Preparation of vanadium phosphate catalysts from VOPO4·2H2O: Effect of VOPO4·2H2O preparation on catalyst performance

Article abstract of DOI:10.1016/j.molcata.2004.02.027

The morphology and structure of vanadium phosphate catalysts prepared from VOPO₄·2H₂O obtained from different preparation methods was studied. The manner in which the initial VOPO₄·2H₂O was prepared could have s

Full text of DOI:10.1016/j.molcata.2004.02.027

Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
Preparation of vanadium phosphate catalysts from VOPO4·2H2O: effect
of VOPO4·2H2O preparation on catalyst performance
Louisa Griesela, Jonathan K. Bartleya, Richard P.K. Wellsb, Graham J. Hutchingsa,∗
a
Department of Chemistry, Cardiff University, PO Box 912, Cardiff, Wales CF10 3TB, UK
b
Department of Chemistry, College of Physical Sciences, University of Aberdeen, Aberdeen AB24 3UE, UK
Received 30 October 2003; accepted 17 February 2004
Available online 20 July 2004
Abstract
The preparation of VOPO4·2H2O is described and discussed. Three samples of the dihydrate are prepared with different ageing times
following the initial reflux of V2O5 with H3PO4 in water for 24 h. The materials were characterised using a combination of powder XRD,
BET surface area measurement, laser Raman spectroscopy and scanning electron microscopy. A sample of VOPO4·2H2O was isolated by
immediate filtration of the reaction mixture and comprised flat oval crystallites with a broad size distribution between ca. 2 and 20 ␮m
in diameter. Materials isolated following ageing of the initial reaction mixture (20 ◦C, 24 h) comprise square platelets again with a very
broad size distribution. Using pyrophosphoric acid as the phosphorus source in place of phosphoric acid also affected the morphology of the
VOPO4·2H2O. The dihydrates were reacted with isobutanol to form VOHPO4·0.5H2O and these were transformed to (VO)2P2O7 by reaction
with 1.7% n-butane in air at 400 ◦C. The most active catalyst was derived from VOPO4·2H2O prepared from ageing a reaction mixture
following the removal of the first crop of crystals. The study shows that the method of preparation of VOPO4·2H2O and, in particular, its
morphology is of importance in the preparation of vanadium phosphate catalysts using the two stage method based on the reaction of the
dihydrate with an alcohol to form the hemihydrate precursor.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Vanadium phosphate; Catalysts; Catalyst performance
1. Introduction
In view of the importance of the morphology of the cata-
lyst precursor, there have been numerous studies concerned
with catalyst preparation. In general, V2O5 is used as a
source of vanadium, and H3PO4 is used as a source of phos-
phorus. Hence, a reducing agent is required to synthesise the
V4+ precursor phase and a broad range of reducing agents
and solvents have been employed [6–22]. Initial catalyst
preparations [5,22] used water as solvent but most studies, in
recent years, have concentrated on the use of alcohols as they
can exhibit the duel role of solvent and reducing agent. In
previous studies [14,23,24] we have shown that very active
catalysts can be prepared using a two stage method based
on VOPO4·2H2O. In particular, we have found that the al-
cohol used in the second step to reduce the VOPO4·2H2O to
VOHPO4·0.5H2O can control the morphology and the best
results are obtained using primary alcohols. In this paper,
we extend this earlier work and present a detailed study of
the morphology and structure of vanadium phosphate cata-
lysts prepared from VOPO4·2H2O obtained from different
preparation methods.
The preparation of selective oxidation of n-butane contin-
ues to receive considerable research attention, and in particu-
lar the preparation of vanadium phosphate catalysts has been
well studied [1–5]. The catalytic performance of vanadium
phosphates depends on the method of preparation of the cat-
alyst precursor, VOHPO4·0.5H2O [6–22], and the reaction
conditions utilised for the in situ activation in n-butane/air
to form the final catalyst [6,7]. The active catalyst comprises
(VO)2P2O7 in combination with some V5+ phosphates, typ-
ically ␣II- and ␦-VOPO4, and the transformation of the pre-
cursor to the final catalyst is topotactic. Hence, the precursor
morphology is of importance in determining the eventual
catalyst morphology and the performance following activa-
tion.
∗
Corresponding author.
E-mail address: hutch@cardiff.ac.uk (G.J. Hutchings).
1381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2004.02.027

114
L. Griesel et al. / Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
2. Experimental
2.1. Catalyst preparation
treated as the previous sample. This material was denoted
o-VPDi-B. A third sample of VOPO4·2H2O was prepared
as follows. V2O5 (5.0 g, Strem) and H3PO4 (30 ml, 85%,
Aldrich) were refluxed in water (120 ml) for 24 h. The
mixture was cooled to ambient temperature (20 ◦C) and
was allowed to stand for 24 h. The crystals formed were
recovered by vacuum filtration, washed and treated as the
previous samples. This material was denoted o-VPDi-C.
Vanadium phosphate catalyst hemihydrate precursors
were prepared using the dihydrates as follows. The dihy-
drate (4 g) was refluxed with isobutanol (80 ml) for 21 h,
and the resulting hemihydrate was recovered by filtration,
dried in air (110 ◦C, 16 h). The hemihydrates were also re-
fluxed in water (9 ml H2O/g solid) for 2 h, filtered hot, and
dried in air (110 ◦C, 16 h).
VOPO4·2H2O was prepared as follows. V2O5 (5.0 g,
Strem) and H3PO4 (30 ml, 85%, Aldrich) were refluxed in
water (120 ml) for 24 h. The yellow solid was recovered
immediately without cooling by vacuum filtration, washed
with hot water (100 ml) and acetone (100 ml) and dried in air
(110 ◦C, 24 h). The material was then left in air at ambient
temperature for 24 h to ensure full hydration. This material
was denoted o-VPDi-A. The filtrate recovered from this
method was allowed to stand at ambient temperature for
24 h during which time further yellow crystals were formed,
these were recovered by vacuum filtration and washed and
Fig. 1. Powder X-ray diffraction patterns of VOPO4·2H2O: (a) o-VPDi-A, (b) o-VPDi-B, (c) o-VPDi-C.
a
b
c
40
50
60
700
800
900
1000
1100
1200
Raman shift (cm-1)
Fig. 2. Raman spectra of VOPO4·2H2O: (a) o-VPDi-A, (b) o-VPDi-B, (c) o-VPDi-C.

L. Griesel et al. / Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
115
2.2. Catalyst testing and characterisation
increases in the order o-VPDi-A > o-VPDi-B > O-VPDi-C.
This demonstrates that the morphology changes with in-
creased ageing time of the dihydrate. The Raman spec-
tra of the three materials (Fig. 2) are also consistent with
VOPO4·2H2O as reported by Ben Abdelouahab et al. [13].
The main band at 953 cm−1 is due to the symmetric stretch
The oxidation of n-butane was carried out in a microreac-
tor with a standard mass of catalyst (0.5 g). n-Butane and air
were fed to the reactor via calibrated mass flow controllers
to give a feedstock composition of 1.7% n-butane in air. The
products were then fed via heated lines to an on-line gas
chromatograph for analysis. The reactor comprised a stain-
less 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 1 ◦C.
Carbon mass balances of ≥97% were typically observed.
Catalyst samples were heated in situ (1.7% n-butane in air)
at 400 ◦C by heating the sample from room temperature at
a rate of 3 ◦C/min.
3−
of P–O in the PO4 tetrahedra and the bands at 987 cm−1
and 1038 cm−1 are due to the V–O and V–O–P stretching
modes, respectively.
The morphology of the three different dihydrate samples
was examined using scanning electron microscopy (Fig. 3).
o-VPDi-A comprises flat oval crystallites with a broad size
distribution between ca. 2 and 20 ␮m in diameter. Both
o-VPDi-B and o-VPDi-C comprise square platelets again
A number of techniques were used to characterise the
catalyst structure. Powder X-ray diffraction (XRD) was per-
formed using an Enraf Nonius FRS 590 X-ray generator with
a Cu K␣ source fitted with an Inel CPS 120 hemispherical
detector. BET surface area measurements using nitrogen ad-
sorption were carried out using a Micromeritics ASAP 2000
instrument. Raman spectra were obtained using a Renishaw
Ramanscope Spectrograph fitted with a green Ar+ laser (␭
= 514.532 nm). Scanning electron microscopy (SEM) was
performed on a Hitachi 326YO-N instrument operating at
20 kV.
3. Results and discussion
3.1. Catalyst precursor preparation and characterisation
o-VPD1-A, o-VPDi-B and O-VPDi-C were prepared as
described and the yields based on the starting V2O5 are
shown in Table 1. Overall the yields, as expected, are the
same since the combined yield of o-VPD1-A (48%) and
o-VPDi-B (24%) is equal to O-VPDi-C (72%). The pow-
der X-ray diffraction patterns correspond to VOPO4·2H2O,
with the dominant reflection at 11.8◦ indexed to the [0 0 1]
plane (Fig. 1). The intensity of the [0 0 1] reflection relative
to the other reflections (notably [0 0 2], [2 0 0] and [1 0 3])
Table 1
Yields from preparation methods
Preparation method
Yielda (%)
o-VPDi-A
o-VPDi-B
o-VPDi-C
p-VPDi-A
p-VPDi-B
VPD-A
VPD-Aw
VPD-B
VPD-C
48
24
72
37
22
41
28
21
64
56
VPD-Cw
Fig. 3. Scanning electron micrographs of VOPO4·2H2O: (a) o-VPDi-A,
a
(b) o-VPDi-B, (c) o-VPDi-C.
Yield based on V2O5.

116
L. Griesel et al. / Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
Fig. 4. Powder X-ray diffraction patterns of VOPO4·2H2O: (a) p-VPDi-A, (b) p-VPDi-B.
with very broad distribution, but o-VPDi-C has much larger
crystallites. The morphologies are consistent with the pow-
der X-ray diffraction data.
VPD-Cw. No changes were observed in the powder X-ray
diffraction pattern or the Raman spectra. The morphology
of the hemihydrates as examined by scanning electron mi-
croscopy also did not reveal any changes to these materials
due to the water reflux. The filtrates were evaporated to dry-
ness and in many cases this gave a crystalline material which
To determine if the phosphorus source was important two
further dihydrate samples were prepared using pyrophospho-
ric acid in place of phosphoric acid using the methodology
used to prepare o-VPDi-A and o-VPDi-B. We have previ-
ously shown that pyrophosphoric acid can be used to pre-
pare high activity catalysts [25]. These materials are denoted
p-VPDi-A and p-VPDi-B. The powder X-ray diffraction pat-
terns (Fig. 4) are similar to those of o-VPDi-B (Fig. 1). The
morphology (Fig. 5) shows that the material recovered im-
mediately following the reflux period (p-VPDi-A) comprises
irregular thin platelets, and consequently are very different
from the material prepared in the same way using phosphoric
acid. The material recovered from the filtrate after standing
for 24 h (p-VPDi-B) comprises mainly flat square platelets
and is very similar to the morphology of o-VPDi-B. These
findings show that the source of phosphorus may be impor-
tant with respect to the morphology of the dihydrate if it is
recovered rapidly after synthesis.
The three dihydrate samples prepared from phosphoric
acid were then reacted with isobutanol to form hemihy-
drates denoted VPD-A, VPD-B and VPD-C and the yields
are given in Table 1. The powder X-ray diffraction patterns
(Fig. 6) can be indexed to VOHPO4·0.5H2O and the [0 0 1]
reflection is most pronounced in VPD-A. The Raman spectra
were all affected by fluorescence, but the main bands at 985,
1007, 1108, 1154 cm−1 of the hemihydrate are observed.
The morphology of the materials was examined using scan-
ning electron microscopy (Fig. 7) and all comprise rosettes
of very thin platelets, with VPD-A having the smaller and
more perfectly formed rosettes.
VPD-A and VPD-C were refluxed for 2 h and recov-
ered by filtration to give materials denoted VPD-Aw and
Fig. 5. Scanning electron micrographs of VOPO4·2H2O: (a) p-VPDi-A,
(b) p-VPDi-B.

L. Griesel et al. / Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
117
[211]
[121
]
[101]
[200]
[040]
[001]
[201]
a
b
c
10
15
20
25
30
35
2 degrees
40
45
50
55
60
Fig. 6. Powder X-ray diffraction patterns of VOHPO4·0.5H2O: (a) VPD-A, (b) VPD-B, (c) VPD-C.
was characterised as ␣I-VOPO4 by powder X-ray diffraction
and Raman spectroscopy. Sometimes a material amorphous
to X-rays was also recovered, but by Raman spectroscopy
this was also determined to be ␣I-VOPO4.
3.2. Catalyst testing and characterisation
The five hemihydrate precursors were treated in situ in the
laboratory microreactor with 1.7% n-butane in air at 400 ◦C.
Fig. 7. Scanning electron micrographs of VOHPO4·0.5H2O: (a) VPD-A, (b) VPD-B, (c) VPD-C.

118
L. Griesel et al. / Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
Table 2
Catalyst performance of vanadium phosphate for the oxidation of n-butanea
Catalyst precursor
Surface area
n-Butane
conversion (%)
Product selectivity (%)
Intrinsic activityb
(10−5 mol MA/m2/h)
Specific activityc
(10−4 mol MA/g/h)
MA
CO
CO2
VPD-A
VPD-Aw
VPD-B
VPD-C
VPD-Cw
17d
21d
17d
13d
7d
13f
–
49
71
71
36
37
12
0
85
72
72
69
63
71
–
8
15
15
17
23
17
–
7
13
13
14
14
13
–
4.43
3.89
5.04
4.44
3.20
2.53g
–
7.52
8.16
8.57
5.77
4.79
3.29
–
e
␣I-VOPO4
Amorphouse
a
Reaction conditions: 400 ◦C, 1.7% n-butane in air, GHSV = 2500 h−1
Intrinsic activity: mol maleic anhydride formed/m2 catalyst/h.
Specific activity: mol maleic anhydride formed/g catalyst/h.
Surface area of the catalyst following activation.
Materials derived from the filtrate after the water reflux of VPD-A and VPD-C.
Surface area of the material prior to catalyst testing.
.
b
c
d
e
f
g
Intrinsic activity based on surface area of the material prior to catalyst testing.
During this time the catalyst performance for the formation
of maleic anhydride steadily improved. The catalyst perfor-
mance data, when steady state had been obtained, are shown
in Table 2 together with the surface areas of the catalysts.
The most active catalysts for maleic anhydride synthesis
were VPD-B and VPD-Aw which gave almost identical per-
formance with surface areas of 17 and 21 m2 g−1. The water
which was prepared from dihydrate that was aged for 24 h
before isolation. For VPD-Aw the relative intensity of the
[2 0 0] reflection relative to the [0 4 2] reflection is signifi-
cantly enhanced, whereas in VPD-Cw the material is much
less crystalline with respect to X-ray diffraction and, al-
though the relative intensities of the [2 0 0] and [0 4 2] re-
flections are similar to VPD-Aw, the reflections are all very
broad. This demonstrates that the morphology of the initial
dihydrate can have significant effects on the subsequent mor-
phology of the steady state activated catalyst, and in particu-
lar that water washing of precursors may not always be ben-
eficial. The morphology effects are to be expected since the
transformations of vanadium phosphates are topotactic and
hence the initial morphology will have major consequences.
The water washing effects require further comment, since for
materials prepared in the one step aqueous or organic solvent
routes the water refluxing of the precursor removes water
soluble VO(H2PO4)2 and this leads to a significant enhance-
reflux step increased the surface area of VPD-A by 3 m2 g−1
,
leading to an increase in conversion and intrinsic activity.
Surprisingly the surface area of VPD-C decreases after the
treatment of water and the maleic anhydride selectivity is
slightly decreased for VPD-Cw.
The powder X-ray diffraction patterns of the catalysts are
shown in Fig. 8 and these can all be indexed to vanadyl py-
rophosphate, (VO)2P2O7, and the Raman spectra were con-
sistent with this. Interestingly, the water washing treatment
has distinctly different effects on VPD-A, prepared from the
dihydrate recovered immediately after synthesis, and VPD-C
[042]
[200]
[202]
[063]
a
b
c
d
e
10
20
30
40
50
degrees
60
70
80
2
Fig. 8. Powder X-ray diffraction patterns of catalysts derived from precursors: (a) VPD-A, (b) VPD-B, (c) VPD-C, (d) VPD-Aw, (e) VPD-Cw.

L. Griesel et al. / Journal of Molecular Catalysis A: Chemical 220 (2004) 113–119
119
[4] G. Centi, F. Trifiro, G. Busca, J. Ebner, J. Gleaves, Faraday Disc.
87 (1989) 215.
[5] G.J. Hutchings, Appl. Catal. 72 (1991) 1.
[6] G.J. Hutchings, A. Desmartin Chomel, R. Olier, J.C. Volta, Nature
368 (1994) 41.
[7] C.J. Kiely, A. Burrows, G.J. Hutchings, K.E. Bere, J.C. Volta, A.
Tuel, M. Abon, J. Chem. Soc., Faraday Disc. 105 (1996) 103.
[8] J.W. Johnson, D.C. Johnston, A.J. Jacobson, J.F. Brody, J. Am.
Chem. Soc. 106 (1984) 8123.
[9] H.S. Horowitz, C.M. Blackstone, A.W. Sleight, G. Teufer, Appl.
Catal. 38 (1988) 211.
[10] E.W. Arnold, S. Sundaresan, Appl. Catal. 41 (1988) 457.
[11] K. Ait-Lachgar, M. Abon, J.C. Volta, J. Catal. 171 (1991) 383.
[12] E.A. Lombardo, C.A. Sanchez, L.M. Conaglia, Catal. Today 15
(1992) 407.
[13] F. Ben Abdelouahab, J.C. Volta, R. Olier, J. Catal. 148 (1994) 334.
[14] I.J. Ellison, G.J. Hutchings, M.T. Sananes, J.C. Volta, J. Chem. Soc.,
Chem. Commun. (1994) 1093.
[15] V.V. Guliants, J.B. Benziger, S. Sundaresan, Chem. Mater. 7 (1995)
1485.
ment in catalyst activity [5]. In the preparation of the hemi-
hydrate using the one step procedures the initial P:V molar
ratio is >1. The high temperature water washing step there-
fore removes the excess phosphorus from the precursors. In
the two step method based on the dihydrate the P:V molar ra-
tio in the step that forms the precursor hemihydrate is unity.
Consequently there is no excess phosphorus that requires re-
moval. The high temperature water washing step in this case
can significantly affect the crystallinity of the material and
consequently the procedure has to be carried out with care.
The two materials recovered from the filtrate following
the water washing procedure were also examined as catalysts
(Table 2). The amorphous material was found to be inac-
tive whereas the more crystalline material did display a low
activity. After activation this material comprised ␣I-VOPO4
and ␣II-VOPO4.
In conclusion this study has shown that the manner in
which the initial VOPO4·2H2O is prepared can have signif-
icant effects on its morphology. This subsequently affects
the catalysts which are derived from the VOHPO4·0.5H2O
formed from the reaction of the dihydrate with isobutanol.
[16] V.V. Guliants, J.B. Benziger, S. Sundaresan, N. Yao, I.E. Wachs,
Catal. Lett. 32 (1995) 379.
[17] V.A. Zazhigalov, J. Haber, J. Storch, L.V. Bogutskaya, I.V.
Bacherikova, Appl. Catal. 135 (1996) 155.
[18] C.J. Kiely, A. Burrows, S. Sajip, G.J. Hutchings, M.T. Sananes, A.
Tuel, J.C. Volta, J. Catal. 162 (1996) 31.
[19] V.V. Guliants, J.B. Benziger, S. Sundaresan, I.E. Wachs, J.-M. Jehng,
J.E. Roberts, Catal. Today 28 (1996) 275.
[20] J. Haber, V.A. Zazhigalov, J. Storch, L.V. Bogutskaya, I.V.
Bacherikova, Catal. Today 33 (1997) 39.
Acknowledgements
[21] W.H. Cheng, W. Wang, Appl. Catal. A 156 (1997) 57.
[22] G.J. Hutchings, R. Higgins, J. Catal. 162 (1996) 153.
[23] G.J. Hutchings, M.T. Sananes, S. Sajip, C.J. Kiely, A. Burrows, I.J.
Ellison, J.C. Volta, Catal. Today 33 (1997) 161.
[24] F.J. Cabello Sanchez, J.A. Lopez Sanchez, R.P.K. Wells, C. Rhodes,
A. Isfahani, G.J. Hutchings, Catal. Lett. 77 (2001) 189.
[25] J.A. Lopez Sanchez, L. Griesel, Bartley, R.P.K. Wells, A. Liskowski,
D. Su, R. Schlögl, J.C. Volta, G.J. Hutchings, Phys. Chem. Chem.
Phys. 5 (2003) 3525.
We thank Sasol and the EPSRC for financial support.
References
[1] G. Centi, Catal. Today 16 (1994) 1.
[2] E. Bordes, Catal. Today 1 (1987) 499.
[3] J.T. Gleaves, J.R. Ebner, T.C. Knechler, Catal. Rev. Sci. Eng. 30
(1988) 49.