The unexpected role of aldehydes and ketones in the standard preparation method for vanadium phosphate catalysts

Article abstract of DOI:10.1006/jcat.2000.3006

Since vanadium phosphate compounds are widely used as heterogeneous catalysts for the production of maleic anhydride for n-butane partial oxidation, their preparation has been extensively studied. The potential role played by aldehyde or ketone by-products in the synthesis of such catalysts was investigated. VO(H₂PO₄)₂ was formed as the exclusive product from the reaction of aldehydes or ketones (C₄-C₁₀) with V₂O₅ and H₃PO₄ whether aqueous (85%) or crystalline (100%) orthophosphoric acid was used. This observation was significant for the synthesis of the commercially important hemihydrate VOHPO₄·0.5H₂O precursor, which is typically synthesized from V₂O₅ and H₃PO₄ using an alcohol as the reducing agent, leading to the formation of aldehydes/ketones as by-products. The use of isobutanol containing a small amount of butanone for the reaction of V₂O₅ and H₃PO₄ (100%) proved that low levels of VO(H₂PO₄)₂ could be formed along with VOHPO₄·0.5H₂O.

Full text of DOI:10.1006/jcat.2000.3006

Journal of Catalysis 195, 423–427 (2000)
doi:10.1006/jcat.2000.3006, available online at http://www.idealibrary.com on
RESEARCH NOTE
The Unexpected Role of Aldehydes and Ketones in the Standard
Preparation Method for Vanadium Phosphate Catalysts
Jonathan K. Bartley, Richard P. K. Wells, and Graham J. Hutchings1
Department of Chemistry, Cardiff University, Cardiff CF10 3TB, United Kingdom
Received May 18, 2000; revised July 6, 2000; accepted July 9, 2000
This method superseded the use of aqueous HCl or HCl
dissolved in alcohols as reducing agents (2), and this
method can now be considered as the standard industrial
preparation method. Most researchers agree that it is
essential that the desired precursor VOHPO4 0.5H2O is
produced in as pure a form as possible. In particular, it
is known that the material VO(H2PO4)2 can be deleterious
when present as an impurity in the catalyst precursor since
it lowers the surface area of the activated catalyst and
consequently these catalysts display poorer catalyst per-
formance. While the use of alcohols as reducing agents has
been extensively studied (1, 5, 6), to date no attention has
been given to the potential role played by the aldehyde or
ketone by-products. We have now addressed this aspect of
catalyst synthesis and have found that, surprisingly, the use
of aldehydes and ketones leads to the exclusive synthesis
of VO(H2PO4)2 which has previously been shown by itself
to exhibit poor catalytic performance (7, 8), and when
present with VOHPO4 0.5H2O can significantly impair (9)
the function of the catalyst for the oxidation of n-butane.
VO(H2PO4)2 is formed as the exclusive product from the reac-
tion of aldehydes or ketones with V2O5 and H3PO4 whether aqueous
(85%) or crystalline (100%) orthophosphoric acid is used. This ex-
clusive product formation has been observed with a broad range
of aldehydes and ketones (C4–C10). This finding casts doubt on the
current commercial preparation of vanadium phosphate catalysts
used for the oxidation of n-butane to maleic anhydride. This catalyst
is derived from a crystalline precursor VOHPO4 0.5H2O, formed
from the reaction of V2O5 and H3PO4 with an alcohol. The alco-
hol acts as a reducing agent forming an aldehyde or ketone. These
aldehydes and ketones, once formed, will lead to the formation of
VO(H2PO4)2 as an impurity, at levels which will be difficult to de-
tect but which are known to affect catalytic performance adversely.
The use of isobutanol containing a small quantity of butanone for
the reaction of V2O5 and H3PO4 (100%) confirmed that low lev-
els of VO(H2PO4)2 can be formed together with VOHPO4 0.5H2O.
c
2000 Academic Press
Key Words: vanadium phosphate catalyst preparation;
VO(H2PO4)2; butane oxidation to maleic anhydride.
INTRODUCTION
METHODS
Vanadium phosphate compounds are used extensively
as heterogeneous catalysts for the production of maleic
anhydride for n-butane partial oxidation. As a result,
vanadium phosphates and their preparation have been
extensively studied (1, 2). The preferred industrial cata-
lyst is synthesised from VOHPO4 0.5H2O to form an
active catalyst comprising (VO)2P2O7 with II-, -, and
-VOPO4 by in situ activation in n-butane/air (3). A major
breakthrough in the synthesis of VOHPO4 0.5H2O was
pioneered by Johnson et al. (4) and involves the reaction
of V2O5 with H3PO4 under reflux with an alcohol as a
reducing agent. When the alcohol is used as a reducing
agent it is oxidised to form an aldehyde or ketone (4).
V2O5 (1.0 g, Strem) and H3PO4 (2.8 g, 85% Aldrich) were
refluxed with an aldehyde/ketone (40 ml, Aldrich) for 16 h.
The resultant pale blue solid was recovered by vacuum fil-
tration, washed with the aldehyde/ketone (50 ml) and ace-
tone (50 ml), and dried in air (110 C, 24 h). Additional ex-
periments were conducted using crystalline H3PO4. V2O5
(1.0 g, Strem) and H3PO4 (2.4 g, 100% Fischer) refluxed
with isobutanol, isobutanol/butanone, and butanone (40ml,
Aldrich) for 16 h. The resultant pale blue solid was recov-
ered by vacuum filtration, washed with the alcohol/ketone
(50 ml), and dried in air (110 C, 24 h).
In a similar set of experiments VOPO4 2H2O (1.0 g)
prepared by standard methods (6) was reacted with
H3PO4 (0.059 g, 85% Aldrich) under reflux with the
aldehyde/ketone (40 ml, Aldrich) for 16 h. The product
1 To whom correspondence should be addressed. E-mail: hutch@
cardiff.ac.uk.
423
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

424
BARTLEY, WELLS, AND HUTCHINGS
was recovered by filtration and washed with the alde- H3PO4 during the alcoholic reduction at levels which are
hyde/ketone (50 ml) and acetone (50 ml) and dried in air fully miscible with the reaction mixtures. We consider that
(110 C, 24 h).
this method, therefore, represents a novel and general syn-
Powder X-ray diffraction was performed using an Enraf thetic route for the synthesis of high purity VO(H2PO4)2.
Nonius FR590 X-ray generator with a CuK source fitted The catalyst performance of VO(H2PO4)2 prepared in this
with an Inel CPS 120 hemispherical detector. Laser Raman way for the oxidation of n-butane was investigated and the
spectra were obtained using a Renishaw Ramanscope spec- results are shown in Table 1. It is clear that the activated
trograph fitted with a green Ar+ laser ( = 514.532 nm). material is not particularly active and poor selectivity to
Thermalgravimetric analysis was performed using a Perkin maleic anhydride is observed.
Elmer TGA 7 instrument.
In a further set of experiments, the reaction of V2O5
Vanadium phosphates were tested for the oxidation of with H3PO4 (100% ) in isobutanol and butanone mixtures
n-butane with air in a standard laboratory microreactor. (2 and 10 mol% ) was investigated. The powder X-ray
Butane and air were fed to the reactor using calibrated diffraction pattern of the resulting materials show that
mass flow controllers and the mixture (1.5 vol% butane VOHPO4 0.5H2O is produced together with VO(H2PO4)2
in air) was passed over the heated catalyst (0.5 g). The exit asa minor phase (Fig. 2a). Thermalgravimetricanalysiscon-
lines from the reactor were heated and the products were firmed this finding (Fig. 2b). Pure VO(H2PO4)2 shows a
analysed using on-line gas chromatography.
weight loss at ca. 400–420 C whereas VOHPO4 0.5H2O
transforms to (VO)2P2O7 at ca. 480 C (2). It is clear that
the pure materials show only a single weight loss whereas
the materials formed from the reaction with isobutanol/
butanone mixtures exhibit two distinct weight losses. These
results show that small amounts of the ketone can lead to
the formation of VO(H2PO4)2 as a detectable impurity.
In previous studies (5, 6), we have shown that VOPO4
2H2O can be reduced with alcohols to form VOHPO4
0.5H2O and this method also leads to the formation of
an aldehyde/ketone. Hence, in a subsequent set of experi-
ments, we investigated the potential role of aldehydes and
ketones in this synthetic method. In this case, for all simple
aldehydes and ketones studied no reaction was observed
with VOPO4 2H2O. This indicates that this preparation
method may be the preferred route for pure samples of
the hemihydrate precursor. However, addition of a small
amount of H3PO4 (molar ratio: VOPO4 2H2O :H3PO4 =
1 : 0.1) led to the synthesis of VO(H2PO4)2 for all aldehy-
des/ketones investigated and the X-ray diffraction patterns
were similar to those shown in Fig. 1a. This observation
leads us to suggest that the aldehyde/ketone acts as a reduc-
ing agent via its enol tautomer, the concentration of which
is promoted by the addition of a Brønsted acid. Some ke-
tones, notably -diketones, exist mainly in the enol form.
RESULTS AND DISCUSSION
In the first set of experiments the reaction of V2O5 and
H3PO4 (85% ) wasinvestigated usingaldehydesand ketones
in place of an alcohol. In this way the standard method of
preparation has been modified to use an aldehyde and ke-
tone. Since aldehydes and ketones are not expected to act
as a reducing agent most researchers would consider that
no reaction would be anticipated. However, a rapid reac-
tion was observed and the exclusive product was observed
to be VO(H2PO4)2 from the full range of C4–C10 aldehy-
des and ketones investigated. The powder X-ray diffrac-
tion patterns and laser Raman spectra for typical samples
of VO(H2PO4)2 are shown in Fig. 1. These diffraction and
spectral data are in full agreement with those published pre-
viously for VO(H2PO4)2 (10, 11). A subsequent set of ex-
periments was carried out using crystalline H3PO4 (100% ),
the starting material that is often used in commercial cata-
lyst preparation, using butanone. Again, VO(H2PO4)2 was
formed as the only isolated reaction product (Fig. 2a) in-
dicating that the water (added in the form of 85% H3PO4)
is not the reason for the formation of VO(H2PO4)2. In-
deed water is formed during the reaction between V2O5 and
TABLE 1
Oxidation of n-Butane over VO(H2PO4)2
Selectivity (% )
Surface area final
n-Butane
Specific activity
(10 5 mol m 2 h
)
1
1
Aldehyde/ketone
catalyst (m2 g
)
conversion (% )
MA
CO
CO2
Octanal
1.3
1.9
0.9
1.7
1.0
21.1
21.6
16.4
28.5
8.5
25
25
25
22
28
45
47
43
53
37
30
28
32
25
35
2.67
1.92
2.98
2.45
1.60
2-Octanone
3-Octanone
Butanone
2-Hexanone
1
Note. Reaction conditions, 385 C, 1.5% n-butane in air, GHSV = 1000 h
.

PREPARATION OF VANADIUM PHOSPHATE CATALYSTS
425
FIG. 1. (a) Powder X-ray patterns for VO(H2PO4)2 prepared from the reaction of V2O5 (85% ) with representative aldehydes/ketones as reducing
agents. The patterns are indexed fro VO(H2PO4)2. (b) The Raman spectra of VO(H2PO4)2 prepared using aldehydes/ketones as reducing agents
(literature values (11): 1151 cm 1, br, m; 935 cm 1; vs; 575 cm 1; m).
For example, 2,4-pentadione isstabilised in the enolform by To demonstrate that ketones/aldehydes act as reducing
the conjugated enone and formation of an intramolecular agents in their enol form in the formation of VO(H2PO4)2,
hydrogen bond:
VOPO4 2H2O was refluxed with 2,4-pentadione for 16 h
in the absence of H3PO4 and VO(H2PO4)2 was formed as
the exclusive product.
The observation that VO(H2PO4)2 can be readily formed
from V2O5 and H3PO4 (85% or 100% ) in a one-step syn-
thesis using aldehydes and ketones is of significance for
the synthesis of the commercially important hemihydrate

426
BARTLEY, WELLS, AND HUTCHINGS
FIG. 2. (a) Powder X-ray diffraction patterns of the material from the reaction of V2O5 with H3PO4 (100% ) using isobutanol/butanone mixtures:
(i) 100% isobutanol, (ii) 98% isobutanol and 2% butanone, (iii) 90% isobutanol and 10% butanone, and (iv) 100% butanone. Key: ᭹ reflections assigned
to VOHPO40.5H2O; ᭜ reflections assigned to VO(H2PO4)2. (b) Thermalgravimetric analysis of the material from the reaction of V2O5 with H3PO4
(100% ) using isobutanol/butanone mixtures: (i) 100% isobutanol, (ii) 98% isobutanol and 2% butanone, (iii) 90% isobutanol and 10% butanone, and
(iv) 100% butanone.
VOHPO4 0.5H2O precursor. The hemihydrate is typically
synthesised from V2O5 and H3PO4 using an alcohol as
the reducing agent (1, 5, 6) which leads to the formation
of aldehydes/ketones as by-products (4). Invariably, this
will lead potentially to the formation of VO(H2PO4)2 as
a by-product, albeit in trace amounts. Indeed, we have
demonstrated this experimentally.

PREPARATION OF VANADIUM PHOSPHATE CATALYSTS
427
VO(H2PO4)2 exhibits poor performance for the oxida- the results of this study confirm that careful consideration
tion of n-butane (7, 8) and additionally we have previ- should be given to the role of reaction by-products in the
ously shown (5, 9) that trace impurities of VO(H2PO4)2 preparation of heterogeneous catalysts.
in VOHPO4 0.5H2O leads to impaired catalytic perfor-
mance on activation. The trace amounts formed by the
action of the aldehyde/ketone impurities in the standard
VOHPO4 0.5H2O preparation would not be expected to
be readily observed by diffraction or other characterisa-
tion methods, and so their formation may have been over-
looked to date in the standard preparation method. How-
ever, if present, VO(H2PO4)2 will lead to a lower surface
area in the activated catalyst and hence lead to poorer
catalyst performance. Fortunately, VO(H2PO4)2 is readily
soluble in hot water, whereas VOHPO4 0.5H2O is insol-
uble, and hence VO(H2PO4)2 can be easily removed by a
simple water extraction procedure (9). Alternatively, re-
searchers should consider the use of the method based
on the reduction of VOPO4 2H2O by alcohols since, in
this case, the aldehyde/ketone by-product is inactive for
the formation of VO(H2PO4)2. This observation may ex-
plain the previous finding that, although catalysts pre-
pared by the reaction of V2O5/H3PO4 or VOPO4 2H2O
with alcohols have the same specific activity, catalysts pre-
pared by the VOPO4 2H2O route can give much higher
surface areas (typically >40 m2 g 1) when compared with
the V2O5/H3PO4 route (typically 30 m2 g 1). In any event,
ACKNOWLEDGMENT
We thank the EPSRC for financial support.
REFERENCES
1. Centi, G. (Ed.), Catal. Today 16 (1994).
2. Hutchings, G. J., Appl. Catal. 72, 1 (1992).
3. Hutchings, G. J., Desmartin Chomel, A., Olier, R., and Volta, J. C.,
Nature 368, 41 (1994).
4. Johnson, J. W., Johmston, D. C., Jacobsen, A. J., and Brody, J. F., J. Am.
Chem. Soc. 106, 8123 (1984).
5. Ellison, I. J., Hutchings, G. J., Sananes, M. T., and Volta, J. C., J. Chem.
Soc., Chem. Commun. 1093 (1994).
6. Sananes, M. T., Ellison, I. J., Sajip, S., Burrows, A., Keily, C. J., Volta,
J. C., and Hutchings, G. J., J. Chem. Soc., Faraday Trans. 92, 137
(1996).
7. Mount, R. A., and Raffleson, J. O., Ger. Offen. 2727617 (1977), as-
signed to Monsanto.
8. Hannur, F. K., Martin, A., Kubias, B., Locke, B., Bordes, E., and
Courtine, P., Catal. Today 40, 263 (1998).
9. Hutchings, G. J., and Higgins, R., U.K. Patent 1601121 (1981), assigned
to Imperial Chemical Industries.
10. Bordes, E., Catal. Today 1, 499 (1987).
11. Guliants, V. V., Benziger, J. B., Sundaresan, S., Wachs, I. E., Jehng,
J. M., and Roberts, J. E., Catal. Today 28, 275 (1996).