Hydrothermal synthesis of Mo-V mixed oxides possessing several crystalline phases and their performance in the catalytic oxydehydration of glycerol to acrylic acid

Article abstract of DOI:10.1016/j.cattod.2017.04.006

The one-step oxydehydration of glycerol to acrylic acid over molybdenum and vanadium mixed oxides was investigated. The Mo-V oxide catalysts were prepared by a simple hydrothermal method under different synthesis and calcination atmospheres and were characterized by in situ XRD, TPD-NH₃, N₂ adsorption/desorption, X-ray absorption near vanadium K-edge spectroscopy and thermogravimetry. The catalytic performance of the samples at different temperatures (290, 320 and 350 °C) and under different gas flow compositions (20% O₂ in N₂, 100% O₂, or 100% N₂) revealed that the arrangement of the crystallographic structures of the active phases directly influenced the catalytic performance. It was found that the catalysts heat-treated in oxidizing atmosphere gave superior catalytic results comparing with the catalysts heat-treated in inert atmosphere due to the equilibrium between the crystalline phases MoVO₅ and Mo_4.65V_0.35O₁₄ that contains V⁺⁴ and V⁺⁵. Catalytic oxydehydration at 320 °C under a flow of 100% O₂ gave the best performance, achieving selectivity of 33.5% towards acrylic acid and 100% conversion of glycerol.

Full text of DOI:10.1016/j.cattod.2017.04.006

Catalysis Today xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Hydrothermal synthesis of Mo-V mixed oxides possessing several crystalline
phases and their performance in the catalytic oxydehydration of glycerol to
acrylic acid
Letícia F. Rasteiro, Luiz H. Vieira, Luiz G. Possato, Sandra H. Pulcinelli, Celso V. Santilli,
⁎
Leandro Martins
Instituto de Química, UNESP—Universidade Estadual Paulista, Rua Prof. Francisco Degni 55, CEP 14800-900 Araraquara, SP, Brazil
A R T I C L E I N F O
A B S T R A C T
Keywords:
The one-step oxydehydration of glycerol to acrylic acid over molybdenum and vanadium mixed oxides was
Glycerol oxydehydration
Acrylic acid
Molybdenum-vanadium mixed oxides
Hydrothermal synthesis
Bifunctional catalysts
investigated. The Mo-V oxide catalysts were prepared by a simple hydrothermal method under different
synthesis and calcination atmospheres and were characterized by in situ XRD, TPD-NH , N adsorption/
3 2
desorption, X-ray absorption near vanadium K-edge spectroscopy and thermogravimetry. The catalytic
performance of the samples at different temperatures (290, 320 and 350 °C) and under different gas flow
2 2 2 2
compositions (20% O in N , 100% O , or 100% N ) revealed that the arrangement of the crystallographic
structures of the active phases directly influenced the catalytic performance. It was found that the catalysts heat-
treated in oxidizing atmosphere gave superior catalytic results comparing with the catalysts heat-treated in inert
0.35
V O
14 that contains V+4
atmosphere due to the equilibrium between the crystalline phases MoVO
5
and Mo4.65
gave the best performance, achieving
selectivity of 33.5% towards acrylic acid and 100% conversion of glycerol.
+5
and V . Catalytic oxydehydration at 320 °C under a flow of 100% O
2
1. Introduction
produce superabsorbent polymers, due to its high liquid absorbance
capacity [9]. It can be produced by glycerol dehydration to form
acrolein, followed by oxidation of acrolein to acrylic acid, by means of a
two-step process occurring in a single or double catalytic bed [10,11].
In 2011 Dubois et al. [12,13] patented a method for producing acrylic
acid from glycerol (US 7,910,771 B2 and US 8,212,070 B2), but using a
double bed reactor with two catalysts. Although presenting a high
selectivity, a high catalyst mass in relation to the glycerol feed is used.
Comparing the productivity of acrylic acid per gram of catalyst per
hour, a bifunctional catalyst is more valuable.
In order to perform the glycerol conversion to acrylic acid using a
single catalysts, this catalyst has to be bifunctional, because acid and
redox active sites are needed for the dehydration and oxidation steps,
respectively [10–21]. The use of a bifunctional catalyst to convert
glycerol not only has engineering and economic advantages, but also
minimizes the use of petrochemical resources as raw materials, hence
providing a competitive route for the production of acrylic acid
[12–14].
The depletion of fossil fuel reserves and the need for cleaner energy
sources have intensified the search for alternative fuels that are cleaner
and less harmful to the environment [1–3]. Biodiesel has proved to be a
good option and its production involves a simple process that essen-
tially occurs by the transesterification of oils and fats derived from plant
and animal sources with short chain alcohols, typically methanol, in the
presence of catalysts [4,5]. Glycerol is a major co-product of this
process, equivalent to 10 wt.% of the biodiesel production. The use of
biodiesel increases every year, and it is forecast that in 2020 biodiesel
production will be nearly 36.9 million metric tons, 10 million metric
tons more than the current production. Therefore, for the success of the
biodiesel production chain, it is essential to find pathways to use or
consume glycerol [6,7]. Due to its non-toxicity, glycerol has many
commercial applications, such as in the manufacture of medicines,
foods, beverages, and textiles [8]. Beyond these applications, other
pathways are being studied at the laboratory scale, and this is the case
of glycerol dehydration and oxidation to produce value-added products
such as acrolein and acrylic acid [7].
Approximately 75% of applications in the chemical industry use
mixed oxide catalysts [22], which have been studied for many gas
phase reactions including ethane oxidation [23,24], ethane oxidative
Acrylic acid is an important industrial bulk chemical used mainly to
⁎
Corresponding author.
E-mail address: leandro@iq.unesp.br (L. Martins).
http://dx.doi.org/10.1016/j.cattod.2017.04.006
Received 10 January 2017; Received in revised form 14 March 2017; Accepted 4 April 2017
0
920-5861/ © 2017 Elsevier B.V. All rights reserved.
Please cite this article as: Rasteiro, L.F., Catalysis Today (2017), http://dx.doi.org/10.1016/j.cattod.2017.04.006

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
dehydrogenation [25], selective oxidation of light alkanes [26], pro-
pane oxidation [27,28], glycerol dehydration to acrolein [29], and
more recently the conversion of glycerol to acrylic acid [10,20,30].
Some works have also reported the transformation of glycerol to
acrolein by a liquid phase reaction, giving a great catalytic activity
2.2. Characterization techniques
The evolution of the crystalline phases during heating of the MoV-
and MoV-O samples was analyzed by in situ X-ray diffraction at the
N
2
2
XPD beamline of the Brazilian Synchrotron Light Laboratory (LNLS),
using a synchrotron radiation wavelength of 0.16522 nm. The thermal
treatment was performed from room temperature up to 500 °C, at a
heating rate of 10 °C/min, in an atmosphere of 5% O or 100% N . Data
2 2
were collected in a 2θ range from 7° to 60°.
The crystalline phases formed in the spent catalysts were analyzed
by conventional X-ray diffraction (XRD), using a Rigaku Miniflex 600
diffractometer with CuKα radiation (0.15418 nm) selected with a
curved graphite monochromator. Data were collected in the 2θ range
from 5° to 60°, with a scan step of 0.02° and counting time of 1 s. All the
crystalline phases of the samples were identified with Crystallographica
Search Match software and were then quantified by the Rietveld
[
31,32]. Although materials such as W-Mo-V oxides [17,18], W-V
oxides [20,30], Mo VO /H SiW12 40/Al [16], and W-V-Nb oxides
14] have shown high capacity to convert acrolein to acrylic acid, only
a few studies have reported good selectivity [14,33,34], Omata et al.
34] was one of the few that achieved a high selectivity to acrylic acid
3
x
4
O
2 3
O
[
[
using W-V-Nb mixed oxides (59.2%), but with the modification of the
surface with phosphoric acid. One of the first studies of mixed oxides
containing vanadium and molybdenum catalysts occurred in 1978,
when Thorsteinyon et al. [35] produced a mixed oxide of vanadium,
molybdenum, and niobium to perform the oxidative dehydrogenation
x y z
of ethane. More recently, Mo V O catalysts have gained attention,
®
because these catalysts not only possess the acid sites that are necessary
for the dehydration of glycerol, but also present oxidative potential,
which is essential for the oxidation step [17,31–33]. Furthermore, only
a few studies have used molybdenum and vanadium as the only metals
in catalysts for the oxydehydration of glycerol to acrylic acid
refinement method using TOPAS 4.2 software.
The amounts of V and Mo present in the catalysts were determined
by chemical analysis of the synthesis solutions (mother liquors) using
an Optima 8000 ICP-OES spectrometer.
Thermogravimetric analysis (TGA) of the spent catalysts was
[
19,30,36], so there is still great potential for exploration of the use
2
conducted under an atmosphere of 100% O at a flow rate of
of Mo catalysts in this catalytic reaction.
x
V
y
O
z
100 mL/min, using an SDT Q600 TGA/DSC thermobalance, with
heating in the range 30–600 °C at a rate of 10 °C/min.
In this study, we synthesized mixed Mo-V oxides possessing several
crystalline phases and assessed their performance in the catalytic
oxydehydration of glycerol to acrylic acid. The influence of the
synthesis procedure on formation of the active catalyst components
and, consequently, on the oxydehydration reaction, was explored using
a combination of several techniques: in situ X-ray diffraction (XRD),
ammonia temperature programmed desorption (NH -TPD), elemental
3
chemical analysis, thermogravimetry, and X-ray absorption near vana-
dium K-edge spectroscopy.
X-ray absorption spectroscopy (XAS) measurements at the vana-
dium K-edge were recorded in transmission mode at the XAFS1 beam-
line of the Brazilian Synchrotron Light Laboratory (LNLS). The XAFS1
beamline was equipped with a monochromator (Si (111) for vana-
dium), operated in Bragg mode, for selection of the chosen range of X-
ray energy. The monochromator was calibrated by setting the first
inflection point at the K-edge spectrum of a metallic foil standard
(5465 eV for vanadium). The powdered samples were pressed into
pellets, whose thickness was chosen so that the absorption jump at the
edge was close to 1. Analysis of the XAS data using the Athena graphical
interface software enabled speciation of the vanadium by means of
linear combinations (LC) of the spectra (from −20 to 120 eV around
the edge), using three merged spectra in order to improve the signal-to-
2. Experimental section
2.1. Catalyst preparation
noise ratio. The references used for quantitative analyses were V
, V , and metallic vanadium foil.
The acid sites of the fresh and spent catalysts were measured by
temperature programmed desorption of ammonia (TPD-NH ). In these
2 3
O ,
The synthesis of the mixed oxides was performed based on a
hydrothermal method described previously by Katou et al. [27]. The
main difference was the control of the synthesis atmosphere by the
V
2
O
4
2 5
O
3
addition of N
was dissolved in 13.9 mL of distilled water. Separately, 1.28 g of
NH Mo 24.4H O was dissolved in 13.9 mL of distilled water. The
two solutions were mixed and stirred for 10 min at 25 °C, resulting in a
solution with pH 2.8 and a Mo/(Mo + V) molar ratio of 0.60, which
corresponded to the optimal catalyst composition to give the highest
conversion of glycerol to acrylic acid, as tested in a previous study [36].
Afterwards, the solution was transferred to a 50 mL Teflon vessel
contained in a stainless steel autoclave reactor. The solution occupied
approximately 70% of the capacity of the vessel. The reactor was
2
or O
2
. In a typical synthesis procedure, 0.84 g of VOSO
4
experiments, a mass of 250 mg of sample was placed under a He flow
(60 mL/min) at 300 °C for 1 h, followed by cooling to 100 °C. At this
(
4
)
6
7
O
2
temperature, the sample was exposed to a 60 mL/min flow of 1% NH
in He for 1 h. The excess and physically adsorbed NH were then purged
3
3
at 100 °C under a flow of He during 60 min. Finally, ammonia was
desorbed in a He flow of 60 mL/min, with heating from 100 to 500 °C at
a rate of 10 °C/min. The amount of ammonia desorbed per gram of
sample was calculated from measurements made using a mass spectro-
meter (PrismaPlus QMG 220, Pfeiffer).
The specific surface areas of the samples were determined by
nitrogen sorption measurements performed at liquid nitrogen tempera-
ture (−196 °C), with relative pressure interval between 0.001 and
0.998, using a Micrometrics ASAP 2010 system. The samples were
2 2
hermetically closed and then purged with the required gas (N or O ) in
successive cycles controlled using a valve. Finally, the reactor was
pressurized to 6 bar with the required gas. The reactor was placed in a
glycerol bath heated at 160 °C for 48 h under static conditions. The
solid formed was separated by centrifugation and dried at 60 °C for
−
6
pretreated under vacuum (∼10 × 10 Pa) for 12 h at 200 °C. The BET
method was used to determine the specific surface area (SBET) [37].
1
2 h. The resulting catalyst precursors synthesized in oxygen and
nitrogen atmospheres were denoted MoV-O and MoV-N , respectively.
The final stage in preparation of the catalysts was calcination of the
MoV-O and MoV-N precursors in a fixed bed reactor at 500 °C for 2 h
under a flow of pure O or pure N . The four different samples obtained
were named according to the synthesis and calcination (in parentheses)
atmospheres as MoV-O (O ), MoV-O (N ), MoV-N (O ), and MoV-
(N ). For the purpose of comparison, pure MoO and V oxides
24.4H O and VOSO
2
2
2.3. Catalytic glycerol oxydehydration
2
2
The catalytic reaction was conducted in the gas phase at 290, 320,
and 350 °C, under atmospheric pressure, using a fixed bed reactor.
Firstly, 200 mg of the calcined catalyst was supported on glass wool
placed inside the reactor. The reactor was heated to the reaction
temperature under continuous flows (30 mL/min) of three different gas
2
2
2
2
2
2
2
2
N
2
2
3
2 5
O
were prepared by calcination of the (NH
precursors.
4
)
6
Mo
7
O
2
4
compositions (100% O
2 2 2 2
, 100% N , or 20% O in 80% N ), and was
maintained at this temperature for 15 min before initiating the reaction.
2

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
A solution containing 10 wt.% of glycerol (Sigma-Aldrich, 99%) in
water was fed to the reactor at a flow rate of 0.05 mL/min. The exact
gas composition was glycerol/oxygen/water = 2/28/70 (% mol) when
diffraction files (PDF). Fig. 2 shows the phases present in the as-
synthesized samples and in the materials calcined in oxidizing and inert
atmospheres. Fig. 3 shows the results of the Rietveld refinement, with a
combination of distinct phases found for the samples. The synthesis of
using 100% O
using 100%
% mol) when using air composition. The products were condensed in
a gas-liquid separator kept at 1 °C and aliquots were withdrawn every
h. A previous test was performed to confirm full recovery of the
2
, glycerol/nitrogen/water = 2/28/70 (% mol) when
N
2
and glycerol/oxygen/nitrogen/water = 2/5/23/70
the mixed oxides in the presence of N
orthorhombic Mo VO phase, which has characteristic peaks at 2θ of
around 9°, 24°, and 49°, as reported in other studies [14,25,29,36–40].
In the case of the synthesis in the presence of O , the X-ray diffraction
fingerprint corresponded to three different phases, with predominance
(70%) of Mo0.88 2.94. When the syntheses were performed in the
presence of N , due to its non-reactivity, the atmosphere didn’t
influence the synthesis. However when O was used, the oxidative
atmosphere influenced in the reaction, that is why almost 70% of the
phase Mo0.88 2.94 was formed.
2
resulted in formation of the
(
3
x
1
2
condensable products outflowing from the reactor, by injecting a
solution of 10 wt.% acetaldehyde (b.p. of 20 °C) in water into the
reactor and determining the mass balance in the gas-liquid separator.
The products were quantified using a gas chromatograph (Shimadzu
GC-2014) equipped with a capillary column (RTX-1, 30 m, 0.32 mm,
V O
0.12
2
2
1
μm) and a flame ionization detector (FID). As part of the products was
0.12
V O
converted to CO and CO , a known mass of n-butanol was added as an
2
When the sample was synthesized in an oxidizing atmosphere and
also submitted to an oxidizing atmosphere during the heat treatment,
internal standard before each injection and compared with a calibration
curve of the products. The glycerol conversion (Xgl) and products
distinct phases were formed, with predominance of the Mo4.65
V
0.35
O
14
selectivity (S
tively.
i
) were calculated according to Eq. (1) and (2), respec-
phase (Fig. 3). There was also a substantial quantity of the MoV
2
O
8
phase, which was reported by Possato et al. [36] to perform well in the
oxydehydration of glycerol. The MoV phase was the main phase
formed (80%) when the MoV-O sample was submitted to an inert
2 8
O
nglinput − ngl
output
Xgl (%) =
Si (%) =
× 100
2
input
ngl
(1)
(2)
atmosphere during the heat treatment.
Highly diverse crystalline phases were also obtained for the MoV-N
2
ni
zi
×
nglinput − ngl
output
sample treated under N
oxidizing atmosphere gave the same phases as found in the MoV-O
2
or O
2
atmospheres. The heat treatment in an
2
(O )
zgl
2
where nglinput and ngloutput are the molar flow rates of glycerol in the
input and in the output (mol/min), n is the molar flow of product i, zgl
is the number of carbon atoms in the glycerol molecule, and z is the
number of carbon atoms in the products [38]. The carbon balance to
obtain the selectivity to CO compounds was calculated from the ratio
sample, but in different quantities. There was almost the same amount
i
of the Mo4.65V0.35O14 phase, while the MoV O8 phase decreased to
2
i
around 5% and the MoO3 phase increased to approximately 50%.
According to Liu et al. [16], the use of MoO3 as catalyst in the
x
oxydehydration of glycerol leads to the formation of CO and CO , so
2
of the molar flow of products and unconverted glycerol leaving the
reactor and the molar flow of glycerol fed to the reactor.
the catalytic potential of the catalyst is decreased. Formation of the
reduced forms of the oxides, V O3 (around 90%) and MoO , was
2
2
observed when the MoV-N
2
sample was treated in N
2
. This was the only
3. Results
sample that did not present crystalline phases of the mixed oxides.
The textural characteristics of the calcined catalysts were assessed
3.1. Characteristics of the catalysts
by means of N
2
adsorption-desorption isotherms. Table 3 summarizes
the BET surface area results for the Mo-V mixed oxides, with values in
2
The hydrothermal synthesis procedure enabled a very high solid
yield to be obtained, with consequently a high incorporation of
molybdenum and vanadium atoms in the solid (Table 1). The synthesis
3 2 5
the range 7.4–12.8 m /g. The surface areas of MoO and V O were in
2
the range 0.5–7.6 m /g. The surface area of the mixed oxides was low
due to the absence of a porous network, as confirmed by the nitrogen
adsorption isotherm profiles (Fig. S1).
performed in the presence of O
2
was much more effective, since almost
100% of the transition metal atoms were incorporated in the solid.
3
The TPD-NH profiles of the catalysts are shown in Fig. S2. Table 3
Elemental chemical analysis of the dried samples showed no significant
incorporation of any other element, indicating that the N and S
elements present in the precursors remained in the mother liquor as
provides the quantification of the acid sites, which were classified
according to three strengths: weak (100–200 °C), medium
(200–300 °C), and strong (> 300 °C). The MoO
quantity of acid sites, while the amount decreased dramatically in the
mixed oxides, reaching a similar value as V . The majority of the acid
3
presented a high
+
2−
ionic pairs of NH
4
and SO
4
. Therefore, the precipitate formed was
free from the precursor salts and was composed of mixed oxides of
molybdenum and vanadium.
The mixed oxides formed in the hydrothermal synthesis were
2 2
further calcined in N or O atmospheres, and the changes of the
2 5
O
sites in the Mo-V oxides possessed weak and medium strengths. The
samples calcined under nitrogen had slightly higher quantities of acid
sites, although the differences were not significant.
crystalline phases during the heat treatment were followed using in situ
X-ray diffraction measurements (Fig. 1). The crystalline phases of the
catalysts produced in the hydrothermal synthesis or after the calcina-
tion procedure were highly dependent on the atmosphere used. Table 2
lists the crystalline structures that were detected during the various
steps of the study, together with the corresponding JCPDS powder
3.2. Catalytic results
Glycerol oxydehydration performed under different conditions
enabled elucidation of the effects of the conditions employed in the
hydrothermal syntheses and for calcination of the molybdenum and
vanadium mixed oxides on the acrylic acid production. The mixed
Table 1
oxides presented superior performance, compared to the pure V
2 5
O and
Hydrothermal synthesis of MoV mixed oxides in N
yields calculated after chemical analysis of the mother liquors.
2 2
and O atmospheres, and the solid
MoO oxides. However, before discussing the results in terms of the
3
active phases that contributed to the conversion of glycerol and the
reaction conditions that maximized production of acrylic acid, there are
several issues that need to be considered:
Sample
pHfinal Solid yield Degree of incorporation
%)
Mo/(Mo + V) molar
ratio in the solid
(
%
Mo
% V
(
a) It is quite possible that the samples contained amounts of amor-
phous phases of molybdenum and vanadium oxides. For example,
MoV-N
MoV-O
2
2.5
1.8
72.4
98.1
70.6
98.8
85.2
97.3
0.55
0.60
2
2 2
Fig. 3 indicates that the MoV-N (N ) sample contained more than
3

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
2 2 2 2 2 2
Fig. 1. In situ XRD patterns of samples (a) MoV-N and (b) MoV-O heat-treated under an oxidizing atmosphere (5% O in N ); and (c) MoV-N and (d) MoV-O heat-treated under an inert
atmosphere (100% N ). *Cooled sample.
2
8
0% crystalline V
2
O
3
, while the chemical analysis revealed that this
320 °C. These results could be explained by the fact that in the
temperature of 290 °C the acrylic acid molecules are more adsorbed
on the surface of the catalyst that resulted in an additional oxidation to
form COx compounds, justifying the decrease in acrylic acid formation
(Fig. S3). The increase of the temperature to 320 °C aided in the
desorption of acrylic acid molecules from the surface of the catalyst.
This temperature seems to be ideal for this system, as the temperature is
sample had a Mo/(Mo + V) molar ratio of 0.55 (Table 1). The
activity of the amorphous phases in this sample is described in the
discussion of the catalytic tests.
(
b) The formation of several phases of Mo-V oxides can lead to
interfacial surfaces that create a unique reactivity. In order to
elucidate this surface phenomenon, sophisticated techniques, in
combination with theoretical studies, are needed to reveal the
configuration and structure/morphology of the surface, as well as
the composition of adsorbed species [42].
increased to 350°C part of the glycerol is decomposed to CO
dehydration to acrolein.
X
before
Fig. 4 compares the catalytic activities of the samples of the mixed
(
c) The synthesis of pure-phase samples was attempted in order to
correlate the catalytic performance individually, but this proved to
oxides of molybdenum and vanadium at 320 °C, during the first hour.
The catalysts showed a selectivity to CO in a range between 59.2-
x
be unfeasible due to the probable formation of V
products of decomposition of the mixed oxides.
2
O
5
and MoO
3
as
73.9% and the lowest selectivities belong to the catalysts heat-treated in
oxygen. These materials are quite oxidative so is expected that they
present a greater amount of CO
best performance in converting glycerol to acrylic acid, with the lowest
formation of CO and CO . As a general tendency, the MoV-N (O ) and
MoV-O (O ) catalysts calcined in an oxygen atmosphere provided
better selectivity towards acrylic acid and lower selectivity towards
CO , while the MoV-N (N ) and MoV‐O (N ) samples calcined in
x 2 2
. The MoV-O (O ) sample showed the
Firstly, in the catalytic tests of glycerol oxydehydration (Scheme 1),
the MoV-O (O ) sample was submitted to three different temperatures
290, 320, and 350 °C) in a 6 h reaction, using a gas flow of 20% O in
, in order to identify the best conditions for conversion of glycerol to
acrylic acid. The differences for each temperature can be seen in Fig. S3
in terms of yield to acrylic acid during the reactions, the error in the
measurements are also shown. The most effective temperature was
2
2
2
2
2
(
N
2
2
2
2
X
2
2
2
2
nitrogen exhibited similar selectivity towards acrylic acid as the pure
oxides.
Table 2
Crystalline phases and stoichiometric oxidation states of the vanadium and molybdenum oxides.
Crystalline compounds
JCPDS PDF
Source
Temperature (°C)
Atmosphere
Stoichiometric oxidation state
V
Mo
(
NH
VOSO
Mo
MoO
MoVO
4
)
6
Mo
7
O
24
70-0957
19-1412
–
Precursor
Precursor
Synthesis
Synthesis/Calcination/Post-reaction
Synthesis/Post-reaction
Synthesis
Calcination
Calcination
RT
RT
160
160/500/320
160/320/350
160
500
500
Air
Air
–
4
–
–
4
5
5.6
5
–
6
–
–
6
6
6
5.6
6
4
–
4
3
VO
x
N
2
O
2
O
2
O
2
O
2
O
2
N
2
N
2
N
2
N
2
3
89-5108
77-0649
81-2414
31-1437
74-0050/74-2366
32-0671
71-0347
82-0661
72-0524
/20% O
/20% O
2
5
2
Mo0.88
Mo4.65
MoV
MoO
V
VO
V
V
0.12
O
O
2.94
0.35
14
2
O
8
2
Calcination/Post-reaction
Calcination
Calcination
500/350
500
500
/20% O
2
O
2 3
3
4
3.3
2
–
–
V
O
3 5
Calcination
500
4

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
Fig. 2. XRD and indexation of the phases for samples MoV-O
2
, MoV-O
2
(O
2
), MoV-O
2
(N
2
), MoV-N
2
, MoV-N
2
(O
2
), and MoV-N
2
2
(N ). The net results of the Rietveld analyses are shown in
red and the deviations from the measured diffractograms are depicted in blue. (For interpretation of the references to color in this figure legend, the reader is referred to the web version
of this article.)
2 2
Fig. 4. Catalytic results for a 1 h reaction at 320 °C under a flow of 20% O in N using the
mixed oxides of molybdenum and vanadium.
The influence of the gas flow composition on the reaction was also
studied using the MoV-O (O ) catalyst in a 6 h reaction at 320 °C
Fig. 5). For the gas flow composed of 100% O , the highest selectivity
towards acrylic acid was achieved after 6 h of reaction. No acrylic acid
was formed when 100% N was used. Lower formation of CO was
observed when 100% N was used, because the absence of an oxidizing
atmosphere prevented complete oxidation of the products. However,
when O was present in the reaction atmosphere, the potential for
formation of oxidation products increased, with a higher O content
resulting in formation of more CO
The importance of feeding O is related to the catalytic redox cycle:
Fig. 3. Crystalline phases found for the mixed oxides of molybdenum and vanadium.
2
2
(
2
Table 3
BET surface areas and quantification of the acid sites determined by means of NH
3
-TPD.
2
x
2
Catalyst TPD-NH BET
3
surface
area
2
2
g−1
(
m
)
2
aWeak
μmol g−1)
bMedium
cStrong
Total
(μmol g−1)
(μmol g−1)
(μmol g−1)
x
.
(
2
MoO
3
51.2(12%)
6.7(10%)
17.5(45%)
10.8(44%)
26.4(58%)
45.4(60%)
205.0(48%)
25.5(38%)
19.8(51%)
9.6(39%)
13.5(30%)
21.7(29%)
170.8(40%)
34.8(52%)
1.6(4%)
4.4(17%)
5.3(12%)
8.5(11%)
427.0
67.0
38.9
24.7
45.2
75.6
0.5
7.6
11.4
7.4
10.8
12.8
acrolein is formed on the acid sites and is then promptly oxidized to
acrylic acid by reducing a vanadium atom of the oxide lattice; the
oxygen vacancy and the vanadium oxidation state are restored by the
V
O
2 5
MoV-O
MoV-N
MoV-N
MoV-O
2
(O
2
(O
2
(N
2
(N
2
)
)
)
)
2
O
2
fed to the reactor. The dynamic reduction and oxidation of
2
5+
4+
4+
5+
vanadium atoms (from V
to V , and then from V
to V ) and
2
the oxidation of acrolein occur according to the Mars-van-Krevelen
redox process [43]. The oxidation state of molybdenum (+6) in the
lattice of the mixed oxide was invariant, as found in a previous study
Acid sites strength: a. Weak (100–200 °C); b. Medium (200–300 °C); c. Strong (> 300 °C).
Scheme 1. Glycerol oxydehydration: dehydration to acrolein followed by oxidation of acrolein to acrylic acid.
5

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
x 2 2
Fig. 5. (a) Glycerol conversion and (b) acrylic acid (closed symbols) and CO (open symbols) selectivity for a 6 h reaction at 320 °C using the MoV-O (O ) sample under three different gas
flows (20% O in N , 100% O , or 100% N ).
2 2 2 2
2 2 2 2 2 2 2 2 2
Fig. 6. Catalytic results for a 6 h reaction at 320 °C under a 100% O gas flow for the samples (a) MoV-O (O ), (b) MoV- N (O ), (c) MoV-O (N ), and (d) MoV-N (N ).
[
36]. As the Mo6+ cation has a similar size as V4+, it has an important
a smaller quantity of MoVO
observed for the catalysts heat-treated in an inert atmosphere, which
presented mainly the MoVO phase. For the MoV-O (O ) catalyst that
was submitted to a flow of 100% O , the quantitative crystallographic
phase analysis revealed 44.3% of MoVO
The performance of the samples showed a correlation with the MoVO
5
. This behavior was the opposite of that
role in accommodating vanadium atoms in the lattice, and conse-
quently in improving the dynamic change of the vanadium oxidation
state and the creation of oxygen vacancies during the catalytic steps.
Fig. 6 shows the catalytic results for the samples under a flow of
5
2
2
2
5
and 35.6% of Mo4.65
V
0.35
O
14
.
1
00% O
2
at 320 °C in a 6 h reaction. As previously, there was
5
/
enhancement of both the selectivity towards acrylic acid and glycerol
conversion, while the samples calcined in an inert atmosphere showed
the poorest results. Table S1 shows the conversion and selectivities to
the catalysts after 6 h on stream. The quantity of other products formed
during the reaction (acetol, propanal, allyl alcohol, 3-hydroxypropanal
and propionic acid) was less than 4%, for that reason these products are
not represented in Fig. 6.
0.35
Mo4.65V O14 molar ratio, as depicted in Fig. 7. The highest selectivity
towards acrylic acid was achieved at a ratio between the two phases of
about 1.
The elemental analysis found negligible amounts of carbon and
hydrogen in the samples, evidencing that coke was not formed in the
catalysts. The thermogravimetric analysis of the spent catalysts (Fig. 8)
showed that all the catalysts presented a weight loss in the 100–200 °C
range, corresponding to elimination of physisorbed water, and then a
weight gain at temperatures above 300 °C, due to the reoxidation of
3.3. Post-reaction characteristics
4+
5+
V
to V . Therefore, the temperature at which the oxidation started
was very important, because it determined the ease with which the
active redox sites were restored. The catalysts heat-treated in N
presented higher reoxidation temperatures (above 350 °C), compared
to the catalysts heat-treated in O (reoxidation temperatures below
20 °C).
The X-ray absorption near vanadium K-edge spectroscopy (XANES
2 2
After the 6 h catalytic tests at 320 °C under a flow of 20% O in N , a
post-reaction investigation of the spent samples was performed using
thermogravimetric measurements, Rietveld refinement of the X-ray
2
diffraction patterns, TPD-NH
edge spectroscopy. Measurements were also performed for the MoV-
(O ) sample after a 6 h reaction at 320 °C under 100% O (this
sample was denoted MoV-O (O )-100%O ). The Rietveld refinement
results are shown in Table 4 and Fig. S4. All the samples formed the
same families of crystallographic phases (MoVO 14, and
MoO ), due to structural reorganization of the mixed oxides under the
glycerol flow. The catalysts that were previously heat-treated in an
oxidizing atmosphere contained mainly the Mo4.65 14 phase, with
3
and X-ray absorption near vanadium K-
2
3
O
2
2
2
region) results for the as-synthesized and spent catalysts are shown in
Fig. 9. The linear combination fittings of the XANES spectra were
2
2
2
performed using V
and showed increased amounts of V
2
O
5
, V
2
O
4
, and V
2
O
3
4
spectra as references (Fig. S5)
5 0.35
, Mo4.65V O
+
for all the spent catalysts,
(N )-
2
)-20% catalysts presented the
3
compared to the fresh ones (Table 4 and Fig. 9). The spent MoV-N
0%, MoV-N (O )-20%, and MoV-O (N
2
2
2
2
2
2
V
0.35
O
6

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
Table 4
4
+
Quantities of the crystalline phases in the spent catalysts, determined by Rietveld refinement, quantities of V
in the fresh and spent catalysts, determined from V K-edge XANES
measurements, and yields of acrylic acid after 6 h of reaction.
Spent catalyst
Phase content (%)
V4+ in the fresh catalysts (%)
V4+ in the spent catalysts (%)
Yield of acrylic acid after 6 h (%)
MoVO
5
Mo4.65
V
0.35
O
14
MoO
3
MoV-O
MoV-O
MoV-N
MoV-O
MoV-N
2
2
2
2
2
(O
(O
(O
(N
(N
2
2
2
2
2
)-100%O
2
37.8
38.3
30.9
71.1
55.6
35.1
49.4
61.6
27.0
21.0
27.1
12.3
7.5
1.9
23.4
36
36
58
55
43
57
89
80
70
29.6
22.9
9.8
10.3
11.3
)-20%O
)-20%O
)-20%O
)-20%O
2
2
2
2
18 (82% of V3+)
2 2
Reaction conditions: 320 °C and 20% O or 100% O as the gas flow.
compared to the fresh one, 21.9 and 38.9 μmol/g, respectively.
4. Discussion
By means of in situ X-ray diffraction patterns, it was possible to
follow the formation of diverse crystalline phases of mixed oxides,
which were dependent on the atmospheres used during synthesis and
calcination. Fig. 3 indicates that the presence of mixed oxide phases is
more abundant in the catalysts treated in oxygen atmosphere (synthesis
or calcination). This is coherent when we compare the oxidation states
4+
4
of vanadium in precursor salt, VOSO (V ), and in the formed mixed
oxide phases, where, basically, all of them presents vanadium in high
oxidation states. The sample synthesized in oxygen atmosphere, which
presents approximately 70% of the mixed oxide phase Mo0.88V O ,
0.12 2.94
when calcined in oxidant or inert atmosphere, despite undergoing a
phase restructuring, tends to preserve a high amount of mixed oxide
phases. When synthesized in nitrogen atmosphere, after calcination in
Fig. 7. Correlation between the acrylic acid selectivity in the last hour of the reaction and
the ratio of the crystallographic phases (MoVO
catalysts.
5 0.35
/Mo4.65V O14) found for the spent
oxidant atmosphere the sample show the formation of Mo4.65
0.35 14
V O
and MoV mixed oxide phases, but a considerable quantity of MoO
2
O
8
3
phase is present in the material. After calcination in inert atmosphere
this sample present just pure oxides phases with vanadium and
molybdenum in reduced forms. The mixed phases observed in this
work are, basically, formed by incorporation of vanadium atoms in the
structure of molybdenum oxides and, comparing these results can be
noted that, in the mild conditions of synthesis proposed in this work,
the vanadium in high oxidation states, that possess a smaller ionic
radius compared do reduced forms of vanadium, like V , can easily be
introduced and accommodated in molybdenum oxide structures. The
oxidant atmosphere plays an important role so that structures of mixed
highest quantities of V4+. Curiously, among all the catalysts, these
materials had the worst yield towards acrylic acid in the last hour of the
2 2 2 2
reaction. The spent MoV-O (O )-20% and MoV-O (O )-100% catalysts
4
+
showed lower quantities of V , with values of 57% and 43%,
respectively, and higher yield towards acrylic acid in the last hour of
the reaction, as shown in Table 3.
Temperature programed desorption of ammonium of the spent
4+
catalyst MoV-O
the acidity comparing with the fresh catalyst MoV-O
in Fig. 10 shows that the total acids sites of the spent catalyst decreases
2
(O
2
)-100% was conducted to observe any changes in
2
(O ). The results
2
5
oxides with reduced forms of vanadium, like MoVO , are not preferen-
Fig. 8. Thermogravimetric and derivative analysis curves for the spent catalysts.
7

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
nal, which consequently dehydrates to acrolein [41,42]. In the glycerol
oxydehydration, the formation of CO and CO
2
is commonly associated
with (Mo,V)O species, the pure oxide phases, that promote complete
X
oxidation of intermediates such as acetol, 3-hydroxypropane, and
acrylic acid [16,45].
For all the samples, the change of the gas flow composition to 100%
O
2
caused increases in the conversion of glycerol and the selectivity
towards acrylic acid. The use of pure O enabled more effective
2
reestablishment of the oxygen vacancies, previously described in the
Mars-van-Krevelen oxidation mechanism. The active sites of the
catalyst were therefore maintained and were readily available to
oxidize acrolein, besides inhibiting the deposition of coke [43–47].
However, the excess of molecular oxygen alone was not sufficient to
explain the increase in catalytic performance. In a previous study, Wang
et al. [48] reported a high selectivity towards acrolein, with low
formation of acrylic acid, in the glycerol reaction over vanadium-based
catalysts, with a gas flow rich in O . Hence, the role of the structure and
2
composition of the mixed oxides in directing the reaction to certain
products is clearly important.
The starting composition of the mixed oxides in the fresh catalysts is
another important point to be highlighted, because it determined the
structural reorganization of the phases during the reaction and the
capacity for changes in the V4+/V
5+
oxidation states, as measured by
X-ray absorption near vanadium K-edge spectroscopy. In the course of
the reaction, the catalysts were restructured, with the Mo4.65
MoVO , and MoO phases being observed in all cases (Fig. S4).
V
0.35 14
O ,
5
3
However, the spent catalysts showed different equilibria between the
phases (Table 4).
Both acid and redox active sites are necessary to accomplish the
oxydehydration of glycerol, but acrylic acid production seems to
depend mainly on the establishment of a suitable equilibrium of the
reduced and oxidized phases (Fig. 7). Maximum productivity was
achieved with a ratio between the phases of about 1. The XANES
measurements (Table 4, Fig. 9) showed that when treated using a gas
Fig. 9. X-ray absorption near vanadium K-edge spectroscopy for the mixed oxides of
molybdenum and vanadium.
flow of 100% O
2
at 320 °C, the MoV-O
2 2
(O ) catalyst presented 43% of
4+
4+
V
. Although this catalyst contained a lower amount of V
,
compared to the other catalysts, it provided the highest selectivity
towards acrylic acid. These results suggest that V4+ in excess did not
benefit the formation of acrylic acid. Nonetheless, equilibrium between
4
+
5+
V
and V
presence of V
acrylate adsorption is impaired, it could be further oxidized to CO
49,50].
The thermogravimetric analysis of the spent catalysts (Fig. 8)
revealed differences in the initial temperature of reoxidation between
the catalysts heat-treated in O and N . For the catalysts heat-treated in
an inert atmosphere, the oxidation of V
temperature than the reaction temperature (320 °C), which must have
been associated with the higher quantity of the reduced MoVO phase.
For the spent MoV-O (O )-20% and MoV-N (O )-20% catalysts, the
seemed to be necessary. It has been reported that the
4+
stabilizes the intermediate acrylate ion and that if
2
[
2
2
4+
5+
to V
began at a higher
Fig. 10. Temperature programmed desorption of ammonia for the fresh and spent
catalyst MoV-O
2
(O
2
).
5
2
2
2
2
temperature of re-oxidation was significantly lower than the reaction
tially formed when an inert atmosphere is used. Due to the varied phase
composition, it was not possible to establish a straightforward relation-
ship between the structure of the mixed oxide and the catalytic activity.
However, the most active catalysts, calcined in oxygen, presented
temperature, favoring the oxidation of V4+–V , associated with
5+
greater formation of the oxidized Mo4.65
0.35
V O14 phase. The spent
MoV-O (O )-100% catalyst presented a reoxidation temperature close
2
2
to the reaction temperature, which could explain the similar quantities
of the reduced and oxidized phases.
approximately 45% of the Mo4.65
was absent in the catalysts calcined in nitrogen. Formation of the
14 crystalline phase was observed elsewhere in MoV/SiW/
catalyst synthesis and was identified as the most active phase in
0.35
V O14 crystalline phase, which
The TPD-NH
3 2 2
analysis of the fresh and spent catalyst MoV-O (O )
0.35
Mo4.65V O
(
Fig. 10) unveil that the acidity after the 6-h reaction decreases, the
2 3
Al O
−1
total amount of acid sites went from 38.9 to 21.9 μmol NH
expected because in XANES measurements the quantity of
increased in the spent catalyst comparing to the fresh one. More V
3
.g . This is
the oxidation of acrolein to acrylic acid [16].
+
+
4
4
V
Although the catalysts did not possess high quantities of total acid
sites, medium and strong acid sites were present, notably in the samples
calcined in oxygen (> 50%), which could be beneficial in the dehydra-
tion step. Several studies have shown that adequate strength of the acid
sites is important because the adsorption of glycerol on relatively
moderate strength acid sites favors the production of 3-hydroxypropa-
leads to a decrease of the acidity of the catalyst. According to Possato
et al. [36] and Takita et al. [51] the reason to the reduction of acid sites
is the influence of the water vapor during the reaction, where the water
molecule can dissociate in the surface of the catalyst making more
+
5
oxygen available for the reaction, what reduces the V
sites and
8

L.F. Rasteiro et al.
Catalysis Today xxx (xxxx) xxx–xxx
Table 5
XPD-18893 and XPD-20150244) and XAFS 1 (20160929) beamlines.
Comparison of glycerol oxydehydration into acrylic acid, glycerol conversion and CO
selectivity with different heterogeneous catalysts in the literature.
X
Appendix A. Supplementary data
Catalyst
Glycerol
conversion
Acr. acid
selectivity
(%)
CO
selectivity
(%)
X
Ref.
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.cattod.2017.04.006.
(
%)
MoV/SiW/Al
2
O
3
100
100
95
100
100
100
99.6
94
100
100
100
100
93.6
100
12
17
55.6
> 60
40
54.7
47
35.1
44.8
20
42.7
35
[16]
[15]
[30]
[19]
[20]
[19]
[19]
[36]
[34]
[18]
[34]
[33]
[43]
This
work
References
V
2
O
5
-MFI
Mo-V-O
V-O
20.1
23.7
26
26.3
28.4
32
33.7
51
59.2
60
W
-
[1] S. Marx, Fuel Process. Technol. 151 (2016) 139–147.
[2] M.R. Nanda, Z. Yuan, W. Qin, C.C. Xu, Catal. Rev. Sci. Eng. 58 (3) (2016) 309–336.
W-V-O
Mo-V-O
Mo-V-Te-Nb-O
Mo-V-O
W-V-Nb-O
W-V-Nb-O
[3] M.R. Nanda, Y. Zhang, Z. Yuan, W. Qin, H.S. Ghaziaskar, C. Xu, Renew. Sustain. Energy
Rev. 56 (2016) 1022–1031.
[4] R. Estevez, S. Lopez-Pedrajas, F. Blanco-Bonilla, D. Luna, F.M. Bautista, Chem. Eng. J. 282
(
2015) 179–186.
[
[
5] V.K. Garlapati, U. Shankar, A. Budhiraja, Biotechnol. Rep. 9 (2016) 9–14.
6] S. Lemos, J. Soares, L. Pires, Quim. Nov. 33 (2010) 1682–1685.
2
.5wt%PO
4
/W2.2
V
0.4Nb2.4
O
14
22.3
13.2
10
[7] B. Katryniok, S. Paul, F. Dumeignil, ACS Catal. 3 (2013) 1819–1834.
[8] A. Talebian-Kiakalaieh, N.A.S. Amin, H. Hezaveh, Renew. Sustain. Energy Rev. 40 (2014)
28–59.
Cs(VO)0.2(PMo)0.5(PW)0.5
Vanadosilicate
Mo-V-O
85.4
33.5
[9] A. Talebian-Kiakalaieh, N.A.S. Amin, Z.Y. Zakaria, J. Ind. Eng. Chem. 34 (2016) 300–312.
59
[
[
10] A. Chieregato, F. Basile, P. Concepción, S. Guidetti, G. Liosi, M.D. Soriano, C. Trevisanut,
F. Cavani, J.M. López Nieto, Catal. Today 197 (2012) 58–65.
11] S. Thanasilp, J.W. Schwank, V. Meeyoo, S. Pengpanich, M. Hunsom, J. Mol. Catal. A
Chem. 380 (2013) 49–56.
decrease the catalyst acidity.
[12] J.-L. Dubois, C. Duquenne, W. Holderich, Method for producing acrylic acid from
glycerol, US 7,910,771 B2, (2011).
Table 5 shows a comparison of this work with others in the
literature for the glycerol oxydehydration to acrylic acid using solid
acid catalysts. Unfortunately, all the catalysts independent of the
reaction condition (temperature or reaction gas composition), pre-
sented a high formation of CO
due to the decomposition of glycerol and the formed products, what is
inevitable as reported in several studies. Although CO formation for
our best catalysts is high (around 60%), acrylic acid was the most
abundant condensable product (33.5%) with a low formation of minor
products (< 4%) and a 100% conversion, not presenting deactivation
after 6 h. Possato et al. [36] in his work showed a Mo-V-O based
catalysts for the glycerol oxydehydration to acrylic acid with 32% of
[
[
[
13] J.-L. Dubois, G. Patience, Method for preparing acrylic acid from glycerol, US 8,212,070
B2, (2012).
14] A. Chieregato, M.D. Soriano, F. Basile, G. Liosi, S. Zamora, P. Concepción, F. Cavani,
J.M. López Nieto, Appl. Catal. B Environ. 150–151 (2014) 37–46.
15] L.G. Possato, W.H. Cassinelli, T. Garetto, S.H. Pulcinelli, C.V. Santilli, L. Martins, Appl.
Catal. A Gen. 492 (2015) 243–251.
X
up to 50% and it is believed that is
[
[
16] L. Liu, B. Wang, Y. Du, Z. Zhong, A. Borgna, Appl. Catal. B Environ. 174–175 (2015) 1–12.
17] Y.S. Yun, K.R. Lee, H. Park, T.Y. Kim, D. Yun, J.W. Han, J. Yi, ACS Catal. 5 (2014) 82–94.
X
[18] A. Chieregato, M.D. Soriano, E. García González, G. Puglia, F. Basile, P. Concepción,
C. Bandinelli, J.M. López Nieto, F. Cavani, ChemSusChem 8 (2015) 398–406.
[
[
19] J. Deleplanque, J.L. Dubois, J.F. Devaux, W. Ueda, Catal. Today 157 (2010) 351–358.
20] M.D. Soriano, P. Concepción, J.M.L. Nieto, F. Cavani, S. Guidetti, C. Trevisanut, Green
Chem. 13 (2011) 2954–2962.
[21] A. Witsuthammakul, T. Sooknoi, Appl. Catal. A Gen. 413–414 (2012) 109–116.
[22] G. Centi, S. Perathoner, Compr. Inorg. Chem. II, (2013), pp. 153–184.
[23] S. Ishikawa, X. Yi, T. Murayama, W. Ueda, Appl. Catal. A Gen. 474 (2014) 10–17.
selectivity to acrylic acid and 20% of CO
catalyst deactivates and in the eighth hour the acrylic acid selectivity
achieves around 5% and CO selectivity 80%.
X
selectivity, however the
[24] S. Ishikawa, D. Kobayashi, T. Konya, S. Ohmura, T. Murayama, N. Yasuda, M. Sadakane,
W. Ueda, J. Phys. Chem. C 119 (2015) 7195–7206.
[25] B. Chu, L. Truter, T.A. Nijhuis, Y. Cheng, Appl. Catal. A Gen. 498 (2015) 99–106.
X
[26] S. Ishikawa, W. Ueda, Catal. Sci. Technol. 6 (2016) 617–629.
[
[
[
[
[
27] T. Katou, D. Vitry, W. Ueda, Chem. Lett. 32 (2003) 1028–1029.
28] W. Ueda, D. Vitry, T. Katou, Catal. Today 99 (2005) 43–49.
29] Y. Choi, H. Park, Y.S. Yun, J. Yi, ChemSusChem 8 (2015) 974–979.
30] L. Shen, H. Yin, A. Wang, X. Lu, C. Zhang, Chem. Eng. J. 244 (2014) 168–177.
31] L. Shen, H. Yin, A. Wang, Y. Feng, Y. Shen, Z. Wu, T. Jiang, Chem. Eng. J. 180 (2012)
5. Conclusions
Mixed oxides of vanadium and molybdenum were prepared by a
277–283.
hydrothermal synthesis procedure and were heat-treated in different
atmospheres. A correspondence was observed between the structural
compositions of the catalysts formed and their catalytic behavior in the
one-step oxydehydration of glycerol to acrylic acid. The catalysts heat-
treated in an oxidizing atmosphere showed the best catalytic perfor-
mance in terms of selectivity towards acrylic acid. Reactions carried out
[
32] L. Shen, H. Yin, A. Wang, X. Lu, C. Zhang, F. Chen, Y. Wang, H. Chen, J. Ind. Eng. Chem.
20 (2014) 759–766.
33] X. Li, Y. Zhang, ACS Catal. 6 (5) (2016) 2785–2791.
[
[
34] K. Omata, K. Matsumoto, T. Murayama, W. Ueda, Catal. Today 259 (2016) 205–212.
[35] E.M. Thorsteinson, T.P. Wilson, F.G. Young, P.H. Kasai, J. Catal. 52 (1978) 116–132.
[
36] L.G. Possato, W.H. Cassinelli, C.I. Meyer, T. Garetto, S.H. Pulcinelli, C.V. Santilli,
L. Martins, Appl. Catal. A Gen. 532 (2017) 1–11.
37] G. Leofanti, M. Padovan, G. Tozzola, B. Venturelli, Catal. Today 41 (1998) 207–219.
[
at 320 °C, with a gas flow composed of 100% O
2
, were found to give the
[38] W. Suprun, M. Lutecki, T. Haber, H. Papp, J. Mol. Catal. A Chem. 309 (2009) 71–78.
[
[
39] S. Ishikawa, X. Yi, T. Murayama, W. Ueda, Catal. Today 238 (2014) 35–40.
40] M. Sadakane, N. Watanabe, T. Katou, Y. Nodasaka, W. Ueda, Angew. Chem. Int. Ed. 46
(2007) 1493–1496.
highest selectivity towards acrylic acid and the best long-term catalyst
stability and the catalysts heat-treated in oxygen presented the best
selectivities, achieving 33.5% of acrylic acid. The explanation for this
[41] S. Ishikawa, M. Tashiro, T. Murayama, W. Ueda, Cryst. Growth Des. 14 (2014)
4553–4561.
was that a gas flow rich in O
formation of phases of mixed oxides containing V
and Mo4.65 14, respectively). The evidence obtained in this study
2
resulted in a suitable equilibrium in the
[42] C. Chiu, T. Vogt, L. Zhao, A. Genest, N. Rösch, Dalton Trans. 44 (2015) 13778–13795.
4
+
5+
and V
(MoVO
5
[43] A. Silva Paula, L.G. Possato, D. Ratero, J. Contro, K. Keinan-Adamsky, R. Soares,
G. Goobes, L. Martins, J.G. Nery, Microporous Mesoporous Mat. 232 (2016) 151–160.
V O
0.35
[44] L.H. Vieira, K.T.G. Carvalho, E.A. Urquieta-González, S.H. Pulcinelli, C.V. Santilli,
concerning the different phases of mixed oxides and the performance of
the catalysts in the glycerol oxydehydration highlights the importance
of controlling the synthesis and calcination atmospheres.
L. Martins, J. Mol. Catal. A Chem. 422 (2015) 148–157.
[45] T. Ressler, A. Walter, Z. Huang, W. Bensch, J. Catal. 254 (2008) 170–179.
[46] J. Tichý, Appl. Catal. A Gen. 157 (1997) 363–385.
[47] L. Giebeler, P. Kampe, A. Wirth, A.H. Adams, J. Kunert, H. Fuess, H. Vogel, J. Mol. Catal.
A Chem. 259 (2006) 309–318.
[
[
48] F. Wang, J. Dubois, W. Ueda, J. Catal. 268 (2009) 260–267.
49] T.G. Kuznetsova, G.K. Boreskov, T.V. Andrusttkevich, Y.A. Grigorkina, N.G. Maksimov,
I.P. Olenkova, L.M. Plyasova, T.P. Gorshkova, Catal. Lett. 19 (1982) 405–409.
Acknowledgements
This work was supported by the Brazilian agencies CNPq (152447/
015-6), CNPq/PVE (401679/2013-6), and FAPESP (2014/20116-6
and 2016/10597-2). The authors also thank the Brazilian Synchrotron
Light Laboratory (LNLS) in Campinas for use of the XPD (proposals
[50] T.G. Kuznetsova, G.K. Boreskov, T.V. Andrushkevich, L.M. Plyasova, N.G. Maksimov,
I.P. Olenkova, Catal. Lett. 12 (1979) 531–536.
2
[51] Y. Takita, J. Catal. 1 (1978) 95–101.
9