You are on page 1of 9

Microporous and Mesoporous Materials 87 (2005) 1–9

www.elsevier.com/locate/micromeso

Nanosize and bimodal porous polyoxotungstate–anatase


TiO2 composites: Preparation and photocatalytic degradation
of organophosphorus pesticide using visible-light excitation
a,b c,* a,* d
Li Li , Qing-yin Wu , Yi-hang Guo , Chang-wen Hu
a
Faculty of Chemistry, Northeast Normal University, Changchun 130024, PR China
b
Faculty of Chemistry and Chemical Engineering, Qiqihar University, Qiqihar 161006, PR China
c
Department of Chemistry, Zhejiang University, Hangzhou 310027, PR China
d
Department of Chemistry, Beijing Institute of Technology, Beijing 100081, PR China

Received 4 December 2004; received in revised form 3 March 2005; accepted 25 July 2005
Available online 16 September 2005

Abstract

New and efficient heterogeneous photocatalytic materials, H3PW12O40/TiO2 and H6P2W18O62/TiO2, with anatase crystalline
phase, dual-pore structures (micro- and meso-porosity), and nanometer sizes (average particle size of 6 nm) were prepared by com-
bination of the methods of sol–gel and hydrothermal treatment at a lower temperature (200 °C, 2 °C/min). Compositions, struc-
tures, physicochemical properties, and surface morphology of the composites were characterized by inductively coupled plasma
atomic emission spectrum (ICP-AES), X-ray diffraction (XRD) patterns, UV diffuse reflectance spectroscopy (UV/DRS), 31P
magic-angle spinning (MAS) NMR, Raman scattering spectra, nitrogen adsorption/desorption determination, and transmission
electron microscopy (TEM), respectively. It indicated that the primary Keggin or Dawson structure remained intact after immobi-
lization of the unit into TiO2 network, and hydrogen bonding and chemical interactions between the unit and the anatase network
existed in the composites. Both of the composites exhibited visible-light photocatalytic activity to degradation of an aqueous orga-
nophosphorus pesticide, parathion-methyl, and the deactivation of the catalysts was hardly observed after five times catalytic
cycling.
Ó 2005 Elsevier Inc. All rights reserved.

Keywords: Hydrothermal treatment; Porous materials; Polyoxometalate; Sol–gel; Visible-light photocatalysis

1. Introduction vides some opportunities for them. Anatase TiO2 is the


most widely investigated heterogeneous semiconductor
Heterogeneous photocatalytic destruction of organic photocatalyst [6–8], while polyoxometalates (POMs)
pollutants in wastewater by using solar light (or light have been studied as a homogeneous photocatalyst
from the visible region of the spectrum) as the excitation [9–13]. Both of TiO2 and POMs share the very similar
energy is an appealing field [1–5]. To meet the need of photocatalytic processes owing to their similar elec-
this, it is important to design and prepare efficient tronic attribute. However, they must be activated in
photocatalytic materials, which may then appear to be the near-UV area due to their wide band gap. In addi-
a challenge to the chemists, at the same time, it also pro- tion, for the purpose of practical application of this
technique, immobilization of soluble POM into some
*
Corresponding authors. Tel.: +86 431 5099371; fax: +86 431
matrixes to prepare solid or supported POM photocata-
5692259. lysts is essential. Our previous work have prepared sev-
E-mail address: guoyh@nenu.edu.cn (Y.-h. Guo). eral water-insoluble POMs by immobilizing them into

1387-1811/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.micromeso.2005.07.035
2 L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9

silica matrix via soft chemical preparation techniques than 10 m2/g, [23]). More importantly, the interactions
including sol–gel, templating, intercalating, and hydro- between the Keggin (or Wells–Dawson) unit and TiO2
thermal treatment [14–18]. These insoluble POM/SiO2 support is chemical rather than physical, which ensure
composites had different pore sizes (i.e., micro-, meso- little leakage of them from TiO2 matrix during the pro-
and macro-pores), and they exhibited much higher cess of catalysis, therefore, the composites can be reused
UV-photocatalytic activity than that of the correspond- more than five times without finding obvious decreases
ing pure or homogeneous POMs. Moreover, it is easy to of their photocatalytic activity. The above described cat-
separate and recover them for recycling purpose. But the alytic result was obtained via the test of visible-light
main disadvantage of these POM/SiO2 composites is photocatalytic degradation of an aqueous parathion-
that they show little activity under visible-light irradia- methyl, a kind of stable organophosphorus pesticide.
tion. In order to improve the photoefficiency of either Parathion-methyl had only absorption at 280 nm, and
POM or TiO2, and immobilization of soluble POM, it did not suffer any degradation with 4 h direct photol-
we here prepared two POM-containing anatase TiO2 ysis under visible-light irradiation. However, its signifi-
composites (H3PW12O40/TiO2 and H6PW18O62/TiO2) cant degradation was obtained under visible-light
via combining the methods of sol–gel and hydrothermal irradiation, the suspension of an aqueous parathion-
treatment with a constant heating rate (2 °C/min to methyl and the POM/TiO2, whereas, the starting
200 °C). The motivation for this design is based on the H3PW12O40 or H6P2W18O62, and commercial available
structural characteristics of either TiO2 or POM. That Degussa P 25 exhibited little visible-light photocatalytic
is, POM has empty d orbits and can be used as good activity to degradation of it. As for the nanosize anatase
electron acceptors. After the addition of POM into TiO2 (average size lower than 10 nm) prepared in the
TiO2 photocatalytic systems, the fast charge-pair (h+– present work, only substrate adsorption rather than deg-
e) recombination on TiO2 can be retarded effectively. radation was observed during the process of visible-light
That is, POM has the ability to enhance the rate of con- irradiation of it.
duction band (CB) electron transfer by accepting e
(photogenerated electrons) to its empty d orbits. Several
research groups have studied this synergistic effect ex- 2. Experimental
isted between TiO2 and POM. For example, Ferry
reported that significant rate enhancement for 1,2- 2.1. Photocatalyst preparation and characterization
dichlorobenzene photooxidation was obtained after
addition of aqueous photoactive POM anions including A typical method of preparation of H3PW12O40/TiO2
PW12 O3 4 4
40 , SiW12 O40 , or W10 O32 into TiO2 suspensions or H6P2W18O62/TiO2 composites is described below.
[19]. Choi reported that the addition of homogenous Titanium tetraisopropoxide (TTIP, 98%, 20 mmol) was
SiW12 O440 into the naked TiO2 suspension greatly in- dissolved with iso-propyl alcohol (392 mmol) while stir-
creased the photocatalytic oxidation rate of As(III) to ring was used. In another container, H3PW12O40 or
As(V), which is less toxic and less mobile in the aquatic H6P2W18O62 (0.13 mmol) was dissolved with water
environment [20]. And they pointed out that the pres- (44 mmol), and then added into the TTIP solution drop
ence of POMs enhances the generation of superoxides, by drop. The resulting mixture was adjusted to pH 1–2
which are the main oxidant of As(III). Choi further by addition of 8 mol/l HCl, and it was stirred at room
proved this synergistic effect occurred between POM temperature for 1 h, and then heated to 45 °C until
and TiO2 through photoelectrochemical investigation homogeneous H3PW12O40/TiO2 or H6P2W18O62/TiO2
[21]. They concluded that POMs successfully transfer hydrogel was formed. This hydrogel was transferred
CB electrons on TiO2 particles to an inert collector elec- into an autoclave, and then heated to 200 °C with a rate
trode with generating photocurrents under UV illumina- of 2 °C/min. Finally, the temperature was kept at 200 °C
tion. The magnitude of photocurrent is directly for 1 h. This hydrogel was cooled to room temperature
proportional to the reduced POM concentration and and dehydrated slowly at 50 °C in a vacuum for 24 h.
markedly decreases in the presence of dissolved O2 due The dried gel was washed with hot water and dried at
to the rapid reoxidation of the reduced POM by O2. room temperature.
However, in the above TiO2–POM systems, POMs were Elemental analysis for H3PW12O40/TiO2: P, 0.21%;
in the homogeneous phases, they dissolved in the reac- W, 14.4%; H3PW12O40, 19.61%. Elemental analysis for
tion systems, thus separating them for recycling applica- H6P2W18O62/TiO2: P, 0.31%; W, 16.28%; H6P2W18O62,
tions is very difficult. Here, as-prepared POM/TiO2 21.52%. The above data were determined with a Leeman
composites are insoluble and they have the characteris- Plasma Spec (I) ICP-AES.
tics such as narrow band gap, dual-pore structures XRD patterns were obtained with a Rigaku D/max
(micro- and meso-porosity), nanometer size (<10 nm), 2000 X-ray diffractometer with Cu-Ka radiation. UV/
and much larger BET surface area (BET surface area DRS were recorded on a Cary 500 UV–VIS–NIR spec-
for anatase TiO2 is 53.2 m2/g [21], and for POM is lower trophotometer. 31P MAS NMR spectra were obtained
L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9 3

on a Varian Unity-400 NMR spectrometer. Raman hydrolysis of titanium TTIP in the presence of acidic
spectra were recorded on a Jobin-Yvon HR 800 instru- inorganic precursor such as H3PW12O40 or H6P2W18-
ment with an Ar+ laser source of 488 nm wavelength in O62. In this step, polycondensation reaction occurred
a macroscopic configuration. TEM micrographs were simultaneously, which induced polymerization forming
obtained on a JEOL JEM-2010 transmission electron higher molecular weight products, i.e., H3PW12O40/
microscope at an accelerating voltage of 200 kV. BET Ti(OTi)x(OH)4x or H6P2W18O62/Ti(OTi)x(OH)4x.
specific surface areas, pore sizes and pore volumes were Second, H3PW12O40/Ti(OTi)x(OH)4x or H6P2W18O62/
calculated from nitrogen adsorption/desorption iso- Ti(OTi)x(OH)4x sol was hydrothermally treated at
therm determined at 196 °C using an ASAP 2010 M 200 °C with a rate of 2 °C/min and kept at this temper-
surface analyzer (the samples were outgassed under vac- ature for 1 h, then the products were dehydrated and
uum at 200 °C). calcined in vacuum, resulting in the final products,
H3PW12O40/TiO2 or H6P2W18O62/TiO2. During this
2.2. Photocatalytic degradation of an aqueous process, TiO2 amorphous-anatase phase transition
parathion-methyl using visible-light excitation occurred, which is supported by XRD measurement
(Fig. 1). It showed both H3PW12O40/TiO2 and H6P2-
The photoreactor was designed with a cylindrical W18O62/TiO2 composites prepared by sol–gel method
glass cell configuration and an internal visible-light following hydrothermal treatment with a constant heat-
source (400 W Xe lamp) surrounded by an optical glass ing rate had a uniform anatase phase (JCPDS File No.
jacket (UV spectrum of Xe lamp was cut out through 21-1272), and the corresponding characteristic diffrac-
this optical glass jacket, which has complete absorption tion peaks of 2h values are summarized in Table 1. Com-
in the range of 190–330 nm), where the suspension of the pared with the pure TiO2 prepared here, changes in the
solid catalyst (H3PW12O40/TiO2 or H6P2W18O62/TiO2, peak positions and widths happened somewhat after
0.2 g) and aqueous parathion-methyl (150 ml, 50 mg l1) formation of the composites. As for the composites pre-
totally surrounded the light source. The acidity of the pared by sol–gel method following hydrothermal treat-
suspension was neutral. The temperature of the suspen- ment at 200 °C without controlling the heating rate
sion was maintained at 35 ± 2 °C by circulating water were amorphous (Fig. 1D), suggesting that the key step
through an external cooling coil, and the system was to obtain polyoxotungstate–TiO2 composites with ana-
open to air. During the process of the photocatalytic tase structure at a lower temperature (200 °C) is to
tests, parathion-methyl was monitored by a Shimadzu maintain a constant heating rate during the process of
LC-8A high pressure liquid chromatography (HPLC) hydrothermal treatment the products. On the other
equipped with a C18 column and a UV detector
(k = 280 nm). Before HPLC determination, the with-
drawn suspensions were centrifuged and filtered with (101)
the membrane. The intermediates and final products
(211)
yielded during the course of photodegradation of the (004) (200) (105)
parathion-methyl were identified by a DX-300 ion chro- (204) (116) (215) A
Intensity (a.u.)

matograph (IC) equipped with a conductivity detector


and AS4A or ICE-ASI column. B

3. Results and discussion C

3.1. Preparation and characterization of H3PW12O40/ D


TiO2 and H6P2W18O62/TiO2 composites
0 20 40 60 80
Preparation of the H3PW12O40/TiO2 or H6P2W18O62/ 2 Theta /degree
TiO2 composite included two steps. At first, H3PW12- Fig. 1. XRD patterns for (A) H6P2W18O62/TiO2, (B) anatase TiO2, (C)
O40/TiO2 or H6P2W18O62/TiO2 sol was obtained via H3PW12O40/TiO2, and (D) amorphous H6P2W18O62/TiO2.

Table 1
Characteristic diffraction peaks of 2h values (20–80 °C) for TiO2, H3PW12O40/TiO2, and H6P2W18O62/TiO2
Sample 101 103 200 105 204 116
TiO2 25.21 37.90 47.96 54.32 63.10 75.38
H3PW12O40/TiO2 25.29 37.94 48.00 54.46 62.90 75.23
H6P2W18O62/TiO2 25.29 36.07 46.80 54.46 62.58 74.90
4 L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9

31
Fig. 2. P MAS NMR for (A) H3PW12O40/TiO2, and (B) H6P2W18O62/TiO2.

hand, at this relatively lower temperature, thermal


decomposition of polyoxotungstates was avoided totally
(polyoxotungstates begin to decompose at approxi-
mately 350 °C [24]). Except the characteristic diffraction
peaks corresponding to the anatase-type structure, the
peaks with poor crystallinity and 2h values at 6.23 Anatase TiO2
(1 0 0), and 7.05 (1 0 1), respectively, in Fig. 1A were orig-
inated from the starting H6P2W18O62 unit (JCPDS File H6P2W18O62/TiO2
No. 72-2414). Similarly, the peak of 2h value at 6.60
(1 1 1) in Fig. 1C was originated from the starting H6P2W18O62
H3PW12O40 unit (JCPDS File No. 76-1815). These are
the strongest diffraction peaks of pure H6P2W18O62 or
H3PW12O40 unit. 200 400 600 800 1000 1200
The structures of polyoxotungstates in the as-pre- A Raman shift (cm-1)
pared POM/TiO2 composites were further studied by
31
P magic-angle spinning (MAS) NMR (Fig. 2), Raman
scattering spectroscopy (Fig. 3) and UV/DRS (Fig. 4).
Unfortunately, we failed to characterize the structures
of the composites with IR spectroscopy. The reason is
that the intense and broad vibration peaks originated
from Ti–O–Ti bonds in TiO2 framework severely Anatase TiO2
covered the bands corresponding to P–O, W@O, or
W–O–W bond vibrations from polyoxotungstates at H3PW12O40/TiO2
the mid-IR region. The determined chemical shift for
H3PW12O40/TiO2 and H6P2W18O62/TiO2 was 14.0
and 12.9 ppm, respectively, approximately consistent H3PW12O40
with the values originated from the starting Keggin or
Wells–Dawson unit [24]. Some changes of the 31P chem- 0 200 400 600 800 1100 1200
ical shifts suggested that interactions between the poly- Raman shift (cm-1)
B
oxotungstate and the anatase TiO2 exist in the
composites, resulting in the changes of chemical envi- Fig. 3. Raman scattering spectra of (A) H6P2W18O62, TiO2,
ronments of P atoms. Raman scattering studies further H6P2W18O62/TiO2, and (B) H3PW12O40, TiO2, H3PW12O40/TiO2.
confirmed the anatase structure of as-prepared pure
TiO2 and POM/TiO2 in the composites. That is, in the From Raman spectra, we also find the Keggin or
case of pure TiO2, characteristic peaks at 154.4 (Eg), Wells–Dawson unit-relating peaks. That is, the starting
403.2 (B1g), 523.4 (B1g), and 645.7 cm1 (Eg) originated H3PW12O40-relating Raman peaks were assigned as
from anatase TiO2 [25]. As for the H3PW12O40/TiO2 or follows: i.e., 1001.6 cm1 (strong), corresponding to
H6P2W18O62/TiO2, we also observed the above four symmetric P–O bond stretching vibrations of the PO4
peaks corresponding to the anatase TiO2. However, sites [26]; 977 cm1(weak), corresponding to W@O bond
the Eg Raman mode at 154.4 cm1 shifted to higher stretching vibrations; and 926 cm1 (weak), correspond-
wave numbers, i.e., 156.4 cm1 for H6P2W18O62/TiO2 ing to W–O–W bond stretching vibrations [4]. After
and 160.1 cm1 for H3PW12O40/TiO2, respectively. impregnation of H3PW12O40 into anatase TiO2 crystal
L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9 5

0.5 only produced moderate widening of electronic band


TiO2 (similar crystal field splitting for both cations, just
0.4
H3PW12O40/TiO2 slightly differing in the d orbital average size). Redshift
H6P2W18O62/TiO2
H3PW12O40 of OMCT (oxygen to metal charge transfer) band indi-
0.3
H6P2W18O62 cates that there is a doping energy level between the con-
Absorbance

duction and valence bands of TiO2, and the electronic


0.2
properties of the composites resulted from the following
two transition forms. One is the transition from the va-
lence band to the conduction band (corresponding to
0.1
wide band gap), and the other is the transition from
the valence band to the doping energy level (correspond-
0.0
ing to narrow band gap).
200 250 300 350 400 450 500 From the above results of XRD, 31P MAS NMR,
Wavelength/nm Raman scattering spectroscopy, and UV/DRS, we con-
clude that the primary Keggin or Dawson structure re-
Fig. 4. UV/DRS of the anatase TiO2, the starting H3PW12O40 and
mained intact after formation of the composites. The
H6P2W18O6 2, and the corresponding composites.
changes including chemical shifts in 31P MAS NMR,
Raman shifts and the peak broadening in Raman scat-
lattice, the peaks corresponding to W@O and P–O tering spectroscopy, and redshift of the OMCT band
bonds vibrations were broadened, therefore, we only ob- indicted that the strong interactions between the
served one wide peak in the range of 935.6–1088.2 cm1. POM anions and TiO2 network existed in the compos-
At the same time, the peak intensity corresponding to ites. According to the structural characteristics of H3-
W–O–W bond vibration became much weaker. Similar PW12O40 or H6P2W18O62 and TiO2 network, we inferred
results were found in H6P2W18O62 and H6P2W18O62/ these interactions between the two components were
TiO2. That is, the peaks at 1022, 990, and 910 cm1 were hydrogen bonding and acid basic. At first, hydrogen
originated from P–O, W@O, and W–O–W bond vibra- bonds were formed in the composites between oxygen
tions, respectively, in pure H6P2W18O62. After forma- atoms of the Keggin or Dawson anion and the surface
tion of H6P2W18O62/TiO2 composite, the peaks hydroxyl groups (Ti–OH) of the anatase TiO2 network,
corresponding to W@O and P–O bond vibrations were which can be expressed in the forms of W@Ot   
broadened and appeared at 1012.7 cm1, and the peak HO–Ti, W–Oc    HO–Ti, and W–Oe  HO–Ti, where
intensity corresponding to W–O–W bond vibration be- Ot and Oc or Oe referred to the terminal oxygen atoms
came very weak. The above changes of Raman shifts and the bridge oxygen atoms, respectively, in the POM
and the decreases of the peak intensities are due to strong units. In addition, chemically active surface Ti–OH
interactions between the Keggin or Wells–Dawson unit groups were protonated in an acidic medium to form
and the anatase TiO2 network, which resulted in the de- Ti–OHþ þ
2 groups. Ti–OH2 group should act as a counter
crease of the symmetry of either TiO2 or POM molecules. ion for H3PW12O40 or H6P2W18O62 unit and yielded
Both Keggin and Wells–Dawson unit exhibited a strong ðTi–OHþ  þ 
2 ÞðH2 PW12 O40 Þ or ðTi–OH2 ÞðH5 P2 W18 O62 Þ by
UV absorption at approximately 190 nm, corresponding acid–base reaction. The process is described in Eqs. (1)
to the charge transfer (CT) from O 2p to W 5d occurred and (2). This interaction also was confirmed in the
at W@O bonds. As for W–O–W bonds of the units, their micro-porous H3PW12O40/SiO2, H4SiW12O40/SiO2, and
CT band was weaker and appeared at approximately Na4W10O32/SiO2 composites [14,17].
260 nm for Keggin unit and approximately 302 nm for
Dawson unit, respectively [27]. Anatase TiO2 prepared Ti–OH þ H3 PW12 O40 ! ðTi–OHþ 
2 ÞðH2 PW12 O40 Þ ð1Þ
in the present work has a strong absorption in the range Ti–OH þ H6 P2 W18 O62 ! ðTi–OHþ 
2 ÞðH5 P2 W18 O62 Þ ð2Þ
of 200–380 nm, corresponding to CT from O 2p to Ti 3d
(i.e., electron excitation from the valence band to con- TEM images (Fig. 5) of the two composites show that
duction bond) [4]. After formation of the POM/TiO2 the particles are sphere-like and exhibited uniform size
composites, their CT band shifted to a higher wavelength distribution with an average size of 6 nm, and aggrega-
compared with the starting POM or TiO2, and an tion among particles was also observed in these two
absorption threshold onset that continuously extended images. The rings recorded from the selected area elec-
to the visible region (420 nm), which corresponded to a tron diffraction (SAED) of the sample were indexed to
relatively narrow band gap. This result suggests that diffraction from the (1 0 1), (0 0 4), (2 0 0), (1 0 5), (2 0 4),
the presence of Keggin or Dawson unit also has an influ- (1 1 6), and (2 1 5) planes of anatase, consistent with the
ence on the electronic properties of anatase TiO2 in as- above XRD results. Formation of these tiny sizes of
prepared composites. Similar energy of Ti 3d and W 5d as-prepared composites is closely related to the current
levels would suggest that the POM doping in an anatase prepared method.
6 L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9

Fig. 5. TEM images of (A) H6P2W18O62/TiO2, and (B) H3PW12O40/TiO2.

Very interestingly, the composites exhibited bimodal 3.2. Visible-light photocatalytic testing
pore systems (micro-mesoporosity) with pore size at
approximately 0.6 and 4.0 nm, respectively, from the Parathion-methyl is a kind of chemically stable
nitrogen adsorption–desorption analysis (Fig. 6). That organophosphorus pesticide that is extensively used in
is, the curves show two well-defined adsorption steps. agriculture recently. Therefore, we select it as a model
The first, at very low relative pressure (p/p0 < 0.1), the contaminant to test the visible-light photocatalytic activ-
isotherms exhibit high adsorption, indicating that the ity of as-prepared composites. Photooxidation of para-
powders contain micro-pores (Type I). In addition, thion-methyl should be based on the following
t-plot (Fig. 6B) obtained from the nitrogen adsorption– reaction (Eq. (3)):

S
visible light
(OH3C)2 P O NO2 CO2 + H2O + PO3- 2- - ð3Þ
4 + SO4 + NO3
H3PW12O40/TiO2 or
H6P2W18O62/TiO2

desorption determination did not cross through zero, At first, visible-light photodecomposition of an aque-
also suggesting some micro-pores existed in the compos- ous parathion-methyl (50 mg l1) in the absence of the
ite. The second, at an intermediate relative pressure (p/ photocatalyst was neglectable, i.e., there were no any
p0 = 0.4–0.8), the curves exhibit a hysteresis loop, indi- changes of the concentrations of it after 4 h visible-light
cating that the powders contain mesopores (Type IV). irradiation. In addition, we also did not find obvious
The BET surface areas of the composites were 201 and parathion-methyl degradation after 4 h visible-light irra-
193 m2 g1, for H3PW12O40/TiO2 and H6P2W18O62/ diation of the starting H3PW12O40 and H6P2W18O62 in
TiO2, respectively, which increased greatly compared homogeneous systems. As for the nanosize anatase
with the starting polyoxotungstates and the Degussa TiO2 prepared in the present work, it totally adsorbed
P25. Higher specific surface areas of the composites are the reactant molecules during the course of mixing it
benefit for improving their photocatalytic activities. with an aqueous parathion-methyl in the dark. There-
According to Kumar et al. [29] and Yu et al. [22], forma- fore, the parathion-methyl in this system was not de-
tion of this bimodal pore size distribution is due to hard tected at the beginning of the photocatalytic reaction.
aggregates in the powders. There are two types of pores However, after visible-light irradiation of the suspension
present in the bimodal pore size distribution. One is the of TiO2 and the reactant exceeded 40 min, the concen-
finer intraaggregated pore, corresponding to micro- tration of the reactant increased slowly. The reason
pores, and the other is the larger interaggregated pore, may be desorption of the parathion-methyl from TiO2
corresponding to mesopores. These aggregates occurred as the reaction carrying on. We further tested the visi-
during the composite catalyst preparation process. ble-light photocatalytic activity of the commercial avail-
L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9 7

140 1.0
A
-1

b
Volumme Adsorbed/cm g
120
3

0.8
100 a
H6P2W18 O62/TiO2
80 0.6 H3PW12 O40 /TiO2

Ct /C0
H3PW12 O40
60 H6P2W18 O62
0.4 TiO2
Deguss P25
40
BET surface area of H3PW12O40/TiO2:201 m2g-1 0.2
20
BET surface area of H6P2W18O62/TiO2:193 m2g-1
0 0.0
0.0 0.2 0.4 0.6 0.8 1.0
-30 -20 -10 0 10 20 30 40
Relative Pressure/P/P0
Time (min)
Fig. 7. Photocatalytic degradation of an aqueous parathion-methyl
150
B (C0 = 50 mg l1) under visible-light excitation. Each photocatalyst
Volume Adsorbed/cm /g

used here was 0.2 g, and Ct refers to the concentration of parathion-


3

120 methyl in the reaction system at t time.

90
The above results are summarized in Fig. 7, and the
60 reaction was performed in the presence of dioxygen dis-
solved in the system from air. Once the reaction was
30 done under an Ar inert atmosphere, neither H3PW12-
O40/TiO2 nor H6P2W18O62/TiO2 exhibited efficient pho-
0 tocatalytic activity to the parathion-methyl degradation.
0 5 10 15 20 At this system, the surface of the catalyst became blue.
Thickness-Harkins & Jura/A The blue composites were analyzed by a UV/DRS
immediately, and a new peak appeared at 750 nm for
2.0 H3PW12O40/TiO2 and 768 nm for H6P2W18O62/TiO2,
b
C was detected respectively, attributed to the interva-
-1

a
Pore volume/cm g

1.6 lence electron transfer (IVCT) occurred in the one-elec-


3

tron reduction product (PW12 O4 7


40 or P2 W18 O62 ) of
1.2 PW12 O40 , or P2 W18 O62 [27]. Formation of PW12 O4
3 6
40
or P2 W18 O7 62 is due to the electron transfer from TiO2
0.8 to PW12 O3 6 4
40 or P2 W18 O62 , while PW12 O40 or P2 W18 -
7
O62 was photoinactive. When the reaction was per-
0.4 formed in the presence of dioxygen, the blue composite
should yield, however, it was difficult to detect the
0.0 absorption peak originated from the blue composite
0 2 4 6 8 10 by UV/DRS since it can be reoxidized rapidly.
Pore diameter/nm Disappearance of the parathion-methyl in the reac-
Fig. 6. (A) Nitrogen adsorption–desorption isotherms, (B) t-plot of tion system does not mean its complete degradation.
the H3PW12O40/TiO2, and (C) BJH pore size distributions for From the HPLC effluent curves we found some new
mesopores. a—H3PW12O40/TiO2, b—H6P2W18O62/TiO2. peaks appeared after the peak corresponding to the
parathion-methyl disappeared, suggesting that minerali-
zation of the parathion-methyl required a reaction time
able photocatalyst, Degassa P-25, to the parathion- longer than for its degradation due to the successive
methyl decomposition. It indicated that the photocata- breaking of P–O, P=S, N–O, C–C, and C–O bonds.
lytic activity of P-25 was low at this system. Only did Some intermediates including acetic acid and formic
H3PW12O40/TiO2 and H6P2W18O62/TiO2 exhibit obvi- acid were detected by IC, and total disappearance of
ous visible-light photocatalytic activity to the degrada- acetic acid and formic acid needed 120, and 160 min,
tion of an aqueous parathion-methyl. That is, the respectively, when H3PW12O40/TiO2 was used as a
concentration of the parathion-methyl in the reaction photocatalyst. The final products yielded during
system decreased rapidly by using H3PW12O40/TiO2 or photodegradation of the parathion-methyl should be
H6P2W18O62/TiO2 as the visible-light-driven-photocata- NO 2 3
3 , SO4 , and PO4 , respectively, but only NO3 and

2
lyst, and it only needed 40 min to completely destruct it. SO4 ions were detected with IC. This may be due to
8 L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9

þ
the fact that highly charged PO3 4 ion strongly adsorbed TiO2 þ hm ! ecb þ hvb ð4Þ
on the surface of the nanoscale photocatalyst. This POM þ e 
cb ! POM ð5Þ
adsorption also led us to hardly determine the concen-
POM þ O ! O þ POM ð6Þ
tration of NO 3 or SO4
2
ion quantitatively and accu- 2 2

rately, and the estimated yield for NO 2 O2 þ e! O2 ð7Þ
3 and SO4 ion cb
hþ  þ
was approximately 88% and 71%, respectively. vb þ H2 O ! OH þ H ð8Þ
After the reaction, the catalysts can be easily recov-
Thus, the rate of oxidation of parathron-methyl in-
ered by sedimentation, and the recovered catalysts were
creased significantly under visible-light irradiation of
treated with hot water (80 °C) and ethanol successively
H3PW12O40 or H6P2W18O62-containing anatase com-
by stirring the suspension vigorously for 3 h. The ob-
posites. As for the homogeneous H3PW12O40 or
tained dried catalysts were analyzed by ICP-AES
H6P2W18O62, it seems that the concentration of superox-
XRD, and UV–vis. The loading of the H3PW12O40 or
ides generated is much smaller than that in a heteroge-
H6P2W18O62 maintained almost unchangeable, and the
neous H3PW12O40/TiO2 or H6P2W18O62/TiO2 system.
structures of the catalysts were not destroyed. In addi-
When the reaction was performed in an Ar atmosphere,
tion, after five catalytic recycling, the leakage of
the stable and photoinactive POM was formed, and it
H3PW12O40 or H6P2W18O62 from TiO2 support was less
was hardly reoxidized in this inert environment, result-
than 1.5%, and deactivation of the catalysts was hardly
ing in the cessation of the parathron-methyl degradation
observed. These results further implied that the interac-
reaction. (iii) Relatively narrow band gap of the com-
tion between the Keggin or Dawson unit and TiO2 sup-
posite. After formation of the POM/TiO2 composites,
port is indeed chemical rather than physical.
their band gap decreased compared to that with either
Here, although as-prepared two composites exhibited
the starting POM or the anatase TiO2. Therefore, the
main absorption in the range from 200 to 420 nm, fol-
energy of the transition from the valence band to the
lowing factors ensure their efficient photocatalytic activ-
doping energy level or from the doping energy level to
ity to aqueous parathion-methyl degradation. (i) The
the conduction band level was lower than that of the va-
porous structure of the composites. Compared with
lence band to the conduction band of TiO2. Thus, it is
the starting POM or TiO2, as-prepared porous POM/
possible to utilize visible-light type excitation more effi-
TiO2 composites showed relatively high BET surface
ciently than classic Ti-only based one [28].
areas, which resulted in an increase in their photocata-
lytic activity by providing more contact areas between
the catalyst and the substrate for the surface-mediated
electron transfer reactions to take place. In addition, 4. Conclusions
the porous catalysts showed a strong adsorption to the
substrates, which also played an important role to im- Two efficient heterogeneous photocatalysts, H3PW12-
prove their photocatalytic activity since the adsorption O40/TiO2, and H6P2W18O62/TiO2, with anatase crystal-
is the first step of the photocatalytic reaction. From line phase, dual-pore structures, narrow band gap, and
Fig. 7 we observed that the concentration of the para- nanometer sizes, were prepared by combination of the
thion-methyl decreased by 10% after 30 min adsorption. methods of sol–gel and hydrothermal treatment at a
Another function for the porous materials is that it pro- lower temperature (200 °C, 2 °C/min). The design idea
vides enhanced mass transport for the reactant mole- was motivated by the electron transfer mediating ability
cules into and out of the pore. (ii) Synergistic effect of POM, therefore, the role of the POM in TiO2 photo-
existed between the POM unit and the anatase TiO2. catalysis is critical in enhancing the overall photoeffi-
Both H3PW12O40and H6P2W18O62 are effect electron ciency by retarding the fast recombination of charge-
acceptors in TiO2 photocatalytic systems, and they acted pair (h+–e) on TiO2 and producing superoxides.
as the electron shuttles in visible-light illuminating TiO2 Results of XRD, 31P MAS NMR, Raman scattering
[21,22], i.e., POM can accept e cb from the surface of spectroscopy, and UV/DRS indicated that the primary
TiO2 and form the reduced POM (PW12 O4 40 or Keggin or Dawson structure remained intact after
P2 W18 O7
62 ), and the reduced POM was reoxidized by immobilization of them into TiO2 network, which was
O2 to yield superoxides (O2 ) and its oxidized form for ensured by the lower temperature preparation process.
another catalytic cycle. Therefore, the role of the POM At the same time, the hydrogen bonding and acid–base
in TiO2 photocatalysis is critical in enhancing the overall interactions existed in the composites. As-prepared
photoefficiency by retarding the fast recombination of composites were successfully used in the process of
charge-pair (h+–e) on TiO2 and produce strong oxi- visible-light photocatalytic degradation of an aqueous
dant, O2 . The presence of POMs enhances the genera- parathion-methyl, attributed to the synergistic effect
tion of O2 , a kind of main oxidant of parathron-methyl between the POM and the anatase TiO2, the porous
except OH radicals [20]. The process is described in Eqs. structures and narrow band gap of the composites.
(4)–(8). The above composite catalysts also had the advantages
L. Li et al. / Microporous and Mesoporous Materials 87 (2005) 1–9 9

of easily being separated and recovered, and deactiva- [10] D. Sattari, C.L. Hill, J. Am. Chem. Soc. 115 (1993) 4649.
tion of the catalysts was hardly observed after five cata- [11] H. Einaga, M. Misono, Bull. Chem. Soc. Jpn. 69 (1996) 3435.
[12] T. Yamase, N. Takabayashi, M. Kaji, J. Chem. Soc., Dalton
lytic recycling. Trans. (1984) 793.
[13] E. Papaconstantinou, Chem. Soc. Rev. 18 (1989) 1.
[14] Y. Guo, Y. Wang, C. Hu, E. Wang, Chem. Mater. 12 (2000)
Acknowledgement 3501.
[15] Y. Guo, D. Li, C. Hu, Y. Wang, E. Wang, Appl. Catal. B 3–4
We thank to Professor Y. Liu and Professor C. Shao (2001) 337.
[16] Y. Guo, Y. Yang, C. Hu, C. Guo, E. Wang, Y. Zhou, S. Feng, J.
for helping determination and explanation the Raman
Mater. Chem. 12 (2002) 3046.
spectra. Also, the NSFC is acknowledged for financial [17] Y. Guo, C. Hu, X. Wang, E. Wang, Y. Zhou, S. Feng, Chem.
support (No. 20271007, 20271045, and 20331010). This Mater. 13 (2001) 4058.
work is also supported by the Program of New Century [18] Y. Guo, C. Hu, C. Jiang, Y. Yang, S. Jiang, X. Li, E. Wang, J.
Excellent Talents in University (NCET-04-0311). Catal. 217 (2003) 141.
[19] R.O. Ruya, J.L. Ferry, J. Phys. Chem. B 106 (2002) 4336.
[20] R. Jungho, W. Choi, Environ. Sci. Technol. 38 (2004) 2928.
[21] H. Park, W. Choi, J. Phys. Chem. B 107 (2003) 3885.
References [22] J. Yu, J.C. Yu, M.K.-P. Leung, W. Ho, B. Cheng, X. Zhao, J.
Zhao, J. Catal. 217 (2003) 69.
[1] T. Ruther, A.M. Bond, W.R. Jackson, Green. Chem. 5 (2003) 364. [23] C. Hu, M. Hashimoto, T. Okuhara, M. Misono, J. Catal. 143
[2] Z. Lei, W. You, M. Liu, G. Zhou, T. Takata, M. Hara, K. (1993) 437.
Domen, C. Li, Chem. Commun. (2003) 2142. [24] M.T. Pope, Hetropoly and Isopoly Oxometalates, first ed.,
[3] S.U.M. Khan, M. Al-Shahry, W.B. Ingler Jr., Science 297 (2002) Springer-Verlag, New York, 1983, p. 8.
2243. [25] L. Miao, S. Tanemura, S. Toh, K. Kaneko, M. Tanemura, J.
[4] A. Fuerte, M.D. Hernandez-Alonso, A.J. Maira, A. Martinnez- Cryst. Growth 264 (2004) 246.
Arias, M. Fernandez-Garcia, J.C. Conesa, J. Soria, G. Munuera, [26] I. Nobuhito, M. Toshiyuki, H. Hidekazu, A. Gin-ya, Chem.
J. Catal. 212 (2002) 1. Mater. 15 (2003) 2289.
[5] W. Ma, J. Li, X. Tao, J. He, Y. Xu, J.C. Yu, J. Zhao, Angew. [27] A. Hiskia, A. Mylonas, E. Papaconstantinous, Chem. Soc. Rev.
Chem. Int. Ed. 42 (2003) 1029. 30 (2001) 62.
[6] J.-M. Herrmann, Catal. Today 53 (1999) 115. [28] Y. Cao, W. Yang, W. Zhang, G. Liu, P. Yue, New. J. Chem. 28
[7] M. Vautier, C. Guillard, J.-M. Herrmann, J. Catal. 201 (2001) 46. (2004) 218.
[8] C. Hu, Y. Wang, H. Tang, Appl. Catal. B 35 (2001) 95. [29] K.N.P. Kumar, J. Kumar, K. Keizer, J. Am. Chem. Soc. 77 (1994)
[9] D. Sattair, C.L. Hill, J. Chem. Soc., Chem. Commun. (1990) 634. 1396.

You might also like