Effect of varying reflux durations on the physico-chemical and catalytic performance of vanadium phosphate catalysts synthesized via vanadyl hydrogen phosphate sesquihydrate

Article abstract of DOI:10.1016/j.apcata.2011.12.007

A series of vanadyl pyrophosphate, (VO)₂P₂O ₇, catalysts prepared via vanadyl hydrogen phosphate sesquihydrate precursors (VOHPO₄·1.5H₂O) was calcined in a reaction flow of 0.75% n-butane in air mixture at 733 K for 18 h. The precursors have been synthesized by refluxing vanadyl phosphate dihydrate (VOPO ₄·2H₂O) with 1-butanol for different lengths of time, i.e. 8, 15 and 24 h, and the produced catalysts were denoted as VPO _s-R8, VPO_s-R15 and VPO_s-R24, respectively. X-ray diffraction (XRD) patterns of the three catalysts showed similar diffraction pattern, comprised of a well-crystallized (VO)₂P ₂O₇ phase. Brunauer-Emmett-Teller (BET) surface area measurements showed that VPO_s-R24 has the highest specific surface area, i.e. 31 m² g^-1 followed by 27 m² g ^-1 and 19 m² g^-1 for VPO_s-R15 and VPO_s-R8, respectively. Inductively coupled plasma (ICP) analyses indicated that the P/V atomic ratios of these catalysts were in the optimum range in producing (VO)₂P₂O₇ phase. A small increment in the average oxidation number of the vanadium was observed as the precursor reflux duration increased. Scanning electron microscope showed the secondary structures of the catalysts with plate-like crystals in different sizes, which were agglomerated into rosette-shape clusters. The total amount of oxygen desorbed from the catalysts increased as the precursor reflux duration increased. Temperature-programmed reduction (TPR) in H₂ profiles of all the catalysts gave three reduction peaks. VPO_s-R8 gave the highest total amount of oxygen removed from V⁵⁺/V⁴⁺ phase followed by VPO_s-R15 and VPO_s-R24. Catalytic tests revealed that the catalyst with lower precursor reflux duration exhibited higher selectivity but lower activity and vice versa.

Full text of DOI:10.1016/j.apcata.2011.12.007

Applied Catalysis A: General 415–416 (2012) 53–58
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Effect of varying reflux durations on the physico-chemical and catalytic
performance of vanadium phosphate catalysts synthesized via vanadyl hydrogen
phosphate sesquihydrate
L.K. Leong , K.S. Chin , Y.H. Taufiq-Yapb
a,∗
a
a
Department of Chemical Engineering, Faculty of Engineering and Science, Universiti Tunku Abdul Rahman, Jalan Genting Kelang, 53300 Setapak, Kuala Lumpur, Malaysia
Centre of Excellence for Catalysis Science and Technology, Faculty of Science, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 March 2011
Received in revised form
A series of vanadyl pyrophosphate, (VO) P O , catalysts prepared via vanadyl hydrogen phosphate
2
2
7
sesquihydrate precursors (VOHPO4·1.5H2O) was calcined in a reaction flow of 0.75% n-butane in air
mixture at 733 K for 18 h. The precursors have been synthesized by refluxing vanadyl phosphate dihy-
drate (VOPO4·2H2O) with 1-butanol for different lengths of time, i.e. 8, 15 and 24 h, and the produced
catalysts were denoted as VPOs-R8, VPOs-R15 and VPOs-R24, respectively. X-ray diffraction (XRD) pat-
terns of the three catalysts showed similar diffraction pattern, comprised of a well-crystallized (VO)2P2O7
phase. Brunauer–Emmett–Teller (BET) surface area measurements showed that VPOs-R24 has the high-
3
0 November 2011
Accepted 2 December 2011
Available online 9 December 2011
Keywords:
2
−1
2
−1
2
−1
est specific surface area, i.e. 31 m g followed by 27 m g and 19 m g for VPOs-R15 and VPOs-R8,
respectively. Inductively coupled plasma (ICP) analyses indicated that the P/V atomic ratios of these
catalysts were in the optimum range in producing (VO)2P2O7 phase. A small increment in the average
oxidation number of the vanadium was observed as the precursor reflux duration increased. Scanning
electron microscope showed the secondary structures of the catalysts with plate-like crystals in different
sizes, which were agglomerated into rosette-shape clusters. The total amount of oxygen desorbed from
the catalysts increased as the precursor reflux duration increased. Temperature-programmed reduction
Vanadyl hydrogen phosphate
sesquihydrate precursor
Vanadyl pyrophosphate
Selective oxidation
Maleic anhydride
(
TPR) in H2 profiles of all the catalysts gave three reduction peaks. VPOs-R8 gave the highest total amount
5
+
4+
of oxygen removed from V /V phase followed by VPOs-R15 and VPOs-R24. Catalytic tests revealed that
the catalyst with lower precursor reflux duration exhibited higher selectivity but lower activity and vice
versa.
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
surface chemistry and bulk properties of the active VPO catalysts
change with the time and condition of calcination were reported
[10]. Also, the catalytic properties of this catalyst could be improved
by the incorporation of specific promoters, which encourage higher
selectivity towards maleic anhydride [11,12].
The high environmental impact and mounting prices of benzene
in the early 1970s have resulted in the finding of other feedstocks
for maleic anhydride [1]. Consistent researches in both academic
and industrial sectors have shed considerable light on the nature
of a catalyst performing an extremely selective, 14-electron oxida-
tive conversion of n-butane to maleic anhydride [2]. This industrial
process is still the only commercialized alkane selective oxidation
by heterogeneous gas-solid catalytic process, which is specifically
based on the presence of vanadyl pyrophosphate, (VO) P O7 [3,4].
In the present study, the latest alternative route in pro-
ducing (VO) P O7 catalyst, which is via vanadyl hydrogen
2
2
phosphate sesquihydrate precursor (VOHPO ·1.5H O) [13–15] has
4
2
been employed. The effect of varying reflux durations in the
preparation of the sesquihydrate precursors towards the physico-
chemical properties and catalytic performances of the catalysts was
examined.
2
2
(
VO) P O7 has been identified as the critical active phase for the
2 2
selective oxidation of n-butane [5].
Recently, different researchers have extensively studied a vari-
ety of modifications for the preparation pathway of vanadium
phosphorus oxide (VPO) catalyst such as calcination conditions
2. Experimental
[
6], incorporating promoters [7] and P/V atomic ratios [8,9]. The
2.1. Preparation of vanadyl pyrophosphate catalysts via
sesquihydrate precursors
∗ Corresponding author. Tel.: +60 341079802; fax: +60 341079803.
E-mail addresses: leonglk@utar.edu.my, lloongkong@yahoo.com (L.K. Leong).
The synthesis of VOHPO ·1.5H O precursor involving vanadyl
4
2
phosphate dihydrate, VOPO ·2H O as an intermediate product,
4
2
0
926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.12.007

5
4
L.K. Leong et al. / Applied Catalysis A: General 415–416 (2012) 53–58
comprised of two reflux stages. In the first stage, vanadium (V)
oxide, V O5 (15.0 g from Merck) and ortho-phosphoric acid, o-
(
2 0 4)
•
1
200
2
(0 2 0)
•
•
(VO) P O
(
2 2 1)
H PO (90 ml, 85% from Merck) were refluxed in distilled water
3
4
•
−1
(
24 ml g
solid) with continuous stirring at 393 K for 24 h. The
1000
resulting yellow solid (VOPO ·2H O) was recovered by centrifu-
4
2
VPO -R24
gation and oven-dried at 358 K for 72 h.
8
6
4
2
00
00
00
00
0
•
•
In the second stage, 10.0 g of the synthesized VOPO ·2H O
•
•
4
2
(
50.5 mmol) was refluxed with 1-butanol (150 ml from Merck) and
•
•
continuously stirred at 393 K for different lengths of time, i.e. 8, 15
VPO -R15
VPO -R8
and 24 h. The resulting whitish-blue solids (VOHPO ·1.5H O pre-
4
2
cursors) were recovered by centrifugation and washed sparingly
with a small amount of acetone and finally oven-dried at 358 K for
7
2 h. These precursors were denoted as VPOs-R8Pre, VPOs-R15Pre
and VPOs-R24Pre, with respect to the different lengths of reflux
duration.
The three sesquihydrate precursors were calcined in a reaction
flow of 0.75% n-butane in air mixture at 733 K for 18 h to gener-
1
0
20
30
40
50
60
2θ / °
ate the active catalysts, (VO) P O7. The activated catalysts were
2
2
Fig. 1. Powder XRD patterns for VPOs catalysts.
denoted as VPOs-R8, VPOs-R15 and VPOs-R24.
and (2 2 1) planes, respectively (JCPDS File No. 34-1381). No V5+
phase could be detected in the XRD profile, which could be due to
the reason that the amount of this particular phase was below the
detection limit of the XRD instrument.
All the three major refection planes (2ꢀ = 22.9 , 28.4 and 29.9 )
exhibited to be more intense with about similar width sizes as
the precursor reflux duration prolonged. Since the I0 2 0/I2 0 4 ratios
remained approximately the same, the degree of crystallinity of
2
.2. Catalysts characterization
The X-ray diffraction (XRD) patterns were obtained by using
a Shimadzu model XRD-6000 diffractometer employing Cu K␣
radiation to generate diffraction patterns from powder crystalline
samples at ambient temperature.
The total surface areas of the catalysts were measured by
Brunauer–Emmett–Teller (BET) method by using nitrogen adsorp-
tion at 77 K. This was done by using a Thermo Finnigan Sorptomatic
◦
◦
◦
(0 2 0) and (2 0 4) was not significantly affected. The estimation of
the crystallite sizes based on X-ray peak broadening was calculated
by using the Debye–Scherrer equation:
1
990 nitrogen adsorption–desorption analyzer.
The bulk chemical composition was determined by using a
sequential scanning inductively coupled plasma-optical emission
spectrometer (ICP-OES) Perkin Elmer Optical Emission Spectrome-
ter Optima 2000 DV.
The average oxidation states of the vanadium in all the catalysts
were determined by redox titration following the method of Niwa
and Murakami [16].
0
.89ꢁ
t =
ˇh k l cos ꢀh k l
where t is the crystallite size for (h k l) phase, ꢁ is the X-ray wave-
length of radiation for Cu K␣, ˇh k l is the full-width at half maximum
FWHM) at (h k l) peak and ꢀh k l is the diffraction angle for (h k l)
phase [17].
The FWHM of the (0 2 0) and (2 0 4) planes was used to deter-
(
Scanning electron microscopy (SEM) analyses were done by
using a Supra 55VP electron microscope. Energy-dispersive X-ray
mine the crystallite size of the catalysts. Increasing the precursor
duration did not show a specific trend to the crystallite sizes for
both the reflection planes, unlike the results found in increasing cal-
cination duration [15]. As tabulated in Table 1, the crystallite sizes
of (0 2 0) reflection plane for VPOs-R8, VPOs-R15 and VPOs-R24
were calculated as 108.28 A˚ , 109.16 A˚ and 103.39 A˚ , respectively
(
EDX) analyses were carried out by using EDAX software.
Temperature-programmed desorption (TPD) of O2 and
temperature-programmed reduction (TPR) in H2 analyses were
done by using a Thermo Electron TPDRO 1100 apparatus utilizing
a thermal conductivity detector (TCD).
whereas the crystallite sizes of (2 0 4) reflection plane for the same
series of catalysts produced were 209.37 A˚ , 206.96 A˚ and 213.34 A˚ ,
2.3. Catalytic oxidation of n-butane to maleic anhydride
respectively.
The oxidation of n-butane was carried out in a fixed-bed
−1
microreactor at 673 K with GHSV = 2400 h using a standard mass
of catalyst (250 mg). A mixture of 1.0% n-butane and air was fed to
the reactor via calibrated mass flow controller. The products were
then fed via heated lines to an on-line gas chromatograph for analy-
sis. The reactor comprised of a stainless steel tube with the catalyst
held in place by plugs of quartz wool. A thermocouple was located in
the centre of the catalyst bed and temperature control was typically
at ± 1 K. Carbon mass balances of ≥95% were typically observed.
3.2. BET surface area measurements and chemical analyses
BET surface areas of the catalysts obtained are as follows:
2
−1
2
−1
2
−1
19 m g for VPOs-R8, 27 m g for VPOs-R15 and 31 m g for
VPOs-R24 (Table 2). Increasing the precursor reflux duration has
Table 1
XRD data of VPOs catalysts.
3
. Results and discussion
a
b
Catalysts
Line width
Line width
(2 0 4) ( )
Crystallite
Crystallite
◦
◦
sizec (0 2 0)
( A˚ )
sizec (2 0 4)
( A˚ )
(
0 2 0) ( )
3.1. X-ray diffraction (XRD)
VPOs-R8
VPOs-R15
VPOs-R24
0.7400
0.7340
0.7750
0.3869
0.3914
0.3797
108.28
109.16
103.39
209.37
206.96
213.34
XRD patterns for VPOs-R8, VPOs-R15 and VPOs-R24 catalysts
are shown in Fig. 1. The studies showed that a similar diffraction
pattern, which only comprised of a well-crystallized (VO) P O7
2
2
a
FWHM of (0 2 0) reflection plane.
FWHM of (2 0 4) reflection plane.
phase was obtained. Three main characteristic peaks appeared at
b
c
◦
◦
◦
2
ꢀ = 22.9 , 28.4 and 29.9 , which are corresponded to (0 2 0), (2 0 4)
Crystallite thickness by means of the Debye–Scherrer’s formula.

L.K. Leong et al. / Applied Catalysis A: General 415–416 (2012) 53–58
55
Table 2
Specific BET surface areas, chemical compositions, average oxidation numbers and percentages of V and V5+ oxidation states present in VPOs catalysts.
4
+
Catalyst
Specific BET surface
P/V atomic ratio
V4+ (%)
V5+ (%)
Average oxidation
number
area (m2
g
−1
)
ICP
EDX
VPOs-R8
VPOs-R15
VPOs-R24
19
27
31
1.03
1.03
1.04
1.12
1.21
1.25
95.81
93.09
90.64
4.19
6.91
9.36
4.0419
4.0691
4.0936
gradually increased the specific surface area of the catalysts. Evi-
dence on higher specific surface area was also shown by the SEM
micrographs and will be discussed in detailed in the following
section. The results were found to be different from the previous
findings on the effect of calcination duration [15].
Chemical analyses using ICP-OES indicated that the P/V atomic
ratios for VPOs-R8, VPOs-R15 and VPOs-R24 catalysts were 1.03,
1
.03 and 1.04, respectively, whereas EDX analyses showed that the
values were in between 1.12 and 1.25 (Table 2). The results were in
close agreement with the optimal P/V atomic ratio range for pro-
ducing the (VO) P O7 phase [5]. The average oxidation numbers of
2
2
5+
4+
the vanadium and percentages of V and V oxidation state are
summarized in Table 2. The average oxidation numbers of the vana-
dium were found to be directly proportional to the precursor reflux
duration. This increment in the average oxidation state of the vana-
dium was due to the presence of V5 phase, as confirmed in the TPR
in H2 analyses (Section 3.5). Since the presence of V5+ phase was
very small, it was not shown in the XRD patterns (Section 3.1) due
to the fact that it was below the XRD detection limit. However, the
increase in the average oxidation numbers of the vanadium found
in this series of catalysts could be highly related to the activation
in an oxidizing feed, i.e. lower than 1.5% n-butane in air balance.
Ultimately, this had caused the longer reflux time precursors to be
more easily oxidized.
+
3.3. Scanning electron microscopy (SEM)
The surface morphologies of all the catalysts prepared via dif-
ferent precursor reflux durations (i.e. 8, 15, 24 h) are shown in
Fig. 2(a)–(c). All the three catalysts showed secondary structures,
consisting of different sizes of plate-like crystals, which were
agglomerated into the characteristic rosette-shape clusters. These
rosette-like crystals were comprised of agglomerates of (VO) P O7
2
2
platelets that were preferentially exposing the (1 0 0) crystal plane
18]. (VO) P O7 catalysts obtained via conventional organic and
[
2
2
dihydrate methods could also find a similar type of platelet arrange-
ment [15,18–20].
The surface of each plate of the crystal, typically in a range of
2
–3 ␮m in lateral dimensions (Fig. 2(a)–(c)) was getting rougher
with cracks as the precursor reflux duration increased. This effect
could have contributed to the increased in specific surface area of
the catalysts, which was found in the BET surface area measure-
ments. Also, the crystal plates were flattened and having folded
edges when the precursor reflux duration increased. This interest-
ing phenomenon was also observed when the calcination duration
increased from 10 to 75 h prepared via the same sesquihydrate
precursor route [15].
3.4. Temperature-programmed desorption (TPD) of O2
TPD of O2 profiles shown in Fig. 3 was obtained by pretreat-
ing the fresh catalysts by heating them to 673 K in an oxygen flow
3
−1
(
1 bar, 25 cm min ) and held at 673 K for 30 min in the same
Fig. 2. SEM micrographs for (a) VPOs-R8, (b) VPOs-R15 and (c) VPOs-R24 catalysts.
stream before cooling to ambient temperature. Then, the flow was
3
−1
switched to helium (1 bar, 25 cm min ) and the temperature was
raised to 1173 K. The peak maxima temperatures, the amount of

5
6
L.K. Leong et al. / Applied Catalysis A: General 415–416 (2012) 53–58
8
6
4
2
0
0
0
0
0
1
001 K
1
000
9
55 K
8
37 K
VPO -R24
996 K
VPO -R24
735 K
740 K
7
5
50
9
40 K
8
8
43 K
1
005 K
00 VPO -R15
VPO -R15
VPO -R8
9
54 K
250
0
47 K
7
44 K
VPO -R8
500
600
700
800
900
1000
1100
5
00
700
900
1100
Temperature / K
Temperature / K
Fig. 4. TPR in H2 profiles for VPOs catalysts.
Fig. 3. TPD of O2 profiles for VPOs catalysts.
desorbed per unit area of VPOs-R15 and VPOs-R24 was found the
same, it would suggest that the surface morphology of these cat-
alysts is about the same and approximately the same amount of
Table 3
Total amount of oxygen atoms desorbed and values of desorption activation energies
obtained by temperature programmed desorption for the VPOs catalysts.
14
oxygen atoms is desorbing from both of them. There are ∼7 × 10
Catalyst
Tm (K)
Desorption
activation
energy, Ed
Total amount
of oxygen
Total amount
of oxygen
oxygen atoms on a (VO)2P2O7 surface. Therefore, between 30
and 40% of the surface oxygen atoms were desorbed (i.e. sub-
desorbed
desorbed
−1
−1
−1
(kJ mol
)
(mol g
)
(atom g
)
monolayer) and a specific V O or V
being desorbed.
O P oxygen appears to be
6.99 × 10−
5
4
4
4.21 × 10
8.22 × 10
9.29 × 10
19
19
19
VPOs-R8
VPOs-R15
VPOs-R24
954
940
955
263.4
259.6
263.7
−
1.37 × 10
1.54 × 10−
3
.5. Temperature-programmed reduction (TPR) in H /N
2 2
desorbed oxygen and the derived desorption activation energies
are summarized in Table 3.
TPR in H2 analyses was used to investigate the redox properties
of the catalysts. Additional information as to the nature and the
oxidizing species available from the catalysts could be obtained.
Fig. 4 shows the TPR profiles for VPOs catalysts in H /N stream
VPOs-R8 gave a broad peak with a peak maximum at 954 K.
However, increasing the precursor reflux duration to 15 h gave a
peak maximum, which occurred at a lower temperature, i.e. 940 K.
VPOs-R24 gave the most intense peak comparatively with a peak
maximum at 955 K. The total amount of oxygen desorbed (Table 3)
from the catalysts were found to be directly proportional to the pre-
cursor reflux duration. The total amount of oxygen desorbed from
2
2
3
−1
(5% H2 in N , 1 bar, 25 cm min ) using fresh catalysts and rais-
2
−
1
ing the temperature from ambient to ∼1173 K at 5 K min in that
stream. Table 4 lists the peak maxima temperatures, the amount of
removed oxygen in each peak and the derived reduction activation
energies.
1
9
−1
VPOs-R8 catalyst was 4.21 × 10 atom g . Increasing the precur-
A few important points were observed in the TPR profiles. All
three catalysts gave three peak maxima in the reduction by H2
sor reflux duration to 15 h and 24 h increased the total amount of
19
−1
19
−1
5+
oxygen desorbed to 8.22 × 10 atom g and 9.29 × 10 atom g
,
(Fig. 4). The first two peaks corresponded to the reduction of V
respectively. The particular trend was also observed in the previous
study on different calcination durations [15]. Besides that, pro-
longed reflux durations had also increased the surface area so that
phase, whereas the third peak was assigned to the removal of lat-
4+
tice oxygen from the active V phase [21,22]. The peak attributed to
4
+
−
2−
V
V
was associated with the removal of O anion and the peak from
2
5+
the amount of oxygen atom desorbed per unit area, i.e. atom/cm
was related to the oxygen species of O [21,23]. The reduction
1
4
14
14
4+
as ∼2.2 × 10 , ∼3 × 10 and ∼3 × 10 for VPOs-R8, VPOs-R15
of V species appeared to be the major peak for all three cata-
lysts. This trend was similarly observed in previous report, whereby
and VPOs-R24, respectively. Since the amount of oxygen atoms
Table 4
Total amount of oxygen atoms removed, values of reduction activation energies and ratio for oxygen removal of V5+/V4+ and V4+/V5+ by reduction in H2/N2 for the VPOs
catalysts.
Catalyst
Peak
Tm (K)
Reduction activation
Amount of oxygen
Amount of oxygen
removed (atom g
Ratio for oxygen
Ratio for oxygen
−
1
−1
−1
)
removal of V /V4+
5+
removal of V /V5+
4+
energy, Er (kJ mol
)
removed (mol g
)
−
−
−
−
4
4
3
3
20
VPOs-R8
1
2
3
Total
744
847
1005
124.4
141.6
168.1
3.53 × 10
4.21 × 10
3.38 × 10
4.16 × 10
2.13 × 10
20
2.53 × 10
21
2.04 × 10
21
2.50 × 10
0.23
0.22
0.20
4.37
4.63
5.07
−
−
−
−
4
4
3
3
20
VPOs-R15
VPOs-R24
1
2
3
Total
740
843
996
123.7
141.0
166.6
2.19 × 10
2.39 × 10
2.12 × 10
2.58 × 10
1.32 × 10
20
1.44 × 10
21
1.28 × 10
21
1.55 × 10
−
−
−
−
4
4
3
3
19
1
2
3
Total
735
837
1001
122.9
140.0
167.4
1.45 × 10
2.48 × 10
1.99 × 10
2.39 × 10
8.71 × 10
20
1.49 × 10
21
1.20 × 10
21
1.44 × 10

L.K. Leong et al. / Applied Catalysis A: General 415–416 (2012) 53–58
57
Table 5
Catalysts performance of VPOs catalysts.
Catalyst
n-Butane conversion (%)
Product selectivity (%)
MA
CO
CO2
VPOs-R8
VPOs-R15
VPOs-R24
20
31
37
67
64
60
1
1
1
32
35
39
catalysts were also produced employing the sesquihydrate precur-
sor route [15]. Three peak maxima were observed for VPOs-R8 in
the rate of hydrogen consumption at 744, 847 and 1005 K with
20
the amount of oxygen removed from each peak were 2.13 × 10
,
Fig. 6. Correlation of n-butane conversion with the amount of oxygen species
removed associated with V /V phase.
4
+
5+
2
0
21
−1
2
.53 × 10 and 2.04 × 10 atom g , respectively. The ratio of oxy-
5+ 4+
gen atom removed from V /V was about 0.23 while the ratio of
oxygen atom removed from V4+/V was about 4.37.
5+
2−
the oxygen species assigned to O was responsible for selectiv-
ity towards MA [24,25]. Also, a plot of n-butane conversion versus
the amount of oxygen species removed associated with V4+/V
phase (Fig. 6) gave an increasing trend. The results obtained were
VPOs-R15 also gave three peak maxima, which appeared at
lower temperatures, i.e. 740, 843 and 996 K. However, the total
5+
5
+
4+
amount of oxygen species associated with V and V phases
20
−1
21
−1
−
had decreased to 2.76 × 10 atom g
and 1.28 × 10 atom g
,
in agreement with the fact that the oxygen species (O ), which was
respectively. Hence, the ratio of oxygenatomremovedfromV /V4+
was slightly declined to 0.22, but an increasing trend was observed
for V4+/V , i.e. 4.63.
5+
4+
associated with V phase was responsible for n-butane activation
of the catalysts [21–23,25]. These correlations were also found to
be in agreement with catalysts produced using the hemihydrate
precursor, employing isobutanol as the reducing agent in previous
findings [22,26,27].
5+
However, VPOs-R24 gave three smaller peaks, which appeared
at 735, 837 and 1001 K. The amount of oxygen species linked to
5
+
20
−1
V
phase was further decreased to 2.36 × 10 atom g , whereas
4+
the amount of oxygen removed from V phase was decreased
4. Conclusions
2
1
5+ 4+
to 1.20 × 10 . The ratio of oxygen atom removed from V /V
was decreased further to 0.20, while the ratio of oxygen atom
1
. Lower precursor reflux duration during the preparation of
sesquihydrate precursor induced higher selectivity but lower
activity on the catalyst as compared to the prolonged reflux
counterparts.
removed from V4 /V increased to 5.07. Higher oxygen species
+
5+
associated with V5+/V4+ and V /V phases removal suggested that
the catalyst tend to show a higher selectivity and higher activity,
respectively, which will be discussed in the following section.
4+
5+
2
. All catalysts exhibited good crystalline with characteristic peaks
of vanadyl pyrophosphate phase and their surface morphologies
were found to be in rosette-shape.
3.6. Catalytic oxidation of n-butane to maleic anhydride
3
. The consumption of H2 in TPR confirmed that more lattice oxy-
gen can be removed from the catalyst which has the shortest
precursor reflux duration and exhibited a higher ratio for oxygen
The catalytic performances of the catalysts for n-butane oxida-
tion to maleic anhydride (MA) have been tested at 673 K, a typical
operating temperature for VPO catalysts. The details of the catalytic
performance data of all the catalysts are shown in Table 5. VPOs-R8
was assigned as the most selective catalyst with 67% MA selectiv-
ity, followed by VPOs-R15 and VPOs-R24 with 64% and 60% MA
selectivity, respectively. However, an opposite trend was observed
for conversion of n-butane, whereby VPOs-R24 was deemed as
the most active catalyst with 37% conversion, followed by VPOs-
R15 with 31% conversion and VPOs-R8 with 20% conversion. This
could be well-supported by the results discussed in Section 3.2 indi-
cating that higher specific surface area could further enhance the
activity of the catalysts [19]. Good relationship between selectivity
towards MA and the amount of oxygen species removed associated
5+ 4+
removal from V /V phase.
4
. The catalytic tests showed that the oxygen species associated
5+ 4+
with V /V phase played an important role in the formation of
4+ 5+
maleic anhydride, whereas the oxygen species linked to V /V
phase was needed for the conversion of n-butane.
Acknowledgment
Financial support from UTAR Research Fund (for K.S. Chin) is
gratefully acknowledged.
References
5
+
4+
with V /V phase (Fig. 5) was observed. This result agreed that
[
1] G. Centi, F. Cavani, F. Trifirò, Selective Oxidation by Heterogeneous Catalysis,
Kluwer Academic/Plenum Publishers, New York, 2001, pp. 141–201.
2] T.R. Felthouse, J.C. Burnett, S.F. Mitchell, M.J. Mummey, Encyclopedia of Chem-
ical Technology, fourth ed., John Wiley & Sons Inc., New York, 1995, p. 893.
3] D.X. Wang, M.C. Kung, H.H. Kung, Catal. Lett. 65 (2000) 9–17.
4] G. Centi, Catal. Today 16 (1993) 147–153.
[
[
[
[
[
[
[
[
5] G. Centi, Catal. Today 16 (1993) 5–26.
6] G.J. Hutchings, R. Higgins, J. Catal. 162 (1996) 153–168.
7] B.T. Pierini, E.A. Lombardo, Catal. Today 107–108 (2005) 323–329.
8] B.K. Hodnett, Ph. Permanne, B. Delmon, Appl. Catal. 6 (1983) 231–244.
9] B.K. Hodnett, B. Delmon, Appl. Catal. 9 (1984) 203–211.
[
[
[
[
[
[
10] K.C. Waugh, Y.H. Taufiq-Yap, Catal. Today 81 (2003) 215–225.
11] G.J. Hutchings, Appl. Catal. 72 (1991) 1–32.
12] B.K. Hodnett, Catal. Rev. Sci. Eng. 27 (1985) 373–424.
13] T. Ishimura, S. Sugiyama, H. Hayashi, J. Mol. Catal. A: Chem. 158 (2000) 559–565.
14] I. Matsuura, T. Ishimura, N. Kimura, Chem. Lett. 24 (1995) 769–770.
15] Y.H. Taufiq-Yap, L.K. Leong, M.Z. Hussein, R. Irmawati, S.B. Abd Hamid, Catal.
Today 93–95 (2004) 715–722.
Fig. 5. Correlation of selectivity to maleic anhydride with the amount of oxygen
species removed associated with V /V phase.
5+
4+
[16] M. Niwa, Y. Murakami, J. Catal. 76 (1982) 9–16.

5
8
L.K. Leong et al. / Applied Catalysis A: General 415–416 (2012) 53–58
[
[
[
[
[
17] P.H. Klug, E. Alexander, X-ray Diffraction Procedures for Polycrystalline and
Amorphous Materials, second ed., John Wiley & Sons, New York, 1974.
18] C.J. Kiely, S. Sajip, I.J. Ellison, M.T. Sananes, G.J. Hutchings, J.C. Volta, Catal. Lett.
[22] Y.H. Taufiq-Yap, C.K. Goh, G.J. Hutchings, N. Dummer, J.K. Bartley, J. Mol. Catal.
A: Chem. 260 (2006) 24–31.
[23] J.C. Volta, C.R. Acad. Sci. Paris Série IIc Chim./Chem. 3 (2000) 717–723.
[24] G.W. Coulston, S.R. Bare, H. Kung, K. Birkeland, G. Bethke, R. Harlow, N. Herron,
P.L. Lee, Science 275 (1997) 191–193.
[25] J.-M. Herrmann, P. Vernoux, K.E. Béré, M. Abon, J. Catal. 167 (1997) 106–117.
[26] C.K. Goh, Y.H. Taufiq-Yap, G.J. Hutchings, N. Dummer, J. Bartley, Catal. Today
131 (2008) 408–412.
3
3 (1995) 357–368.
19] Y.H. Taufiq-Yap, M.H. Looi, K.C. Waugh, M.Z. Hussein, Z. Zainal, R. Samsuddin,
Catal. Lett. 74 (1–2) (2001) 99–104.
20] Y.H. Taufiq-Yap, M.H. Looi, K.C. Waugh, M.Z. Hussein, Z. Zainal, Asian J. Chem.
1
4 (2002) 1494–1502.
21] M. Abon, J.M. Herrmann, J.C. Volta, Catal. Today 71 (2001) 121–128.
[27] Y.H. Taufiq-Yap, J. Nat. Gas Chem. 15 (2006) 144–148.