Você está na página 1de 11

View Article Online / Journal Homepage / Table of Contents for this issue

F E AT U R E A R T I C L E

Journal of

Department of Chemistry, Cardiff University, P.O. Box 912, Cardiff, UK CF10 3TB.
E-mail: hutch@cardiff.ac.uk

Materials
Chemistry

Graham J. Hutchings

www.rsc.org/materials

Vanadium phosphate: a new look at the active components of


catalysts for the oxidation of butane to maleic anhydride

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

Received 26th March 2004, Accepted 30th July 2004


First published as an Advance Article on the web 27th September 2004

Vanadium phosphate catalysts have been extensively


studied since the 1960s, when it was discovered that they
are effective catalysts for the oxidation of butane to
maleic anhydride. They represent the sole example of a
commercialised material for the catalytic oxidation of an
alkane. To date several hundred papers have been
devoted to their study. Yet, virtually all studies have
concentrated on two specific phases, namely the vanadyl
hemihydrate precursor, VOHPO4?0.5H2O, which is
converted into vanadyl pyrophosphate, (VO)2P2O7,
which most researchers consider to be the active
component of vanadium phosphate catalysts. In this
feature article the nature of the active catalyst
components is revisited and new approaches being
pioneered in the synthesis of vanadium phosphates are
described. In particular, detailed comments are made
concerning the method of preparation and the role of
crystalline and amorphous vanadium phosphate catalysts.

Introduction

DOI: 10.1039/b404610m

Heterogeneous catalysis is a research field in which materials


science can be expected to make a major impact. The
preparation of active heterogeneous catalysts often involves
fairly complex methods1 that have to be carefully controlled
and which often result in the formation of mixtures of phases
from which it is difficult to determine the nature of the active
components. Significant progress has been achieved in the
study of microporous and mesoporous materials.2 Techniques
Graham is a graduate of University College London (1972) and
he spent his early career in ICI with positions in research and
production in both the UK and South Africa. It was in ICI that
he was introduced to the topic of vanadium phosphate catalysts
and he has worked in this area since 1975, making early
discoveries on the role of impurity phases and the recognition of
Mo as a catalyst promoter. In 1984, he left industry to take up
an academic career and has held
three chairs of chemistry, sequentially at the Universities of the
Witwatersrand, Liverpool and
Cardiff. Since 1997 he has been
Head of the Chemistry Department at Cardiff University. His
current research interests include
oxidation catalysis, enantioselective catalysis and catalysis by
gold. He has published over 400
papers on the topic of heterogeneous catalysis, including 47
papers and 8 patents on vanadium
Professor Graham
phosphate catalysts.
Hutchings

that are normally associated with the study of the bulk of a


solid rather than the surface, e.g. powder X-ray diffraction,
solid state NMR spectroscopy and EXAFS have been used to
great effect with these microporous and mesoporous materials
since, in many cases, all the atoms are surface atoms or, at the
very least, the ratio of surface to bulk atoms is enhanced. In
many cases the nature of the active sites has been determined.3
The situation is less advanced in the study of the majority of
oxides and metal-supported oxides that are used as industrial
catalysts, and to a large extent these have received much less
research attention from materials scientists. In writing this
feature article it was decided to select a catalyst of industrial
importance that has been the subject of several hundred papers
and patents and yet there is still considerable controversy
concerning the nature of the active catalyst components. The
article will therefore deal with a topic that is ideal for renewed
detailed studies by materials scientists in combination with
theoreticians, catalysis chemists and engineers, and in so doing
these new studies can be expected to make a breakthrough of
significance in the field of alkane activation. This article will
deal with vanadium phosphate catalysts for the selective
oxidation of butane to maleic anhydride. Vanadium phosphates comprise a large class of compounds including: aI-, aII-,
b-, c-, d-and V-VOPO4, VOPO4?2H2O, VOHPO4?0.5H2O,
VO(H2PO4)2 and (VO)2P2O7. Vanadium phosphate catalysts
represent perhaps the most well studied heterogeneous catalyst
since their discovery as an effective catalyst in 1966,4 and the
first academic paper on the topic.5 Worldwide, they are used
commercially for the production of maleic anhydride and have
been extensively studied,69 yet questions remain as to how they
work, the optimal method of their preparation and, perhaps
most significantly, the nature of the active vanadium phosphate
phase. The purpose of this feature article is to introduce this
fascinating topic and to introduce new ideas concerning the
method of preparation, the role of promoters and the relative
role of crystalline versus amorphous vanadium phosphates as
active catalysts.
Preparation of vanadium phosphate catalysts
In the early patent literature on this topic a large range of
preparation methods were evaluated, but, in general, they all
prepared the same vanadium phosphate.4,10,11 Until the 1980s
the structure of this material was unknown and it was referred
to simply as phase A. This material is now known to be the low
temperature precursor vanadyl hemihydrate, VOHPO4?
0.5H2O.12 Heat treatment of this precursor formed a new
phase that was associated with active catalysts, denoted phase
B in the early patent literature, and now known to be vanadyl
pyrophosphate, (VO)2P2O7. This heat treatment process,
during which the precursor is transformed to the active
catalyst is usually carried out in situ in the reactor. Virtually
all industrial processes and academic studies for the oxidation

This journal is The Royal Society of Chemistry 2004

J. Mater. Chem., 2004, 14, 33853395

3385

View Article Online

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

of butane to maleic anhydride use butaneoxygen mixtures


below the lower explosion limit, denoted fuel-lean conditions,
and so this pre-treatment often involves heating the hemihydrate in 1.5% butane in air. However, the transformation of
the hemihydrate to the pyrophosphate can also be achieved by
heat treatment in an inert atmosphere (N2, He) and so can be
achieved prior to use in an industrial reactor. To date, many
researchers consider vanadyl pyrophosphate to be the active
phase, and there is no doubt that this is a component of active
catalysts for butane oxidation. However, as will be discussed in
this paper the structure of active catalysts is considerably more
complex than that presented by one precursor and one active
vanadium phosphate phase.
Under fuel-lean conditions (1.5% butane in air), the catalysts
have to be operated at high conversion to achieve a significant
yield of maleic anhydride. At any temperature, selectivity is a
function of conversion (Fig. 1) and selectivity can be significantly increased if the catalysts can be operated at a lower
temperature whilst maintaining high conversion (Fig. 2). In
view of this, preparation methods that produce high-activity
catalysts are favoured, since these permit access to lower
operational temperatures, thereby giving improved selectivity.
Although the earlier industrial research concerning vanadium phosphates explored a large number of preparation
strategies for the synthesis of VOHPO4?0.5H2O, most recent
literature has concentrated on the reaction between V2O5,
H3PO4 and an alcohol and a related reaction of VOPO4?2H2O
with an alcohol. This is largely due to the seminal work of
Horowitz et al.13 and Johnson et al.12 Horowitz et al.13
explored the preparation and, in particular, the effect of the
reactant P : V molar ratio and the nature of the alcohol. They
showed that the preparation of the precursor was significantly
affected by the nature of the alcohol, e.g. refluxing
VOPO4?2H2O in isobutanol, a primary alcohol, gave a rosette
structure, whereas with sec-butanol a platelet morphology was
obtained. The subsequent transformation to (VO)2P2O7 was
also studied and they concluded that the morphology of the
precursor was a controlling factor in the catalytic properties.
Johnson et al.12 also showed the importance of the alcohol in
controlling the morphology of the hemihydrate precursor and
the pyrophosphate generated on activation.
The mechanism by which the morphology of the
VOPO4?0.5H2O rosettes are formed is only now being
addressed. Recently, a tentative crystallisation mechanism of
VOHPO4?0.5H2O in organic solvents was proposed by
OMahony et al.14 They proposed that the initial precursor
was VOPO4?2H2O which is rapidly formed when H3PO4 is
reacted with V2O5. The dihydrate subsequently transforms to
monohydrate which acts as a nucleation centre for the
crystallisation of VOHPO4?0.5H2O, which grows epitaxially
into rosette structures. Hodnett and co-workers15 have now
made detailed studies of the crystallisation process by using
in situ energy dispersion X-ray diffraction and significant

Fig. 1 Typical plot for selectivity to maleic anhydride as a function of


conversion at constant temperature for a standard VPO catalyst.
3386

J. Mater. Chem., 2004, 14, 33853395

Fig. 2 Selectivity as a function of temperature required for 90%


conversion for a standard VPO catalyst. Key: & VPA; r VPO; .
promoted catalysts.

insights into this stage of catalyst preparation can now be


expected.
Based on the seminal studies of Horowitz et al.13 and
Johnson et al.,12 we subsequently investigated and contrasted
three preparation methodologies outlined in Scheme 1.16
The VPA method was used in early patent literature6 and
uses water as the solvent. In this method V2O5 is refluxed with
hydrochloric acid and in this step V5+ is reduced to V4+. H3PO4
is then added to the solution (P : V molar ratio 1.0) and
following a further reflux and evaporation a blue green
precursor is obtained that comprises mainly the hemihydrate
VOHPO4?0.5H2O. However, significant amounts of an impurity VO(H2PO4)2 are also obtained. Using an organic solvent,
typically an alcohol, in place of water together with anhydrous
HCl significantly decreases the amount of impurity formation
and produces a more active final catalyst.11 An interesting
observation is that during the solvent removal stage the
hemihydrate is produced steadily as a precipitate, whereas in
aqueous HCl no precipitate forms and the solid is only
obtained on complete solvent removal. This is due to the
relative solubility of VOHPO4?0.5H2O in the respective
reaction media. The VO(H2PO4)2 impurity transforms to
VO(PO3)2 and amorphous vanadium phosphates on heat
treatment which are associated with low activity catalysts.17
However, VO(H2PO4)2 is soluble in water, whereas VOHPO4?0.5H2O is insoluble, and so can be readily removed from
catalyst precursors by extraction with hot water and this results
in a much more active catalyst.18 Water extraction of catalyst
precursors is now a common preparation procedure.
The VPO method is considered to be the standard
preparation method and is used in most academic studies.12,13
It is a variant of the VPA method and is based on the
observation that an alcohol can act as a reducing agent and
consequently the HCl is superfluous. In this method V2O5 is
refluxed with H3PO4 (P : V molar ratio 1.0) with an alcohol
(alcohol : V molar ratio 50) and a blue precursor is obtained
as a precipitate that comprises almost exclusively the hemihydrate VOHPO4?0.5H2O. Many alcohols have been tried but

Scheme 1

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

isobutanol is one of those used. In this reaction the alcohol


reduces V5+ to V4+ and is oxidised. However, the alcohol
oxidation product is not a simple aldehyde or ketone since the
replacement of the alcohol by an aldehyde or ketone leads to
the exclusive formation of VO(H2PO4)2,17,19 and this is not
observed.
The VPD method was first disclosed by Horowitz et al.13 and
further described by Johnson et al.,12 and has subsequently
been investigated in detail.20,21 This method is based on the
observation that the reaction of V2O5 with H3PO4 with water
as solvent, i.e. in the absence of the alcohol used in the VPO
method, leads to the formation of the V5+ dihydrate phase
VOPO4?2H2O. The dihydrate is recovered and dried and then
refluxed in a second step with an alcohol to form the
hemihydrate. The structure of the alcohol determines the
morphology of the hemihydrate precursor, and primary
alcohols produce rosette clusters of thin hemihydrate platelets,
whereas secondary alcohols produce lower surface area thicker
platelets. With primary alcohols, materials with surface areas in
excess of 40 m2 g21 can be prepared.21
All three preparation methods can be used to produce
relatively pure samples of VOHPO4?0.5H2O although there are
differences in their surface areas and morphologies. Subsequent heat treatment of the three precursors in butaneair
mixtures give active catalysts for the oxidation of butane to
maleic anhydride for which there is a linear relationship
between specific n-butane conversion (mol butane converted
per g per h) with the catalyst surface area (Fig. 3). This implies
that the surface structure of the activated catalysts are very
similar and the activity differences are just due to the higher
surface area VPD and VPO catalyst having a higher number of
active sites per unit mass of catalyst. It is, therefore, very
surprising that the bulk structure of the three activated
catalysts are very different (Fig. 4) as determined using
powder X-ray diffraction and 31P NMR spectroscopy by
spin echo mapping.16 The VPA activated material comprises
mainly VOPO4 phases, the VPO activated material is mainly
amorphous to X-rays but by the NMR method is found to
comprise a mixture of VOPO4 and (VO)2P2O7 together with
non-crystalline vanadium phosphates, whereas the VPD
activated material is principally highly crystalline (VO)2P2O7.
It is therefore apparent that, for the activity to be solely a
function of the surface area, the surfaces exposed on these
different materials must all be the same even though their bulk
structures are completely different.
These observations lead to the first key point with respect to
vanadium phosphate catalysts, namely that the surface area is
the factor that controls the activity of catalysts prepared using
VOHPO4?0.5H2O as the precursor. Commercial catalysts
typically have surface areas in the range 2030 m2 g21.
Consequently, experimental strategies are typically aimed at
finding routes to higher area vanadium phosphates. A

Fig. 3 Relationship between specific butane conversion (mol butane


converted per g per h) with catalyst surface area for VPA, VPO and
VPD.

particular advantage of higher area catalysts is that they can


be operated at lower temperatures which leads to an enhanced
selectivity and hence yield in maleic anhydride being obtained
(Fig. 2). A number of exciting new approaches have been
investigated and these are described in the next section.
Preparation of high area vanadium phosphate materials
The data presented in Fig. 3 for the VPA, VPO and VPD
catalysts indicate that it is important to investigate preparation
methods that generate high area materials. One relatively
straightforward approach, first utilised almost thirty years ago,
is to ball mill the precursor prior to heat treatment.22,23 More
recently, this approach has been used by Haber and coworkers.24 This tribomechanical method can readily produce
materials with surface area in excess of 40 m2 g21 and these
materials can be used at temperatures as low as 340 uC giving a
significant enhancement in the yield of maleic anhydride.22
Schlogl and co-workers25 have also studied this method for the
preparation of a Bi-promoted vanadyl phosphate. They found
that ball milling in air for relatively short times could generate a
material that was amorphous in powder X-ray diffraction.
They further noted that ball milling decreases the yield and
conversion per unit surface area and, consequently, the full
benefits that could be expected from the enhanced surface area
are, unfortunately, not realised.
Another standard approach to enhancing the surface area of
the active component is to increase its dispersion by supporting
it on an inactive material. Until recently, this has not been a
particularly successful method for vanadium phosphates since,
in many earlier studies, this has led to a loss of catalytic
performance.26 However, recently Ledoux and co-workers
have pioneered the preparation of mesoporous carbon and
silicon carbide nanotubes as supports and nanosized reactors.2729 Ledoux et al.30 have studied the use of thermally
conducting supports (SiC and BN) for vanadium phosphate
catalysts. The support material is added during the preparation
step, e.g. in the VPA or VPO methods the support is added
together with H3PO4. These supported vanadium phosphates
have considerably improved catalytic performance at high
temperatures (w425 uC) when compared with non-supported
materials. In this way, they give much higher yields of maleic
anhydride when used for butane oxidation. This demonstrates
that, by careful control of the preparation of supported
vanadium phosphates, together with careful selection of the
support material, high activity supported vanadium phosphates can be prepared.
A further approach in the design of high area vanadium
phosphates is based on the observation that VOPO4?2H2O and
VOHPO4?0.5H2O are layer structures that can be exfoliated by
the intercalation of alcohols and other molecules within their
structure. Kamiya et al.31 have shown that 1-propanol and
1-butanol can be intercalated into crystallites of VOPO4?2H2O
and these subsequently exfoliate when the dihydrate is reduced
to the hemihydrate. The intercalation is carried out at low
temperatures, i.e. lower than the reflux temperature of the
alcohol, and the intercalate is not isolated prior to the
reduction step, which is carried out under reflux in the alcohol.
The alcohol was retained within the exfoliated hemihydrate
and, for example, the chemical formula of the material
obtained by exfoliationreduction in 1-butanol was VO[nC4H9)0.16H0.84]PO4?0.8H2O. Activation of these precursors
gave (VO)2P2O7 with enhanced surface areas (w40 m2 g21)
which gave high selectivities to maleic anhydride from butane
oxidation.
Recently, Dasgupta et al.32 have shown that the hemihydrate
VOHPO4?0.5H2O can be exfoliated using dimethylformamide/
water mixtures. The exfoliated hemihydrate formed a novel
mesostructured layered VPO phase.
Guliants, Cavani and co-workers3338 have used a two-stage
J. Mater. Chem., 2004, 14, 33853395

3387

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

Fig. 4

Characterisation of the activated catalysts from VPA, VPO and VPD.

process in the presence of charged surfactants to generate


vanadium phosphates with surface areas as high as 250 m2 g21.
In this method, VOSO4 and H3PO4 were reacted together in the
presence of cationic surfactants. The initial precipitate was
extracted with toluene and then thermally activated in N2
at 400450 uC. The lamellar to hexagonal, hexagonal to
disordered and cubic to amorphous transformations were
studied using in situ Raman spectroscopy. However, the
mesostructured vanadium phosphates were not thermally
stable under the reaction conditions used for butane oxidation
and relatively low yields of maleic anhydride were obtained.
Carreon and Guliants39,40 have also shown that high surface
area vanadium phosphates can be synthesised by introducing
colloidal polystyrene spheres as templates. Using this methodology, very high surface area forms of (VO)2P2O7 were
produced (4475 m2 g21) with preferential exposure of the
(100) planes, which are proposed to be the active surface planes
of the pyrophosphate.41
Clearfield and co-workers42,43 have also investigated intercalation of VOPO4?2H2O as a starting point for the generation
of high-area vanadium phosphates. Using alkylamines in an
alcohol solvent as the intercalating agent, surface areas as
high as 400 m2 g21 could be obtained with pore diameters of
12.3 nm. These intercalators have yet to be evaluated as
oxidation catalysts.
At present, most approaches for the synthesis of high-area,
high-activity catalysts make use of organic media, since waterbased preparations tend to lead to low-surface area materials.
However, with the modern thrust towards green chemistry, the
use of environmentally benign solvent, such as water, leads us
to ask how high-area materials can be obtained using aqueous
media.
High-activity non-promoted vanadium phosphates prepared using
water as solvent
As noted above, there have been numerous studies concerned
with the preparation of vanadium phosphates. In general, since
V2O5 is used as a source of vanadium and H3PO4 is used as a
3388

J. Mater. Chem., 2004, 14, 33853395

source of phosphorus, a reducing agent is required to synthesise


the V4+ precursor phase and a broad range of reducing agents
and solvents have been employed.1253 When water is used as
a solvent, the final catalyst tends to have low surface area
(4 m2 g21) and, as catalyst activity is directly proportional to
surface area (Fig. 1), these catalysts exhibit poor performance
for butane oxidation. In contrast, catalysts prepared with
organic solvents, e.g. alcohols, tend to exhibit high surface
areas (w20 m2 g21) and these have become preferred for the
preparation of commercial catalysts. Indeed, using recent
methodologies, surface areas in the range 250400 m2 g21 have
been reported. To some extent, the synthesis of vanadium
phosphate catalysts has been complicated by the many studies
using alcohols as both solvents and reducing agents. In view of
this, we studied the use of water as solvent with either V2O4,
i.e. a pre-reduced vanadium source, or H3PO3 as a reducing
agent.54 This approach was useful since it did not add any
further components into the catalyst preparation and only
sources of V and P together with water were therefore present.
The amount of H3PO3 added was sufficient to ensure reduction
of V5+ to V4+ and the additional P required to give a V : P molar
ratio of ca. 1 was added as either H3PO4 or H4P2O7. A series of
catalyst precursors was prepared using H3PO3 as the reducing
agent or V2O4 as the pre-reduced V source and the preparations
were carried out in small autoclaves at 145 uC and crystalline
VOHPO4?0.5H2O started to form after 20 h synthesis time and
was fully formed after 72 h synthesis. Interestingly, no traces of
VO(H2PO4)2, an impurity noted to be readily formed under
traditional aqueous preparation conditions,18 were observed to
be present.
When V2O5 is used as the source of vanadium, it is clear that
reduction of V5+ A V4+ must occur prior to the formation of
the precursor compound VOHPO4?0.5H2O. It is interesting to
consider whether the reduction to V4+ is a required initial step
in the catalyst synthesis, or whether the sequence in which
reduction occurs is not of key importance. To investigate this
aspect of the catalyst preparation, three experiments were
carried out. First, V2O5 was reacted in an autoclave with the
required amount of H3PO3 to reduce fully the V5+ to V4+, in

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

water at 145 uC. A green solid was formed which was isolated
and found by powder X-ray diffraction to be amorphous. The
green solid was then reacted with H3PO4 (V : P molar ratio of
1 : 1), in water at 145 uC. The blue solid isolated from the twostep preparation was confirmed by powder X-ray diffraction to
be VOHPO4?0.5H2O. Second, the same experiment was carried
out, except the V2O5 was first reacted with H3PO4 to form an
unidentified green solid as characterised by powder diffraction
which, on subsequent reaction with H3PO3, formed
VOHPO4?0.5H2O. These experiments indicate that the order
of reagent addition is not important for the formation of
VOHPO4?0.5H2O. In particular, the sequence in which
vanadium phosphates are formed and the reduction of
V5+ A V4+ are also not viewed as being specifically important.
The catalyst precursors prepared from V2O4 and H3PO4 or
V2O5 and H3PO3 were activated in situ with 1.7% n-butane in
air at 400 uC and the plot of the specific n-butane conversion
(mol butane converted per g catalyst per h) based on the mass
of catalyst present (Fig. 5). It is clear that the specific activity is
primarily dependent upon the catalyst surface area. However,
it is possible that some of the materials may have a slightly
higher specific activity than that expected from the correlation
between activity and surface area.
Two examples of the small group that gave significantly
higher specific activity than expected from the surface area
correlation (denoted P9 and P10 in Fig. 5) were characterised in
detail using HREM (Fig. 6). VPOP9, prepared by reaction of
V2O4 and H3PO4, comprised mainly (VO)2P2O7, most particles
examined provided higher resolution images with either perfect
or highly ordered (VO)2P2O7. A typical image of the
microstructure taken perpendicular to the rectangular platelet
is given as Fig. 6a, shown in the [100] projection, with the (012)
and (012) lattice planes of (VO)2P2O7 evident [d = 0.626 nm,
angle between (012) and (012) = 82u]. Some particles show
regions with defect structures which are probably due to
irradiation damage and an example of this is marked by the
arrows in Fig. 6b. This may be indicative of microdomains of
more beam sensitive V5+ planes being present within the
(VO)2P2O7 crystallites. The HREM images from VPOP10,
prepared by reaction of V2O4 and H4P2O7, reveal two distinct
microstructures (Figs. 6c and 6d). Some exhibit rectangular
crystallites which comprise (VO)2P2O7 (Fig. 6c). The second
distinct morphology present in VPOP10 (Fig. 6d) was assigned
to aII-VOPO4 and the figure shows a [001] projection in which
the (110) and (110) lattice planes are imaged [d = 0.424 nm,
angle between (110) and (110) = 90u]. These aII-VOPO4
crystallites were particularly beam sensitive. Hence, it was
concluded that VPOP9 and VPOP10 catalysts comprise differing

Fig. 5 Relationship between specific butane conversion (mol butane


converted per g per h) with catalyst surface area for catalysts prepared
using water as solvent with V2O4 or H3PO3 as reagents. Key: P1, P2 :
V2O5 + H3PO3 + H3PO4; P3, P4, P13 : V2O5 + H3PO3 + H4P2O7; P5,
P7, P9, P11 : V2O4 + H3PO4; P6, P8, P10, P12, P14, P15 : V2O4 +
H4P2O7, VPA1, VPA2, VPO, VPD as described in Scheme 1.

Fig. 6 HREM images for VPOP9 and VPOP10 catalysts. (a) VPOP9:
image from (VO)2P2O7 [100] projection; (b) VPOP9: image of
(VO)2P2O7 [100] projection showing amorphous region indicated by
the arrows; (c) VPOP10: image from (VO)2P2O7 [100] projection; (d)
VPOP10 image from aII-VOPO4 [001] projection.

amounts of V4+ and V5+ phases. Consequently, it is possible


that the slightly enhanced specific activity observed with the
catalyst derived from VPOP10, and similar samples (Fig. 5) may
be due to the combination of a higher proportion of V5+ phases
in the bulk of the catalyst crystallites. This effect has been noted
previously in vanadium phosphate catalysts prepared using
water as solvent.18 In addition, Coulston et al.55 have shown
that the concentration of V(V) sites in vanadium phosphate
catalysts is of importance with respect to the synthesis of maleic
anhydride and this may also be related to this effect.

Transformation of the precursor to the activated catalyst


One of the key steps in the preparation of activated catalysts is
the transformation of the crystalline hemihydrate precursor to
the crystalline activated catalyst comprising (VO)2P2O7 in
combination with some V5+ phosphates, typically aII- and
d-VOPO4, and the transformation of the precursor to the final
catalyst is topotactic.56 Hence, the morphology of the
precursor is of crucial importance in determining the eventual
catalyst morphology and the performance following activation.
During the activation process the butane conversion and
selectivity to maleic anhydride gradually increase until the
steady state activity is obtained, typically this takes 2472 h.
One of the surprising observations of two detailed studies on
this transformation process we have carried out56,57 concerns
the involvement of significant amounts of amorphous vanadium phosphates in this process. Hence the transformation is
not represented by the transformation of one well crystalline
material into another, rather the greater majority of the
vanadium phosphate becomes amorphous on heating in
butaneair mixtures and the crystallisation takes place
relatively slowly. The first of these studies involved the use
of in situ laser Raman spectroscopy for the transformation of
the VPA precursor (Fig. 7). Laser Raman spectra were taken
continuously as the hemihydrate was heated in butaneair.
Initially, the intensity of the Raman bands of the hemihydrate
decreased in intensity. However, coincident with the onset of
maleic anhydride formation there was an almost total loss of
the Raman spectrum. Subsequently the material slowly
crystallised, over a period of ca. 24 h, giving a mixture of
J. Mater. Chem., 2004, 14, 33853395

3389

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

Fig. 7 In situ laser Raman spectra and catalytic performance for butane oxidation to maleic anhydride during the transformation from crystalline
VOHPO4?0.5H2O (prepared by the VPA method) to a largely amorphous material, T denotes the sample temperature in flowing 1.5% butaneair,
CBut denotes the butane conversion and SMA denotes the selectivity to maleic anhydride, the bands in the Raman spectra have been assigned where
possible: H,VOHPO4?0.5H2O; P, (VO)2P2O7; aII, aII-VOPO4; c, c-VOPO4; d, d-VOPO4.

(VO)2P2O7 in combination with some V5+ phosphates, typically


aII- and d-VOPO4. We confirmed the existence of the
amorphous material in a later detailed HREM, 31P NMR
spectroscopy study of the transformation of the VPO
precursor.56 In this case samples were isolated after specific
activation times at 400 uC in flowing butaneair, cooled in
nitrogen and then characterised ex situ (Fig. 8). After activation
for 5 min the hemihydrate crystallite was transformed to a

Fig. 8 Bright-field electron micrographs of platelet morphologies in


the transformation of VOHPO4?0.5H2O (7 m2 g21) in 1.6% butane,
18% oxygen and 80.4% helium at 400 uC. (a) 0.1 h, (10.5 m2 g21,
intrinsic activity 0.43 6 1028mol cm22), (b) 8 h, (7.6 m2 g21, intrinsic
activity 0.99 6 1028mol cm22), (c) 84 h, (14.8 m2 g21, intrinsic activity
1.48 6 1028mol cm22), (d) 132 h, (19.4 m2 g21, intrinsic activity 1.39 6
1028mol cm22).
3390

J. Mater. Chem., 2004, 14, 33853395

mixture comprising mainly amorphous vanadium phosphate,


together with non-transformed hemihydrate and small amounts
of (VO)2P2O7 located at the edge of the crystallites and
d-VOPO4 located in the bulk of the crystallites. The electron
diffraction pattern showed reflections for (VO)2P2O7,
d-VOPO4 and VOHPO4?0.5H2O but it is the detailed microscopy that confirms the role of the amorphous material (Fig. 8).
As the transformation proceeds the (VO)2P2O7 crystallites
grow at the edges at the expense of the amorphous material and
the surface area gradually increases. Initially, the d-VOPO4 also
forms at the expense of the amorphous material. However,
subsequently, in the interior of the crystallites the d-VOPO4
also transforms topotactically to (VO)2P2O7 and at the end of
the process very little d-VOPO4 remains non-transformed.
During these processes the catalyst performance gradually
increases over the 132 h experimental period.56 There are
therefore two topotactic transformations occurring during the
formation of the active catalyst under in situ treatment with
butaneair mixtures:
at the crystallite edges: VOHPO4?0.5H2O A (VO)2P2O7
in the crystallite interior: VOHPO4?0.5H2O A d-VOPO4 A
(VO)2P2O7
The activation is therefore not a simple single process. These
studies emphasised the importance of in situ studies since the
role played by the amorphous vanadium phosphate would not
have been apparent.
Promoted vanadium phosphate catalysts
The activity of vanadium phosphates is often enhanced by the
addition of low concentrations of metal cations known as
promoters. The subject of promotion of vanadium phosphate

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

catalysts, until recently, is almost the exclusive preserve of the


patent literature, and this was reviewed in detail in 1991.58
Since then, there has not been extensive interest in this topic,
yet it is a subject that still requires study in order that the effects
are fully understood. Promoters tend to be added as transition
metal cations in the preparation of the hemihydrate precursor
and positive effects have been claimed for a number of cations,
including Bi, Co, U, Fe, Mo, Nb, Zn, Zr.18,58 However, in the
earlier review58 the effects of the promoters were classified into
two groups. This was based on a simplified kinetic analysis of
the available data. As noted earlier, virtually all catalytic data
for the oxidation of butane to maleic anhydride is obtained
with dilute butaneair mixtures. Under these conditions carbon
oxides and water are the only notable by-products and the
reaction data can be satisfactorily described by simple seriesconsecutive pseudo first order reaction scheme, where k1, k2
and k3 are rate constants.18,58

The first promotion effect, denoted type 1, was considered to


be purely structural in nature and ensured that a high surface
area final catalyst was obtained. In this case the added cation
functions as a phosphorus scavenger ensuring that deleterious
phases, e.g. VO(H2PO4)2, are not formed. In many catalyst
preparation methods excess of the phosphorus compound was
used to ensure that the hemihydrate was formed, but this meant
that VO(H2PO4)2 was also formed which, if it was not removed
by water extraction, resulted in a loss of surface area on
activation. Hence, the intrinsic activity of the catalyst was not
affected by the addition of the promoter and the excess
phosphorus was often associated with phosphates of the added
cation in the activated catalyst.18,58 This effect is best
demonstrated by Fig. 9 which presents data for catalysts
prepared using the VPA method,18 where it is apparent that for
many promoters the effect is solely to enhance the surface area.
In some cases, notably Sb, Ag, Cs, Fe, although higher surface
areas are produced the intrinsic activity is decreased and it is

therefore difficult to imply that these additives act as structural


promoters. However, the effects of various promoters can
depend on the preparation method, and for catalysts prepared
using the VPO method Fe does act as a promoter.59 However,
it is also apparent from Fig. 9 that both Co and Mo enhance
the intrinsic activity for butane activation, and this represents
a true promotion effect. This was classified as type 2 promotion
as in this case the cation was incorporated into the vanadyl
pyrophosphate in solid solutions of the type ((VO)12xMx)2P2O7
where M is a cation that substitutes for VO2+, and the effect
was only observed for relatively low concentrations of the
promoter cation. Although there have been a small number of
type 2 promoter effects observed to date,59 most notably with
Co2+, Fe3+ and Al3+ this is a subject that has yet to be fully
explored. The addition of molybdenum compounds has been
found to enhance the selectivity of maleic anhydride formation
by decreasing the rate of over-oxidation18,60 (i.e. by suppressing
k3 relative to k1 + k2) and this is shown in Fig. 10. To date Mo
appears to be unique in this effect and it is often incorporated
into commercial catalysts.
Recent detailed microscopy studies have shown that the
origin of the type 2 promotion effect is more complex than first
thought, particularly for Co-doped catalysts prepared using
the VPO method.59 Addition of 1 at% Co, based on V, gives
a significant promotion effect both structurally in terms
of enhancement in surface area and electronic in terms of
enhancement in the intrinsic activity. However, addition of
5 atom % Co did not give these effects and the Co in this case
was associated with cobalt phosphate in the activated catalyst.
The cations therefore have limited solubility in vanadium
phosphates. Detailed electron microscopy characterisation of
the 1 atom % Co-doped material showed that the hemihydrate
precursor had the characteristic rhomboidal plate-like morphology (Fig. 11(a)). Chemical analysis using energy dispersive
X-ray analysis indicated that the Co was homogeneously
dispersed throughout the hemihydrate crystal. Activation of
the material was also followed using microscopy and as in the
undoped material (Fig. 8) (VO)2P2O7 nucleated at the periphery of the platelet. Even after extensive activation in dilute
butaneair for ca. 250 h, the activated 1 atom % Co-doped
material remained largely disordered (Fig. 11(b)) with some
(VO)2P2O7 crystallites (30150 nm in size). It is therefore clear
that the presence of the Co enhances the stability of the
amorphous material in the activated catalyst. Chemical

Fig. 9 Dependence on the rate constant for butane conversion (k1 + k2) on the activated catalyst surface area: % non-promoted VPA, # nonpromoted VPA made using isobutanol in place of water, evaporated to dryness, non-promoted VPA made using isobutanol in place of water,
collected by filtration, & promoted catalysts.14
J. Mater. Chem., 2004, 14, 33853395

3391

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

Fig. 12 Schematic of the apparatus for the precipitation of vanadium


phosphates using supercritical or liquid CO2. BPR, back pressure
regulator; PV, precipitation vessel; P, pump.

Fig. 10 Effect of Mo on the oxidation of maleic anhydride relative to


butane, VPA method using isobutanol in place of water, 385 uC. # Mo
added as MoO3, $ Mo added separately prior to activation by
grinding MoO3 with VOHPO4?0.5H2O.14

analysis using energy dispersive X-ray analysis, of the


disordered material and an isolated (VO)2P2O7 crystallite
(Fig. 11(c)) showed that within the limits of detection, the Co
phase separated during activation and remained preferentially
with the amorphous component of the activated catalyst.59 It
appears that there is very limited solubility of Co within
(VO)2P2O7. In this case the origin of the enhancement in
intrinsic activity appears to be associated with the Co and the
amorphous material, and consequently appears to represent a
third type of catalyst promotion.
Preparation of wholly amorphous vanadium phosphates
As noted earlier, a number of studies have implicated
amorphous vanadium phosphates as components in active
catalysts. We have found that rapid precipitation of vanadium
phosphate from an alcohol solution using supercritical CO2 as
an antisolvent provides a preparation route for an amorphous
vanadium phosphate.61,62 In this method a solution of H3PO4
in isopropanol was refluxed with VOCl3 for 16 h to give a blue
solution. The resulting isopropanol solution was processed
using supercritical CO2 to precipitate a vanadium phosphate
using the apparatus shown schematically in Fig. 12 with the
following methodology. The isopropanol solution was pumped
through a fine capillary (220 mm id) into a precipitation vessel
containing concurrently flowing CO2. The CO2 can act as an
effective antisolvent when it is either a liquid (w42 bar at 20 uC)
or a supercritical fluid (T c = 31.3 uC, Pc = 72 bar). A material

denoted VPOSCP1 was prepared using supercritical CO2 as the


antisolvent (PCO2 = 110 bar, 60 uC). A catalyst was also
prepared using the same methodology but using liquid CO2
(PCO2 = 60 bar, 15 uC) and this was denoted VPOLP. In
addition a solid was prepared by allowing the isopropanol
solution to evaporate slowly in a Schlenk line and this is
denoted VPOEP. Electron microscopy and electron diffraction
studies showed that VPOSCP1 comprised discrete amorphous
spheroidal particles ranging from 75 nm to 5 mm in diameter.
These particles showed no diffraction contrast (only thickness
contrast), no lattice fringes or nanocrystalline order. These
three precursors were evaluated as catalysts for the partial
oxidation of butane to maleic anhydride and the results are
shown in Fig. 13 for the first 72 h operation. For comparison,
the results for typical catalysts prepared by the standard VPA,
VPO and VPD procedures are also shown.
It is apparent that none of the three catalysts requires an
activation period to establish the steady state catalyst
performance which is generally associated with vanadium
phosphate catalysts.6 During this activation period, it is usual
that the VPA, VPO and VPD catalysts derived from crystalline
hemihydrate VOHPO4?0.5H2O undergo a structural transformation to (VO)2P2O7 and VOPO4 phases.56,57 The catalysts
derived from VPOSCP1, VPOLP and VPOEP all have low surface
areas of 4, 6 and 8 m2 g21, respectively, when compared with
the standard VPO (14 m2 g21) and VPD (43 m2 g21) materials.
From the data in Fig. 13 it is clear that the catalyst derived
from VPOSCP1 has a significantly higher intrinsic activity for
the production of maleic anhydride. We have repeated this
preparation method many times and the enhanced activity is
always observed. Interestingly, the catalyst derived from the
precursor prepared in liquid CO2 which has an intrinsic activity
higher than the VPA/VPO/VPD catalysts, whereas the catalyst
derived from the sample prepared by evaporation, VPOEP,

Fig. 11 (a) Bright field TEM of 1% Co-doped VOHPO4?0.5H2O, (b) bright field TEM of the activated catalyst, (c) energy dispersive X-ray analysis
of the amorphous platelet interior of the activated catalyst.42
3392

J. Mater. Chem., 2004, 14, 33853395

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online

Fig. 13 The intrinsic activity for maleic anhydride (mol h21 m22) with
time on stream. For VPA/VPO/VPD catalysts GHSV = 1200 h21, for
all other catalysts GHSV = 2400 h21. Key: 6 VPO, # VPA, % VPD,
$ VPOSCP1, & VPOLP, + VPOEP

exhibited a similar intrinsic activity to the VPA/VPO/VPD


catalysts.
Characterisation of the catalysts following activation
showed that the material synthesised using supercritical CO2
as an antisolvent remained wholly amorphous.61,62 However,
the materials derived from liquid CO2, VPOLP, and from
evaporation, VPOEP, crystallised during activation to a complex
mixture comprising (VO)2P2O7 and aII-VOPO4, b-VOPO4,
c-VOPO4 and d-VOPO4.
Using 31P spin echo mapping NMR spectroscopy we have
confirmed62 that both the precursor and catalyst prepared from
VPOSCP1 are distinctly different from the conventional
VOHPO4?0.5H2O derived precursors and activated catalysts.
However, at present, the major problems that require attention
concern the surface area of the amorphous material, the
selectivity to maleic anhydride and the levels of adventitious
promoters that are present due to the preparation method used.
Hence, although this method appears to produce a novel highactivity vanadium phosphate, significant advances are still
required for it to be used as a commercial catalyst.
The role of amorphous vanadium phosphates
A number of previous studies41,50,56,57,6379 have suggested that
amorphous VPO material can play an important role in the
selective oxidation of butane. Many studies have suggested that
an amorphous overlayer on the crystalline VPO subsurface
may be the active surface for this reaction.56,57,64,67,68,7578 This
has also been observed on related vanadium carbide catalysts
for butane oxidation.70 As noted earlier, the transformation
from well crystalline VOHPO4?0.5H2O to the final active
catalysts involves the formation and transformation of
amorphous material. This was first clearly demonstrated in
our in situ laser Raman spectroscopy study of the pre-treatment
step of the catalyst preparation.57
The amorphous overlayer on crystalline vanadium phosphate catalysts is readily seen in many TEM studies, for
example a clear example is shown in Fig. 14, but it is also
clearly visible in the micrographs in Fig. 6a and Fig. 6c.
However, all the evidence presented from detailed transmission
electron microscopy studies concerning the presence or absence
of amorphous overlayers has to be viewed with great care. It is
possible that the observed amorphous overlayer may result
from a combination of one or more of the following processes :
(i) exposure of the crystalline surfaces to the reactor feed
producing amorphous material in situ; (ii) electron beam
damage of the (VO)2P2O7 crystallites (which are known to
completely amorphise after 2030 s under typical electron
beam irradiation conditions in the microscope) and/or (iii)
preferential electron beam sensitivity of a crystalline surface
layer (e.g. VOPO4 which is known to completely amorphise in a
few seconds in the electron beam of the microscopy). Hence, it
is not possible, with transmission electron microscopy alone, to

Fig. 14 TEM micrograph showing the amorphous overlayer


observed commonly on (VO)2P2O7 crystallites activated VPO catalysts82

determine unequivocally whether the presence of a surface


amorphous layer is simply an artefact due to beam damage or
is, indeed, the genuine catalytically active phase. However, in
the recent study in which a wholly amorphous vanadium
phosphate catalyst was prepared using supercritical CO2 as an
antisolvent, which retains its amorphous nature throughout the
reactor studies, presents clear evidence of the potential central
importance of amorphous material. Furthermore, the amorphous material prepared in this way requires no activation
period and is more active, on a surface area basis, than the
crystalline counterparts. In the extreme, it is possible that for
many years the detailed characterisation studies involving
methodology that probes the bulk structure rather than the
surface have been studying an elegant support matrix for the
active vanadium phosphate catalyst.
Additional evidence in support of the proposal that the
surface layer of the crystallites of vanadium phosphate is
significantly different from the bulk is provided by the
following observations : (a) X-ray photoelectron spectroscopy6
consistently shows phosphorus enrichment in the surface layers
(P : V 1.5) indicating that the surface is significantly different
from the bulk structure of the crystallites. Fully crystalline
structures of (VO)2P2O7 and VOPO4 would not be able to
provide this degree of phosphorus enrichment. Furthermore,
the phosphorus enrichment is enhanced during the transformation of the precursor VOHPO4.0.5H2O to the final catalyst; (b)
catalysts prepared from VOHPO4.0.5H2O prepared by the
different VPA, VPO and VPD routes are observed to give very
different relative bulk compositions of (VO)2P2O7 and VOPO4
phases, as determined by microscopy, diffraction and spectroscopy,16 but the catalysts will generally have very similar
intrinsic activities for maleic anhydride production (Figs. 3 and
5). This suggests that these different composition catalysts all
exhibit the same active sites on the surface although the bulk
structures are different; (c) the Co promoter for a vanadium
phosphate catalyst, prepared using the VPO route, was found
to phase segregate to, and stabilise, the amorphous material.
Co was not found to be present (to the analytical detection
limit) in the crystalline (VO)2P2O7, yet the intrinsic activity of
the Co-containing catalyst was enhanced by a factor of 3. This
indicates that Co promotes the catalytic performance of the
amorphous material derived from this preparation method. It
is interesting to note that Co does not have this effect for the
amorphous vanadium phosphate catalysts prepared using
supercritical CO2 as an antisolvent;80 (e) oxidation of
J. Mater. Chem., 2004, 14, 33853395

3393

View Article Online

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

(VO)2P2O7 has been shown to enhance the selectivity to maleic


anhydride and oxidise the surface of the catalyst.71,77 This
oxidation treatment also correlates with an increase in the
depth of the amorphous overlayer as observed on edge-on the
(VO)2P2O7 crystallites viewed using very low illumination
transmission electron microscopy.
However, it must be noted that this, at present, remains a
controversial subject. One of the major problems preventing
progress in this debate, until recently, was that, in all these
attempts to unravel the complexity of the role of the
amorphous material, the catalysts all contained substantial
quantities of crystalline phases in addition to the amorphous
material. Indeed, a number of researchers have proposed that
the amorphous material is detrimental to catalyst performance41,50,79,81 or that the amorphous material is a transient
phase and that, following extensive reaction, only wholly
crystalline VPO material is present. Guliants et al.,41,50,79 in
particular, have stated that the amorphous layer terminating
the (200) planes of (VO)2P2O7 observed in fresh catalysts was
not observed in equilibrated catalysts obtained after many days
of reaction. They conclude, in agreement with Albonetti et al.65
and Ebner and Thompson81 that the best catalytic performance
is obtained for catalysts containing only (VO)2P2O7 with the
highest degree of stacking order. In addition, these researchers
consider that VOPO4 phases are detrimental to the catalyst
performance. Indeed, Volta and co-workers45 have shown that
pure VOPO4 phases are less selective for maleic anhydride than
(VO)2P2O7. However, a number of studies have shown that
a combination of VOPO4 phases, together with (VO)2P2O7
can give enhanced catalyst performance.18,71 There remains,
therefore, a considerable amount of debate concerning the
nature of the active sites in vanadium phosphate catalysts,
the preferred composition of the catalyst and the role of the
amorphous material.

Concluding remarks
Vanadium phosphates present a fascinating topic for study by
material scientists, and many problems remain to be resolved.
In this paper a number of new exciting routes to prepare active
oxidation catalysts have been described, against the background of the significant body of work that has been carried
out on vanadium phosphate catalysts. However, even after all
these investigations the nature of the active surface remains
unclear. In this respect it is possible that the best or optimal
materials are yet to be discovered and a number of preparation
strategies can be envisaged to attempt to prepare improved
catalytic materials. There remains plenty of scope for
innovation in the subject of vanadium phosphates and their
use as oxidation catalysts.

References
1 G. J. Hutchings and J. C. Vedrine, in Basic Principles in Applied
Catalysis, ed. M. Baerns, Springer, Berlin, 2004, p. 215.
2 R. Raja, G. Sankar and J. M. Thomas, Angew. Chem. Int. Ed.,
2000, 39, 2313.
3 J. M. Thomas, G. N. Greaves, G. Sankar, P. A. Wright, J. Chen,
A. J. Dent and L. Marchese, Angew. Chem., Int. Ed. Engl., 1994,
106, 1922.
4 R. L. Bergman and N. W. Frisch, U.S. Pat., 3 293 268, 1966.
5 E. Bordes and P. Courtine, J. Catal., 1979, 57, 236.
6 G. Centi, Catal. Today, 1994, 16.
7 E. Bordes, Catal. Today, 1987, 1, 499.
8 J. T. Gleaves, J. R. Ebner and T. C. Knechler, Catal. Rev. Sci.
Eng., 1988, 30, 49.
9 G. Centi, F. Trifiro`, G. Busca, J. Ebner and J. Gleaves, Faraday
Discuss., 1989, 87, 215.
10 J. P. Harrison, U.S. Pat., 3 985 775, 1976.
11 R. A. Scheider, U.S. Pat., 384 280, 1975, U.S. Pat., 4 043 943,
1977.
3394

J. Mater. Chem., 2004, 14, 33853395

12 J. W. Johnson, D. C. Johnston, A. J. Jacobson and J. F. Brody,


J. Am. Chem. Soc., 1984, 106, 8123.
13 H. S. Horowitz, C. M. Blackstone, A. W. Sleight and G. Teufer,
Appl. Catal., 1988, 38, 211.
14 L. OMahony, J. Henry, D. Sutton, T. Curtin and B. K. Hodnett,
Appl. Catal., A, 2003, 253, 409.
15 L. OMahony, T. Curtin, D. Zemlyanov, M. Mihov and
B. K. Hodnett, J. Catal., 2003, 90, 171.
16 C. J. Kiely, A. Burrows, S. Sajip, G. J. Hutchings, M. T. Sananes,
A. Tuel and J. C. Volta, J. Catal., 1996, 162, 31.
17 J. K. Bartley, C. Rhodes, C. J. Kiely, A. F. Carley and
G. J. Hutchings, Phys. Chem. Chem. Phys., 2000, 2, 4999.
18 G. J. Hutchings and R. Higgins, J. Catal., 1996, 162, 153.
19 J. K. Bartley, R. P. K. Wells and G. J. Hutchings, J. Catal., 2000,
195, 423.
20 I. J. Ellison, G. J. Hutchings, M. T. Sananes and J. C. Volta,
J. Chem. Soc., Chem. Commun., 1994, 1093.
21 M. T. Sananes, I. J. Ellison, S. Sajip, A. Burrow, C. J. Kiely,
J. C. Volta and G. J. Hutchings, J. Chem. Soc., Faraday Trans.,
1996, 92, 137.
22 R. Higgins and G. J. Hutchings, U.S. Pat., 4 317 777, 1982.
23 G. J. Hutchings and R. Higgins, Appl. Catal., A, 1997, 154, 103.
24 V. A. Zazhigalov, J. Haber, J. Stoch, L. V. Bogutskaya and
I. V. Bacherikova, Appl. Catal., A, 1996, 135, 155.
25 I. Ayub, D. S. Su, M. Willinger, A. Kharlamov, L. Ushkalov,
V. A. Zazhigalov, N. Kirillova and R. Schlogl, Phys. Chem. Chem.
Phys., 2003, 5, 970.
26 M. Ruitenbeck, A. J. van Dillen, D. C. Koningsberger and
J. W. Geus, Stud. Surf. Sci. Catal., 1998, 118, 549.
27 J. M. Nhut, L. Pesant, J. P. Tessonnier, G. Wine, J. Guille,
C. Pham-Huu and M. J. Ledoux, Appl. Catal., A, 2003, 254, 345.
28 N. Keller, C. Pham-Huu, G. Ehret, V. Keller and M. J. Ledoux,
Carbon, 2003, 41, 2131.
29 J. M. Nhut, R. Vieira, L. Pesant, J. P. Tessonnier, N. Keller,
G. Ehret, C. Pham-Huu and M. J. Ledoux, Catal. Today, 2002, 76,
11.
30 M. J. Ledoux, B. Heinrich, J. J. Lerou, C. Crouzet, C. Bouchy and
K. Kourtakis, U.S. Pat. 6 660 681, 2003.
31 Y. Kamiya, S. Veki, N. Hiyoshi, N. Yamamoto and T. Okuhara,
Catal. Today, 2003, 78, 281.
32 S. Dasgupta, M. Agarwal and A. Datta, Microporous Mesoporous
Mater., 2004, 67, 229.
33 M. A. Carreon and V. V. Guliants, Microporous Mesoporous
Mater., 2002, 55, 297.
34 M. A. Carreon and V. V. Guliants, Stud. Surf. Sci. Catal., 2002,
141, 301.
35 M. A. Carreon and V. V. Guliants, Stud. Surf. Sci. Catal., 2002,
141, 309.
36 M. A. Carreon and V. V. Guliants, Catal. Today, 2003, 78, 303.
37 M. A. Carreon, V. V. Guliants, M. Olga Guerrero-Perez and
M. Banares, Microporous Mesoporous Mater., 2004, 71, 57.
38 M. A. Carreon, V. V. Guliants, F. Pierelli and F. Cavani, Catal.
Lett., 2004, 92, 11.
39 M. A. Carreon and V. V. Guliants, Chem. Commun., 2001, 1438.
40 M. A. Carreon and V. V. Guliants, Chem. Mater., 2002, 14, 2670.
41 V. V. Guliants, J. B. Benziger, S. Sundaresan, I. E. Wachs,
J.-M. Jehng and J. E. Roberts, Catal. Today, 1996, 28, 275.
42 B. G. Shpeizer, X. Ouyang, J. M. Heising and A. Clearfield, Chem.
Mater., 2001, 13, 2288.
43 A. M. Clearfield, J. M. Heising, B. G. Shpeizer and X. Ouyang,
Int. J. Inorg. Mater., 2001, 3, 215.
44 E. W. Arnold and S. Sundaresan, Appl. Catal., 1988, 41, 457.
45 K. Ait-Lachgar, M. Abon and J. C. Volta, J. Catal., 1991, 171, 383.
46 E. A. Lombardo, C. A. Sanchez and L. M. Conaglia, Catal.
Today, 1992, 15, 407.
47 F. Benabdelouahab, J. C. Volta and R. Olier, J. Catal., 1994, 148,
334.
48 V. V. Guliants, J. B. Benziger and S. Sundaresan, Chem. Mater.,
1995, 7, 1485.
49 V. V. Guliants, J. B. Benziger, S. Sundaresan, I. E. Wachs
and J.-M. Jehng, Chem. Mater., 1995, 7, 1485.
50 V. V. Guliants, J. B. Benziger, S. Sundaresan, N. Yao and
I. E. Wachs, Catal. Lett., 1995, 32, 379.
51 V. A. Zazhigalov, J. Haber, J. Storch, L. V. Bogutskaya
and I. V. Bacherikova, Appl. Catal., 1996, 135, 155.
52 J. Haber, V. A. Zazhigalov, J. Storch, L. V. Bogutskaya and
I. V. Bacherikova, Catal. Today, 1997, 33, 39.
53 W. H. Cheng and W. Wang, Appl. Catal., A, 1997, 156, 57.
54 J. A. Lopez-Sanchez, L. Griesel, J. K. Bartley, R. P. K. Wells,
A. Liskowski, D. Su, R. Schlogl, J.-C. Volta and G. J. Hutchings,
Phys. Chem. Chem. Phys., 2003, 5, 3525.

Published on 27 September 2004. Downloaded by UTSA Libraries on 18/06/2013 11:01:31.

View Article Online


55 G. W. Coulston, S. R. Bare, H. Kung, K. Birkeland, G. K. Bethke,
R. Harlow, N. Herron and P. L. Lee, Science, 1997, 275, 191.
56 C. J. Kiely, A. Burrows, G. J. Hutchings, K. E. Bere, J. C. Volta,
A. Tuel and M. Abon, J. Chem. Soc., Faraday Disc., 1996, 105,
103.
57 G. J. Hutchings, A. Desmartin Chomel, R. Olier and J. C. Volta,
Nature, 1994, 368, 41.
58 G. J. Hutchings, Appl. Catal., 1991, 72, 1.
59 S. Sajip, J. K. Bartley, A. Burrows, M.-T. Sananes Schulz, A. Tuel,
J. C. Volta, C. J. Kiely and G. J. Hutchings, New J. Chem., 2001,
25, 125.
60 G. J. Hutchings, U.S. Pat. 4 147 661, 1979.
61 G. J. Hutchings, J. K. Bartley, J. M. Webster, J. A. LopezSanchez, D. J. Gilbert, C. J. Kiely, A. F. Carley, S. M. Howdle,
S. Sajip, S. Caldarelli, C. Rhodes, J. C. Volta and M. Poliakoff,
J. Catal., 2001, 197, 232.
62 G. J. Hutchings, J. A. Lopez-Sanchez, J. K. Bartley, J. M. Webster,
A. Burrows, C. J. Kiely, A. F. Carley, C. Rhodes, M. Havecker,
A. Knop-Gericke, R. W. Mayer, R. Schlogl, J. C. Volta and
M. Poliakoff, J. Catal., 2002, 208, 197.
63 M. R. Thompson, A. C. Hess, J. B. Nicholas, J. C. White,
J. Anchell and J. R. Ebner, Stud. Surf. Sci. Catal, 1994, 82, 167.
64 H. Berndt, K. Buker, A. Martin, A. Bruckner and B. Lucke,
J. Chem. Soc., Faraday Trans., 1995, 91, 725.
65 S. Albonetti, F. Cavani, F. Trifiro, P. Venturoli, G. Calestani,
M. L. Granados and J. L. G. Fierro, J. Catal., 1996, 160, 52.
66 S. Zeyss, G. Wendt, K. H. Hallmeier, R. Szargan and G. Lippold,
J. Chem. Soc., Faraday Trans., 1996, 92, 3273.
67 A. Bruckner, A. Martin, N. Steinfeldt, G. U. Wold and B. Lucke,
J. Chem. Soc., Faraday Trans., 1996, 92, 4257.

68 A. Bruckner, B. Kubias and B. Lucke, Catal. Today., 1996, 32, 215.


69 W. H. Cheng and W. Wang, Appl. Catal., A, 1997, 156, 57.
70 F. Meunier, P. Delporte, B. Heinrich, C. Bouchy, C. Crouzet,
C. Phamhuu, P. Panissod, J. L. Leroud, P. L. Mills and
M. J. Ledoux, J. Catal., 1997, 169, 33.
71 K. Aitlachgar, A. Tuel, M. Brun, J. M. Herrmann, J. M. Krafft,
J. R. Martin, J. C. Volta and M. Abon, J. Catal., 1998, 177, 224.
72 A. Bruckner, A. Martin, B. Kubias and B. Lucke, J. Chem. Soc.,
Faraday Trans., 1998, 94, 2221.
73 A. Martin, G. U. Wolf, U. Steinike and B. Lucke, J. Chem. Soc.,
Faraday Trans., 1998, 94, 2227.
74 P. Delichere, K. E. Bere and M. Abon, Appl. Catal., A, 1998, 172,
295.
75 M. Ruitenbeck, A. J. van Dillen, A. Barbon, E. E. van Faassen,
D. C. Koningsberger and J. W. Geus, Catal. Lett., 1998, 55, 133.
76 P. Ruiz, Ph. Bastians, L. Caussin, R. Reuse, L. Daza, D. Acosta
and B. Delmon, Catal. Today, 1993, 16, 99.
77 H. Morishige, J. Tamaki, N. Msura and N. Yamazoe, Chem. Lett.,
1990, 1513.
78 S. Sajip, C. Rhodes, J. K. Bartley, A. Burrows, C. J. Kiely and
G. J. Hutchings, in Catalytic Activation and Functionalisation of
Light Alkanes, ed. E. G. Derouane, Kluwer, Dordrecht, 1998,
p. 429.
79 V. V. Guliants, S. A. Holmes, J. B. Benziger, P. Heaney, D. Yates
and I. E. Wachs, J. Mol. Catal. A, 2001, 172, 265.
80 J. A. Lopez-Sanchez, J. K. Bartley, A. Burrows, C. J. Kiely,
M. Havecker, R. Schogl, J. C. Volta, M. Poliakoff and
G. J. Hutchings, New J. Chem., 2002, 26, 1811.
81 R. Ebner and M. R. Thompson, Catal. Today, 1993, 16, 51.
82 G. J. Hutchings and M. S. Scurrell, CATTECH, 2003, 7, 90.

J. Mater. Chem., 2004, 14, 33853395

3395

Você também pode gostar