Preparation and heterogeneous photocatalytic activity of mesoporous H3PW12O40/ZrO2 composites

Article abstract of DOI:10.1016/j.molcata.2006.08.026

A series of mesoporous zirconia supported Keggin units with different H₃PW₁₂O₄₀ loading levels (5-20 wt.%) and phase structures (amorphous, tetragonal, and the mixed phases of tetragonal and monoclinic) are prepared by a modified wet impregnation method. The phase structures, chemical structures, optical absorption properties, surface physicochemical properties, and morphologies of the composites are well-characterized via X-ray diffraction (XRD) patterns, FT-IR spectroscopy, Raman scattering spectroscopy, UV-vis diffuse reflectance spectroscopy (UV-vis/DRS), nitrogen adsorption/desorption determination, and field emission scanning electron microscopy (FESEM), indicating that the primary Keggin structure remained intact after formation of the composites, and impregnation of the Keggin unit into the surface of ZrO₂ framework has an influence on the electronic properties of the ZrO₂. The heterogeneous photocatalytic activity of as-prepared composites is studied via the model reactions of degradation of an aqueous 4-nitrophenol (4-NP) and dye methylene blue (MB) under aerobic condition, indicating that the photocatalytic activity of the composites was influenced by the factors such as phase structures, optical absorption properties, and surface physicochemical properties of the composites.

Full text of DOI:10.1016/j.molcata.2006.08.026

Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
Preparation and heterogeneous photocatalytic activity of
mesoporous H3PW12O40/ZrO2 composites
Xuesong Qua, Yihang Guoa,∗, Changwen Hub,∗∗
a Faculty of Chemistry, Northeast Normal University, Changchun 130024, PR China
b Department of Chemistry, Beijing Institute of Technology, Beijing 100081, PR China
Available online 22 August 2006
Abstract
A series of mesoporous zirconia supported Keggin units with different H3PW12O40 loading levels (5–20 wt.%) and phase structures (amorphous,
tetragonal, and the mixed phases of tetragonal and monoclinic) are prepared by a modified wet impregnation method. The phase structures, chemical
structures, optical absorption properties, surface physicochemical properties, and morphologies of the composites are well-characterized via X-ray
diffraction (XRD) patterns, FT-IR spectroscopy, Raman scattering spectroscopy, UV–vis diffuse reflectance spectroscopy (UV–vis/DRS), nitrogen
adsorption/desorption determination, and field emission scanning electron microscopy (FESEM), indicating that the primary Keggin structure
remained intact after formation of the composites, and impregnation of the Keggin unit into the surface of ZrO2 framework has an influence on
the electronic properties of the ZrO2. The heterogeneous photocatalytic activity of as-prepared composites is studied via the model reactions of
degradation of an aqueous 4-nitrophenol (4-NP) and dye methylene blue (MB) under aerobic condition, indicating that the photocatalytic activity
of the composites was influenced by the factors such as phase structures, optical absorption properties, and surface physicochemical properties of
the composites.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Composites; Photocatalytic activity; Wet impregnation; Physicochemical properties
1. Introduction
In this paper, a series of novel solid Keggin-type POMs are
prepared by using ZrO2 as the support. ZrO2 is an important
material widely used in ceramics technology [13] and in het-
erogeneous catalysis [14,15]. In the field of catalysis, ZrO2 can
be served as a catalyst or catalyst support. In addition, as the
same as TiO2, ZrO2 is an n-type semiconductor oxide that has
also been tested for performing photocatalytic reactions [16,17].
The photocatalytic activity of ZrO2 has been successfully tested
although it has been well established that the photoactivity per
unit surface area is far less than that shown by TiO2 [16]. This
is due to high band gap energy of ZrO2 (5.0 eV corresponding
to an absorption edge to close to 250 nm [18]). It is expected
that impregnation of photoactive Keggin unit into the surface
of ZrO2 framework can improve its photocatalytic activity. So
far, the results concerning about addition of POM (mainly Keg-
gin unit) on the stabilized or hydrated ZrO2 support has been
reported [19,20], and their acid-catalyzed property has been
investigated. The studies pointed out that the chemical interac-
tions between the Keggin unit and the ZrO2 support are a matter
of great interest because a strong interaction could fix the Keg-
gin unit into the carrier, avoiding the leaching of the Keggin unit
in liquid-phase reactions [21,22]. Here, the H3PW12O40/ZrO2
Heterogeneous photocatalysis by solid polyoxometalate
(POM) is a new branch in the field of photocatalytic chemistry,
and some attractive researching progresses have developed in
recent years. Although POMs exhibit UV-light photocatalytic
activity in homogeneous system, the main disadvantages of them
as photocatalysts lie in their low surface area and difficulty of
separation from the reaction mixture. Therefore, development
of novel solid catalysts with advanced characteristics such as
surface area and porosity has been a challenge for a long time.
RecentstudiesshowedthatimmobilizationofsolublePOMunits
with different structures into the suitable supports (i.e., silica,
mesoporous molecular sieve MCM-41 or MCM-48, NaY zeo-
lite, activated carbons, layered double hydroxide, amorphous or
anatase TiO2) are an effective method to overcome the men-
tioned problems [1–12].
∗
∗∗
Corresponding author. Tel.: +86 431 5099762; fax: +86 431 5099762.
Corresponding author. Tel.: +86 10 62828869; fax: +86 10 62828869.
E-mail addresses: guoyh@nenu.edu.cn (Y. Guo), cwhu@bit.edu.cn (C. Hu).
1381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.08.026

X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
129
Table 1
composites with different H3PW12O40 doping levels and phase
structures are prepared by a modified wet impregnation method.
Considering the thermal stability of H3PW12O40 and avoiding
its decomposition during the calcination process, ZrO2 (rather
than H3PW12O40/ZrO2) supports with different phase struc-
tures are firstly prepared via calcination of the hydrous ZrO2 at
different temperatures. The H3PW12O40/ZrO2 composites are
produced by incorporation of the H3PW12O40 unit into the dif-
ferent ZrO2 supports. The phase structures, chemical structures,
optical absorption properties, surface physicochemical proper-
ties, and morphologies of the H3PW12O40/ZrO2 composites
are well-characterized via XRD patterns, FT-IR spectroscopy,
Raman scattering spectroscopy, UV–vis/DRS, nitrogen adsorp-
tion/desorption determination, and FESEM. The heterogeneous
photocatalytic activity of as-prepared composites with different
phase structuresis studied systematically via the modelreactions
of UV-light photodegradation of an aqueous 4-NP and MB.
Purpose of this study is to develop novel supported POM
photocatalytic materials, for example, ZrO2-supported Keggin
unit with different phase structures (amorphous, tetragonal, and
the mixture of tetragonal and monoclinic). The photocatalytic
activity of as-prepared materials is evaluated afterwards, and the
relationships between the heterogeneous photocatalytic activity
and their phase structures, optical absorption properties, surface
physicochemical properties, or H3PW12O40 doping levels are
studied.
H3PW12O40 loading level and characteristic IR absorption peaks of the
H3PW12O40 in the H3PW12O40/ZrO2 composites
Catalyst
H3PW12O40
loading level
(%)
Wavenumber (cm−1
)
P(Zr)-Oa
W
Od
W–O–W W–O–W
5 TZ-623
4.62
9.10
14.12
18.3
3.91
8.04
13.56
17.8
3.25
6.87
9.57
10.85
1069.2
1086.7
1081.3
1089.4
1083.1
1085.6
1084.5
1083.8
1080.4
1082.3
1081.7
1080.1
1103.8
1080.0
979.2
978.6
967.4
955.1
976.2
974.3
973.5
969.1
980.9
978.7
976.8
979.1
874.9
871.9
883.9
900.7
879.4
878.6
876.8
873.1
879.2
875.9
876.0
873.8
798.1
794.7
786.8
798.0
796.2
795.4
796.3
794.1
798.6
797.9
792.7
790.4
10 TZ-623
15 TZ-623
20 TZ-623
5 TZ-773
10 TZ-773
15 TZ-773
20 TZ-773
5 TZ-923
10 TZ-923
15 TZ-923
20 TZ-923
ZrO2-623
H3PW12O40
a
a
a
a
–
–
–
–
100
988.2
894.6
811.9
a
Not found.
2.2. Characterization
The loading levels of the H3PW12O40 in the final com-
posites are determined by a Leeman Prodigy inductively cou-
pled plasma atomic emission spectrometer (ICP-AES), and the
results are summarized in Table 1. XRD patterns of the samples
(pure powders) are obtained with a Rigaku D/max-3c X-Ray
diffractometer with Cu K␣ radiation. UV–vis/DRS (pure sam-
ple powder mixing with BaSO4 powder) are recorded on a Cary
500 UV–VIS-NIR spectrophotometer. FI-IR spectra are taken
in a Nicolet magna 560 IR spectrophotometer. FESEM images
are obtained on a XL-30 ESEM FEG scanning electron micro-
scope, and the samples are dispersed into absolute ethanol firstly.
Raman spectra are recorded on a Jobin-Yvon HR 800 instrument
withanAr+ lasersourceof488 nmwavelengthsinamacroscopic
configuration. BET specific surface area and BJH pore size are
calculated from nitrogen adsorption/desorption isotherms deter-
mined at −196 ◦C using an ASAP 2010 M surface analyzer (the
samples were outgassed under vacuum at 200 ◦C).
2. Experimental
2.1. Catalyst preparation
Zirconia supports with different phase structures are prepared
by sol–gel method, and the process is described below. The
mixture of zirconium n-butoxide (Zr(n-OC4H9)4, 21 mmol) and
ethanol (870 mmol) is stirred vigorously for 1 h at room temper-
ature. In addition, water (84 mmol) and HCl (2 mmol) are added
dropwise into the above Zr(n-OC4H9)4 solution, respectively.
The resulting mixture is stirred until the transparent Zr(OH)4
sol is obtained. After aging for 12 h, the gel particulate is dried
in air at 373 K for 24 h. Eventually, the product is calcined at
different temperatures (623, 773, and 923 K, respectively) for
3 h.
2.3. Photocatalytic testing
The H3PW12O40/ZrO2 composites are prepared by a mod-
ified wet impregnation method, and the process is described
below. Zirconia support (1 g) was suspended in the different
amountsofethanolsolutioncontaining2.5 mmolofH3PW12O40
in order to obtain a series of catalysts with different H3PW12O40
loading levels (i.e., 5, 10, 15, 20 wt.%). The above suspension is
stirred at room temperature for 12 h while reflux is used. Ethanol
is removed at 353 K. The product is dried at 373 K for 2 h, and
then it is washed for three times with hot water. The final prod-
uct is obtained by drying at 333 K for 24 h in air. Before the
photocatalytic test, the product is calcined at 473 K in a vacuum
oven for 2 h. As-prepared catalysts are represented by xTZ-t,
where x refers to weight percentage of H3PW12O40 (wt.%), T
H3PW12O40, Z zirconia, and t refers to calcination temperature
(K) of preparation of the ZrO2 support.
The photoreactor is designed with a cylindrical quartz cell
configuration and an internal light source surrounded by a quartz
jacket where the suspension of catalyst and aqueous organics
completelysurroundedthelightsource. Anexternalcoolingflow
of water was used to maintain the temperature of the system at
30 2 ◦C. UV-light energy is provided by a 50 W high pressure
mercury lamp (HPML).
The photocatalytic activity of as-prepared catalysts is tested
by the model reactions of photodecomposition of an aque-
ous 4-NP and dye MB (its structure is shown in Scheme 1).
In a typical reaction, the solid catalyst (0.1 g) is suspended
in an aqueous solution of 4-NP (Co = 0.36 mmol l−1, 100 ml)
or MB (Co = 0.065 mmol l−1, 100 ml), and then ultrasonicate
for 10 min to make the catalyst homogeneously disperses into
the reaction solution. Before UV-light irradiation, the suspen-

130
X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
Scheme 1. Structure of methylene blue (MB).
sion is stirred in the dark for 30 min to ensure establishment
of adsorption–desorption equilibrium. The suspension is vig-
orously stirred and the system opens to air during the whole
process. The concentration of 4-NP is monitored by a Shi-
madzu LC-8A high-pressure liquid chromatography (HPLC)
using UV detector at 210 nm. The concentration of MB is
analyzed by a Cary 500 UV–VIS-NIR spectrophotometer at
λmax = 664 nm. Before determination, the withdrawn suspen-
sions are centrifuged and filtered with the microporous mem-
brane.
3. Results and discussion
3.1. Catalyst preparation and characterizations
Hydrous zirconia support is prepared by a sol–gel method
via hydrolysis of the Zr(n-OC4H9)4 under acidic condition.
After calcination of the resulting hydrous ZrO2 at 373, 623,
773, and 923 K, respectively, zirconia with amorphous, tetrag-
onal, and the mixture of tetragonal and monoclinic phases
are produced (Fig. 1). Impregnation of the Keggin unit
into zirconia support is obtained via suspending the support
into the H3PW12O40–ethanol solution. During this process,
H3PW12O40/ZrO2 composites are formed via electrostatic inter-
actions and hydrogen bonding. The hydrogen bonding are
formed in the composites between the oxygen atoms of the
Keggin anion and the surface hydroxyl groups of the zirco-
nia network ( Zr–OH), which can be expressed in the forms
ofW Ot· · ·HO–Zr , W–Oc· · ·HO–Zr , andW–Oe· · ·HO–Zr ,
where Ot and Oc or Oe refers to the terminal oxygen atoms
and the bridge oxygen atoms, respectively, in the Keggin
unit [23,24]. These two interactions would ensure to fix the
H3PW12O40 unit into the ZrO2 support firmly, so that the leach-
ing of H3PW12O40 in liquid phase reactions may be avoided.
Fig. 1. XRD patterns of the ZrO2 support and H3PW12O40/ZrO2 composites
with different calcination temperatures.
support and the corresponding H3PW12O40/ZrO2 composites
(ZrO2-773, 20 TZ-773, and 5 TZ-773, Fig. 1b). The new
3.1.1. XRD patterns
◦
Phase structures of ZrO2 supports obtained by calcination of
hydrous ZrO2 at different temperatures and their corresponding
H3PW12O40/ZrO2 composites are analyzed by XRD (Fig. 1).
Both the ZrO2 support obtained via calcination at 373 K (ZrO2-
373) and its corresponding H3PW12O40/ZrO2 composite (15
TZ-373) are in the amorphous phase. Increase the calcination
temperature to 623 K, both the ZrO2 support (ZrO2-623) and
its corresponding H3PW12O40/ZrO2 composite (5 TZ-623)
are mainly in the tetragonal phase with the characteristic 2θ
values at 30.1◦ (1 1 1), 35.2◦ (2 0 0), 50.4◦ (2 2 0), and 59.9◦
(3 1 1), respectively (Fig. 1a, JCPDS file 17-0923). Continuous
increasing the calcination temperature to 773 K, the mixed
tetragonal and monoclinic phases are formed for both ZrO2
diffraction peaks locate at 24.0◦ (0 1 1◦), 28.2◦ (1 1 1), 31.4
¯
◦
¯
◦
◦
¯
(1 1 1), 3◦4.3 (0 0 2), 40.7 (2 1 1), 44.5 (1 1 2), 55.5 (0 1 3),
and 60.1 (3 0 2), respectively, corresponds to the diffractions
of monoclinic ZrO2 phase (JCPDS file 86-1450). As the
ZrO2 support is calcined to 923 K, the mixed tetragonal and
monoclinic phases are also formed. The crystallinity of the
composites (5 TZ-923 and 15 TZ-923) increases compared with
the TZ-773 samples (Fig. 1c). In addition, no indication of any
crystalline phase related to H3PW12O40 appears even though
the loading level of the H3PW12O40 reaches to 20 wt.%. This
result implies that the Keggin unit homogeneously disperses
into the surface of ZO2 framework, which will be benefit to
enhance the catalytic activity of the composites. On the other

X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
131
to saturate in the composite after H3PW12O40 loading level is
up to ca. 10 wt.%. With the increase of the H3PW12O40 load-
ing level, the excessive H3PW12O40 physically adsorbs on the
surface of the catalyst, at least partially, resulting in partial seg-
regation of the Keggin unit from the ZrO2 support. As for the 15
TZ-373, its OMCT (233 nm) band is shorter than that of 15 TZ-
623 or 20 TZ-773, suggesting that the different phase structures
have an effect on their optical absorption properties. By using
the same ZrO2 support (ZrO2-623) but different H3PW12O40
doping levels, the positions and shapes of the OMCT bands are
very similar, implying that different H3PW12O40 doping lev-
els have little effect on the optical absorption properties of the
composites (Fig. 2b).
The above UV–vis/DRS results indicate that impregnation of
the Keggin unit into the ZrO2 framework (ZrO2-623 and ZrO2-
773) had an influence on the electronic properties of the ZrO2,
resulting in red shift of the OMCT band. This is due to forma-
tion of a doping energy level (or moderate widening of electronic
band) between the conduction (Zr 4d) and valence bands (O 2p)
of ZrO2. The electronic transitions from the valence band (O
2p) to dopant level or from the dopant level to the conduction
band (W 5d + Zr 4d) can effectively red shift the band edge
absorption threshold. This optical absorption property will be
beneficial to improve the photocatalytic activity of ZrO2 since
the transition from the valence band to the dopant level or from
the dopant level to the conduction band become easier, resulting
in more photogenerated electrons (e−) and holes (h+) to par-
ticipate the photocatalytic reactions. In addition, the band gap
of the crystalline ZrO2 is narrower than that of the amorphous
one.
Fig. 2. (a) UV/DR spectra of 15 TZ-t composites with the different ZrO2 calcina-
tion temperature, where t = 373, 623, 773, and 923 K, respectively. (b) UV/DR
spectra of the xTZ-623 composites with different H3PW12O40 doping levels,
where x = 5, 10, 15, and 20, respectively.
3.1.3. FT-IR and Raman scattering spectroscopies
The integrity of the Keggin structure in as-prepared com-
posites is confirmed by FT-IR (Table 1 and Fig. 3) and Raman
scattering spectroscopies (Fig. 4). That is, the composites exhib-
ited four characteristic IR absorption peaks in the range from
700 to 1100 cm−1, corresponding to the stretching vibrations of
P–O, W O, and W–Oc/e–W bonds of the Keggin unit, respec-
tively [25]. Specifically, the peak appears at ca.1080 cm−1 is
originated from the vibration of P–O bonds, the peak appears
hand, it can infer that different H3PW12O40 loading levels have
little effect on the ZrO2 phase structures.
3.1.2. Optical absorption properties
The optical absorption properties of the ZrO2 supports and
the corresponding H3PW12O40/ZrO2 composites are studied
by UV–vis/DR spectroscopy (Fig. 2). At first, with the same
H3PW12O40 doping level (15 wt.%) but different phase struc-
tures, the electronic spectra of the H3PW12O40/ZrO2 are obvi-
ously different (Fig. 2a). That is, O 2p to W 5d charge transfer
(CT) bands of the starting H3PW12O40 appear at ca. 200 (W Ot
bond) and 260 nm (W–Oc/e–W bond), respectively. O 2p to Zr
4d CT band of ZrO2-623 appears at 237 nm. After formation of
the crystalline H3PW12O40/ZrO2 composites (i.e., 15 TZ-623
and 20 TZ-773), the original O 2p to W 5d or O 2p to Zr 4d CT
bands disappear, whereas, a broad absorption peak appear with
the red shift of the CT band compared with ZrO2 support or the
starting Keggin unit. However, as for the 15 TZ-923, two solitary
absorption peaks corresponding to O 2p to W 5d and O 2p to Zr
4d charge transfer appear again. This is due to the surface prop-
erties of the ZrO2 support. When the ZrO2 support is calcined at
923 K, both the amount of hydroxyl groups of the ZrO2 and its
specific surface area decreased, accordingly, H3PW12O40 tends
Fig. 3. FT-IR spectra of the starting H3PW12O40 and 15 TZ-623.

132
X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
Fig. 4. Raman scattering spectra of H3PW12O40, ZrO2-623, and 15 TZ-623.
at ca. 980 cm−1 corresponds to the vibrations of W O bonds,
and two peaks appear at ca. 880 and 800 cm−1 were due to
the vibrations of the two types of W–Oc/e–W bridge bonds.
Compared with the parent Keggin unit, both shifts of the IR
peak positions and broadening of the peak shapes are ascribed
to the interactions between the Keggin unit and the ZrO2
support.
Raman scattering studies further support the above result.
That is, three Raman scattering peaks appear at 1001, 977, and,
926 cm−1 are observed in the starting H3PW12O40 unit, cor-
responding to stretching vibrations of P–O bond of the PO4
sites, W Ot and W–Oc/e–W bonds, respectively [8]. As for
the H3PW12O40/ZrO2 composites, only two Raman scattering
peaks at 997 and 966 cm−1 are detected, and their peak intensi-
ties decrease compared with those of the starting H3PW12O40.
´
The similar results have been reported by Lopez-Salinas [26].
The red shift of the peak positions and weakening of the peak
intensity in comparison with the starting Keggin unit may be
due to formation of some W–O–Zr bond on the surface of ZrO2
support [27]. This weakening of the W Ot bond may also due
to hydrogen bonding between the W Ot bonds and surface
Zr–OH groups. The peak corresponding to W–Oc/e–W bond
disappears in the H3PW12O40/ZrO2 composite, and the reason
is unclear in current study. In addition, the weak peaks located
at 200–700 cm−1 are the characteristic of tetragonal phase of
zirconia, originating from the vibrations of Zr–O or Zr–O–Zr
bonds [26].
Fig. 5. N2 adsorption/desorption isotherms and pore size distribution curves
(insert) of as-prepared catalysts. (a) ZrO2-623 K, (b) 15 TZ-623 K and (c) 20
TZ-773 K.
p/p0 = 0.4–0.6 (Fig. 5b). However, the BET surface area, pore
diameter, and pore volume decrease, respectively, compared
with the ZrO2 support. As for the 20 TZ-773, its BET surface
area further decrease, at the same time, the volume adsorbed
3.1.4. Surface physicochemical properties
Table 2
The surface physicochemical properties of as-prepared ZrO2
and corresponding H3PW12O40/ZrO2 composites are charac-
terized by N2 adsorption/desorption analyses, and the represen-
tative N2 adsorption/desorption isotherms and BJH pore size
distribution curves are shown in Fig. 5. The calculated BET
specific surface area, pore size, and pore volume are summa-
rized in Table 2. The ZrO2 support calcined at 623 K (ZrO2-623)
exhibits typical mesoporosity according to IUPAC definition,
and its isotherm is type IV with the hysteresis loop appearing at
anintermediaterelativepressure(p/p0 = 0.4to0.6, Fig. 5a). After
grafting the Keggin unit into the surface of ZrO2 framework,
the isotherm still is type IV with the hysteresis loop showing at
Textural properties of the ZrO2 supports and corresponding H3PW12O40/ZrO2
composites
Sample
SAa (m2 g−1
)
PDb (nm)
PVc (cm3 g−1
)
ZrO2-623
15 TZ-623
ZrO2-773
20 TZ-773
ZrO2-923
10 TZ-923
111.6
69.4
54.3
39.4
29.8
21.0
3.4
2.0
8.2
6.9
8.6
6.2
0.26
0.09
0.29
0.11
0.12
0.08
a
BET surface area.
Average pore diameter.
Pore volume.
b
c

X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
133
Fig. 7. Photodegradation of an aqueous 4-NP (0.36 mmol l−1, 100 ml) by UV-
light irradiation of ZrO2 (0.1 g) with different phase structures.
3.2. Photocatalytic testing
3.2.1. UV-light photocatalytic activity of the ZrO2 supports
It is well known that the photoactivity of the catalyst is
influenced by its crystal phase structures. For example, the
photoactivity of TiO2 in the anatase phase is higher than that of
other phases. However, little work has been done on studies of
the photocatalytic behavior of the ZrO2 in the different phase
structures. Here, 4-NP is chosen as one of the model reactant
to investigate the relationships between the heterogeneous
photocatalytic activity and the ZrO2 phase structures. 4-NP is an
intermediate product that involves in the process of degradation
of pesticides and hardly removal because of its high stability.
Photodegradation of an aqueous 4-NP (Co = 0.36 mmol l−1
,
100 ml) in the absence of the photocatalyst is neglectable
because there is no any change of its concentration after 4 h
UV-light irradiation. Fig. 7 shows change of 4-NP conversion
under UV-light irradiation of the different ZrO2 supports. It can
be clearly seen that the concentration of the 4-NP decreases
in some extents by using different ZrO2 supports. ZrO2-773
exhibits the highest photocatalytic activity, while ZrO2-373
exhibits the lowest photocatalytic activity among the four tested
ZrO2 supports. The above results indicate that the four ZrO2
supports are photoactive, and the crystalline ZrO2 are more
photoactive than that of the amorphous one. As for the ZrO2-
923, it has the similar phase structure with that of the ZrO2-773,
however, its photocatalytic activity is relatively low. The reason
may be due to its lower BET surface area. In subsequent studies,
only ZrO2-773 is selected as the support of H3PW12O40.
Fig. 6. FESEM images of the H3PW12O40/ZrO2 composites. (a) 15 TZ-623 K,
(b) 15 TZ-773 K, and (c) 15 TZ-923 K.
decrease obviously. However, the isotherm of 20 TZ-773 is
still type IV with the hysteresis loop showing at p/p0 = 0.4–0.8
(Fig. 5c). The above results indicate that the BET surface areas
of the H3PW12O40/ZrO2 composites are lower than that of the
ZrO2 support. This is due to the preparation method used here.
After grafting the large Keggin unit on the surface of the ZrO2
support, some of the pores of ZrO2 are blocked up with the
decrease of its BET surface area and pore volume. In addition,
with the increase of the calcination temperature, the BET sur-
face area of the corresponding composites decreases owing to
the crystallinity of the product being increased.
3.2.2. UV-light photocatalytic activity of the
H3PW12O40/ZrO2 composites
For the sake of enhancement of the photocatalytic activ-
ity of ZrO2 and H3PW12O40, the TZ-773 composites with
different H3PW12O40 doping levels are prepared, and their
photocatalytic activity is evaluated by degradation of aqueous
4-NP (Fig. 8) and MB (Fig. 9). Homogeneous H3PW12O40 and
zirconia support are also tested for comparison under the same
conditions. Homogeneous H3PW12O40 exhibits low photocat-
alytic activity to 4-NP degradation, however, obvious 4-NP
3.1.5. Morphology
FESEM images of the composites are shown in Fig. 6. It is
observed that the particles are spherical and exhibited uniform
distribution with an average size of 25–30 nm, at the same time,
aggregation among particles occurred.

134
X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
tion, the photocatalytic activity of the starting H3PW12O40 or
ZrO2-773 was low to MB degradation. However, significant MB
degradation is observed by UV-light irradiation of the TZ-773
composites, and the 15 TZ-773 exhibits the highest photocat-
alytic activity among the six tested photocatalytic materials. The
above results are in agreement with those of the 4-NP degrada-
tion, further confirming the enhanced photocatalytic activity by
impregnation of the Keggin unit into the ZrO2 support.
It must be emphasized that the concentration of 4-NP or
MB has decreased before UV-light irradiation, indicating that
adsorption of the ZrO2 or H3PW12O40/ZrO2 to the 4-NP or MB
molecule occurred. This adsorption results in more easily degra-
dation of the 4-NP or MB molecules. In a heterogeneous pho-
tocatalysis system, photo-induced molecular transformations or
reactions take place at the surface of a catalyst, and the adsorp-
tion of the reactant molecules on the surface of the catalyst is the
first step of photocatalytic reaction [30]. Difference of adsorp-
tion ability between the ZrO2 and H3PW12O40/ZrO2 may be
explained below. The hydroxyl group of the zirconia can’t be
protonated or deprotonated at pH 6–7 [31], therefore, 4-NP or
MB is difficult to be adsorbed on the surface of the ZrO2. In
the case of the H3PW12O40/ZrO2 composites, 4-NP or MB
molecule may adsorbed on their surface by hydrogen bonds
formed between the catalyst and the reactant molecules, which
may be an essential premise for the photocatalytic degradation
of 4-NP or MB.
On the basis of the above experimental results and the charac-
teristics of as-prepared H3PW12O40/ZrO2 composites in phase
structures, surface physicochemical properties, and electronic
attributes, we infer that the following two factors could con-
tribute to this increased photocatalytic efficiency compared with
that of the ZrO2 or H3PW12O40. On the one hand, relatively
narrow band gap of the composites and increase the amount of
photochemistry at longer wavelengths play an important role to
this enhanced photocatalytic activity. Therefore, electron tran-
sition from the valence band to the dopant level or from the
dopant level to the conduction band become easier, resulting in
more photogenerated electrons (e−) and holes (h+) to partici-
pate the photocatalytic reactions. Higher photocatalytic activity
of the crystalline ZrO2 than that of the amorphous ZrO2 is also
due to this fact. On the other hand, the surface physicochemi-
cal properties of the mesoporous materials also play significant
role to this enhanced photocatalytic activity. Compared with the
starting POM, as-prepared composites show relatively high BET
surface area. Therefore, more contact areas between the catalyst
and the substrate are provided for the surface-mediated electron-
transfer reactions to take place. In addition, the porous catalysts
show a strong adsorption to the substrates, which is beneficial for
the photocatalytic reactions. At the same time, more substrates
are transported to catalytic active sites, thus the higher turnover
number was obtained (Table 3).
Fig. 8. Photodegradation of an aqueous 4-NP (0.36 mmol l−1, 100 ml) by UV-
light irradiation of the different H3PW12O40/ZrO2 composites (0.1 g).
degradation is observed by using the H3PW12O40/ZrO2 as the
photocatalysts. Moreover, the photocatalytic activity improves
with increasing H3PW12O40 loading levels from 5 to 15 wt.%.
However, once the H3PW12O40 loading level is up to 20 wt.%,
the photocatalytic activity decreases obviously. This may be
due to the fact that the Keggin unit homogeneously disperses
into the framework of ZrO2 if the H3PW12O40 loading levels
are less than 15 wt.% [28,29]. This homogeneous dispersion
of the Keggin unit into the ZrO2 framework plays important
role to improve the photocatalytic activity of the composites.
When the H3PW12O40 loading level is higher than 20%, partial
segregation of the Keggin unit from the ZrO2 support occurs,
accordingly, the photocatalytic activity decreases.
In order to further test the photocatalytic activity of as-
prepared composites, MB is selected as another model reactant.
MB is a heteropolyaromatic dye with large and complicated
structure (Scheme 1) that is difficult to be degraded. All the
experimental conditions used are same with those of 4-NP
degradations. Fig. 9 shows change of MB conversion by
UV-light irradiating the suspension including an aqueous MB
(0.065 mmol l−1, 100 ml) and the TZ-773 composites (0.1 g)
with different H3PW12O40 loading levels. MB direct photolysis
results shows that disappearance of MB is negligible. In addi-
After the reactions finished, the catalysts are separated by
centrifugation, and the clear solutions are analyzed by UV–vis
spectrophotometer and ICP-AES. There is no light absorption
in the range from 190 to 300 nm for these solutions. In addition,
detected concentration of W in these solutions is in the range
from 1.1 to 3.3 mg l−1 (the initial concentration of W is in the
Fig. 9. Photodegradation of an aqueous MB (0.065 mmol l−1, 100 ml) by UV-
light irradiation of the different H3PW12O40/ZrO2 composites (0.1 g).

X. Qu et al. / Journal of Molecular Catalysis A: Chemical 262 (2007) 128–135
135
Table 3
90406002), and Specialized Research Fund for the Doctoral
Program of Higher Education of China (20030007014). This
work is also supported by the Program of New Century
Excellent Talents in University (NCET-04-0311).
Photocatalytic activity of H3PW12O40/ZrO2 in the photodegradation of aqueous
4-NP and MBa
Catalyst
TON for degradation
of 4-NPb
TON for degradation
of MB
References
5 TZ-773
10 TZ-773
15 TZ-773
20 TZ-773
14.6
9.5
6.9
4.1
1.5
1.5
1.3
1.4
[1] A. Molinari, R. Amadelli, L. Andreotti, A. Maldotti, J. Chem. Soc. Dalton
Trans. (1999) 1203.
[2] R.R. Ozer, J.L. Ferry, Environ. Sci. Technol. 35 (2001) 3242.
[3] R.R. Ozer, J.L. Ferry, J. Phys. Chem. B 106 (2002) 4336.
[4] B. Yue, Y. Zhou, J. Xu, Z. Wu, X. Zhang, Y. Zou, S. Jin, Environ. Sci.
Technol. 36 (2002) 1325.
[5] Y. Guo, C. Hu, X. Wang, E. Wang, Y. Zhou, S. Feng, Chem. Mater. 13
(2001) 4058.
a
Initial concentration of 4-NP 0.36 mmol l−1; initial concentration of MB
0.065 mmol l−1; the system opened to air; the temperature of the suspension
30 2 ◦C; catalyst 0.1 g; UV light (50 W HPML) irradiation time 90 min.
b
TON (turnover number) is based on mmol of 4-NP or MB consumed/mmol
of active H3PW12O40 site.
[6] Y. Guo, C. Hu, S. Jiang, C. Guo, Y. Yang, E. Wang, Appl. Catal. B 36
(2002) 9.
[7] Y. Guo, C. Hu, C. Jiang, Y. Yang, S. Jiang, X. Li, E. Wang, J. Catal. 217
(2003) 141.
[8] L. Li, Q. Wu, Y. Guo, C. Hu, Micropor. Mesopor. Mater. 87 (2005) 1.
[9] Y. Yang, Q. Wu, Y. Guo, C. Hu, E. Wang, J. Mol. Catal. A 225 (2005) 203.
[10] Y. Yang, Y. Guo, C. Hu, Y. Wang, E. Wang, Appl. Catal. A 273 (2004) 201.
[11] D. Li, Y. Guo, C. Hu, L. Mao, Appl. Catal. A 235 (2002) 11.
[12] D. Li, Y. Guo, C. Hu, J. Mol. Catal. A 207 (2004) 181.
[13] D.A. Ward, E.I. Ko, Chem. Mater. 5 (1993) 956.
[14] H.H. Kung, Stud. Surf. Sci. Catal. 45 (1989) 721.
[15] A. Corma, Chem. Rev. 95 (1995) 559.
range from 36.7 to 110.2 mg l−1), therefore, the leakage of the
Keggin unit from the support is neglectable. The separated cat-
alyst is used for another catalytic run, and the photoactivity is
almost unchangeable for three times’ run, suggesting that the
strong interactions between the Keggin unit and ZrO2 support
ensure the little leaching of the Keggin unit from the support.
4. Conclusions
[16] J.-M. Herrman, J. Disdier, P. Pichat, J. Chem. Soc. Faraday Trans. 77 (1981)
67.
[17] J.A. Nav´ıo, G. Colo´n, Stud. Surf. Sci. Catal. 82 (1994) 325.
[18] A. Nav´ıo, M.C. Hidalgo, G. Colo´n, S.G. Botta, M.I. Litter, Langmuir 17
(2001) 202.
Novel solid Keggin-type POM, H3PW12O40/ZrO2 with dif-
ferent H3PW12O40 loading levels and phase structures, are pre-
paredbythemodifiedwetimpregnationmethod. Thecomposites
exhibit mesoporosity with average pore diameter in the range of
2.0–6.9 nm, and their BET surface area is higher than that of
the starting H3PW12O40 but lower than the corresponding ZrO2
support. Impregnation of the Keggin unit into the ZrO2 frame-
work has an influence on the electronic properties of the ZrO2,
namely, the H3PW12O40/ZrO2 composites prepared by using the
ZrO2-623 (tetragonal phase) and ZrO2-773 (the mixed phases
of tetragonal and monoclinic) as the supports exhibit red shifted
OMCT band, corresponding to relatively narrow bandgap.
Heterogeneous photocatalytic tests show that the increased
photocatalytic activity compared with homogeneous
H3PW12O40 or ZrO2 support is obtained with H3PW12O40
loading levels not higher than 15 wt.%, mainly resulting
from the increase the amount of photochemistry at longer
wavelengths, phase structures, and surface physicochemical
properties of the mesoporous materials.
[19] K. Saito, M. Nomura, S. Ohgoshi, Y. Akai, Prep. Symp. Am. Chem. Soc.
Div. Petro. Chem. 42 (1997) 712.
[20] E. Lo´pez-Salinas, J.G. Herna´ndez-Co´rtez, M. Corte´z, J. Navarrete, M.
Yanos, A. Va´zquez, H. Armdaris, T. Lo´pez, Appl. Catal. 175 (1998) 43.
[21] Y. Izumi, K. Urabe, Chem. Lett. (1981) 663.
[22] M.A. Schwegler, H. Van Bekkum, N.A. De Munck, Appl. Catal. 74 (1991)
191.
[23] M. Valigi, D. Gazzoli, G. Ferraris, E. Bemporad, Phys. Chem. Chem. Phys.
21 (2003) 4974.
[24] J.B. Moffat, in: B. Imelik, et al. (Eds.), Catalysis by Acids and Bases,
Elsevier, Amsterdam, 1985, p. 157.
[25] C. Rocchiccioli-Deltcheff, M. Fournier, R. Franck, R. Thouvenot, Inorg.
Chem. 22 (1983) 207.
[26] E. Lo´pez-Salinas, J.G. Herna´ndez-Co´rtez, I. Schifter, E. Torres-Garc, T.
Lo´pez, Appl. Catal. A 193 (2000) 215;
J. Navarrete, A. Gutie´rrez, T. Lo´pez, P.P. Lottici, D. Bersani, Appl. Catal.
A 193 (2000) 215.
[27] C. Wu, X. Zhao, Y. Ren, Y. Yue, W. Hua, Y. Cao, Y. Tang, Z. Gao, J. Mol.
Catal. A 229 (2005), p. 229, 233.
[28] B.M. Devassy, S.B. Halligudi, S.G. Hegde, A.B. Halgeri, Chem. Commun.
10 (2002) 1074–1075.
Acknowledgments
[29] B.M. Devassy, F. Lefebvre, J. Catal. 231 (2005) 1.
[30] A.L. Linsebigler, G. Lu, J.T. Yates, Chem. Rev. 95 (1995) 735.
[31] A.A. Parks, Chem. Rev. 65 (1965) 177.
The Natural Science Found Council of China is acknowl-
edged for financial support (NSFC 20331010, 20271007, and