Novel modifications in preparing vanadium phosphorus oxides and their applications for partial oxidation of n-butane

Article abstract of DOI:10.1016/S1381-1169(03)00422-9

One novel modification in preparing high surface area VPO catalyst was established by applying polyethylene glycols (PEGs) as the unique additives. The PEG-derived samples possessed considerably high surface areas of 41-54 m²/g, with the averag

Full text of DOI:10.1016/S1381-1169(03)00422-9

Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
Novel modifications in preparing vanadium phosphorus oxides and
their applications for partial oxidation of n-butane
b
a,1
a
a,∗
a
a
Xiaoshu Wang , Lijun Xu , Xi Chen , Weijie Ji , Qijie Yan , Yi Chen
a
Department of Chemistry, Nanjing University, Nanjing 210093, China
b
Center for Materials Analysis, Nanjing University, Nanjing 210093, China
Received 13 March 2003; accepted 6 May 2003
Abstract
One novel modification in preparing high surface area VPO catalyst was established by applying polyethylene glycols
2
(
PEGs) as the unique additives. The PEG-derived samples possessed considerably high surface areas of 41–54 m /g, with
the average particle size of 100–200 nm and well crystallized (VO)2P2O7 phase, showing notably improved performances
for the partial oxidation of n-butane to maleic anhydride (MA) even at lower reaction temperature. The principle function of
PEG in preparation chemistry and the possible origins of the enhanced performance were preliminarily discussed. Another
modification in preparing high surface area VPO catalyst adopted in this study was the combination of synthesizing high
surface area VPO precursor by well-controlled addition of phosphoric acid, thermal pre-treating the precursor at certain
temperature and ball milling in cyclohexane. The so-obtained catalysts also showed pronounced catalytic performance. The
applied milling process not only effectively increased the surface areas of the milled samples and consequently the available
number of active sites, but also somehow changed other sample characteristics such as the local environment and the redox
behavior; which may also be responsible for the overall performance enhancement.
©
2003 Elsevier B.V. All rights reserved.
Keywords: VPO; Polyethylene glycol; Ball milling; n-Butane; Maleic anhydride
1
. Introduction
tion between the surface and bulk elements, and the
difficulty in detecting intermediate products in cat-
alytic reaction [1]. After years of work, it has been
recognized that preparation chemistry plays an in-
comparable role in determining the structure of VPO
catalysts and, accordingly, the catalytic behavior.
There is also consensus that (VO)2P2O7 is the major
component of active VPO catalysts [2], but different
phase components were also announced in papers or
patents [3,4]. In recent years, there is a worldwide
trend that the selectivity is more emphasized than the
overall conversion, in order to decrease the amount
of undesired by-products, especially the harmful ones
[5]. For partial oxidation of lower paraffins including
n-butane, the challenge is how to further increase the
Vanadium phosphorus oxides are widely used in
partial oxidation of n-butane to maleic anhydride
MA). Although laboratory and industry efforts have
(
been exerted on this catalyst system, lots of details are
still not thoroughly elaborated, especially the active
phase/sites, the surface vanadium oxidation states,
the detailed catalytic mechanism and so on, which
is may be ascribed to both the complicated interac-
∗
Corresponding author. Fax: +86-25-331-7761.
E-mail address: jiwj@nju.edu.cn (W. Ji).
1
Present address: Department of Chemistry, University of Wash-
ington, Seattle, WA 98195-1700, USA.
1
381-1169/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1381-1169(03)00422-9

2
62
X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
selectivity of target product, otherwise costly reactor
design and engineering would be required to mini-
mize the inefficiency of the process. Currently used
industrial VPO catalysts showed the butane conver-
sion beyond 80%, with MA selectivity of around
catalyst (symbolized as P1) was prepared similarly
except that no PEG additive was introduced.
2.2. Preparation of the ball milled VPO catalysts
6
5 mol%. Large portion of n-butane is burned to car-
The VPO precursors were prepared in organic
bon oxides. It is known that MA selectivity is strongly
associated with n-butane conversion and can be in-
creased at the expense of conversion. Although a few
attempts have been made to improve MA selectivity,
it is hardly beyond 80% with adequate n-butane con-
version. Dupont reported that a higher level of MA
selectivity could be achieved using a rising reactor in
which the oxidation of n-butane and the regeneration
of the catalyst were operated in the separate zones
medium by employing the reaction of vanadium
pentoxide (32.9 g) with isobutanol (120 ml) in ben-
zyl alcohol (120 ml). After the reaction mixture was
refluxed for 6 h, phosphoric acid (29.3 g, 85 wt.%)
was added into the hot black mixture at a rate of
0.6 ml/min, with the V/P ratio being 1.0/1.2. After re-
fluxed for another 6 h, the turbid reaction mixture was
filtered and the obtained blue–green precipitate, which
was typically defined as VPO catalyst precursor, was
dried at 393 K in air. The dried precursor was then
pre-conditioned in flowing nitrogen at 573 K for 2 h,
after that the received material was subjected to ball
milling in agate pots with the solvent of cyclohexane
for 5–20 h. Once the milling procedure completed,
the agate pots were dried at 393 K in air to remove
the solvent, and the collected sample was activated
in the mixture of 1.5% n-butane/air (gas space hour
[6]. This process is rather attractive for high MA se-
lectivity, but has not yet been widely applied possibly
due to its specialty in engineering requirement and
in preparation of the exclusive catalyst. Alternatively,
improving catalyst performance through the approach
of preparation chemistry is still very helpful. In this
study, two modifications in VPO catalyst preparation
have been studied in order to generate more efficient
VPO systems.
−
1
velocity, GSHV = 1200 h ) at 673 K for 6–24 h.
2.3. Sample characterization and evaluation
2
. Experimental
BET surface areas of the samples were measured
by nitrogen adsorption at 77 K on a Micromeritics
ASAP2000 adsorption apparatus. X-ray diffraction
2
.1. Preparation of the VPO catalysts in the
presence of polyethylene glycols (PEGs)
(XRD) was conducted on a Shimadzu XD-3A diffrac-
The details for synthesizing VPO catalyst in the
presence of polyethylene glycols (PEGs) with differ-
ent molecular weight can be referred in a Chinese
patent [7]. Briefly, the catalyst precursor was prepared
by refluxing of vanadium pentoxide in a mixture of
iso-butanol/benzyl alcohol, then appropriate amount
of PEGs (PEG2000–PEG20000) and phosphoric acid
tometer with graphite-filtered Cu K␣ radiation. The
morphology and particle size of the samples were ex-
amined on a TEM-200CX transmission electron mi-
croscopy (TEM). XPS measurement was performed
on an X-ray photoelectron spectrometer—VG ES-
CALAB MK II, with the X-ray energy of 1253.6 eV
(Mg K␣). The high voltage and emission current
are 12 kV and 20 mA, respectively. The binding en-
ergy (BE) for each element was referred to the C
1s signal (284.6 eV). Removal of O 1s satellite and
curve-fitting of the V 2p3/2 peak were performed by
using the SCAKAB software. The surface concentra-
tion of the elements was estimated on the basis of the
corresponding peak area which was normalized by
using the Wagner Factor database. Infrared spectra
were collected on a Nicolet 170SX Fourier transform
(
85 wt.%) were added in for further refluxing, with
the controlled vanadium/phosphorus atomic ratio of
.0/1.2. The resulting suspension was cooled down
1
to room temperature and filtered. The precipitate
was washed with iso-butanol and acetone and dried
at 393 K for 24 h. The derived precursor was acti-
vated in the mixture of 1.5% n-butane/air at a rate of
2
K/min from room temperature to 673 K and kept at
this temperature for 15 h. The obtained samples are
symbolized as P2–P5. For comparison, another VPO
−1
infrared spectrometer, with the resolution of 4 cm
.

X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
263
2
Temperature-programmed reduction (TPR) was per-
formed in the temperature range of 298–993 K. An
amount of 0.2–0.5 g sample was charged in a quartz
reactor and reduced in the mixture of 10% H2 in
the balanced N2 at a rate of 10 K/min. The catalytic
performance was evaluated in a quartz micro-reactor
in the temperature range of 653–698 K. The reaction
mixture consists of 1.5% n-butane, 17.0% oxygen and
balanced nitrogen. The GSHV is 1200–1800 h . Two
on-line gas chromatography systems were adopted
to analyze the outlet reaction mixture and the carbon
balance was generally better than 95%.
areas (41–56 m /g), with rather small particles (ca.
100–200 nm) of the well-crystallized (VO)2P2O7
phase. More importantly, the catalysts showed great
performance for partial oxidation of n-butane to MA.
The samples prepared by applying PEG2000,
PEG6000, PEG10000 and PEG20000 had their sur-
2
face areas of 41, 56, 54 and 52 m /g, respectively.
Except for the samples prepared by using the SCFD
process [11], they are among the highest surface area
VPO analogues. Note that the P1 sample derived in
the absence of PEG only had considerably low surface
−
1
2
area of 19 m /g. XRD results indicated that all the pre-
cursors prepared in both the presence and the absence
of PEG showed the typical VO(HPO4)·0.5H2O struc-
ture. TGA measurement revealed that the PEG-derived
sample showed additional ca. 4% weight loss, which
could be attributed to the residual PEG on the pre-
cursor surface. In the following activation procedure
3
. Results and discussion
3
.1. The effects of polyethylene glycols (PEGs) on
preparing VPO catalysts
(at a rate of 2 K/min from room temperature to
It is recognized that the efficient VPO catalysts
are generally associated with the presence of vanadyl
pyrophosphate, (VO)2P2O7, and the catalyst perfor-
mance strongly depends on the preparation history
673 K in the reaction mixture and kept at this tem-
perature for 15 h), it was observed that the residual
PEG would be gradually removed from the sample.
The finally activated catalysts were presented as the
well-crystallized vanadyl pyrophosphate phase, seen
in Fig. 1. Note that although the P1 and the P3–P5
samples have essentially identical phase composition,
the relative diffraction intensity for the (2 0 0) and the
(0 4 2) crystalline planes changed notably from the P1
sample to the P3–P5 ones, suggesting there might be
[3,8,9]. In one standard method, the catalyst can be
prepared by using HCl as the reducing agent, and
2
typically has low surface areas (<10 m /g) [3]. In an-
other preparation route, the catalyst with surface area
2
of ∼20 m /g can be obtained by using alcohols as the
solvent and reducing agent, and usually shows better
catalytic performance [10]. Since the surface area is
closely related to the active sites available for the reac-
tion, it is desirable to prepare a high surface area VPO
system while still to retain the essential active compo-
nent/phase for the reaction. Different approaches have
been studied in our laboratory to synthesizing small
particle, high surface area VPO system, including
application of the SCFD process [11] and mechanical
milling the adopted VPO catalysts [12]. The sample
subjected to the SCFD process could have a BET sur-
P5
P4
P3
2
face area beyond 50 m /g, but the main sample phase
deviated from the vanadyl pyrophosphate [11]. In ad-
dition, the SCFD process cannot be easily operated,
especially for large quantity production. Here, we re-
ported a novel modification in preparing high surface
area VPO by using PEGs (PEG2000–PEG20000) as
the unique additives. It was found that this preparation
method was quite promising and the PEG-derived
VPO samples possessed significantly large surface
P1
5
10
15
20
25
30
35
40
45
50
55
2
θ/degree
Fig. 1. The XRD patterns of the PEG-derived VPO catalysts,
among them P1 was prepared in the absence of PEG while P3–P5
were prepared in the presence of PEG6000, PEG10000 and
PEG20000, respectively.

2
64
X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
1
00
9
8
7
6
5
4
0
0
0
0
0
0
Fig. 2. The TEM photograph of the VPO sample prepared in the
presence of PEG20000. The magnification is 100,000×.
6
00
620
640
660
680
700
Reaction temp.(K)
Fig. 3. Temperature dependence of the performances on the P1 and
P2 catalysts: P1 (᭝), P2 (᭺). The filled marks are for selectivity
and the open ones are for conversion.
local structure/morphology deviation between these
two kinds of the samples. TEM examination re-
vealed that the PEG-derived samples exhibited small
particle size (ca. 100–200 nm) and fairly even size
distribution. The representative picture was shown in
Fig. 2. Infrared measurement (not shown) indicated
that all the activated VPO samples had the IR bands
served in the IR spectra, suggesting that the residual
PEG was entirely removed and the activated catalyst
was PEG-free. XPS measurement demonstrated that
the P element was slightly enriched on the catalyst
surfaces, seen in Table 1, and this was in agreement
with other observations made [14,18]. The binding
energies of V 2p3/2 and O 1s were also close to those
of (VO) P O reported [14].
−
1
−1
at 1244 cm (νPO3), 1060–1069 cm (νPO3) and
−1
9
60 cm (νV=O), which were the characteristic IR
features for the (VO)2P2O7 phase [13]. On the other
hand, the PEG-derived samples (P2–P5) showed a
2
2
7
−
1
much more intensive IR band at 1247 cm , giving
another indication of local structural deviation be-
tween the two kinds of samples. After a period of
The temperature dependence of the catalyst per-
formance on the selected two samples was shown
in Fig. 3. In the temperature range of 593–698 K,
MA selectivity was decreased slowly while butane
1
5 h activation, no C–H and C–O bands can be ob-
Table 1
The XPS results and the catalytic performances of the four VPO catalysts prepared in the absence (P1) and presence (P3–P5) of PEGs
Catalyst
XPS results
P/V ratio
1.6
Reaction data
Temperaturea (K)
658
V 2p3/2 (eV)
P 2p (eV)
133.9
O 1s (eV)
531.4
Cb (%)
Sc (mol%)
Yd (mol%)
P1
P3
P4
517.0
71
84
71
66
51
55
6
73
658
73
658
73
658
73
1.5
1.5
1.5
516.9
517.2
517.2
133.8
134.0
134.0
531.2
531.5
531.6
87
92
78
72
68
66
6
85
93
75
71
64
66
6
P5
94
97
72
69
68
67
6
a
Reaction temperature.
n-Butane conversion.
Selectivity for maleic anhydride.
Yield for maleic anhydride.
b
c
d

X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
265
conversion was increased drastically at elevating tem-
perature on the P2 sample. Even at low reaction tem-
perature of 593 K, catalyst P2 showed 43% conversion
and ∼90 mol% selectivity. At 658 K, 90% butane con-
version with 74 mol% MA selectivity can be achieved,
resulting in 67 mol% MA yield. Comparatively, only
non-selective for MA formation. Zazhigalov et al.
activated various promoted and non-promoted VPO
catalysts by means of “mechano-chemistry” [16]. It
was found that the applied process was beneficial for
the promoted samples while almost ineffective for the
non-promoted one. In another application of milling
process [17], Hutchings and Higgins studied the ef-
fects of milling on the physico-chemical properties
of VPO and observed positive effects in general. In
their case, however, more than one special additive
was usually used in the milling media, which in-
troduced the complexity in both the milling process
and the related chemistry. It is important to note that
not only the milling process itself but also the na-
ture of the precursor/sample for which the milling
was applied determined the effectiveness of milling
process. In this study, we did not directly fracture
the crystalline (VO)2P2O7, instead, ball-milled the
VOHPO4·0.5H2O precursor which was synthesized
through a controlled preparation procedure in the
organic media. Cyclohexane, with its non-polar char-
acter, would be a preferable choice as a milling sol-
vent. Since the VOHPO4·0.5H2O precursor derived
through a common preparation route generally had
5
1 mol% MA yield was obtained on catalyst P1 un-
der the same reaction conditions, with 71% conver-
sion and 71 mol% selectivity, respectively. Note that
the performance of catalyst P2 at 658 K is still notice-
ably better than that of P1 at 673 K.
The representative reaction data for the other PEG-
derived samples (P3–P5) as well as the compared
one (P1) were summarized in Table 1. The results
clearly demonstrated that the samples prepared in the
presence of PEGs showed considerably higher activ-
ity and selectivity than that derived in the absence of
PEG, and the former catalysts also exhibited better
low-temperature performance.
The application of PEGs seems to provide an effec-
tive and satisfactory medium to produce high surface
area, good performance VPO catalyst. It is thought that
the presence of PEG molecules can decrease the sur-
face energy of nucleation of VO(HPO4)·0.5H2O pre-
cursor and promotes the uniform precipitation of the
fine VO(HPO4)·0.5H2O particles. The PEG molecule
adsorbed on the precursor surface may also effectively
hinder the agglomeration of the precursor particles
formed. The gradual removal of PEG covering on the
precursor surface in the following activation procedure
is likely beneficial in generating the high surface area
2
low surface area (<10 m /g), in order to enhance the
effect of ball-milling, it is desirable to prepare a high
surface area VPO precursor. It was found that careful
controlling the speed of addition of phosphoric acid
into the hot reaction mixture was quite helpful for this
purpose. The as-synthesized precursor had a surface
2
area of ca. 15 m /g with the typical VOHPO4·0.5H2O
(VO)2P2O7. This kind of catalyst appears much more
phase, notably larger than some values reported [17].
Interestingly, when the precursor was dried at 393 K
in air and then directly applied for ball milling, the
increment of surface area of the activated sample was
quite limited. However, if the dried precursor was
subjected to a short preconditioning in N2 at 573 K
prior to ball milling, the surface area of the activated
active on the basis of per unit mass sample due to
much more active sites available, meanwhile it is also
more selective possibly due to the unique structure fea-
ture/morphology formed in the PEG medium, which
was partly revealed by XRD and IR investigations.
2
3
.2. The effects of precursor preparation,
sample, being in the range of 30–38 m /g, was con-
pre-treatment and ball milling
siderably larger than that of the VPO without such
a pretreatment, seen in Table 2. Although XRD ex-
amination revealed that shortly treating the precursor
in N2 at 573 K did not change the phase composition
of VOHPO4·0.5H2O, this thermal-treating process
together with the following ball milling procedure
apparently favored increment of surface area. Under
the milling conditions applied, it was found that the
surface area can be further enlarged by extending the
Okuhara et al. first modified the morphology of a
VPO catalyst by means of fracturing the plate form
crystalline (VO)2P2O7 in order to understand the
origin of active sites [15]. They found that the frac-
ture process did not notably alter the catalyst surface
area and catalytic performance, which made them
to conclude that the created new side faces were

2
66
X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
Table 2
The physical characteristics of the VPO precursors and the final VPO catalysts with or without milling
Sample
P-VPO (air/393 K)
P-VPO (N2/573 K
milling/10 h)
VPO
(milling/0 h)
VPO
(milling/5 h)
VPO
(milling/10 h)
2
BET surface areas (m /g)
XRD principal phase
15
30
20
30
38
VOHPO4·0.5H2O
VOHPO4·0.5H2O
(VO)2P2O7
1.5
26/74
6.3
(VO)2P2O7
1.4
15/85
6.1
(VO)2P2O7
1.4
15/85
6.0
XPS surface atomic P/V ratios
–
–
–
–
–
–
5
+
4+
XPS surface atomic V /V ratios
XPS surface atomic O/V ratios
variation in local structure could occur in the milled
samples. The TEM results (seen in Fig. 5) demon-
strated clearly the particle size/morphology changed
with extended milling time. XPS examination re-
vealed that the surface atomic P/V ratio of the milled
sample was ca. 1.5, almost identical to that of the
non-milled one. As mentioned above already, slight
enrichment of the P element on catalyst surfaces was
commonly observed on the VPO systems [14,18].
The binding energies of O 1s, V 2p3/2 and P 2p
of the milled samples are around 531.6, 517.2 and
134.1 eV, respectively, and close to those reported for
the (VO)2P2O7 phase [14]. The XPS curve-fitting re-
sults for the surface V4 and V species (according
+
5+
Fig. 4. The XRD patterns of the activated VPO samples milled
for different period of time.
4+
5+
to [19] the binding energies for V and V species
are assumed to be 516.9 and 518.0 eV, respectively),
however, indicated a notable difference in surface
state between the milled and non-milled samples (see
milling time to 10 h, but little increment was achieved
as the milling was extended beyond 10 h. Therefore,
the milling time for this study is generally 5–10 h.
XRD results indicated that the milled samples after
activation retained the typical phase composition of
the data presented in Table 2). The surface V5 /V
+
4+
ratio was remarkably decreased after 5 h-milling, but
essentially unchanged as the milling time was ex-
tended to 10 h, which was consistent with the trends
in the surface areas and the reaction data of the milled
samples.
The catalytic performances for the milled and
non-milled samples were presented in Table 3. On the
milled samples, even at lower reaction temperature
(VO)2P2O7 [3], seen in Fig. 4, suggesting the ball
milling process should not alter the essential phase
structure. On the other hand, the ball milling indeed
broadened the major diffraction peaks and also in-
creased the intensity of some minor diffraction peaks,
indicating both a reduction of particle size and a fine
Fig. 5. The TEM photographs of the non-milled and milled VPO samples: (a) non-milled, (b) milled for 5 h, (c) milled for 10 h.

X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
267
Table 3
The reaction performance of the milled and non-milled catalysts
Sample
VPO (0 h)
VPO (5 h)
VPO (10 h)
Reaction temperature (K)
Conversion (%)
658
51
673
63
698
83
658
74
673
85
698
95
658
74
673
84
698
93
Selectivity (mol%)
70
67
59
77
75
64
79
78
59
(
658 K), 77–79 mol% MA selectivity can be achieved
way of cleavage and accordingly, no unusually active
and/or selective planes were generated.
with the butane conversion of ca. 74%. MA selectivity
declined to ca. 62 mol% as the reaction temperature
was raised to 693 K at which the butane conversion
reached ca. 95%. The best performance, i.e. 78 mol%
MA selectivity with 84% butane conversion, was
achieved on the 10 h-milled sample at the optimal
reaction temperature of 673 K. Since the non-milled
sample only showed 67 mol% MA selectivity and
The enhancement of the milled catalyst perfor-
mance can be principally attributed to sliding of
precursor platelets away from one another, exposing
the certain surfaces that will further transfer into the
active and selective faces. This transformation was
known to be topotatic, and the morphology of the
catalyst precursor will control the morphology of the
final catalyst [20]. Besides the increment of surface
areas and available reaction sites, other factors may be
more or less relevant. Preliminary TPR investigation
6
3% butane conversion under the same reaction con-
ditions, it was clear that the combination of thermal
pre-treating the precursor and ball-milling process
did enhance the catalyst performance remarkably in
the temperature range of 658–673 K. At the highest
reaction temperature of 698 K, somewhat sintering of
the fine VPO particles likely occurred, which in turn
leaded to a slightly declined performance, seen on the
revealed that reduction of V5 and V species, cor-
responding to the shoulder at lower temperature and
the main peak at higher temperature, respectively in
the TPR profile, shifted to the lower temperature end
as the milling time was extended (seen in Fig. 6). It
seems that the milling treatment leads to a higher re-
+
4+
1
0 h-milled sample in Table 3.
In literature the P/V ratio of a VPO catalyst was
5
+
activity of the lattice oxygen associated with the V
4
+
thought to be one of the key factors determining the
catalytic performance [1], therefore, three precursors
with different bulk P/V ratio of 1.0, 1.2 and 1.5, re-
spectively, were prepared, and then followed the same
thermal treatment and 5 h ball milling. The typical
reaction data derived on these milled samples were
presented in Table 4. It was found that the best per-
formance was again obtained on the milled sample
with the bulk P/V ratio of 1.2, being consistent with
the observations made on the other non-milled VPO
catalysts [1]. It seems that the difference in bulk P/V
ratio did not bring about a significant change in the
and V species. Moreover, the amount of reducible
5+
V
species also declines with extended milling, be-
ing in agreement with the XPS curve-fitting results
mentioned above. Although a quantitative correlation
β
α
C
β
α
B
β
α
A
Table 4
The reaction performance of the milled samples with different
bulk P/V ratios
Sample
VPO
1.0 h)
VPO
(1.2 h)
VPO
(1.5 h)
100
300
500
700
Temp./ C
0
(
Reaction temperature (K)
Conversion (%)
Selectivity (mol%)
673
77
67
673
86
76
673
83
66
Fig. 6. The H2-TPR profiles of the non-milled and milled VPO
samples. Catalyst A: non-milled, catalysts B and C: milled for 5
and 10 h, respectively.

2
68
X. Wang et al. / Journal of Molecular Catalysis A: Chemical 206 (2003) 261–268
can hardly be established, it is clear that the applied
milling process not only effectively increased the sur-
face areas of the milled samples and consequently the
available number of active sites, but also somehow
changed the other characteristics such as the local en-
vironment and the redox behavior to a certain extent;
and both of them may account for the overall en-
hancement of performance on the ball milled samples.
References
[1] J. Ebner, V. Franchetti, G. Centi, F. Trifiro, Chem. Rev. 88
(1988) 251.
[2] G. Centi, Catal. Today 16 (1993) 5.
[3] B.K. Hodnett, Catal. Rev.-Sci. Eng. 27 (1985) 373.
[4] G. Centi, I. Manenit, A. Riva, F. Trifiro, Appl. Catal. 9 (1984)
177.
[5] P.T. Anastas, J.C. Warner, Green Chemistry: Theory and
Practice, Oxford University Press, Oxford, 1998.
[
6] R.M. Contractor, H.E. Bergna, H.S. Horowitz, C.M.
Blackstone, H. Chowdhry, A.W. Sleight, Catalysis 1987,
Elsevier, Amsterdam, 1988, p. 645.
4
. Concluding remarks
[
[
[
7] W. Ji, X. Wang, Q. Yan, Y. Chen, Chinese Patent,
CN1311058A.
8] G. Centi, F. Trifiro, J.R. Ebner, V.M. Franchetti, Chem. Rev.
A successful approach to preparing large surface
area VPO system by using PEGs as the unique ad-
ditive has been established and the resulting materi-
als showed notably enhanced performances. Another
interesting feature of the PEG-derived sample was
its better low-temperature performance. The principle
function of PEG in preparation chemistry and the pos-
sible causes for the enhanced performance were briefly
discussed. Another approach to preparing high surface
area VPO system, namely, well-controlled addition of
phosphoric acid to generate a high surface area VPO
precursor, short preconditioning of the precursor in ni-
trogen at certain temperature and final ball milling the
as-treated precursor, was also set up. The evaluation
of target reaction confirmed its effectiveness. In ad-
dition to the increment of surface areas and available
active sites, other factors induced by milling process
may also be responsible for the overall enhancement.
88 (1988) 55.
9] Catal. Today 16 (1) (1993) (the special issue for VPO
catalysts).
[10] G.J. Hutchings, Appl. Catal. 72 (1991) 1.
[11] W. Ji, L. Zeng, Y. Chen, Chinese Patent, CN1264619A;
Z. Ling, J. Wei-Jie, C. Yi, Chem. J. Chin. Univ. 20 (1999)
595.
12] W. Ji, L. Xu, A. Wang, Q. Yan, Y. Chen, Catal. Today 74
2002) 101.
[13] H. Igarashi, K. Tsuji, T. Okuhara, M. Misono, J. Phys. Chem.
97 (1993) 7065.
14] P. Delichere, K.E. Bere, M. Abon, Appl. Catal. A: Gen. 172
1998) 295.
15] T. Okuhara, K. Inumaru, M. Misono, in: S.T. Oyama,
J.W. Hightower (Eds.), Catalytic Selective Oxidation,
ACS Symposium Series 523, American Chemical Society,
Washington, DC, 1993, p. 156.
16] V.A. Zazhigalov, J. Haber, J. Stoch, L.V. Bogutskaya, I.V.
Bacherikova, Appl. Catal. A: Gen. 135 (1996) 155.
17] G.J. Hutchings, R. Higgins, Appl. Catal. A: Gen. 154 (1997)
1
[
(
[
(
[
[
[
[
[
[
103.
18] L.M. Cornaglia, C. Caspani, E.A. Lombardo, Appl. Catal. 74
(1991) 15.
19] M. Abon, K.E. Bere, A. Tuel, P. Delichere, J. Catal. 156
1995) 28.
20] J. Ziolkowiski, E. Bordes, P. Courtine, J. Catal. 122 (1990)
26.
Acknowledgements
(
The authors acknowledge the financial support
from the Ministry of Science and Technology of
China (Grant 2000048009) and the Natural Science
Foundation of Jiangsu Province (Grant BK2001201).
1