Catalytic selective oxidation of isobutane over Csx(NH 4)3-xHPMo11VO40 mixed salts

Article abstract of DOI:10.1039/c4cy00518j

A series of mixed Keggin-type heteropolysalts Cs_x(NH ₄)_3-xHPMo₁₁VO₄₀ with various ammonia/caesium ratios was prepared by the precipitation method and characterized by TGA, N₂ adsorption/desorption, XRD, FT-IR, and NH₃-TPD techniques. Correlations between the ammonia/caesium ratio and the specific surface area, as well as with the total number of acid sites, were established. Furthermore, the introduction of Cs to the catalytic formulation was beneficial to the stabilization of the Keggin structure and helped limiting the elimination of the V atoms from the primary structure. The as-prepared samples were applied in the catalytic selective oxidation of isobutane at 340 °C under atmospheric pressure. The best results were obtained over Cs_1.7(NH₄)_1.3HPMo ₁₁VO₄₀ with an isobutane conversion of 9.6% and a total selectivity to valuable products (methacrylic acid and methacrolein) of 57.1%. This was explained by the well-balanced acidity and specific surface of this catalyst, promoting the C-H bond activation (adequate acid properties) over a large number of accessible active sites (high acid sites density). the Partner Organisations 2014.

Full text of DOI:10.1039/c4cy00518j

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Catalytic selective oxidation of isobutane over
Csx(NH4)3−xHPMo11VO40 mixed salts†
Cite this: Catal. Sci. Technol., 2014,
4, 2938
Fangli Jing,ab Benjamin Katryniok,abc Franck Dumeignil,abde Elisabeth Bordes-Richardabf
abc
*
and Sébastien Paul
A series of mixed Keggin-type heteropolysalts Csx(NH4)3−xHPMo11VO40 with various ammonia/caesium
ratios was prepared by the precipitation method and characterized by TGA, N2 adsorption/desorption,
XRD, FT-IR, and NH3-TPD techniques. Correlations between the ammonia/caesium ratio and the specific
surface area, as well as with the total number of acid sites, were established. Furthermore, the intro-
duction of Cs to the catalytic formulation was beneficial to the stabilization of the Keggin structure and
helped limiting the elimination of the V atoms from the primary structure. The as-prepared samples were
applied in the catalytic selective oxidation of isobutane at 340 °C under atmospheric pressure. The best
results were obtained over Cs1.7(NH4)1.3HPMo11VO40 with an isobutane conversion of 9.6% and a total
selectivity to valuable products (methacrylic acid and methacrolein) of 57.1%. This was explained by the
well-balanced acidity and specific surface of this catalyst, promoting the C–H bond activation (adequate
acid properties) over a large number of accessible active sites (high acid sites density).
Received 21st April 2014,
Accepted 14th May 2014
DOI: 10.1039/c4cy00518j
www.rsc.org/catalysis
reaction. Hereby, the acidity plays an important role in acti-
vating the C–H bound, and the redox properties are responsi-
1. Introduction
The development of an efficient catalyst for the selective oxi-
dation of isobutane would provide a very interesting shortcut
for the production of methacrolein (MAC) and methacrylic
acid (MAA), which could be further used in the synthesis of
polymethyl methacrylate polymer (PMMA). The conventional
process for producing PMMA is based on the acetone-
cyanohydrin (ACH) route,1 which has two main drawbacks:
the use of a highly toxic raw material – namely, hydrogen
cyanide – and the production of large amounts of contami-
nated by-products (mainly ammonium bisulphate). In com-
parison with the traditional techniques, the advantages of
the direct catalytic isobutane selective oxidation are obvious:
simplicity of the process, low cost of the raw materials, environ-
mental friendliness and, last but not least, less by-products
generation. Bifunctional catalysts possessing simultaneously
acid and redox properties are necessary for catalysing such a
ble for the oxygen insertion in the hydrocarbon skeleton to
generate the carboxylic acid function.
In a previous study, we demonstrated that (NH4)3HPMo11VO40
(APMV) supported on Cs3PMo12O40 (CPM) enables high catalytic
performance in the selective oxidation of isobutane.2,3 Thus
came the idea of studying the Cs+/NH4 mixed salts of 1-vanado-
+
11-molybdophosphoric acid for the selective oxidation of isobu-
tane. Various studies on (mixed) salts of heteropolycompounds
such as, for instance, (NH4)3PW12O40,4 CsxH3−xPW12O40,5
Csx(NH4)3−xPW12O40 (ref. 6) or Csx(NH4)3−xPM12O40 (ref. 7,8)
have already been published, but these solids were not
used as catalysts for the isobutane selective oxidation to
methacrolein and/or methacrylic acid. Herein, we report the
preparation and the application of Cs+/NH4 mixed acid
+
salts of 1-vanado-11-molybdophosphoric corresponding to
the formula Csx(NH4)3−xHPMo11VO40, and using various ratios
between caesium and ammonium cations with x = 3, 2.5, 1.7,
1.5, 0.5 and 0. Next to the influence of the counter-cations
ratio on the catalytic performances, the effect of the composi-
tion on the structural, the textural and the acidic properties of
the solids was also studied.
a Univ. Lille Nord de France, F-59000, Lille, France
b Unité de Catalyse et de Chimie du Solide, UCCS (UMR CNRS 8181), Cité
Scientifique, F-59650, Villeneuve d'Ascq, France
c Ecole Centrale de Lille, ECLille, F-59655 Villeneuve d'Ascq, France.
E-mail: sebastien.paul@ec-lille.fr; Fax: +33 3 20 33 54 99; Tel: +33 3 20 33 54 57
d Université Lille 1 Sciences et Technologies, USTL, F-59652 Villeneuve d'Ascq, France
e Institut Universitaire de France, Maison des Universités, 103 Bd St-Michel,
Paris, 75005, France
f Ecole Nationale Supérieure de Chimie de Lille, ENSCL, F-59659
Villeneuve d'Ascq, France
2. Experimental
2.1 Catalysts preparation
The synthesis of the parent heteropolyacid (H4PMo11VO40)
was performed as reported elsewhere.9 The mixed salts
† Electronic supplementary information (ESI) available: Raman spectra, XRD
diffractograms, N2 adsorption–desorption, NH3-TPD. See DOI: 10.1039/c4cy00518j
2938 | Catal. Sci. Technol., 2014, 4, 2938–2945
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Csx(NH4)3−xHPMo11VO40 (x = 3, 2.5, 1.7, 1.5, 0.5 and 0) were
prepared by a precipitation method. In a typical synthesis
procedure, the appropriate amount of H4PMo11VO40 acid was
dissolved under stirring in deionized water to obtain a
0.05 M solution at 45 °C. Then, solutions of Cs2CO3 (0.05 M)
and of NH4NO3 (0.05 M) were added using a pump with a
constant flowrate of 2 mL min−1 under vigorous stirring.
The reaction mixture was kept as such for 18 h for ageing,
and, then, the so-obtained precipitate was recovered by evap-
orating the solvent at 70 °C under reduced pressure. After-
wards, the solid was dried at 70 °C in an oven for another
24 h. The final catalysts were obtained after thermal treat-
ment at 350 °C for 3 h under static air to decompose the
nitrate and carbonate precursors in order to obtain the mixed
salts. The samples were denoted as Csx(NH4)3−xH (x = 3, 2.5,
1.7, 1.5, 0.5 and 0) according to the theoretical catalyst com-
position. In the case of x = 0, the sample corresponds to the
ammonium acidic salt of the phospho-vanado-molybdic acid,
which is denoted APMV in the following.
saturation (evaluated from the thermal conductivity detector –
TCD – signal). The TPD profiles were monitored by a TCD and
recorded from 130 to 700 °C at the heating rate of 10 °C min−1.
The effluent gas was identified by a mass spectrometer.
2.3 Catalytic selective oxidation of isobutane
The catalytic performances in the selective oxidation of isobu-
tane to methacrolein and methacrylic acid were evaluated in
a fixed-bed reactor at 340 °C under atmospheric pressure.
0.8 g of catalyst diluted in 0.8–1.6 g SiC (0.21 mm particle
size) were loaded in the middle of the reactor. The reactants
[27% isobutane, 13.5% O2, 10% H2O and 49.5% He (mol.%)]
were fed at a constant total flow-rate of 10 mL min−1. All the
products – except water – were analysed by online gas
chromatography (GC). The incondensable compounds such
as CO, CO2, O2 and isobutane (IBAN) were analysed by a GC
equipped with a TCD and 2 packed columns in series
(Porapak Q, 3.2 mm diameter, 3 m length; molecular sieve
13X, 3.2 mm diameter, 3 m length). The less volatile products
like methacrolein (MAC), methacrylic acid (MAA), acetic acid
(AA) and acrylic acid (ACA) were analysed by a GC equipped
with a flame ionization detector (FID) and an Alltech EC1000
semi-capillary column (30 m length, 0.53 mm diameter,
1.2 μm film width). The reaction data were obtained after
18 h under stream, which corresponds to steady state condi-
tions (typically reached within 2 h).
2.2 Catalysts characterization
Thermogravimetric analyses (TGA) were performed on fresh
catalysts using a TA instruments (Model 2960 SDT, V3.0F).
The catalysts were heated from room temperature to 700 °C
with an increasing rate of 3 °C min−1 under air flow.
The textural properties of the calcined catalysts were deter-
mined by the N2 physisorption/desorption technique, using a
Micromeritics ASAP 2010 analyser at the liquid nitrogen tem-
perature. The specific surface area (SBET) was calculated using
the BET method.10 The total pore volume (Vp) was deter-
mined using the value observed at the relative pressure (P/P0)
of 0.995. The samples were outgassed at 150 °C for 3 h prior
to analysis.
Infrared spectra (FT-IR) were recorded on a Thermo Nicolet
480 equipped with a MCT detector. The samples (1 wt.%) were
pressed to form KBr pellets for analysis, and the spectra were
recorded from 400 to 4000 cm−1 with a spectral resolution of
4 cm−1 and 256 scans.
The crystalline structure of the catalysts was resolved by
X-ray diffraction (XRD) technique on a Bruker D8 Advance
diffractometer, using the CuKα radiation (λ = 1.5506 Å) as
X-ray source. A range from 10° < 2θ < 80° was scanned
with a step of 0.02° s−1 and 2 s of acquisition time. The crys-
tallite size was estimated by the Debye–Scherrer's equation,
namely L = 0.89λ/β(θ)cos θ, where L is the crystallite size,
λ is the wavelength of the radiation used, θ is the Bragg
diffraction angle and β(θ) is the full width at half maxi-
mum (FWHM).11,12
Ammonia temperature-programmed desorption (NH3-TPD)
was employed in order to determine the amount and the
strength of the acid sites over the solids. A typical acquisition
was carried out as follows: 50 mg of catalyst were pre-treated
in a He flow (30 mL min−1) at 250 °C for 2 h in order to
remove the physisorbed and the crystal water. Then, NH3 was
absorbed at the surface by pulsed injections at 130 °C until
3. Results and discussion
3.1 Thermal stability
+
The thermal analyses of two samples, one without NH4
+
cations (x = 3) and another one with NH4 cations (x = 1.7,
taken as an example) were carried out under air flow to study
the structural evolution during the calcination (Fig. 1). A con-
tinuous weight loss was observed from the TG curves (solid
lines) from the starting temperature to 450 °C for both
samples. A more precise picture can be obtained from the
DTG curves (dash lines). The initial stage between 100–205 °C
was ascribed to the removal of physisorbed and crystal water.
Afterwards, a broad peak between 175 °C and 500 °C
for the Cs3(NH4)0H sample was assigned to the loss of one
proton as constitution water.13 The calculated weight loss for
this step is 0.41%, which is close to the theoretical value
(0.48%). In the case of Cs1.7(NH4)1.3H, a multi-step decompo-
sition between 205 °C and 430 °C was observed. As already
reported,2 the total removal of ammonium cations consists of
2 steps: the release of NH3 to give the residual acidic protons,
and subsequent loss of the residual protons as constitution
water. The weight loss of 1.86% observed on the TG curve
was in rather good agreement with the theoretical value
(2.04%), which suggests almost complete loss of proton and
ammonium cations. For both samples, Keggin type com-
pounds with oxygen vacancies were generated after the elimi-
nation of constitutional water as it is well-known from the
literature.14 Finally, the Keggin structure collapsed to form
the simple P2O5, V2O5 and MoO3 oxides at 630 °C for the
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2938–2945 | 2939

View Article Online
Paper
Catalysis Science & Technology
996 cm−1, which are considered as characteristic of the sub-
stitution of one Mo atom by one V atom in the Keggin anion
3− 18
PMo12O40
.
In fact, the exchange of one Mo by one V
reduces the symmetry, leading to a weak distortion of the
PO4 tetrahedron. A third shoulder at 1045 cm−1 also present
on both samples is assigned to the asymmetric stretching
mode of P–O. The other samples with lower Cs contents
(Fig. 2c–f) do not exhibit these three shoulders observed for
Cs3(NH4)0H and Cs2.5(NH4)0.5H. Instead, a weak peak appears
at 1034 cm−1, which is assigned to the v VO band of
V2O5.19 The characteristic band of N–H can be observed at
+
1415 cm−1 in the samples containing NH4 (Fig. 2b–f) except
+
for Cs2.5(NH4)0.5H, in which the NH4 may have been
removed as NH3 during the thermal treatment. The band at
lower wavenumber of 595 cm−1, observed on all the samples,
can be attributed to MoO3,20 while the band at 1384 cm−1,
observed on all the N-containing samples, is related to the
Fig. 1 Thermogravimetric (TG, solid lines) and derivative thermogravimetric
analysis (DTG, dash lines) [a: Cs3(NH4)0H; b: Cs1.7(NH4)1.3H].
− 7
presence of residual NO3 .
Cs1.7(NH4)1.3H sample, and at 654 °C for the Cs3(NH4)0H
sample. Simultaneously, the P2O5 began to sublimate,
resulting in further weight loss. From these results, it can be
deduced that the active phase can partially release NH3 at the
studied reaction temperature (namely 340 °C), leading to a
protonated heteropolycompound species.
These results indicate that the structure of the synthesized
solids significantly depends on the number of Cs atoms. For
a high Cs loading (namely for x = 3 and 2.5), the structural
features due to V-substituted Keggin anions are still clearly
visible. When the Cs atoms number progressively decreases
to finally reach zero for APMV, the aforementioned structural
features concomitantly disappear to the profit of new features
from V2O5, suggesting the elimination of V from the primary
structure. At the same time, the detection of MoO3 by IR
suggests a partial decomposition during calcination. This
latter interpretation was also confirmed by Raman analysis
(Fig. S1†).
Except for Cs3(NH4)0H the FT-IR spectra of the spent cata-
lysts exhibit the nearly the same features as those of the fresh
samples but an additional band at 629 cm−1 attributed to
MoO2 (Fig. 3; S2† for full-scale spectra). It was explained by
the fact that the MoO3 species issued from the partial decom-
position of Keggin anions (band at 595 cm−1) were partially
reduced to MoO2 by isobutane under reaction conditions.7,21
Since the intensity of the 629 cm−1 band increased with the
decrease in the Cs loading, this supports the hypothesis that
the introduction of Cs stabilizes the Keggin structure under
the reaction conditions.
3.2 Structural features
The structural features of the catalysts were determined by
FT-IR spectroscopy and X-ray diffraction. The FT-IR spectra of
the calcined catalysts are reported in Fig. 2. The vibration
bands at 1061, 963, 865 and 788 cm−1 are attributed to
the asymmetric stretching of heteroatom-oxygen (vas P–O),
the peripheral metal atoms–terminal oxygen (vas MOt),
the stretching mode of inter- and intra-octahedral bridges
(vas M–Ob–M and vas M–Oc–M), respectively.15–17 In addition,
for the Cs3(NH4)0H (Fig. 2a) and Cs2.5(NH4)0.5H samples
(Fig. 2b), two shoulders can be clearly observed at 1080 and
X-Ray diffraction was used to determine the crystalline
phases present in the Csx(NH4)3−xH (x = 0–3) mixed salts
samples. The patterns of the calcined and used catalysts are
presented in Fig. 4 and S3,† respectively. All the samples
exhibit three strong diffraction peaks at 2θ = 26.4°, 10.6° and
36.0°, which are characteristic of the Keggin structure.22
When introducing Cs into the catalysts, the patterns show
the trend to be shifted to lower angles, which is characteristic
of an increase in the d-spacing according to the Bragg's law.
+
In fact, when NH4 cations are partially substituted by Cs+
cations, the inter-planar spacing is expanded, simply because
+
Cs+ has a larger ionic radius than NH4 . Thus, the d-spacing
of plane (222) for the calcined samples increased from 0.3377 nm
Fig.
2 FT-IR spectra of the calcined samples [a: Cs3(NH4)0H,
for the Cs-free APMV sample to 0.3416 nm for the Cs2.5(NH4)0.5
sample (Table 1). On the other hand, a slight decrease in the
H
b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H, d: Cs1.5(NH4)1.5H, e: Cs0.5(NH4)2.5H,
f: APMV].
2940 | Catal. Sci. Technol., 2014, 4, 2938–2945
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
In order to gain more information about the structural
evolution of the catalysts during the reaction, the cell param-
eters a of the cubic phases were also calculated. We selected
six diffraction planes including (110), (211), (310), (222),
(400) and (332), which give the characteristic XRD lines of
the Keggin type heteropolycompounds. The relationship
between the cell parameter a and the Cs loading, represented
by ‘x’ in the Csx(NH4)3−xHPMo11VO40 formula, is reported in
Fig. 5. For the samples before reaction, a linear relationship
between a and x is observed, meaning that the partial substi-
tution of ammonia by caesium cations obeys the Vegard's law.
The cell parameter a changes irregularly after reaction,
reflecting
a complex structural evolution of the hetero-
polycompounds under the reaction conditions. However, a linear
relationship is obtained for the Cs contained catalysts, and the
observed evolution of the a parameter between the fresh and
the used samples depends on the Cs loading. The catalyst with
the lower Cs content [namely Cs0.5(NH4)2.5H] shows an increase
of the a value between the fresh and the spent sample (11.76 Å
vs. 11.74 Å). In contrast, almost no change is observed for the
Cs1.5(NH4)1.5H sample, with a a value around 11.79 Å before
and after reaction. In contrast, for the samples with the higher
contents in Cs such as Cs1.7(NH4)1.3H, Cs2.5(NH4)0.5H and
Cs3(NH4)0H, a is clearly lowered after reaction. It seems that
these samples evolve towards the stable Cs4PMo11VO40 phase,
which exhibits a a value of 11.75 Å (PDF#46-0481), but probably
under a protonated form. It can be concluded combining TGA
and FT-IR results that: i) the successive loss of ammonium
cations and residual protons as NH3 and water may take place
under the reaction conditions; ii) the introduction of Cs in
the catalysts stabilizes the V-containing Keggin structure, thus
protecting V against the elimination from the primary structure.
To this respect, we can remark that the a cell parameter observed
for Cs0.5(NH4)2.5H is much closer to that of Cs4PMo11VO40
than the other Cs-containing samples, as it exhibits a weaker
ability to stabilize the structure. For the APMV Cs-free
sample, the value decreases to 11.68 Å after test, which is
quite close to that of (NH4)3PMo12O40 (a = 11.67 Å, PDF#43-0315).
This evidences the fact that in absence of caesium the
V-containing Keggin structure cannot be stabilized.
Fig.
3 FT-IR spectra of the spent catalysts [a: Cs3(NH4)0H,
b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H, d: Cs1.5(NH4)1.5H, e: Cs0.5(NH4)2.5H,
f: APMV].
Additionally, the influence of the Cs introduction on the
crystallites size was also investigated. The crystallite size of the
catalysts was estimated using the Debye–Shererr's equation
from the values of the FWHM of the (110), (222) and (332) dif-
fraction planes.5,12 The APMV sample showed the largest crys-
tallite size (>100 nm) under the calcined and the spent form.
The introduction of caesium in the catalyst favoured the forma-
tion of crystallites with a smaller size, which is in agreement
with the textural properties of the solids (vide infra). Further-
more, it was found that the crystallite size slightly increased
after reaction, thus suggesting the sintering of the catalyst.
Fig.
4 XRD patterns of the different catalysts before reaction
[a: Cs3(NH4)0H, b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H, d: Cs1.5(NH4)1.5H,
e: Cs0.5(NH4)2.5H, f: APMV]. The diffraction line observed at 18° is due
to the Teflon holder.
d-spacing could be observed for the fully substituted sample
Cs3(NH4)0H. It is important to note that two diffraction peaks
at 2θ = 15.1° and 21.4° only appear for the catalysts
containing NH4 . The intensities of these peaks weaken as
the number of NH4 cations decreases, and they are
+
+
ultimately not detected at all over the Cs3(NH4)0H sample.
Therefore, these two diffraction lines could be considered as
characteristic of the presence of NH4 cations-containing
3.3 Textural properties of the calcined samples
+
The textural properties of the calcined samples including the
polyoxometalates.
specific surface area (SBET), the total porous volume (V), and
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2938–2945 | 2941

View Article Online
Paper
Catalysis Science & Technology
Table 1 Average crystallite size & related parameters of the samples, as determined using XRD patterns
Before reaction
Sample
2θ for hkl (222)
d(222), nm
Average crystallite size, nm
Cs0(NH4)3H (APMV)
26.37
26.29
26.18
26.12
26.06
26.10
0.3377
0.3387
0.3400
0.3408
0.3416
0.3411
>100 (>100)a
45.7 (47.2)
39.4 (42.2)
38.8 (42.3)
36.9 (44.9)
41.9 (46.4)
Cs0.5(NH4)2.5
Cs1.5(NH4)1.5
Cs1.7(NH4)1.3
Cs2.5(NH4)0.5
Cs3(NH4)0H
H
H
H
H
The values d(222) for (NH4)3PMo12O40 (PDF#43-0315), Cs4PMo11VO40 (PDF#46-0481) and H4PMo11VO40 (PDF#52-1117) are 0.3370, 0.3393 and
0.3377 nm, respectively.a The values in parenthesis correspond to the spent catalysts.
in Fig. 6 (circles). The specific surface area of the catalyst
increases linearly as the Cs number increases from 1.5 to 3.
On the other hand, when the Cs number is as low as 0.5, the
correlation does no longer fit. However, the value is still
higher than that of the Cs-free sample, namely Cs0(NH4)3H
(APMV). These results show that it is possible to increase
the specific surface area of the solids by introducing some
caesium cations in the catalysts formulation.
On the other hand, an inverse linear relationship is found
between the average pore diameter (square symbols in Fig. 6)
and the caesium content. Nevertheless, the BJH model evi-
dences a rather broad pore size distribution (Fig. S4†). Only
in the case of Cs1.5(NH4)1.5H and Cs1.7(NH4)1.3H, a discrete
distribution is observed with a maximum at 4 nm. This
mesoporous feature is further underlined by the isotherms
corresponding to an IUPAC type IV with a H3 hysteresis
(Fig. S5†). On the other hand, no clear picture is obtained
from the evolution of the total pore volume with the Cs
content.
Fig. 5 Correlation between the cell parameter a and the x number of
Cs atoms in the Csx(NH4)3−xHPMo11VO40 catalysts. The cell parameter
a was calculated using the following formula: a = d(hkl)·√(h2 + k2 + l2).
the average pore diameter (D) are reported in Table 2. The
+
textural properties are strongly depending on the Cs+/NH4
ratio. The APMV sample – without Cs cations – shows a low
SBET (3.5 m2 g−1), which suggests the formation of a bulk,
non-porous-like material. Upon progressive replacement of
+
the NH4 cations by Cs+ cations, the surface area gradually
increases. When the number of Cs cations (x) in the catalyst
increases from 0.5 to 1.7, the specific surface area increases
from 5.8 to 16.4 m2 g−1. Then, the surface area reaches the
maximum value of 87.6 m2 g−1 when all NH4 cations are
+
substituted by Cs+ cations [Cs3(NH4)0H]. The correlation
between the Cs content and the surface area is represented
Table 2 Textural properties of the calcined mixed salts samples
x
Sample
SBET, m2 g−1
V, cm3 g−1
D, nm
0
Cs0(NH4)3H (APMV)
3.5
5.8
8.4
16.4
58.6
87.6
0.020
0.014
0.018
0.033
0.040
0.035
23.1
11.1
7.1
6.5
4.2
0.5
1.5
1.7
2.5
3.0
Cs0.5(NH4)2.5
Cs1.5(NH4)1.5
Cs1.7(NH4)1.3
Cs2.5(NH4)0.5
Cs3(NH4)0H
H
H
H
H
Fig. 6 Plots of surface area (circles) and pore diameter (squares) as a
function of the Cs loading expressed as x in the Csx(NH4)3−xHPMo11VO40
calcined samples [a: Cs3(NH4)0H, b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H,
d: Cs1.5(NH4)1.5H, e: Cs0.5(NH4)2.5H, f: APMV].
4.5
2942 | Catal. Sci. Technol., 2014, 4, 2938–2945
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
3.4 Acid properties (NH3-TPD)
The selective oxidation of isobutane to methacrolein and
methacrylic acid is significantly impacted by the nature and
the strength of the acid sites, which have to activate the C–H
bond, promoting the conversion of MAC to MAA and driving
the desorption of the formed acid products like MAA from
the catalysts surface.23–25 Therefore, it is worth investigating
the acidic properties of the prepared catalysts. The NH3-TPD
profiles were collected between 130 to 700 °C (Fig. S6†) and
the amount of acidic sites was quantified by deconvolution/
integration of the profiles (Table 3 and Fig. S8†). The amount
of desorbed ammonia was corrected by the amount of ammo-
nia yielding from the decomposition of the ammonium
cations. The detailed method used is described elsewhere.2
Three types of acidic sites with different strengths were
defined, corresponding to weak sites in the 130–300 °C
range, medium sites in the 300–450 °C range and strong
acidic sites in the 450–560 °C range, respectively. Strong
acidic sites are only observed for the catalyst having the
lowest Cs content [Cs0.5(NH4)2.5H] (1.16 mmol gcat−1), this
sample being also the catalyst exhibiting the largest total
amount of acid sites of the series (1.84 mmol gcat.−1). The
other samples only exhibit medium sites together with a very
small amount of weak sites.
Fig.
7
Plot of the NH3 uptake as
a
function of the Cs loading
expressed as x in the Csx(NH4)3−xHPMo11VO40 formula [a: Cs3(NH4)0H,
b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H, d: Cs1.5(NH4)1.5H, e: Cs0.5(NH4)2.5H].
However the highest yield in MAC + MAA (5.5%) was
obtained over the Cs1.7(NH4)1.3H sample, corresponding to
an isobutane conversion of 9.6% and an oxygen conversion
of 52.7%. Cs1.7(NH4)1.3H and Cs2.5(NH4)0.5H catalysts showed
similar activity in isobutane oxidation, but different product
The amount of acid sites per mass of catalyst linearly
increased when x decreased (Fig. 7). As the specific surface
area also increased when x decreased (Table 2), the highest
surface acidic sites density (317.55 μmol mcat.−2) was accord-
ingly observed over the Cs0.5(NH4)2.5H catalyst.
distributions were observed. In fact, whereas Cs1.7(NH4)1.3
H
exhibited a higher selectivity to both MAC (13.7%) and MAA
(43.4%), and relatively low selectivity to COx (25.6%), Cs2.5(NH4)0.5H
exhibited a very high selectivity to COx (53.1%). By comparing
3.5 Catalytic performance in the reaction of isobutane
partial oxidation
the converted reactants, it was noteworthy that Cs2.5(NH4)0.5
H
The prepared catalysts were used to catalyse the selective
oxidation of isobutane to MAC and MAA. The results are
listed in Table 4. The Cs-free APMV exhibited rather low con-
versions in both oxygen and isobutane (10.4% and 2.0%,
respectively). The introduction of Cs in the catalysts formulas
dramatically affected both the conversion of reactants and
the products distribution. Over the catalyst with the lowest
Cs content, namely Cs0.5(NH4)2.5H, the isobutane conversion
is 4.1%, more than 2 times as high as that obtained over
APMV, and the yield of the desired products (MAC + MAA)
was more than tripled. With the increase in x up to 2.5, the
maximum conversion of isobutane (9.9%) was achieved, and,
simultaneously, full conversion of oxygen was observed.
shows very high ratio of oxygen/isobutane conversion (10/1)
compared to the other samples (ca. 5/1). We shown in a
previous study26 that consecutive combustion of the desired
products mainly occurs in the case of MAC, whereas MAA
is rather stable under the reaction conditions. In terms of
competitive reaction to convert MAC to MAA or to COx,
Cs2.5(NH4)0.5H exhibited a greater ability in promoting the
degradation reaction instead of the selective oxidation to
MAA. The Cs3(NH4)0H sample gave the lowest activity as a
consequence of the low amount of acid sites, resulting in
difficulty for activating the C–H bond of isobutane.
Finally, the influence of the acidity on the catalytic activity
is reported in Fig. 8. The sample Cs3(NH4)0H having the
Table 3 Amount of acidic sites and strength distribution over the samples, determined from the NH3-TPD profiles
−1
Acidity, NH3 uptake, mmol gcat.
−1
Total acidit−y2mmol gcat.
Catalyst
130–300 weak
300–450 medium
450–560 strong
(μmol mcat. )
Cs0.5(NH4)2.5
Cs1.5(NH4)1.5
Cs1.7(NH4)1.3
Cs2.5(NH4)0.5
Cs3(NH4)0H
H
H
H
H
0.00
0.01
0.02
0.07
0.07
0.68
0.89
0.71
0.23
0.10
1.16
—
—
—
—
1.84
0.90
0.72
0.30
0.17
(317.55)
(107.03)
(44.16)
(5.06)
(1.94)
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2938–2945 | 2943

View Article Online
Paper
Catalysis Science & Technology
Table 4 Catalytic performances in IBAN selective oxidationa
Conversion, %
O2 IBAN
10.4
17.1
29.2
52.7
100.0
6.2
Selectivity, %
Carbon
balance
Catalyst
MAC
AA
ACA
MAA
COx
Y(MAC+MAA), %
Cs0(NH4)3H (APMV)
2.0
4.1
5.8
9.6
9.9
1.2
11.5
29.3
22.1
13.7
9.7
11.9
9.3
10.7
13.4
15.5
19.9
6.2
3.1
3.3
3.9
2.6
9.0
32.9
41.5
44.8
43.4
19.1
31.9
37.5
16.8
19.1
25.6
53.1
23.0
0.9
2.9
3.9
5.5
2.9
0.6
0.992
0.995
0.997
0.997
0.998
0.996
Cs0.5(NH4)2.5
Cs1.5(NH4)1.5
Cs1.7(NH4)1.3
Cs2.5(NH4)0.5
Cs3(NH4)0H
H
H
H
H
16.2
a
Reaction conditions: temperature = 340 °C, atmospheric pressure, contact time = 4.8 s, time on stream = 18 h. Here IBAN = isobutane,
MAC = methacrolein, AA = acetic acid, ACA = acrylic acid and MAA = methacrylic acid.
previous work,3 and their relationship is depicted in Fig. 9. A
pronounced increase in activity from 0.02 to 0.25 μmol
mcat. min−1 can be seen as x decreases from 3 to 2.5, corre-
−2
sponding to an increase in surface acid density from 1.9 to
−2
5.1 μmol mcat. (points labelled ‘a’ and ‘b’ on the figure).
−2
Then, the activity continues to increase to 0.88 μmol mcat.
min−1 when x further decreases to 1.7, although the slope
becomes a little bit less pronounced. Further, the activity still
increases, but with a quite slow slope until it reaches a
plateau around 1.07 μmol mcat. min−1, for the low x value
−2
0.5 with a surface acid density of 317.6 μmol mcat.−2. The
sharp increase in activity from Cs3(NH4)0H to Cs2.5(NH4)0.5
H
indicates the importance of the acidic sites in the activation
step. However, it is not the unique important factor that
governs the catalytic performances, since the quite low
surface area also restricted the conversion of isobutane, as
observed over Cs1.5(NH4)1.5H and Cs0.5(NH4)2.5H catalysts,
despite the fact that they possess a relatively high number of
acid sites. The highest activity is obtained neither over the
catalyst with the strongest acidity (labelled e in Fig. 9) nor
over the catalyst with largest specific surface area (labelled a
in Fig. 9). Since the specific surface area and the acidity show
an opposite trend when varying the Cs loading (and, thus
Fig.
8 Correlation between the catalytic activity and acidity
[a: Cs3(NH4)0H, b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H, d: Cs1.5(NH4)1.5H,
e: Cs0.5(NH4)2.5H].
lowest amount of acid sites (0.17 mmol g−1 NH3 uptake)
−1
shows the lowest activity (1.81 μmol gcat. min−1). As the
number of acid sites increases to 0.30 mmol g−1 for
Cs2.5(NH4)0.5H, the activity for isobutane oxidation increase
−1
sharply to 14.92 μmol gcat.
min−1. A similar value is
obtained for the Cs1.7(NH4)1.3H catalyst which exhibits an
increased acidity of 0.72 mmol g−1. On the other hand, a
further increase in the number of acid sites, as for the
Cs1.5(NH4)1.5H and Cs0.5(NH4)2.5H samples do not improve
the activity anymore, which may be related to the low surface
area of the samples. As a matter of fact, the selective oxi-
dation of isobutane over heteropolycompounds catalysts was
classified as a surface-type reaction.22
It has been proved that the mechanism of isobutane
oxidation involves an oxidative dehydrogenation step to give
isobutene,27–29 which is subsequently converted to MAC and
MAA, 500 times faster than it is formed. The C–H bond acti-
vation, which is therefore considered as the rate-limiting
step, is promoted by the surface acidic sites.24,26 Activity
and acidity were then accounted per unit of catalyst surface
(with the notion of surface acid density discussed in a
Fig. 9 Relationship between the activity and the surface acid density
[a: Cs3(NH4)0H, b: Cs2.5(NH4)0.5H, c: Cs1.7(NH4)1.3H, d: Cs1.5(NH4)1.5H,
e: Cs0.5(NH4)2.5H].
2944 | Catal. Sci. Technol., 2014, 4, 2938–2945
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
when varying x), it then become crucial to find the “key
point” in order to gain the appropriate acidity and surface
area, which ensures high isobutane conversion. According to
our results, the optimum is then between 1.7 and 2.5 Cs
atom per Keggin unit.
4 K. Inumaru, T. Ito and M. Misono, Microporous Mesoporous
Mater., 1998, 21, 629–635.
5 T. Okuhara, H. Watanabe, T. Nishimura, K. Inumaru and
M. Misono, Chem. Mater., 2000, 12, 2230–2238.
6 J. S. Santos, J. A. Dias, S. C. L. Dias, F. A. C. Garcia,
J. L. Macedo, F. S. G. Sousa and L. S. Almeida, Appl. Catal., A,
2011, 394, 138–148.
7 C. Marchal-Roch, N. Laronze, R. Villanneau, N. Guillou,
A. Tézé and G. Hervé, J. Catal., 2000, 190, 173–181.
8 J. He, H. Pang, W. Wang, Y. Zhang, B. Yan, X. Li, S. Li and
J. Chen, Dalton Trans., 2013, 42, 15637–15644.
4. Conclusions
Mixed salts Csx(NH4)3−xHPMo11VO40 (x = 0–3) were prepared
by the precipitation method. Their properties such as texture,
nature, crystallinity and acidity were determined and success-
fully correlated to the Cs substitution extent represented by
x in the previous formula. We suggest that the introduction
of caesium in the catalysts formula stabilized the Keggin
structure and limited the expelling of vanadium from the
primary structure, avoiding mono-oxide phase segregation.
Furthermore, the specific surface area increases with the Cs
content in the catalyst. However, as a consequence of the
9 T. Ressler, U. Dorn, A. Walter, S. Schwarz and A. H. P. Hahn,
J. Catal., 2010, 275, 1–10.
10 P. Visuvamithiran, K. Shanthi, M. Palanichamy and
V. Murugesan, Catal. Sci. Technol., 2013, 3, 2340–2348.
11 T. Okuhara, H. Watanabe, T. Nishimura, K. Inumaru and
M. Misono, Chem. Mater., 2000, 12, 2230–2238.
12 F. Jing, Y. Zhang, S. Luo, W. Chu, H. Zhang and X. Shi,
J. Chem. Sci., 2010, 122, 621–630.
+
Cs+/NH4 ratio increase, a decrease in the number of acid
sites was concomitantly observed. This opposite trend
implies that a trade-off has to be found to keep a sufficiently
high acid site density, which is of main importance for the
catalytic activity since the isobutane selective oxidation takes
place on acid active sites at the surface of the catalyst. Thus,
proper balance between acidity and specific surface area
enables getting an optimised yield, which was observed over
Cs1.7(NH4)1.3H. This latter exhibited an isobutane conversion
of 9.6%, and a total selectivity to the desired products
MAC + MAA of 57.1%.
13 I. Kozhevnikov, J. Mol. Catal. A: Chem., 2009, 305, 104–111.
14 I. V. Kozhevnikov, J. Mol. Catal. A: Chem., 2007, 262, 86–92.
15 X.-K. Li, J. Zhao, W.-J. Ji, Z.-B. Zhang, Y. Chen, C.-T. Au,
S. Han and H. Hibst, J. Catal., 2006, 237, 58–66.
16 M. Sultan, S. Paul, M. Fournier and D. Vanhove, Appl. Catal.,
A, 2004, 259, 141–152.
17 C. Knapp, T. Ui, K. Nagai and N. Mizuno, Catal. Today,
2001, 71, 111–119.
18 C. Rocchiccioli-Deltcheff and M. Fournier, J. Chem. Soc.,
Faraday Trans., 1991, 87, 3913–3920.
19 N. Mizuno, D. J. Suh, W. Han and T. Kudo, J. Mol. Catal. A:
Chem., 1996, 114, 309–317.
Acknowledgements
20 N. Ballarini, F. Candiracci, F. Cavani, H. Degrand,
J. L. Dubois, G. Lucarelli, M. Margotti, A. Patinet, A. Pigamo
and F. Trifiro, Appl. Catal., A, 2007, 325, 263–269.
21 C. Marchal-Roch, C. Julien, J. F. Moisan, N. Leclerc-Laronze,
F. X. Liu and G. Herve, Appl. Catal., A, 2004, 278, 123–132.
22 N. Mizuno, M. Tateishi and M. Iwamoto, J. Catal., 1996, 163,
87–94.
The authors thank O. Gardoll and G. Cambien for TG and N2
adsorption/desorption analyses, respectively. The Fonds
Européen de Développement Régional (FEDER), CNRS, Région
Nord Pas-de-Calais and Ministère de l'Education Nationale
de l'Enseignement Supérieur et de la Recherche are acknowl-
edged for fundings of X-ray diffractometers. The staffs and
PhD students in Ecole Centrale de Lille are also acknowledged
for their kind supports. F. Jing thanks the China Scholarship
Council (CSC) for providing the financial support during his
stay in France.
23 F. Cavani, Catal. Today, 1998, 41, 73–86.
24 G. Busca, F. Cavani, E. Etienne, E. Finocchio, A. Galli,
G. Selleri and F. Trifiro, J. Mol. Catal. A: Chem., 1996, 114,
343–359.
25 N. Mizuno and M. Misono, Chem. Rev., 1998, 98, 199–217.
26 S. Paul, V. LeCourtois and D. Vanhove, Ind. Eng. Chem. Res.,
1997, 36, 3391–3399.
References
1 N. Mizuno and H. Yahiro, J. Phys. Chem. B, 1998, 102,
437–443.
27 Q. Huynh, Y. Schuurman, R. Delichere, S. Loridant and
J. M. M. Millet, J. Catal., 2009, 261, 166–176.
2 F. Jing, B. Katryniok, E. Bordes-Richard and S. Paul, Catal.
Today, 2013, 203, 32–39.
28 G. P. Schindler, T. Ui and K. Nagai, Appl. Catal., A,
2001, 206, 183–195.
3 F. Jing, B. Katryniok, F. Dumeignil, E. Bordes-Richard and
S. Paul, J. Catal., 2014, 309, 121–135.
29 L. Jalowiecki-Duhamel, A. Monnier, Y. Barbaux and
G. Hecquet, Catal. Today, 1996, 32, 237–241.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2938–2945 | 2945