SO2 Poisoning of Ceria

It has been suggested that ceria influences the steam-reforming and WGS reactions through a redox mechanism similar to that shown below for WGS [22,26]  

a represents an adsorption site on the metal catalyst. The key point in this mechanism is that ceria is oxidized by H2O, allowing the oxygen from steam to be used in the oxidation of CO or other reducing agents. 

There is significant support for this mechanism, both from kinetic and spectroscopic data. For example, the oxidation of CO adsorbed on a precious metal by oxygen from the ceria has been observed to occur below 400 K in temperature-programmed-desorption (TPD) measurements [27]. Indeed, recent spectroscopic data has shown that Pd particles are oxidized by their ceria-zirconia support, beginning at 470 K [28], Evidence for the other important step in the reaction, the oxidation of reduced ceria by steam, has also been presented [29]. Finally, the kinetic rate expression for the WGS reaction over ceria-supported precious metals, in which the reaction is zeroth order in CO, agrees with expected rate expression for the mechanism shown above [29]. 

If one accepts the above mechanism for the WGS reaction, poisoning of the catalyst by SO2 must either block the oxidation of reduced ceria by water (Reaction (2)) or decrease the rate at which adsorbates on the metal can be oxidized by ceria (Reaction (3)). In either case, it is necessary to understand how SO2 interacts with ceria in any attempt to synthesize sulfur-tolerant catalysts. 

---

However, for the relatively low sulfur concentrations to which three-way catalyst are typically exposed, the precious metals are usually unaffected. Poisoning of ceria is much more serious than poisoning of the precious metal at the levels of 5 to 20 ppm SO2 currently present in the typical automotive exhaust. At these concentrations, SO2 interacis primarily with the ceria-containing component of in the catalytic converter and it is this poisoning of ceria that appears to be the primary problem associated with sulfur poisoning [6-11]. The evidence for this is strong. For... 

Clearly, understanding the interaction between SO2 and ceria is extremely important in three-way. emissions-control catalysis. This chapter will review what is known about the nature of sulfur poisoning of ceria and how this affects the catalytic properties. 

In the absence of SO2, Pt supported on silica and alumina show about the same activity for CO oxidation (blue squares) Upon addition of 20 ppm of SOj in the gas feed, the ignition temperature increases only 5°C during the first run in the alumina-supported sample, compared to 3UC in the case of the silica-supported catalyst. Because both samples have similar metal dispersion, the differences in LOT can only be attributed to the nature of the supports. In the literature, it has been proposed that the presence of SOj leads to the formation of sulfates on the alumina surface, which thus becomes a sulfur trap. This would allow the Pt surface to appear not to be affected as strongly as in the case of the silica support in this short time on stream (T-O-S) run. The increase in LOT is relatively small for the ceria-promoted alumina catalyst, probably for the same reason as for unpromoted alumina. Addition of cerium oxide to Pt/silica significantly decreases the LOT (by 26°C compared to Pt(0.63)/SiO2). In the presence of S, however, the activity is the same as that of the Pt without Ce, indicating poisoning of the Ce promotional effect. 

In a more recent demonstration that titration methods do not properly account for OSC, it has been shown that Pd/ceria and Pd/ceria-zirconia catalysts are capable of reversibly releasing more oxygen after they have been poisoned with SO2 than they are before being poisoned with SOj [17]. An example of this is given in the pulse data from a Pd/ceria-zirconia catalyst at 773 K, reported in Figure I l.l, with data for the unpoisoned catalyst shown on left and data for the poisoned catalyst on the right. 

Also, Marsh and co-workers [145] showed that gold on cobalt oxide particles, supported on a mechanical mixture of zirconia-stabilised ceria, zirconia and titania remains active in a gas stream containing 15 ppm SO2. Haruta and co-workers [207] found that although the low-temperature CO oxidation activity of Ti02-supported Au can be inhibited by exposure to SO2, the effect on the activity for the oxidation of H2 or propane is quite small. Venezia and co-workers [208] reported that bimetallic Pd-Au catalysts supported on silica/alumina are resistant to sulphur poisoning (up to 113 ppm S in the form of dibenzothiophene) in the simultaneous hydrogenation of toluene and naphthalene at 523 K. 

Only surface sulfate groups are observed on zirconia by SO2 oxidation, even in presence of platinum, while bulklike species are also formed on Ce02-Zr02 mixed oxides, with or without platinum, as on pure ceria (3,4). Then the sulfate poisoning can be as important on Pt/Ce02 -Zr02 catalysts as on Pt/Ce02. 

SOFC anodes Ni-gadolininm doped ceria (Ni/GDC) are studied for sulfur poisoning under operando conditions by Nurk et al. [62], The molecular structure of sulfur species formed on the anodes in the temperature range 250-550 °C is studied by XANES (K shell). With Fl2 fuel containing 5 ppm H2S, several sulfur species in different oxidation states (6-1-, 4- -, 0, —2) are detected the species could either relate to -S04 or SO3 (g), -S03 or SO2 (g), S2 (g) or surface-adsorbed S atoms, and Ni or Ce snlfides. These results do not agree with thermodynamic phase calculations, in particnlar the formation of sulfate species is not expected at the highest temperatures this can be related to the difference between eqnilibrium conditions and the steady-state conditions in the fuel ceU that are determined by kinetic-controlled processes. 

---