Reduction of benzaldehyde on alkaline earth metal oxides

Article abstract of DOI:10.1016/j.molcata.2006.05.029

Benzaldehyde reduction was studied over alkaline earth metal oxide catalysts and carried out in a nitrogen or a dihydrogen atmosphere at the 100-350 °C of the range reaction temperature. The results show that benzyl alcohol is produced by the Cannizzaro reaction under nitrogen flow on MgO oxide and by direct hydrogenation under dihydrogen on irreducible oxides such as CaO, BaO and SrO. The Cannizzaro reaction involves the basic hydroxyl groups and is followed by the reduction of benzoate species by H₂ to benzaldehyde. Toluene and benzene were produced at high reaction temperature under N₂ flow over alkaline earth oxides from a reaction involving benzylate species. Under H₂ flow, toluene and benzene were also formed by means of consecutive reduction reactions. The present work deals with the acid-base properties of alkaline earth metal oxide catalysts (MgO, CaO, BaO, SrO) and benzaldehyde reduction was used as probe reaction for characterizing the surface properties.

Full text of DOI:10.1016/j.molcata.2006.05.029

Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
Reduction of benzaldehyde on alkaline earth metal oxides
a,∗
a
b,1
A. Saadi , Z. Rassoul , M.M. Bettahar
a
Laboratoire de Chimie du Gaz Naturel, Facult e´ de Chimie, USTHB, BP32 El-Alia, 16111 Bab-Ezzouar, Algeria
b
Laboratoire de Catalyse H e´ t e´ rog e` ne, UMR-CNRS 7565, Universit e´ Henri Poincar e´ , Nancy-I, Facult e´ des Sciences,
BP 239, 54506, Vandoeuvre-l e` s-Nancy-Cedex, France
Received 1 April 2006; received in revised form 7 May 2006; accepted 11 May 2006
Available online 16 June 2006
Abstract
Benzaldehyde reduction was studied over alkaline earth metal oxide catalysts and carried out in a nitrogen or a dihydrogen atmosphere at the
◦
1
00–350 C of the range reaction temperature. The results show that benzyl alcohol is produced by the Cannizzaro reaction under nitrogen flow on
MgO oxide and by direct hydrogenation under dihydrogen on irreducible oxides such as CaO, BaO and SrO. The Cannizzaro reaction involves the
basic hydroxyl groups and is followed by the reduction of benzoate species by H to benzaldehyde. Toluene and benzene were produced at high
reaction temperature under N flow over alkaline earth oxides from a reaction involving benzylate species. Under H flow, toluene and benzene
were also formed by means of consecutive reduction reactions.
2
2
2
The present work deals with the acid–base properties of alkaline earth metal oxide catalysts (MgO, CaO, BaO, SrO) and benzaldehyde reduction
was used as probe reaction for characterizing the surface properties.
©
2006 Elsevier B.V. All rights reserved.
Keywords: Reduction; Alkaline earth metal oxide; Benzaldehyde; Acid–base properties; Benzyl alcohol; Toluene; Benzene
1
. Introduction
vokes the direct hydrogenation of benzaldehyde. In this case, the
presence of surface adsorbed hydrogen over alkali earth metal
oxides has been proved [16]. The experimental evidence that
In the past few years, various reduction reactions in the gas
or the liquid phases have been studied to synthesize valuable
intermediate products or fine chemicals [1–5]. Benzaldehyde
reduction is one of these reactions which can provide infor-
mation about nature and properties of surface centers [1–13].
A large number of papers have been published concerning the
transformationof benzaldehyde on oxide catalysts [2,4,6,11,14].
Alkaline earth metal oxides, which were reported to have basic
properties on the surface [14], were used as catalysts.
It is well known that benzaldehyde is reduced to benzyl alco-
hol or toluene on metal oxides by two main process [2,6,15]:
the first one seems to be the formation of the reduced com-
pounds, at the absence of dihydrogen flow, by the Cannizzaro
reaction accompanied by that of oxygenated products such as
benzoate species and the active sites are mobile hydroxyl groups
the low-coordination sites are responsible for H dissociation
2
has also been confirmed by theoretical investigations [17]. The
presence of a totally dehydration surface is therefore necessary
for both heterolytic dissociative hydrogen adsorption [16]. As
to metal oxides as catalysts, one of us have shown that their cat-
alytic properties in the hydrogenation of benzaldehyde to benzyl
alcohol depends on their reducibility and, also, on their surface
acid–baseproperties[2]. Tolueneandbenzeneby-productsarose
from benzyl alcohol and benzaldehyde hydrogenolysis, respec-
tively [1,2,7,11,18,19].
Adsorption studies of benzaldehyde over several metal
oxides [20] strengthen the hypothesis of the contribution of the
acid–base properties of the SiO or Al O supports. Indeed, an
2
2
3
IRTF study [20] showed that the adsorption of benzaldehyde on
silica occurs through a hydrogen bond of the oxygen atom of the
C O group to the surface silanol groups whereas for the rather
basic alumina it involves a nucleophilic attack at the carbonyl
center by surface oxygen (or hydroxyl), producing surface ben-
zoate species [21]. Moreover, stable surface oxygenated inter-
mediates have also been detected in the hydrogenation of CO
[
2]. In the second process, the presence of dihydrogen flow pro-
∗
Corresponding author. Fax: +213 21 24 73 11.
E-mail addresses: adel saadi@yahoo.fr (A. Saadi),
bettahar@lcah.u-nancy.fr (M.M. Bettahar).
1
Fax: +33 3 83 912181.
1
381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.05.029

6
0
A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
on nickel-based catalysts [22]. They were identified as methoxy
species on the alumina support but not found on the silica support
Before testing, the oxides were in situ pre-treated for 2 h at
◦
350 C in a current of dihydrogen (H2) or in a current of nitrogen
3
−1
[
23–26].
Benzaldehyde can be used as, model reaction by the charac-
(N2) with a flow rate of 20 cm min
.
Benzaldehyde transformation was performed in dihydrogen
atmosphere or in nitrogen atmosphere at atmospheric pressure
terization of metal oxide surfaces [2,6]. Therefore, comparison
of benzaldehyde conversion on metal oxide catalysts, in the
presence or absence of dihydrogen atmosphere would allow the
classification of the metal oxides according to their hydroxyl
mobility (basicity).
◦
and in the whole range of reaction temperature 100–350 C.
Gaseous benzaldehyde (3.2 Torr) was obtained by bubbling N2
(250 Torr) in liquid benzaldehyde maintained at constant tem-
◦
perature (50 C) in a suitable saturator.
The present paper reports the correlation between the cat-
alytic activity and the surface property, the nature of active sites
and the reaction mechanism.
The gaseous reactant and products were heated up-stream and
outstream in order to avoid their condensation and analyzed on
line by a FID gas chromatograph (Delsi IGC 121 ML) equipped
with a CP-Sil8CB/Chromosorb WAW column.
2
. Experimental
3. Results
2
.1. Catalysts preparation, activation and characterization
The metal oxides used are reported in Table 1. CaO, BaO
3.1. BET and XRD characterizations
and SrO oxides were prepared by calcinating the carbonate
precursors (MCO3: M = Ca, Ba, Sr) at 1200 C for 3 h in air,
whereas MgO oxide was prepared from the hydroxide precur-
sor (Mg(OH)2) by thermal decomposition (Table 1).
The characteristics of alkaline earth metal oxides are reported
◦
◦
in Table 1. The calcined simple at 1200 C showed a very small
2
−1
surface areas in the range 7.0–15.0 m g and the pore vol-
ume (VP) decreased with the decrease of the specific surface
area. These results indicate that high thermal treatment process
strongly affected the microstructure of the alkaline earth metal
oxides.
The specific area measurements and average pore radius were
performed by the classical BET method on a Coultronics 2100E
apparatus. The data were interpreted using the BET equation and
˚
an effective cross-sectional area of 16.2 A for N2. The standard
On the other hand, the XRD spectra showed the characteris-
tic intense bands of the different oxides and, the characteristic
weak bands of the corresponding carbonates (CaCO3, SrCO3,
BaCO3)orhydroxide(Mg(OH)2)phases. Thepatternwasthatof
crystallized materials with well-defined bands and high degree
of crystallinity of the diffraction lines. The metal phase was of
‘CFC’ structure for all the oxides. Calculations showed that the
mean crystallite size of metal oxide was not the same, and vary
in the range of 60–150 nm (Table 1).
pre-treatment consisted of heating the sample under dynamic
◦
vacuum at 300 C for 3 h in order to remove adsorbed water
and other impurities. The measurement was then made at liquid
nitrogen temperature with nitrogen as adsorbate.
Phase analysis and crystallite size of the samples were per-
formed by X-ray powder diffraction employing a PW1710
Philips diffractometer. Patterns were using Ni-filtered copper
◦
˚
radiation (λ = 1.5406 A) at a scanning rate of 2 at 2θ/min.
The average size of the metal particle was calculated from the
linewidth at half height of the MO (M = Mg, Ca, Sr, Ba) peaks
using the Scherrer equation with Warren’s correction for instru-
mental line broadening [27].
3.2. Catalytic activity
3.2.1. Benzaldehyde reduction under nitrogen atmosphere
(N2)
2
.2. Catalyst testing
Before reaction under N2 flow, all oxides were pre-treated for
◦
h at 350 C in a current of the same gas (N2). Fig. 1 reports
2
Benzaldehyde (Aldrich, 99.88%) was degassed under nitro-
genpurgebeforeuse. H2 andN2 (Airliquid)werepassedthrough
Supelco molecular sieve traps oxytraps (Altech).
activity results with time on stream in the whole range of reac-
tion temperature. It can be noted in this figure that benzaldehyde
reactivity decreased with time on flow at all reaction tempera-
◦
The catalytic performances were carried out in the same reac-
tor with 0.2 g samples at atmospheric pressure and a total flow
tures, generally tending to zero (TR < 250 C) or to a low residual
◦
value (TR > 250 C). In this case, benzaldehyde was reduced
3
−1
.
rate of 50 cm min
stoichiometrically in the range of temperature to benzyl alco-
Table 1
Characteristics of alkaline earth metal oxide catalysts
◦
2
2
−1
3
−1
a
Oxide
Origin
SBET (m g
)
VP (cm g
)
XRD species identification
Crystallite size (nm)
E
(Volt)
→0
MgO
CaO
SrO
Fluka
Fluka
Fluka
Fluka
15.0
11.0
10.0
7.0
0.073
0.045
0.039
0.032
MgO, Mg(OH)2
CaO, CaCO3
SrO, SrCO3
65
77
108
139
−1.86
−2.26
−2.59
−1.59
BaO
BaO, BaCO3
a
Determined after calcination step by XR diffraction using Debby-Scherrer equation.

A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
61
◦
Fig. 1. Benzaldehyde conversion vs. time on stream on alkaline earth metal oxides under a nitrogen atmosphere. Range of reaction temperature 100–350 C.
Pre-treatment: N2/350 C.
◦
◦
hol, toluene and benzene. At low reaction temperature (100 C)
and after 10 min on nitrogen flow, all oxides were inactive with
exception of MgO (6.2%).
• Benzyl alcohol was appeared only on the MgO oxide in the
◦
whole range of reaction temperature 100–350 C (Table 2).
The results were showed that benzyl alcohol decreased pro-
gressively as a function of the reaction temperature for to
Generally, the conversion of benzaldehyde after 10 min
of reaction shows that the reactivity increases as function
of reaction temperature on the following order (Fig. 2):
MgO < CaO < SrO < BaO. On the other hand, and at given reac-
tion temperature, the inspection of the Fig. 2 shows that the
benzaldehyde conversion increased with the basicity character
of metal oxide and the increase being more pronounced in SrO
◦
reach at 350 C a value of 92.6%. In the whole range of
◦
the reaction temperature 100–350 C, benzyl alcohol was
Table 2
Catalytic results for benzaldehyde reduction under nitrogen flow after 10 min of
stream over alkali earth metal oxides
◦
◦
and BaO oxides at 250 C and 350 C. A gap of activity was
◦
Oxide
Treaction
( C)
Conv
(%)
Pre-treatment/N2/350 C,
◦
◦
◦
a
◦
observed between 250 C and 350 C in the case of CaO, SrO
and BaO oxides; a conversion of which were multiplied by 1.4
reaction/PhCHO/N /T
C,
2
reaction
selectivity %
(Table 2).
Benzyl alcohol
Toluene
Benzene
The reaction products were benzyl alcohol, toluene and ben-
zene. No aromatic ring hydrogenation was observed:
1
1
2
3
00
50
50
50
6.2
8.7
10.3
11.5
100
100
95.9
92.6
0
0
0
0
0
0
4.1
7.4
MgO
CaO
SrO
1
1
2
3
00
50
50
50
0
9.4
12.6
17.5
0
0
0
0
0
0
28.1
19.5
6.2
71.9
80.5
93.8
1
1
2
3
00
50
50
50
0
0
0
0
0
0
0
0
0
0
100
100
100
10.4
20.8
25.4
1
1
2
3
00
50
50
50
0
0
0
0
0
0
0
0
0
0
100
100
100
11.4
21.3
27.9
BaO
◦
Fig. 2. Benzaldehyde conversion vs. reaction temperature on metal oxides in a
nitrogen atmosphere after 10 min under flow.
Pre-treatment: 350 C/N2.
a
After 10 min of reaction.

6
2
A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
gressively with time on stream after introduction of dihydrogen
over CaO, SrO and BaO, whereas, MgO oxide remained com-
pletely inactive.
After 8 h of the beginning of the reaction, CaO, SrO and BaO
oxides exhibited various activities. In relation to the reaction
under nitrogen atmosphere, the influence of dihydrogen atmo-
sphere depended on metal oxide used (Fig. 3). Comparatively
to the reaction under nitrogen flow, the results show that the
magnesia activity was not re-established and conversely, the
benzaldehyde conversion was relatively lowest on CaO, SrO
and BaO. No change was observed in the selectivities.
Fig. 3. Influence of the nitrogen replacement by dihydrogen as gas carrier on
benzaldehyde conversion at 350 C.
◦
3.2.2. Benzaldehyde reduction under dihydrogen
atmosphere (H2)
Before testing under a dihydrogen flow, the oxides were in
situactivatedfor2 hat350 CinacurrentofH2. Fig. 4showsthat
obtained as a result of the Cannizzaro reaction only for the
MgO oxide.
◦
benzaldehyde was reduced catalytically in the whole range of
reaction temperature. The alkaline earth metal oxides exhibited
initial activation before the steady state was established and the
duration (1–3 h) of activation depended on the nature of metal
oxide and reaction temperature.
•
•
Toluene was obtained only on the CaO oxide and the same
feature was observed, since the toluene production decreased
◦
◦
from 28.1% (at 150 C) to 6.2% (at 350 C).
In contrast, benzene was produced with all the oxides and the
yield was higher to raised temperature.
For all oxides, the inspection of Fig. 4 shows the increase of
the steady state activity with the reaction temperature. The level
and order of activities depended on both reaction temperature
and nature of oxide. The activity results with time of stream
showed that MgO oxide was remained inactive in benzaldehyde
hydrogenation in all temperature reaction. The SrO and BaO
After the benzaldehyde reaction under a current of nitro-
gen, the latter was replaced by dihydrogen flow. Previous work
showed some influence of the gas reaction on benzaldehyde
reduction on the alkaline earth metal oxides. Thus, after a total
disappearance of catalytic activity under nitrogen atmosphere,
Fig. 3 shows that the benzaldehyde conversion increased pro-
◦
oxide were inactive at 100 C and the activity increased with
the increasing of reaction temperature. On the other hand, the
◦
Fig. 4. Benzaldehyde conversion vs. time on stream on alkaline earth metal oxides under a dihydrogen atmosphere. Range of reaction temperature 100–350 C.
Pre-treatment: H2/350 C.
◦

A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
63
Fig. 5. Benzaldehyde conversion vs. reaction temperature on metal oxides in a
dihydrogen atmosphere after 10 min under flow.
Fig. 6. Influence of the dihydrogen replacement by nitrogen as gas carrier on
benzaldehyde conversion at 350 C.
◦
◦
maximum of conversion was obtained at 350 C for CaO oxide
14.3%), whereas it reached lower values (<11%) for SrO and
BaO oxides. Strikingly, a similar gap of activity was observed
(
except SrO who produced the toluene with high amount at
◦
only150 Creactiontemperature(Table3). Ontheotherhand,
◦
◦
for the CaO, SrO and BaO oxides between 150 C and 250 C:
the activity was multiplied by 2.4 for CaO, 2.5 for SrO and 2.7
for BaO (Table 3, Fig. 5).
the increase of reaction temperature leads to the decrease of
toluene selectivity.
•
High selectivity to benzene was observed for the SrO (100% at
3
Under dihydrogen atmosphere, the reaction products
obtained are the same as those obtained for the transformation
under nitrogen flow: benzyl alcohol, toluene, benzene and traces
of condensed products (not analyzed). Of course, the selectivi-
ties are different:
◦ ◦ ◦
50 C), CaO (99.7% at 350 C) and BaO (98.9% at 350 C).
Incontrasttobenzylalcoholandtoluene, thebenzeneselectiv-
ity increased with the reaction temperature and the maximum
of selectivity (100%) was obtained at a high reaction temper-
ature (Table 3). On SrO oxide, only benzene formation was
◦
observed in the range 250–350 C.
◦
◦
•
Benzyl alcohol was produced between 100 C and 150 C
◦
on CaO oxide and only at 150 C on SrO and BaO oxides
After 4 h of reaction under dihydrogen flow, the same gas
(
Table 3). The elevation of reaction temperature is not favor-
was substituted by the nitrogen gas. Fig. 6 shows that benzalde-
hyde conversion decreased with time on stream and after 8 h of
reaction, all oxides became totally inactive.
◦
able to benzyl alcohol formation. Indeed, after 150 C this
product was totally disappeared.
•
In the range of reaction temperature, low selectivity values to
toluene (<11%) was observed for the CaO and BaO oxides
4. Discussion
Table 3
Catalytic results for benzaldehyde hydrogenation under H2 flow after 10 min of
stream over alkali earth metal oxides
4.1. Benzaldehyde reduction under nitrogen atmosphere
(N2)
◦
Oxide
Treaction
Conv
(%)
Pre-treatment/H2/350 C,
◦
a
◦
(
C)
reaction/PhCHO/H2/Treaction C,
4.1.1. Activity
selectivity %
In the discussion of the results of benzaldehyde reduction
under nitrogen flow, we have to bear in mind the nature of active
sites and organic molecules. Let us then discuss in more detail
the obtained results in this study. We first consider the results
Benzyl alcohol
Toluene
Benzene
1
1
2
3
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
50
50
50
MgO
CaO
SrO
obtainedinthereductionofbenzaldehydeunderN flowbecause
2
they give useful information on acid–base properties of the cat-
alysts.
100
150
250
350
2.8
4.2
10.4
14.3
100
100
0
0
0
10.7
0.7
0
0
89.3
99.3
The correlation of the basicity of the alkaline earth metal
oxides with the activity and the increase in activity with atomic
number seem to indicate that the active sites are mainly basic.
In the whole range of the reaction temperature, the observed
activities: MgO < CaO < SrO < BaO mean that this order is in
good agreement with the bulk basicity in the periodic table.
It was reported by several workers that the basic site is a sur-
0
100
150
250
350
0
0
100
0
0
0
9.4
1.3
0
0
90.6
98.7
3.1
7.8
10.7
0
100
150
250
350
0
0
56.7
0
0
43.3
0
0
0
100
100
face lattice oxygen O2 [28–30] but the only presence of basic
sites was not sufficient for activated the C O bond. Indeed,
Tanabe and Saito [14] were showed that the use of the solids
−
2.1
5.7
8.3
BaO
0
0
◦
2 2 2
NaOH/SiO , SrO and BaO , which exhibited only basic prop-
erty with phenolphthalein, does not show any catalytic activity
Pre-treatment: 350 C/H2.
a
After 10 min of reaction.

6
4
A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
and the catalytic activity of MgO and CaO was lost by the addi-
tion of pyridine. These results were suggested that the Lewis
acid site (metal ions of MgO and CaO) is considered to be also
important as an active site. The active species formed by the
reaction of benzaldehyde with basic and acidic sites were ben-
zylate and benzoate species and their existence on MgO and
CaO oxides was confirmed by infrared study [14,31–33]. The
raise of the reaction temperature increased the intensity bands
of surface benzoates and decreased the intensity bands of sur-
face benzylates, suggesting that the benzylate species change to
benzoate species [14,33].
In this process, the obtained results reflect the surface mobil-
ity (or nucleophilicity) of the hydroxyl groups on MgO oxide
[2,6,19] and the formation of one benzyl alcohol molecule con-
sumes one surface hydroxyl group, thus explaining the decrease
in reactivity with time. This result involves the disappearance
of surface basic hydroxyl groups and the formation of adsorbed
benzoate species. The number of active sites can be estimated
tentatively for MgO oxide from extrapolation to t → 0 of the
curve conversion versus time (Fig. 1). Taking into account the
amount of benzyl alcohol formed with time, we can deduced the
number of hydroxyl groups reacts with benzaldehyde.
After total inhibition of reaction with time on flow, the
replacement of nitrogen by dihydrogen as the carrier gas entailed
the progressively apparition of the activity on CaO, SrO and BaO
oxides, whereas MgO oxide was completely inactive. However,
although the raise of the activity with reaction time has been
observed, one notices that the initial activity gotten under cur-
rent of nitrogen was not recovered. This result suggests that the
regeneration of active species such as basic hydroxyl groups
However, the analysis results show that the increase in the
reaction temperature leads to the increase the number of the sur-
face active hydroxyl species. At 100 C, the number of hydroxyl
◦
groups, which were consumed after reaction are approximately
2
2
◦
18 OH/nm and this number increased to 25 OH/nm at 150 C.
Haffad et al. [2] were showed that the determined OH number
not reflects the real basicity of surface metal oxide and may
be different from the total number of hydroxyl groups present
on the surface. The observed activity to benzyl alcohol would
only depend on the number of strong basic hydroxyls, referred
to as type I, is a terminal OH coordinated to one metal cation
(M OH). It corresponds to the highest wave number in the IR
spectrum. The presence of basic OH groups was proven by the
formation of hydrogen carbonate species with the adsorption
CO2 [34] or by CS2 and COS hydrolysis [35].
(
OHsurface) or others intermediate species are probably more
difficult.
4
4
.1.2. Selectivity
.1.2.1. Benzyl alcohol selectivity. After N2 pretreatment, ben-
zyl alcohol was selectively obtained only on MgO oxide in
the whole range of reaction temperature (Table 2). It was also
◦
obtained like only reaction product at TR < 250 C and the reac-
On the other hand, on the basis of the basic and acidic nature
of alkaline earth metal oxides and the kinetic results [14], it is
inferred that benzylate and benzoate species are formed over
CaO oxide:
tivity decreased with time on flow (Fig. 1), tending to zero order.
Analogous to results obtained in homogeneous medium [21],
Benzaldehyde reacts with basic sites O2 of CaO oxide to
−
benzyl alcohol should be produced by the Cannizzaro reaction:
−
−
form an intermediate (I), while it reacts with Lewis acid sites
2
PhCHO + OH → PhCOO + PhCH2OH
2+
Ca to form an intermediate (II). Calcium benzoate (III) and
benzylate (IV) are formed by the hydride ion transfer form (I)
to (II). Benzyl alcohol is considered to be formed by a proton
transfer of surface hydroxyl group to benzylate (IV).
The decrease of the benzyl alcohol formation with the time on
flow leads to the decrease in the number of the hydroxyl surface
species andconsequently, in benzaldehyde conversion. TheCan-
nizzaro reaction stoichiometrically consumes surface hydroxyl
groups (OHsurf) then ends up. The mechanism on oxide surfaces
can be written as follows:
On CaO, SrO and BaO oxides, no benzaldehyde conversion at
◦
◦
100 C and no benzyl alcohol formation above 100 C are prob-
ably due to the strong adsorption of benzylate species formed
on the surface oxide, and the increase in the reaction tempera-
ture changes the surface benzylate species to benzoate species
[14]. Benzaldehyde is known to dissociatively adsorb as ben-
PhCHOg + OH
ꢀ PhCH(OH)Oads
surf
PhCHOg
−
→ PhCOOsurf + PhCH2OHg

A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
65
zoate and benzylate species on basic oxides [2,20,21,36] and
subsequent heating leads to the disproportionation to toluene
toluene formation is accompanied to benzoate species. The ele-
vation of reaction temperature decomposed the benzoate species
to benzene and CO2.
All results strongly suggest that benzoate surface species
always appear during the formation of benzyl alcohol, toluene
or benzene. A possible intermediate is a hemiacetal specie Ph-
CH(OH)Oads [39]. It may be regarded as a hydride producer for
reductions (via benzylate species) or a benzene precursor.
[
[
15], or to the decomposition of these species to benzene
20,37–38].
4
.1.2.2. Toluene selectivity. In such conditions, only CaO oxide
produced selectively the toluene (Table 2). It appears in the range
◦
1
50–350 C and decreases progressively with a reaction temper-
ature. Formally, toluene formation requires two hydrogen atoms
from benzaldehyde and four hydrogen atoms if water appears
4
.2. Benzaldehyde reduction under dihydrogen
atmosphere (H2)
[
2]. In the absence of dihydrogen flow, hydrogen comes from
surface hydroxyl or aldehydic groups. Sreekumar and Pillai [15]
showed that toluene product is formed by disproportionation
from benzyl alcohol:
4
.2.1. Activity
The use dihydrogen flow, as the carrier gas can provokes not
only the direct hydrogenation of benzaldehyde but it can also
modify the surface properties [2]. Indeed, the comparison of the
obtained results with the two reaction processes revealed that, in
the range reaction temperature, the activity was the most weak
under atmosphere hydrogen. Thus, the dihydrogen pretreatment
is expected to affect the hydroxyl groups. The dihydrogen pro-
duced some oxygen vacancies by moderate reduction of the
surface [2,24].
However, the activity of the irreducible alkaline earth metal
oxides was modified by the dihydrogen atmosphere process and
decrease in the order: CaO > SrO > BaO > MgO = 0 (Table 3).
In the whole temperature reaction, no conversion was observed
over the irreducible magnesia and the H2 pretreatment dras-
tically inhibited the Cannizzaro reaction. One infers that this
activation mode provokes the modification of hydroxyl groups
and, as a result, the total inhibition of the dismutation reaction
and mechanism of dehydroxylation is probably complex. On the
other hand, the observed activities were not related directly to
the reducibility of the metal oxides. The order of the standard
potentials redox E0 (ox/red) [40]:
2
PhCH2OH → PhCH3 + PhCHO + H2O
On CaO oxide, no trace of benzyl alcohol was detected in the
reaction mixture, suggesting that alkoxy species do not desorb
(
reaction 2) and probably react directly by a new surface step
to toluene [2]. By analogous to Sreekumar and Pillai reaction,
toluene may arise from successive reactions.
Benzaldehyde dismutation to surface benzylate species (Can-
nizzaro reaction):
2
PhCHOg → PhCOOads + PhCH2Oads
and the presence of two nearby benzylate species could trans-
formation to toluene (Pillai reaction):
2
PhCH2Oads + 2Hads → PhCH3 + PhCHOg + H2Og
(1)
or a benzylate species and a benzaldehyde molecule (reaction
2
):
PhCH2Oads + PhCHOg → PhCH3 + PhCOOads
(2)
BaO/Ba(−1.59 V) > MgO/Mg(−1.86 V)
> CaO/Ca(−2.26 V) > SrO/Sr(−2.59 V)
To differentiate mechanisms (1) and (2), we had recourse to TCD
detector in GC analysis. The results reveal the totally absence
of the water formation and the presence in the reaction mixture
the concomitant formation of CO2. These results confirm the
high benzene selectivity obtained on CaO oxide, and suggest
that mechanism (2) prevails.
is different from the order of activity of the alkaline earth metal
oxides. Under the experimental conditions, probably the hydro-
gen molecules could be not activated by the irreducible MgO
oxide because of its small crystallite size, whereas CaO, SrO,
BaO oxides that exhibits thick crystallite size can probably
encourage the activation of hydrogen molecules (Table 1). This
result was confirmed when the dihydrogen was replaced by the
nitrogen as carrier gas. The benzaldehyde conversion decreased
with time on stream until the total reaction inhibition on all
oxides, what shows that the reaction is become stoichiometric
and the presence of a nondissociative or dissociative hydrogen
adsorption on the surface oxides has been completely consumed.
4
.1.2.3. Benzene selectivity. Some benzene appears on all alka-
line earth metal oxides and the rate increase with the increasing
of basicity in order: MgO < CaO < SrO < BaO (Table 2). Over
SrO and BaO oxides, the high selectivity to benzene (100%)
◦
◦
between 150 C and 350 C, and the increased benzaldehyde
conversion to benzene with the increasing reaction temperature
suggests that the final product should be dependant only the
acidic or redox surface properties [2]. Benzene is usually con-
sidered to be the result of the decomposition of surface benzoate
species with carbon dioxide formation [37]:
4.2.2. Selectivity
The fact that the aromatic nucleus is unreactive while the
carbonyl group undergoes hydrogenation and hydrogenolysis
suggest that the aromatic aldehyde reacts principally with the
catalyst via carbonyl function [19]. In such case, the adsorbed
hydrogen preferably attack the carbonyl group due to the energy
barriers [18,41,42].
C H COO + Hads → C H + CO2g
The high activity and benzene selectivity over SrO and BaO
suggest that a subsequent heating leads to the decomposition of
the benzoate species into benzene and CO2. On CaO oxide, the
6
5
ads
6
6g

6
6
A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
4
.2.2.1. Benzyl alcohol selectivity. At low reaction temperature
function on a Br o¨ nsted acid site (H–O–Msurf) followed by
hydrogenolysis of the protonated specie on a adjacent metal
site (M H):
◦
(
<250 C), benzyl alcohol was produced with high selectivity at
low conversion on CaO (100%), SrO (100%) and BaO (56.7%)
oxides. The elevation of reaction temperature seems to indicate
that benzyl alcohol disappeared progressively with the increase
of benzaldehyde conversion. This result reveals that the benzyl
alcohol is favored only at low benzaldehyde conversion.
In such conditions, benzyl alcohol is the primary product dur-
ing the gas-phase hydrogenation of benzaldehyde over alkaline
earth metal oxides. In the presence of dihydrogen flow, these
oxides favored the carbonyl reduction compound to the corre-
sponding alcohol, even at low conversion. As to the chemical
process of benzyl alcohol formation, it can be supposed to be
•
In the case of BaO oxide, during the steady-state setting-up at
low reaction temperature (150 C), toluene appeared as a sec-
◦
ondary product. The presence of benzyl alcohol and toluene
products in the reaction mixture suggests the existence of
another reaction path than that described above. This path
probably involves very active sites on which toluene is fastly
formed in the early stages of the reaction where the partial
pressure of both reactants and products were low. These sites
were then poisoned at higher conversion by the one or the
other reactant molecule, even water. The possible route would
be the benzylalcohol dismutation to toluene, benzaldehyde
and water on the support [47–49]:
1
–2 nucleophilic addition, with a high polarization of the tran-
sition state, as in the case of the hydrogenation of conjugated
carbonyl compounds [41–44]:
Therefore, hydrogen surface species may be considered as
active intermediates for the direct hydrogenation of benzalde-
hyde to benzyl alcohol [2,5,19,45]. It was reported by several
workers that the infrared studies have shown on different oxides,
the heterolytic and homolytic dissociation of dihydrogen [45].
The surface oxide has been totally dehydrated prior by H2
adsorption and the OH group is likely to originate from the dis-
sociative adsorption of the hydrogen molecule occurring on the
thoroughly dehydrated oxide [16].
It is worth noting that MgO bare oxide, which is considered to
be basic, no benzyl alcohol was observed in the exit gas probably
because, more strongly held on the oxide surface or the hydroxyl
groups, active in the Cannizzaro reaction, has been removed by
the dihydrogen pretreatment.
4.2.2.3. Benzene selectivity. In the present case, Benzene is
formed only at high reaction temperature (>250 C) on CaO
◦
(>89%), SrO (>90%) and BaO (100%) oxides. It is believed to
be the product of the hydrogenolysis of the external C(aryl)–C
bond of benzaldehyde [18,19]. Vannice and Poondi [1] were
showed that benzene formation was produced directly from the
reactant molecule and not from the alcohol or toluene intermedi-
ates. Carbon monoxide was identified as co-product of benzene
[1]. However, Herskowitz [46] was showed that benzene is the
sequential product of the consecutive hydrogenation of the alco-
hol to toluene; it appears only when essentially all aldehyde has
disappeared.
In our study, using of TCD detector in GC analysis has shown
that no CH4 was detected. The results were revealed in the reac-
tionmixturethepresenceofcarbonmonoxideasco-productben-
zene. These results confirm the highbenzene selectivity obtained
on CaO, SrO and BaO oxides. In the following sequence, we can
be simplified the benzaldehyde hydrogenolysis to benzene:
4
.2.2.2. Toluene selectivity. It is generally considered that
toluene is the product of a consecutive reaction of hydrogenoly-
sis of benzyl alcohol [1,2,19,46]. Indeed, flowing the CaO, SrO,
andBaOoxideswiththebenzylalcoholinsteadofbenzaldehyde,
selectively led to toluene production. In a parallel experiment,
replacing benzaldehyde by benzylalcohol in the reactant flow
showed the formation of toluene with a selectivity of 100%.
However, benzyl alcohol is a complex molecule possessing sev-
eral active centers such as the acidic proton and basic oxygen
of the O H function and the reductive hydrogen of the C H
exocyclic function. All these centers may interact with the oxide
surface or dihydrogen molecule [19]. Also several pathways can
be envisaged for toluene formation.
C H CHOg → C H + CO
6
5
6
6
ads
The exact nature of the reaction paths of benzaldehyde
hydrogenolysis to benzene and carbon monoxide is not yet
reported to our best knowledge. It probably implies a bifunc-
tional site: the adsorption of the reactant molecule on an
acidic/cationic site then dissociation of the aldehydic C H bond
on a metal site. Further work is needed to obtain more informa-
tion’s on the real involved reaction paths.
•
In the case of the CaO and SrO oxides, the totally disap-
pearance of benzyl alcohol product and the appearance of
toluene product with low selectivity in the mixture reaction
supposed that the hydrogenation of benzaldehyde to benzyl
alcohol can continue to toluene. In such a case, the activity
in toluene should reflect the aptitude of the surface oxide
to both adsorb the alcohol and dissociate H2. The toluene
formation probably resulted in the protonation of the alcohol
5. Conclusion
Reduction of benzaldehyde has been chosen as a model reac-
tion for determining the surface properties of alkaline earth

A. Saadi et al. / Journal of Molecular Catalysis A: Chemical 258 (2006) 59–67
67
metal oxides. Under nitrogen atmosphere, the results show that
the active sites are basic hydroxyl groups. Only MgO oxide
produced selectively the benzyl alcohol, whereas toluene and
benzeneareusuallyformedathighreactiontemperatureonCaO,
SrO and BaO oxides. On MgO, the mechanism suggests that
benzaldehyde reduction to benzyl alcohol is the result of the
Cannizzaro dismutation. The reaction requires basic hydroxyl
groups. The observed conversions depend only on the initial
number and strength of these entities.
Under dihydrogen flow, a stationary state is reached on CaO,
BaO and SrO, whereas MgO is inactive. The presence of H2 flow
not affects the benzyl alcohol formation on CaO oxide. This
result suggests that the presence of totally dehydrated surface
is probably necessary for both dissociative and nondissocia-
tive adsorption of dihydrogen, as consequence, benzaldehyde is
directly reduced to benzyl alcohol. Toluene product arose from
the hydrogenolysis of benzyl alcohol on probably bifunctional
sites involving metal hydride species. As to benzene, it resulted
from benzaldehyde hydrogenolysis probably through a surface
hydrogenate specie. Useful information on the surface proper-
ties of the alkaline earth metal oxides and the reaction course
were given by comparing the transformations of bnezaldehyde
under N2 or H2 atmosphere.
[13] J.A. Schreifels, P.C. Maybury, W.E. Swartz, J. Org. Chem. 46 (1981) 1263.
[
[
[
14] K. Tanabe, K. Saito, J. Catal. 35 (1974) 247–255.
15] R. Sreekumar, C.N. Pillai, Catal. Lett. 19 (1993) 281.
16] M.C. Paganini, M. Chiesa, E. Giamello, S. Collucia, G. Marta, D.M. Mur-
phy, G. Pacchioni, Surf. Sci. 421 (1999) 246–262.
[17] E.V. Stefanovich, T.N. Truong, J. Chem. Phys. 102 (1995) 5071.
[18] M.A. Keane, Appl. Catal. A 118 (1997) 261.
[
19] A. Saadi, Z. Rassoul, M.M. Bettahar, J. Mol. Catal. A: Chem. 164 (2000)
05–216.
2
[
[
20] C.A. Koustall, P.A.J.M. Angeware, V. Ponec, J. Catal. 143 (1993) 573.
21] J. March, Advanced Organic Chemistry, Intern. Stud. Edit., Mc Grawhill,
Tokyo, 1997.
[22] T.F. Mao, J.L. Falconer, J. Catal. 123 (1990) 443.
[
[
[
[
[
23] B. Sen, J.L. Falconer, J. Catal. 117 (1989) 404.
24] B. Sen, J.L. Falconer, J. Catal. 113 (1988) 444.
25] B. Sen, J.L. Falconer, J. Catal. 123 (1990) 45.
26] B. Chen, J.L. Falconer, K.M. Bailey, B. Sen, Appl. Catal. 66 (1990) 283.
27] A. Guiner, X-Ray Diffraction in crystals, Imperfect Crystals and Amor-
phous Bodies, Dover Publications, NY, 1994.
[
[
[
28] T. Iizura, H. Hattori, Y. Ohno, J. Sohma, K. Tanabe, J. Catal. 22 (1971)
1
30.
29] H. Hattori, N. Yoshii, K. Tanabe, Proc. 5th Int. Congr. Catal. 10 (1973)
33.
30] Y. Fukuda, K. Tanabe, Bull. Chem. Soc. Jpn. 46 (1973) 1616.
2
[31] D. Haffad, M.M. Bettahar, A. Chambellan, J. Lamotte, J.C. Lavalley, Act.
th Coll. Franc. Maghr. Catal. T1 (1996) 57–82.
4
[
[
[
[
32] A.E.T. Kuiper, J. Medema, J.J.G.M. Van Bokhoven, J. Catal. 29 (1973) 40.
33] L.D. Chapman, M.L. Hair, Proc. 3rd Int. Congr. Catal. 1964 (1965) 1091.
34] C. Morterra, L. Orio, Mater. Chem. Phys. 24 (1990) 247.
References
35] P.E. Hoggan, A. Aboulayt, A. Pieplu, P. Nortier, J.C. Lavalley, J. Catal. 149
(
1994) 300.
[
[
1] M.A. Vannice, D. Poondi, J. Catal. 169 (1997) 166.
2] D. Haffad, U. Kameswari, M.M. Bettahar, A. Chambellan, J.C. Lavalley,
J. Catal. 172 (1997) 85.
[36] F. Lepeltier, P. Chaumette, J. Saussey, M.M. Bettahar, J.C. Lavalley, J. Mol.
Catal. A: Chem. 122 (1997) 131.
[37] S.T. King, E.J. Strojny, J. Catal. 76 (1982) 274.
[38] M. Coudurier, G. Wieker-Coudurier, C. Naccache, B. Vital-Mathieu, Bull.
Soc. Chim. Fr. 4 (1969) 1121.
[39] G. Busca, J. Lamotte, J.C. Lavalley, V.J. Lorenzelli, J. Am. Chem. Soc. 209
(1987) 5197.
[40] M. Pourbaix, Atlas des Equilibres Electroniques, Gauthiers-Villars, Paris,
1963.
[41] A. Chambers, S.D. Jackson, D. Stirling, G. Webb, J. Catal. 168 (1997) 301.
[42] R. Hubault, J.P. Bonnelle, J. Mol. Catal. 55 (1989) 170.
[43] J. Jenck, J.E. Germain, J. Catal. 65 (1980) 141.
[44] R. Hubault, M. Daage, J.P. Bonnelle, Appl. Catal. 22 (1986) 231.
[45] J. Saussey, J.C. Lavalley, J. Lamotte, T.J. Arai, J. Chem. Soc. Chem. Com-
mun. 278 (1982).
[
[
[
[
[
[
[
3] A. Saadi, Z. Rassoul, M.M. Bettahar, J. Barrault, J. Alg. Chem. Soc. 8 (2)
(
1998) 224.
4] A. Aboulayt, A. Chambellan, M. Marzin, J. Saussey, F. Maug e´ , J.C. Laval-
ley, C. Mercier, R. Jacquot, Stud. Surf. Sci. Catal. 78 (1993) 131.
5] A. Saadi, M.M. Bettahar, Z. Rassoul, Stud. Surf. Sci. Catal. 130 (2000)
2
261.
6] A. Saadi, R. Merabti, M.M. Bettahar, Z. Rassoul, J. Alg. Chem. Soc. 11
2) (2001) 231–240.
7] R. Merabti, A. Saadi, Z. Rassoul, M.M. Bettahar, J. Alg. Chem. Soc. 16
1) (2006) 65–73.
8] A. Saadi, R. Merabti, Z. Rassoul, M.M. Bettahar, J. Mol. Catal.: A Chem.
53 (2006) 79–85.
(
(
2
9] P.N. Rylander, Catalytic Hydrogenation in Organic Synthesis, Academic
Press, London, 1979.
[46] M. Herskowitz, in: M. Guisnet (Ed.), Heterogeneous Catalysis and Fine
Chemical, Elsevier, Amsterdam, 1991, p. 105.
[
[
[
10] P.N. Rylander, Hydrogenation Methods, Academic Press, London, 1985.
11] J. Jenck, J.E. Germain, J. Catal. 65 (1980) 141.
12] W.H. Carothers, R. Adams, J. Am. Chem. Soc. 47 (1925) 1047.
[47] E.A.I. Heiba, P.S. Landis, J. Catal. 3 (1964) 471.
[48] M. Jayamani, C.N. Pillai, J. Catal. 82 (1983) 485.
[49] M. Jayamani, N. Murugesan, C.N. Pillai, J. Catal. 85 (1984) 527.