Low-temperature CO oxidation over Co3O4/Al2O3

Article abstract of DOI:10.1006/jcat.2000.2924

The oxidation of CO over Co₃O₄/γ-Al₂O₃ at 21°C was studied in a continuous-flow reactor system with oxidizing and reducing pretreatment, at varying stoichiometric ratios (S), and by using isotope labeled ¹⁸O₂. TPR and temperature-programmed oxidation were used to characterize the fresh catalyst. The preoxidized catalyst showed higher activity for CO oxidation than the prereduced catalyst. By raising the temperature to 50°C, no deactivation was observed over the preoxidized catalyst for S ≥ 2:1, which can be explained by the proposed reaction mechanism. When the CO oxidation proceeded, at least two forms of carbonaceous species were deposited on the catalyst, i.e., carbonates (both on the cobalt oxide and on the alumina support) and the graphite-like carbon.

Full text of DOI:10.1006/jcat.2000.2924

Journal of Catalysis 194, 55–60 (2000)
doi:10.1006/jcat.2000.2924, available online at http://www.idealibrary.com on
Low-Temperature CO Oxidation over Co3O4/Al2O3
Jonas Jansson1
Competence Centre for Catalysis, Department of Chemical Reaction Engineering, Chalmers University of Technology, S-412 96 Go¨teborg, Sweden
Received November 15, 1999; revised May 1, 2000; accepted May 5, 2000
Al2O3 was sieved to obtain a powder with particles between
125 and 210 m and a BET area of 165.4 m2/g. A 13.6 g por-
The oxidation of CO over Co3O4/ -Al2O3 at room temperature
(21 C) has been studied in a continuous-flow reactor system with
oxidising and reducing pretreatment, at varying stoichiometric ra-
tios, and by using isotope-labeled 18O2. The fresh catalyst was char-
acterised by temperature-programmed reduction and temperature-
programmed oxidation. CO deactivates the oxidised catalyst but the
rate of deactivation can be suppressed by having a high O2/CO ratio.
The isotope study showed that the oxygen participating in the oxi-
dation of CO comes from oxygen bound to the cobalt oxide surface.
Finally, a mechanism for the CO oxidation over Co3O4/ -Al2O3 is
c
tion of Co(NO3)2 6H2O of pro analysi grade from Merck,
Germany, was diluted in distilled water and 15.0 g of
-
Al2O3 powder was added to form a slurry. A 0.5 ml aliquot
of 25% NH3(aq) was added to raise the pH. The slurry was
freeze dried for 24 h, calcined at 550 C for 90 min, and then
ground to reobtain the original particle size. The BET sur-
face of the 20 wt% Co3O4/ -Al2O3 catalyst thus produced
was found to be 70.6 m2/g.
proposed.
2000 Academic Press
Reactor System
Key Words: CO oxidation; low temperature; cobalt oxide; isotope-
labeled oxygen; reaction mechanism; catalyst pretreatment; deacti-
vation.
The catalyst was tested in a continuous-flow reactor sys-
tem consisting of a vertical quartz tube reactor (6 mm
inner diameter) placed in an oven, where the tempera-
ture could be controlled between 20 and 600 C. The tem-
perature was measured with thermocouples upstream of
the reactor bed as well as in the reactor bed. The argon
was 99.9997% pure, O2, H2, and CO2 99.998% pure, and
CO 99.997% pure, all supplied by Air Liquide. The 18O2
was prepared by mixing 18O2 (isotope content 96.26% 18O)
from Larodan Fine Chemicals AB with 99.999% pure ar-
gon from AGA to obtain a 6 vol% 18O2/argon gas mixture.
The gases in the outlet of the reactor were detected by a
Balzers QMS 200 quadrupole mass spectrometer. M/e = 2
(H2), 28 (CO), 32 (16O2), 36 (18O2), 40 (Ar), 44 (C16O2), 46
(C16O18O), and 48 (C18O2) were followed, and the signals
were calibrated by flowing different concentrations of H2,
CO, O2, and CO2 through the empty reactor. The M/e = 28
signal was corrected for contribution from CO2 to obtain
the CO concentration.
INTRODUCTION
The low-temperature CO oxidation over Co3O4 and
other metal oxide-based catalysts has been studied by sev-
eral groups (1–9). Studies of the oxidation of hydrocarbons
over cobalt oxide-based catalysts also exist in the litera-
ture (10–13). Our group has previously observed that it
is only preoxidised cobalt oxide catalysts that show high
low-temperature activity (14). The activity over prereduced
cobalt oxide is much lower. Thorma¨hlen et al. (14) re-
ported a T50 (the reactor inlet temperature at which the
conversion of the reactant is 50% ) of 63 C for preoxidised
CoOx/Al2O3 compared to 157 C for the prereduced cata-
lyst. We have also seen that the surface layer of the reduced
catalyst is easily oxidised by O2(g) even at room tempera-
ture (14). The purpose of this work is to further understand
the mechanisms behind the low-temperature activity of the
Co3O4 catalyst and explain the difference in activity be-
tween the preoxidised and the prereduced catalyst.
Catalyst Testing
Six different types of experiments (see Table 1) were per-
formed:
METHODS
Catalyst Preparation
(1) Temperature-programmed reduction (TPR) and
temperature-programmed oxidation (TPO) repeated one
after the other several times on the fresh catalyst.
(2) CO oxidation over the prereduced and the
preoxidised catalyst, while heating it in a temperature
ramp.
-Al2O3 was prepared by calcining boehmite (AlOOH,
Disperal from Condea, Germany) at 600 C for 8 h. The
-
1 Fax: +46 31 772 30 35. E-mail: Jonas.Jansson@cre.chalmers.se. Web:
http://www.cre.chalmers.se/.
55
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

56
JONAS JANSSON
TABLE 1
Experimental Conditions
Experiment
Pretreatment
Inlet gas composition
Temperature
(1a) TPR
(1b) TPO
(2a) Light-off
Nonea/(1b)
(1a)
Reduction in 5% H2 at 550 C
Oxidation in 10% O2 at 550 C
Oxidation in 10% O2 at 550 C
5% H2
10% O2
21 C to 550 C at 10 K/min
21 C to 550 C at 10 K/min
21 C to 550 C at 10 K/min
21 C to 550 C at 10 K/min
21 C and 50 C
2% CO + 1.17% O2
2% CO + 1.17% O2
2% CO + x% O2;
x = 0.83, 1.0, 1.17, 1.5, 2.0,b 3.0
2% C16O + 1.17% 16O2
(2b) Light-off
(3) CO oxidation with different
oxygen concentrations
(4) CO oxidation over 18O-labeled
catalyst
(5) CO2 exchangec
(6a) Deactivation
(6b-1) TPD
Reduction in 5% H2 followed by
oxidation in 6% 18O2 at 550 C
Same as in (4)
Same as in (4)
(6a)
21 C
1.5% C16O2
21 C
21 C for 90 min
21 C to 550 C at 10 K/min
21 C to 550 C at 10 K/min
1.67% C16O + 0.98% 18O2
100% Ar
(6b-2) TPO
(6a)
10% O2
Note. All experiments were performed with 68 2.2 mg of Co3O4/Al2O3. The total gas flow was 20 nml/min and argon was used as balance.
a The first TPR was performed on the fresh catalyst.
b x = 2.0 only performed at 50 C.
c This experiment was also performed on /Al2O3 (BET = 165.4 m2/g).
(3) CO oxidation with several different oxygen concen- Fig. 1a) shows one large peak at about 432 C and a small
trations, over the preoxidised catalyst. The experiments broad peak at about 140 C. When the catalyst was oxi-
were performed at 21 C and at 50 C. (See Table 3.)
(4) CO oxidation at 21 C over a cobalt oxide catalyst shifted to 370 C (curve 3). The small broad peak at 140 C
preoxidised with isotope-labeled 18O.
was still present. The two subsequent oxidations followed
(5) Exchange of oxygen in gas-phase CO2 over a cobalt by TPR did not change the TPR profile substantially. The
oxide catalyst preoxidised with isotope-labeled 18O.
TPR curves were integrated to get the total H2 consump-
dised and the second TPR was recorded, the main peak
(6) Temperature-programmed desorption (TPD) and tion during the reduction. The result is presented in Table 2.
temperature-programmed oxidation (TPO) of the deacti- The TPO profile of the catalyst that has been reduced once
vated catalyst.
(curve 2in Fig. 1b) showsone large peak at about 245 C. The
two subsequent reductions followed by TPO did not change
the TPO profile, apart from a slight decrease in peak max-
imum (curves 4 and 6). The TPO curves were integrated
to get the total O2 consumption during the oxidation. The
result is presented in Table 2.
RESULT AND DISCUSSION
1. TPR and TPO of the Fresh Catalyst
The results from the TPR and TPO experiments are pre-
As seen from the area under the TPR graph, the cobalt
sented in Fig. 1. The TPR of the fresh catalyst (curve 1 in is only reduced to 55–62% of total reduction (Table 2).
FIG. 1. (a) Temperature-programmed reduction (TPR) and (b) temperature-programmed oxidation (TPO) of the fresh Co3O4/Al2O3 catalyst. The
H2 and O2 consumptions are defined as inlet minus detected outlet gas-phase mole fractions. The numbers indicate the order in which the experiments
were conducted.

LOW-TEMPERATURE CO OXIDATION OVER Co3O4/Al2O3
57
TABLE 2
sult isin contradiction to some previousresults(8, 13) where
the opposite was shown, but in agreement with newer re-
sults (14). The explanation for these contradicting results
may lie in the cobalt oxide catalyst’s high sensitivity to wa-
ter. Cunningham et al. (5) have shown that only 3 ppm of
water in the gas feed increases the T50 by about 100 K. In
the present experiments the gases were very pure and con-
tained less than this amount. Another explanation might
be the difference in volume flow rate between the present
experiments and the previous ones. Here the volume flow
rate (at standard conditions) per milligram of cobalt oxide
was about 70 times lower than that reported in (13). An
increase in flow rate will shift the T50 in the light-off experi-
ments to higher temperatures, and the shift will be different
for reduced and oxidised catalysts if the activation energies
are different.
Hydrogen Consumption in TPR and Oxygen Consumption in
TPO Calculated as Percentage Reduction of Co3O4 to Co0 for TPR
and as Percentage Oxidation of Co0 to Co3O4 for TPO
Difference
(Red-Ox)
TPR
Reduction (% )
TPO
Oxidation (% )
1
3
5
7
61.6
55.4
58.0
55.4
2
4
6
53.2
45.9
47.9
8.4
9.5
10.1
Note. The numbers 1–7 correspond to the run numbers in Fig. 1.
This corresponds to a mean oxidation state of 1.0–1.2 for
the cobalt after reduction at 550 C. To obtain reduction
to metallic cobalt, the temperature must be raised above
550 C. Comparing the areas under the TPR and the TPO
profiles, one notices a systematic difference (Table 2) be-
tween the degree of reduction and the degree of oxida-
tion. The degree of reduction seems to be higher than the
corresponding oxidation. The explanation for this differ-
ence is that, in the TPO experiment, oxygen is consumed
already at room temperature during the preconditioning
before the temperature ramp was started, as has been re-
ported by Thorma¨hlen et al. (14).
3. CO Oxidation with Varying O2 Concentration
The result of the reaction between CO and O2 with vary-
ing O2 concentration is presented in Table 3. The value
S indicates the stoichiometric ratio for the reaction, i.e.,
S = 2[O2]/[CO]. S = 1 thus corresponds to a stoichiometric
mixture. It is clear that the higher the oxygen concentra-
tion (higher S) the higher is the CO conversion. The cata-
lyst deactivates with time, and the deactivation is faster
when the stoichiometric ratio is low. It should be noted,
however, that due to the integral reactor behaviour, less
deactivation would be seen at higher conversion levels. At
21 C, even with a high excess of oxygen, a certain amount of
deactivation occurred, but at 50 C almost no deactivation
occurred with S 2.0. This behaviour is consistent with a
reaction mechanism where CO and O2 are competing for
the same sites on the catalyst surface, and where CO irre-
versibly blocks the sites if it cannot react with oxygen. The
strength of CO blocking under reaction conditions is not
Arnoldy and Moulijn (15) assigned reduction in the
region 440–500 C to surface Co3+ ions or mixed Co3+
–
Al3+ oxidic crystallites and reduction in the region 280–
380 C to Co3O4 crystallites. This corresponds well with the
main peak in the first TPR profile (curve 1 in Fig. 1a), re-
spectively the main peak for the repeated TPR profile af-
ter the first reduction and oxidation (curves 3, 5, and 7 in
Fig. 1a). One explanation for the shift in temperature for
the main peak in the TPR profile is that the first reduc-
tion followed by oxidation transforms Co3+ ions or mixed
Co3+–Al3+ oxidic crystallites into Co3O4 crystallites. The
Co3O4 crystallites are then stable in the sense that, upon
subsequent reduction, the cobalt will return to crystallites
again once the catalyst is oxidised. It is also possible that the
difference between the first TPR (curve 1 in Fig. 1a) and
the subsequent TPR (curves 3, 5, and 7 in Fig. 1a) might
be due to the fact that the fresh catalyst contains impu-
rities that are removed during the first TPR and/or TPO.
These impurities might be nitrates not completely removed
during the calcination, or water or hydrocarbons from the
air.
TABLE 3
Conversion in CO Oxidation Initially and after 60 Min of Reaction
Reaction at 21 C
Reaction at 50 C
Initial CO
conversion
(% )
CO conversion
after 60 min
(% )
Initial CO
conversion
(% )
CO conversion
after 60 min
(% )
S
0.83
1.0
1.17
1.5
2.0
3.0
74
87
90
68
75
79
77
91
97
98
98
98
73
85
91
95
97
97
2. Light-Off
92
84
a
a
The prereduced catalyst lights off with a temperature of
92
88
50% conversion, T50, of 150 C. Full conversion is reached at
Note. The reaction was performed at 21 C and at 50 C over the Co3O4/
Al2O3 catalyst first oxidised in 10% O2 at 550 C. The reaction mixture
consisted of 2% CO and different concentrations of O2, as indicated by
the stoichiometric number S. S = 2[O2]/[CO].
about 265 C. The preoxidised catalyst lights off with T50
=
80 C. The light-off curve shows a shoulder and reaches full
conversion at about 215 C. Preoxidation of the catalyst thus
lowers the T50 by 70 K compared to prereduction. This re-
a Not measured.

58
JONAS JANSSON
known, but from reduced Co/Al2O3, CO desorbs at about 5. Exchange of Oxygen between CO2
100 C, as known from CO-TPD (13, 16). At 21 C it will
thus probably form a strongly adsorbed unreactive species,
but at 50 C the rate of CO desorption is high enough to
prevent CO from irreversibly blocking the cobalt sites. If
the oxygen concentration is high enough, O2 will be able to
compete with CO for the free sites, whereby the rate of CO
blocking is limited.
and the 18O-Preoxidised Catalyst
When the C16O2 is turned on, there is a formation of
C16O18O that decreases after about 10 min. The C16O2
curve levels off at the inlet concentration, and the amount
of C18O2 formation was negligible. The area below the
C16O18O curve for these first 10 min corresponded to
4.5 molofC16O18O and was, asin the previousexperiment,
related to the total number of cobalt atoms in the catalyst,
giving C16O18O/Co = 0.027. This ratio gives a measure of
the fraction of cobalt atoms involved in the exchange reac-
tion. CO2 is proposed to adsorb on the cobalt oxide surface
and exchange oxygen with the surface via the formation of a
carbonate species. Formation ofcarbonate specieson cobalt
oxide has been observed by Finocchio et al. (11) with FTIR
after exposing Co3O4 to CO2. When the same experiment
was performed over pure -Al2O3, formation of C16O18O
was also observed during the first 10 min, corresponding to
3.1 mol of C16O18O. It is known that carbonates may form
on alumina upon CO2 introduction (17), and exchange be-
tween CO2 in the gas phase and oxygen on alumina has
been reported at 27 C (18). This may explain the C16O18O
on pure Al2O3.
4. CO Oxidation over an Isotope-Labeled
Cobalt Oxide Catalyst
The result from this experiment with an 18O-preoxidised
catalyst is presented in Fig. 2. When the reaction mixture
16
(CO + O2, S = 1.17) is switched on, a certain amount of
isotope-labeled C16O18O is formed in addition to the for-
mation of C16O2. The amount of C18O2 formed is negligible.
After about 10 min the C16O18O decreases to zero. As no
18O is introduced in the gas phase, this means that one of
the oxygens in the CO2 formed must come from the preoxi-
dised catalyst surface, which will be partially reduced when
reacting with CO. When the surface oxygen on the catalyst
is consumed, no more C16O18O can be formed, which ex-
plains why the curve decreases to zero after about 10 min.
The area belowthe curve for these first 10min corresponded
to 6.4 mol of C16O18O and was related to the total number
of cobalt atoms in the catalyst, giving C16O18O/Co = 0.038.
As in the previous experiment, where the deactivation of
the catalyst was studied, the amount of CO2 formed de-
creased with time. The presence of the about 8 times larger
amount of C16O2 compared to C16O18O in Fig. 2 may indi-
cate that only a part of the surface atoms are involved in
the reaction, since the gas-phase isotope fraction should be
proportional to the active site abundance.
6. TPD and TPO of the Deactivated Catalyst
The reaction between C16O and 18O2 over the 18O-
oxidised catalyst resulted in the formation of the CO2 iso-
topes in the proportions 29% C16O2, 50% C16O18O, and
21% C18O2. These ratios stayed constant during the 90 min
that the reaction took place. Asin the previousexperiments,
a deactivation of the catalyst as a function of time could be
seen. From the results in experiment 4, where CO reacted
with one oxygen in the surface cobalt oxide, one would
have expected only C16O18O formation during the reac-
tion. Instead, large amounts of both C18O2 and C16O2 were
formed besides the C16O18O. This might be due to exchange
of oxygen in gas-phase CO2 and surface oxygen (either on
cobalt oxide or on the alumina), asseen in the CO2 exchange
experiment.
The results from the TPD and the TPO are presented in
Fig. 3. Both TPD and TPO profiles have one peak at about
100 C. This peak contains 31% C16O2, 49% C16O18O, and
20% C18O2 in the TPD, and 27% C16O2, 52% C16O18O,
and 21% C18O2 in the TPO. The different CO2 isotope
peaks are shifted somewhat in relation to each other in the
order Tpeak(C16O2) < Tpeak(C16O18O) < Tpeak(C18O2). The
TPO profile also has a peak at about 450 C, which con-
tains only C16O2. This peak is not present in the TPD
profile. No CO formation was observed in either TPD or
TPO.
FIG. 2. Gas phase mole fractions of C16O2, C16O18O, and C18O2 in the
The first peak in TPD and TPO profiles (at about 100 C)
of the catalyst deactivated by the reaction between 18O2
and C16O is similar in the two experiments (Fig. 3), i.e.,
outlet from the reactor as a function of time for the reaction between C16
O
and 16O2 at 21 C. The Co3O4/Al2O3 catalyst was first preoxidised with 18O.
The shaded area corresponds to C16O18O/Co = 0.038.

LOW-TEMPERATURE CO OXIDATION OVER Co3O4/Al2O3
59
FIG. 3. (a) Temperature-programmed desorption (TPD) and (b) temperature-programmed oxidation (TPO) of a mildly deactivated Co3O4/Al2O3
catalyst. The catalyst was first preoxidised with 18O and then deactivated by reacting 1.67% C16O and 0.98% 18O2 for 90 min over it before TPD or
TPO was performed.
the presence of this peak is independent of the presence of
(b) a CO can adsorb to the site and thus deactivate it.
gas-phase oxygen. This indicates that it is a pure desorption It is possible that the adsorbed CO thereafter undergoes
peak. It is suggested that this peak is due to decomposition dissociation or reacts with another CO to form CO2 and C
of surface carbonates formed during the reaction between (CO disproportionation).
C16O and 18O2 prior to the TPD/TPO. Carbonates probably
(4) Gas-phase CO2 can react with the oxidised surface
also form on the alumina support. The different CO2 iso- to form a carbonate species. The carbonate may exchange
tope peaks are shifted somewhat in relation to each other. oxygen with the surface and desorb as CO2 again.
This is probably due to a kinetic isotope effect in the de-
Mechanisms similar to this have previously been pro-
composition of the carbonate to CO2 (19, 20).
posed by, e.g., Mergler et al. (8) and Haruta et al. (3). The
The second peak at about 450 C contained only C16O2
main difference is that the catalysts investigated also con-
and was present in the TPO profile but not in the TPD. This
tained Pt (8) or Au (3). The authors propose that the CO is
is consistent with oxidation of surface carbon, Csurf, with
adsorbed to the Pt or Au site, while oxygen issupplied bythe
gas-phase O2. Csurf might be formed in the CO dispropor-
CoOx. In this investigation it was shown that the reaction
tionation reaction (2CO → C + CO2), which was reported
may take place without the presence of a noble metal, lead-
to occur over partially reduced CeO2 at room temperature
ing to the above-proposed mechanism. This redox mecha-
(21) and over Pt/CeO2/Al2O3 at 300 C (22). It is also possi-
nism is in accordance with what has previously been found
ble that another form of C, for example carbonate, during
for CO oxidation over metal oxide catalysts (1, 23).
the TPO is transformed into graphitic C during heating of
the catalyst sample.
CONCLUSIONS
Proposed Mechanism
The preoxidised catalyst shows higher activity for CO
oxidation than the prereduced catalyst. Deactivation takes
place at 21 C for the preoxidised catalyst, but the deactiva-
tion can be slowed by increasing the O2/CO ratio. By raising
the temperature to 50 C, no deactivation was seen over the
preoxidised catalyst for S 2.0. This behaviour is well ex-
plained by the above-proposed reaction mechanism. When
the CO oxidation reaction proceeds, at least two forms of
carbonaceous species are deposited on the catalyst. These
are probably carbonates (both on the cobalt oxide and on
the alumina support) and graphite-like C.
From the experimental results, the following mechanism
for the low-temperature CO oxidation over Co3O4/ -Al2O3
is proposed:
(1) CO(g) isadsorbed on an oxidised cobalt site (probably
Co3+).
(2) The adsorbed CO reacts with an oxygen linked to the
active Co3+. CO2 is formed and desorbs quickly. The result
is a partially reduced site. The partially reduced site may
consist of two Co2+ ions or may be regarded as an oxygen
vacancy.
(3) The partially reduced site can undergo two different
reactions.
ACKNOWLEDGMENT
(a) It can either be reoxidised by gas-phase oxygen to
I am grateful to Prof. Bengt Andersson for his valuable comments
on the manuscript. The Competence Center for Catalysis is financially
an active Co3+ site, or

60
JONAS JANSSON
supported by the Swedish National Energy Administration and the mem- 11. Finocchio, E., Busca, G., Lorenzelli, V., and Escribano, V. S., J. Chem.
ber companies: AB Volvo, Johnson Matthey CSD, Saab Automobile AB,
Perstorp AB, MTC AB, and Eka Chemicals AB.
Soc., Faraday Trans. 92(9), 1587 (1996).
12. Finocchio, E., Willey, R. J., Busca, G., and Lorenzelli, V., J. Chem.
Soc., Faraday Trans. 93(1), 175 (1997).
13. To¨rncrona, A., Skoglundh, M., Thorma¨hlen, P., Fridell, E., and Jobson,
E., Appl. Catal. B 14, 131 (1997).
REFERENCES
14. Thorma¨hlen, P., Skoglundh, M., Fridell, E., and Andersson, B., J. Catal.
188, 300 (1999).
1. Shelef, M., Otto, K., and Gandhi, H., J. Catal. 12, 361 (1968).
2. Yu Yao, Y.-F., J. Catal. 33, 108 (1974).
15. Arnoldy, P., and Moulijn, J. A., J. Catal. 93, 38 (1985).
16. Bridge, M. E., Comrie, C. M., and Lambert, R. M., Surf. Sci. 67, 393
(1977).
17. Cabrejas Manchado, M., Guil, J. M., Pe´rez Masia´, A., Ruiz Paniego,
A., and Trejo Menayo, J. M., Langmuir 10(3), 685 (1994).
18. Morikawa, Y., and Amenomiya, Y., J. Catal. 48, 120 (1977).
19. Gates, S. M., Russell, J. N., and Yates, J. T., Surf. Sci. 146, 199
(1984).
3. Haruta, M., Tsubota, S., Kobayashi, T., Kageyama, H., Genet, M. J.,
and Delmon, B., J. Catal. 144, 175 (1993).
4. Haruta, M., Yamada, N., Kobayashi, T., and Iijima, S., J. Catal. 115,
301 (1989).
5. Cunningham, D. A. H., Kobayashi, T., Kamijo, N., and Haruta, M.,
Catal. Lett. 25, 257 (1994).
6. Meng, M., Lin, P., and Fu, Y., Catal. Lett. 48, 213 (1997).
7. Mergler, Y. J., van Aalst, A., van Delft, J., and Nieuwenhuys, B. E.,
Stud. Surf. Sci. Catal. 96, 163 (1995).
20. Houtman, C., and Barteau, M. A., J. Phys. Chem. 95, 3755 (1991).
21. Li, C., Sakata, Y., Arai, T., Domen, K., Maruya, K., and Onishi, T.,
J. Chem. Soc., Chem. Commun. 410 (1991).
8. Mergler, Y. J., van Aalst, A., van Delft, J., and Nieuwenhuys, B. E.,
Appl. Catal. B 10, 245 (1996).
22. Holmgren, A., and Andersson, B., J. Catal. 178, 14 (1998).
23. Boreskov, G. K., in “Catalysis: Science and Technology” (J. R.
Anderson and M. Boudart, Eds.), Vol. 3, p. 39. Springer-Verlag, Berlin,
1982.
9. Mergler, Y. J., Hoebink, J., and Nieuwenhuys, B. E., J. Catal. 167, 305
(1997).
10. Garbowski, E., Guenin, M., Marion, M.-C., and Primet, M., Appl.
Catal. 64, 209 (1990).