxide. Under certain conditions, CO combines with the metal in the particles to form volatile metal carbonyls, such as nickel carbonyl. These volatile species carry the metal out of the reactor, resulting in a rapid net loss of metal. In some cases, metal is not carried out of the reactor, but new metal particles form at pore mouths, blocking them and effectively deactivating the catalyst. 

Activity-versus-time curves shown in Fig. 25 for alumina-supported Ni and Ni bimetallic catalysts show two significant facts (1) the exponential decay for each of the curves is characteristic of nonuniform pore-mouth poisoning, and (2) the rate at which activity declines varies considerably with metal loading, surface area, and composition. Because of large differences in metal surface area (i.e., sulfur capacity), catalysts cannot be compared directly unless these differences are taken into account. There are basically two ways to do this (1) for monometallic catalysts normalize time in terms of sulfur coverage or the number of H2S molecules passed over the catalysts per active metal site (161,194), and (2) for mono- or bimetallic catalysts compare values of the deactivation rate constant calculated from a poisoning model (113, 195). 

Dautzenberg, F.M. van Klinken, J. Pronk, K.M.A. Sie, S.T. Wijffels, J.-B. Catalyst Deactivation through Pore Mouth Plugging during Residue Desulfurization presented at the 5th International Symposium on Chemical Reaction Engineering Houston, March 13-15,1978. 

The catalyst pore-mouth plugging can cause rapid deactivation. 

They considered deactivation to occur by either pore-mouth (shell-progressive) or uniform (homogeneous) poisoning and examined the effect these types of deactivation had on overall activity and production rates for a single catalyst pellet. Analytical solutions were obtained for the production per pore by considering the time dependence of activity. Their results will be used here as the basis for the development of models for deactivation in fixed bed reactors. 

Intermediate Catalyst Deactivation. Except the coke-controlled regime, an increase in the metal layer thickness in the pore mouth may be a major cause of the slow deactivation following the initial fast deactivation period as pointed out, though Figure 4 shows a slight decrease in the active sites due to metal poisoning during this period. 

Coke on the catalyst is, thus, largely responsible for catalyst deactivation by loss of surface area, and this could be minimized by increasing the hydrogen pressure. However, increasing pressure has been reported to increase vanadium deposition more near the exterior surface of the catalyst pellet (13,14). In essence, an increase in the hydrogen pressure has a beneficial effect in suppressing coke formation, but can lead to shorter catalyst life due to rapid accumulation of vanadium at pore mouths. 

Blocking the pore mouth and reducing the diffiisivities of the xylenes does not change this overall picture for toluene methylation, but enhances the p- selectivity [258]. As a negative side effect the catalysts deactivate and this has to be balanced with higher reaction temperatures. The higher reaction temperatures are required to open new reaction channels (dealkylation, transalkylation, disproportionation) to drain products fi om the pores as the longer residence times lead to polymethylated products that are unable to leave the zeolite pores and would eventually block all acid sites [258]. 

---