reÑectance spectrum of the sulfur-treated catalyst shows the
When comparing the global activities of sulfur-treated cata-
lysts, alumina loaded with chromium alone had the same
activity as the mixed catalyst. Thus the minute amounts of
palladium which were not deactivated by sulfur adsorption,
were sufficient to ensure complete oxidation of CO. An extra
experiment performed over 60 h with a feed containing hydro-
gen sulÐde showed that almost no deactivation was observed.
The similar catalytic behaviour found for Cr/Al O and
almost complete reduction of CrO 2~ ions to Cr3`. This
4
reduction is almost unobservable in the same measurement
performed without H S. Both the position and intensity of
2
bands seem to correspond to a bulk oxide containing chro-
mium ions rather than to isolated Cr3` ions spread out onto
the alumina surface. Previous work performed on chromium
mixed oxides, such as perovskite29 and alumina doped with
Cr3` ions,8 has shown that UV band positions are slightly
sensitive to the structure in which chromium is involved.
Along with reduction of chromate to Cr3` species hydrogen
sulÐde is expected to be oxidized with the formation of sulfur
oxides probably adsorbed as sulfate on the surface of the cata-
lyst. Cr3` is known to be acidic and a rather mild oxidizer.
However, its activity is generally lower than that of the other
transition-metal oxides like Co O or MnO .37,38 It seems
2
3
PdÈCr/Al O catalysts strongly suggests that the same active
2
3
phases are present at the surface of the solids.
The di†use reÑectance spectrum of the mixed catalyst is also
similar to the spectrum of the Cr/Al O sample. Almost all
2
3
the chromate ions have been reduced to Cr3`, whereas palla-
dium species were undetectable.
As far as selectivity towards CO is concerned, the intro-
duction of small amounts of palladium is sufficient to com-
pletely suppress the formation of carbon monoxide, meaning
that palladium is not totally poisoned. As in the case of the
fresh catalyst, palladium is probably in a strong interaction
with the chromium-containing phase, i.e., Cr3` or chromium
oxide. Such chromium species possess an acidic character and
are able to withdraw electron density from palladium oxide
particles. As for the Pd/Al O catalyst H S oxidation leads to
3
4
2
that the acidic properties of Cr3` ions induces dehydroge-
nation or decomposition of intermediate species leading to
CO formation. We can suppose that catalytic combustion,
occurring at fairly low temperature proceeds on the surface via
a partially oxidized species, e.g. formate. If no further oxida-
tion proceeds decomposition of these formate species before
total oxidation may lead to CO.39,40 Cr3` has not been
reported as a very efficient combustion catalyst8 and minute
amounts of this species, not detected either by DRS or by
XPS, may be responsible for the production of CO at low
temperature: at higher temperature chromate are sufficiently
active in removing CO by oxidation. When traces of sulfur are
present in the feed a signiÐcant reduction of chromate occurs
limiting the performance in combustion, whereas the presence
of excess Cr3` leads to partial transformation of methane into
CO. In that sense deposition of Cr3` onto an alumina may
appear as a methane reforming catalyst.
2
3
2
the formation of SO and then to sulfate groups linked to
x
palladium surface atoms. Because of the electronic interaction
previously postulated in bimetallic catalysts and evidenced by
XPS, the decrease in electron density of palladium atoms must
be associated with a decrease in the strength of the palladiumÈ
sulfur bond. As a consequence, sulfate groups bonded to pal-
ladium are less stable when the noble metal is supported by
chromium oxide.
Conclusion
The calcination of Cr/Al O in air leads to the formation of
Pd–Cr/Al O solid
2
3
2
3
supported chromate species CrO 2~ and not to Cr O . This
4
2 3
When discussing the activity of catalysts in state 1, the conver-
sion vs. temperature gives a curve almost identical to that
observed with the Cr/Al O catalyst. All the solids containing
prevents any Cr3` formation avoiding further ion migration
inside the bulk of alumina. The solid obtained is very active in
methane combustion. The formation of CO, probably by a
side-reaction such as decomposition of partially oxidized
species strongly limits the use of such a catalyst in catalytic
combustion.
2
3
chromium present the same catalytic activity and the same
DRS spectrum, suggesting that similar species are present at
the same concentration. STEM analysis leads to similar con-
clusions when the analysis was performed on an alumina
crystal. The catalyst activity is that of the most active species,
i.e., chromate ions. This is also demonstrated by the value of
activation energies which are comparable to the value of
Cr/Al O , and very di†erent from the value of Pd/Al O
Introduction of H S in the feed leads to (i) a slight decrease
2
in the activity and (ii) the formation of large amount of CO at
low temperature. The catalyst shows some resistance towards
poisoning by sulfur, but partial oxidation occurs owing to the
transformation of CrO 2~ into Cr3`, which is less active.
2
3
2 3
4
(
Table 2). In addition the activities of the catalyst as well as
Catalysts consisting of palladium supported on alumina are
the apparent energies of activation are similar in state 1 and in
state 2 and close to those of Cr/Al O .
The fundamental di†erence is the absence of CO because
palladium allows the oxidation of CO into CO . The Ðrst
conclusion that can be deduced is that palladium supported
efficient for the combustion of methane. Even if palladium is
in a concentration as low as 500È1000 ppm (wt.%) supported
on alumina, such catalysts are able to enhance the combustion
2
3
of methane. In the presence of H S, oxidation of reduced
2
2
sulÐde leads to sulfate adsorbed on ionic palladium species
on Cr/Al O is a very efficient catalyst for carbon monoxide
and is responsible for a strong deactivation of the catalyst.
When palladium is deposited on a chromium-covered
alumina solid a new catalyst is obtained. The global activity,
which is due to the major species (chromium based), is
unchanged but now the selectivity is entirely in favour of
2
3
combustion. These bimetallic catalysts mainly consist of a
mixture of ionic species: well dispersed chromate ions and sin-
tered palladium oxide particles.
After a run up to 100% conversion at 873 K the bimetallic
catalyst shows no activation in contrast to what is observed
for palladium-supported alumina. Indeed a slight deactivation
is observed, mirroring to the behaviour of CrO 2~-supported
alumina. The non-observation of any activation suggests that
CO , meaning that the palladium is not totally poisoned. This
2
mixed solid constituted by palladium and chromium oxide
appears to be a valuable active phase for methane and natural
gas combustion.
4
reconstruction of palladium does not occur. On the contrary,
palladium is present in the form of large PdO particles. The
reconstruction postulated for small metallic Pd particles
deposited on alumina does not occur in the case of
PdÈCr/Al O because of the large size of the palladium oxide
References
1
2
D. L. Trimm, Appl. Catal., 1983, 7, 249.
R. Prasad, L. A. Kennedy and E. Ruckenstein, Catal. Rev. Sci.
Eng., 1984, 26, 1.
2
3
crystallites.
In the presence of H S in the feed T was increased by ca.
3
4
R. B. Anderson, K. C. Stein, J. J. Feenan and L. J. E. Hoger, Ind.
Eng. Chem., 1961, 53, 809.
2
50
7
0 K, i.e. by the same value as for the Cr/Al O
3 catalyst.
G. J. Firth and H. B. Holland, T rans. Faraday Soc., 1969, 65, 11.
2
2222
J. Chem. Soc., Faraday T rans., 1997, V ol. 93