Long-term catalyst deactivation

Rahimpour MR, Bayat M, Rahmani F Dynamic simulation of a cascade fluidized-bed membrane reactor in the presence of long-term catalyst deactivation for methanol synthesis, Chem Eng Set 65 4239—4249, 2010. 

Table 8.4 contains the deactivation parameters estimated from the long-term catalyst stability test and the feedstock evaluation with HCO. d and d are the parameters of the hyperbolic function that describes the initial activity decay caused by coke formation, whereas 3 is the exponent of the power-type function that represents the slow deactivation process by metal deposition (see Equation 8.21). Each set of... 

FIGURE 8.9 Comparison between experimental deactivation data and model predictions, (a) Long-term catalyst stability test (b) Feedstock evaluations with HCO (—) 5%. 

Catalyst Selection. The choice of catalyst is one of the most important design decisions. Selection is usually based on activity, selectivity, stabiUty, mechanical strength, and cost (31). StabiUty and mechanical strength, which make for steady, long-term performance, are the key characteristics. The basic strategy in process design is to minimize catalyst deactivation, while optimizing pollutant destmction. 

The long term tests in the SASOL plant as well as in the Schwechat plant were run with outlet temperatures of 450°C, but both plants were also operated with higher loads that caused reactor outlet temperatures of 470°C or even higher. In comparison with the test run at 450°C, only a slight increase in deactivation rate was detectable which demonstrates the thermostability of the catalyst. From the aspect of thermostability, outlet temperatures of 450°-470°C are acceptable. Further considerations including the possibility of overload operation, the SNG specification to be achieved in final methanation, end-of-run conditions, and cost of reactor material will affect the selection of optimum outlet temperature. 

The long-term stability of the Ru/Ti02 catalyst was studied under various reaction conditions and the spent catalysts were characterized for assessing the reasons of deactivation. It was observed that the rate exhibits a rapid reduction at the initial several hours of reaction, followed by a slow and continuous deactivation. Analysis of the spent catalyst, by H2 adsorption after removing surface carbon, showed that the initial rapid reduction of activity is mainly due to metal sintering, while the continuous and slow deactivation is related to the occurrence of the SMSl phenomenon at the later part of the catalyst bed, where reducing conditions prevail. In order to avoid these processes which lead to catalyst deactivation, Ti02... 

The activity and stability of catalysts for methane-carbon dioxide reforming depend subtly upon the support and the active metal. Methane decomposes to carbon and hydrogen, forming carbon on the oxide support and the metal. Carbon on the metal is reactive and can be oxidized to CO by oxygen from dissociatively adsorbed COj. For noble metals this reaction is fast, leading to low coke accumulation on the metal particles The rate of carbon formation on the support is proportional to the concentration of Lewis acid sites. This carbon is non reactive and may cover the Pt particles causing catalyst deactivation. Hence, the combination of Pt with a support low in acid sites, such as ZrO, is well suited for long term stable operation. For non-noble metals such as Ni, the rate of CH4 dissociation exceeds the rate of oxidation drastically and carbon forms rapidly on the metal in the form of filaments. The rate of carbon filament formation is proportional to the particle size of Ni Below a critical Ni particle size (d<2 nm), formation of carbon slowed down dramatically Well dispersed Ni supported on ZrO is thus a viable alternative to the noble metal based materials. 

Stable performance was demonstrated to 4,000 hours with Nafion membrane cells having 0.13 mg Pt/cm and cell conditions of 2.4/5.1 atmospheres, H2/air, and 80°C (4000 hour performance was 0.5 V at 600 mA/cm ). These results mean that the previous problem of water management is not severe, particularly after thinner membranes of somewhat lower equivalent weight have become available. Some losses may be caused by slow anode catalyst deactivation, but it has been concluded that the platinum catalyst "ripening" phenomenon does not contribute significantly to the long-term performance losses observed in PEFCs (5). 

Prior to the kinetic experiments, possible deactivation phenomena of the catalytic system were checked by recycling experiments with prenal and citral as substrates. These results provide not only important hints on the form of the rate equation, but also on which reaction is convenient for long-term investigations in the loop reactor. After the reaction, the aqueous and organic phases were separated and the catalyst phase was reused without further purification. Results on the hydrogenation of prenal are shown in Fig. 7. The reaction rate clearly decreases if the catalyst phase is reused. According to GC analysis and H-NMR studies, this can be attributed to the fact that the product of the reaction, prenol, is highly soluble in water. Consequently, a simple phase... 

Although the standard Rh(I)/tppts catalyst has extremely good long-term stability, a slight loss of activity occurs after several years use. The deactivation mechanism of the Ruhrchemie/Rhone-Poulenc catalyst has been clarified in detail (63) and is presented in Scheme 5. 

The low-pressure methanol synthesis process utilizes ternary catalysts based on copper, zinc oxide, and another oxide, such as alumina or chromia, prepared by coprecipitation. Cu-Zn0-Al203 and Cu-Zn0-Cr203 are usually the most important industrial catalysts. A significant advance was made when a two-stage precipitation was suggested in which ZnAl2C>4, a crystalline zinc aluminate spinel, was prepared prior to the main precipitation of copper-zinc species.372 This alteration resulted in an increase in catalyst stability for long-term performance with respect to deactivation. Catalyst lifetimes industrially are typically about 2 years. 

Schultze pointed out (18) that whenever the enhanced catalytic activity of the catalyst is due to so-called active sites, that is, exposed crystal defects or dislocations, these sites will only be active long term if processes that would lead to healing or recrystallization and accordingly to deactivation have an activation energy in excess of 100 kJ/mol. Such high activation energies would yield at 100°C a retardation of any surface relaxation processes to effective relaxation times in excess of several years. 

A collection of 66 catalyst combinations was operated for 24 h to check the long-term behavior. Fig. 4.2 demonstrates that all combinations show deactivation, especially the most active catalysts. This clearly demonstrated that screening also depends very much on the selection of the appropriate time scale for a reaction (in this study at least 24 h). 

The catalyst support impacts the rate of a catalyzed reaction, the reaction pathway (quantities and species of intermediates and products) and the resistance of the catalyst to deactivation. In DBCP reactions, powders had higher rate constants than beads, presumably due to reduced mass transfer limitations alumina yielded a faster rate than C, which had a faster rate than PEI/silica. Sorptive capabilities of the supports may also play an important role Kovenklioglu found that supports which sorbed 1,1,2-TCA more strongly had higher reaction rates, and Farrell concluded that TCE sorption to Fe cause higher reaction rates on Pd/Fe electrodes than on pure Pd electrodes. It is also clear that supports influence reaction products, but the correlation between a given support and pathways/products it promotes is not yet understood. The choice of support can also affect its resistance to deactivation this implies that catalyst supports may be tailored to maximize activity over the long term. 

Reaction conditions investigated in long-term deactivation runs are shown in Table 4. The chromia/alumina catalyst was studied... 

---