88
M. Mrad et al. / Applied Catalysis A: General 471 (2014) 84–90
03-065-3288) with Cu0 (JCPDS: 00-004-0836) phase are present
besides the peaks assigned to ZnO.
owning the worst dispersion (59%).
The presence of reduced species of copper after test is revealed
by the decreasing intensity of EPR signal compared to that before
test (Table 2).
Same to xCu10Ce, the presence of less dispersed copper induces
the best selectivity of H2 with absence of CO and CH4 which was
attributed to 5Cu10Ce10Al (59%).
The addition of zinc on the catalysts does not have the same
effect on the activities and the dispersions. The impregnation of
Zn on 5Cu10Ce10Al causes a decrease in the methanol conver-
sion (96%), the dispersion (32%) and the carbon assessment (0.73),
besides 0.1% CO was formed. On the other hand, the impregna-
tion of Zn on 5Cu10Ce improves the dispersion (28%), the methanol
conversion (99.9%) and the carbon assessment (0.99).
There are several theories describing the nature of the interac-
tion between Cu–ZnO, some of them consider that the major role of
ZnO in a Cu based catalysts is a promoter and this role is explained
by different mechanisms [28]. Furthermore, our results correlate
with the theory explaining the marked effect of the interactions
species [29]. Somehow, those interactions seem to depend directly
on the nature of the support.
EPR technique can serve as a method for carbon species inves-
tigation [30,31]. Several studies have shown that carbon species
and its derivatives show that an EPR signal constituted by a single
line at different widths depending on the carbon species [30,31].
The EPR parameters of this signal are g = 2.002–2.005 and the line-
width ꢀH = 7–100 G. The paramagnetic centers originate by mobile
unpaired electrons within the carbon structure or at the surface
forming free radicals. It was indicated that the EPR line-width
and the unpaired spin density may be related to the surface area,
the molecular structure, the particle size and defect characteris-
tics. The increase in line-width is attributed to the formation of
carbon–oxygen complexes.
Thus, during test the CuO species tend to be reduced into Cu0
and Cu+. In their study, Pfeifer et al. [32] have shown that the
high catalytic activity of the Cu/CeO2 catalysts in the SRM reac-
tion is directly related to the stabilization of the Cu2O species in
the presence of CeO2. Otherwise, Reitz et al. [33], did not allocate
any activity to the Cu+ species. It is well known that the reducibil-
ity of CuO species increases when CeO2 is used as a support or
even when it is added to another mixed oxide support [34–38].
This fact could probably enhance the catalytic activity [34,37].
In our case, the higher catalytic activity of 5%Zn5Cu10Ce cata-
lyst could be ascribed to the formation of an optimum Cu0–Cu2O
species stabilized by chemical interaction on ceria support, and
in the presence of zinc, which is in correlation with Idem and
Bakhshi [11].
But this was not the case of 5%Zn5Cu10Ce10Al where the addi-
tion of 5% of ZnO has limited the formation of reduced copper
species and has decreased the methanol conversion and the carbon
assessment. On contrary of all other catalysts, CuO species are not
CuO peaks in these catalysts before test suggests that these particles
remaining on the surface of CeO2 may form a solid solution which
escapes the XRD detection. In addition, the 5Cu10Ce10Al pattern
shows the presence of CuAl2O4 (JCPDS: 00-033-0448) species. Patel
and Pant [39] substantiate that the presence of alumina improves
the stabilization of isolated Cu2+ ions in their matrix, and moder-
ately by the formation of spinel like CuAl2O4. After test, the spinel
like CuAl2O4 species becomes more identified with the presence of
Cu0 (JCPDS: 00-004-0836).
After adding 5% of ZnO over 5Cu10Ce10Al, the peaks assigned
to the CuAl2O4 species are always visible without any detection of
CuO and ZnO species. Unlike the 5Cu10Ce10Al, no Cu0 is detected
over the 5%Zn5Cu10Ce10Al after test showing that the reducibility
of the copper in the presence of alumina is significantly affected by
the addition of the zinc.
The absence of Cu0 species in the XRD pattern of the
5%Zn5Cu10Ce10Al catalysts after test could explain the decrease of
the methanol conversion and the carbon assessment values com-
pared to the 5Cu10Ce10Al where Cu0 species are detected. In order
on the catalysts before and after test.
Fig.
2 shows an EPR signal of 10Ce10Al after test with
∼6 G. The EPR parameters of this signal are consistent with localized
paramagnetic spins on the carbon particles [31]. Thus, the presence
of this EPR signal showing the existence of carbon species adsorbed
on the 10Ce10Al support and explains the low rate of carbon (0.35).
In parallel, Fig. 2 shows an EPR six lines spectrum of high spin
Mn2+ ions with electronic configuration 3d5. It has been attributed
to Mn2+ ions (g = 2.001, A = 87 G) located in a weak axially dis-
torted octahedral crystal field. These Mn2+ species are present as
impurities (6 ppm) in the sample and do not have any catalytic
role.
XRD patterns of 5Cu10Ce, 5%Zn5Cu10Ce, 5Cu10Ce10Al and
5%Zn5Cu10Ce10Al catalysts are shown in Fig. 3. It can be seen
that fluorite type oxide structure of CeO2 (JCPDS: 00-034-0394)
is observed in all samples and remained in oxide form after test.
Besides, aluminum oxide was not shown over the pattern because
of its amorphous phase.
The XRD pattern of the 5Cu10Ce (Fig. 3c) catalyst before test,
shows the presence of diffraction peaks related to CuO species
(JCPDS: 00-045-0937) that remains even after test. After test,
diffraction peaks corresponding to Cu0 (JCPDS: 00-004-0836) were
detected which signify the reduction of CuO species during test. The
results of EPR confirm this fact because the intensity of EPR signal
rise after test due to the dissociation of the bulk CuO which become
more dispersed and detected by EPR.
After adding 5% of ZnO over 5Cu10Ce, CuO species are more
visible over this catalyst before test with the presence of small
diffraction peaks related to ZnO (JCPDS: 00-036-1451). After test,
diffraction peaks related to CuO species disappear and Cu2O (JCPDS:
and Fig. 4) and the H2 consumption calculated theoretically and
measured experimentally before and after test for species that can
be reduced in samples such as CeO2, ZnO and CuO. As it is shown
in Table 3 for the 10Ce10Al, ceria is not completely reduced during
TPR before test (0.59 mmol g−1). In fact, previous studies done in
our laboratory have shown that the high load of ceria lower its dis-
persion over the alumina surface and diminish its reduction [40].
quantity of ceria is not reduced during test (0.27 mmol g−1).
While adding copper (5Cu10Ce10Al), the reduction of ceria was
kept almost the same (0.62 mmol g−1) but 82% of the added copper
was reduced by TPR (1.56 mmol g−1). After test, TPR measurements
(Fig. 4) reveal that a large part of copper was reduced during the test
with the decreasing intensity of the EPR signal and the presence of
Cu0 detected by XRD. The non-reduced fraction during test could
be related to the CuAl2O4 stable phase detected by XRD after test.
After adding 5% of ZnO, no H2 consumption (Table 3) or reduc-
tion peak (Fig. 4) characterizing the reduction of the ZnO species
was revealed. Compared to the 5Cu10Ce10Al, the impregnation of
Zn enhances the reducibility of ceria (1.34 mmol g−1) but decreases