Transformation of nano-sized vanadyl hydrogen phosphate hemihydrate crystallites to vanadyl pyrophosphate during activation in the presence of n-butane and oxygen

Article abstract of DOI:10.1016/j.jcat.2008.02.002

The transformation of nano-sized VOHPO₄s0.5H₂O crystallites (340 nm × 40 nm) to (VO)₂P₂O₇ was examined. The crystalline structure of VOHPO₄s0.5H₂O quickly collapsed to form an oxidized amorphous phase within 1 h (reaction conditions: 1.5% n-butane, 17% O₂, and 81.5% He at 663 K), followed by gradual crystallization to (VO)₂P₂O₇, accompanied by the formation of sharply angular nano-sized crystallites (about 50 nm). No crystalline phases other than (VO)₂P₂O₇ were formed during this transformation. This is quite different from the transformation of large and thick VOHPO₄s0.5H₂O crystallites, in which α_II-VOPO₄ and δ-VOPO₄ have been detected in the resulting catalyst. The selectivity to maleic anhydride increases during the transformation as a result of the formation of (VO)₂P₂O₇. The activity increases in the early stages of the transformation, then decreases with time. The decrease in activity is caused by gradual reduction of the catalyst surface during the transformation.

Full text of DOI:10.1016/j.jcat.2008.02.002

Journal of Catalysis 255 (2008) 213–219
www.elsevier.com/locate/jcat
Transformation of nano-sized vanadyl hydrogen phosphate hemihydrate
crystallites to vanadyl pyrophosphate during activation in the presence of
n-butane and oxygen
Hiroyuki Imai, Yuichi Kamiya ∗, Toshio Okuhara
Research Faculty of Environmental Earth Science, Hokkaido University, Sapporo 060-0810, Japan
Received 19 December 2007; revised 5 February 2008; accepted 7 February 2008
Available online 10 March 2008
Abstract
The transformation of nano-sized VOHPO ·0.5H O crystallites (340 nm × 40 nm) to (VO) P O was examined. The crystalline structure of
7
2 2
4
2
VOHPO ·0.5H O quickly collapsed to form an oxidized amorphous phase within 1 h (reaction conditions: 1.5% n-butane, 17% O , and 81.5%
4
2
2
He at 663 K), followed by gradual crystallization to (VO) P O , accompanied by the formation of sharply angular nano-sized crystallites (about
7
2
2
50 nm). No crystalline phases other than (VO) P O were formed during this transformation. This is quite different from the transformation of
7
2
2
large and thick VOHPO ·0.5H O crystallites, in which α -VOPO and δ-VOPO have been detected in the resulting catalyst. The selectivity to
II
4
2
4
4
maleic anhydride increases during the transformation as a result of the formation of (VO) P O . The activity increases in the early stages of the
7
2
2
transformation, then decreases with time. The decrease in activity is caused by gradual reduction of the catalyst surface during the transformation.
© 2008 Elsevier Inc. All rights reserved.
Keywords: n-Butane oxidation; Vanadyl pyrophosphate; Nano-crystallites; Transformation; Microstructure
1. Introduction
Abon et al. [6] proposed that in the initial stage, VOHPO4·
0.5H2O is in part transformed into (VO)2P2O7 through topotac-
tic transformation and in part oxidized to δ-VOPO4, which is
then reduced to (VO)2P2O7 in the reaction gas (n-C4H10/O2/He
= 1.6/18/80.4) at 673 K, becoming completely reduced after
a long time on stream. They also suggested that the V4+/V5+
ratio on the surface of the catalyst is an important factor in con-
trolling the activity; that is, a catalyst with a high V4+/V5+
ratio exhibits high activity. Hutchings et al. [7] observed the
transformation through in-situ laser Raman spectroscopy; as
temperature was increased from room temperature to 667 K,
the peaks due to VOHPO4·0.5H2O completely disappeared at
643 K, and peaks attributed to (VO)2P2O7 as well as αII-, γ -,
and δ-VOPO4 appeared instead, until only a strong peak for
(VO)2P2O7 was observed at 667 K. O’Mahony et al. [10] re-
ported that during the activation process in n-butane/air mix-
ture, (VO)2P2O7 was formed at 573 K with no intermediate
amorphous phase. Conte et al. [11], using in-situ XRD, ob-
served the formation of ε-VOPO4 from VOHPO4·0.5H2O in
N2 and its rapid transformation into δ-VOPO4 through expo-
sure to n-butane at the reaction temperature.
Vanadyl pyrophosphate, (VO)2P2O7, is a major active com-
ponent of a catalyst used for selective oxidation of n-butane
to maleic anhydride (MA) [1–3]. The (VO)2P2O7 catalyst is
derived from the precursor VOHPO4·0.5H2O by thermal treat-
ment under an inert atmosphere. An active and selective catalyst
also has been produced by treating VOHPO4·0.5H2O in an n-
butane/air mixture at 600–750 K for more than 100 h [1,2,4,5].
Under these reaction conditions, the transformation from
VOHPO4·0.5H2O to (VO)2P2O7 proceeds through a compli-
cated process involving dehydration, oxidation, and reduction
[6–10]. Because the microstructure of the solid sample changes
with treatment time, so do the selectivity and activity in n-
butane oxidation. Thus, investigation of the transformation pro-
vides valuable information on factors influencing the selectivity
and activity of the (VO)2P2O7 catalyst.
*
Corresponding author. Fax: +81 11 706 4513.
E-mail address: kamiya@ees.hokudai.ac.jp (Y. Kamiya).
0021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.02.002

214
H. Imai et al. / Journal of Catalysis 255 (2008) 213–219
Kiely et al. [8] observed the formation of catalysts contain-
for 1 h at each temperature under stirring to form a homoge-
neous solution containing exfoliated VOPO4 sheets as the prod-
uct of intercalation, followed by exfoliation of VOPO4·2H2O
[16]. To the resulting homogeneous 2-butanol solution, ethanol
(14.4 cm3, Wako Pure Chem. Ind., Ltd.) was added at room
temperature, and then this solution was refluxed for 20 h (reduc-
tion) to form a light-blue precipitate. This precipitate was sepa-
rated by centrifugation, washed with acetone, and dried at room
temperature overnight to obtain nano-sized VOHPO4·0.5H2O
crystallites.
ing (VO)2P2O7 and crystalline VOPO4 (V5+) phases in cer-
tain proportions from precursors prepared by different prepa-
ration methods. These results prompted an investigation of
the effect of VOHPO4·0.5H2O crystallite size on the crys-
talline phase and microstructure of the resulting catalysts.
The data clearly demonstrated that micrometer-sized crystal-
lites (1000 nm × 110 nm) of VOHPO4·0.5H2O transformed
into a single phase of (VO)2P2O7, whereas large crystallites
(10,000 nm × 415 nm) transformed into catalyst particles con-
sisting of peripheral (VO)2P2O7 and internal αII-VOPO4 under
reaction conditions of 1.5% n-butane, 17% O2, and 81.5% He
at 663 K for 300 h [12]. The catalyst derived from micrometer-
sized VOHPO4·0.5H2O crystallites showed relatively high ac-
tivity and selectivity for the oxidation of n-butane compared
with those from the large crystallites.
2.2. Transformation of nano-sized VOHPO4·0.5H2O into
catalyst
Transformation of nano-sized VOHPO4·0.5H2O to the cat-
alyst was performed at 663 K in a flow reactor (Pyrex tube,
10 mm i.d.) under atmospheric pressure. After VOHPO4·
0.5H2O powder (0.20 g) was placed in the reactor, reactant
gas consisting of 1.5% n-butane, 17% O2, and 81.5% He was
added at a flow rate of 10 cm3 min−1. The temperature was
raised from room temperature to 663 K at a rate of 5 K min−1
and then maintained at 663 K. To observe structural changes
of the solids during the transformation, solid samples were pre-
pared under different treatment times (1, 5, 10, 15, 31, 120, and
300 h after the temperature reached 663 K). After the predeter-
mined time, the reactor was cooled rapidly to room temperature.
The samples were stored in vacuo until characterized. A sample
treated for x h is designated Sxh.
During the transformation, the gas at the outlet of the reactor
was analyzed by online gas chromatography. A Shimadzu GC-
8A gas chromatograph equipped with flame ionization detector
and a Porapak QS column (1 m) was used to analyze n-butane
and MA; an Area M200 high-speed gas chromatograph with
Porapak Q and molecular sieve 5A columns was used to analyze
CO, CO2, and O2.
Much smaller and thinner VOHPO4·0.5H2O crystallites
(340 nm × 40 nm) have been successfully synthesized by
exfoliation–reduction of VOPO4·2H2O in a mixture of 2-
butanol/ethanol [13,14]. The (VO)2P2O7 catalyst derived from
nano-sized VOHPO4·0.5H2O crystallites exhibited extremely
high selectivity to MA (about 84% at 47% conversion and 78%
even at 85% conversion at 663 K). Interestingly, in contrast to
micrometer-sized VOHPO4·0.5H2O, the resulting catalyst was
composed of crystallites about 50 nm in length, much smaller
than those of the starting VOHPO4·0.5H2O crystallites (340 nm
long).
The present study examined changes in the microstruc-
ture of solid samples during transformation of nano-sized
VOHPO4·0.5H2O crystallites to (VO)2P2O7 using X-ray dif-
fraction (XRD), Raman spectroscopy, X-ray photoelectron
spectroscopy (XPS), redox titration, N2 adsorption–desorption
isotherms, scanning electron microscopy (SEM), and tempera-
ture-programmed reduction (TPR). Furthermore, changes in the
activity and selectivity for n-butane oxidation during the trans-
formation in relation to the physicochemical properties of the
catalyst (including microstructure and redox properties) were
evaluated, to provide insight into factors affecting the catalyst
activity and selectivity during the oxidation of n-butane.
2.3. Characterization
The powder XRD patterns of the solid samples were mea-
sured with a Rigaku Miniflex XRD diffractometer with CuKα
radiation. Raman spectra of the solid samples were obtained
with a JASCO RMP-230 Raman spectrometer. Adsorption–
desorption isotherms of nitrogen were obtained at 77 K with a
BEL Japan Belsorp 28SA automatic adsorption apparatus after
evacuation of the chamber containing the sample at 423 K for
1 h. Specific surface area was calculated using the BET method.
SEM images were obtained with a Hitachi FE-SEM S-4800
instrument. The average oxidation number of V in the sam-
ple bulk was determined using a redox-titration method with
KMnO4 [17]. XPS was carried out on a Shimadzu XPS-7000
instrument with MgKα radiation. The average oxidation num-
ber of V on the surface was estimated as described previ-
ously [6].
2. Experimental
2.1. Preparation of nano-sized VOHPO4·0.5H2O crystallites
VOPO4·2H2O, which served as the raw material for nano-
sized VOHPO4·0.5H2O crystallites, was prepared as follows.
A mixture of V2O5 (24 g, Wako Pure Chem. Ind., Ltd.), aque-
ous 85% H3PO4 (223 g, Wako Pure Chem. Ind., Ltd.), and H2O
(577 cm3) was refluxed for 16 h. The resulting precipitate was
separated by filtration, washed with acetone, and dried under
ambient atmosphere. The resulting yellow solid was identified
as VOPO4·2H2O by XRD and IR [15].
Nano-sized VOHPO4·0.5H2O crystallites were prepared
by exfoliation–reduction of VOPO4·2H2O in a mixture of 2-
butanol and ethanol, as reported previously [13]. A mixture of
VOPO4·2H2O (1.0 g) and 2-butanol (50 cm3, Wako Pure Chem.
Ind., Ltd.) was heated stepwise at 303, 323, 343, and 363 K
TPR profiles were obtained using a custom apparatus
equipped with an Anelva M-200Q quadrupole mass spectrom-
eter. A sample of approximately 0.02 g was fixed in a quartz
U-shaped tube and pretreated under He flow (20 cm3 min−1) at

H. Imai et al. / Journal of Catalysis 255 (2008) 213–219
215
Fig. 1. Time course changes for selective oxidation of n-butane during transfor-
mation of nano-sized VOHPO ·0.5H O. The reaction was conducted at 663 K
4
2
in a mixture of 1.5% n-butane, 17% O , and 81.5% He. (!) Conversion of
2
n-butane; (1) selectivity to MA; (2) yield of MA; (Q) yield of CO .
x
423 K for 1 h. After the pretreatment, the gas was switched to
H2 and the temperature increased at a rate of 10 K min−1 while
monitoring signals of m/e = 18.
3. Results
Fig. 1 shows the time course for the selective oxidation of
n-butane at 663 K. The conversion of n-butane was low ini-
tially (0 h), then increased rapidly up to 30–40 h. After reaching
maximum conversion, the level of conversion decreased gradu-
ally, until a constant value was achieved at 300 h. Selectivity to
MA was very low at the beginning of the reaction but increased
dramatically to 75% within the first 40 h, finally reaching a con-
stant value of 82% after 100 h. As a result, the yield of MA
showed a trend similar to that of the conversion rate.
Fig. 2 shows changes in the XRD patterns of the samples
during transformation from the precursor VOHPO4·0.5H2O to
the catalyst, as well as changes in the intensities of diffraction
lines attributed to the (200), (042), and (023) planes. The pre-
cursor was confirmed to be crystalline VOHPO4·0.5H2O. The
intense diffraction lines attributed to VOHPO4·0.5H2O disap-
peared completely, and no diffraction line was observed at 1 h,
indicating the amorphous nature of S1h. At 5 h, very weak dif-
fraction lines attributed to the (VO)2P2O7 phase appeared. The
intensities of the diffraction lines due to (VO)2P2O7 increased
rapidly up to 31 h and more slowly thereafter (Fig. 2B). After
300 h, only the XRD pattern attributed to the highly crystalline
(VO)2P2O7 was observed.
Fig. 2. Changes in: (A) XRD patterns of the samples during the transforma-
tion [" and 1 represent VOHPO ·0.5H O and (VO) P O , respectively]; (B)
7
2 2
4
2
intensity of diffraction lines as a function of treatment time [(!) (200) plane;
(") (042) plane; (1) (023) plane]. The transformation was conducted at 663 K
in a mixture of 1.5% n-butane, 17% O , and 81.5% He.
2
Fig. 3 shows changes in the Raman spectra of the sam-
ples during the transformation. These changes are consistent
with those seen in the XRD patterns. Only peaks attributed to
VOHPO4·0.5H2O (985, 1106, and 1150 cm−1) were observed
for the precursor. These peaks disappeared completely at 1 h,
and no peaks were observed until 10 h. A peak assigned to
(VO)2P2O7 appeared at 926 cm−1 after 15 h and increased in
intensity.
Fig. 4 shows SEM images of the precursor and samples
treated over different durations. The precursor was composed of
small, thin, leaf-like crystallites with an average lateral length
of 300 nm. The thickness was estimated as 40 nm from the SEM
images. The shape and size of the sample crystallites were sim-
ilar to those of the precursor up to 10 h; however, at 15 h, many
cracks appeared on the crystallites, and very small crystallites
with a rectangular shape began to appear. Finally, at 300 h, these
crystallites were completely transformed into angular, nano-
sized crystallites of approximately 50 nm on the narrow side
and 100 nm on the long side.
Fig. 5 shows changes in surface area (Fig. 5A) and oxidation
number of V on the surface and in the bulk (Fig. 5B) as a func-
tion of treatment time (with values at 0 h representing those of
the precursor). The precursor had a surface area of 23 m2 g−1
,
which increased nearly twofold after treatment for 300 h, reach-
ing a value of 39 m2 g−1. As shown in Fig. 5B, the V oxidation

216
H. Imai et al. / Journal of Catalysis 255 (2008) 213–219
files of the samples at 31 and 300 h (Fig. 6). In the spectra, the
peaks at 780 and 920 K are assignable to the reduction of V5+
and V4+, respectively. As shown in Fig. 6, the V5+ peak was
observed for S31h. In contrast, the peak at 780 K was absent for
S300h, clearly indicating that the sample was more highly oxi-
dized at 31 h than at 300 h.
4. Discussion
Previous studies have reported on the transformation of
VOHPO4·0.5H2O, which tends to have relatively large crys-
tallites, to an active catalyst form [6–12]. The present study in-
vestigated the transformation of nano-sized VOHPO4·0.5H2O
crystallites (300 nm long, 40 nm thick). The resulting catalyst
showed excellent selectivity to MA, as was described previ-
ously [13,14].
Whereas the nano-sized precursor VOHPO4·0.5H2O was
highly crystalline, its crystal structure quickly collapsed into an
amorphous phase just after the start of treatment (1 h; Figs. 2A
and 3), accompanied by deep oxidation of the surface and bulk.
This finding indicates that the surface and bulk of the crystal-
lites were both oxidized during the initial stage of transforma-
tion, whereas any change in morphology during this process
(VOHPO4·0.5H2O → partially oxidized amorphous) was small
and any change in surface area was negligible. In terms of cat-
alytic properties for the amorphous material, S5h and S10h ex-
hibited decent activity and selectivity to MA (Fig. 1), whereas
they were basically amorphous. Because the amorphous phase
was formed simply by the dehydration and partial oxidation of
VOHPO4·0.5H2O, we can assume that crystallographic regu-
larity of VOHPO4·0.5H2O, such as V–O–V dimer structure,
Fig. 3. Changes in Raman spectra of samples during the transformation from
the precursor to the catalyst. The transformation was conducted at 663 K in a
mixture of 1.5% n-butane, 17% O , and 81.5% He.
2
number increased dramatically to a maximum after 1 h (V4.78+
for the surface and V4.40+ for the bulk). The oxidation numbers
then gradually decreased and become constant at 300 h (V4.18+
for the surface and V4.06+ for the bulk). The surface was more
highly oxidized at 31 h than at 300 h, whereas the oxidation
state of the sample bulk was nearly the same at 31 h as at 300 h.
To further investigate this difference, we measured TPR pro-
Fig. 4. SEM images of samples at different times. The transformation was conducted at 663 K in a mixture of 1.5% n-butane, 17% O , and 81.5% He.
2

H. Imai et al. / Journal of Catalysis 255 (2008) 213–219
217
Fig. 4. (continued)
remained partly in the amorphous phase. This may have led to
the decent activity and selectivity for these amorphous materi-
als. Hutchings et al. [18,19] reported that an amorphous VPO
material (VPOLP) prepared using liquid CO2 partially crystal-
lized under the reaction conditions of the n-butane oxidation,
but another amorphous VPO (VPOSCP) synthesized in super-
critical CO2 did not crystallize even under the such conditions
over a long period. Because our amorphous material was gradu-
ally transformed into crystalline (VO)2P2O7 under the reaction
conditions, it may have a similar structural nature to the VPOLP
.
Crystalline phases other than (VO)2P2O7 were not observed
by XRD (Fig. 2A) or Raman spectroscopy (Fig. 3) for sam-
ples obtained during transformation from the partially oxidized
amorphous (S1h) to (VO)2P2O7 (S300h). These results suggest
that the nano-sized VOHPO4·0.5H2O crystallites could trans-
form to crystalline (VO)2P2O7 as follows: VOHPO4·0.5H2O
→ partially oxidized amorphous phase → highly crystalline
(VO)2P2O7.
The SEM images (Fig. 4) show the formation of much
smaller crystallites after the appearance of cracks on the crys-
tallites at 15 h. These smaller crystallites steadily transform
into much smaller nano-sized rectangular crystallites with in-
creasing treatment time. Because the formation of these rec-
tangular crystallites was coincident with the appearance of the
(VO)2P2O7 phase in the XRD and Raman spectra, the rec-
tangular crystallites observed in S15h were (VO)2P2O7 crys-
tallites. As reported previously [13], TEM and selected-area
electron diffraction (ED) of S300h demonstrated that the rectan-
gular nano-sized crystallites of S300h consisted of single crystals
of (VO)2P2O7. At the onset of crystallization of (VO)2P2O7
(S15h), an amorphous phase remains, as shown by the very
Fig. 5. Changes in (A) surface area and (B) average oxidation number of V
[(") on the surface and (!) in bulk] as a function of treatment time. The trans-
formation was conducted at 663 K in a mixture of 1.5% n-butane, 17% O , and
2
81.5% He.
broad line at around 20–25◦ of 2θ in Fig. 2A. The morpholog-
ical heterogeneity observed in the SEM image of S15h supports
this result. These findings suggest that (VO)2P2O7 was partially
formed in the crystallite, leading to the smaller crystallites.

218
H. Imai et al. / Journal of Catalysis 255 (2008) 213–219
rior of the crystallites, whereas the (VO)2P2O7 phase is formed
directly on simple dehydration of VOHPO4·0.5H2O (topotactic
transformation) at the periphery of the crystallites. Similarly,
Abon et al. [6] demonstrated that VOHPO4·0.5H2O is partially
transformed into (VO)2P2O7 through topotactic transformation
and partially into (VO)2P2O7 via the δ-VOPO4 phase. Kamiya
et al. [12] reported that crystallite size affects the transforma-
tion and that the subsequent microstructure and phases of the
resulting catalysts depend on the size of the VOHPO4·0.5H2O
crystallites. The micrometer-sized VOHPO4·0.5H2O crystal-
lites (1000 nm × 110 nm) transform into a single phase
(VO)2P2O7 with a high specific surface area, whereas the
catalyst derived from the large VOHPO4·0.5H2O crystallites
(10,000 nm × 415 nm) comprises particles with a double-layer
structure, consisting of peripheral (VO)2P2O7 and internal αII-
VOPO4.
In those previous studies, the catalyst particles contained
a double-layer structure, comprising peripheral (VO)2P2O7
crystallites and internal crystallites including (VO)2P2O7, δ-
VOPO4, or αII-VOPO4, whereas the apparent shape and size
(morphology) of the catalyst particles were very similar to
those of the starting VOHPO4·0.5H2O. In contrast, the nano-
sized VOHPO4·0.5H2O (300 nm long, 40 nm thick) crystal-
lites observed in the present work, which were much smaller
than those reported previously, transformed into a bare, sin-
gle crystal of (VO)2P2O7 that did not have a double-layer
structure. During the transformation from VOHPO4·0.5H2O
to (VO)2P2O7, dehydration of crystalline water is inevitable
regardless of the treatment conditions (reactant gas or in-
ert gas). For large VOHPO4·0.5H2O crystallites, water at the
periphery of the crystallites likely can readily escape, lead-
ing to topotactic transformation from VOHPO4·0.5H2O to
Fig. 6. Temperature-programmed reduction profiles for samples treated for 31
and 300 h at 663 K.
The surface area increased considerably after 15 h (Fig. 5A)
due to morphological changes in the solid caused by forma-
tion of crystalline (VO)2P2O7. Schematic models of the crys-
tallites during transformation of nano-sized VOHPO4·0.5H2O
to (VO)2P2O7 are depicted in Fig. 7.
The striking difference between the results of the present
study and those of previous studies of the transformation
of VOHPO4·0.5H2O is the observation of crystalline phases
formed during the intermediate stages of transformation. No
crystalline phases were formed from nano-sized VOHPO4·
0.5H2O, whereas VOPO4 phases, including αII-VOPO4 and
δ-VOPO4, are known to be temporarily formed from larger
VOHPO4·0.5H2O crystallites [6–9,12]. Using TEM, XRD, and
31P NMR spin-echo mapping, Kiely et al. [9] demonstrated that
the transformation from VOHPO4·0.5H2O (2 µm long, 100 nm
thick) to (VO)2P2O7 proceeds via formation of δ-VOPO4 as an
intermediate, followed by reduction to (VO)2P2O7, in the inte-
Fig. 7. Schematic models of crystallite structures during transformation from nano-sized VOHPO ·0.5H O to (VO) P O .
7
2 2
4
2

H. Imai et al. / Journal of Catalysis 255 (2008) 213–219
219
(VO)2P2O7 at the periphery but preventing the water inside
5. Conclusion
the crystallites from escaping [9]. This promotes the forma-
tion of VOPO4 phases (V5+), such as δ-VOPO4, inside the
particle, because VOHPO4·0.5H2O readily transforms to δ-
VOPO4 under the reaction conditions [6]. In contrast, for nano-
sized VOHPO4·0.5H2O crystallites, dehydration proceeded
uniformly throughout the crystallites due to their small size,
preventing the formation of oxidized VOPO4 phases. Conse-
quently, nano-sized VOHPO4·0.5H2O was transformed into
crystalline (VO)2P2O7 without the formation of any crystalline
phases other than (VO)2P2O7 under the reaction conditions.
As shown in Fig. 1, selectivity to MA increased rapidly dur-
ing the initial 20 h and then increased more slowly with in-
creasing treatment time. This trend closely follows the growth
of crystalline (VO)2P2O7 as confirmed by XRD (Fig. 2A) and
Raman spectra (Fig. 3). Because (VO)2P2O7 is considered
indispensable for the formation of MA, although the active
phases involving δ-VOPO4, the X1 phase, and the interface
of V4+/V5+ remain controversial, evolution of the (VO)2P2O7
phase is a major reason for the increased selectivity to MA with
treatment time.
Nano-sized precursor VOHPO4·0.5H2O crystallites (300 nm
long, 40 nm thick), were transformed into highly crystalline
(VO)2P2O7 through a deeply oxidized amorphous intermedi-
ate at 663 K in an n-butane/air mixture over 300 h. During
this transformation, no crystalline phases other than (VO)2P2O7
were formed, a result that differs significantly from the trans-
formation in micrometer-sized VOHPO4·0.5H2O crystallites.
The selectivity to MA increased monotonically with treatment
time due to the formation of crystalline (VO)2P2O7. In contrast,
the activity increased during the early stages of transformation
but then decreased over time. During the transformation, the
catalyst surface oxidation state decreased gradually, causing a
decrease in activity.
Acknowledgment
This work was supported by a grant from Core Research for
Evolution Science and Technology (CREST) of the Japan Sci-
ence and Technology Corporation (JST).
In contrast to the trend in selectivity, the conversion in-
creased to a maximum after approximately 40 h but then de-
creased with further treatment time. n-Butane is considered
to be selectively activated on (VO)2P2O7 [1–3]; thus, the in-
creasing activity up to 40 h likely is due to formation of the
(VO)2P2O7 phase. But the decrease in activity over time can-
not be explained by changes in the crystalline phases or sur-
face area, because the formation of crystalline (VO)2P2O7 pro-
ceeded monotonically with treatment time while the surface
area continued to increase after 40 h. The selective oxidation of
n-butane over (VO)2P2O7 proceeds via a redox mechanism in-
volving V4+ and V5+ at a few surface layers [20–22]. n-Butane
is oxidized to MA over the partially oxidized phase (V5+),
and, consequently, the partially oxidized phase transforms to
the reduced phase [V4+; (VO)2P2O7]. The reduced phase is
then reoxidized by O2 to regenerate the oxidized phase. In this
mechanism, a catalyst containing a large amount of the oxidized
phase should be more active. In addition, if the rate of reoxida-
tion is slower than the rate of reduction, then the oxidation state
of the surface will gradually decrease with treatment time, caus-
ing a deterioration in activity. Indeed, as shown in Figs. 5B and
6, the oxidation state of the surface decreased gradually with
treatment time. The activity also decreased with a reduction in
oxidation state. These results are consistent with the considera-
tions described above.
References
[1] G. Centi, F. Trifiró, J.R. Ebner, V.M. Franchetti, Chem. Rev. 88 (1988) 55.
[2] G. Centi, F. Cavani, F. Trifirò, in: Selective Oxidation by Heterogeneous
Catalysis, Kluwer Academic, New York, 2001, p. 143.
[3] G.J. Hutchings, J. Mater. Chem. 14 (2004) 3385.
[4] B.K. Hodnett, Catal. Rev. Eng. 17 (1985) 373.
[5] S. Albonetti, F. Cavani, F. Trifiró, Catal. Rev. Sci. Eng. 38 (1996) 413.
[6] M. Abon, K.E. Bere, A. Tuel, P. Delochere, J. Catal. 156 (1995) 28.
[7] G.J. Hutchings, A. Desmartin-Chomel, R. Olier, J.C. Volta, Nature 368
(1994) 41.
[8] C.J. Kiely, A. Burrows, S. Sajip, G.J. Hutchings, M.T. Sananes, A. Tuel,
J.C. Volta, J. Catal. 162 (1996) 31.
[9] C.J. Kiely, A. Burrows, G.J. Hutchings, K.E. Bere, J.C. Volta, A. Tuel, M.
Abon, Faraday Discuss. 105 (1996) 103.
[10] L. O’Mahony, T. Curtin, J. Henry, D. Zemlyanov, M. Mihov, B.K. Hod-
nett, Appl. Catal. A 285 (2005) 36.
[11] M. Conte, G. Budroni, J.K. Bartley, S.H. Taylor, A.F. Carley, A. Schmidt,
D.M. Murphy, F. Girgsdies, T. Ressler, R. Schlögl, G.J. Hutchings, Sci-
ence 313 (2006) 1270.
[12] Y. Kamiya, N. Hiyoshi, N. Ryumon, T. Okuhara, J. Mol. Catal. A 220
(2004) 103.
[13] Y. Kamiya, N. Ryumon, H. Imai, T. Okuhara, Catal. Lett. 111 (2006) 159.
[14] H. Imai, Y. Kamiya, T. Okuhara, J. Catal. 251 (2007) 195.
[15] T. Nakato, Y. Furumi, N. Terao, T. Okuhara, J. Mater. Chem. 10 (2000)
737.
[16] N. Yamamoto, N. Hiyoshi, T. Okuhara, Chem. Mater. 14 (2002) 3882.
[17] B.K. Hodnett, P. Permannce, B. Delmon, Appl. Catal. 6 (1983) 231.
[18] G.J. Hutchings, J.K. Bartley, J.M. Webster, J.A. Lopez-Sanchez, D.J.
Gilbert, C.J. Kiely, A.F. Carley, S.M. Howdle, S. Sajip, S. Caldarelli, C.
Rhodes, J.C. Volta, M. Poliakoff, J. Catal. 197 (2001) 232.
[19] G.J. Hutchings, J.A. Lopez-Sanchez, J.K. Bartley, J.M. Webster, A. Bur-
rows, C.J. Kiely, A.F. Carley, C. Rhodes, M. Hävecker, A. Knop-Gericke,
R.W. Mayer, R. Schlögl, J.C. Volta, M. Poliakoff, J. Catal. 208 (2002) 197.
[20] M. Pepera, J.L. Callahan, M.J. Desmond, E.C. Milberger, P.R. Blum, N.J.
Bremer, J. Am. Chem. Soc. 107 (1985) 4883.
Kamiya et al. [23] demonstrated that highly crystalline
(VO)2P2O7 has a low ratio of reoxidation to reduction rate
(rOX/rRE); that is, the surface is difficult to reoxidize. The
present catalyst system demonstrated the growth of crystalline
(VO)2P2O7 with treatment time. The growth of crystalline
[21] T. Okuhara, M. Misono, Catal. Today 16 (1993) 61.
[22] G. Koyano, T. Okuhara, M. Misono, J. Am. Chem. Soc. 120 (1998) 767.
[23] Y. Kamiya, E. Nishikawa, T. Okuhara, T. Hattori, Appl. Catal. A 206
(2001) 103.
(VO)2P2O7 is a likely reason for the reduction in rOX/rRE
,
which results in decreased catalytic activity with treatment
time.