Production of hydrogen by steam reforming of methanol on CeO2 promoted Cu/Al2O3 catalysts

Article abstract of DOI:10.1016/S1381-1169(02)00464-8

Catalytic production of hydrogen by steam reforming of methanol reaction has been developed on a series of co-precipitated Cu/Al₂O₃ catalysts promoted with CeO₂ at atmospheric pressure in a microreactor. Effects of CeO₂ content, reaction temperature, methanol space velocity and H₂O/CH₃OH molar ratio on the catalytic activity have been investigated. CeO₂ promoted Cu/Al₂O₃ catalysts exhibited higher activity and stability as compared to the unpromoted ones. The catalyst containing 20 wt.% of CeO₂ was the most active one with a methanol conversion of 95.5 mol%, H₂ selectivity of 99.9 mol% and the outlet CO concentration of 0.14 mol% at 250°C. After 200 h of reaction, methanol conversion was still over 90.0% with the catalyst containing 20 wt.% of CEO₂, while Cu/Al₂O₃ catalyst deactivated rapidly after 100 h of reaction. Results of X-ray diffraction (XRD) and the surface element distribution of catalysts showed that CeO₂ not only greatly improved the surface copper dispersion and prevented copper crystallites from conglomeration or sintering, but also made copper crystallites relatively smaller. The improvement in activity and stability of the promoted catalysts was attributed to higher copper dispersion and smaller copper crystallites, and the synergetic effect of ceria.

Full text of DOI:10.1016/S1381-1169(02)00464-8

Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
Production of hydrogen by steam reforming of methanol
on CeO promoted Cu/Al O catalysts
2
2
3
∗
Xinrong Zhang , Pengfei Shi
Department of Applied Chemistry, Harbin Institute of Technology, P.O. Box 411, No. 92,
Xidazhi Street, Harbin Heilongjiang 150001, China
Received 15 March 2002
Abstract
Catalytic production of hydrogen by steam reforming of methanol reaction has been developed on a series of co-precipitated
Cu/Al2O3 catalysts promoted with CeO2 at atmospheric pressure in a microreactor. Effects of CeO2 content, reaction temper-
ature, methanol space velocity and H2O/CH3OH molar ratio on the catalytic activity have been investigated. CeO2 promoted
Cu/Al2O3 catalysts exhibited higher activity and stability as compared to the unpromoted ones. The catalyst containing
2
0 wt.% of CeO2 was the most active one with a methanol conversion of 95.5 mol%, H2 selectivity of 99.9 mol% and the
◦
outlet CO concentration of 0.14 mol% at 250 C. After 200 h of reaction, methanol conversion was still over 90.0% with the
catalyst containing 20 wt.% of CeO2, while Cu/Al2O3 catalyst deactivated rapidly after 100 h of reaction. Results of X-ray
diffraction (XRD) and the surface element distribution of catalysts showed that CeO2 not only greatly improved the surface
copper dispersion and prevented copper crystallites from conglomeration or sintering, but also made copper crystallites rela-
tively smaller. The improvement in activity and stability of the promoted catalysts was attributed to higher copper dispersion
and smaller copper crystallites, and the synergetic effect of ceria.
©
2002 Elsevier Science B.V. All rights reserved.
Keywords: Hydrogen production; Steam reforming of methanol; Cu/Al2O3 catalyst; CeO2 promoted
1
. Introduction
the reformed gases deteriorate a Pt electrode and the
cell performance is worsened [4]. An ideal method
to produce hydrogen with lower amount of CO from
steam reforming of methanol greatly requires a high
performance catalyst, which must be highly active
and selective for hydrogen production and also sta-
ble for a long period in a continuous operation. Now
the most widely used catalysts for this reaction are
copper containing catalysts since copper has been
found to be high activity and selectivity for hydrogen
production [5–9]. Copper containing catalysts are
used in many catalytic reactions, such as CO shift
at low temperature [10] and CO hydrogenation for
methanol [11], steam reforming of methanol reaction
gives copper containing catalysts new application.
Fuel cell powered electric vehicles and power
plants using hydrogen as fuel are currently being de-
velopment in an effort to protect the environment and
sustainable development [1,2]. Hydrogen produced
by steam reforming of methanol is an increasing
worldwide interest, the equilibrium conversion of
steam reforming of methanol reaction reaches around
◦
1
00% at 150 C at atmospheric pressure [3]. Unfortu-
nately, a considerable amount of CO (>100 ppm) as a
by-product is produced during the reaction. As for the
application of PEFC, even traces of CO (>20 ppm) in
∗
Corresponding author.
1
381-1169/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(02)00464-8

1
00
X. Zhang, P. Shi / Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
Furthermore, the high oxygen mobility, strong inter-
action with certain metals and the modifying ability
make CeO2 to be a very interesting promoter for cat-
alysts [12–14], CeO2 promoted catalysts have been
rapidly developed in recent years [15].
The objective of the present investigation is to de-
velop an efficient catalytic system based on CeO2
promoted Cu/Al2O3 catalysts by co-precipitation
method for production of hydrogen by steam re-
forming of methanol reaction. The effects of catalyst
composition, reaction temperature, methanol space
velocity, H2O/CH3OH molar ratio on the performance
of Cu/CeO2/Al2O3 catalysts, and the analysis of the
surface element distribution and crystal phases and
crystallite sizes of catalysts will be reported.
of radiation (1.5418 Å), β is the line broadening of
the peak due to small crystallites (rad 2θ), and θ is
the corresponding angle of the diffraction peak. The
full width at half maximum (FWHM) of (1 1 1) reflec-
tion of copper was measured for calculating crystallite
sizes [16]. The instrumental broadening was corrected
by β = (B − b ) , where B is the total broaden-
ing, and b is the instrumental broadening. The surface
element distribution of catalysts was analyzed with a
PHI-5000C ESCA system (Perkin-Elmer) with Al K␣
radiation.
2
2 0.5
2.3. Activity measurements
Catalytic activity test experiments were made in a
continuous flow fixed-bed microreactor (6 mm i.d.)
placed in an electric furnace. The furnace temperature
was controlled by a PID temperature controller with
a K-type thermocouple inserted in the furnace. A sep-
arate thermocouple was used to monitor the tempera-
ture of the catalyst bed. This arrangement was capable
2
. Experimental
2
.1. Catalyst preparation
◦
CeO2 promoted Cu/Al2O3 catalysts were prepared
of ensuring a temperature accuracy of ± 1 C for the
by co-precipitation method. A mixed aqueous solu-
tion of copper nitrate 3-hydrate (Cu(NO3)2·3H2O),
aluminum nitrate 9-hydrate (Al(NO3)3·9H2O) and
cerium nitrate 6-hydrate (Ce(NO3)3·6H2O), and a so-
lution of sodium carbonate were added slowly and si-
catalyst bed. All tests were conducted in a reaction
◦
temperature range of 180–280 C at atmospheric pres-
sure with 500 mg of the catalyst loaded in the reactor.
The catalysts were reduced in situ with a premixed
H /Ar (5/95 (v/v)) gas flow, which had a speed of
2
◦
−1
multaneously into 100 ml of deionized water at 60 C
80 ml min and was heated from room temperature
◦
◦
−1
◦
with vigorous stirring. The pH was kept constant at
to 300 C at a rate of 1 C min , and kept at 300 C
for 3 h. A typical test experiment for methanol steam
reforming reaction at a methanol space velocity of
3.28 h with a 1:1 molar mixture of methanol and wa-
ter was made as follows. The catalyst was loaded into
the reactor and activated using designated activation
procedure. The premixed water and methanol was then
◦
7
.0–7.2. The precipitates were aged at 60 C for 30 min
with vigorous stirring, then filtered and thoroughly
washed with warm deionized water. The precipitates
−
1
◦
were dried overnight in air at 110 C and calcined in
◦
a muffle oven at 500 C for 3 h. Each calcined cata-
lyst was pelletized in a hydraulic press, crushed, and
sieved into a particle size of 0.45–0.55 mm, and used
as catalysts for steam reforming of methanol reaction.
◦
pumped to the vaporizer maintained at about 240 C.
The vaporized feed entered the reactor with a stream
−
1
of Ar gas, which had a speed of 50ml min , and then
began the steam reforming reaction at the designated
reaction temperature. The reaction products were an-
alyzed on-line by HP5890 gas chromatography with
thermal conductivity detector. The GC equipped with a
3 m long Porapak-Q column and a 3 m long Porapak-R
column was able to detect both the liquid products,
such as water, methanol, formaldehyde, methyl for-
mate and dimethyl ether, and the gaseous products,
such as H2, CO2, and CO, respectively. The catalytic
activity was evaluated from the data recorded between
2
.2. Catalyst characterization
Powder X-ray diffraction (XRD) patterns of cata-
lyst samples were obtained using a Riggku D/Max-3B
with Cu K␣ radiation, a scanning angle (2θ) range of
◦ ◦ −1
1
0–80 , a scanning speed of 6 min , and a voltage
and current of 35 kV and 30 mA, respectively, for the
analysis of crystal phases and crystallite sizes. Crystal-
lite sizes of Cu were calculated using Debye–Scherrer
equation: t = 0.9λ/β cos θ, where λ is the wavelength

X. Zhang, P. Shi / Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
101
2
and 4 h of the continuous operation. In order to
crystal state. The diffraction peak of copper is low-
ered remarkably for the catalyst containing 20 wt.%
of CeO2, the higher the CeO2 concentration, the lower
the diffraction peak shows. In contrast to the catalyst
containing 20 wt.% of CeO2, the diffraction peak of
Cu/Al2O3 catalyst shows relatively sharp. It indicated
that CeO2 greatly promoted surface copper dispersion.
Fig. 2 shows that only Cu diffraction peak is present
after reaction. It indicated that Cu was the main ac-
tivation center of catalysts. The crystallite sizes of
copper (DCu) are estimated from the half width of
check the stability of catalysts, the reaction was also
performed for a period of 100–200 h of the continu-
ous operation at 250 C. Blank run conducted with an
empty reactor in a temperature range of 180–280 C
did not show any detectable methanol conversion.
◦
◦
3
. Results and discussion
3
.1. Characterization of calcined catalysts
(
1 1 1) reflection of Cu by Debye–Scherrer equation.
The calculation results (DCu lowered from 23.9 to
9.8 nm) showed that copper crystallite sizes lowered
XRD patterns of Cu/Al2O3 and Cu/CeO2/Al2O3
1
calcined catalysts are shown in Fig. 1, and XRD
patterns of both catalysts after reaction are shown in
Fig. 2. Fig. 1 shows that no alumina and ceria phases
are detected, implying that aluminum and ceria phases
are probably present in an amorphous-like or micro-
crystallite state in catalysts, copper is present in a
with the addition of CeO2. XRD results showed that
CeO2 had a significant influence on surface copper
dispersion and crystallite sizes.
The methanol decomposition and reforming reac-
tion can be considered as the reverse of the methanol
synthesis reaction using CO/H2 or CO2/H2 as a feed
on copper containing catalysts [17]. For copper con-
taining catalysts, Cu is the active site of methanol
synthesis from CO2 hydrogenation [18,19], so the
dispersity and the surface element distribution of Cu
acting specie are the factors determining the per-
formance of catalysts. Table 1 shows the surface
element distribution of catalysts. Copper content is
1
9.83 mol% and Cu/Al atom ratio is 0.25 on the sur-
face of fresh Cu/Al2O3 catalyst as compared with
4.83 mol% and 0.37 on the surface of fresh catalyst
2
containing 20 wt.% of CeO2. After100 h of reaction,
copper content on the surface of Cu/Al2O3 catalyst is
Fig. 1. XRD patterns of calcined catalysts (1) w(CeO2) = 20%;
17.26 mol% and Cu/Al atom ratio is 0.20, however,
(
2) w(CeO2) = 0%.
on the surface of the catalyst containing 20 wt.% of
CeO2, even after 200 h of reaction, copper content is
Table 1
The surface element distribution of catalysts
Catalysts
Cu
Al
Cu/Al
(mol%) (mol%) (atom ratio)
Cu/Al2O3
Fresh
After
19.83
17.26
80.17
82.74
0.25
0.20
100 h
reaction
Cu/CeO2/Al2O3 Fresh
After
24.83
19.93
66.73
71.61
0.37
0.28
2
00 h
Fig. 2. XRD patterns of catalysts after reaction (1) w(CeO2)
20%; (2) w(CeO2) = 0%.
reaction
=

1
02
X. Zhang, P. Shi / Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
1
9.93 mol% and Cu/Al atom ratio is 0.28. It indicated
Thus, the main reactions of the proposed route may
that surface copper dispersion was much improved
by CeO2, which would improve the catalytic perfor-
mance. It can also be seen that Al content increases
from 66.73 to 71.61 mol% on the surface of the cat-
alyst containing 20 wt.% of CeO2 as compared with
that from 80.17 to 82.74 mol% on the surface of
Cu/Al2O3 catalyst. A higher Al content is favorable
for the stabilization of copper structure on the catalyst
surface. A XPS study [9,20] of Cu/Al-based catalysts
showed that catalysts with higher copper dispersion
might have exhibit Cu/Al interactions on their sur-
faces of calcined samples. The presence of Al would
contribute the dispersion of oxidized copper species
to a significant extent on the surface of calcined sam-
ples due to the formation of certain surface phases
between Cu and Al, and the stabilization of isolated
Cu species. During the subsequent reduction, these
highly dispersed cations might behave as basic points
for the formation of Cu particles with a higher surface
area, consequently, which would have a significant
influence on the catalytic performance.
be represented by Eqs. (1)–(3).
CH3OH + H2O
ꢀ
3H2 + CO2 (methanol steam reforming)
◦
−1
ꢀH
= +49.4 kJ mol
(1)
(2)
(3)
298
CH3OH ꢀ CO + 2H2 (methanol decomposition)
◦
−1
ꢀH
= +92.0 kJ mol
2
98
CO + H O ꢀ CO + H (water gas shift)
2
2
2
◦
−1
ꢀH
= −41.1 kJ mol
2
98
A systematic experiment was then undertaken after
reducing the catalyst.
3.2.1. Effect of CeO concentration
2
Table 2 shows a typical dependence of methanol
conversion on CeO concentration in catalysts. An in-
2
crease in methanol conversion with increasing CeO2
concentration can be seen. Beyond 20.0 wt.% of CeO2,
methanol conversion begins to decrease. The promoted
catalysts display high selectivity for hydrogen and low
selectivity for CO. It can also be seen that the outlet CO
concentration is less than 0.2 mol%. Results showed
that the promoted Cu/Al2O3 catalysts exhibited better
catalytic performance as compared to the unpromoted
ones, it indicated that CeO2 had an important influence
on improving catalytic activity and decreasing the out-
let CO concentration. Rosynek [22] considered that
CeO2 could oxidize hydrocarbon by itself. Although
the catalytic activity of CeO2 is lower than that of
the oxides of other transition metals, the catalytic ac-
tivity of Cu/CeO2/Al2O3 catalyst is greatly improved
3
.2. Catalytic activity
Catalytic activity was evaluated in terms of
methanol conversion (mol%). Methanol steam re-
forming reaction on CeO2 promoted Cu/Al2O3 cata-
lysts in a wide temperature range showed that it was
◦
still active at a temperature as low as 180 C. Anal-
ysis of the effluent gas indicated that H2 and CO2
were major components with a minor amount of CO.
Other products such as formaldehyde, formic acid,
methyl formate and dimethyl ether formed during
reactions of methanol on Cu-based catalysts could
not be detected under the reaction conditions [9,21].
Table 2
Effect of CeO2 concentration on catalytic activity
Number
w(CeO2) (%)
X(CH3OH) (mol%)
Y(H2) (mol (h·g)−1)
S(H2) (%)
y(CO) (mol%)
1
2
3
4
5
6
0
5
10
15
20
25
81.4
87.5
90.0
93.0
95.5
91.8
0.2509
0.2697
0.2774
0.2866
0.2944
0.2829
99.7
99.9
99.9
99.9
99.9
99.9
0.37
0.19
0.17
0.15
0.14
0.16
◦
Reaction conditions: P = 0.1 MPa; t = 250 C; X the methanol conversion; Y the hydrogen yield; S the selectivity; and y is the outlet CO
concentration.

X. Zhang, P. Shi / Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
103
by the synergetic effect of ceria. Furthermore, CeO2
can improve the water gas shift reaction and promote
CO conversion by this reaction, thus hydrogen with
lower amount of CO can be produced on the promoted
catalysts. On the other hand, results of XRD and the
surface element distribution of catalysts showed that
CeO2 promoted catalysts had high copper dispersion
and small crystallite sizes. The improvement in the
catalytic performance of the promoted catalysts was
partly attributed to such a copper dispersion and crys-
tallites. Table 2 shows that there is an optimum CeO2
concentration in the promoted catalysts. Such a maxi-
mum occurred for methanol conversion and CeO2 con-
centration is explained as below. An increase in CeO2
concentration means a decrease in the overall copper
concentration in catalysts, consequently, a decrease in
Fig. 4. Effect of WHSV on catalytic activity.
concentration increases with increasing reaction tem-
perature. It indicates that CO is produced by the
reverse reaction (Eq. (3)) at a higher temperature. Be-
cause the water gas shift reaction is exothermic, the
increase in reaction temperature is not favorable for it.
◦
amounts of Cu in activated catalysts. The overall ef-
fect is the existence of a maximum ratio, inferring
that, beyond the concentration limit for CeO2, its syn-
ergetic effect begins decreasing while its role as an
active site for methanol steam reforming reaction be-
comes significant. In comparison with copper species,
CeO2 is less active if used solely as a catalyst [22].
3.2.3. Effect of methanol space velocity
Fig. 4 shows the effect of methanol space velocity
(WHSV) on the catalytic performance of CeO2 pro-
3
.2.2. Effect of reaction temperature
The effect of reaction temperature on catalytic
moted catalysts. Methanol conversion and the outlet
CO concentration decrease with increasing methanol
space velocity, and hydrogen yield has a maximum
in the experiment conditions. On the other hand,
methanol space velocity does not affect the selectivity
of H2, which remains around 99.9% throughout the
experiment. In the present study, a methanol space
performance of CeO2 promoted catalysts is shown
in Fig. 3. Methanol conversion and hydrogen yield
increase with increasing reaction temperature, while
methanol is converted almost completely into H2,
CO2, and CO up to 280 C. In the temperature range
of 180–280 C, hydrogen selectivity remains almost
◦
◦
−1
velocity of 3.28 h has been chosen for the evalua-
unchanged, and the outlet CO concentration is less
than 0.4 mol%. On the other hand, the outlet CO
tion of the performance of catalysts during the steam
reforming reaction.
3.2.4. Effect of H2O/CH3OH molar ratio
It is known from Eqs. (1)–(3) that excess H2O
promotes methanol conversion and reduces CO con-
centration by shifting the equilibrium (Eq. (3)) toward
the right. The effect of H2O/CH3OH molar ratio on
the catalytic performance during the steam reforming
◦
of methanol reaction at 250 C is presented in Fig. 5.
It can be seen that methanol conversion increases
remarkably with increasing the H2O/CH3OH molar
ratio below 1.0, while the increase slows down be-
yond the ratio of 1.0, and hydrogen yield increases
slowly with increasing H2O/CH3OH molar ratio. An
outlet CO concentration of 0.78 mol% is determined
Fig. 3. Effect of reaction temperature on catalytic activity.

1
04
X. Zhang, P. Shi / Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
deactivates rapidly after 100 h and methanol conver-
sion decreases from 81.4 to 75.8%, approximately a
decrease of 6.87%. There is an initial deactivation of
the catalyst containing 20 wt.% of CeO2 before 20 h
of continuous operation. After 20 h of that, no signifi-
cant deactivation of the catalyst is observed even after
200 h, and methanol conversion remains almost un-
changed throughout this period. As far as the product
selectivity is concerned, there is no change in the se-
lectivity of H2, which is around 99.9 mol%. The result
showed that CeO2 improved the stability of catalysts.
Fig. 5. Effect of H2O/CH3OH molar ratio on catalytic activity.
4. Conclusions
when the H2O/CH3OH molar ratio is 0.6. It should
be noted from Fig. 5 that the outlet CO concentra-
tion decreases to about 0.1 mol% with increasing the
H2O/CH3OH molar ratio up to 1.5. Hence, the results
showed that higher H2O/CH3OH molar ratio was
favorable for reducing the outlet CO concentration
due to the enhancement of WGS reaction. Therefore,
an H2O/CH3OH molar ratio between 1.0 and 1.5
offered a better catalytic performance in the present
experimental conditions.
The addition of CeO2 to Cu/Al2O3 catalyst in-
creases its activity and stability for hydrogen pro-
duction from methanol steam reforming. Methanol
conversion increases with increasing CeO2 concentra-
tion below 20 wt.%. Increase in reaction temperature
and H2O/CH3OH molar ratio also improves methanol
conversion. On the other hand, methanol conversion
decreases with increasing methanol space velocity.
Hydrogen selectivity remains unchanged in the exper-
imental conditions, the outlet CO concentration is less
than 0.8 mol%. Results of XRD and the surface ele-
ment distribution of catalysts showed that CeO2 could
enhance the surface dispersion of copper on catalysts,
and prevent copper crystallites from sintering or con-
glomerating, and make copper crystallites relatively
smaller. It is suggested that high activity, selectivity
and stability of CeO2 promoted catalysts have been
resulted from a higher copper dispersion and smaller
copper crystallites, and the synergetic effect of ceria.
3
.2.5. Effect of reaction time
In order to investigate the stability of catalysts dur-
ing the steam reforming reaction, the continuous op-
◦
erations were performed at 250 C and 0.1 MPa with
catalysts containing 0 wt.% of CeO2 and 20 wt.% of
CeO2 for a period of 100–200 h, and the results are
shown in Fig. 6. It can be seen that Cu/Al2O3 catalyst
Acknowledgements
This work was supported in part by Guangyu Bat-
tery Co., China.
References
[
[
1] W. Wies, B. Emonts, J. Power Sources 84 (1999) 187.
2] J.C. Amphlett, R.F. Mann, J. Power Sources 71 (1998) 179.
Fig. 6. Effect of reaction time on catalytic activity (1) w(CeO2)
[3] J.C. Amphlett, M.J. Evans, R.A. Jones, et al., Can. J. Chem.
Eng. 59 (1981) 720.
=
20%; (2) w(CeO2) = 0%.

X. Zhang, P. Shi / Journal of Molecular Catalysis A: Chemical 194 (2003) 99–105
105
[
[
[
[
[
[
4] H.F. Oetjen, V.M. Schmidt, U. Stimming, et al., J.
Electrochem. Soc. 143 (1996) 3838.
5] J.C. Amphlett, R.F. Mann, R.D. Weir, Can. J. Chem. Eng.
[13] P. Fornasiero, G. Balducci, R.D. Monte, et al., J. Catal. 164
(1996) 173.
[14] L. Fan, K. Fujimoto, J. Catal. 172 (1997) 238.
[15] Y.Y. Liu, T. Hayakawa, K. Suzuki, et al., Catal. Commun. 2
(2001) 195.
[16] C.J.G. Grift, A.F.H. Wielers, B.P.J. Joghi, J. Catal. 131 (1991)
178.
[17] S. Velu, K. Suzuki, M.P. Kapoor, et al., Appl. Catal. A 213
(2001) 47.
6
6 (1988) 950.
6] C.J. Jiang, D.L. Trimm, M.S. Wainwright, Appl. Catal. A 93
1993) 245.
7] R.O. Idem, N.N. Bakkhshi, Ind. Eng. Chem. Res. 33 (1994)
056.
8] S. Velu, K. Suzuki, T. Osaki, et al., J. Catal. 194 (2000)
73.
9] S. Velu, K. Suzuki, T. Osaki, Catal. Lett. 62 (1999) 159.
(
2
3
[18] S.G. Neophytides, A.J. Marchi, G.F. Froment, Appl. Catal.
A 86 (1992) 45.
[
[
[
10] M.J.L. Gines, N. Amadeo, M. Laborde, et al., Appl. Catal.
A 131 (1995) 283.
11] J.R. Jennings, M.V. Twigg, Crit. Rev. Appl. Chem. 13 (1985)
[19] M. Saito, T. Fujitani, M. Takeuchi, et al., Appl. Catal. A 138
(1996) 311.
[20] R.T. Figueiredo, A. Martinez-Arias, M.L. Granados, J. Catal.
178 (1998) 146.
[21] I.E. Wachs, R.J. Madix, J. Catal. 53 (1978) 208.
[22] M.P. Rosynek, Catal. Rev. Sci. Eng. 16 (1977) 112.
1
02.
12] M. Pijolat, M. Prin, M. Soustelle, J. Chem. Soc., Faraday
Trans. 91 (1995) 3941.