Benzaldehyde reduction over Cu-Al-O bimetallic oxide catalyst. Influence of pH during hydrothermal synthesis on the structural and catalytic properties

Article abstract of DOI:10.1016/j.molcata.2014.09.037

The Cu-Al-O bimetallic oxide was prepared by varying the pH during synthesis. The samples were characterized by several techniques (BET, TGA, XRD, LRS, EDX and XPS). Their behavior during reduction by hydrogen was examined by in situ X-ray diffraction and by X-ray photoelectron spectroscopy. The catalysts were studied in the reduction of benzaldehyde in the 150-200 °C temperature range. The choice of pH value of the material preparation influenced strongly the textural and structural properties of the bimetallic system, as well as their reactivity. When the catalyst was prepared in highly basic medium (pH = 12), benzyl alcohol was obtained with high selectivity (96%) and at high conversion (78%), while toluene was the only product observed when preparation was carried out at pH = 8. The marked divergence in behavior from the two systems may be accounted for by their acid-base and reducibility properties, dispersion and environment of copper.

Full text of DOI:10.1016/j.molcata.2014.09.037

Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Benzaldehyde reduction over Cu–Al–O bimetallic oxide catalyst.
Influence of pH during hydrothermal synthesis on the structural and
catalytic properties
N. Haddada,∗, A. Saadia, A. Löfbergb, R.N. Vannierb, E. Bordes-Richardb, C. Rabiaa
a Laboratoire de Chimie du Gaz Naturel, Faculté de Chimie, USTHB, BP 32 El Alia, 16111 Bab Ezzouar, Alger, Algeria
b Unité de Catalyse et de Chimie du Solide, UMR CNRS 8181, Université Lille1, ENSCL, Cité Scientifique, 59655 Villeneuve d’Ascq, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The Cu–Al–O bimetallic oxide was prepared by varying the pH during synthesis. The samples were
characterized by several techniques (BET, TGA, XRD, LRS, EDX and XPS). Their behavior during reduc-
tion by hydrogen was examined by in situ X-ray diffraction and by X-ray photoelectron spectroscopy.
The catalysts were studied in the reduction of benzaldehyde in the 150–200 ◦C temperature range. The
choice of pH value of the material preparation influenced strongly the textural and structural proper-
ties of the bimetallic system, as well as their reactivity. When the catalyst was prepared in highly basic
medium (pH = 12), benzyl alcohol was obtained with high selectivity (96%) and at high conversion (78%),
while toluene was the only product observed when preparation was carried out at pH = 8. The marked
divergence in behavior from the two systems may be accounted for by their acid–base and reducibility
properties, dispersion and environment of copper.
Received 16 July 2014
Received in revised form
26 September 2014
Accepted 29 September 2014
Available online 7 October 2014
Keywords:
Copper aluminum oxide
Hydrothermal method
Reduction of benzaldehyde
Benzyl alcohol
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
of the surface area, of the copper particle size and of its dispersion
regarding its catalytic activity were highlighted in the literature.
Benzyl alcohol is an important chemical product, used as an
intermediate in the synthesis of pharmaceuticals and fragrances
[1,2]. It is also used as raw material in the manufacture of various
esters in both cosmetics and flavor industries [3].
In the case of glycerol hydrogenolysis over co-precipitated Cu–ZnO
catalysts, Wang and Liu [12] reported that the crystallite size of
copper is very sensitive to the pH during preparation of the mate-
rial. Indeed, the smaller Cu particles found to be formed in basic
solution (pH = 12), lead to higher conversion and higher propylene
glycol selectivity. In the case of methanol synthesis a dependence
of the activity on the copper specific surface area was proven [13]
and today, it is widely accepted that metallic copper clusters are
the active sites [13–15].
It is well-known that adding another metal is a means to
enhance the catalytic properties with respect to monometal-
lic systems [14]. Aluminum as alumina is a good promoter for
copper-based catalysts, it plays a structural role, thereby minimiz-
ing or preventing the sintering of Cu active phases [16]. Several
authors demonstrated that mixed Cu–Al mesoporous oxides are
very efficient catalysts for the selective hydrogenation of conju-
gated unsaturated carbonyl compounds. In the case of the selective
hydrogenation of cinnamaldehyde (CNA), Valange et al. compared
the performance of Cu–Al mesoporous oxides to that of classical
Cu-impregnated alumina catalysts [17]. The authors attributed the
remarkable selectivity to unsaturated alcohol (hydrogenation of
the carbonyl group of CNA) to the strong interaction of nanometer
The catalytic reduction of benzaldehyde to benzyl alcohol in the
gas phase is an interesting alternative to liquid phase processes.
Nowadays, most heterogeneous catalysts are based on transition
metals [4–6], alkaline-earth oxide and layered double hydroxides
[7,8]. Systems involving supported noble metals and other sup-
ported metal oxides have been also reported [4,7]. However, over
all these systems, benzyl alcohol selectivity does not exceed 80%
due to the benzaldehyde hydrogenolysis that generates toluene
and benzene. Therefore, the selectivity to the hydrogenation of the
aldehyde group remains a challenge [7].
Copper is suggested as a very effective metal to catalyze the
hydrogenation of aromatic aldehydes to alcohols [9–11]. It is usu-
ally supported on Al2O3, SiO2, TiO2, CeO2 and ZrO2 in order to
increase the dispersion of the active phase [11]. Recently, the roles
∗
Corresponding author. Tel.: +213 21 24 79 64; fax: +213 21 24 73 11.
E-mail address: akrour naima@hotmail.com (N. Haddad).
http://dx.doi.org/10.1016/j.molcata.2014.09.037
1381-1169/© 2014 Elsevier B.V. All rights reserved.

208
N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
0.1 ◦C/s heating rate between two temperatures. The sample was
displayed on a gold sheet and N2 or 3% H2 in N2 was flowed in the
chamber (5 L/h). After measurement, all samples were cooled down
to room temperature at a cooling rate of 0.3 ◦C/s.
Laser Raman spectra (LRS) were recorded on a Labraminfinity
laser Raman spectrometer (XY-DILOR) equipped with an optical
multichannel charge coupled device liquid nitrogen-cooled detec-
tor. The laser intensity (Ar+, 514.5 nm) was reduced by various
filters (<1 mW) and the data were treated by the Lasbpec software.
The spectral resolution and the accuracy in the Raman shifts are
estimated to be 2 cm−1. A tenth particles was examined for each
sample to check its homogeneity.
sized Cu0 “islets” with the Al mesoporous walls. The conjugated
C bond is more readily hydrogenated on larger Cu0 clusters,
which are weakly interacting with the support.
C
Aluminum is usually used as alumina support but we found
interesting to add it in smaller amounts, like a promoter. In this
paper, the synthesis of bimetallic oxide, Cu–Al–O, by hydrothermal
method is proposed, instead of the more conventional impregna-
tion or ionic exchange. The effect of the pH of preparation on the
structural, textural and surface properties of the system is exam-
ined by different techniques. The catalytic performance of these
systems in the benzaldehyde reaction with or without hydrogen is
studied and discussed.
Scanning electron microscopy (SEM) and X-ray energy dis-
persive microanalysis (EDS) were carried out on HITACHI 4100S
apparatus at 6 kV. Catalysts were ground as fine particles and
mechanically dispersed on an electrically conductive carbon tape
which was placed on an aluminum disk.
2. Experimental
2.1. Preparation of the bimetallic Cu–Al–O oxide
X-ray photoelectron spectroscopy (XPS) was carried out on
Escalab 220 XL spectrometer (Vacuum Generators), using a focused
monochromatic Al K␣ radiation (13 kV, 20 mA). Samples were pre-
liminary degassed under ultra-high vacuum (10−7 Torr) during one
night. Post-reaction catalyst samples were sealed in their reac-
tor tubes under He and transferred to a glove box under inert
atmosphere. Samples were then loaded into the analysis cham-
ber without exposure to atmospheric air. In order to observe the
effect of reduction of Cu, the samples were treated by a reduc-
ing mixture (H2/N2 = 40/60) for 60 min at 100, 150 and 200 ◦C in
the pre-treatment chamber of the VG-ESCALAB apparatus and then
introduced (without exposure to air) inside the vacuum chamber
for acquisition of XPS spectrum.
The catalysts were prepared by precipitation of copper and alu-
minum precursors via hydrothermal method. Given amounts of
cuprite oxide (Cu2O, purity >99.9%), aluminum nitrate nonahy-
drate (Al (NO3)3·9H2O, purity >99.9%) in the 1:1 atomic ratio were
dissolved in deionized water. The mixture was stirred at ambient
temperature for 15 min. Then a precise volume of sodium hydrox-
ide solution (NaOH 2.5 M) was added to the mixture to adjust the
pH to 8 or 12, and to control metal oxide solubility. The stirring
was continued for 45 min. Then the as-obtained solution was trans-
ferred into a 40 ml Teflon-lined autoclave, which was sealed and
maintained at 220 ◦C for 65 h. At the end, the autoclave was let
to cool down to room temperature. The precipitate was filtered
and washed several times alternatively with absolute alcohol and
deionized water. The product was dried overnight at 80 ◦C. The
obtained solid precursors were heated at 500 ◦C in nitrogen flow
for 3 h. The samples were named CuAl-8 and CuAl-12 for pH = 8
and pH = 12, respectively.
2.3. Catalytic testing
The catalytic transformation of benzaldehyde was carried out
in a glass tubular fixed-bed reactor using 0.1 g of sample pow-
der at atmospheric pressure and 100–200 ◦C range temperature.
Pure hydrogen or pure nitrogen was fed at a total flow rate of
2.2. Physicochemical analysis
50 cm3 min−1
.
A number of physico-chemical methods were used for charac-
terization of solids before and after heat treatment.
The composition of the samples was measured through induc-
tively coupled plasma-optique emission spectroscopy (ICP-OES)
with a 700 Series, Agilent Technologies apparatus.
The specific surface area (SSA) of the catalyst was determined
by nitrogen adsorption at −196 ◦C with a Micromeritics ASAP2010
apparatus.
Thermal analyses (TGA/DSC) were carried out with a TA-
Instrument SDT-2960 Thermobalance in argon flow up to 600 ◦C.
Ten to fifteen milligrams of solid were heated at 5 ◦C/min heating
rate.
Gaseous benzaldehyde (4.8 Torr) was delivered by bubbling N2
(250 Torr) in liquid benzaldehyde maintained at constant temper-
ature (50 ◦C) in a saturator. The gaseous reactant and products
were heated up-stream and out-stream to avoid condensation.
They were analyzed on-line by a FID gas chromatograph (Delsi
IGC 121 ML) equipped with a 10% CP-SIL 8 CB/Chromosorb W col-
umn. Each reaction temperature was maintained constant until the
corresponding steady-state was reached as indicated by the gas
chromatograph analysis of the exit gases (ca. 2 h). Before testing,
the catalyst was in situ pre-treated for 2 h at 300 ◦C in nitrogen
flow at 20 cm3 min−1
.
Temperature-programmed reduction patterns were obtained
on 2910 Micromeritics apparatus. Thirty milligrams of catalyst
were loaded and were first treated in argon at room tempera-
ture for 1 h. The samples were then contacted with H2/Ar mixture
(H2/Ar = 1.5; total flow 30 cm3 min−1) and heated at 6 ◦C/min up to
1000 ◦C. The hydrogen consumption was analyzed by on-line gas
chromatography.
X-ray powder diffraction (XRD) was performed on a D8 advance
Bruker AXS diffractometer equipped with a Cu anode (CuK␣) and a
1D LynxEye PSD detector. Patterns were collected at room temper-
ature, in the 10–90◦ 2ꢀ range, with a 0.02◦ step and 10 s counting
time per step. EVA software was used for phase identification.
X-ray diffraction at variable temperature under N2 atmosphere
(N2-HT-XRD) or hydrogen (H2-HT-XRD) was carried out on the
same apparatus equipped with XRK 900 chamber and a Vantec
detector up to 800 ◦C. Diagrams were collected every 25 ◦C with
3. Results and discussion
3.1. Thermal analysis of CuAl-8 precursor
The decomposition of CuAl-8 precursor (sample after overnight
drying at 80 ◦C) was studied by TG-DTA under argon in 25–600 ◦C
range (Fig. 1). The TG curves show that the CuAl-8 precursor is
decomposed in two well-defined steps. The first step at 33 to
ca.100 ◦C can be attributed to the dehydroxylation of boehmite
phase (observed weight loss: ∼1.5%). The second decomposition
step which starts at 198 ◦C, ends at 483 ◦C (observed weight loss:
4.7%), is attributed to the hydroxyl and nitrate ions departure.
Above 483 ◦C, no weight loss is observed, indicating that the oxides
are formed from ca. 500 ◦C.

N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
209
Fig. 1. Thermal analyses (TG-DTA) in N2 of CuAl-8 precursor.
Fig. 3. XRD patterns of CuAl-8 and CuAl-12 catalysts after heat treatment at 500 ◦C.
♦: CuO tenorite, *: Al2O3 alumina.
To identify the phase transformations, the CuAl-8 precursor
was also studied by HT-XRD up to 600 ◦C under nitrogen flow. At
room temperature, the precursor consisted of three crystal phases:
boehmite AlO(OH) (JCPDS 00-021-1307), rouaite or gerhardtite (the
patterns are very close) Cu2NO3(OH)3 (JCPDS 04-011-9699) and
tenorite CuO (JCPDS 04-007-1375). With increasing temperature,
the peaks corresponding to the rouaite phase disappeared com-
pletely at 175 ◦C, followed by that of the boehmite phase at 425 ◦C.
At this temperature (425 ◦C) only CuO (tenorite) was observed with,
in addition, the appearance of one peak at ca. 2ꢀ = 47◦ attributed to
Al2O3 (JCPDS 01-074-2206) (Fig. 2).
CuAl-12 compared to CuAl-8 or commercial CuO. To account for
that, the formation of Cu–O–Al mixed phase, not detected by XRD,
but that could be present as microdomains, is proposed.
3.3. Chemical composition and textural properties
The Cu/Al atomic ratio of CuAl-8 and CuAl-12 catalysts obtained
by ICP-OES analysis after heat treatment at 500 ◦C is shown in
Table 1. As it can be seen, the Cu/Al atomic ratio is almost similar to
that expected from the initial concentrations of the solutions used
during synthesis (Cu/Al = 1). This result confirms that practically all
the metals introduced in hydrothermal preparation are preserved
in the catalyst.
3.2. Structural properties of CuAl-8 and CuAl-12 after heat
treatment
The textural properties differ according to the pH during
preparation. The N2 adsorption–desorption isotherms and the cor-
responding data for CuAl-8 and CuAl-12 are given in Fig. 5 and
Table 1. The isotherms are typical of type II with an H3 hysteresis
loop [15]. This type of hysteresis with relative pressures between
0.6 and 1 is characteristic of mesoporous materials containing
aggregates of platelike particles [15,16]. The pore diameters of the
two systems are indeed in the mesoporous range (Table 1), how-
ever, systems probably contain not only mesopores but also some
macropores.
It is also worth noting that the desorption rate for the CuAl-12
system is relatively low at its beginning compared to that of CuAl-
8 system. This observation is likely due to the different shape of
pores. This difference is confirmed by the texture of the bimetallic
sample with SSA of 106 m2/g for CuAl-8 against 54 m2/g for CuAl-
12. The same tendency is observed for the total pore volume which
decreases from 0.33 to 0.25 cm3/g. The formation of aggregates at
The XRD patterns of CuAl-8 and CuAl-12 after heat treatment
in N2 at 500 ◦C are shown in Fig. 3. CuO tenorite and Al2O3 were
detected in the two samples. The crystallite size of CuO in CuAl-8
and CuAl-12, determined using Scherrer equation, increased with
increasing the pH during preparation from 48 nm (CuAl-8) to 60 nm
(CuAl-12). Thus, it can be concluded that the interaction of Cu2+
with Al3+ species was weaker at high pH, leading to larger crystallite
sizes.
The Raman spectra of CuAl-8 and CuAl-12 catalysts were the
same whichever the observed particles (except the relative inten-
sity of lines that could be different), indicating that, in first
approximation, the same phases are present but with different
relative populations (Fig. 4). The Raman spectra of CuAl-8 and CuAl-
12 were similar to the spectrum of commercial CuO powder. It is
noticeable that all lines of CuAl-12 were shifted by 9 cm−1. This shift
is probably related to a different chemical environment of CuO in
CuAl-12
CuAl-8
CuO
100
300
500
700
900
1100
1300
Wavenumber (Cm-1)
Fig. 2. XRD patterns of CuAl-8 precursor at 25 and 575 ◦C. ♦: CuO tenorite; ◦:
Boehmite AlO(OH); •: Rouaite/gerhardtite Cu2NO3(OH)3; *: alumina Al2O3.
Fig. 4. Raman spectra of commercial CuO powder, of CuAl-8 and CuAl-12 catalysts.

210
N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
Table 1
Chemical composition, textural and structural properties of CuAl-8 and CuAl-12 after heat treatment.
a
Catalysts
SBET (m2/g)
ØBJH (nm)
Pore volume (cm3/g)
EDS
Cu/Al
Chemical composition from ICP-OES
Atomic ratio Cu/Al
CuAl-8
CuAl-12
106
54
10.9
14.5
0.33
0.25
0.36
0.40
1.01
0.84
a
Mean pore diameter determined from BJH desorption dV/dD pore volume.
ending at ca. 320 ◦C. The shoulder could account for the two-step
reduction (Cu2+ → Cu+ → Cu), with two maxima at ca. 243 and
267 ◦C, that is 16 ◦C higher than for CuAl-8, with a consumption
of H2 which should be the same for each step. Another explanation
could be the existence of a dual population of large and small sized
particles for CuAl-12, as seen by SEM.
250
Adsorption CuAl-8
Desorption CuAl-8
200
Adsorption CuAl-12
Desorption CuAl-12
150
Large CuO domains are known to be less readily reducible as the
reduction rate is controlled by the diffusion of hydrogen through
the particles [17,18]. It has been reported that Al2O3 as a support
promotes the reduction of finely dispersed CuO surface species
[19,20], and that the smaller the CuO particles, the faster their
reduction [21]. For copper oxide supported on CeO2, two reduction
peaks were reported by Stephanopoulos et al. [22,23]. The reduc-
tion of copper oxide clusters, strongly interacting with ceria, was
observed in the range of 125–175 ◦C, while larger CuO particles,
non-associated with ceria, were reduced at ∼200 ◦C.
100
50
0
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
Relative pressure (P/P0)
To get more information, the reducibility of the two systems
was examined by HT-XRD under H2 up to 575 ◦C. The series of
diffraction patterns obtained at various temperatures shows that
their structural properties under reducing atmosphere are similar
(Fig. 8). The copper oxide present at room temperature is stable up
to 200 ◦C. Above this temperature (250 ◦C), the line characteristic
of Cu2O appears at 2ꢀ = 35.9◦ and those of metallic Cu appear at
2ꢀ = 43.06◦, 50.11◦ and 73.4◦. The intensity of lines characteristic
of CuO decreases and disappears completely at 275 ◦C. This result
confirms that the reduction of CuO proceeds in two steps. At 250 ◦C
a part of CuO is reduced to Cu2O then to Cu, which is the only phase
present at 500 ◦C. It is worth noting that the lines of alumina at 2ꢀ
around 47 and 68◦ are present, as found at room temperature, and
Fig. 5. N2 adsorption–desorption isotherms of CuAl-8 and CuAl-12 after heat treat-
ment at 500 ◦C.
high basic pH could be responsible for the filling of the pores and a
decrease of SSA. This is in agreement with XRD results which show
the formation of large CuO crystallite for the CuAl-12 sample.
Another example of the influence of pH is the morphology of the
catalyst as studied by SEM. As shown in Fig. 6, smaller particles, with
uniform size distribution and relatively low crystallization degree
can be observed for CuAl-8, at variance with CuAl-12 where several
particles are agglomerated to form block-like particles.
3.4. Reducibility of catalysts by H2-TPR, H2-HT-XRD and
H2-HT-XPS
that they remain up to 500 ◦C. At 250 ◦C, all copper species (Cu2+
,
Cu+, Cu) are present with alumina (Fig. 9). The same results were
obtained with CuAl-8 (figure not shown).
The reduction behavior of CuAl-8 and CuAl-12 catalysts was
studied by H2-TPR, in situ HT-XRD and XPS at variable temperature
under H2 atmosphere.
H2-TPR profiles of CuAl-8 and CuAl-12 are different (Fig. 7).
The onset for CuAl-8 is at 200 ◦C and the reduction ends at 300 ◦C
after peaking at 251 ◦C. CuAl-12 exhibits a broader reduction peak
XPS spectra were carried out to get information on the composi-
tion and chemical states of the copper species on the surface. Fig. 10
shows the Cu-2p3/2 photopeak for the heat treated (in N2) and
reduced CuAl-8 and CuAl-12 samples, as well as for used catalysts
(after catalytic experiments).
Fig. 6. SEM images of (a) CuAl-8 and (b) CuAl-12 catalysts.

N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
211
being almost the same, it is difficult to distinguish one from each
other on the basis of Cu 2p level. The same features were observed
on examining the catalysts after reaction with benzaldehyde at
200 ◦C. As a partial conclusion, amounts of Cu2+ species remained
over the surface of all the reduced and used catalysts, together with
reduced copper species.
The atomic ratios and stoichiometry of CuAl-8 and CuAl-12 cat-
alysts obtained by in situ XPS analysis after heat treatment at 500 ◦C
(a), reduction by hydrogen at 100 ◦C (b), 150 ◦C (c), and 200 ◦C (d)
and after catalytic test at 200 ◦C (e) are shown in Table 2. The Cu/Al
atomic ratio on the surface of CuAl-8 and CuAl-12 is strongly lower
than the theoretical ratio whatever the conditions (0.05–0.18 and
0.07–0.09 for CuAl-8 and CuAl-12 respectively against 1.0). This
implies that a high fraction of CuO on the catalysts is masked
by Al2O3. On the other hand, the variation of Cu/Al ratio in the
case of CuAl-8 is more pronounced, particularly after benzaldehyde
reduction (0.05–0.18), compared to that of CuAl-12 (0.07–0.09). The
same tendency is observed with Al/O ratio. These results suggest
a remarkable stability of CuAl-12 system in reduction or reaction
conditions.
The Na/Al atomic ratio after heat treatment show that the sur-
face of CuAl-12 is more Na rich than CuAl-8 (0.13 vs. 0.004), result in
accordance with the higher pH value. After reduction by hydrogen
from 100 to 200 ◦C, Na/Al increases from 0.13 to 0.17 for CuAl-
12, while it decreases from 0.014 to 0.07 for CuAl-8. These results
indicate that the reducing pretreatment at different temperatures
promotes the migration of sodium species from the bulk to the sur-
face, maintaining the Cu/Al ratio constant, for CuAl-12 contrarily to
CuAl-8. We suggest that the Na enrichment of the surface results in
the increase of interaction between the different surface species,
leading to the stabilization of surface composition in reduction
conditions.
Fig. 7. H2-TPR profiles of CuAl-8 and CuAl-12 catalysts.
The Cu2p spectrum of the two N2-heat treated catalysts
contained a shake-up peak of weak intensity at 941–944 eV.
The binding energy of Cu 2p3/2 peak was centered at about
934.1–934.7 eV. These values are higher than in CuO (933.6 eV)
[24], but they are similar to that of CuAl2O4 spinel (ca. 934.6 eV)
[25,26]. Because no XRD line of CuAl2O4 was found, we suggest, in
agreement with the LRS study, that Cu–O–Al surface linkages are
formed, making the environment of Cu2+ species different from that
of CuO.
After H2 reducing pre-treatment at 150 and 200 ◦C before acqui-
sition of XPS spectra, the Cu2+ satellites progressively disappeared,
and Cu 2p3/2 peak shifted to 932.5 eV indicated the existence of Cu+
and Cu0 metal [27]. The binding energies for Cu0 and Cu+ species
Cu2+/Cutot, the atomic ratio after heat treatment at 500 ◦C and
reduction by hydrogen at 100 ◦C is equal to 1 for both systems,
indicating that the samples are not reduced at 100 ◦C. Cu2+/Cutot
decreases from 0.66 to 0.33 and from 0.85 to 0.31 for CuAl-8 and
CuAl-12 respectively, when the reduction temperature increases
from 150 to 200 ◦C, corresponding to the presence of more reduced
copper species (Cu+ and Cu0). These results also show that CuAl-8 is
more reducible than CuAl-12 at 150 ◦C. However, the two systems
have the same behavior with respect to their reduction by hydrogen
or benzaldehyde when the reaction takes place at 200 ◦C.
Thus these XPS results are in agreement with those obtained by
H2-TPR and H2-HT-XRD studies.
3.5. Catalytic activity
Fig. 8. XRD patterns of in situ reduction of CuAl-12 in H2/N2 at various temperatures.
The reaction of benzaldehyde reduction over CuAl-8 and CuAl-
12 was examined as a function of reaction temperature and nature
of pretreatment gas (N2 or H2). Benzyl alcohol and toluene were
the only reaction products observed.
◊: CuO
♦: Cu
♦
◊
*: Al2O3
•: Cu2O
◊
3.5.1. Benzaldehyde reaction under nitrogen atmosphere (N2)
Before reaction in N2 flow, the CuAl-8 and CuAl-12 samples were
pre-treated for 2 h at 300 ◦C in N2 flow. The results obtained show
that the two catalysts begin to be active from 150 ◦C (Fig. 11) and
that CuAl-12 is the most active. Their activity decreased with time
on flow and the conversion of benzaldehyde reaches zero at 150 ◦C.
At 200 ◦C, CuAl-12 catalyst remains stable after 4 h but at lower
conversion (∼12%).
♦
♦
•
◊
◊
◊
*
*
◊
CuAl-12
CuAl-8
The reaction product obtained is strongly sensitive to the
pH during catalyst preparation. In the presence of CuAl-12,
benzaldehyde is stoichiometrically reduced to benzyl alcohol at
150 and 200 ◦C. At variance, no trace of alcohol was detected with
CuAl-8 but benzaldehyde was converted exclusively to toluene.
10
20
30
40
50
60
70
80
Fig. 9. XRD patterns of CuAl-8 and CuAl-12 during reduction by H2/N2 at T = 250 ◦C.

212
N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
Fig. 10. XPS photopeaks of Cu 2p3/2 of: (i) CuAl-8 and (ii) CuAl-12 catalysts after (a) heat treatment at 500 ◦C, reduction by hydrogen at (b) 100 ◦C, (c) 150 ◦C, (d) 200 ◦C, and
after catalytic test at 200 ◦C (e).
The enhancement of the catalytic activity observed when
comparing catalysts prepared at increasing pH could be attributed
to many factors. First, several authors related the activity of
the catalyst to their acid–base properties. They found that an
increase of surface hydroxyl groups led to an enhancement of
its activity [6–8]. In addition, Tanabe et al. suggested that the
basic sites are surface lattice oxide ions (O2−) but that only their
presence is not sufficient to activate the C O bond [28,29]. In
this way, Saadi et al. [7] reported that, besides the surface O2−
species, the Lewis acid site of metal ions would have an important
role on the activation of carbonyl group. In our case a better
activity for the system prepared in the most basic conditions
is confirmed with benzyl alcohol as main product for CuAl-12
and can be explained by the surface mobility of the hydroxyl
groups [6–8]. Moreover, the consumption and progressive dis-
appearance of these OH groups at the surface may explain the
decrease in reactivity with time on stream observed on both
systems.
It is admitted that in absence of hydrogen, the formation
of benzyl alcohol proceeds via the Cannizzaro reaction that
Table 2
Atomic ratios and stoichiometry of CuAl-8 and CuAl-12 catalysts obtained by in situ XPS analysis after: (a) heat treatment at 500 ◦C, reduction by hydrogen at (b) 100 ◦C, (c)
150 ◦C, (d) 200 ◦C and (e) after catalytic test at 200 ◦C.
Catalyst
Cu/Al
Na/Al
Al/O
Cu2+/Cutot
Surface Stoichiometrya
CuAl-8
a
b
0.08
0.07
0.004
0.014
0.56
0.54
1
1
Cu2+0.08Al3+O1.58
Cu2+0.07Al3+O1.57
c
0.06
0.010
0.55
0.66
Cu2+0.04Cu1+0.02Al3+O1.55
d
e
0.05
0.18
0.007
0.013
0.54
0.47
0.33
0.77
Cu2+0.016Cu00.033Al3+O1.51 + Cu2+0.016Cu+10.033Al3+O1.53
Cu2+0.138Cu1+0.041Al3+O1.65
CuAl-12
a
b
c
d
e
0.09
0.08
0.08
0.07
0.07
0.13
0.13
0.15
0.17
0.05
0.54
0.54
0.54
0.55
0.55
1
1
0.85
0.31
0.75
Cu2+0.09Al3+O1.59
Cu2+0.08Al3+O1.58
Cu2+0.068Cu1+0.012Al3+O1.57
Cu2+0.021Cu00.048Al3+O1.52 + Cu2+0.021Cu+10.048Al3+O1.54
Cu2+0.052Cu1+0.017Al3+O1.56
a
Sodium omitted.

N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
213
Fig. 11. Benzaldehyde conversion vs. time on stream on CuAl-8 and CuAl-12 under a nitrogen atmosphere, pre-treatment: N2/300 ◦C.
Table 3
stoichiometrically consumes surface OHsurf groups (Eq. (1)), [6]. The
Catalytic results for benzaldehyde hydrogenation at steady state over CuAl-8 and
CuAl-12 catalysts.
mechanism can be written as follows
PhCHO
g
Treaction (◦C)
Conv. (%)
Pre-treatment/N2/300 ◦C,
reaction/PhCHO/H2/Treaction ◦C,
selectivity %
PhCHOg + OHsurf → PhCH(OH)Oads −→ PhCOOads + PhCH2OHg
(1)
Catalyst
Benzyl alcohol
Toluene
The toluene may come from the benzaldehyde dismutation
to benzoate and benzylate species on the basic surface (Eq. (2)),
followed by the reaction between benzylate species with benzal-
dehyde molecule (Eq. (3)), as suggested by Sreekumar and Pillai
reaction [30].
CuAl-8
150
200
3
54
0
0
100
100
CuAl-12
150
200
22
78
100
96
0
4
2Ph-CHOg → Ph-COOads + Ph-CH2Oads
Ph-CH2O Ph−-C→HOgPh-CH3 + Ph-COO
(2)
(3)
trace of benzyl alcohol was observed with CuAl-8 which converted
benzaldehyde exclusively to toluene.
ads
ads
The comparison of the obtained results with the two reaction
processes (under N2 or H2) reveals that, as expected, dihydrogen
enhances considerably the catalytic activity and the stability of CuAl
system as compared to N2. The heterolytic and homolytic disso-
ciation of H2 molecule on the surface of metallic oxide can lead
to the creation of active hydrogen surface species as intermedi-
ates for direct activation of benzaldehyde molecule. Several authors
[6–8,30] suggested that the chemical process of benzyl alcohol for-
mation is a 1–2 nucleophilic addition, with a high polarization of the
transition state, as in the case of the hydrogenation of benzaldehyde
[6].
The formation of toluene with 100% selectivity on CuAl-8, sug-
gests that the benzylate species do not desorb and probably react
directly by a new surface step to toluene [6,30]. A high pore
volume with a small pore diameter (BJH method) observed for
CuAl-8 compared to CuAl-12 could delay the desorption of the
benzylate species and thereby inhibit the formation of benzyl
alcohol.
3.5.2. Benzaldehyde reduction under dihydrogen
Before testing under H2 flow, the oxidic catalysts were in situ
activated for 2 h at 300 ◦C in N2 flow. The contribution of
(4)
Whereas, two pathways are generally proposed in the litera-
ture for the toluene formation. In the first case, toluene is formed
by a consecutive reaction, the reaction (4) followed by (5) (benzyl
alcohol hydrogenolysis) [6–8,31]. On metallic oxide, the active site
should be bifunctional where the metal cation and hydride entities
interact with the oxygen and carbon atoms of the alcohol molecule,
respectively.
gas-phase initiated reactions was tested by conducting experi-
ments using an empty-volume reactor. Benzaldehyde conversion
was null, confirming that gas-phase reactions were negligible
under our experimental conditions.
The conversion of benzaldehyde at 150 and 200 ◦C is plotted
against the time on-stream in Fig. 12. The reaction products (ben-
zyl alcohol and toluene) were the same than for the reaction in
nitrogen. Again, CuAl-12 was far more active than CuAl-8 (22% of
conversion vs. 3% and 78% vs. 54% at 150 and 200 ◦C respectively)
and its stability also higher whatever the temperature. The CuAl-12
system led to the formation of benzyl alcohol with 100% selectivity
at 150 ◦C, and 96% at 200 ◦C (4% of toluene) (Table 3). In contrast, no
(5)

214
N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
Fig. 12. Benzaldehyde conversion vs. time on stream on CuAl-8 and CuAl-12 under a hydrogen atmosphere, pre-treatment: N2/300 ◦C.
The benzyl alcohol dismutation to toluene, benzaldehyde and
water could be a second pathway for the toluene formation (Eq.
(6)) [32–34]:
(due, for example, to the presence of CuAl2O4), the Cu disper-
sion is improved. This phase, no detected by any analysis method,
could be present as microdomains. Moreover, XPS showed the
(6)
enrichment of the CuAl-12 surface by sodium species after reduc-
tion and reaction, whereas the Cu/Al ratio remains unchanged.
Sodium species could be directly or indirectly responsible for the
high selectivity of benzyl alcohol (100% at 150 ◦C and 96% at 200 ◦C)
observed on CuAl-12. Indeed, sodium species could have a possible
“site isolation” effect on copper species that probably increase the
desorption rate of benzyl alcohol.
The reducibility of CuAl-12 and CuAl-8 studied by H2-TPR could
be correlated to their catalytic behavior as far as copper species
would be the active sites. The dual population of large and small
particles observed for the system prepared at high pH (CuAl-12)
could favor both the activity and the benzyl alcohol selectivity. The
small particles, easily reduced, could favor the activity, while the
CuO large particles in weaker interaction with other components
(Al2O3) could favor more readily the C O hydrogenation of benz-
aldehyde, resulting in the high alcohol selectivity. In contrast, the
medium sized particles observed for the system CuAl-8, that are
in relatively high interaction with Al2O3, could convert the benzyl
alcohol before its desorption, resulting in high toluene selectivity.
In the case of CuAl-8 oxide, the absence of benzyl alcohol in
the reaction mixture could be explained by the fact that benzyl
alcohol is more strongly absorbed on the catalyst surface, and may
be subsequently chemically converted before its desorption.
The divergence of the catalytic behavior of CuAl-8 and CuAl-12
may be accounted for by their acid–base and oxidative characters,
as well as by the dispersion and environment of copper, and the
microstructure which are different.
The role of surface acid–base properties of catalysts in the
benzaldehyde hydrogenation is proved. Many authors agree that
the adsorption and activation of unsaturated carbonyl molecule
are related to acid–base interactions (reactant-catalyst), and that
the adsorption and desorption rates of products are related to
acid–base properties of catalyst. The choice of the basic pH value
(8 and 12) in the catalyst preparation and the existence of Lewis
acid/base sites (Cu–O) would explain the strong reactivity of the
two systems (CuAl-8 and CuAl-12) toward the highly polarized C
bond of benzaldehyde molecule.
O
Moreover, it is known that the active site of copper in the
hydrogenation reactions is the association of Cu(I) in an octahedral
environment with hydride ion [35–37]. The in situ XPS analy-
sis showed that amounts of Cu2+ associated with Cu+ and/or Cu0
species were still present over the surface of reduced samples and
of catalysts after reaction. On the one hand, the existence of such
Cu(I) center would explain the strong reactivity of the two systems
[6–8,30] and on the other hand, the coexistence of different states of
copper on the catalysts surface (confirmed by HT-XPS and HT-XRD)
would have a beneficial electronic effect necessary for the dissoci-
ation of hydrogen molecule and for the activation of benzaldehyde
molecule.
The catalytic activity enhancement of the CuAl-12 as compared
to that of CuAl-8 could be due to a different environment of Cu.
Indeed, both LRS and XPS analyses showed that the environment of
CuO observed on CuAl-12 is different from those of CuAl-8 and com-
mercial CuO. This result could suggest the formation of Cu–O–Al
linkage on the CuAl-12 sample. Several authors [19,38] hypoth-
esized that when direct Cu–O–Al surface linkages are formed
4. Conclusion
Cu–Al–O bimetallic oxide prepared by hydrothermal method
exhibits interesting catalytic properties in the benzaldehyde reduc-
tion. The basic pH value during preparation has a strong influence
on the system activity. In reaction conditions, the coexistence of
Lewis acid site (metal ions of CuO) in different states (Cu2+, Cu+/Cu0)
would have a beneficial electronic effect necessary for the activa-
tion of the highly polarized C O double bond of benzaldehyde. Both
catalytic activity and benzyl alcohol selectivity are improved when
the pH value is increased from pH = 8 to pH = 12. The increase of pH
value would favor Al(OH)3 (pH ca. 8) that may precipitate while at
pH 12 soluble Al(OH)4− species could be present. The interaction of
these respective species with copper during drying and heat treat-
ment could be responsible for modifications of the microstructures.
When pH = 12, the Cu2+ reducibility is enhanced (TPR) and due to
dispersion of Cu species at the catalyst surface and the Cu–O–Al

N. Haddad et al. / Journal of Molecular Catalysis A: Chemical 396 (2015) 207–215
215
linkage is formed (XPS, LRS). Moreover, a remarkable enrichment of
the surface by sodium in reaction conditions is obtained (HT-XPS),
which may help to “isolate” clusters of Cu oxo species, resulting to
an increase of the desorption rate of product and thus an increase
of selectivity to benzyl alcohol.
[14] I.M.J. Vilella, S.R. De Miguel, O.A. Scelza, J. Mol. Catal. A: Chem. 284 (2008)
161–171.
[15] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309–319.
[16] T.A. Zepeda, B. Pawelec, J.L.G. Fierro, A. Montesinos, A. Olivas, S. Fuentes, T.
Halachev, Micropor. Mesopor. Mater. 111 (2008) 493–506.
[17] A. Jones, B.D. McNicol, Temperature Programmed-Reduction for Solid Materials
Characterization, Dekker, New York, 1986.
[18] M. Hartmann, S. Racouchot, C. Bischof, Micropor. Mesopor. Mater. 27 (1999)
309–320.
[19] S. Valange, J. Barrault, A. Derouault, Z. Gabelica, Micropor. Mesopor. Mater.
44–45 (2001) 211–220.
[20] H. Jiang, H. Bongard, W. Schmidt, F. Schüth, Micropor. Mesopor. Mater. 164
(2012) 3–8.
[21] S. Zhang, W. Huang, X. Qiu, B. Li, X. Zheng, S. Wu, Catal. Lett. 80 (2002) 41–46.
[22] W. Liu, M.F. Stephanopoulos, Chem. Eng. J. 64 (1996) 283–294.
[23] L. Kundakovic, M. Stephanopoulos, Appl. Catal. A: Gen. 171 (1998) 13–29.
[24] L. Wang, W. Ding, Y. Liu, W. Fang, Y. Yang, J. Nat. Gas Chem. 19 (2010) 487–492.
[25] G. Petrini, F. Garbassi, J. Catal. 90 (1984) 113–118.
Acknowledgements
L. Burylo, N. Djellal, M. Frère and O. Gardoll (UCCS) are acknowl-
edged for discussions and XRD, ATG, XPS and TPR experiments,
respectively.
References
[26] K.T. Jacob, C.B. Alcock, J. Am. Ceram. Soc. 58 (1975) 192–195.
[27] G.C. Chinchen, K.C. Waugh, D.A. Whan, Appl. Catal. 25 (1986) 101–107.
[28] T. Iizura, H. Hattori, Y. Ohno, J. Sohma, K. Tanabe, J. Catal. 22 (1971) 130–139.
[29] H. Hattori, N. Yoshii, K. Tanabe, Proc. 5th Int. Congr. Catal. 10 (1973) 233–239.
[30] R. Sreekumar, C.N. Pillai, Catal. Lett. 19 (1993) 281–291.
[31] M. Herskowitz, Heterogeneous Catalysis and Fine Chemicals, Elsevier, Amster-
dam, 1991, pp. 105–112.
[1] J. Scognamiglio, L. Jones, D. Vitale, C.S. Letizia, A.M. Api, Food Chem. Toxicol. 50
(2012) S140–S160.
[2] Y. Zhou, J. Liu, X. Li, X. Pan, X. Bao, J. Nat. Gas Chem. 21 (2012) 241–245.
[3] M. Li, X. Wang, N. Perret, M.A. Keane, Catal. Commun. 46 (2014) 187–191.
[4] M.A. Vannice, D. Poondi, J. Catal. 169 (1997) 166–175.
[5] M.A. Keane, J. Mol. Catal. A: Chem. 118 (1997) 261–269.
[6] D. Haffad, U. Kameswari, M.M. Bettahar, A. Chambellan, J.C. Lavalley, J. Catal.
172 (1997) 85–92.
[32] E.A.I. Heiba, P.S. Landis, J. Catal. 3 (1964) 471–476.
[33] M. Jayamani, C.N. Pillai, J. Catal. 82 (1983) 485–488.
[34] M. Jayamani, N. Murugesan, C.N. Pillai, J. Catal. 85 (1984) 527–529.
[35] R. Bechara, G. Wrobel, M. Daage, J.P. Bonnelle, Appl. Catal. A: Gen. 16 (1985)
15–27.
[36] S. Siegel, J. Catal. 30 (1973) 139–145.
[37] R. Hubaut, M. Daage, J.P. Bonnelle, Appl. Catal. A: Gen. 22 (1986) 231–241.
[38] B.K. Kwak, D.S. Park, Y.S. Yun, J. YI., Catal. Commun. 24 (2012) 90–95.
[7] H. Zazoua, A. Saadi, K. Bachari, D. Halliche, C. Rabia, Res. Chem. Intermed. (2013)
931–946.
[8] A. Saadi, Z. Rassoul, M.M. Bettahar, J. Mol. Catal. A: Chem. 258 (2006) 59–67.
[9] J. Jenck, J.E. Germain, J. Catal. 65 (1980) 141–149.
[10] R. Hubault, J.P. Bonnelle, J. Mol. Catal. 55 (1989) 170–183.
[11] A. Saadi, Z. Rassoul, M.M. Bettahar, J. Mol. Catal. A: Chem. 164 (2000) 205–216.
[12] S. Wang, H. Liu, Catal. Lett. 117 (2007) 62–67.
[13] M. Kurtz, N. Bauer, C. Büscher, H. Wilmer, O. Hinrichsen, R. Becker, S. Rabe, K.
Merz, M. Driess, R.A. Fischer, M. Muhler, Catal. Lett. 92 (2004) 49–52.