Ammoxidation of ethylene over low and over-exchanged Cr-ZSM-5 catalysts

Article abstract of DOI:10.1016/j.apcata.2011.12.021

Catalytic performances of Cr-ZSM-5 catalysts (5 wt.% of Cr, Si/Al = 26), prepared by solid-state reaction and aqueous exchange, from Cr nitrate and Cr acetate precursors, were evaluated in the selective ammoxidation of ethylene into acetonitrile in the temperature range 425.500 °C. Catalysts were characterized by chemical and thermal analysis, XRD, N₂ physisorption, ²⁷Al MAS NMR, TEM, UV-vis DRS, Raman, DRIFTS and H₂-TPR. Characterization results shown that solid-state exchange was favorable for Cr₂O₃ formation, while exchanging chromium in aqueous phase led, essentially, to Cr(VI) species. Catalysts were actives and selectives in the studied reaction, and among them, those, prepared from aqueous exchange, exhibited the highest acetonitrile yields (23±0.5%, at 500 °C). Improved catalytic properties can be correlated with the chromium species nature. In fact, mono/di-chromates and/or polychromate species, sited in the charge compensation positions, were definitively shown, as being, the active sites. Furthermore, during solid-state reaction, the agglomeration of Cr₂O₃ oxide should be avoided since these species inhibit the catalyst activity.

Full text of DOI:10.1016/j.apcata.2011.12.021

Applied Catalysis A: General 415–416 (2012) 132–140
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Ammoxidation of ethylene over low and over-exchanged Cr–ZSM-5 catalysts
a,∗
a
b
b
c
a
F. Ayari , M. Mhamdi , J. Álvarez-Rodríguez , A.R. Guerrero Ruiz , G. Delahay , A. Ghorbel
a
Laboratoire de Chimie des Matériaux et Catalyse, Département de Chimie, Faculté des Sciences de Tunis, Campus Universitaire Tunis ElManar, 2092 Tunis, Tunisia
Unidad Asociada Group for Desig. and Appl. of Heter. Catal. UNED-ICP (CSIC), Dpto. Inorganic and Technical Chemistry, Faculty of Sciences, UNED, 28040 Madrid, Spain
Institut Charles Gerhardt Montpellier, UMR 5253, CNRS-UM2-ENSCM-UM1, Eq. “Matériaux Avancés pour la Catalyse et la Santé”, ENSCM (MACS – Site la Galéra), 8, rue Ecole
b
c
Normale, 34296 Montpellier Cedex 5, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 October 2011
Received in revised form
Catalytic performances of Cr–ZSM-5 catalysts (5 wt.% of Cr, Si/Al = 26), prepared by solid-state reaction
and aqueous exchange, from Cr nitrate and Cr acetate precursors, were evaluated in the selective ammox-
◦
idation of ethylene into acetonitrile in the temperature range 425–500 C. Catalysts were characterized by
2
4 November 2011
27
chemical and thermal analysis, XRD, N2 physisorption, Al MAS NMR, TEM, UV–vis DRS, Raman, DRIFTS
Accepted 12 December 2011
Available online 18 December 2011
and H2-TPR. Characterization results shown that solid-state exchange was favorable for Cr2O3 formation,
while exchanging chromium in aqueous phase led, essentially, to Cr(VI) species. Catalysts were actives
and selectives in the studied reaction, and among them, those, prepared from aqueous exchange, exhib-
Keywords:
Cr–ZSM-5
Ammoxidation
Acetonitrile
Solid-state exchange
Aqueous exchange
◦
ited the highest acetonitrile yields (23 ± 0.5%, at 500 C). Improved catalytic properties can be correlated
with the chromium species nature. In fact, mono/di-chromates and/or polychromate species, sited in the
charge compensation positions, were definitively shown, as being, the active sites. Furthermore, during
solid-state reaction, the agglomeration of Cr2O3 oxide should be avoided since these species inhibit the
catalyst activity.
Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
1
. Introduction
In the past few decades, a large number of catalytic systems have
The catalytic activity of Cr–ZSM-5 catalysts in the ammoxi-
dation of ethylene lies in the variability of oxidation states of
chromium, the degree of polymerization (mono/di-chromates or
polychromates, etc.), the aggregation state of oxide species and
their crystallographic structure (amorphous or crystalline). Thus, it
has been accepted that the improved catalytic activity of Cr–ZSM-5
catalysts is due to the presence of (poly)-chromate species and Cr
been developed in our laboratory for light hydrocarbons ammox-
idation into acetonitrile. In fact, numerous works have dealt with
the shape-selective ammoxidation of ethylene using either micro-
porous materials (zeolites) or transition metal oxides as catalyst
supports.
ions in charge compensation positions, while agglomerated Cr O3
2
For example, the catalytic performance of Co/ZSM-5 catalysts
oxide should be avoided [6].
[
1–4] has been evaluated in the ammoxidation of ethylene into ace-
Cr O oxide species provided essentially from the fact that,
2
3
tonitrile. Sol–gel derived Co/Al O and Cr/Al O catalysts have also
been used [5].
during solid-state exchange, a non-negligible fraction of Cr ions
does not diffuse inside the zeolite channels but remains on the
outer surface of the grain. In this way, an optimization of catalysts
preparation method would thus consist in tuning the operating
conditions and the nature of chromium salt to ensure the effective
diffusion of Cr ions inside the zeolite.
In the case of zeolite based catalysts, fundamental studies [7–9]
have provided a strong dependence of the exchange level, i.e. the
preparation method, and the metal speciation in a wide range of
catalytic reactions.
2
3
2
3
More recently, we have reported the performance of Cr–ZSM-5
catalysts, issued from solid-state exchange, in the ammoxida-
tion of ethylene after varying the source of chromium [6].
The selectivity and the yield of acetonitrile largely depended
on the nature of the applied catalysts. For example, 26% of
C2H4 conversion and 95% of selectivity toward acetonitrile have
◦
been reached at 500 C after choosing an appropriate chromium
precursor.
The aim of this work is the comparison between the catalytic
behaviors of under- and over-exchanged Cr–ZSM-5 catalysts in the
ammoxidation of ethylene into acetonitrile. It should be underlined
that a large part of the work has been devoted to the physicochem-
ical properties of the prepared materials.
∗ Corresponding author. Tel.: +21650646504; fax: +216 71885008.
E-mail address: lcmc fa@yahoo.fr (F. Ayari).
0
926-860X/$ – see front matter. Crown Copyright © 2011 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.12.021

F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
133
2
. Experimental
H -TPR analysis was performed with a Micromeritics Autochem
2
2
910 analyzer, in a Pyrex U-tube reactor and an on-line thermal
◦
2.1. Catalysts preparation
conductivity detector. The catalyst (70 mg) was dried at 500 C for
1
◦ ◦
h and reduced from 50 to 1000 C (15 C/min) with H (3%)/Ar
2
Solid-state exchange was performed as follows: 1.5 g of
flow.
NH +–ZSM-5 (Zeolyst, Si/Al = 26) was mixed in a mortar and pestle
Ammoxidation of ethylene was studied with 100 mg of catalyst
4
◦
with Cr acetate (Strem Chemicals) or Cr nitrate (99%, Acros Organ-
ics) in the desired molar ratio Cr/Al = 1, which corresponds to 0.2 g
of Cr acetate and 0.36 g of Cr nitrate. The finely ground powders
between 425 and 500 C using a down flow tubular glass reactor. In
all cases, the inlet reactants composition was 10% O2 (Air Liquide
99.995%), 10% C H4 (Air Liquide 99.995%) and 10% NH3 (Air Liq-
uide 99.96%). The total flow rate was maintained at 100 cm /min
by balancing with helium (Air Liquide 99.998%). The analysis of the
outlet flow was recorded on-line by two chromatographic units,
one operated with a flame ionization detector, while the other was
equipped with a thermal conductivity detector.
2
◦
3
obtained by mechanical mixture was heated for 12 h at 500 C in
3
◦
helium (30 cm /min, heating rate 2 C/min).
Aqueous exchange was performed as follows: 1 g of NH4+–ZSM-
5
(Si/Al = 26) was exchanged with 100 ml of 0.01 M Cr acetate
◦
solution (pH = 4.6) or Cr nitrate (pH = 3.5) for 24 h at 70–80 C. After
three identical exchanges, the resulting zeolite slurry was filtered,
washed with 500 ml of de-ionized water and then filtered again.
The conversion, selectivity and yield are defined as follows:
Conversion of C H ,
2
4
◦
Finally, the zeolite was dried at 110 C overnight and then treated
ꢀ
◦
3
◦
yin
i
in helium at 500 C for 3 h (30 cm /min, heating rate 2 C/min)
Catalysts were labelled Cr–P-PM, where P refers to the Cr pre-
cursor nature (A: Cr acetate and N: Cr nitrate), while PM refers to the
preparation method: AE (aqueous ion exchange) and SE (solid-state
exchange).
i
X =
ꢀ
yEnE +
yini
i
Selectivity of product P (carbon basis),
i
2
.2. Catalysts characterization
y n
i
i
S = ꢀ
i
yini
The elemental analysis (Cr, Si, Al) content of different solids was
i
determined by ICP at the Vernaison Center of the CNRS (France).
DTA/TGA–MS analyses were performed using a SDT Q600 appa-
ratus with ∼30 mg of zeolite/precursor mixture (5 wt.% of Cr) and a
precursor mass, which also corresponds to 5 wt.% of Cr. The thermal
Yield of product P (carbon basis), Y = XS where y and y are the
E
i
i
i
i
mole fractions of product P and C H , respectively; n and n are
E
i
2
4
i
the number of carbon atoms in each molecule of product P and
C H , respectively.
i
3
treatment was performed under helium (30 cm /min) between 30
2
4
◦
◦
and 700 C at a heating rate of 5 C/min. The chemical composition
of gaseous products was determined using a mass spectrometer
piloted with Quadstar 32 Bits software.
3
. Results
N2 adsorption–desorption isotherms were determined at 77 K
with an automatic ASAP 2000 apparatus from Micromeritics after
pretreatment under vacuum at 200 C for 5 h. Specific surface area
3.1. Chemical analysis
◦
Table 1 gives the elemental analysis of the parent zeolite and
Cr–P-PM catalysts. The Si/Al molar ratio values, determined by ICP,
corresponded to the data provided by the zeolite manufacturer.
Cr–P-SE solids exhibit high Cr/Al molar ratios, which indicate
that the metal was retained after solid-state exchange. On the other
hand, Cr–P-AE solids exhibit the lowest Cr contents (thus the lowest
Cr/Al molar ratios) because of diffusional limitation which occurs
during the exchange in aqueous phase.
was determined by BET method, microporous volume by t-plot
method and porous volume is the volume adsorbed at P/P = 0.98.
XRD measurements were performed on an X’Pert Pro
X-ray diffractometer from PANalytical with CuK␣ radiation
◦
(
ꢀ = 1.54060 A˚ ), generator setting of 40 kV and 40 mA, a scanning
◦
speed of 0.05 /min, and scanning region of 2–70. The diffractometer
◦
◦
was operated at 1.0 diverging and 0.1 receiving slits and a con-
tinuous intensity trace was recorded as function of 2ꢁ. Structural
data identification was performed using EVA software.
Besides elemental analysis, the resulting ion exchange degrees
3+
are presented in Table 1. Considering that Cr ions are only the
TEM study was performed with a JEOL JEM-2000 FX microscope
type, operating at 200 kV. The samples were prepared by grind-
ing and ultrasonic dispersal into an acetone solution, placed on the
copper TEM grid, and then evaporated.
possible exchange species, the exchange degree is defined by
3
00 × Cr/Al (mol/mol). The data shows that the catalysts issued
from solid-state exchange contain much more chromium than
required for 100% total ion-exchange. These samples are regarded
as ‘over-exchanged’ materials contrary to Cr–P-AE catalysts,
which exhibit very low exchange degrees. Kouwenhoven [10]
reported that in ZSM-5 zeolite type, the positions of tetrahedrally
27
Al MAS NMR spectra were recorded at 78.20609 MHz on a
Bruker WB spectrometer using AlClO ·6H O as reference. An over-
3
2
all 4096 free induction decays were accumulated. The excitation
pulse and recycle time were 6 ␮s and 0.06 s, respectively.
UV–vis DRS spectra were recorded at room temperature in the
wavelength range 900–200 nm on a Perkin Elmer Lambda 45 spec-
trophotometer equipped with a diffuse reflectance attachment.
Parent zeolite was the reference material.
3+
coordinated Al ions are spatially much separated and therefore,
Table 1
Chemical analysis results.
Sample
Cr (wt.%)
Al (wt.%)
Si/Al
Cr/Al
Ion
Raman measurements were carried out on a confocal Thermo
Scientific DXR Raman Microscopy system using the visible line at
(
mol/mol) (mol/mol) exchange
degree
(%)
5
32 nm and an incident power of 10 mW.
DRIFTS spectra were recorded on a Bruker IFS 55 spectropho-
NH4+–ZSM-5
–
1.40
1.30
1.33
1.62
1.41
27.38
27.02
27.53
23.23
26.23
–
–
366
252
63
tometer equipped with a Thermo Spectra Tech reacting cell at a
Cr–A-SE
Cr–N-SE
Cr–A-AE
Cr–N-AE
3.06
2.16
0.65
0.09
1.22
0.84
0.21
0.03
−
1
spectral resolution of 4 cm
and accumulating 200 scans. Sam-
◦
◦
ples were treated in situ at 500 C with helium (5 C/min, flow:
3
10
3
0 cm /min).

1
34
F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
a) 100
30
(b) 100
50
(
TGA
DTA
TGA
DTA
9
8
7
6
5
4
3
0
0
0
0
0
0
0
0
exo
2
1
0
0
0
80
60
40
20
Exo
-50
-100
-150
-200
-250
-10
-
300
100
200
300
400
500
600
700
1
00 200 300 400 500 600 700
Temperature (°C)
Temperature (°C)
Fig. 1. DTA/TGA curves of (a) Cr3(OH)2(CH3COO)7 and (b) Cr(NO3)3·9H2O.
bare Cr6+ and/or Cr3+ cations cannot exist since it would be nec-
essary to have 3 or 6 aluminum neighbors. Similarly to Fe/ZSM-5
system [11], it is possible to have the presence of chromium
of the residue color, from dark green to green, which reflect the
decomposition of Cr acetate into Cr2O3.
TG curve of Cr acetate/zeolite mixture (Fig. 2a) displays three
(
hydr)oxo-cation species like [OH Cr
O
Cr OH]2+ species,
weight losses of 2.79, 0.44 and 6.92%. The first one (30–182 C,
◦
◦
for example.
endothermic peak at 65 C) corresponds to the dehydration of both
zeolite and precursor as confirmed by MS signal profile charac-
terizing the evolution of water. The second weight loss, situated
◦
3.2. Thermal analysis
between 182 and 237 C, corresponds to the loss of ammonia
molecules, while the last one evidencing the solid-state reaction
◦
Fig. 1(a and b) illustrates DTA/TGA curves of chromium acetate
between Cr acetate and zeolite between 250 and 500 C. The evolu-
tion of MS fragment profiles of CH4+, CO2+ and CH COOH (results
not shown) reflect the solid-state exchange, which can be simpli-
fied as follows:
+
and chromium nitrate, respectively. TG curve of Cr acetate shows
different steps of weight losses: the first one, situated at relatively
3
◦
low temperatures (30–250 C), corresponds to the elimination of
◦
adsorbed water, while the second one starts at ∼300 C and corre-
+
−
2
(NH , O − Zeo + Cr(CH COO) · xH O
4
3
3
2
sponds to the salt decomposition. TG curve of Cr nitrate displays
◦
◦
three weight losses at relatively low temperatures (<200 C) and
30–250 C
2+
−
→
2CH COOH + 2NH + xH O + (CrCH COO)
( O
Zeo)2
◦
3
3
2
3
another one at ∼450 C.
Fig. 2(a and b) illustrates DTA/TGA curves of chromium
salts/zeolite mixtures. Fig. 2a and b indicates that the solid-
(CrCH3COO)2
+
( O
Zeo)2 + H2O
◦
state reaction is achieved before 500 C. On the other hand,
◦
there is a similarity between DTA/TGA curves of precursors and
those of mixtures, evidencing that the exchange process might
involve the intermediate compounds of Cr precursors decompo-
sition.
250–500 C
2+
−→
Cr(OH) + (
O
Zeo) + CH + CO
2
4
2
The thermal decomposition of Cr nitrate (Fig. 1b) occurs accord-
ing to two principal steps: the first one (below 250 C) corresponds
to the elimination of water molecules and the degradation of NO3
◦
After dehydration, (weight loss of 8.62%, endothermic peak at
◦
−
6
8 C), Cr acetate begins to decompose (Fig. 1a), leading to the loss
◦
of various molecules such as acetic acid and water (weight loss of
groups (weight loss of ∼74%), while the second step, above 350 C,
◦
5
3.61%, DTA peaks at 347, 383, 407 and 444 C). This phenomenon
corresponds to the oxygen removal from Cr O system (weight loss
of 5.15%) [12,13].
is accompanied by the evolution of MS fragments such as: m/z = 14,
+
+
+
+
1
(
5 and 16 (CH ), 17 and 18 (H O ), 28 (CO ), 30 (CH O ), 43
TG curve of Cr nitrate/zeolite mixture (Fig. 2b) displays two
weight losses of 14.75% (30–212 C) and 0.49% (400–480 C). On
4
2
2
+
+
+
+
◦
◦
CH CO ), 44 (CO ), 45 (COOH ) and 60 (CH COOH ) and a change
3
2
3
a) 100
40
40
(
(b) 100
TGA
DTA
TGA
DTA
Exo
30
9
9
9
9
9
8
8
6
4
2
0
8
Exo
2
0
0
9
9
8
5
0
5
2
1
0
0
0
-
20
-10
-20
-40
700
80
1
00
200
300
400
500
600
100
200
300
400
500
600
700
Temperature (°C)
Temperature (°C)
Fig. 2. DTA/TGA curves of (a) Cr acetate and (b) Cr nitrate–zeolite mixtures.

F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
135
Fig. 3. TEM images of (a) Cr–N-SE and (b) Cr–A-SE catalysts.
the other hand, MS study reveals the presence of the follow-
3.4. TEM study
ing fragments: NH3 (m/z = 15), O (m/z = 16), HO+ (m/z = 17),
+
+
H O (m/z = 18), N2+ (m/z = 28), NO+ (m/z = 30), O2 (m/z = 32),
+
+
TEM images of Cr–P-SE catalysts are presented in Fig. 3.
2
N O (m/z = 44), NO2+ (m/z = 46) and HNO3+ (m/z = 63). From both
+
Randomly shaped oxide particles, which are heterogeneously
dispersed, are observed in the TEM images of Cr–N-SE catalyst.
Furthermore, there are some surface regions which contain oxide
particles having a significant size (see image #FX2945).
On the other hand, TEM images of Cr–A-SE reveal some surface
regions which are poor in chromium (see image #FX2931). At the
opposite, in the TEM images #FX2932 and FX2933, some surface
regions contain metallic species without any apparent phases.
It seems that the thermal treatment during solid-state reaction
2
thermal analysis and mass spectrometry results, Cr nitrate decom-
poses according to the mechanism previously described [12,13].
Moreover, such mechanism is also suitable for Cr nitrate/zeolite
◦
mixture. Firstly, the precursor melts into Cr O (30–212 C) then, at
2
3
◦
relatively high temperatures (400–480 C), the desorption of oxy-
gen atoms leads to Cr(III) ions, which may reach the exchange sites
after diffusion.
favored the formation of Cr O3 oxide particles, and the dispersion
of these metallic species is governed by the nature of chromium
precursor.
2
3
.3. Textural properties
Cr–P-AE catalysts as well as the zeolite support exhibit a type
I isotherm evidencing a microporous texture, while Cr–P-SE cat-
alysts exhibit a type IV isotherm, indicative of the mesoporous
character of such solids. Textural properties are summarized in
Table 2.
3
.5. Structural properties
X-ray diffraction patterns of the investigated materials are
shown in Fig. 4. Most catalysts exhibit diffraction peaks which are
typical of crystalline ZSM-5 [15] (Fig. 4a). This result indicates that
Cr species, stabilized after thermal treatment, are dispersed over
the zeolite in the form of small oxide clusters (less than 4 nm)
After aqueous exchange, the zeolite support exhibits a slightly
decrease of its specific surface area and (micro)porous volumes.
This effect can be induced by the presence of oxide aggregates
and/or extra-framework aluminum species dispersed in the chan-
nels or deposited on the outer surface of the zeolite. On the other
hand, Cr–P-SE catalysts exhibit SBET and porous volume values
which exceed the zeolite support ones. Gervasini [14] reported a
similar behavior with cobalt and nickel loaded ZSM-5 zeolite and
considered that such solids exhibit some degree of mesoporosity. In
our study, according to DTA/TGA/MS results, Cr ions were merely
deposited on the external surface of the ZSM-5 zeolite as Cr O3
2
oxide aggregates, and such metallic phase might contribute to the
textural data calculus.
Table 2
Textural properties of the investigated solids.
2
Sample
SBET (m /g) Microporous
Microporous
volume (cm /g) (cm /g)
Porous volume
2
3
3
surface (m /g)
NH4+–ZSM-5 312
239
197
212
161
213
0.139
0.113
0.121
0.092
0.121
0.240
0.525
0.528
0.198
0.225
Cr–A-SE
Cr–N-SE
Cr–A-AE
Cr–N-AE
336
353
248
291
Fig. 4. XRD patterns of (a) NH4+–ZSM-5, (b) Cr–A-SE, (c) Cr–N-SE, (d) Cr–A-AE and
(
e) Cr–N-AE, (*) Cr2O3.

1
36
F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
tetrahedral coordinated framework aluminum [16]. The weak sig-
nal situated at ∼0 ppm corresponds to octahedral aluminum atoms
extracted from the framework [16].
According to NMR spectrum of the zeolite support, the inten-
sity of tetrahedral aluminum peak decreases after metal loading. It
n+
has been reported that exchanging paramagnetic species (like Cr
,
+
2+
with n < 6) with aluminosilicates [17], and Na , Ca cations with
zeolites [18] modify the tetrahedral aluminum environment and
decreases the corresponding peak intensity. On the other hand, the
octahedral aluminum peak intensity does not exhibit any signifi-
cant change after chromium loading. As XRD patterns have clearly
showed up, there is no significant damage of the zeolite structure.
3.6. DRS and Raman studies
UV–vis DRS spectra of prepared catalysts are illustrated in Fig. 6.
Besides the support band (<230 nm), Cr–A-SE catalyst spectrum
Fig. 5. 27Al MAS NMR spectra of both Cr–P-SE catalysts and support.
exhibits DRS bands at 278, 324, 346, 404, 466 and 582 nm, while
bands at 279, 307, 349, 465 and 603 nm were observed in the
DRS spectrum of Cr–N-SE catalyst. In both spectra, the three first
bands are typical for chromate O2 → Cr charge transfer, while
bands situated at ∼465 nm evidencing the presence of polychro-
mate species [19,20]. Finally, DRS bands situated at 582 and 603 nm
are assigned to the d–d transition of Cr3+ ions (probably situated
−
6+
in the exchange sites) or to Cr O3 [19,21]. In the case of Cr–A-SE
2
catalyst, we were unable to explain the presence of the DRS band
situated at 404 nm.
DRS spectra of Cr–P-AE catalysts exhibit two bands at 271 and
3
78 nm, typical for chromate species, and a shoulder at 454 nm evi-
dencing the presence of dichromate and/or polychromate species
[
19,21–23].
Raman spectra of Cr–N-PM catalysts are illustrated in Fig. 7. Zeo-
lite support does not exhibit any Raman features in the region of
−
1
2
00–1100 cm and, therefore, all the observed bands are assigned
to chromium species vibrations.
Cr–A-SE and Cr–A-AE catalysts do not exhibit any Raman fea-
tures, because of fluorescence phenomenon.
Fig. 6. DRS spectra of Cr–P-SE catalysts, Fig. in-set: Cr–P-AE catalysts spectra.
In our previous study [6], we reported Raman spectra of
mechanical mixtures of CrO –ZSM-5 and Cr O –ZSM-5 (5 wt.% of
3
2
3
that do not show X-ray diffraction. Nevertheless, XRD pattern of
Cr). Raman spectrum of CrO –ZSM-5 solid exhibited two bands
at 374 and 1000 cm , attributed to the symmetric vibrations
of monochromate species and Cr O vibrations of polychromates,
3
−
1
Cr–N-SE catalyst (Fig. 4c) reveals clear additional peaks at 2ꢁ = 34,
◦
3
6, 42, 51 and 56 , attributable to ␣-Cr O [13]. Based on ther-
2
3
mal analysis results, the decomposition of Cr nitrate led to Cr O3
respectively [21,24]. Nevertheless, Raman spectrum of Cr O –ZSM-
5 solid exhibited a band at 553 cm , ascribed to Cr O Cr
2
2 3
−
1
oxide which exhibited an agglomeration because of the potential
oxidizing power of nitrate ions.
vibrations of Cr O3 [21,24,25].
2
2
7
Al MAS NMR spectra of both Cr–P-SE catalysts and zeolite
Raman spectrum of Cr–N-SE catalyst exhibits two bands at 372
+
−1
support are shown in Fig. 5. NMR spectrum of NH4 –ZSM-5 zeo-
and 555 cm , attributable to Cr O vibrations of monochromate
lite shows a strong signal at ı ∼60 ppm which corresponds to
and octahedral Cr(III) species, respectively [21,24,26].
Fig. 7. Raman spectra of Cr–N-PM catalysts.

F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
137
Fig. 8. DRIFTS spectra of NH4+–ZSM-5 and chromium loaded catalysts.
Table 3
TPR results.
Such results are in good agreement with DRS ones. Nevertheless,
in the case of Cr–N-AE catalyst, two bands dominate the corre-
sponding Raman spectrum at 377 and 811 cm , evidencing the
presence of monochromate species [22,24].
−
1
◦
◦
◦
◦
Catalyst
T1 ( C)
T2 ( C)
T3 ( C)
T4 ( C)
H2/Cr
H2 amount
mmol H2/g)
(
Cr–A-SE
Cr–N-SE
Cr–A-AE
Cr–N-AE
294
281
–
361
350
–
419
–
–
460
460
471
436
0.44
0.29
1.68
1.73
0.26
0.12
0.21
0.03
3
.7. DRIFTS study
–
–
–
DRIFTS spectra of Cr–ZSM-5 catalysts and parent zeolite are
reported in Fig. 8.
The spectrum of zeolite support exhibits two bands. The first
evidencing that a part of Cr ions is exchanged with strong Brönsted
acid sites, while the other part is grafted on silanol groups.
In the DRIFTS spectrum of Cr–A-SE catalyst, the band situated
−1
one, situated at 3591 cm , is assigned to the vibration of strong
Brönsted acid sites [27,28], while the second band, situated at
−1
3
725 cm , is attributed to terminal silanol groups Si OH [29,30].
Exchanging chromium in aqueous phase (Cr–P-AE catalysts)
−1
at 3725 cm , attributable to terminal silanol groups, does not
undergo any significant change, suggesting that Si OH groups do
not participate significantly in the solid-state reaction. Neverthe-
−
1
involves a decrease of 3591 and 3725 cm
band intensities,
−
1
less, the 3591 cm band intensity decreases, evidencing that Cr
ions are exchanged with Brönsted acid sites, in agreement with
27
Al MAS NMR results which indicate the modification of Al atoms
environment.
+
Starting from different Cr precursors and H –ZSM-5 zeolite
(Si/Al = 15), similar results have also been reported [6].
3
.8. H -TPR study
2
TPR profiles of reference oxides (CrO3 and Cr O ) and Cr–P-PM
2
3
catalysts are shown in Figs. 9 and 10, respectively. TPR results are
compiled in Table 3.
TPR profile of Cr O oxide exhibits two reduction peaks at 288
2
3
◦
and 348 C. The first one is attributed to the reduction of Cr(VI)
present in the surface of chromia [6,31], while the second one cor-
responds to the reduction of Cr(III) to Cr(II) [6,31]. TPR profile of
Fig. 9. TPR profiles of reference oxides.
Fig. 10. TPR profiles of chromium loaded catalysts.

1
38
F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
Fig. 11. Ethylene conversion (ꢀ) and CH3CN selectivity (ꢁ) as a function of reaction temperature. Conditions: 0.1 g catalyst, 10% NH3, 10% C2H4, 10% O2, total flow rate:
3
1
00 cm /min.
◦
CrO3 oxide exhibits three reduction peaks at 289, 474 and 587 C,
which corresponds to the reduction of Cr(VI) to Cr(III) [32–34].
of Cr–P-AE catalysts, bridging hydroxyls groups have been replaced
by Cr ions after aqueous exchange. Such catalysts, which are less
acidic, contain only one type of cationic Cr(VI) species located at
the exchange sites of the zeolite. Weckhuysen et al. [19] reported
that dichromates or larger Cr(VI) clusters are more easily reduced
if they are located at the external surface of the zeolite.
TPR profile of Cr–A-SE catalyst exhibits reduction peaks at 294,
◦
3
61, 419 and 460 C. According to Ilieva et al., the peak situated
◦
at 294 C corresponds to the reduction of Cr(VI) species anchored
to the surface of chromia [6,31]. The reduction peak situated at
◦
3
61 C corresponds to the reduction of Cr(III) to Cr(II) [31], in agree-
ment with the TPR study performed with Cr O reference oxide
3.9. Catalytic activity
2
3
◦
(
Fig. 9). TPR peak at 419 C could be attributed to the reduction of Cr
◦
oxo-cations [6]. At relatively high temperature (460 C), the reduc-
tion of chromate species strongly bounded to the support occurs
Catalytic performances of prepared catalysts, in ethylene
ammoxidation (C H + O + NH → CH CN + 2H O), as a function of
2
4
2
3
3
2
[
19,32–34].
reaction temperature, are compiled in Table 4 and Figs. 11 and 12.
Data were collected under stationary conditions after a stabilization
period of 1 h.
TPR profile of Cr–N-SE catalyst exhibited three reduction peaks
◦
at 281, 350 and 460 C, attributable to the reduction of Cr(VI)
species anchored to the chromia surface [6,31], Cr(III) ions [31]
and chromate species strongly bounded to the zeolite [19,32–34],
respectively.
The zeolite support does not exhibit any activity in ethy-
lene ammoxidation. Furthermore, Cr–P-PM catalysts are relatively
TPR quantitative study performed with Cr–P-SE catalysts
(
Table 3) indicates that H /Cr molar ratio values are lower than
2
theoretical value required for a total reduction of Cr(VI) to Cr(III)
according to: 2CrO + 3H → Cr O + 3H O. Probably, a part of Cr
3
2
2
3
2
species seems to be in exchange sites and reduced with H2 accord-
ing to:
Cr OH]2 + H → 2[Cr(OH)]+ + H O
+
[
HO Cr
O
2
2
In this context, Weckhuysen et al. reported [19] that Cr species
were hardly reduced because of the residual acidity of Cr–MOR and
Cr–Y catalysts. According to DRIFTS study, the amount of bridging
hydroxyls might be higher on Cr–A-SE catalyst since a small amount
of OH groups have been replaced by Cr ions. Consequently, the dif-
ficult reduction of Cr(VI) species is indicated by the presence of
residual hydroxyls.
TPR profiles of Cr–P-AE catalysts exhibit unique zone of H con-
2
◦
sumption between 250 and 600 C, with a maximum at 471 and
36 C, respectively for Cr–A-AE and Cr–N-AE catalyst. In addition,
◦
4
H /Cr molar ratio values are close to the theoretical value required
for a total reduction of Cr(VI) to Cr(III). According to DRIFTS spectra
2
Fig. 12. CH CN yields as a function of reaction temperature. Conditions: 0.1 g cata-
lyst, 10% NH3, 10% C2H4, 10% O2, total flow rate: 100 cm /min.
3
3

F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
139
Table 4
Ethylene ammoxidation results.
Catalyst
Ethylene conversion (X) (%)
CH3CN selectivity (S) (%)
CH3CN Yield (%)
◦
Temp. ( C)
425
450
475
500
425
450
475
500
425
450
475
500
Cr–A-SE
Cr–N-SE
Cr–A-AE
Cr–N-AE
7
10
13.7
16.7
18
16.1
19.5
25.7
23.7
77
22
27
5
84
45.5
66
90
83
86
74
92.5
92.5
92
5.4
1.2
1.1
0
8.4
4.4
6.7
1.7
12.3
13.8
15.5
8.3
14.8
18
23.6
22.5
5.5
4. 2
0.1
9.7
10.2
5
11.2
35
95
3
Conditions: 0.1 g catalyst, 10% NH3, 10% C2H4, 10% O2, total flow rate: 100 cm /min.
active and selective toward acetonitrile and do not exhibit any sig-
nificant change in catalytic properties even at 500 C, after 10 h on
During this process, besides Brönsted acid sites, Cr ions were
anchored with silanols groups according to DRIFTS results. Dry-
◦
stream.
ing procedure results in a decrease of the ligand field strength of
◦
At 425 C, Cr–A-AE is a poor active catalyst in ammoxidation,
3+
Cr(H O)6 complex which migrates to cationic sites on the sur-
2
with both low C H4 conversion (4.2%) and acetonitrile selectivity
◦
2
face of the zeolite after a thermal treatment at 500 C. DRS and
(
27%), whereas Cr–A-SE catalyst exhibited, at the same tempera-
Raman studies indicate that monochromates, dichromate and/or
polychromate species were founded at the zeolite surface.
The difference between catalytic properties of low and over-
ture, slightly higher C H4 conversion (7%) and acetonitrile as main
2
reaction product (acetonitrile selectivity of 77%). On the other hand,
◦
catalysts issued from Cr nitrate exhibited, at 425 C, very low cat-
exchanged catalysts is obvious. At relatively low temperature
alytic activity. In fact, at 5.5% of ethylene conversion, Cr–N-SE
catalyst yielded only 1.2% of acetonitrile, while Cr–N-AE catalyst
is inactive.
◦
(
425 C), catalysts prepared in aqueous medium exhibited substan-
tially low activity. Since Cr(III) species were not detected in the
case of Cr–P-AE catalysts, it is more likely that these species are
actives at low temperatures. This idea confirms the agreement that
the improved catalytic properties of Cr–A-SE catalyst are mainly
related to the dispersed Cr(III) species. Characterization by DRS and
◦
At 450 C, Cr–A-AE catalyst exhibited 66% of selectivity toward
acetonitrile at 10.2% of ethylene conversion. Such catalytic prop-
erties are similar to those obtained in the presence of Cr–A-SE
catalyst, but they slightly exceed those of Cr–N-PM ones.
H -TPR is clearly in favor of the formation of such species. How-
2
◦
At 500 C, it is clear that Cr–P-AE catalysts yielded higher C H
2
4
ever, Cr–N-SE catalyst, having a higher concentration of Cr(III) sites,
consequently yielded CO2 as main product at low temperatures.
Probably, residual nitrate ions poisoned these sites for further CO2
formation. Similar results have been reported in our previous study
conversion and acetonitrile yields when compared to Cr–P-SE
solids.
[
6].
At 450 C, the catalysts issued from Cr nitrate exhibited low
4
. Discussion
◦
catalytic activity since the negative effect of nitrate ions was not
entirely prevented.
Solid-state exchange of Cr(III) ions into ZSM-5 has been studied
by the mean of thermal analysis coupled to MS. The mechanism
of such exchange depended on the nature of chromium precur-
sor. For example, because of the low thermal stability of Cr nitrate
As the reaction temperature increases, the prepared solids
improve their catalytic properties. Among them, catalysts issued
from aqueous exchange, which are very poor on chromium, exhib-
ited substantially higher ethylene conversions, nitrile yields and
selectivities. This behavior can be correlated with the nature of Cr
species which are, essentially, mono/di and polychromate species.
In the case of Cr–P-SE catalysts, besides Cr3+ ions and
mono/poly-chromates, Cr oxo-cations could be considered as active
+
salt, solid-state exchange with NH4 –ZSM-5 zeolite occurs at rel-
atively low temperatures. After decomposition of Cr nitrate, Cr
ions (present mostly in the form of Cr3+) balancing the negatively
charged ZSM-5 framework and occupy the exchange sites. The
production of the strong oxidizing agent NO2 caused these Cr3+
ions to be either oxidized to Cr(VI) species, or aggregated into
sites. In fact, the catalyst we reported earlier [6], e.g. issued
(
nano)crystals of Cr O3 (detectable by XRD).
When a mixture of Cr acetate and zeolite is heated, Cr ions
2
+
3+
from H –ZSM-5 (Si/Al = 15) and CrCl , was very active in ethylene
3
ammoxidation. We have attributed the improved catalytic proper-
ties of this catalyst, which is also poor on chromium, to the presence
of Cr oxo-cations at the exchange sites.
Nevertheless, higher concentration of Cr(III) oxide species
around the active sites thereby creating a diffusion block for
ethylene molecules, and inhibits the catalytic activity of Cr–P-SE
catalysts.
might have migrated either to the external surface of the zeolite
or continued to reside at the internal surface. During solid-state
reaction herein, volatile compounds such as acetic acid and water
were produced. The hydrate water of the mixture provides surface
hydroxyl groups that act as binder making Cr(III) crystals aggre-
gates. We believe that the external surface of the zeolite might
contain Cr(III) oxide as characterization techniques revealed. Rais-
ing of the temperature and the prolongation of solid-state reaction
time caused that a part of Cr(III) ions migrates inside the zeolite
channels to be exchanged with Brönsted acid sites and stabilized
5. Conclusion
in the form of (hydro)oxo-cations such as [HO Cr
mono and polychromate species.
O
Cr OH]2+
,
The exchanged chromium content and the nature of metallic
species introduced in the ZSM-5 zeolite depended on the catalysts
preparation method. In fact, upon aqueous exchange, low exchange
levels were reached. Sitting of chromium ions in exchange sites led
to the consumption of both Brönsted acid sites and silanol groups
as DRIFTS spectra have showed up, and the thermal treatment
led essentially to the formation of chromate species with differ-
ent degrees of polymerization as DRS and Raman results revealed.
The catalysts prepared in aqueous phase, using zeolite support
and 0.01 M solutions of Cr precursors exhibited low metal retention
because of diffusional restrictions. After ion exchange, hydrated
3+
Cr cation, such as Cr(H O)6 [19], could be present in the cavities
2
and/or channels of the zeolite. The following way can be proposed
19]:
[
Indeed, the reduction of different Cr ions with H was demonstrated
2
3+
+ n O Zeo → Cr3+ + −(O Zeo)n(H O)m + (6-m)H O
−
Cr(H O)
2
6
2
2
by H -TPR analysis.
2

1
40
F. Ayari et al. / Applied Catalysis A: General 415–416 (2012) 132–140
After solid-state reaction, higher exchange levels were reached.
[7] R.S. Da Cruz, A.J.S. Mascarenhas, H.M.C. Andrade, Appl. Catal. B 18 (1998)
23–231.
8] J.S. Olgen, R.S. Wyatt, J. Chem. Soc., Dalton Trans. 4 (1987) 859–865.
9] T. Komatsu, M. Nunokawa, I.S. Moon, T. Takahara, S. Namba, T. Yashima, J. Catal.
148 (1994) 427–437.
2
DRS, Raman, TPR and DRIFTS results showed that Cr species are
essentially oxo-cations, mono and/or polychromates, and Cr(III)
ions either in the zeolite exchange sites or located on the outer
[
[
[
[
10] H.W. Kouwenhoven, Adv. Chem. 121 (1973) 529–539.
11] G. Delahay, D. Valade, A.G. Vargas, B. Coq, Appl. Catal. B: Environ. 55 (2005)
surface as Cr O oxide (XRD and TEM results).
2
3
The prepared materials exhibited generally good catalytic prop-
erties in the studied reaction. Among them, those prepared from
aqueous exchange exhibited the highest activity. Improved cat-
alytic properties were not related to the level of metal exchanged,
as expected, but depended on chromium speciation. It seems that
Cr(VI) species play a key role in the ammoxidation of ethylene,
whereas, in the case of Cr–P-SE catalysts, higher concentration of
Cr(III) oxide species, around the active sites, inhibits the catalytic
activity.
1
49–155.
[12] A. Małecki, B. Małecka, R. Gajerski, S. Łabus, J. Therm. Anal. Calorim. 72 (2003)
135–144.
13] L. Li, Z.F. Yan, G.Q. Lu, Z.H. Zhu, J. Phys. Chem. B 110 (2006) 178–183.
14] A. Gervasini, Appl. Catal. A: Gen. 180 (1999) 71–82.
[15] M.M.J. Treacy, J.B. Higgins, Collection of Simulated XRD Powder Patterns for
Zeolites, 4th ed., 2001.
[
[
[
[
[
16] C.A. Fyfe, Y. Feng, H. Grondey, G.T. Kokotailo, H. Gies, Chem. Rev. 91 (1991)
525–1543.
1
17] B.M. Weckhuysen, L.M. De Ridder, R.J. Grobet, R.A. Schoonheydt, J. Phys. Chem.
99 (1995) 320–326.
18] P. Sarv, Ch. Fernadez, J.P. Amoureux, K. Keskinen, J. Phys. Chem. B 100 (1996)
1
9223–19226.
[19] B.M. Weckhuysen, H.J. Spooren, R.A. Schoonheydt, Zeolites 14 (1994) 450–457.
Acknowledgments
[
20] B.M. Weckhuysen, P. Vand Der Voort, G. Catana, Spectroscopy of Transition
Metal Ions on Surface, Leuven University Press, 2000, p. 243.
The authors thank Pr. E.M. Gaigneaux and P. Eloy (Université
catholique de Louvain) for Raman and DRIFTS analysis.
[21] B.M. Weckhuysen, I.E. Wachs, R. Schoonheydt, Chem. Rev. 96 (1996)
3327–3349.
[
22] B.M. Weckhuysen, L.M. De Ridder, R.A. Schoonheydt, J. Phys. Chem. 97 (1993)
756–4763.
4
References
[23] B.M. Weckhuysen, A.A. Verberckmoes, A.L. Buttiens, R.A. Schoonheydt, J. Phys.
Chem. 98 (1994) 579–584.
[
[
[
[
[
[
1] M. Mhamdi, S. Khaddar-Zine, A. Ghorbel, Stud. Surf. Sci. Catal. 142 (A) (2002)
[24] M.A. Vuurman, I.E. Wachs, J. Phys. Chem. 96 (1992) 5008–5016.
[25] L. Zhang, Y. Zhao, H. Dai, H. He, C.T. Au, Catal. Today 131 (2008) 42–54.
[26] M. Cherian, M.S. Rao, A.M. Hirt, I.E. Wachs, G. Deo, J. Catal. 211 (2002) 482–495.
[27] J. Ward, J. Catal. 9 (1967) 225–236.
[28] A.W. Chesr, R.M. Dessarn, L.B. Alemani, G.J. Woolevy, Zeolites 6 (1986) 14–16.
[29] U. Lhose, E. Löffler, M. Hunger, J. Stöckner, V. Patzelove, Zeolites 7 (1987) 11–13.
[30] A. Jentys, G. Rumplmayr, J.A. Lercher, J. Appl. Catal. 53 (1989) 299–312.
[31] L.I. Ilieva, D.H. Andreeva, Thermochim. Acta 265 (1995) 223–231.
[32] K. Jó z´ wiak, W. Ignaczak, D. Dominiac, T.P. Maniecki, Appl. Catal. A: Gen. 258
(2004) 33–45.
9
35–942.
2] M. Mhamdi, S. Khaddar-Zine, A. Ghorbel, E. Marceau, M. Che, Y. Ben Taarit, F.
Villain, Stud. Surf. Sci. Catal. 158 (2005) 749–756.
3] M. Mhamdi, S. Khaddar-Zine, A. Ghorbel, React. Kinet. Catal. Lett. 88 (2006)
1
49–156.
4] M. Mhamdi, S. Khaddar-Zine, A. Ghorbel, Appl. Catal. A: Gen. 357 (1) (2009)
2–50.
5] F. Ayari, M. Mhamdi, G. Delahay, A. Ghorbel, J. Sol–Gel Sci. Technol. 49 (2) (2009)
70–179.
4
1
6] F. Ayari, M. Mhamdi, D.P. Debecker, E.M. Gaigneaux, J. Álvarez-Rodríguez, A.R.
Guerrero- Ruiz, G. Delahay, A. Ghorbel, J. Mol. Catal. A: Chem. 339 (2011) 8–16.
[33] B.L. Chamberland, Solid State Mater. Sci. 7 (1997) 1–31.
[34] N.E. Fouad, J. Therm. Anal. 31 (1986) 825–834.