Reflux method as a novel route for the synthesis of MoVTeNbOx catalysts for selective oxidation of propane to acrylic acid

Article abstract of DOI:10.1016/j.molcata.2011.04.009

This work reports the applicability of reflux to the synthesis of Mo-V-Te-Nb-O metal oxide catalysts for the first time. The well-formed and defined orthorhombic M1 phase, i.e. Te₂M₂₀O₅₇ (M = Mo, V, Nb), was found to be better developed when a longer time was used to reflux the mixed metal clusters. The catalysts efficient for propane selective oxidation to acrylic acid, reaching ca. 60% of selectivity to acrylic acid at propane conversion of ca. 50% and at a reaction temperature of 693 K. In addition, a parallel between catalytic activity and catalyst reducibility is also concluded.

Full text of DOI:10.1016/j.molcata.2011.04.009

Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Reflux method as a novel route for the synthesis of MoVTeNbOx catalysts for
selective oxidation of propane to acrylic acid
Irmawati Ramlia,∗ , Pablo Botellab , Francisco Ivarsb , Woi Pei Menga , Siti Murni Mohamad Zawawia ,
Hossein Abbastabar Ahangara, Selene Hernándezb, José M. López Nietob,∗
a Centre of Excellence for Catalysis Science and Technology, Department of Chemistry, Faculty of Science, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
b Instituto de Tecnología Química, UPV-CSIC, Campus Universidad Politécnica de Valencia, Avenida de los Naranjos s/n, 46022 Valencia, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
This work reports the applicability of reflux to the synthesis of Mo–V–Te–Nb–O metal oxide catalysts
for the first time. The well-formed and defined orthorhombic M1 phase, i.e. Te2M20O57 (M = Mo, V, Nb),
was found to be better developed when a longer time was used to reflux the mixed metal clusters. The
catalysts efficient for propane selective oxidation to acrylic acid, reaching ca. 60% of selectivity to acrylic
acid at propane conversion of ca. 50% and at a reaction temperature of 693 K. In addition, a parallel
between catalytic activity and catalyst reducibility is also concluded.
Received 29 November 2010
Received in revised form 4 March 2011
Accepted 17 April 2011
Available online 22 April 2011
Keywords:
Propane
© 2011 Elsevier B.V. All rights reserved.
Selective oxidation
Acrylic acid
Reflux method
Mo–V–Te–Nb mixed metal oxide
1. Introduction
[11–23] and the isolation of the active sites [11,15–18,20,21,23].
In the former, different authors have concluded that at least in the
MoVTeNbOx mixed oxide is reportedly the most promising cat-
alyst system for the selective oxidation of propane to acrylic acid.
These catalysts were developed and patented by Mitsubishi Chem-
icals [1], who reports the highest acrylic acid yield (ca. 50%) up until
today. Academic and industrial communities in the area of partial
oxidation are enthusiastic about improving its catalyst perfor-
mance. This is easily evidenced by a significant amount of work on
these materials [1–24]. It is debatably suggested that their unique
catalytic properties are due to the presence of two major structures,
termed as orthorhombic M1 phase, Te2M20O57 (M = Mo, V, Nb),
and hexagonal M2 phase, Te0.33MO3.33 (M = Mo, V, Nb) [9]. Other
case of the (amm)oxidation of propane, the performance of the M1
phase could be improved by cooperation of the pseudohexagonal
Te0.33MO3.33 phase (M = Mo, V, Nb) [11,16–18,20,21,23].
Despite the effort expended in synthesizing and characteriz-
ing mixed metal oxides, only a few methods have been used
to synthesize these catalysts. Although Slurry [1,4,7,11,12] and
hydrothermal [3,4,6,11,13,23,24] methods are the most common
options, microwave-assisted hydrothermal synthesis has been also
proposed [24]. Undoubtedly, these methods have produced the
desired catalysts. This is especially true for the hydrothermal
method, through which it is possible to obtain, in the appropri-
ate conditions, the monophasic MoVTeNbO material with M1-like
crystalline structures. The hydrothermal synthesis of these mate-
rials is usually performed at 448 K in a nitrogen atmosphere with a
synthesis time of 48 h [11,23,24] and is mainly characterized by its
high solvent effect and equilibrium conditions, which lead to the
formation of the most favored crystalline phase. On the contrary,
the slurry method uses a non-solution precursor, which implies a
complexity in the formation of the active phase. For this reason, the
slurry method contributes to the multiphase MoVTeNbO system in
which the species is always present in different valence states due
to its non-equilibrium state [1,4,7,11,12].
phases such as (Mo0.93V0.07)5O14, MoO3, and TeMo5O16 can also
be present depending on the preparation and activation [18,19].
The M1 phase alone can activate propane and selectively convert
it to acrylic acid [18,20]. On the contrary, the M2 phase is inactive
for propane oxidation, but selectively oxidizes free intermediately
formed propylene to desired products [11,19]. Two additional top-
ics have been stressed in the literature: the cooperation of phases
∗
Corresponding authors at: Centre of Excellence for Catalysis Science and Tech-
nology, Department of Chemistry, Faculty of Science, University Putra Malaysia,
43400 UPM Serdang, Selangor, Malaysia. Tel.: +60 389466786; fax: +60 389432508.
E-mail addresses: irmawati@science.upm.edu.my (I. Ramli), jmlopez@itq.upv.es
(J.M. López Nieto).
In the present work, a reflux method has been developed for
the synthesis of MoVTeNbOx catalysts. The catalysts were prepared
by refluxing the aqueous synthesis gel with a mixture of metals.
1381-1169/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.04.009

I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
51
This method has the advantage of allowing intimate mixing of
corresponding species in a homogeneous state under atmospheric
pressure and at the solvent’s boiling point. This is the first reported
work on the utilization of this method in synthesizing MoVTeNbOx
catalysts for propane selective oxidation. The method has been
found successful for the synthesis of (VO)2P2O7 [25] based cata-
lyst employed in the transformation of butane to maleic anhydride.
The effect of the reflux period on the catalyst structural develop-
ment was also studied. The activity of these catalysts in the propane
conversion to acrylic acid was studied.
from room temperature to 1250 K following the evolution of
water continuously on a mass spectrometer. The temperature was
increased linearly at 10 K/min.
2.3. Catalytic test
Propane selective oxidation was carried out at atmospheric
pressure in a fixed bed quartz tubular reactor (i.d. 12 mm, length
400 mm). The weight of the catalyst (with homogeneous particle
size in the 0.3–0.5 mm range) was varied in order to achieve dif-
ferent contact times, Silicon carbide was added in order to keep
a constant volume in the catalytic bed. The gas feed consisting of
C3H8:O2:H2O:He with a 4:8:30:58 (%) molar composition was sup-
plied at flow rates of 30–100 ml/min. Experiments were carried out
in the 613–693 K range. Reactants and products were analyzed by
online gas chromatography using two packed columns: (i) molec-
2. Experimental
2.1. Catalyst preparation
Commercial chemicals of ammonium heptamolybdate, AHM,
(NH4)6Mo7O24·4H2O (Merck), vanadyl sulphate, VOSO4 (Aldrich),
telluric acid, Te(OH)6 (Aldrich), and ammonium niobium oxalate,
(NH4)2Nb2 (C2O4)5 (CBMM), were used as the metal sources for Mo,
V, Te and Nb respectively. The gel presented a Mo/V/Te/Nb atomic
ratio of 1/0.3/0.17/0.12. The preparation process begins by dissolv-
ing AHM in deionized water, heated at 353 K and was followed by
the addition of telluric acid to form the first solution. In another
beaker, a second solution was prepared by dissolving VOSO4 in
deionized water heated at 353 K, and then added to the first solu-
tion, producing a dark green solution. A third solution was obtained
by dissolving niobium oxalate in deionized water solution, which
was added to the former mixture. The resulting dirty green mixture
was left to stir for 10 min before being refluxed. The effect of reflux
time was studied by conducting reflux for 1, 4, 7, and 14 days. Fol-
lowing the reflux, the solid phase was filtrated and a purplish-black
solid was obtained which was then dried overnight in an oven at
373 K. The dried samples were heat-treated in flowing N2 at 873 K
for 2 h. The corresponding materials were denoted as A-1, A-4, A-7,
and A-14 (Table 1).
˚
ular sieve 5 A (3 m); and (ii) Porapak Q (3 m).
3. Results and discussion
Table
1 summarizes some of the main characteristics of
MoVTeNbOx catalysts. No significant differences in the chemical
composition can be seen in the activated samples, which are close
to the atomic ratios used in the synthesis gel. On the other hand,
heat-treated samples present BET surface areas ranging from 1.9 to
9.3 m2 g−1 (Table 1), within the range of those previously described
for similar materials [21,26,34]. A correlation between surface area
and reflux time is observed, in that the longest reflux time is
obtained by the highest surface area.
3.1. Characterization of as-synthesized samples
XRD patterns of the as-synthesized samples (before heat
treatment) are displayed in Fig. 1A. The reflections set with a
peak at 2ꢁ = 22◦, and two broad peaks at 2ꢁ < 10◦ (low intense)
and at 2ꢁ = 27◦ are similar to those found in the precursor
of the Mo/W orthorhombic and tetragonal phases [11,29,30].
The peak appeared at 2ꢁ of ca. 22◦ is related to a kind of
Mo-based oxides with layer-type structures closely related to
the ReO3-family [27,28]. Overlapping another set of reflections
are observed only in the XRD patterns of as-synthesized sam-
ples prepared from 1 to 7 days, which indicate the presence
of heteropoly-compounds (HPC) with Anderson-type structure
(NH4)6TeMo6O24·7H2O [11], (NH4)7TeMo5VO24·8H2O [12], and/or
(NH4)8(V19O41 (OH)9)(H2O)11 [ICSD 063213]. The narrow diffrac-
tions become broad bands during the reflux, which means that
HPCs slowly turn into pseudo-crystallines with increasing time of
up to 14 days’ synthesis resulting in a homogeneous distribution of
crystalline nano-domains without a trace of heteropoly-compound
reflections (Fig. 1A, d).
2.2. Catalyst characterization
BET specific surface area measurements were performed using
a ThermoFinnigan Sorptomatic apparatus, with adsorption of N2 at
77 K. Bulk composition of the catalyst was determined by inductive
coupled plasma-atomic emission spectrometer (ICP-AES).
Powder XRD patterns were collected by employing a Philips
X’Pert diffractometer equipped with a graphite monochromator,
operating at 40 kV and 30 mA and using Ni filtered CuK␣ radiation
with wavelength, ꢀ = 0.1542 nm. Samples were ground and put on
a horizontal sample holder and XRD patterns were recorded in the
5–60◦ range.
Infrared spectra were recorded at room temperature in the
300–3900 cm−1 region with a Nicolet 205xB spectrophotometer,
equipped with a Data Station, at a spectral resolution of 1 cm−1
with an accumulation of 128 scans. The pellets were prepared with
20 mg of sample mixed with 100 mg of dry KBr and pressed into
disks.
Raman spectra were obtained with an “in via” Renishaw spec-
trometer, equipped with an Olympus microscope. The exciting
wavelength was 785 nm from a Renishaw HPNIR laser with a power
of ca. 15 mW on the sample.
Scanning electron microscopy (SEM) was performed using JEOL
Model JSM 6400 to record the image at an accelerating voltage of
15 kV.
Raman spectra of as-synthesized samples, before the heat-
treatment, are shown in Fig. 2A. The A-1 sample presents a
homogeneous composition with practically identical Raman spec-
tra obtained at different discrete points of the sample, characterized
by the presence of a band at ca. 950 cm−1. This band is related
to Mo–Ot stretching vibration from Anderson-type HPC structures
(more or less shifted depending on hydration), in agreement with
XRD results [31,32]. On the other hand, the sample obtained at 4
days of reflux (A-4 sample) presents a heterogeneous composition
with different Raman spectra at different selected areas of the sam-
ple (Fig. 2A, spectra b1–b3), similar to spectra of samples prepared
in 1 day (indicating the presence of Anderson HPC). However, there
Temperature programmed reduction of hydrogen (H2-TPR) was
achieved by using a ThermoFinnigan TPDRO1110 apparatus, uti-
lizing a thermal conductivity detector (TCD). Hydrogen (5% in Ar,
1 bar, 25 ml/min) flowed over the samples at temperatures ranging
is also a new Raman spectrum at 874 cm−1, which can be associated
6−
with the presence of pentagonal [M6O21
]
(M = Mo, V, Nb) units
[33]. In the case of the sample A-14 (Fig. 2A, spectrum c), only one
type of Raman spectrum is observed with a wide band centered

52
I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
Table 1
General characteristics of MoVTeNbOx catalysts prepared by reflux.
Catalysts
Reflux time (days)
Bulk compositiona
BET surface area (m2 g−1
)
TPR results
b
Tmax (K)
Er (kJ mol−1
)
O2 (atom g−1
)
A-1
A-4
A-7
1
4
7
Mo1V0.22Te0.17Nb0.08
Mo1V0.25Te0.11Nb0.10
Mo1V0.27Te0.12Nb0.11
1.9
4.5
4.7
941c
157
2.5 × 1021
2.5 × 1021
911d
152
3.2 × 1021
3.2 × 1021
1.7 × 1021
1.7 × 1021
3.4 × 1021
824
883
138
148
A-14
14
Mo1V0.28Te0.14Nb0.11
9.3
839
887
140
148
1.8 × 1021
1.1 × 1021
2.9 × 1021
a
Chemical composition determined by ICP-AES. In all cases, the atomic ratio in the synthesis gel was Mo/V/Te/Nb = 1/0.3/0.17/0.12.
Reduction activation energies in H2-TPR experiments.
A shoulder at 890 K is also observed.
b
c
d
Shoulders at 820 and 890 K are also observed.
at 850 cm−1, which is formed by at least two bands overlapping
with maximum peaks at 840 and 875 cm−1, the latter related to the
the A-14 solid IR spectrum. In contrast with the others, the latter
spectrum is quite similar to those obtained after heat treatment
at 873 K in N2 (Fig. 3B, d) with slight differences due to improved
crystallization and a different level of hydration, as indicated by
the peak at 1600 cm−1 which is also present in the latter sample
but with a lower intensity. This peak associated to H–O–H bending
modes also is present in the remaining heat-treated samples.
structural pentagonal units [33], and a shoulder at 930 cm−1
.
Fig. 3A shows the IR spectra of as-synthesized samples in the
range of 500–1800 cm−1. All of these spectra show two bands at
ca. 1600 and 1400 cm−1, associated with the presence of H2O and
ammonium ions respectively [34], forming part of the solids. On
the other hand, the sample prepared in 1 day shows IR bands at
ca. 620, 670, 800, 890, 937 cm−1 which is related to the presence
of a tellurium molybdate with Anderson-type structures [31], in
accordance with XRD and Raman results.
The most highlighted difference among IR spectra of as-
synthesized samples is an intensity increase of the bands at 858
and 890 cm−1. The intensity of bands centered at ca. 800 and
820 cm−1 reduces as reflux time increases and as a consequence,
the resolution of IR spectra becomes lower in A-4 and is practi-
cally indistinguishable in the A-7 sample. The band at 937 cm−1
also decreases with the reflux time until it is practically absent in
3.2. Characterization of catalysts
As a consequence of the different as-synthesized solids, several
crystalline phases are observed by XRD in catalysts, i.e. materials
heat-treated at 873 K in N2 (Fig. 1B), depending on time of reflux.
Thus, the TeM5O16 (M = Mo, V, Nb) phase is mainly observed in the
case of A-1 catalysts (Fig. 1B, pattern a), although the presence
of M5O14 cannot be completely ruled out. However, new peaks
begin to appear at 2ꢁ = 6.6, 7.9, 9.0, 22.1, and 27.2◦, characteris-
tic of orthorhombic M1 phase [26], Te2M20O57 (M = Mo, V, Nb), in
A
B
d
d
c
c
b
b
a
a
5
15
25
35
45
5
15
25
35
45
2θ
2θ
Fig. 1. XRD patterns of as-synthesized (A) and heat-treated (B) MoVTeNbOx materials from solids prepared by reflux during: 1 (a), 4 (b), 7 (c), or 14 (d) days. Symbols: (ꢀ)
M1 phase (Te2M20O57), (ꢀ) M2 phase (Te0.33MO3.33), (ꢁ) TeM5O16, (+) M5O14. (M = Mo, V, Nb).

I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
53
A
B
c
b3
b2
c
b
b1
a
a
0
400
800
1200
1600
0
400
800
1200
1600
Raman shift (cm-1)
Raman shift (cm-1)
Fig. 2. Raman spectra of as-synthesized (A) and heat-treated (B) MoVTeNbOx materials prepared by reflux during 1 day (a), 4 days (b), or 14 days (c). b1, b2, b3 indicate
Raman spectra at different selected particles in the same sample.
catalysts achieved from precursors with refluxing time of 4 days
(Fig. 1B, patterns c–e, respectively). In addition to these, peaks at
2ꢁ = 22.3, 25.8, 28.3, 36.3, 45.0, and 50.0◦, related to M2 phase,
Te0.33MO3.33 (M = Mo, V, Nb), were also observed to a minor extent.
It should be noted that TeM5O16 was not observed in A-14 cata-
lyst, which is mainly of the M1 phase with the minority presence
of M2 phase (Fig. 1B, pattern d). Additionally, it must be noted that
the longest reflux time during the preparation of catalyst precursor
provides the best scenario for the M1 phase crystallization at a high
temperature during activation (Fig. 1B).
of low intensity. The intensity of the latter Raman peaks increases
for the rest of catalysts, overlapping and forming a wide Raman
band with maxima and shoulders at ca. similar wave numbers, save
the maximum at 843 cm−1 (in A-1 catalyst) which shifts to lower
energies with increasing days in reflux, up to 800 cm−1 in the A-
14 catalyst. Regarding the zone of less absorption, the presence of
477 cm−1 in A-14 sample suggests the main presence of M1 phase
[35], in agreement with XRD results.
Fig. 3B shows the IR spectra of catalysts. In the case of the A-
1 catalyst (Fig. 3B, spectrum a), the IR spectrum shows bands at
904, 727, 650 and 540 cm−1 which are similar to those described
for the TeMo5O16 phase [36]. In addition to these, a shoulder
at 950 cm−1, associated with stretching Mo-O frequencies in the
Fig. 2B shows the Raman spectra of heat-treated samples. The
Raman spectrum of the A-1 catalyst presents two main peaks at
843 and 910 cm−1 with shoulders at 680, 720, 880 and 980 cm−1
A
B
d
c
b
d
c
a
b
a
1000
600
1800
1400
1000
600
1800
1400
cm-1
cm-1
Fig. 3. IR spectra of as-synthesized (A) and heat-treated (B) MoVTeNbOx materials prepared by reflux during 1 day (a), 4 days (b), 7 days (c) or 14 days (d).

54
I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
Fig. 4. SEM micrography of MoVTeNbOx catalysts (heat-treated at 873 K in N2) achieved from solid prepared by reflux during 1 day (a), 4 days (b) or 14 days (c).
Mo5O14 phase [37], is also observed. We must indicate that the
shoulder at 950 cm−1 is not present in catalysts prepared from 4 to
14 days in reflux, for which Mo5O14 was not observed by XRD. How-
ever, the intensity of IR bands in the 700–900 cm−1 range increases
with reflux time. In this way, the IR spectrum of the A-14 catalyst
ping TPR signals, the TPR patterns of samples prepared at synthesis
times of 7 and 14 days have two maximum peaks at 820 and 890 K
(Fig. 5, patterns c and d), similar to previous results for hydrother-
mally prepared MoVTeNb catalysts containing M1 and M2 phases
[39,40]. Note that the reduction of both Mo6+ and V5+ which are
octahedrally coordinated in our catalysts are reduced more easily
than in pure MoO3 and V2O5 [41]. However, for the 1-day refluxed
sample, the maximum peak happens at 941 K with a shoulder at
883 K (Fig. 5a). This reduction peak is higher than that of other cat-
alysts and is probably due to the reduction of TeMo5O16 and/or
M5O14 phases (M = Mo, V, Nb) [40].
presents a set of bands at 905, 880, 775, 715, 630 and 575 cm−1
,
characteristic of the crystalline M1 phase [19]. The presence of
the M2 phase cannot be completely ruled out, as suggested by the
hinted shoulders at 924 and 750 cm−1 [11a].
The catalysts were studied by scanning electron microscopy
(SEM). In general, the heat-treated catalysts obtained from materi-
als prepared with four or more days of reflux showed crystals with
pointed corners and sharp edges, which indicate highly crystallined
materials. This is in good agreement with XRD results, which
showed highly intense and sharp diffraction peaks. A-14 catalyst
is presented in Fig. 4c as rod-shape crystals, commonly observed
for M1 phase crystals [11b,38], with a rather homogeneous size,
along with more particles in an apparent quasi-amorphous state.
In contrast, the sample prepared in one day contains larger crystals
(Fig. 4a). According to Ueda and Oshikara [14], the catalytic prop-
erties are derived from the cross-sections of the rods. Therefore,
materials were ground before being tested in catalyst reaction in
order to enhance their catalytic performance.
Fig. 5 shows the temperature programmed reduction profiles
(H2-TPR) of heat-treated samples. The catalysts display changes
in the TPR patterns depending on the synthesis time. At low syn-
thesis time, i.e. 1 day, the catalyst mainly shows two reduction
peaks at 890 and 940 K (Fig. 5a). Moreover, a new reduction peak at
824–841 K increases and the reduction peaks at 940 K shift to lower
temperatures when reflux time is increased during catalyst syn-
thesis. This is in accordance with the presence of crystalline phases
which are octahedrally more distorted as shown by XRD results
(Fig. 1). Although resolution is low due to the wide and overlap-
d
c
b
a
400
600
800
1000
1200
Temperature (K)
Fig. 5. H2-TPR profiles of MoVTeNbOx catalysts (heat-treated at 873 K in N2)
achieved from solid prepared by reflux method during 1 day (a), 4 days (b), 7 days
(c) or 14 days (d).

I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
55
Table 2
Catalytic results obtained during propane oxidation at 673 K.
Catalysts
Conv. (%)a
Selectivity (%)
AAb
Propene
Acroleinc
HAcb
CO
CO2
A-1
A-4
A-7
A-14
2.5
8.5
14.6
28.9
0.0
12.8
31.1
56.2
77.3
39.8
38.9
11.3
0.4
0.0
0.2
0.2
0.6
11.2
24.4
15.5
15.4
11.1
22.8
14.3
16.5
t
t
t
−1
a
Propane conversion at 673 K and a contact time, W/F, of 200 gcat h molC
Selectivity to acrylic acid (AA) and acetic acid (HAc).
Traces.
.
H
3
8
b
c
The onset of reduction reaction occurs at ∼ 650 K, which is the
reflux, respectively), but selectivity improved greatly with the A-14
catalyst. A-14 catalyst was the most active and selective in propane
partial oxidation to acrylic acid. The selectivity towards propy-
lene also followed a similar trend, while an inverse tendency was
observed for the selectivity to both acetic acid and COx.
same for all catalysts. This onset temperature coincides approxi-
mately with the temperature for propane activation. Therefore, it is
suggested that the oxygen removed in the anaerobic H2-TPR could
be the same lattice oxygen species for the selective oxidation of
propane. In this way, a rough estimation on the amount of oxygen
removed was made by deconvoluting the H2-TPR curves (Table 1).
All of the samples released roughly the same number of oxygen
atoms by reduction, i.e. ∼3.0 × 1021 atom g−1. However, the max-
imum for H2-uptake change depends on the synthesis time or, in
other words, on the nature of crystalline phases. Thus, samples A-7
and A-14 presented the higher numbers of oxygen atoms reduced
at low temperature (Table 1) with a maximum peak at 820 K.
On the other hand, monolayer surface oxygen species accounted
for roughly 1.0 × 1015 atom g−1 cm−2, according to the Arrhenius
law, so lattice oxygen species are also involved in H2-TPR results.
All of the samples showed rather high reduction activation energy
(Er >100 kJ mol−1).
Characterization and catalytic reaction results confirm that it
is necessary for the presence of the M1 phase to obtain active
and selective catalysts for propane oxidation to acrylic acid, as
was previously reported [1,14,22,23]. In this way, the materials
prepared from 4 to 14 days of reflux present mainly M1 crys-
talline phase, with small differences in the phases distribution
among them, though important differences in the catalytic behav-
ior are observed. Therefore, there seems to be a strong dependence
between catalytic behavior and the crystallization of the M1 phase
[23]. Reflux time improves the crystalline growth of the M1 phase
but also minimizes the presence of other undesired crystalline
phases that are inactive for propane oxidation and/or unselective
for acrylic acid in consecutive oxidation of reaction intermediates.
Poor activity for the propane oxidation of A-1 is due to the
absence of M1 phase, which may be a consequence of the pres-
ence of the samples’ disordered structure since the reflux time
employed for its synthesis was insufficient to develop the appropri-
ate solid precursor, which leads to the crystallization of the active
and selective M1 phase.
On the other hand, the catalytic activity is clearly related to
the behavior of catalysts in the TPR experiments. Thus, the higher
catalytic activity was observed in samples prepared at high synthe-
sis times, which presented a first reduction peak at 820 K (Fig. 5).
Moreover, the samples prepared with a lower synthesis time show
the same number of oxygen atoms in the 600–1000 K range, but
with lower reducibility, presenting the maximum H2-uptake in the
900–940 K range.
In the catalysts with M1 and M2 phases, Mo and V ions may
undergo charge transfer on the Mo and V sites: V5+ 2− ꢁ V4+O−
O
and Mo5+ − ꢁ Mo6+O2−. The lattice oxygen species O2− could be
O
involved in the ˛-H abstraction in propane (V-sites) and with the
insertion into the allylic intermediate (Mo-sites). Some studies
revealed an association between the V O activity and the dehy-
drogenation of propane [40–42], whereas Mo O is involved in
selective oxidation of propylene [16,43–44].
3.3. Catalytic test
Results obtained during propane oxidation with the prepared
catalysts are summarized in Table 2. The main reaction products
were acrylic acid, propylene and COx, though acrolein and acetic
acid were also detected as minority compounds. Catalytic results
show that A-4, A-7 and A-14 catalysts are active and selective for
propane selective oxidation to acrylic acid, with a catalytic behavior
improving with refluxing time during catalyst synthesis. However,
the catalytic properties drastically change for A-1 catalyst, which
presents an extremely low propane conversion.
Fig. 6 shows the variation in both the catalytic activity (Fig. 6A)
and the selectivity to acrylic acid (Fig. 6B) with the reaction temper-
ature achieved during the propane oxidation on catalysts prepared
by reflux. In general, it is observed that the refluxing time during
catalyst synthesis exerts an important influence on the catalytic
activity of the final catalyst, so that the propane conversion is pro-
gressively higher with reflux time. The highest gap is observed for
the catalyst prepared with 14 days in reflux, which is the most
active one. It is important to note that the catalyst with only 1 day
in reflux (sample A-1) is practically inactive for propane conversion
in all the range of temperatures tested, in that only traces of acrylic
acid were found during propane oxidation with this material.
A similar trend was observed in the selectivity to acrylic acid,
whereby the catalysts prepared from 4 days in reflux presented
significant acrylic acid selectivity. The selectivity to acrylic acid
slightly increased from A-4 to A-7 catalyst (from 4 to 7 days in
4. General remarks
According to characterization results, as reflux time increases,
the as-synthesized solid gradually transforms into a precursor with
similar characteristics to that obtained by employing a similar syn-
thesis gel composition through hydrothermal synthesis, according
to XRD [11], IR [11b] and Raman [35] results. In this way, reflux pro-
vides an excellent opportunity to study in detail, the formation of
the different phases during the synthesis of MoVTeNb mixed metal
oxides, which will contribute to an improved understanding of the
molecular processes that lead to the formation of an optimal pre-
cursor. Additionally, reflux provides the advantage of interactions
at the desired moment in order to incorporate changes and modi-
fications at a time that could lead to new materials with improved
catalytic properties.
It must be taken into account that hydrothermal and slurry syn-
theses are the methods most often used to prepare these kinds
of catalysts. So the procedures involving them have been widely
studied and optimized for more than a decade [1].
According to XRD results (Fig. 1B), the nature of crystalline
phases in catalysts is a consequence of the transformation of the

56
I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
60
B
50
60
50
40
30
20
10
0
A
40
30
20
10
0
380
400
420
380
400
420
Temperature, ºC
Temperature, ºC
Fig. 6. Variation of propane conversion (A) and selectivity to acrylic acid (B), during partial oxidation of propane at different temperatures over MoVTeNbOx catalysts prepared
by reflux: A-1 (ꢀ); A-4 (᭹); A-7 (ꢂ); A-14 (ꢃ). Contact time, W/F, of 200 gcat h molC3−1. Molar ratio fed of C3H8/O2/He/H2O = 4/8/58/30.
different as-synthesized solids (Fig. 1A), which strongly depends
on refluxing time. The presence of hexagonal domains of Anderson-
type HPC led to the formation of TeMo5O16 (presenting hexagonal
performed through reflux. This is the first ever report on the appli-
cation of reflux to synthesize these mixed metal oxide catalysts.
The M1 orthorhombic phase, active and selective to acrylic acid
had significant results when the initial synthesis gel was refluxed
for at least 4 days, and was optimum for a reflux synthesis for 14
days.
channels), while the presence of both hexagonal domains and pen-
6−
tagonal [M6O21
]
(M = Mo, V, Nb) units allow for the formation of
M1 crystalline phase, which is characterized by its pentagonal rings
which join with the hexagonal ones and allows for the arrangement
of octahedral-forming heptagonal channels [33].
Through reflux, it is possible to combine the advantages of
both the slurry and hydrothermal methods, the most commonly
employed methods to prepare these kinds of MoVTeNbO catalysts.
Reflux favors solvent power and chemical transport, and provides
a controlled synthesis ambience for the steady oxidation states
of the different metal elements. All these factors lead to the for-
mation of meta-stable crystalline phases with higher purity and
homogeneous crystal size. The latter feature is similar to the results
obtained in the hydrothermal method, but without the inconve-
nience a closed system. Accordingly, reflux could be a promising
method to prepare this kind of catalyst, since it offers the possibil-
ity for observation and involvement, allowing interaction and more
thorough inspection of the reaction medium, whose chemical and
physical properties may change over the course of the reaction.
The method also favors solubility and provides the conditions and
controlled synthesis that welcomes steady oxidation states of the
different metals, which allows for the formation and stabilization
of meta-stable phases (in pseudo-minimum energy state) as is the
case of the M1 phase.
The XRD pattern of as-synthesized samples prepared by reflux
at 80 ◦C for 14 days (i.e. the precursor of A-14 catalyst) suggests the
presence of pseudo-amorphous materials (Fig. 1A, d). This is quite
similar to as-synthesized MoVTeNb materials prepared hydrother-
mally at 175 ◦C (48 h), with a low intense broad peak at 10◦, a peak
at 22◦ and a broad peak at around 27◦ [11,38]. However, it is differ-
ent from those obtained by the slurry method, which had only two
broad and low intense peaks at 2ꢁ around 10◦ and 27◦ [34].
On the other hand, the XRD pattern of as-synthesized samples
prepared by reflux at 80 ◦C and for 1 day, i.e. the precursor of A-
1, is quite similar to precursor materials of TeMo5O16-containing
solids prepared by slurry [34], presenting crystalline heteropoly-
compounds (HPC).
Similar conclusions can be suggested from the IR results. Thus,
Anderson type-HPC are clearly observed in the as-synthesized A-
1 sample (Fig. 3A, spectrum a) as well as in the precursors of
TeMo5O16-containing materials [31,34]. It must be indicated that
IR [45] and Raman [46] spectra similar to those of as-synthesized
A-1 samples are observed by other authors by drying or filtrating
the synthesis gel prior to hydrothermal synthesis.
On the contrary, the IR spectrum of precursors obtained after
14 days in reflux (the precursor of A-14 catalyst) is quite similar
to those obtained for well-crystallized M1 phase containing solids
prepared by other methods such as hydrothermal [11b,19] or slurry
[34] and are characterized by the presence of mixed metal oxides.
Refluxing for one day was not enough to form the desired pre-
cursor, with a poorly formed final activated material. A minimum
of four days of reflux was needed for the correct arrangement of
molecular building blocks in aqueous solution to enable M1 phase
crystallization during heat-treatment. We must indicate that the
XRD pattern of A-14 catalyst, achieved after heat-treatment of
as-synthesized sample prepared by reflux at 80 ◦C for 14 days, is
quite similar and presents catalytic behavior comparable to those
achieved for heat-treated samples prepared from hydrothermal
synthesis (at 175 ◦C, 2 days) [3,4,6,11,13,23,24], or microwave-
assisted hydrothermal synthesis (175 ◦C, 2 h) [24].
There is a clear parallel between redox and catalytic properties,
suggesting that the reducibility of crystalline phases is a key factor
in the partial oxidation of propane. In this way, the M1 phase is an
active and selective crystalline phase.
Acknowledgements
The authors would like to thank the Malaysian government (Sci-
ence Fund 03-01-04-SF0642 and research fellowship grants) and
DGICYT in Spain (Project CTQ2009-14495) for their financial sup-
port.
References
[1] T. Ushikubo, H. Nakamura, Y. Koyasu, S. Wajiki, US Patent 5,380,933 (1995).
[2] M. Lin, T. Desai, F. Kaiser, P. Klugherz, Catal. Today 61 (2000) 223–229.
[3] F. Ivars, B. Solsona, E. Rodriguez-Castellón, J.M. López Nieto, J. Catal. 262 (2009)
35–43.
[4] B. Deniau, J.M.M. Millet, S. Loridant, N. Christin, J.L. Dubois, J. Catal. 260 (2008)
30–36.
[5] X. Tu, N. Furuta, Y. Sumida, M. Takahashi, H. Niiduma, Catal. Today 117 (2006)
259–264.
[6] R. Feng, X.J. Yang, W.-J. Ji, H.-Y. Zhu, X.-D. Gu, Y. Chen, S. Han, H. Hibst, J. Mol.
Catal. A: Chem. 267 (2007) 245–254.
[7] M.M. Lin, M.W. Linsen, EP 962,253 (1999).
5. Conclusions
[8] Y. Koyasu, H. Watanabe, JP 143,244 (2000).
[9] T. Ushikubo, K. Oshima, A. Kayou, M. Hatano, Stud. Surf. Sci. Catal. 112 (1997)
473–480.
Synthesis of highly active and selective Mo–V–Te–Nb–O cat-
alysts for propane oxidation to acrylic acid was successfully

I. Ramli et al. / Journal of Molecular Catalysis A: Chemical 342–343 (2011) 50–57
57
[10] M. Aouine, J.L. Dubois, J.M.M. Millet, Chem. Commun. 13 (2001) 1180–1181.
[11] (a) P. Botella, B. Solsona, A. Martinez-Arias, J.M. López Nieto, Catal. Lett. 74
(2001) 149–154;
[27] M. Lundberg, M. Sundberg, Ultramicroscopy 52 (1993) 429–435.
[28] P. Botella, E. García-González, J.M. López Nieto, J.M. González-Calbet, Solid State
Sci. 7 (2005) 507–519.
(b) P. Botella, J.M. López Nieto, B. Solsona, A. Mifsud, F. Marquez, J. Catal. 209
(2002) 445–455.
[29] G. Mestl, Top. Catal. 38 (2006) 69–82, and references therein.
[30] P. Botella, E. García-González, B. Solsona, E. Rodríguez-Castellón, J.M. González-
Calbet, J.M. López Nieto, J. Catal. 265 (2009) 43–53.
[31] I.L. Botto, C.I. Cabello, H.J. Thomas, Mater. Chem. Phys. 47 (1997) 37–45.
[32] C. Martin, C. Lamonier, M. Fournier, O. Mentré, V. Harlé, D. Guillaume, E. Payen,
Inorg. Chem. 43 (2004) 4636–4644.
[33] M. Sadakane, K. Yamagata, K. Kodato, K. Endo, K. Toriumi, Y. Ozawa, T. Ozeki,
T. Nagai, Y. Matsui, N. Sakaguchi, W.D. Pyrz, D.J. Buttrey, D.A. Blom, T. Vogt, W.
Ueda, Angew. Chem. Int. Ed. 48 (2009) 3782–3786.
[34] J.M. Oliver, J.M. López Nieto, P. Botella, A. Mifsud, Appl. Catal. A 257 (2004)
67–76.
[12] H. Watanabe, Y. Koyasu, Appl. Catal. 194–95 (2000) 479–485.
[13] J.M. Oliver, J.M. López Nieto, P. Botella, Catal. Today 96 (2004) 241–249.
[14] W. Ueda, K. Oshihara, Appl. Catal. A: Gen. 200 (2000) 135–143.
[15] (a) R.K. Grasselli, Catal. Today 49 (1999) 141–153;
(b) R.K. Grasselli, Catal. Today 99 (2005) 23–31.
[16] (a) R.K. Grasselli, J.D. Burrington, D.J. Buttery, P. Desanto Jr., C.G. Lugmair Jr.,
A.F. Volpe Jr., T. Weingand, Top. Catal. 23 (2003) 5–22;
(b) P. DeSanto, D.J. Buttreya, R.K. Grasselli, C.G. Lugmair, A.F. Volpe, B.H. Tobyc,
T. Vogt, Top. Catal. 23 (2003) 23–38.
[17] (a) H. Tsuji, Y. Koyasu, J. Am. Chem. Soc. 124 (2002) 5608;
(b) H. Tsuji, K. Oshimna, Y. Koyasu, Chem. Mater. 15 (2003) 2112–2114.
[18] M. Baca, A. Pigamo, J.L. Dubois, J.M.M. Millet, Top. Catal. 23 (2003) 39–46.
[19] J.M. López Nieto, P. Botella, B. Solsona, J.M. Oliver, Catal. Today 81 (2003)
87–94.
[20] E. García-González, J.M. López Nieto, P. Botella, J.M. González-Calbet, Chem.
Mater. 14 (2002) 4416–4421.
[21] J.M.M. Millet, H. Roussel, A. Pigamo, J.L. Dubois, J.C. Jumas, Appl. Catal. A: Gen.
232 (2002) 77–92.
[22] M. Baca, M. Aouine, J.L. Dubois, J.M.M. Millet, J. Catal. 233 (2005) 234–
241.
[23] (a) D. Vitry, Y. Morikawa, J.L. Dubois, W. Ueda, Appl. Catal. A: Gen. 251 (2003)
411–424;
(b) W. Ueda, D. Vitry, T. Katou, Catal. Today 96 (2004) 235–240.
[24] N.R. Shiju, V.V. Guliants, ChemPhysChem 8 (2007) 1615.
[25] E. Bordes, Catal. Today 1 (1987) 499–526.
[26] P. Beato, A. Blume, F. Girgsdies, R.E. Jentoft, R. Schlögl, O. Timpe, A. Trunschke,
G. Weinberg, Q. Basher, F.A. Hamid, S.B.A. Hamid, E. Omar, L. Mohd Salim, Appl.
Catal. A: Gen. 307 (2006) 137–147.
[35] B. Solsona, M.I. Vázquez, F. Ivars, A. Dejoz, P. Concepción, J.M. López Nieto, J.
Catal. 252 (2007) 271–280.
[36] J.C.J. Bart, F. Cariati, A. Sgamelloti, Inorg. Chim. Acta 36 (1979) 105–112.
[37] F. Cariati, J.C.J. Bart, A. Sgamellotti, Inorg. Chim. Acta-Articles 48 (1981) 97.
[38] F. Ivars, P. Botella, A. Dejoz, j.M. López Nieto, P. Concepción, M.I. Vázquez, Top.
Catal. 38 (2006) 59–67.
[39] P. Botella, A. Dejoz, M.C. Abello, M.I. Vázquez, L. Arrúa, J.M. López Nieto, Catal.
Today 142 (2009) 272–277.
[40] X.J. Yang, R.M. Feng, W.J. Ji, C.T. Au, J. Catal. 253 (2008) 57–65.
[41] F. Arena, N. Giordano, A. Parmaliana, J. Catal. 167 (1997) 66–76.
[42] Y. Zhu, W. Lu, H. Li, H. Wan, J. Catal. 246 (2) (2007) 382–389.
[43] C. Huang, W. Guo, X. Yi, W. Weng, H. Wan, Catal. Commun. 8 (2) (2007) 162–166.
[44] J.M. Wu, H.P. Yang, Y.N. Fan, B.L. Xu, Y. Chen, J. Fuel Chem. Technol. 35 (6) (2007)
684–690.
[45] G.Ya. Popova, T.M. Andrushkevich, L.S. Dovlitova, G.A. Aleshina, Yu.A. Chesalov,
A.V. Ishenko, E.V. Ishenko, L.M. Plyasova, v.V. Malakhov, M.I. Khramov, Appl.
Catal. A 353 (2009) 249–257.
[46] A. Celaya Sanfiz, T.W. Hansen, F. Girgsdies, O. Timpe, E. Rödel, T. Ressler, A.
Trunschke, R. Schlögl, Top. Catal. 50 (2008) 19–32.