Characterization and reactivity of 11-molybdo-1-vanadophosphoric acid catalyst supported on zirconia for dehydration of glycerol to acrolein

Article abstract of DOI:10.1007/s12039-014-0586-z

A series of vanadium-substituted phosphomolybdic acid (HPA) catalysts supported on zirconia were prepared by impregnation method with varying the HPA active phase content from 10 to 50 wt% on the support. The calcined catalysts were characterized by X-ray diffraction, Raman spectroscopy, temperature-programmed desorption of NH₃, FT-IR spectra of pyridine adsorption and surface area measurements. XRD results suggest that the active phase of heteropolyacid is present in a highly dispersed state at lower loadings and as a crystalline phase at higher HPA loadings and these findings are well-supported by the results of FT-IR and Raman spectra. Calcination of the samples did not affect the Keggin ion structure of HPA. The ammonia TPD results suggest that acidity of the catalysts was found to increase with increase of HPA loading up to 40 wt% and decreases at higher loadings. FT-IR spectra of pyridine adsorption show that the Bronsted acidic sites increase with increase of HPA loadings up to 40 wt% catalyst. However, Lewis acid sites decrease with increase of HPA loading. Catalytic properties were evaluated during vapour phase dehydration of glycerol to acrolein. The catalyst with 40 wt% HPA has exhibited excellent selectivity towards acrolein formation with complete conversion of glycerol at 225°C under atmospheric pressure. Catalytic performances during dehydration of glycerol are well-correlated with acidity of the catalysts.

Full text of DOI:10.1007/s12039-014-0586-z

c
J. Chem. Sci. Vol. 126, No. 2, March 2014, pp. 445–454. ꢀ Indian Academy of Sciences.
Characterization and reactivity of 11-molybdo-1-vanadophosphoric acid
catalyst supported on zirconia for dehydration of glycerol to acrolein
BALAGA VISWANADHAM, AMIRINENI SRIKANTH and KOMANDUR V R CHARY∗
Catalysis Division, Indian Institute of Chemical Technology, Hyderabad 500 007, India
e-mail: kvrchary@iict.res.in
MS received 18 September 2013; revised 27 December 2013; accepted 30 December 2013
Abstract. A series of vanadium-substituted phosphomolybdic acid (HPA) catalysts supported on zirconia
were prepared by impregnation method with varying the HPA active phase content from 10 to 50 wt% on
the support. The calcined catalysts were characterized by X-ray diffraction, Raman spectroscopy, temperature-
programmed desorption of NH3, FT-IR spectra of pyridine adsorption and surface area measurements. XRD
results suggest that the active phase of heteropolyacid is present in a highly dispersed state at lower loadings and
as a crystalline phase at higher HPA loadings and these findings are well-supported by the results of FT-IR and
Raman spectra. Calcination of the samples did not affect the Keggin ion structure of HPA. The ammonia TPD
results suggest that acidity of the catalysts was found to increase with increase of HPA loading up to 40 wt%and
decreases at higher loadings. FT-IR spectra of pyridine adsorption show that the Brønsted acidic sites increase
with increase of HPA loadings up to 40 wt% catalyst. However, Lewis acid sites decrease with increase ofHPA
loading. Catalytic properties were evaluated during vapour phase dehydration of glycerol to acrolein. The cata-
lyst with 40 wt% HPA has exhibited excellent selectivity towards acrolein formation with complete conversion
of glycerol at 225◦C under atmospheric pressure. Catalytic performances during dehydration of glycerol are
well-correlated with acidity of the catalysts.
Keywords. H4PMo11VO40/ZrO2; acidity; Raman; glycerol dehydration; acrolein.
1. Introduction
to acrolein is accompanied by side reactions leading
to the formation of some by-products such as hydroxy
propanone, propanaldehyde, acetaldehyde, acetone and
adducts of acrolein to form polyaromatic compounds.
This leads to the formation of coke over the active
sites and causes deactivation of the catalyst. However
the presence of these by-products with acrolein neces-
sitates high recovery cost for separation and purifica-
tion of acrolein. Various solid acid catalysts includ-
ing sulphates, phosphates, zeolites, heteropolyacids and
solid phosphoric acid (SPA) have been tested for the
dehydration of glycerol in either gaseous or liquid
phase.6–21 Glycerol is usually produced as a mixture
with water. Usage of glycerol with water is advanta-
geous over pure glycerol for the production of acrolein
and also a solid acid catalyst with a high redox pro-
perty of catalyst would be beneficial for obtaining the
better catalytic activities which minimizes the coke
Exploitation of glycerol for the production of value-
added fine chemical intermediates is a topic of enor-
mous interest in the recent past because glycerol is a
by-product formed in huge amounts during the pro-
duction of biodiesel from renewable sources. Various
processes such as hydrogenolysis, dehydration, acetyla-
tion, etc., have been investigated for converting glyce-
rol into value-added products. Among these processes,
dehydration of glycerol is one of the interesting and
challenging route and to produce acrolein in the pres-
ence of solid acid catalysts. Acrolein is an impor-
tant and versatile chemical intermediate used in many
industrial applications and it is the starting material for
the production of acrylic acid, acrylic acid ester, glu-
taraldehyde and methionine.1–5 This process is widely
investigated by many researchers to convert glycerol
into value-added chemicals via environmentally benign
catalytic pathways. In general, dehydration of glycerol
formation.
22,23
Usage of ZrO2
as a catalyst support has seve-
ral advantages over other conventional oxide sup-
ports such as alumina, silica and titania. Advantages
of using zirconia as a support include: (i) interacts
strongly with the active phase, (ii) possesses high
thermal stability and more chemically inert than the
∗For correspondence
445

446
Balaga Viswanadham et al.
conventional supported oxides and (iii) it has acidic 2. Experimental
and redox properties. Zirconia-supported heteropoly-
2.1 Catalyst preparation
acids have high surface area and high thermal stabi-
lity. By varying the amount of H4PMo11VO40 (HPA)
on the support leads to an enhancement of the
catalytic performance.21,24–26 Design of solid acid
catalysts containing heteropolyacids is one of the key
technologies to establish environment-friendly catalytic
processes. The bulk phosphomolybdic acid with pri-
mary Keggin unit is well-recognized as the oxida-
tion and also solid acid catalyst in which the redox
and acidic functionalities can be tuned by substitu-
tion of vanadium metal in the Keggin units. Syn-
thesis of vanadium-containing phosphomolybdic acid
is found to be an interesting and promising he-
teropolyacid catalyst wherein, the molybdenum metal
is partially replaced by vanadium metal. Vanadium-
containing phosphomolybdic acid possesses positive
reduction potentials compared to pure phosphomolyb-
dic acid and enhances the redox properties. Thus, incor-
poration of V into the structure of phosphomolybdic
acid (PMA) enhances the catalytic performance (acid
and redox catalysts).27,28 Vanadium-containing HPAs
have shown excellent redox properties because substitu-
tion of vanadium stabilizes the LUMOs.29,30 Vanadium-
containing HPA catalysts are the basic components of
several oxidative and acidic reactions in homogenous
and heterogeneous catalysis.26,31–35
In the present study, we report dehydration of gly-
cerol to acrolein over a series of vanadium-containing
HPA catalyst as an active phase supported on zirco-
nia. The aim of this investigation is to study the effect
of HPA loading on zirconia during the vapour phase
dehydration glycerol under mild reaction conditions.
We also report the comparison of different supports
with HPA as an active phase on the catalytic pro-
perties during dehydration of glycerol. The purpose of
this study is to estimate acidity of vanadium-substituted
phosphomolybdic acid supported on zirconia as a func-
tion of HPA loading and to identify the structural
changes of HPA with increase of active phase load-
ing and also to understand the relation between selec-
tivity and acidic sites. The calcined catalysts are char-
acterized by XRD, BET surface area, FT-IR, Raman,
TPD of NH3 and FT-IR of pyridine adsorbed samples
to obtain the structural and acidic properties of the
active species and relate it to the catalytic functionali-
ties during vapour phase dehydration of glycerol. Pro-
duct distributions during dehydration of glycerol were
investigated as a function of different reaction vari-
ables such as the effect of active phase loading, reac-
tion temperature, time on stream and the nature of the
support.
2.1a Synthesis of ZrO2 support: Various metal
oxides employed during the catalysts preparation are
vanadium(V) oxide, zirconium(IV) isopropoxide and
titanium(IV) isopropoxide supplied by Aldrich. Molyb-
denum(VI) oxide procured from Fluka Chemie and
phosphoric acid from S D Fine-Chem. Ltd. Aluminum
oxide supplied by Engelhard Corporation. Zirconia sup-
port was prepared using zirconium(IV) isopropoxide as
a precursor. About 40 g of zirconium(IV) isopropoxide
was hydrolysed by slow addition of 20 ml of distilled
water until the formation of a white precipitate. This
precipitate was filtered and washed with distilled water
and then dried at 100◦C for 10 h. The solid product of
zirconium(IV) hydroxide is calcined at 500◦C for 5 h.
Titania support was also prepared by using titanium(IV)
isopropoxide as a precursor by similar procedure.
2.1b Synthesis of H4PMo11VO40/ZrO2 catalyst: Pre-
paration of H4PMo11VO40 has been reported else-
where.26 H4PMo11VO40 is prepared by adding requisite
quantities of MoO3, V2O5 and H3PO4 solutions to dis-
tilled water and refluxed at 100◦C. The resulting solid is
dried at 60◦C for 16 h. A series of H4PMo11VO40 cata-
lysts with H4PMo11VO40 loadings ranging from 10–
50 wt% supported on ZrO2 was prepared by the impreg-
nation method. A required amount of HPA was dis-
solved in distilled water, and then ZrO2 was added to
the above solution. The resultant solution was stirred
at room temperature for 5 h. The catalysts were subse-
quently dried at 110◦C for 10 h and calcined in air at
250◦C for 4 h.
2.2 Catalyst characterization
2.2a X-ray diffraction: X-ray powder diffraction
patterns of the samples were obtained with a model: D8
Diffractometer (Advance, Bruker, Germany), using Cu
Kα radiation (1.5406 Å) at 40 kV and 30 mA. Mea-
surements were recorded in steps of 0.045◦ with a count
time of 0.5 s in the range of 2–65◦.
2.2b BET surface area: Specific surface area of the
catalysts were estimated using N2 adsorption isotherms
at −196◦C by the multipoint BET method taking
0.162 nm2 as its cross-sectional area using Autosorb 1
(Quantachrome instruments).
2.2c Raman spectroscopy: Raman spectra of the
catalyst samples were collected with a Horbia-Jobin
Yvon LabRam-HR spectrometer equipped with a

Reactivity of HPA catalysts
447
which serve as pre-heater zone to increase evaporation
of liquid glycerol feed before reaching the catalyst bed.
The catalyst was pretreated with N2 gas at 60 ml/min at
same reaction temperature for 1 h. An aqueous glyce-
rol (10 wt%) solution (0.5 ml/h) was fed from the top
of the reactor through inlet in a flow of N2 gas at
6 ml/min. Reaction products were collected in a cold
trap every hour and were analysed by Gas Chromato-
graph (Shimadzu GC-2014) equipped with flame io-
nization detector using DB-wax 123-7033 capillary col-
umn and methanol as internal standard. Glycerol con-
version and product selectivities are calculated by the
following equations.
confocal microscope, 2400/900 grooves/mm gratings,
and a notch filter. Visible laser excitation at 532 nm
(visible/green) is supplied by a Yag doubled diode
pumped Laser (20 mW). Scattered photons were dried
and focused on to a single-stage monochromator and
measured with a UV-sensitive LN2-cooled CCD detec-
tor (Horbia-Jobin Yvon CCD-3000 V).
2.2d Temperature-programmed desorption of ammo-
nia: Temperature-programmed desorption (TPD)
studies of NH3 were conducted on Auto Chem 2910
(Micromeritics, USA) instrument. In a typical experi-
ment, ca. 100 mg of calcined H4PMo11VO40/ZrO2
sample is taken in a U-shaped quartz cell. The catalyst
sample is packed in one arm of the sample tube on
a quartz wool bed. Prior to TPD studies, the catalyst
sample is pre-treated by passing high purity helium
(50 ml/min) at 200◦C for 1 h. After pre-treatment, the
sample was saturated by passing (50 ml/min) high
purity anhydrous ammonia at 80◦C for 1 h and sub-
sequently flushed with He flow (50 ml/min) at 150◦C
for 1 h to remove physisorbed ammonia. TPD analysis
was carried out from ambient temperature to 600◦C at
a heating rate of 10◦C/min. Ammonia concentration
in the effluent stream was monitored with thermal
conductivity detector, and the area under the peak was
integrated using the software GRAMS/32 to determine
the amount of desorbed ammonia.
Moles of glycerol reacted
Glycerol conversion (%)=
Product selectivity (mol %)=
× 100
Moles of glycerol fed
Moles of carbon in a product defined
Moles of carbon in glycerol reacted
× 100
3. Results and discussion
3.1 Characterization techniques
3.1a X-ray diffraction: X-ray diffraction pat-
terns of the pure zirconia and zirconia-supported
H4PMo11VO40 catalysts are presented in figure 1. In
general, ZrO2 exists in three crystallographic poly-
morphs namely monoclinic, tetragonal and cubic.
X-ray diffraction pattern of pure zirconia used in the
present study was crystalline in nature and having
both monoclinic and tetragonal phases. Sharp X-ray
diffraction lines at d = 3.19, 3.73 and 2.86 Å are due
to m-ZrO2 and X-ray diffraction lines at d = 2.98
and 1.85 Å are due to the t-ZrO2.33 As HPA loading
increases on the zirconia support, intensity of X-ray
diffractions peaks at 2θ = 6.8◦, 10.9◦, 27.5◦ and 33.5◦
also increase. This result clearly suggests formation of
Keggin ion structure of heteropolyacid (HPA) on zir-
conia support. At lower HPA loadings, the above XRD
peaks are absent suggesting that HPA is well-dispersed
on the zirconia support. However, at lower loadings, the
presence of HPA crystalline species of a size less than
4 nm which is beyond the detection limit of powder
X-ray diffraction technique, cannot be ruled out.
2.2e Fourier trans infrared spectroscopy: FT-IR
spectra of the catalysts were recorded on a IR (Model:
GC-FT-IR Nicolet 670) spectrometer by KBr disc
method at room temperature. Ex situ experiments of FT-
IR spectra of pyridine adsorbed samples were carried
out to determine the nature of acidity (Brønsted and
Lewis acid sites). Before recording the IR spectra, pyri-
dine adsorption experiments were carried out by plac-
ing a drop of pyridine on 10 mg of the HPA sample fol-
lowed by evacuation in air for 1 h at room temperature
to remove reversibly adsorbed pyridine on the surface
of the catalyst.35
2.3 Dehydration of glycerol
Vapour phase dehydration reaction of glycerol was car-
ried out at 200–250◦C under atmospheric pressure in a
vertical fixed-bed pyrex glass reactor of 36 cm length
and 0.5 cm internal diameter with a catalyst bed set at
the middle of reactor. The reactor is placed in an elec- 3.1b BET surface area: Surface areas of the ZrO2-
trically heated furnace. Temperature was controlled by supported HPA catalysts are presented in table 1. Spe-
a thermocouple which is located near the catalyst bed cific surface area of the catalysts decreases with HPA
and N2 gas was used as a carrier gas. The glass reactor loading. This decline of surface area with increasing of
was packed with 0.3 g of catalyst, and above the cata- HPA loading is due to blocking the pores of the support
lyst bed, the reactor was packed with ceramic beads by crystalline phase of HPA.

448
Balaga Viswanadham et al.
(e)
(d)
P-O
1100
Mo-Oc-Mo
Mo-Oe-Mo
Mo-O
1000
700
800
900
Wave number (cm-1
)
(c)
(b)
(a)
Mo-O
P-O
Mo-Oc-Mo
Mo-Oe-Mo
700
750
800
850
900
950
1000
1050
1100
Wave number (cm-1)
Figure 1. X-ray diffraction patterns of various wt% of
H4PMo11VO40/ZrO2 catalysts. (a) Pure ZrO2, (b) 10, (c) 20,
(d) 30, (e) 40 and (f) 50.
Figure 2. FT-IR spectra of various wt% of H4PMo11VO40/
ZrO2 catalysts. (a) 10, (b) 20, (c) 30, (d) 40 and (e) 50.
3.1c FT-IR spectroscopy: Figure 2 shows FT-IR
spectra of different wt% of H4PMo11VO40 loadings
supported on ZrO2 recorded in the range of 1100–
700 cm−1. Pure HPA material was also characterized
by FT-IR spectroscopy for better comparison of results
zirconia support. FTIR results are good agreement with
XRD studies.
and it is also shown in figure 2 (inset). Figure 2 shows 3.1d Raman spectroscopy: Raman spectra of tetra-
that all the samples exhibit four well-defined infrared gonal zirconia is characterized by bands at 263, 302,
bands in the range of 1100–700 cm−1at 1056 cm−1, 534 and 645 cm.−1 However, monoclinic zirconia show
950 cm−1, 857, and 790 cm−1.36,37 These IR bands Raman bands with frequencies at 331, 374, 476, 530,
are attributed to vibrational bands of P-O, Mo-O, Mo- 539 and 616 cm−1.38 Laser Raman spectra of pure
Oc-Mo (Oc = corner sharing oxygen) and Mo-Oe-Mo H4PMo11VO40 and various loadings of H4PMo11VO40
(Oe = edge sharing oxygen), respectively. It can be supported on ZrO2 is shown in figures 3 and 4, respec-
seen from figure 2 that the intensity of these four skele- tively. Laser Raman spectra of pure H4PMo11VO40
tal vibrations of Keggin ion are found to increase with and various loadings of HPA/ZrO2 catalysts showed
HPA loading on ZrO2 support. At higher HPA loading Raman bands at 1000, 842, 676 and 300 cm−1. These
(50 wt%), IR bands of Mo-Oc-Mo and Mo-Oe-Mo are Raman bands are assigned to frequencies of υs(Mo-Ot),
clearly observed suggesting the formation of bulk-like υas(Mo-Ot), υs(Mo-Oc-Mo) and υs(Mo-Oa), respec-
HPA on the ZrO2 support. Characteristic bands of four tively. As the HPA loading increases from 10 to 50 wt%
skeletal vibrations observed in all the materials con- on zirconia support, intensity of aforementioned Raman
firm that primary Keggin ion structure is retained on bands increase indicating that the Keggin structure of
Table 1. Results of BET surface area and TPD-NH3 of H4PMo11VO40/ZrO2 catalysts.
NH3 uptake (mmol/g)
HPA loading
BET surface area (m2/g)
Weaka
Moderateb
Strongc
Total NH3 uptake (mmol/g)
10
20
30
40
50
56
42
28
19
8
0.252
0.298
0.340
0.493
0.696
0.230
0.524
2.563
4.850
3.512
0.114
0.078
–
–
–
0.596
0.900
2.903
5.343
4.208
aWeak acidic sites = 150◦–300◦C
bModerate acidic sites = 300◦–450◦C
cStrong acidic sites = 450◦–550◦C

Reactivity of HPA catalysts
449
of HPA indicates bulk-like formation of HPA on zirco-
nia. These findings are in good agreement with results
of H4PMo11VO40/SiO2 reported by Kanno et al.26 Thus,
the above results confirm that at lower loadings, the
Keggin ion of HPA is well-dispersed on the zirconia
support compared to higher loading (50 wt%) and bulk
HPA (shown in figure 3).
Pure H4PMo11VO40
3.1e Temperature-programmed desorption of ammo-
nia: Temperature-programmed desorption (TPD)
profiles of ammonia for H4PMo11VO40/ZrO2 with
H4PMo11VO40 loadings ranging from 10 to 50 wt%
are shown in figure 5 and the amount of NH3 des-
orbed is reported in table 1. It can be seen from
figure 5 that the desorption peak appearing between
150◦ and 300◦C is attributed to weak acidic sites, the
peak noticed between 300◦ and 450◦C is attributed to
moderate acidic sites and the peak observed between
450◦ and 550◦C is attributed to strong acidic sites of
the catalyst. TPD profiles of lower HPA loading of
catalyst showed three desorbed peaks indicating that
it has three types of acidic sites (weak, moderate and
strong). As the HPA loading increases on zirconia
support, weak acidic sites increase gradually from 10–
50 wt% H4PMo11VO40 and strong acidic sites decrease
with increasing HPA loading. Whereas the moderate
acidic sites increase with increase in HPA loading up
to 40 wt%, thereafter, it decreases at higher loadings
(50 wt%). Ammonia uptake values of different wt%
HPA on support are summarized in table 1. It is inter-
esting to see that total acidity of catalyst increases up
to 40 wt% and it is decreases at 50 wt% HPA loading.
Decrease in acidity at higher HPA loading might be due
200
400
600
800
1000
1200
Wave number (cm-1)
Figure 3. Raman spectra of pure H4PMo11VO40 acid.
heteropolyacid remains intact in all the calcined sam-
ples. However, the sample with 40 wt% HPA showed
clear characteristic Keggin ion band at 1000 cm−1 and
this band is shifted to 1005 cm−1 at higher loadings.
Stretching frequencies of υas(Mo-Ot) and υs(Mo-Oa)
observed clearly at 846 and 300 cm−1 at higher loading
(f)
(e)
(d)
(e)
(d)
(c)
(b)
(c)
(b)
(a)
(a)
200
400
600
800
1000
1200
0
100
200
300
400
500
600
Wave number (cm-1)
Temperature (oC)
Figure 4. Raman spectra of various wt% of H4PMo11VO40/
ZrO2 catalysts. (a) Pure ZrO2, (b) 10, (c) 20, (d) 30, (e) 40
and (f) 50.
Figure 5. TPD-NH3 of various wt% of H4PMo11VO40/ZrO2
catalysts. (a) 10, (b) 20, (c) 30, (d) 40 and (e) 50.

450
Balaga Viswanadham et al.
to formation of high crystalline HPA (bulk-like) and it
was decomposed at higher loadings leading to decrease
in acidity. This was also noticed from the previously
described results of FT-IR and Raman studies. There is
a considerable amount of strong acidic sites at lower
loadings which decreased considerably at higher load-
ings. TPD of ammonia shows a clear shoulder peak at
moderate temperature region and a deconvoluted peak
(figure 5). The deconvoluted peak indicates that mode-
rate acidity increases with increase in HPA loading and
decreases at higher HPA loadings.
B + L
L
B
(c)
(b)
3.1f FT-IR spectroscopy of pyridine adsorption: The
nature of acidic sites (Brønsted and Lewis acidic sites)
of HPA supported on ZrO2 is also examined by ex
situ FT-IR of pyridine-adsorbed samples. Ex situ pyri-
dine FT-IR spectra of various wt% of H4PMo11VO40
on ZrO2 catalysts in the range of 1600–1400 cm−1 are
shown in figure 6. The IR band appearing at 1537 cm−1
is due to Brønsted acidic sites, the IR peak appearing
at 1486 cm−1 is assigned to both Brønsted and Lewis
(B+L) acidic sites and the IR band at 1443 cm−1 band is
assigned exclusively to Lewis acidic sites. As seen from
(a)
1600
1550
1500
1450
1400
Wave number (cm-1)
Figure 7. FT-IR spectra of pyridine adsorption of 40 wt%
of H4PMo11VO40 on various supports of (a) ZrO2, (b) TiO2
and (c) Al2O3.
in figure 6, Brønsted acidity increases with increase in
HPA loading up to 40 wt% and decreases at higher
loading. However, Lewis acidity decreases gradually
up to 40 wt% of HPA and increases at higher load-
ings (>40 wt% catalyst). Intensity of IR bands of both
Brønsted and Lewis acidic sites at 1486 cm−1 increases
with HPA loading up to 40 wt% and at higher load-
ing, both the B+L acidic sites decrease. These results
are in good agreement with previously described acidity
measurements by ammonia TPD method.
B + L
B
L
(e)
A comparative study of acidity exhibited by 40 wt%
HPA on different supports (ZrO2, TiO2 and Al2O3) are
examined by ex situ FT-IR spectra of pyridine in the
range of 1600–1400 cm−1 (figure 7). As seen in fig-
ure 7, intensity of IR peaks for B and B + L acidic sites
of H4PMo11VO40/ZrO2 is found to be higher compared
to H4PMo11VO40/TiO2 and H4PMo11VO40/Al2O3.
However, intensity of L acidic sites peak of
H4PMo11VO40/ZrO2 is found to be much lower than
H4PMo11VO40/TiO2 and H4PMo11VO40/Al2O3. These
results clearly demonstrate that H4PMo11VO40 sup-
ported on ZrO2 exhibits higher amount of Brønsted
acidic sites than HPA supported on TiO2, Al2O3.
(d)
(c)
(b)
(a)
3.2 Catalytic activity studies
1600
1500
1400
Wave number (cm-1)
3.2a Effect of HPA loading on ZrO2: Catalytic pro-
perties exhibited by various HPA catalysts-supported
ZrO2 during vapour phase dehydration of glycerol are
shown in figure 8 and the product distribution in table 2.
Figure 6. FT-IR spectra of pyridine adsorption of various
wt% of H4PMo11VO40/ZrO2 catalysts. (a) 10, (b) 20, (c) 30,
(d) 40 and (e) 50.

Reactivity of HPA catalysts
451
selectivity also increases with HPA loading in a simi-
lar way. Therefore, Brønsted acidic sites play an impor-
tant role in glycerol dehydration to form acrolein selec-
tively. A considerable decrease in selectivity of acrolein
is noticed at 50 wt% HPA due to presence of more
number of Lewis acidic sites and also less amount of
moderate acidic sites (table 1). The 40 wt% HPA/ZrO2
exhibited 89% acrolein selectivity with complete con-
version of glycerol. Acetic acid selectivity decreases
with increase in HPA loading up to 40 wt% of HPA on
the support and increases at higher loadings (table 2).
This is due to decrease in intensity of L acidic sites peak
up to 40 wt% and increases at higher loading (figure 6).
Results of catalytic properties are well-correlated with
the nature of acidic sites and also with total acidity of
catalysts.
110
100
90
80
70
60
50
40
30
20
10
0
110
100
90
80
70
60
50
40
30
20
10
0
% Conversion of glycerol
% Acrolein selectivity
% Acetic acid selectivity
10
20
30
40
50
HPA loading (wt %)
Figure 8. Conversion of aqueous glycerol/acrolein selec-
tivity and wt% of H4PMo11VO40/ZrO2 catalyst at 225◦C.
As seen in figure 8 conversion of glycerol did not appre- 3.2b Effect of reaction temperature: Effect of reac-
ciably change with HPA loading. It might be due to tion temperature on dehydration of glycerol was exa-
low concentration of glycerol employed in the feed. mined over 40 wt% HPA on zirconia support and the
However, HPA loading in the catalyst considerably product distribution results are reported in table 3. As
influences acrolein selectivity during glycerol dehy- seen from the results in table 3, selectivity of acrolein
dration. It is found that acrolein selectivity increases and acetic acid varies with reaction temperature. How-
with increase in HPA loading up to 40 wt% HPA ever, conversion of glycerol increases with increase in
and decreases at higher HPA loadings. It is interest- reaction temperature up to 225◦C and remains con-
ing to note from the ammonia TPD results that mode- stant at higher temperatures. Presence of vanadium in
rate or total acidity increases with increase in HPA heteropolyacid prevents deactivation of the catalyst at
loading up to 40 wt% and further decreases at higher higher temperatures. This is due to the oxidative abi-
loading. Thus, the above results clearly suggest that lity of vanadium-containing heteropolyacids during the
acrolein selectivity depends mainly on acidity of the reaction at higher reaction temperature.26,28,31–34,36,39–42
catalyst. The generally accepted reaction for glycerol As the reaction temperature increases from 200◦ to
dehydration is shown in scheme 1.
250◦C, acrolein selectivity increases from 65% to 89%
According to literature, Brønsted acidic sites are and thereafter decreases to 76%, whereas acetic acid
responsible for the formation of acrolein.3 As seen in selectivity increases from 2% to 10% during the reac-
figure 6 that Brønsted acidic sites increase up to 40 wt% tion. Significant changes in product selectivity are prob-
HPA loading and decreases at higher loadings. Acrolein ably due to variation in the nature of acidic sites of
Table 2. Product distribution results of glycerol dehydration over various wt% of H4PMo11VO40/ZrO2 at 225◦Cx.
Selectivity (mol%)
HPA loading on ZrO2
Cgly(%)a
Acb
Acec
Acetad
Othery
10
20
30
40
50
100
100
100
100
100
35
48
65
89
76
38
25
21
3
10
8
4
4
2
17
19
10
4
16
6
xResults obtained after 4 h. Reaction conditions: catalyst weight = 0.3 g, feed = 0.5 mLh−1, 10 mLmin−1 gas flow rate (N2),
10 wt% aqueous glycerol solution, reaction temperature = 225◦C
aConversion of glycerol
bAcrolein selectivity
cAcetic acid selectivity
dAcetaldehyde selectivity
yHydroxyacetone, acetone, other products

452
Balaga Viswanadham et al.
O
Brønsted acidic sites
H
- 2H2O
OH
H4PMo11VO40/ZrO2
Acrolein
HO
OH
O
Lewis acidic sites
OH
O
O
Glycerol
OH
Acetic acid
H
- H2O
Hydroxy acetone
Acetaldehyde
Scheme 1. Glycerol dehydration over H4PMo11VO40/ZrO2 catalysts.
vanadium-containing heteropolyacid anion at higher properties depend on the nature of support and inter-
reaction temperature.
action between the support and HPA. Thus, ZrO2 as a
support exhibits much better catalytic performance than
other high surface area Al2O3 and TiO2 supports indi-
cate that ZrO2 support strongly interacts with primary
Keggin unit of HPA than other supports.
3.2c Effect of support: Superior catalytic perfor-
mance of HPA/ZrO2 catalysts is also compared with
HPA supported on TiO2 and Al2O3 and the results are
reported in table 4. Acrolein selectivity is found to be
higher in the case of ZrO2-supported HPA compared
to TiO2 and Al2O3-supported catalysts. Enhancement 3.2d Effect of time on stream: Effect of time on
in the selectivity of acrolein over HPA/ZrO2 is due to stream on catalytic properties during glycerol dehy-
availability of more number of Brønsted acidic sites dration by using 40 wt% H4PMo11VO40 on different
(figure 7). However, it is observed that acetic acid selec- supports is shown in figure 9. As seen in figure 9,
tivity is high in the case of Al2O3-supported catalysts HPA/ZrO2 exhibited complete conversion of glycerol
compared to other supports such as TiO2 and ZrO2 with stable activity over a period of 10 h on stream
(table 4). This might be due to the presence of more with high acrolein selectivity (10 wt% aqueous glycerol
number of L acidic sites in HPA/Al2O3 than the other solution). However, HPA/TiO2 shows stable activity up
supports as revealed from the FT-IR spectra of pyridine to 8 h and decreases gradually with time on stream.
adsorption of 40 wt% on various supports (figure 7). It Similarly, acrolein selectivity also decreases with time
is interesting to note that irrespective of support used on stream. In the case of HPA/Al2O3, stable activity is
in dehydration of glycerol, there is no change in the noticed only up to 7 h and further it decreased. Sta-
glycerol conversion due to redox nature of vanadium- ble activities of ZrO2-supported catalyst compared to
containing HPA.26,28,31–34,36,39–42 However, the supports other supported catalysts are due to strong interaction
considerably affect selectivity of acrolein and other between active phase of HPA and zirconia supports than
products. Acrolein selectivity of H4PMo11VO40 sup- other supports such as alumina and titania. Decrease in
ported on different supports (table 4) are well-correlated glycerol conversion and acrolein selectivity over HPA
with the results of FT-IR spectra of pyridine-adsorbed supported on Al2O3 and TiO2 is due to significant deac-
samples (figure 7). This results show that catalytic tivation of HPA catalyst or decomposition of Keggin ion
Table 3. Product distribution results of glycerol dehydration over 40 wt% H4PMo11VO40/ZrO2 catalyst at 200◦–250◦CX.
Selectivity (mol%)
Reaction temperature (◦C)
Cgly(%)a
Acb
Acec
Acetad
Othery
200
225
250
91
100
100
65
89
78
2
3
10
12
4
2
21
4
10
xResults obtained after 4 h. Reaction conditions: catalyst weight = 0.3 g, feed = 0.5 mLh−1, 10 mLmin−1 gas flow rate (N2),
10 wt% aqueous glycerol solution, reaction temperature = 200◦–250◦C
aConversion of glycerol
bAcrolein selectivity
cAcetic acid selectivity
dAcetaldehyde selectivity
yHydroxyacetone, acetone products

Reactivity of HPA catalysts
453
Table 4. Product distribution results of glycerol dehydration over 40 wt% H4PMo11VO40 on different support catalyst at
225◦CX.
Selectivity (mol%)
HPA on support
BET surface area (m2/g)
Cgly(%)a
Acb
Acec
Acetad
Othery
ZrO2
TiO2
Al2O3
19
15
146
100
100
100
89
76
53
3
18
28
4
3
5
4
3
13
xResults obtained after 4 h. Reaction conditions: catalyst weight = 0.3 g, feed = 0.5 mLh−1, 10 mLmin−1 gas flow rate (N2),
10 wt% aqueous glycerol solution, reaction temperature = 225◦C
aConversion of glycerol
bAcrolein selectivity
cAcetic acid selectivity
dAcetaldehyde selectivity
yHydroxyacetone, acetone, other products
180
160
140
120
100
80
180
160
140
120
100
80
40 wt% HPA loading possesses higher amount of mode-
rate acidic sites and also high total acidity compared to
other HPA loadings of the catalysts. Ex situ FT-IR of
pyridine adsorption results suggests that 40 wt% HPA
has higher amount of Brønsted acidity than other load-
ings. Amount of Brønsted acidity is also higher in HPA
supported on ZrO2 than Al2O3 and TiO2-supported cata-
lysts. Acrolein selectivity during vapour phase dehy-
dration of glycerol is well-correlated with the mode-
rate or total acidity and Brønsted acidity of the cata-
lysts. Furthermore, this study reveals optimum compo-
sition of H4PMo11VO40 over zirconia catalyst (40 wt%)
for obtaining good catalytic performance in dehydra-
tion of glycerol to acrolein at 225◦C under atmospheric
pressure (mild reaction conditions).
% Conversion of glycerol on HPA/ZrO2
% Acrolein selectivity on HPA/ZrO2
% Conversion of glycerol on HPA/TiO2
% Acrolein selectivity on HPA/TiO2
% Conversion of glycerol on HPA/Al2O3
% Acrolein selectivity on HPA/Al2O3
60
60
40
40
20
20
0
1
2
3
4
5
6
7
8
9
10
11
Time on stream (h)
Figure 9. Time on stream studies of 40 wt% HPA on ZrO2,
TiO2 and Al2O3 support at 225◦C.
Acknowledgements
structure of HPA on the support during the course of the
reaction.
The authors thank the Director, IICT, Hyderabad. BV
thanks the Council of Scientific and Industrial Research
(CSIR), New Delhi for the award of Senior Research
Fellowship. This work is supported by CSIR, New
Delhi under the Network projects of XII -5 year plan
(Indus Magic-WP3).
4. Conclusion
Vanadium-containing phosphomolybdic acid catalysts
supported on ZrO2 polymorphs are found to be highly
active and selective for dehydration of glycerol to
acrolein compared to HPA supported on Al2O3 and
TiO2. This is mainly due to the presence of highly dis-
persed H4PMo11VO40 Keggin structure on ZrO2 support
leading to strong interaction between primary structure
of Keggin ion and ZrO2 support. XRD results suggest
that HPA is present in a highly dispersed state at lower
loadings. FT-IR and Raman spectroscopy confirm that
the characteristic Keggin ion of HPA remains intact
with the support. TPD of ammonia results confirms that
References
1. Zhou C H, Beltramini J N, Fana Y X and Lu G Q 2008
Chem. Soc. Rev. 37 527
2. Jia C J, Liu Y, Schmidt W, Lu A H and Schüth F 2010 J.
Catal. 269 71
3. Katryniok B, Paul S, Baca V B, Rey P and Dumeignil F
2010 Green. Chem. 12 2079
4. Pagliaro M, Ciriminna R, Kimura H, Rossi M and Pina
C D 2007 Ang. Chem. Int. Ed. 46 4434
5. Behr A, Eilting J, Irawadi K, Leschinski J and Lindner
F 2008 Green. Chem. 10 13

454
Balaga Viswanadham et al.
6. Kim Y T, Jung K D and Park E D 2011 Appl. Catal. B: 24. Chai S H, Wang H P, Liang Y and Xu B Q 2009 Appl.
Environ. 107 177
7. Tao L Z, Chai S H, Zuo Y, Zheng W T, Liang Y and Xu
B Q 2010 Catal. Today 158 310
8. Atia H, Armbruster U and Martin A 2011 Appl. Catal.
A: Gen. 393 331
Catal. A: Gen. 353 213
25. Devassy B M and Halligudi S B 2005 J. Catal. 236 313
26. Kanno M, Yasukawa T, Ninomiya W, Ooyachi K and
Kamiya Y 2010 J. Catal. 273 1
27. Barteau K P, Lyons J E, Song I K and Barteau M A 2006
Top. Catal. 41 55
28. Song I K and Barteau M A 2004 J. Mol. Catal. A: Chem.
212 229
9. Deleplanque J, Dubois J L, Devaux J F and Ueda W
2010 Catal. Today 157 351
10. Ulgen A and Hoelderich W F 2011 Appl. Catal. A: Gen.
400 34
11. Cavani F, Guidetti S, Marinelli L, Piccinini M, Ghedini
E and Signoretto M 2010 Appl. Catal. B: Environ. 100
197
29. Weber R S 1994 J. Phys. Chem. 98 2999
30. Baghernejad B, Heravi M M, Oskooie H A and
Bamoharram F F 2012 Bull. Chem. Soc. Ethiop. 26 145
31. Kanno M, Miura Y K, Yasukawa T, Hasegawa T,
Ninomiya W, Ooyachi K, Imai H, Tatsumi T and
Kamiya Y 2011 Catal. Commun. 13 59
32. Benadji S, Eloy P, Léonard A, Su B L, Bachari K, Rabia
C and Gaigneaux E M 2010 Micro. Meso. Mater. 130
103
33. Romanelli G P, Villabrille P I, Cáceres C V, Vázquez P
G and Tundo P 2011 Catal. Commun. 12 726
34. Erfle S, Armbruster U, Bentrup U, Martin A and Brück-
ner A 2011 Appl. Catal. A: Gen. 391 102
35. Kaba M S, Barteau M A, Lee W Y and Song I K 2000
Appl. Catal. A: Gen. 194–195 129
12. Pathak K, Reddy K M, Bakhshi N N and Dalai A K 2010
Appl. Catal. A: Gen. 372 224
13. Suprun W, Lutecki M, Gläser R and Papp H 2011 J. Mol.
Catal. A: Chem. 342–343 91
14. Alhanash A, Kozhevnikova E F and Kozhevnikov I V
2010 Appl. Catal. A: Gen. 378 11
15. Wang F, Dubois J L and Ueda W 2010 Appl. Catal. A:
Gen. 376 25
16. Garbay P L, Millet J M M, Loridant S, Baca V B and
Rey P 2011 J. Catal. 280 68
17. Yoda E and Ootawa A 2009 Appl. Catal. A: Gen. 360 66
18. Gu Y, Cui N, Yu Q, Li C and Cui Q 2012 Appl. Catal.
A: Gen. 429–430 9
19. Sancho C G, Tost R M, Robles J M, González J S, López
A J and Torres P M 2012 Appl. Catal. A: Gen. 433–434
179
36. Zhang J, Tang Y, Li G and Hu C 2005 Appl. Catal. A:
Gen. 278 251
37. Ilkenhans T, Herzag B, Braun T and Schlogl R 1995 J.
Catal. 153 275
38. Shi L, Tin C K and Wong B N 1999 J. Mater. Sci. 34
3367
20. Wang F, Dubois J L and Ueda W 2009 J. Catal. 268 260
21. Chai S H, Wang H P, Liang Y and Xu B Q 2008 Green.
Chem. 10 1087
22. Chary K V R, Ramesh K, Naresh D, Rao P V R, Rao A
R and Rao V V 2009 Catal. Today 141 187
23. Chary K V R, Ramesh K, Vidyasagar G and Rao V V
2003 J. Mol. Catal. A: Gen. 198 195
39. Collins F M, Lucy A R and Sharp C 1997 J. Mol. Catal.
A: Chem. 117 397
40. Barteau M A, Lyons J E and Song I K 2003 J. Catal. 216
236
41. Liu H and Iglesia E 2003 J. Phys. Chem. B107 10840
42. Choi J H, Park D R, Park S and Song I K 2011 Catal.
Lett. 141 826