Effect of composition and pretreatment parameters on activity and stability of Cu-Al catalysts for water-gas shift reaction

Article abstract of DOI:10.1002/cctc.201301049

We investigated various Cu species responsible for highly efficient Cu-Al oxide catalyst for the water-gas shift reaction (WGSR). The formation of various Cu species was achieved by systematically varying the Cu-Al composition in the coprecipitated mixed Cu-Al oxides. The Cu-Al composition of 70:30 (Cu-Al-7) was the best for WGSR using the reformate gas composition. In addition, the Cu-Al-7 catalyst reduced under 100 % H₂, was relatively stable with time on stream of 100 h, at higher gas hourly space velocity of 36 201 h ^-1.The structural investigation of our coprecipitated catalysts with varying Cu-Al compositions revealed the formation of nonzero oxidation state copper and metallic Cu to be essential for the observed WGSR activity. In addition, the highest activity and stability of Cu-Al-7 catalysts reduced under 100 % H₂ at lower temperature was attributed to particle-size stabilization and a lower extent of Cu aggregation by Cu₂O and boehmite phases, respectively, along with the formation of various Cu species during the activation protocol for 12 h. Complete CO₂ selectivity without methanation was observed for all the Cu-Al compositions irrespective of their pretreatment conditions. Teamwork brings success: The coexistence of various nonzero oxidation state and metallic copper species in a Cu-Al catalyst with 70 atom % Cu yields the highest CO conversion with a stable time on stream (TOS) of 100 h in the water-gas shift reaction (WGSR). PEM=proton-exchange membrane.

Full text of DOI:10.1002/cctc.201301049

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201301049
Effect of Composition and Pretreatment Parameters on
Activity and Stability of Cu–Al Catalysts for Water–Gas
Shift Reaction
Rasika B. Mane,[a] Dae-Woon Jeong,[b] Atul V. Malawadkar,[a] Hyun-Seog Roh,*[b] and
Chandrashekhar V. Rode*[a]
We investigated various Cu species responsible for highly effi-
cient Cu–Al oxide catalyst for the water–gas shift reaction
(WGSR). The formation of various Cu species was achieved by
systematically varying the Cu–Al composition in the coprecipi-
tated mixed Cu–Al oxides. The Cu–Al composition of 70:30
(Cu–Al-7) was the best for WGSR using the reformate gas com-
position. In addition, the Cu–Al-7 catalyst reduced under 100%
H2, was relatively stable with time on stream of 100 h, at
higher gas hourly space velocity of 36201 hꢀ1.The structural in-
vestigation of our coprecipitated catalysts with varying Cu–Al
compositions revealed the formation of nonzero oxidation
state copper and metallic Cu to be essential for the observed
WGSR activity. In addition, the highest activity and stability of
Cu–Al-7 catalysts reduced under 100% H2 at lower tempera-
ture was attributed to particle-size stabilization and a lower
extent of Cu aggregation by Cu2O and boehmite phases, re-
spectively, along with the formation of various Cu species
during the activation protocol for 12 h. Complete CO2 selectivi-
ty without methanation was observed for all the Cu–Al compo-
sitions irrespective of their pretreatment conditions.
Introduction
Increased concern towards the clean energy source proposed
hydrogen as the ideal fuel having potential applications in vari-
ous fields. Fuel cell is one of the primary technologies utilizing
hydrogen and is considered as an attractive energy system be-
cause of its high power density, low-temperature operation,
low emissions of NOx and greenhouse gases.[1] Hydrogen ob-
tained from hydrocarbons either by steam reforming or auto-
thermal reforming was used as a fuel for polymer electrolyte
fuel cells (PEFCs). This reformed gas contains 1–10% CO, to
which Pt electrode of PEFC is not tolerant.[2] Thus it is necessa-
ry to lower the CO level to <10–50 ppm for fuel cell applica-
tion. The water–gas shift reaction (WGSR, CO+H2O) and pref-
erential CO oxidation (2CO+O2) not only reduce the CO con-
tent but also maximize the H2 yield.[3,4] WGSR is mildly exother-
mic (DH=ꢀ40.6 kJmolꢀ1) and thermodynamically favored at
lower temperatures, but kinetically the reactant gases are not
active to reach the chemical equilibrium at lower tempera-
tures. In addition, WGSR is not affected by the pressure be-
cause there is no change in volume from reactant to product.
Industrially, WGSR is performed by using high- (320–4508C)
and low- (190–2508C) temperature-shift catalysts such as
Fe3O4/Cr2O3 and Cu/ZnO/A2O3, respectively.[5] However, these
catalysts are not viable for small-scale operation because of
the long and tedious pretreatment procedure and are likely to
be pyrophoric.[6] Therefore, there is a need to design highly ef-
ficient, air-tolerant, cost-effective, and low-temperature-active
catalysts for WGSR. Ceria possessing high oxygen storage ca-
pacity has been effectively used as a support for both noble
(Au, Rh, Pt, Pd) as well as non-noble metals (Cu, Ni) for low-
temperature WGSR.[7–10] Nevertheless, cost-intensive factors re-
strict the application of noble metals, hence, the focus is more
on the development of non-noble metal catalysts. Among
these, Cu was the most attractive choice because of its higher
activity for low-temperature WGSR. However, copper catalysts
exhibit lower stability to oxidant gases,[11–13] hence, a challenge
still exists to develop highly active and stable Cu-based cata-
lysts. Studying the effect of different precipitants, Li et al. re-
ported that CuO/CeO2 catalyst using NH3 rather than carbo-
nates, gave much higher CO conversion (ꢁ90% at 200–2508C)
for WGSR for a feed 10% CO, 60% H2, 12% CO2, and balance
N2 at much lower gas hourly space velocity (GHSV) of
3000 hꢀ1.[14] Nanoclusters of noble metals as well as Cu individ-
ually encapsulated in mesoporous ceria prepared by using ex-
pensive SBA-15 as a template were also evaluated for a feed
composition of 5.6% CO and 22.44% H2O different form that
of a reformate gas.[15] Studies on bimetallic CuꢀNi on La-doped
mesoporous ceria exhibited a synergetic effect of Cu in sup-
pressing the methanation reaction and high WGS activity of
Ni.[16] A major drawback associated with ceria is its irreversible
reduction in the presence of excess H2 leading to catalyst de-
[a] R. B. Mane, A. V. Malawadkar, Dr. C. V. Rode
Chemical Engineering and Process Development Division
CSIR-National Chemical Laboratory
Pune 411008 (India)
Fax: (+91)20-2590-2621
E-mail: cv.rode@ncl.res.in
[b] D.-W. Jeong, Prof. H.-S. Roh
Environmental Engineering
Yonsei University (South Korea)
Fax: (+82)33-760-2571
E-mail: hsroh@yonsei.ac.kr
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201301049.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
activation,[17,18] hence, most of the studies using ceria were per-
formed under synthetic gaseous composition without H2.
Besides CeO2, commercially used support, ZnO for the Cu
catalyst has been also studied by modifying it with alkali
metals as well as by adding Al2O3, SiO2, and MnO.[19–23] The ef-
fects of catalyst calcination and reduction temperatures, Cu/
Mn molar ratio, and process parameters such as CO and CO2
concentration on the activity of spinel Cu/MnO were studied
for WGSR under the realistic conditions of fuel cell.[24] Among
various CuB2O4 spinel catalysts, only Cu–Zn–Al having easily re-
ducible Cu gave higher activity for WGSR for H2O/CO (=4.4)
and a gas flow rate of 33 mLminꢀ1.[25] Under the conditions of
relevance to fuel-cell application, the incorporation of CeO2
and ZnO in Cu–Al2O3 did not show any promotional effect and
Cu was the only active site for WGSR.[26] In another recent at-
tempt to enhance the Cu activity, Lin et al. developed well dis-
persed Cu in a bimetallic Cu–Ni catalyst that inhibited the
methanation reaction leading to high catalyst activity in both
medium- and high-temperature ranges (>3508C).[27,28] Cu–Al–
Ox catalysts prepared by various methods were reported to
impart higher activity than commercial Cu/ZnO/Al2O3 catalysts
even for daily start-up and shut-down (DSS) operation in
WGSR, as a result of boehmite-stabilized Cu particles.[29] By cor-
relating the activity and basicity of Cu catalysts, it was shown
that increase in weak basic sites on Cu–AlOx and Cu–ZnOx cata-
lysts increased the WGS activity per Cu0 surface area.[30]
BET surface areas of 70:30 Cu–Al samples after calcination
and reduction under various conditions and those of used cat-
alysts were measured and the results are given in Table 2. The
sample reduced under 5% H2 +N2 exhibited a slight increase
in surface area compared to that of the calcined sample
Table 2. BET surface areas of calcined, reduced, and used Cu–Al-7 cata-
lyst.
Entry
Catalyst
BET surface area [m2 gꢀ1
]
1
2
3
4
5
calcined
reduced (5% H2)
reduced (100% H2)
used Cu–Al after 25 h stable activity
used Cu–Al after 100 h stable activity
32.4
37.1
19.7
9.5
8.2
whereas the sample reduced under 100% H2 exhibited a much
lower surface area of 19 m2 gꢀ1. This is owing to the use of
100% H2 for reduction causing aggregation of Cu particles.[31]
However, the activity results could not be explained based on
catalyst surface area alone (as discussed below).
H2 temperature-programmed reduction (H2–TPR) profiles of
various Cu–Al catalysts in Figure 1 have very interesting fea-
tures. The lowest Cu containing catalyst (10 atom%, Cu–Al-1)
exhibited two peaks, the first prominent in the region of 350–
4008C and another small one at approximately 6008C. Both
Although several attempts were made to enhance stability
and activity of copper catalysts for WGSR, most of these stud-
ies were performed at very low GHSV and/or with non-refor-
mate gas composition (with CO concentration <5%). Never-
theless, structural aspects of Cu–Al2O3 catalyst system seemed
to be more promising, hence, we investigated various Cu spe-
cies formed by varying the Cu–Al composition in the coprecipi-
tated Cu–Al oxides. Screening of these catalysts revealed that
the Cu–Al composition of 70:30 was the best for WGSR using
reformate gas composition at higher GHSV (36201 hꢀ1). This
composition was structurally investigated in detail for under-
standing its higher activity, complete CO2 selectivity, and stabil-
ity.
Results and Discussion
Figure 1. TPR profiles of calcined Cu–Al catalysts of different compositions.
BET surface areas of Cu–Al catalysts with varied compositions
are presented in Table 1. With the increase in Cu loading from
10 to 30 atom%, BET surface area increased considerably, how-
ever, it remained almost constant in the range of 38–43 m2 gꢀ1
with further increase in Cu loading up to 80 atom%.
peaks appeared at higher temperatures than those of all the
other samples indicating that most of the 10 atom% Cu load-
ing diffused into the Al lattice leads to the formation of
CuAl2O4, which is rather difficult to reduce.[32] The shift of H2
consumption peaks to lower temperatures (<3008C) for 30
and 50 atom% Cu loadings (Cu–Al-3 and Cu–Al-5, respectively)
owed to surface enrichment with highly dispersed and, hence,
easily reducible CuO species.[24] The catalyst with 70 atom% Cu
(Cu–Al-7) exhibited a small peak at approximately 250–3008C
and a continued broader peak up to 4008C with a maximum
at 3408C. The presence of the lower temperature peak repre-
sents the reduction of CuO species, but the broader peak in
the high-temperature range indicates the aggregation of the
Table 1. BET surface areas of Cu–Al catalysts.
Entry
Catalyst
BET surface area [m2 gꢀ1
]
1
2
3
4
5
Cu–Al-1
Cu–Al-3
Cu–Al-5
Cu–Al-7
Cu–Al-8
27.7
39.0
38.4
32.4
43.6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
mixed oxide species of Cu. Similarly, a broad reduction peak
observed at approximately 3408C in the case of reduced Cu–
Al-7 sample also indicates the lower extent of reduction of Cu
oxide species. Further increase in Cu loading to 80 atom%
(Cu–Al-8) resulted in surface enrichment with the easily reduci-
ble CuO at lower temperature of 2008C and bulky CuO and/or
isolated Cu species reduced at higher temperature of 2508C.
From these results, the ratio 70:30 was found to be an opti-
mum Cu–Al composition giving the most efficient WGS cata-
lyst.
61.78 (220) [JCPDS file no. 77-0199], which were attributed to
CuO and Cu2O phases, respectively. These results proved that
the stepwise reduction of CuO was started but its complete re-
duction to Cu0 was not observed. Further increase in Cu load-
ing to 80 atom% resulted in the appearance of new peaks at
2q=43.28 (111), 50.48 (200), 74.18 (220) as a result of forma-
tion of metallic Cu phase [JCPDS file no. 85-1326]. Along with
this, Cu2O phase formation was also evident from a broad
peak at 2q=36.78 (111), however, the presence of CuO could
not be ruled out because of the merging of the corresponding
peaks at 2q=35.58 and a very small one at 38.88. Thus, the
extent of reduction was higher, although not complete, in the
catalyst with higher Cu content than that in other catalysts.
These results also corroborate with the TPR characterization.
To understand the difference in activity of Cu–Al-7 catalyst
reduced under different conditions of H2 composition and tem-
perature, their XRD characterization results were compared
with each other and with those of calcined and used samples
(Figure 3). As expected, calcined samples displayed the pres-
ence of only CuO phase evident from the reflections at 2q=
The reduced Cu–Al catalysts were evaluated for WGS reac-
tion, and their XRD patterns are shown in Figure 2. The cata-
lysts with the lower Cu loadings of 10 and 30 atom% did not
display any distinct copper phases indicating fine dispersion of
Cu. However, in the case of 10 atom% Cu loading, the peaks
at 2q=23.88 (111), 27.68 (200), 39.68 (220) [JCPDS file nos. 77-
ꢀ
35.58 (002), 38.88 (111), 48.88 (202), 66.28 (311) (Figure 3A).
XRD pattern of the catalyst reduced under pure hydrogen re-
vealed the peaks at 2q=35.58 (002), 38.88 (111), 48.88 (202)
corresponding to CuO phase. Along with these, predominant
peaks at 2q=36.68 (111), 42.18 (200), 61.78 (220) and less in-
tensity peaks at 2q=43.28 (111), 74.18 (220) attributable to
Cu2O and Cu0, respectively, suggest the stepwise reduction of
CuO. However, under 5% H2 +N2 conditions, the absence of
Cu0 phase and lower intensity peaks of Cu2O confirmed the in-
complete reduction of CuO. As TPR studies (Figure 1) of Cu–Al-
7 showed the maximum at 3408C, the XRD pattern of its
sample reduced under this condition is shown in Figure 3B. As
expected, the peaks at 2q=43.38 (111), 50.48 (200), 74.38
Figure 2. XRD patterns of reduced (5% H2 +N2) Cu–Al catalysts of different
compositions.
2151, 77-2176] could be attribut-
ed to residual K2O (precipitating
agent used was K2CO3). As Cu
loading increased from 50 to
80 atom%, the intensity as well
as the width of the various
peaks also increased indicating
a more crystalline nature of the
catalysts. The sample with
50 atom% Cu exhibited two
peaks at 2q=35.58 (002) and
38.88 (111) indicating the pres-
ence of only CuO [JCPDS file no.
80-1917] that remained unre-
duced even under conditions of
5% H2 at 2008C.
Cu–Al-7
sample
(Cu
70 atom%) exhibited peaks at
2q=35.58 (002), 38.88 (111),
48.88 (202), 58.58 (202), 66.18
(311), 68.18 (220) and less inten-
Figure 3. XRD patterns of Cu–Al-7 catalyst: A) calcined, reduced at 2008C, and used (24 h); B) 5% H2 reduced at
sity peaks at 2q=36.48 (111), 3408C.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
(220) [JCPDS file no. 85-1326] corresponding to metallic Cu
phase were more intense than those for the 5% H2 and 100%
H2 reduced Cu–Al-7 samples at 2008C. The medium- and low-
intensity peaks at 2q=36.48 (111), 61.78 (220) and 38.88 (111)
indicate the presence of Cu2O and a very small amount of un-
reduced CuO phases. In spite of a higher extent of metallic Cu
in Cu–Al-7 reduced at high temperature, its activity was much
lower than that of the low-temperature-reduced samples.
In the case of used catalyst (100% H2 reduced at 2008C),
high-intensity peaks at 2q=43.38 (111), 50.48 (200), 74.18
(220) and 36.68 (111), 61.78 (220), 29.48 (110) confirmed the
progressive reduction of CuO to Cu0 through Cu2O under
WGSR conditions (Figure 3A). However, complete reduction of
Cu2O to Cu0 is still not possible as water is known to oxidize
Cu to Cu2O but not further to CuO.[33] The presence of stable
CuO species (peaks at 2q=35.58 (002), 38.88 (111), 48.88
(202)) was owing to the inaccessibility by surrounding Alx–Oy.
Thus, coexistence of various reduced and oxidized Cu species
contribute to the activity of the developed Cu–Al-7 catalyst.
Used Cu–Al-7 also displayed diffraction peaks at 2q=29.68
(040), 49.38 (200), 67.68 (171) [JCPDS file no. 72-0359] corre-
sponding to boehmite (AlOOH) phase formed under WGS con-
ditions under which steam is one of the prevailing compo-
nents.[34] Thus, not only Cu aggregation is prevented by the
formation of boehmite but also the Cu particle size becomes
stabilized as a result of the presence of Cu2O phase in 100%
H2 reduced and used samples, which was also evidenced from
the high-resolution TEM images (see Figure 5).[35]
tion, because high-temperature reduction was mainly responsi-
ble for the formation of reduced Cu species, which is also evi-
dent from the XRD studies (Figure 3B). If the potential scan
was reversed, the reduction of anode current peak AIII occur-
ring as CIII at 0.4 V and CI peak at ꢀ0.350 V could be related to
metallic Cu species. The maximum reduction peak potential
(Ic,max) for the sample reduced with 100% H2 at 2008C was
higher than those of the samples reduced with 5% H2 at 200
and 3408C owing to the presence of electrochemically active
sites in the former sample, which had maximum WGSR activity.
The effect of reduction conditions on morphology and Cu
aggregation was studied by field-emission SEM and the results
are shown in the Supporting Information, Figure S1. Spherical
Cu particles were observed in a sample reduced under 5% H2
at 2008C (Figure S1A). In contrast, the sample reduced at the
same temperature but with 100% H2 exhibited an appreciable
aggregation of Cu particles (Figure S1B). The spherical mor-
phology was completely transformed to a dendrite-like struc-
ture (Figure S1C) probably owing to a still higher extent of ag-
gregation of smaller Cu particles at 3408C reduction tempera-
tures. A similar observation was made by a comparative study
of HRTEM images as shown in Figure S2. These structural
changes contributed to the difference in activity of the rele-
vant samples.
The effect of different conditions on the particle size, mor-
phology, and fringe patterns of Cu–Al-7 catalyst was investigat-
ed in more detail by HRTEM characterization (Figure 5). As
shown in Figure 5A, Cu–Al-7 catalyst reduced under 5% H2
consists of bigger size (14–16 nm) particles than the sample re-
duced under 100% H2 (5–7 nm, Figure 5B). It is most likely
that 100% H2 reduction led to the formation of smaller partial-
ly and completely reduced Cu2O and metallic Cu particles, re-
spectively, as confirmed by XRD analysis. Under 5% H2 reduc-
tion conditions, the catalyst mainly exists as CuO confirmed by
the fringe pattern in Figure 5C that matches very well with the
XRD results discussed above. Reduction under 100% H2
formed the catalyst with a majority of smaller size Cu2O and
Cu0 particles, as seen from the fringe pattern in Figure 5D. In
addition, the needlelike typical morphology of boehmite was
also observed (Figure S3B–D).[37] Even after the reaction time
of 24 and 100 h, particle size was stable in the range of 4–
7 nm as shown in Figure S4A. The fringe pattern (Figure S4B)
of 24 h used sample indicates that metallic Cu phase was
formed during the reaction. In the case of the 100 h used
sample, a number of small Cu metal particles (5 nm) were seen
on the boehmite (Figures 5E, S3D), the latter helped to pre-
vent the aggregation of metal particles. The fringe pattern of
this sample shown in Figure 5F confirmed the existence of
metallic Cu, Cu2O, and boehmite (AlO(OH)) phases.[29] The for-
mation of the boehmite phase was also evident from the SEM
images without any change in needlelike morphology in the
used samples of 24 and 100 h reaction time, as shown in
Figure 6. Cu metal aggregation was visible for the particles
separated from the boehmite phase.
The active sites present in Cu–Al-7 catalyst reduced under
various conditions were also estimated by cyclic voltammetry.
In Figure 4, cyclic voltammograms of 100% and 5% H2 re-
duced Cu–Al-7 catalysts are shown at 200 and 3408C. The sam-
ples reduced with 100% H2 at 2008C and 5% H2 at 3408C ex-
hibited three anodic current peaks, namely, AI, AII, and AIII at
ꢀ0.46, ꢀ0.20, and 0.157 V related to the formation of Cu2O,
CuCl salt layer, and CuO, respectively.[36] In contrast, the sample
reduced with 5% H2 at 2008C exhibited a single shallow
anodic current peak AIII at ꢀ0.20 V. The current density and the
potential of anodic peak AI increased with the increase in re-
duction temperature from 200 to 3408C even for 5% H2 condi-
The inhibition of Cu metal particle aggregation by in situ for-
mation of boehmite leads to high stability of the catalysts, as
discussed below.
Figure 4. Cyclic voltammograms of Cu–Al-7 catalysts reduced at 200 and
3408C.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
4008C (70%). Further increase in
Cu loading to 80 atom% low-
ered the CO conversion dramati-
cally to 60% at 4008C, which is
even lower than that of Cu–Al-5
(Cu loading=50 atom%). We
strongly believe that, although
Cu is an active component in
the catalyst, its appropriate con-
centration along with Al to form
active species is necessary for
the optimum activity. Similar ob-
servations were made in the
case of Cu/MnO and other Cu-
based catalysts and it was con-
cluded that the WGSR activity
not only depends on Cu metal
but also on MnO and the weak
basicity of the support (ZnOx,
AlOx).[24,30] Cu–Al-7 catalyst exhib-
ited
maximum
conversion
among all the catalysts, at all
temperatures. Generally, CO con-
version is limited by temperature
<3508C, with higher surface
coverage by CO2 at lower tem-
peratures retarding the catalyst
activity predominantly.[24] Hence,
achieving higher CO conversion
is obvious at higher tempera-
ture. At the same time, another
contributing feature of our cata-
lyst was the higher extent of re-
duction of Cu species to metallic
Cu at higher temperatures, as
shown by the TPR results. Inter-
estingly, complete selectivity to
CO2 without formation of meth-
ane was achieved irrespective of
Cu loading and the reaction
temperature (Figure 7B).
From the activity screening
studies, Cu–Al-7 catalyst was
found to give the best per-
formance, hence, further investi-
gation on the effect of catalyst
pretreatment parameters on
WGSR activity was performed
Figure 5. HRTEM images of Cu–Al-7 catalysts: A) 5% H2 reduced, showing particle size of 14–16 nm; B) 100% H2
reduced, showing particle size of 5.6–6.8 nm; C) fringe pattern of 5% H2 reduced sample; D) fringe pattern of
100% H2 reduced sample; E) used 100 h sample; F) fringe pattern of used 100 h sample.
In Figure 7A, the activity of Cu–Al catalysts having different
compositions is shown, for CO conversion at different tempera-
tures. Cu–Al-1 catalyst with the lowest Cu loading of
10 atom% exhibited the least activity with CO conversion in
the range of 5–30% for the rise in temperature from 200–
4008C. The CO conversion increased linearly from 30 to 69%
with increase in Cu loading from 10–50%, at 4008C. In con-
trast, for 70 atom% Cu loading, maximum conversion of 76%
was observed at 3608C while lower conversion obtained at
over the same catalyst. In Figure 8A, the comparison of activi-
ties is shown of calcined Cu–Al-7 catalyst are shown as well as
that of the reduced samples using 5% and 100% H2 at 200
and 3408C conditions. Cu–Al-7 reduced under 5% H2 gave
higher CO conversion than calcined and 100% H2 reduced
sample, for the temperature range of 200–3008C. With increase
in temperature to 4008C, CO conversion was the same at 73%
for all the three catalysts. This observation and the fact that
the calcined sample showed a similar or better performance
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
5
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
than 100% H2 reduced sample
at lower temperature indicate
that the active species formed
may be similar under these reac-
tion conditions. However, at
3608C, Cu–Al-7 reduced at 5%
and 100% H2 exhibited higher
CO conversion of 76 and 79%,
respectively, than that of the cal-
cined sample. The higher reduc-
tion temperature of 3408C was
responsible for the much higher
extent of metallic Cu in Cu–Al-7
as evidenced from the charac-
terization
results
discussed
above. However, interestingly,
this highly reduced sample ex-
hibited the lowest WGSR activity
(Figure 8A) partly because of
the high-temperature aggrega-
tion of Cu particles. In a control
experiment, very poor activity of
pure Cu metal undoubtedly
confirmed that mixed Cu spe-
cies contributed to the efficient
WGSR performance. However,
CO2 selectivity was not influ-
enced by the catalyst pretreat-
ment conditions (Figure 8B).
Figure 6. SEM images of Cu–Al-7 catalysts, 100% H2 reduced samples: A) reduced; B) used 24 h; C) used 100 h.
Therefore, as discussed above, changes in the reduction con-
ditions affecting BET surface area, extent of reduction, and Cu
particle size, together affect the WGSR activity of the same cat-
alyst. The observed highest WGSR activity and selectivity of
Cu–Al-7 catalyst demonstrate the efficient redox mechanism
operating as evident from the coexistence of different Cu spe-
cies (Cu2+, Cu1+, and Cu0) in reduced and recovered sam-
ples.[38] Such nonzero oxidation states of Cu are possible as
a result of the presence of Al species, which, in turn, facilitate
the electron acceptance ability of active copper. From the pres-
ence of CuO in the calcined and 5% H2 reduced samples, CuO
and Cu2O phases in 100% H2 reduced, and Cu2O and Cu0 in
the used (after 24 h at 3608C) 100% H2 samples, the reaction
mechanism seen in Scheme 1 is proposed. In this scheme,
WGS reaction involving CO!CO2 and H2O!H2 is also respon-
sible for the interconversion of various Cu species.[39]
As the samples of Cu–Al-7 reduced under 5% and 100% H2
gave comparable WGSR activity at 3608C, their stability was
tested for continuous operation of 25 h and the results are
shown in Figure 9A. It is evident that CO conversion decreased
appreciably with time on stream for 5% H2 reduced sample
compared to the sample reduced with 100% H2. The 100% H2
reduced sample was further tested for extended time of 100 h
(Figure 9B), which exhibited marginal decrease in activity from
79 to 73% for the initial 24 h as a result of a significant de-
crease in BET surface area from 19.7 to 9.5 m2 gꢀ1 (Table 2).
After 25 h, the activity was approximately constant because
the surface area did not vary considerably (8.2 m2 gꢀ1).The con-
Figure 7. Activity of Cu–Al catalysts of different compositions: A) CO conver-
sion; B) CO2 and CH4 selectivity.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
6
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 9. Time-on-stream activity of reduced Cu–Al-7 catalysts: A) compari-
son of 5% and 100% H2 reduced catalysts; B) 100 h stability of 100% H2 re-
duced catalyst.
Figure 8. Effect of pretreatment conditions on the activity of Cu–Al-7 cata-
lysts: A) CO conversion with temperature over calcined and 5% and 100%
H2 reduced samples; B) CO2 and CH4 selectivity.
100% H2 at 2008C exhibited relatively stable activity with time
on stream. The coexistence of nonzero oxidation state copper
and metallic Cu in this catalyst, as proved by H2 temperature-
programmed reduction, XRD, and high-resolution transmission
electron microscopy, was an essential aspect for the observed
WGSR activity. The change in reduction conditions played an
important role for the variation in WGSR activity of the same
catalyst. The contributing factors for the highest activity and
stability of 100% H2 reduced Cu–Al-7 catalyst were as follows:
particle stabilization owing to the Cu2O phase, formation of
zero and nonzero oxidation state Cu species, and the lower
extent of Cu aggregation owing to boehmite phase. Complete
CO2 selectivity without methanation was observed for all the
Cu–Al compositions irrespective of their pretreatment condi-
tions.
Scheme 1. Redox species of Cu in WGSR reaction over Cu–Al catalyst.
sistent activity and stability of 100% H2 reduced sample was
mainly owing to the presence of various redox active Cu spe-
cies (mainly Cu2O and Cu0), particle size stabilization, and
a lower extent of Cu aggregation owing to the formation of
Cu2O and boehmite phases.
Experimental Section
Conclusions
Materials
Mixed Cu–Al oxides were prepared by a coprecipitation
method with varying Cu–Al compositions for efficiently catalyz-
ing the water–gas shift reaction (WGSR). Although all the cata-
lysts studied in this work were selective to CO2 formation by
suppressing methanation reaction, only Cu–Al-7 (70:30) cata-
lyst exhibited the highest CO conversion within the tempera-
ture range from 200 to 4008C at the gas hourly space velocity
of 36201 hꢀ1. In addition, Cu–Al-7 catalyst reduced under
Nitrate precursors of copper aluminum and potassium carbonate
were purchased from Loba Chemie, Mumbai, India.
Catalyst preparation
A series of Cu–Al catalysts were prepared by a coprecipitation
method involving simultaneous addition of nitrate precursor solu-
tions of Cu, Al, and 0.2m aqueous K2CO3 in a round bottom flask
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
7
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
containing a volume of 5–10 mL of water at room temperature.
The obtained precipitate was left to digest for 4–5 h and then fil-
tered and washed with deionized water to remove the traces of
potassium. The precipitate was dried in an oven at 1008C for 5–
8 h. The catalysts with different atom% of Cu, namely, 10, 30, 50,
70, and 80 atom% were prepared and designated as Cu–Al-1, Cu–
Al-3, Cu–Al-5, Cu–Al-7, and Cu–Al-8, respectively. All the prepared
catalysts were calcined at 4008C for 3 h.
methane (SRM: H2O+CH4 =3H2 +CO) to avoid coke formation.[40]
GHSV of 36201 hꢀ1 was used to screen the catalysts in this study.
Water was fed by using a syringe pump and was vaporized at
1508C upstream of the reactor. The reformed gas was chilled,
passed through a trap to condense the residual water, and then
flowed to the on-line micro gas chromatograph (Agilent 3000).
Acknowledgements
One of the authors, RBM, thanks the Council of Scientific and In-
dustrial Research (CSIR) New Delhi, for the award of a senior re-
search fellowship.
Catalyst characterization
The BET surface area was measured by nitrogen adsorption at
ꢀ1968C using an ASAP 2000 (Micromeritics).
Field-emission SEM of catalysts samples was performed on a Hitachi
S-4200 instrument.
Keywords: aluminum · copper · fuel cells · hydrogen ·
reduction
TPR experiments were performed in a BEL-CAT (BEL Japan Inc.).
Typically, calcined sample (0.1 g) was loaded into a quartz reactor,
and TPR was performed by using 10% H2 in Ar from 20 to 6008C
with a heating rate of 108C minꢀ1. The sensitivity of the detector
was calibrated by reducing a known mass of NiO.
[1] J. A. Rodriguez, P. Liu, J. Hrbek, J. Evans, M. Perez, Angew. Chem. Int. Ed.
2007, 46, 1329–1332; Angew. Chem. 2007, 119, 1351–1354.
[2] T. Shishido, M. Yamamoto, I. Atake, D. Li, Y. Tian, H. Morioka, M. Honda,
T. Sano, K. Takehira, J. Mol. Catal. A 2006, 253, 270–278.
[3] C. H. Kim, L. T. Thompson, J. Catal. 2005, 230, 66–74.
[4] P. Panagiotopoulou, D. I. Kondarides, J. Catal. 2004, 225, 327–336.
[5] C. Martos, J. Dufour, A. Ruiz, Int. J. Hydrogen Energy 2009, 34, 4475–
4481.
[6] W. Ruettinger, O. Ilinich, R. J. Farrauto, J. Power Sources 2003, 118, 61–
65.
[7] S. Hilaire, X. Wang, T. Luo, R. J. Gorte, J. Wagner, Appl. Catal. A 2001,
215, 271–278.
[8] R. Si, Y. W. Zhang, S. J. Li, B. X. Lin, C. H. Yan, J. Phys. Chem. B 2004, 108,
12481–12488.
[9] D. W. Jeong, H. S. Potdar, H.-S. Roh, Catal. Lett. 2012, 142, 439–444.
[10] Q. Fu, W. Deng, H. Saltsburg, M. Flytzani-Stephanopoulos, Appl. Catal. B
2005, 56, 57–68.
[11] J. Wu, M. Saito, J. Catal. 2000, 195, 420–422.
[12] T. Shishido, S. Nishimura, Y. Yoshinaga, K. Ebitani, K. Teramura, T. Tanaka,
Catal. Commun. 2009, 10, 1057–1061.
[13] Y. Tanaka, T. Takeguchi, R. Kikuchi, K. Eguchi, Appl. Catal. A 2005, 279,
59–66.
Cyclic voltammetry of the catalysts was performed on an Autolab
PGSTAT30 (Eco-Chemie) instrument having an electrochemical cell
with a three-electrode system. Test cell with Hg/Hg2SO4 as the ref-
erence electrode and a platinum foil as the counter electrode. For
the preparation of the working electrode (WE), the catalyst slurry
was made by sonicating a 5 mg sample of the catalyst in a 3:2
EtOH/H2O mixture (1 mL). The catalyst slurry was drop-coated on
a previously polished (by using 0.3 mm alumina slurry) glassy
carbon (GC) electrode surface (geometrical area=0.07065 cm2) and
0.01 wt% Nafion solution (2.5 mL) was applied on the whole sur-
face of the electrode to yield a uniform thin film. The electrode
was dried for 2 h in air and was used as the WE for the electro-
chemical investigations. Measurements were made in 3 wt% KCl
solution at a scan rate of 20 mVsꢀ1 over the potential range of
ꢀ0.8–0.5 V.
Powder XRD patterns of all the samples were recorded on a PANa-
lytical X’pert Pro diffractometer using CuKa (1.5418 ꢁ) radiations
with a Ni filter and a X’celerator solid-state detector.
[14] L. Li, L. Song, H. Wang, C. Chen, Y. She, Y. Zhan, X. Lin, Q. Zheng, Int. J.
Hydrogen Energy 2011, 36, 8839–8849.
[15] C. Wen, Y. Zhu, Y. Ye, S. Zhang, F. Cheng, Y. Liu, P. Wang, F. Tao, ACS
Nano 2012, 6, 9305–9313.
The particle size and morphology of the catalyst were determined
by using a transmission electron microscope (HRTEM), model JEOL
1200 EX. A small amount of the solid sample was sonicated in eth-
anol for 1 h. A drop of the prepared suspension was deposited on
a Ni grid coated with a carbon layer and the grid was dried at RT
before analysis. Scanning electron micrographs were obtained on
a JEOL JSM 500 model.
[16] J. Lin, P. Biswas, V. V. Guliants, S. Misture, Appl. Catal. A 2010, 387, 87–
94.
[17] J. M. Zalc, V. Sokolovskii, D. G. Loffler, J. Catal. 2002, 206, 169–171.
[18] S. Takenaka, T. Shimizu, K. Otsuka, Int. J. Hydrogen Energy 2004, 29,
1065–1073.
[19] C. H. Bartholomew, R. J. Farrauto, Fundamentals of Industrial Catalytic
Processes, Wiley, Hoboken, NJ, 2006.
[20] T. Shishido, Y. Yamamoto, H. Morioka, K. Takai, K. Takehira, Appl. Catal. A
2004, 263, 249–253.
[21] Y. Tanaka, T. Utaka, R. Kikuchi, K. Sasaki, K. Eguchi, Appl. Catal. A 2003,
238, 11–18.
Catalytic activity
[22] H. Schaper, E. B. M. Doesburg, L. L. Van Reijen, Appl. Catal. A 1983, 7,
211–220.
[23] M. S. Spencer, Top. Catal. 1999, 8, 259–266.
[24] Y. Tanaka, T. Utaka, R. Kikuchi, T. Takeguchi, K. Sasaki, K. Eguchi, J. Catal.
2003, 215, 271–278.
[25] S. Boumaza, A. Auroux, S. Bennici, A. Boudjemaa, M. Trari, A. Bouguelia,
R. Bouarab, React. Kinet. Mech. Catal. 2010, 100, 145–151.
[26] N. A. Koryabkina, A. A. Phatak, W. F. Ruettinger, R. J. Farrauto, F. H. Ri-
beiro, J. Catal. 2003, 217, 233–239.
[27] J. Lin, V. V. Guliants, ChemCatChem 2011, 3, 1426–1430.
[28] J. Lin, V. V. Guliants, ChemCatChem 2012, 4, 1611–1621.
[29] S. Nishimura, T. Shishido, K. Ebitani, K. Teramura, T. Tanaka, Appl. Catal. A
2010, 387, 185–194.
Activity tests were performed from 200 to 4008C under atmos-
pheric pressure in a fixed-bed microtubular quartz reactor with an
inner diameter of 4 mm. The catalyst charged was 48 mg. T-union
was employed at the exit of quartz reactor to install a thermocou-
ple. A thermocouple was inserted into the catalyst bed to measure
the reaction temperature. Prior to each catalytic measurement, the
catalyst was reducꢀe1d in 5% H2/N2 from RT to 2008C at a heating
rate of 3.38C min
and the temperature was maintained for 2 h.
The simulated reformed gas consisted of 6.5 vol% CO, 7.1 vol%
CO2, 0.7 vol% CH4, 42.4 vol% H2, 28.7 vol% H2O, and 14.5 vol% N2.
The feed H2O/(CH4+CO+CO2) ratio was fixed at 2.0 because of the
fact that a H2O/CH4 ratio is typically 3.0 in steam reforming of
[30] K. Sagata, N. Imazu, H. Yahiro, Catal. Today 2013, 201, 145–150.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
8
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[31] C. Rhodes, G. J. Hutchings, A. M. Ward, Catal. Today 1995, 23, 43–58.
[32] F. Severino, J. L. Brito, J. Laine, J. L. G. Fierro, A. L. Agudo, J. Catal. 1998,
177, 82–95.
[33] D. F. Shriver, P. W. Atkins, Inorganic chemistry, Oxford University Press,
3rd ed., 1999.
[38] C. S. Chen, T. W. Lai, C. C. Chen, J. Catal. 2010, 273, 18–28.
[39] J. B. Ko, C. M. Bae, Y. S. Jung, D. H. Kim, Catal. Lett. 2005, 105, 157–161.
[40] H.-S. Roh, H. S. Potdar, D.-W. Jeong, K. Kim, J. Shim, W. Jang, K. Y. Koo,
W. L. Yoon, Catal. Today 2012, 185, 113–118.
[34] X. Y. Chen, S. W. Lee, Chem. Phys. Lett. 2007, 438, 279–284.
[35] R. B. Mane, A. M. Hengne, A. A. Ghalwadkar, V. Subramanian, P. H.
Mohite, H. S. Potdar, C. V. Rode, Catal. Lett. 2010, 135, 141–147.
[36] Y.-G. Chun, S.-I. Pyun, C.-H. Kim, Mater. Lett. 1994, 20, 265–270.
[37] H. Ivankovic, E. Tkalcec, R. Nass, H. Schmidt, J. Eur. Ceram. Soc. 2003, 23,
283–292.
Received: December 9, 2013
Revised: February 17, 2014
Published online on && &&, 0000
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
9
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
R. B. Mane, D.-W. Jeong,
A. V. Malawadkar, H.-S. Roh,* C. V. Rode*
&& – &&
Effect of Composition and
Pretreatment Parameters on Activity
and Stability of Cu–Al Catalysts for
Water–Gas Shift Reaction
Teamwork brings success: The coexis-
tence of various nonzero oxidation state
and metallic copper species in a Cu–Al
catalyst with 70 atom% Cu yields the
highest CO conversion with a stable
time on stream (TOS) of 100 h in the
water–gas shift reaction (WGSR). PEM=
proton-exchange membrane.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&10
&
ÞÞ
These are not the final page numbers!