Porous zirconium phosphate supported tungsten oxide solid acid catalysts for the vapour phase dehydration of glycerol

Article abstract of DOI:10.1016/j.molcata.2014.09.018

Solid acid catalysts containing WO_x/ZrP with varying the active component loading (5-40 wt%) on porous ZrP support have been investigated for the vapour phase dehydration of glycerol to acrolein. The catalyst containing 30 wt% WO_x/ZrP has shown high selectivity to acrolein (about 82%) with a total conversion of glycerol at 300 °C in the presence of water. The calcined catalysts were characterized by X-ray diffraction, pore size distribution, FT-IR, UV-DRS, pyridine absorbed FT-IR and NH₃-temperature programmed desorption to elucidate the structural and acidic properties of the catalysts. The XRD results suggest that WO_x is found to be present in a highly dispersed state at lower loadings (<30 wt%) and crystalline WO_x at higher loadings. NH₃-TPD and FT-IR results of adsorbed pyridine suggest that the total acidity and number of Br?nsted acidic sites are found to increase with WO₃ loading on the support. Further, there is no significant change of acidity was noticed at higher loadings. The conversion of glycerol and the selectivity toward acrolein mainly depend on the fraction of moderate acidic sites and Br?nsted acidic sites. The Glycerol dehydration functionalities are explained in terms of the acidity and structural properties of WO_x/ZrP catalysts. In addition, the positive effect due to addition of air to N₂ feed flow suppresses the coke formation on the surface of catalyst was also investigated during dehydration reaction.

Full text of DOI:10.1016/j.molcata.2014.09.018

Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Porous zirconium phosphate supported tungsten oxide solid acid
catalysts for the vapour phase dehydration of glycerol
Ginjupalli Srinivasa Rao, N. Pethan Rajan, M. Hari Sekhar,
S. Ammaji, Komandur V.R. Chary∗
Catalysis Division, Indian Institute of Chemical Technology, Hyderabad, India
a r t i c l e i n f o
a b s t r a c t
Article history:
Solid acid catalysts containing WOx/ZrP with varying the active component loading (5–40 wt%) on porous
ZrP support have been investigated for the vapour phase dehydration of glycerol to acrolein. The catalyst
containing 30 wt% WOx/ZrP has shown high selectivity to acrolein (about 82%) with a total conversion of
glycerol at 300 ◦C in the presence of water. The calcined catalysts were characterized by X-ray diffrac-
tion, pore size distribution, FT-IR, UV-DRS, pyridine absorbed FT-IR and NH3-temperature programmed
desorption to elucidate the structural and acidic properties of the catalysts. The XRD results suggest that
WOx is found to be present in a highly dispersed state at lower loadings (<30 wt%) and crystalline WOx
at higher loadings. NH3-TPD and FT-IR results of adsorbed pyridine suggest that the total acidity and
number of Brønsted acidic sites are found to increase with WO3 loading on the support. Further, there is
no significant change of acidity was noticed at higher loadings. The conversion of glycerol and the selec-
tivity toward acrolein mainly depend on the fraction of moderate acidic sites and Brønsted acidic sites.
The Glycerol dehydration functionalities are explained in terms of the acidity and structural properties
of WOx/ZrP catalysts. In addition, the positive effect due to addition of air to N2 feed flow suppresses the
coke formation on the surface of catalyst was also investigated during dehydration reaction.
© 2014 Elsevier B.V. All rights reserved.
Received 10 April 2014
Received in revised form
11 September 2014
Accepted 14 September 2014
Keywords:
Porous zirconium phosphate (ZrP)
Tungsten oxide
Glycerol
Acrolein
Dehydration
1. Introduction
formation of 3-hydroxypropionaldehyde and 1-hydroxyacetone [6]
(Scheme 1). The highest selectivity to acrolein reported so far in
Renewable biomass resources are promising alternatives for
the sustainable synthesis of chemical intermediates and liquid
fuels [1]. The by-product produced in bio-diesel synthesis is glyc-
erol and its effective utilization will be a key issue to promote
the bio-diesel commercialization. Therefore, new uses of glycerol
need to be explored for its valorization. The crude glycerol can be
converted to acrolein, which is an important chemical interme-
diate for the production of acrylic acid, acrylic acid esters, super
absorbers and polymers [2,3]. Acrolein is also used for the manu-
facture of methionine, 1,3-propanediol, glutaraldehyde, pyridines,
flavors, and fragrances [4]. Compared to the petroleum-based pro-
cesses [5], the dehydration of glycerol to acrolein has received a
great deal of attention in the recent past as a significant route by
virtue of its being an environmentally benign process.
both the vapour phase and liquid phase dehydration process is
65–90%. It was found that a solid acid with Hammet acidity (Ho)
between −10 and −16 is the most suitable catalyst for the dehy-
dration of glycerol than the catalysts having lower acidity with
H0 between −2 and −6 [7]. However, such strong acidic catalysts
are known to deactivate rapidly due to deposition of carbonaceous
species on the catalyst surface [8]. The acidic properties in combina-
tion with textural properties of the catalysts in the presence of large
number of micro pores also play an important role in determining
the catalytic performance [9–13]. The selectivity and deactivation
during glycerol dehydration will be strongly affected by diffusion
constraints due to coke formation.
The dehydration of glycerol to acrolein in the gaseous phase over
a solid acid catalyst leads to a sufficiently high dehydration activity.
However the catalyst deactivation is a major concern due to severe
reaction conditions, coke formation [8,14–16], sintering or leaching
of active phase in the reaction media [17] and formation of large
amount of by-products.
The dehydration of glycerol in gas phase on acidic catalysts is
a typical example of double dehydration reaction proceeds via the
Among various solid acid catalysts reported, WO3/ZrO2 cat-
alysts represent one of the highly active catalysts for vapour
∗
Corresponding author. Tel.: +91 40 27193162; fax: +91 40 27160921.
E-mail addresses: kvrchary@iict.res.in, Charykvr@gmail.com (K.V.R. Chary).
http://dx.doi.org/10.1016/j.molcata.2014.09.018
1381-1169/© 2014 Elsevier B.V. All rights reserved.

G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
487
Scheme 1. Schematic representation of the dehydration of glycerol to acrolein and acetol.
phase dehydration of glycerol, as they exhibit show unique cat-
alytic performance in the dehydration of glycerol [18–20]. Ulgen
and Hoelderich [18] have evaluated the catalytic performance of
WO3/ZrO2 catalysts with various WO3 loadings and noticed an
increase of acrolein selectivity in 73% with 19 wt% catalyst. Similar
acrolein selectivity was also reported at 300 ◦C with total conver-
sion using a commercial WO3/ZrO2 catalyst [19]. On the other hand
improvement of tungstated zirconia catalyst with SiO2 doping,
exhibited better acrolein selectivity (80%) with long catalyst life
and thermal stability to the catalyst [20]. There are many investiga-
tions reported on the structural and acidic properties of WOx/ZrO2
solid acids for glycerol dehydration reaction. However, not many
studies reported so far on the interaction of WOx with porous zir-
conium phosphate support for the gas phase glycerol dehydration
reaction.
In the present study we report the synthesis, characterization
and application of porous zirconium phosphate (ZrP) supported
WOx catalysts for the dehydration of glycerol to acrolein. Our
results provide mainly a basis for correlating the catalyst acidity
by varying the tungsten oxide content and the effect of reaction
temperature in glycerol dehydration. In addition, we also report
the positive effect due to addition of air to reactant flow in the gas
phase dehydration of glycerol.
Pore size distribution measurements were performed on
Autosorb-1 instrument (Qunta chrome, USA) using by nitrogen
physisorption.
The UV–vis diffused reflectance spectra were recorded on
a GBC UV–visible Cintra 10e spectrometer with an integrating
sphere reflectance accessory. The spectra were recorded in air at
room temperature and the data was transformed according the
Kubelka–Munk equation f(R) = (1 − R˛) 2/2r˛.
NH3-TPD experiments were conducted on the AutoChem 2910
(Micromeritics) instrument. Prior to TPD analysis the sample was
pretreated by passing high purity (99.999%) helium (50 ml/min)
at 300 ◦C for 1 h. After pretreatment, the sample was saturated
with 10% NH3 balance He mixture (75 ml/min) at 80 ◦C for 1 h and
subsequently flushed at 150 ◦C for 1 h to remove the physisorbed
ammonia. TPD analysis was carried out from ambient temperature
to 700 ◦C at a heating rate 10 ◦C/min. The amount of NH3 desorbed
is calculated using GRAMS/32 software.
The ex situ experiments of FT-IR spectra of pyridine adsorbed
samples were carried out to find the Brønsted and Lewis acid sites.
Pyridine was adsorbed on the activated catalysts at 200 ◦C until
saturation. Prior to adsorption experiments the catalysts were acti-
vated in N2 flow at 300 ◦C for 1 h to remove adsorbed water in
the samples. After such activation, the samples were cooled to
room temperature. The IR spectra were recorded using IR (Model:
GC-FT-IR Nicolet 670) spectrometer by KBr disk method at room
temperature. FT-IR spectra of the catalysts were recorded on IR
(Model: GC-FT-IR Nicolet 670) spectrometer by KBr disk method
at room temperature.
2. Experimental
Porous zirconium phosphate support was prepared from zirco-
nium n-propoxide precursor and 85% phosphoric acid following the
procedure described elsewhere [21]. About 0.01 mol of zirconium
n-propoxide, (70 wt% solution in 1-propanol, Aldrich) was added
drop wise to a 60 mL solution of H3PO4 (0.1 mol L−1) under stirring.
After 2 h of stirring at room temperature, the obtained mixture was
transferred into a teflon lined autoclave and aged statically at 80 ◦C
for 24 h. The final material was filtered, dried and calcined at 400 ◦C
for 5 h. A series of WOx/ZrP catalysts with WOx loadings ranging
from 5 to 40 wt% supported on ZrP were prepared by impregnation
method by adding aquous solution of ammonium metatungstate
to the calcined ZrP support. The catalysts were subsequently dried
at 100 ◦C for 12 h and calcined in a muffle furnace at 400 ◦C for
5 h.
Thermogravimetry analysis (Shimadzu TGA-51) was measured
at a heating rate of 10 ◦C/min from 25 ◦C up to 800 ◦C under the flow
of air.
The gas-phase dehydration of glycerol was conducted in the
reaction temperature ranging from 280 to 340 ◦C under atmo-
spheric pressure in a vertical fixed-bed quartz reactor (400 mm
length, 9 mm i.d.) using 0.2 g of catalyst. Before the reaction,
the catalyst was pretreated at 320 ◦C for 1 h in flow of dry N2
(30 mL min−1). An aqueous solution containing 20 wt% glycerol was
fed into the reactor by a micro-syringe pump at a flow rate of
0.5 mL/h (WHSV-2.6 h−1). The reaction products were condensed
in an ice–water trap and collected hourly for the analysis using
a gas chromatograph GC-2014 (Shimadzu) equipped with a DB-
wax 123-7033 (Agilent) capillary column (0.32 mm i.d., 30 m long)
and a flame ionization detector (FID). The oven temperature was
set from 56 ◦C to 119 ◦C (heating rate 5 ◦C/min, isothermal step at
119 ◦C, 3 min), then from 119 ◦C to 240 ◦C (heating rate 15 ◦C/min,
final isothermal step at 240 ◦C, 6 min).
X-ray powder diffraction patterns of the samples were obtained
by a model: D8 Diffract meter (Advance, Bruker, Germany), using
˚
Cu K␣ radiation (1.5406 A) at 40 kV and 30 mA. The measurements
were recorded in steps of 0.045◦ with a count time of 0.5 s in the
range of 2–40◦.

488
G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
200
180
(a)
(b)
160
(c)
(d)
140
(e)
120
100
80
60
40
Fig. 1. Powder XRD patterns of pure and tungstated ZrP catalysts. (a) P-ZrP, (b)
10 wt% WOx/ZrP, (c) 20 wt% WOx/ZrP, (d) 30 wt% WOx/ZrP and (e) 40 wt% WOx/ZrP.
20
0.0
0.2
0.4
0.6
0.8
1.0
P/Po
3. Results and discussion
Fig. 2. N2 adsorption–desorption isotherms of pure and tungstated ZrP catalysts. (a)
P-ZrP, (b) 10 wt% WOx/ZrP, (c) 20 wt% WOx/ZrP, (d) 30 wt% WOx/ZrP and (e) 40 wt%
WOx/ZrP.
3.1. Catalyst characterization results
X-ray diffraction patterns of various WOx/ZrP catalysts with
tungsten oxide loadings ranging from 10 to 40 wt% are shown in
Fig. 1. XRD results suggest that the synthesized ZrP is found to be X-
ray amorphous. However, the samples with tungsten oxide loading
less than 30 wt% WO3, did not show any XRD peaks suggesting that
the tungsten oxide is present in a well dispersed amorphous state.
However, the samples above 30 wt% WO3 loadings have shown XRD
reflections with poor crystallinity. Additional peaks are observed
for the samples with tungsten oxide loading of 40 wt% in the 2ꢀ
range of 23–25◦ [JCPDS-43-1035]. According to the previous studies
[22–24], these XRD reflections can be assigned to micro crystal-
lites of monoclinic WO3. It has been suggested previously that the
agglomeration of WOx species, leads to WO3 micro crystallites on
the support surface when the tungsten coverage exceeds that of a
monolayer [25].
Brunauer–Emmett–Teller (BET) surface area, average pore
diameter, and pore volume for pure zirconium phosphate support
(ZrP) and tungstated zirconium phosphate (WOx/ZrP) materials are
determined from their respective adsorption isotherms and the
results are given in Table 1. The BET surface area of ZrP is found to be
400 m2 g−1 with pore volume of 0.36 cc g−1 is comparable to that of
previously reported surfactant-templated mesoporous zirconium
dioxide [26]. Fig. 2 shows the N2 adsorption–desorption isotherm
of pure ZrP support calcined at 400 ◦C. The isotherm exhibits a
high uptake of N2 at low relative pressures (P/Po) of 0.1–0.4, sug-
gesting the presence of pore sizes ranging between micropore and
mesopore region.
WOx loading increases on the ZrP support, the mean pore diameter
of the sample is also increasing. However, the isotherms of WOx/ZrP
materials can be classified as type IV, which is a characteristic fea-
ture of the mesoporous materials [27]. The isotherm exhibits low
uptake of N2 at low relative pressures (P/Po) of 0.1–0.4, indicating
the decrease of the number of micropore and mesopores compare
to pure ZrP. This might be due to filling of the micro pores of ZrP sup-
port and these are not included in the measurement. These findings
are further confirmed by the pore size distribution profiles shown
in the Fig. 2, which reveals a narrow pore size distribution (Fig. 3)
centered around 2.2 nm.
UV–vis spectra of all the supported WOx samples are presented
in Fig. 4, along with the spectrum of pure ZrP support. The parent
ZrP showed a band at 300 nm due to Zr (IV) cations interacting
with the phosphate counter anions in the framework [28]. The
UV spectrum of ZrP support is strongly modified by the WOx
species. UV–visible diffuse reflectance spectroscopy was used to
probe tungsten oxide species dispersed on zirconia surface. The
size effect can be reflected from the shift of the absorption edge
[29]. As the WOx loadings increases on ZrP, the intensity of the
charge transfer transition band at 450 nm is also increased, due to
characteristic nature of crystalline WO3. As described earlier these
(a)
(b)
(c)
(d)
(e)
Following impregnation with WOx, a progressive decrease of
surface area and pore volume is observed with increasing WOx
loading (10–40 wt%). Although the surface area and pore volume
decreases with WOx loadings, a high surface area 117 m2 g−1 can
be still obtained even after a 30 wt% WOx loading. This decrease of
surface area with increasing WOx loading probably due to the added
WOx components are occupied in the pores of the ZrP support. As
Table 1
BET surface area and pore size distribution data of pure and tungstated ZrP catalysts.
S. no.
WOx loadings
(wt%)
BET SA (m2 g−1
)
Total pore
volume (cc/g)
Mean pore
diameter (Å)
01
02
03
04
05
06
P-ZrP
400
317
236
183
117
99
0.36
0.26
0.20
0.16
0.11
0.09
31.54
32.53
33.40
34.35
36.70
41.20
0
50
100
150
200
250
5W-ZrP
10W-ZrP
20W-ZrP
30W-ZrP
40W-ZrP
BJH Porediameter (Α0
)
Fig. 3. BJH pore diameter profile of pure and tungstated ZrP catalysts. (a) P-ZrP, (b)
10 wt% WOx/ZrP, (c) 20 wt% WOx/ZrP, (d) 30 wt% WOx/ZrP and (e) 40 wt% WOx/ZrP.

G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
489
(f)
P-ZrP
(e)
(d)
(c)
(b)
200 300 400
500
600
700 800
Wave length (nm)
(a)
(d)
(c)
(b)
(a)
100
200
300
400
500
600
700
Temperature (0C)
200
300
400
500
600
700
800
Fig. 6. NH3-TPD profiles of pure and tungstated ZrP catalysts. (a) P-ZrP, (b) 5 wt%
WOx/ZrP, (c) 10 wt% WOx/ZrP, (d) 20 wt% WOx/ZrP, (e) 30 wt% WOx/ZrP and (f)
40 wt% WOx/ZrP.
Wave length (nm)
Fig. 4. UV DRS profiles of of pure and tungstated ZrP catalysts. (a) 10 wt% WOx/ZrP,
(b) 20 wt% WOx/ZrP, (c) 30 wt% WOx/ZrP and (d) 40 wt% WOx/ZrP.
from a Zr–O–W stretching vibration as postulated by Chauveau
et al. [33].
The surface acidity is an important characteristic property of a
solid acid catalyst to assess the dehydration functionality of glycerol
to acrolein [13]. The ammonia TPD profiles of various loadings of
WOx catalysts supported on ZrP are presented in Fig. 6. The amounts
of ammonia desorbed during TPD analysis by various catalysts are
shown in Table 2. The pure ZrP support exhibited one sharp peak in
the weak acidic region and one broad peak in the moderate acidic
region. It can be seen from the Fig. 6 that TPD profiles showed only
one broad peak in WOx/ZrP catalysts. It was difficult for the decon-
volution of TPD profiles into weak, moderate and strong acidic sites
as the TPD profiles are strongly overlapped. Therefore, the surface
acidity was calculated as total acidity and expressed per mol of NH3
desorbed per gram of catalyst. Additionally, the acidity is given as
mol of NH3 desorbed per m2 of surface area (Table 2).
The TPD profile shown in Fig. 6 reveals that the desorption peaks
are due to weak acidic sites which are present only in ZrP support
and no peaks were noticed in WOx/ZrP catalysts. The total acid-
ity values of supported WOx on ZrP support are increasing from
5 wt% to 30 wt% WOx/ZrP and decreases at higher loadings (40 wt%
WOx/ZrP). This could be possibly due to condensation of the W–OH
groups and formation of 3-D WOx clusters at higher WOx loadings
reduces the acidity [34]. However, ammonia is mainly desorbed
between 200 and 550 ◦C, in all the WOx/ZrP catalysts, suggesting
the presence of medium and strong acidic sites. The acidity of the
WOx/ZrP catalysts can be attributed to the P–OH and well dispersed
WOx groups present on the surface [35]. TPD analysis suggests that
the amount of acidity and strength of acidic sites are increasing
upon addition of WOx ions to the ZrP support.
results are qualitatively in accord with the results obtained from
X-ray diffraction.
FT-IR spectra of pure ZrP support and various WO3/ZrP cata-
lysts are shown in Fig. 5. The spectra of all the samples have shown
an absorption band at 1000–1100 cm−1 corresponding to strong
Zr–O–P vibration [30]. The IR spectra also show the bands in the
region 3400–2400 cm−1 ascribed due to surface OH groups and
the IR band at 1630 cm−1 is assigned to bending adsorbed mode
of water molecule by the catalyst [31]. The absence of band at
750 cm−1 suggests the nonexistence of P–O–P (poly phosphate)
like groups in pure ZrP [32]. An additional IR band was observed
at 820 cm−1 in all the impregnated samples, attributable to the
W–Oc–W vibration of two connecting WO6 octahedra. It is also
noteworthy that an IR band was observed at 640 cm−1 in all the
impregnated samples which can be assigned due to a Zr–O vibra-
tion. Since this IR band was absent in ZrP support, as it might arise
The previous discussion does not contain any information on the
nature of acidic sites, because the ammonia-TPD cannot discrimi-
nate Brønsted and Lewis acid sites. For characterizing the nature of
Table 2
Temperature programmed desorption of NH3 of pure and tungstated ZrP catalysts.
S. no.
WOx loadings Tmax (◦C)
(wt%)
NH3 desorbed NH3 desorbed
(mmol/g)
(␮mol/m2)
01
02
03
04
05
06
P-ZrP
120
320
350
350
345
345
3.30
3.72
3.84
4.02
4.91
4.10
8.25
11.73
16.25
21.95
41.90
41.00
5W-ZrP
10W-ZrP
20W-ZrP
30W-ZrP
40W-ZrP
Fig. 5. FT-IR profiles of pure and tungstated ZrP catalysts. (a) P-ZrP, (b) 10 wt%
WOx/ZrP, (c) 20 wt% WOx/ZrP, (d) 30 wt% WOx/ZrP and (e) 40 wt% WOx/ZrP.

490
G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
120
100
80
Con of glycerol
sel to acrolein
sel to acetaldehyde
sel to hydroxy acetone
sel to acetic acid
60
15
10
5
0
280
300
320
340
Reaction temperature (oC)
Fig. 8. Effect of reaction temperature on glycerol conversion and selectivity to reac-
tion products on 30 wt% WOx/ZrP catalyst (reaction conditions: 0.2 g of catalyst was
used and diluted with 3.0 g of quartz. The aqueous glycerol (20% wt/wt) was fed at
the speed of 0.5 g h−1 by B. Braun syringe pump under 7 mL min−1 N2 + 3 mL min−1
air atmosphere).
Fig. 7. FT-IR Pyridine adsorption profiles of pure and tungstated ZrP catalysts. (a)
P-ZrP, (b) 10 wt% WOx/ZrP, (c) 20 wt% WOx/ZrP, (d) 30 wt% WOx/ZrP and (e) 40 wt%
WOx/ZrP.
kinetically favored subsequent consecutive steps. Higher amount
of hydroxyacetone observed at lower reaction temperature is pos-
sibly due to the difference in the activation energies of reaction
pathways as depicted in Scheme 1. While the high temperature
favors the decomposition of hydroxyacetone, as a result the selec-
tivity of acetaldehyde and acetic acid also increases with reaction
temperature.
surface acidic sites, we have employed ex situ the pyridine adsorbed
FT-IR analysis. Prior to ex situ FT-IR analysis the catalyst samples
were saturated with pyridine vapour at 200 ◦C for 2 h. The FT-IR
analysis reveals that the IR bands appeared at 1540–1548 cm−1
and 1445–1460 cm−1 are characteristic of Brønsted (B) and Lewis
(L) acid sites. Furthermore, the IR bands correspond to the com-
bination of both Brønsted and Lewis (B+L) acid sites are appeared
at 1490–1500 cm−1. It should be noted that the intensity of the IR
bands is proportional to the concentration of acidic sites.
The FTIR spectra of pure ZrP support and various WO3/ZrP cata-
lysts are illustrated in Fig. 7. All the catalysts have shown IR bands
at 1444 cm−1 corresponding to Lewis sites and another IR band
appeared at 1550 cm−1 is attributed to Brønsted sites. It can also
be seen from Fig. 7, that all the catalysts contain both Lewis and
Brønsted acidic sites in different proportions depending upon the
WOx loadings on the support.
It is interesting to note that with increasing WOx loadings on
the support, the intensity of IR band at 1450 cm−1 corresponding
to the Lewis acidity decreases marginally. The IR absorption peak
at 1498 cm−1 is attributed to combination of both Brønsted (B) and
Lewis (L) acid sites. This IR peak intensity is increased with increas-
ing WOx loadings on the support up to 30 wt% WOx/ZrP. The pure
ZrP shows the band at 1550 cm−1 due to Brønsted acidic sites. The
Brønsted acidity of WOx/ZrP catalysts increases initially and did not
change appreciably until 30 wt% WOx/ZrP. Furthermore, in 40 wt%
WOx/ZrP a slight decrease of intensity of IR bands is noticed in both
the types of acidic sites.
At 300 ◦C 30 wt% WOx/ZrP catalyst has shown maximum con-
version and selectivity. There is no significant effect on the glycerol
conversion is noticed with the increase of temperature (Fig. 8).
However, at higher temperature (340 ◦C) the acrolein selectivity
decreased due to increase in the formation of acetaldehyde, acetic
acid and also a few unidentified products. These results clearly
demonstrate that at higher temperature (340 ◦C) the catalyst favors
the C C bond cleavage and also oxidation reaction instead of
dehydration of glycerol. The major side product obtained is the
acetaldehyde, which is formed due to C C bond cleavage and oxi-
dation of acetaldehyde leads to the formation of acetic acid. In the
view of these findings, 300 ◦C was selected as optimum reaction
temperature for screening of the catalysts for further investigation.
The results of the effect of WOx loading on ZrP support, as well
as nature of feed flow (only N2 and N2 along with air) during the
vapour phase dehydration of glycerol are shown in Table 3. The
results clearly suggest that the tungstated zirconium phosphate
(WOx/ZrP) catalysts are found to be more active than pure ZrP
support under both the gas flow conditions (only N2 and N2 along
with air). Glycerol conversion and acrolein selectivity are found to
increase with addition of WOx up to 30 wt% WOx/ZrP and decreases
at higher loadings (40 wt% WOx/ZrP). The catalysts with 20 wt%,
30 wt% WOx/ZrP have shown higher glycerol conversion than the
other catalysts. However, the sample with 30 wt% WOx/ZrP exhib-
ited highest selectivity to acrolein (82% using Air + N2 and 70% using
N2).
3.2. Catalytic results
The dehydration of glycerol exhibited by various WOx/ZrP cat-
alysts are illustrated in Figs. 8 and 9 and the product distribution
results are reported in Table 3. The influence of the reaction temper-
ature (280–340 ◦C) on the dehydration of glycerol was examined
over 30 wt% WOx/ZrP by passing air along with N2, and the results
are shown in Fig. 8. At 280 ◦C the conversion of glycerol is found
to be 89% and reaches a maximum at higher temperature. A high
selectivity to acrolein formed at moderate temperature can be
explained due to the selective activation of the stronger adsorp-
tion mode of glycerol through the secondary OH group and the
Pure ZrP support exhibited 90% glycerol conversion with 52%
selectivity toward acrolein under N2 along with air feed flow. The
other products obtained by using pure ZrP support are hydroxyace-
tone, acetic acid and acetaldehyde as the major by-products along
with 32% containing allylic alcohol, acetone, methanol, hydrogenol-
ysis products and also trace amount of unidentified products. The
decrease of dehydration activity can be explained by the acidity of
P-ZrP support determined with NH3-TPD method. The TPD profile

G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
491
Table 3
Product distribution results of glycerol dehydration over pure and tungstated ZrP catalysts.
S. no.
WOx loadings (wt%)
Conversion (%)
Selectivity (%)
Acrolein
Hydroxy-acetone
Aceticacid
Acetaldehyde
Others
01
02
03
04
05
06
P-ZrP
90 [80]
97 [92]
100 [95]
100 [96]
100 [96]
95 [90]
52 [40]
57 [49]
62 [55]
69 [61]
82 [70]
74 [68]
12 [16]
10 [19]
7 [17]
4 [15]
4 [10]
6 [12]
2 [1]
7 [3]
12 [3]
10 [3]
5 [2]
2 [6]
10 [8]
10 [10]
7 [8]
4 [9]
5 [7]
32 [37]
16 [21]
9 [15]
10 [13]
5 [9]
5W-ZrP
10W-ZrP
20W-ZrP
30W-ZrP
40W-ZrP
7 [2]
8 [11]
Reaction conditions: 0.2 g catalyst was used and diluted with 3.0 g of quartz. The aqueous glycerol (20% wt/wt) was fed at the speed of 0.5 g h−1 by syringe pump under
7 mL min−1 N2 + 3 mL min−1 Air atmosphere and reaction temperature was 300 ◦C. The values reported in paranthesis are under similar reaction condition except gas flow of
10 mL min−1 N2.
of P-ZrP exhibits lower amount of acidity and also the desorption
peak appeared at lower temperature region compared to other
samples. While zirconia based supports were reported as a neu-
tral materials [36]. It was found that basic sites can exist on its
surface along with acidic sites [37]. However, ZrP catalysts have
shown excellent acidic properties in many acid catalyzed reactions
[32] and it might also contain some basic sites on its surface [38].
The presence of these weak basic sites enables higher selectivity of
undesirable products and hydroxyacetone [20,39]. At lower load-
ings (5 wt% WOx) these by-products and other products are formed
in higher amount because the ZrP support is not fully covered by
the WOx.
The selectivity to hydroxyacetone is found to be very low in
WOx/ZrP catalysts compared to the pure ZrP support suggesting
that the addition of WOx enhances acidity of the catalysts resulting
an increase of selectivity of acrolein. The better catalytic perfor-
mance of these catalysts could imply that the Brønsted acid sites
and also catalysts containing both Lewis and the Brønsted acid
sites are advantageous for stable acrolein production during glyc-
erol dehydration. Characterization of the WOx/ZrP catalysts has
suggested that the active species correspond to P-OH and W-OH
species, which generates Brønsted acid sites active and selective
for the dehydration reaction [15,40]. Whereas decrease of cata-
lyst activity at high WO3 (40 wt %) loading has been attributed to
condensation of W–OH groups and formation of 3-D WOx clusters
which lowers the acidity [41]. The XRD results also further support
these findings as crystalline WOx appeared for 40 wt% WOx/ZrP.
WOx/ZrO2 [34,41] and WOx/ZrP [42] have been used as solid acid
catalysts for various reactions, such as esterification, hydrolysis,
dehydration, and hydration. The WOx impregnated catalysts con-
tain stronger acidity and also large number of acidic sites [15,40].
However, such strong acidic catalysts tend to deactivate rapidly
as carbonaceous deposits will be formed which blocks the cata-
lyst surface [8,14–16]. Therefore, several attempts were made to
control the coke deposition by the co-feeding of molecular oxygen
with the feed [43,44]. The positive effect of oxygen gas has been
reported previously [45] with the addition of oxygen to the reactant
flow. Ulgen and Hoelderich [18] studied the glycerol dehydration
reaction over WOx/ZrO2 catalyst by passing molecular oxygen with
nitrogen gas, and reported that there is an increase in acrolein
selectivity and decrease in the main side product selectivity. Wang
et al. [43] studied the glycerol dehydration reaction over pure VPO
catalyst by passing molecular oxygen with nitrogen gas to main-
tain the glycerol conversion, acrolein yield and greatly reduces the
side product formation. In the present investigation the dehydra-
tion of glycerol was carried out over different loadings of WOx on
ZrP support by passing air along with N2 flow and also only with
N2 feed flow. The effective addition of air in the N2 flow had a
remarkable effect on the catalytic results of glycerol dehydration
(Table 3). The results clearly suggest that conversion of glycerol and
selectivity to acrolein improved significantly irrespective of active
component loading on the support. The presence of air can greatly
decreases the selectivity of main side product i.e., hydroxyacetone
and other products (mainly allyl alcohol). The selectivity of acetic
acid is increasing as it is the oxidation product obtained from sec-
ondary reactions. However, there is no significant decrease in the
acrolein selectivity. The allyl alcohol is formed by the hydrogena-
tion of acrolein in the reaction medium. The hydrogen is mainly
produced from the degradation of glycerol into the coke and addi-
tion of air prevents the degradation of glycerol. The hydroxyacetone
is produced through the dehydration at the terminal hydroxyl
group of glycerol. The oxygen presents in the air restricts the
hydrogenation of acrolein and favors the dehydration at the central
hydroxyl group of the glycerol.
(A)
100
90
80
Conversion of glycerol
70
Sel to acrolein
sel to acetaldehyde
sel to hydroxyacetone
sel to acetic acid
8
sel to others
4
0
0
10
20
30
40
50
TOS (h)
60
70
80
90
100
100
90
80
70
60
(B)
Conversion of glycerol
Sel to acrolein
50
20
Sel to acetaldehyde
Sel to hydroxyacetone
Others selectivity
15
10
5
5
10
15
20
25 30
TOS (h)
35
40
45
50
Fig. 9. Evolution of the glycerol conversion and selectivity of reaction products
as a function of time over 30 wt% WOx/ZrP catalyst. (A). By passing 7 mL min−1
N
2 + 3 mL min−1 air flow. (B) By passing 10 mL min−1 N2. (Reaction conditions:
0.2 g of catalyst was used and diluted with 3.0 g of quartz. The aqueous glycerol
(20% wt/wt) was fed at the speed of 0.5 g h−1 by B. Braun syringe pump at 300 ◦
temperature.).
C

492
G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
Fig. 11. TG-DTG curve of spent 30 wt% WOx/ZrP catalysts after 50 h TOS.
Table 4
B.E.T surface area, pore size distribution and CHNS analysis data of NH3 of fresh and
spent 30 wt% WOx/ZrP catalyst.
Fresh catalyst Spent catalyst Spent catalyst
under Air + N2 under N2 flow
Fig. 10. FT-IR pyridine adsorption profiles of fresh and spent 30 wt% WOx/ZrP cata-
lysts after 50 h TOS.
BET surface area (m2 g−1
BJH pore volume (cc/g)
)
117.0
0.11
55.7
0.07
32.4
0.04
BJH average pore diameter (Å) 36.7
H/C (CHNS analysis)
49.10
1.48
61.60
0.96
–
The advantage of adding air in the dehydration reaction is not
only enriches the selectivity of desired product but also increase
life of the catalyst by reducing carbon deposition on the surface
of catalyst. Strongly adsorbed hydrocarbon species on the catalyst
surface can be quickly removed by oxidation before the formation
of intensive carbon coke, and catalytic active sites are recovered.
The effect of feed flow (only N2 and air along with N2) over
30 wt% WOx/ZrP catalyst at 300 ◦C as a function of reaction time is
shown in the Fig. 9 A and B. The results indicate that the conver-
sion of glycerol is found to be greater than 90% even after 100 h of
reaction time. Thus by passing N2 along with air flow enhances the
stability of the catalysts during the vapour phase dehydration reac-
tion. However, under only N2 gas flow (Fig. 9B) a rapid deactivation
can be noticed with time on stream (TOS: 50 h-70% conversion).
The deactivation phenomenon can be explained by the formation
of coke.
During the reaction under only N2 flow the selectivity to hydrox-
yacetone increases from 7% to 18% with time on stream. While
under the reaction conditions using N2 along with air flow the selec-
tivity of hydroxyacetone is found to be >6% even after 100 h TOS.
However, the acetic acid formation is found to increase in the pres-
ence of air in the feed. The change in hydroxyacetone selectivity
might depend on the nature of the active sites, which can be mod-
ified in situ by the presence of oxygen in the air flow [18]. It is also
possible that hydroxyacetone can get oxidize to other products.
The selectivities to acetaldehyde initially rather high and decreased
with time on stream confirming, that acetaldehyde was formed
from consecutive reactions of acrolein or hydroxyacetone in both
cases.
occurred at 350 ◦C after 50 h TOS by passing air along with N2. How-
ever the spent catalysts under N2 flow have shown significantly
more weight loss (22%) at 450 ◦C temperature. TGA results further
suggest that formation of coke deposited is higher on the surface
of 30 wt% WOx/ZrP catalyst under non aerobic conditions.
Analysis of atomic H/C ratio in the coke over the catalysts reveals
that the results are in agreement with findings of the TGA-DTG. The
atomic H/C ratio in the coke measured by CHNS elemental analy-
sis (Table 4) was found to be higher for the spent catalyst (30 wt%
WOx/ZrP) under air along with N2 flow than the spent catalyst
under only N2 flow. The higher H/C ratio of spent catalyst under
air along with N2 flow implies that the carbon deposits have a less
condensed and aromatized structure over the active sites.
The results reported in Table 4 suggest that an increase in the
average pore diameter is observed in spent 30 wt% WOx/ZrP cat-
˚
alysts than the fresh catalyst (36.7 A). This could be possibly due
to the pore condensation during the glycerol dehydration reaction
leading to deactivation of the catalyst [46]. Interesting observation
can be seen from the Table 4 that the average pore diameter sig-
˚
nificantly increased in the spent catalyst under N2 (61.6 A) than
the spent catalyst under N2 along with air flow (49.1). Further, the
addition of air to N2 flow reduces the loss of surface area due to
sintering.
4. Conclusions
This study has demonstrated that a series of porous zirco-
nium phosphate supported WOx catalysts prepared by the wet
impregnation method are effective for the gas phase dehydration of
glycerol to produce acrolein. The characterization results have con-
firmed that the WOx species is well dispersed on ZrP support. The
supported WOx catalyst showed better conversion and selectivity
toward acrolein than pure ZrP support. The catalyst performance
for the dehydration reaction is significantly affected by the loading
of WOx over the ZrP surface, which induces the changes in the mod-
erate surface acidity and number of Brønsted acidic sites. Improved
acidity has been noticed upon addition of WO3 to ZrP by using
NH3-TPD and Py-FT-IR techniques. The 30 wt% WOx/ZrP catalyst
has shown the best catalytic performance during glycerol dehydra-
tion with 100% conversion and 82% acrolein selectivity compared to
Further, the characterization of spent catalysts has provided bet-
ter explanation on the effect of feed flow during the gas phase
dehydration of glycerol. Ex situ pyridine adsorbed FT-IR analysis
(Fig. 10) was also performed to examine the acidic properties of the
spent (30 wt% WOx/ZrP) catalysts. The peak areas corresponding to
Brønsted (B) and B + L acidic sites of spent catalysts were decreased
compared with those of fresh 30 wt% WOx/ZrP catalyst. However,
the spent catalyst under air along with N2 flow has shown more
intense peaks than the spent catalyst under only N2 flow. This indi-
cates that the addition of air to feed flow regenerate the active sites
which are involved in the deactivation during reaction.
The TG–DTG curves of the spent 30 wt% WOx/ZrP catalysts are
shown in Fig. 11. The results suggest that about 10% of weight loss

G. Srinivasa Rao et al. / Journal of Molecular Catalysis A: Chemical 395 (2014) 486–493
493
other catalyst. The addition of air during the reaction considerably
reduces the formation of by-products and enhances the stability of
the catalysts.
[18] A. Ulgen, W. Hoelderich, Catal. Lett. 131 (2009) 122–128.
[19] J.L. Dubois, C. Duquenne, W. Hoelderich, J. Kervennal, WO2006087084, 2006.
[20] P.L. Garbey, S. Loridant, V.B. Baca, P. Rey, J.M.M. Millet, Catal. Commun. 16
(2011) 170–174.
[21] T.Z. Ren, Z.Y. Yuan, B.L. Su, Chem. Commun. (23) (2004) 2730–2731.
[22] A. Martinez, G. Prieto, M.A. Arribas, P. Concepcion, Appl. Catal. A: Gen. 309
(2006) 224–236.
Acknowledgment
[23] A. Martinez, G. Prieto, M.A. Arribas, P. Concepcion, J.F. Sanchez- Royo, J. Catal.
248 (2007) 288–302.
[24] S. Sarish, B.M. Devassy, S.B. Halligudi, J. Mol. Catal. A: Chem. 235 (2005) 44–51.
[25] N. Liu, S. Ding, Y. Cui, N. Xue, L. Peng, X. Guo, W. Ding, Chem. Eng. Res. Des. 91
(2013) 573–580.
Author thanks to Council of Scientific and Industrial Research
(CSIR-India), New Delhi for the award of Senior Research Fellow-
ship.
[26] Z.Y. Yuan, A. Vantomme, A. Leonard, B.L. Su, Chem. Commun. (13) (2003)
1558–1559.
References
[27] J. JimØnez-JimØnez, P. Maireles-Torres, P. Olivera-Pastor, E. Rodríguez-
Castellón, A. JimØnez-López, D.J. Jones, J. Rozire, Adv. Mater. 10 (1998) 812–815.
[28] M. Ziyad, M. Rouimi, J.L. Portefaix, Appl. Catal. A: Gen. 183 (1999) 93–105.
[29] G.A. Ozin, S. Ozkar, J. Phys. Chem. 94 (1990) 7556–7560.
[30] H.N. Kim, S.W. Keller, T.E. Mallouk, J. Schmitt, G. Decher, Chem. Mater. 9 (1997)
1414–1421.
[31] A. Clearfield, D.S. Thakur, Appl. Catal. 26 (1986) 1–26.
[32] A. Sinhamahapatra, N. Sutradhar, B. Roy, A. Tarafdar, H.C. Bajaj, A. Baran Panda,
Appl. Catal. A: Gen. 385 (2010) 22–30.
[33] F. Chauveau, J. Eberle, J. Lefebvre, Nouv. J. Chim. 9 (1985) 315–319.
[34] A. Kumara, A. Alia, K.N. Vinoda, A. Kumar Mondal, H. Hegde, A. Menon, B.H.S.
Thimmappa, J. Mol. Catal. A: Chem. 378 (2013) 22–29.
[35] T.Y. Kim, D.S. Park, Y. Choi, J. Baek, J.R. Park, J. Yi, J. Mater. Chem. 22 (2012)
10021–10028.
[36] J. Goscianska, M. Ziolek, E. Gibson, M. Daturi, Catal. Today 152 (2010) 33–41.
[37] K. Pokrovski, K.T. Jung, A.T. Bell, Langmuir 17 (2001) 4297–4300.
[38] D. Spielbauer, G.A.H. Mekhemer, T. Riemer, M.I. Zaki, H. Kno1zinger, J. Phys.
Chem. B 101 (1997) 4681–4690.
[39] A.K. Kinage, P.P. Upare, P. Kasinathan, Y.K. Hwang, J.S. Chang, Catal. Commun.
11 (2010) 620–623.
[40] Y.T. Kim, K.D. Jung, E.D. Park, Appl. Catal. B: Environ. 107 (2011) 177–187.
[41] S.T. Wong, T. Li, S. Cheng, J.F. Lee, C.Y. Mou, J. Catal. 215 (2003) 45–56.
[42] K. Narasimha Rao, A. Sridhar, A.F. Lee, S.J. Tavener, N.A. Young, K. Wilson, Green
Chem. 8 (2006) 790–797.
[43] F. Wang, J.L. Dubois, W. Ueda, Appl. Catal. A: Gen. 376 (2010) 25–32.
[44] A. Corma, G.W. Huber, L. Sauvanaud, P. O’Connor, J. Catal. 257 (2008) 163–171.
[45] J.L. Dubois, C. Duquenne, W.H. Hoelderich, 2005, F Patent 2,882,052 to Arkema,
SA.
[1] S.E. Koonin, Science 311 (2006) 435.
[2] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C.D. Pina, Angew. Chem. Int. Ed.
46 (2007) 4434–4440.
[3] B. Katryniok, S. Paul, M. Capron, F. Dumeignil, ChemSusChem 2 (2009) 719–730.
[4] D. Arntz, A. Fischer, M. Höpp, S. Jcobi, J. Sauer, T. Ohara, T. Sato, N. Shimizu, H.
Schwind, Ullmann’s Encyclopedia of Industrial Chemistry, 17, 7th ed., Wiley-
VCH, Weinheim, 2012, pp. p329.
[5] C. Zhao, I.E. Wachs, Catal. Today 118 (2006) 332–343.
[6] M.J. Antal, W.S.L. Mok, J.C. Roy, A.T. Raissi, J. Anal. Appl. Pyrol. 8 (1985) 291–303.
[7] S. Chai, H. Wang, Y. Liang, B. Xu, Green Chem. 9 (2007) 1130–1140.
[8] J. Lourenco, M.I. Macedo, A. Fernandes, Catal. Commun. 19 (2012) 105–109.
[9] B. Katryniok, S. Paul, V. Bellière-Baca, P. Rey, F. Dumeignil, Green Chem. 12
(2010) 2079–2098.
[10] C.J. Jia, Y. Liu, W. Schmidt, A.H. Lu, F. Schüth, J. Catal. 269 (2010) 71–79.
[11] Y.T. Kim, K.D. Jung, E.D. Park, Micropor. Mesopor. Mater. 131 (2010) 28–36.
[12] E. Tsukuda, S. Sato, R. Takahashi, T. Sodesawa, Catal. Commun. 8 (2007)
1349–1350.
[13] B. Katryniok, S. Paul, M. Capron, C. Lancelot, V. Bellière-Baca, P. Rey, F.
Dumeignil, Green Chem. 12 (2010) 1922–1930.
[14] N.R. Shiju, D.R. Brown, K. Wilson, G. Rothenberg, Top. Catal. 53 (2010)
1217–1220.
[15] A. Alhanash, E.F. Kozhevnikova, I.V. Kozhevnikov, Appl. Catal. A: Gen. 378
(2010) 11–18.
[16] A. Witsuthammakul, T. Sooknoi, Appl. Catal. A: Gen. 413–414 (2012)
109–116.
[17] C.H. Zhou, J.N. Beltramini, Y.X. Fan, C.Q. Lu, Chem. Soc. Rev. 37 (2008) 527–
549.
[46] C. Hulteberg, A. Leveau, J.G. Meo Brandin, Top. Catal. 56 (2013) 813–821.