Sulfur deactivation, methanation

Reaction Studies. Despite the large body of work on the effect of sulfur on Ni-based and precious metal catalysts for reforming of methane, there is relatively little work on sulfur deactivation in the reforming of liquid fuels. 

The space velocity was varied from 2539 to 9130 scf/hr ft3 catalyst. Carbon monoxide and ethane were at equilibrium conversion at all space velocities however, some carbon dioxide breakthrough was noticed at the higher space velocities. A bed of activated carbon and zinc oxide at 149 °C reduced the sulfur content of the feed gas from about 2 ppm to less than 0.1 ppm in order to avoid catalyst deactivation by sulfur poisoning. Subsequent tests have indicated that the catalyst is equally effective for feed gases containing up to 1 mole % benzene and 0.5 ppm sulfur (5). These are the maximum concentrations of impurities that can be present in methanation section feed gases. 

Agrawal et al.33 performed studies of Co/A1203 catalysts using sulfur-free feed synthesis gas and reported a slow continual deactivation of Co/A1203 methanation catalysts at 300°C due to carbon deposition. They postulate that the deactivation could occur by carburization of bulk cobalt and formation of graphite deposits on the Co surface, which they observed by Auger spectroscopy. 

Agrawal, P. K., Katzer, J. R., and Manogue, W. H. 1981. Methanation over transition metal catalysts. II. Carbon deactivation of cobalt/alumina in sulfur-free studies. J. Catal. 69 312-26. 

This decrease of H2 and CO, coupled with an increase in methane concentration, is attributed to deactivation of the shift catalysts and increased met-hanation reaction. However, this does not explain the increase in CO2 concentration, and it is difficult to envision that the addition of sulfur would increase the methanation activity. The reformate produced in the ATR step was not analyzed, making it difficult to interpret these results. 

Methanation catalysts are not usually deactivated by thermal sintering. The principal reason for any loss of activity is poisoning. Sulfur compounds will poison methanation catalysts, but sulfur is not present unless the low temperature shift catalyst is by-passed. The poisons most likely to occur under normal operating conditions are those originating from the carbon dioxide removal system that precedes the methanator. Carry-over of a small amount of liquid into the methanator is not serious. Large volumes of liquid will have a... 

On the other hand, markedly different behavior was observed during reaction for H2S adsorption on Ni 140) and for S02 adsorption on Pt (81, 131). Erekson and Bartholomew (140) observed that during methanation of CO on polycrystalline Ni, the number of sulfur atoms required to deactivate one nickel atom decreased from 1.0 to 0.23 as the mol % of CO (H2 diluent h2s — 0-2 ppm) was increased from 2 to 20%. During those experiments H2S and COS appeared in the exit stream when the CO concentration was 5% or above. At 20 mol % CO only 1/4 of the reactant H2S (and apparently none of the COS) adsorbed on the catalyst, suggesting that adsorbed CO or carbon inhibits H2S adsorption. 

Most of the previous studies mentioned to this point did not meet the experimental requirements outlined in Section V,A and involved H2S concentrations in the ppm range. In view of the previously discussed adsorption studies showing that reversible adsorption of H2S occurs only at ppb levels under typical methanation conditions, it is reasonable to expect that full saturation coverage of sulfur occurred on the catalysts used in these previous studies hence their steady-state activities should not vary significantly for a given metal such as Ni. To allow quantitative measurements of the rates of catalyst deactivation, of the steady-state activity in the presence of ppb concentrations of H2S, and to determine the catalytic activity as a function of surface coverage by sulfur, Katzer and co-workers (99-101,147, 205-208), used a reactor and catalyst configuration which satisfied all the... 

The turnover numbers and activation energies for methanation over all four metals freshly reduced and under sulfur-free conditions (Table XVI) were in good agreement with values reported for supported metal catalysts (220). At 673 K, Ni and Ru exhibited only very slow losses in activity apparently due to slow carbon deposition, whereas Co and Fe underwent rapid, severe carbon deactivation after maintaining their fresh catalyst activity for a few hours. After rapid deactivation the final steady-state activity, which was about 100-fold lower than the activity of the fresh catalyst, was approached slowly this activity region was referred to as the lower pseudosteady state. Likewise, the fresh catalyst behavior was referred to as the upper pseudo-... 

Comparison of the methanation activity in Fig. 28 with the H2S concentration transients in Fig. 29 indicates a prolonged adsorption of H2S during the 13-ppb deactivation with the H2S level reaching a steady state at about the same time as the methanation activity. The difference between the sulfur fed to the reactor and that in the effluent represents the sulfur that was adsorbed on the Ni surface, since the alumina support and the quartz reactor did not adsorb sulfur. Time integration of the amount of sulfur adsorbed, obtained from the effluent H2S concentration data, gives the total amount of sulfur adsorbed at equilibrium with the gas-phase H2S concentration. 

That the activation energies and rate concentration dependences for methanation over sulfur-poisoned Ni and Ru are the same as those observed under sulfur-free conditions (99-100,140,147,209) provides further evidence that sulfur poisoning of Ni and Ru involves geometric blockage of active metal sites. However, the observations that (i) the activation energies for methanation over sulfur-poisoned and carbon-deactivated Co are both 50 kJ/mol lower than that for fresh Co and (ii) the rate dependence on CO partial pressure is positive order for both carbon-deactivated Co and sulfur-poisoned Co, suggest that the deactivations of Co by sulfur and carbon are similar and may involve electronic as well as geometric effects. 

It must be emphasized that high purity syngas is essential for the methanol synthesis catalyst, which is easily poisoned or deactivated by components such as chlorides, sulfur and heavy metals, i.e., mercury. Also, large quantities of inerts, namely, nitrogen and methane, which have no effect on the methanol catalyst, can reduce the overall carbon efficiency which will increase the amount of syngas required per unit make of methanol. 

Among the different species that natural gas can contain, sulfur compounds can have a dramatic effect on the catalytic activity. Pd is known to be extremely sensitive to even very small amounts of sulfur. Nevertheless, the deactivation rate in presence of sulfur species in the feed can be reduced when the metal particles are impregnated onto a sulfating support [10]. In that case, the support acts as a trap for the sulfur species. The sulfates are preferably formed on the support material and the palladium particles are then protected. The formation of Pd sulfate is reported to be responsible for the loss of catalytic activity in total oxidation of methane [15], the active species being PdO. 

The purpose of the present paper is to investigate the behavior of high temperature stable support in complete oxidation of methane, when sulfur species are present in the feed. Various characterization methods have been employed here in order to give a general picture of the deactivation of supported Pd catalysts in catalytic combustion of methane. This work will examine the reactivity of supported Pd catalysts, when the metal particles are dispersed over a thermally stable material, containing rare earth elements. Moreover, we will focus on the behavior of the catalysts when sulfur species are added to the feed, as sulfur can be present in significant amounts in natural gas. 

---