Pt

KD

a

ARRAA

KBPLKS

1

tgrrsdw

mTMdartcZ

Cbml

0d

Applied Catalysis A: General 399 (2011) 22–27

Contents lists available at ScienceDirect

Applied Catalysis A: General

journa l homepage: www.e lsev ier .com/ locate /apcata

hotocatalytic activity and its stacking order dependence of transparent 12ungsto(VI) phosphoric acid-brookite hybrid films

unchaya Pruethiarenun, Toshihiro Isobe, Sachiko Matsushita, Akira Nakajima ∗

epartment of Metallurgy and Ceramic Science, Graduate School of Science and Engineering, Tokyo Institute of Technology, 2-12-1 O-okayama, Meguro-ku, Tokyo 152-8552, Japan

r t i c l e i n f o

rticle history:eceived 4 January 2011eceived in revised form 8 March 2011ccepted 15 March 2011

a b s t r a c t

Transparent hybrid films of brookite (TiO2) and 12 tungsto(VI) phosphoric acid (PW12) were preparedvia layer-by-layer (LBL) processes on a quartz substrate with various orders. Their photocatalytic acti-vities were evaluated using gaseous 2-propanol (IPA) decomposition. Pure brookite film showed higherphotocatalytic activity than the pure PW12 one; the combination of PW12 to brookite increased the pho-

vailable online 23 March 2011

eywords:rookitehotocatalystayer-by-layer

tocatalytic activity. The electron scavenger effect of PW12 against brookite under UV illumination isimportant for photocatalytic activity improvement. The degree of improvement depends on the film sta-cking order. Their photocatalytic activity was the highest when PW12 was arranged on the film’s topmostsurface.

© 2011 Elsevier B.V. All rights reserved.

EGGINtacking order

. Introduction

Titanium dioxide (TiO2) has been used widely as a photocataly-ic material [1,2]. Ultraviolet light (UV) illuminating a TiO2 surfaceenerates electrons and holes, which diffuse to the TiO2 surface andespectively reduce and oxidize adsorbed oxygen and water, the-eby producing radical species such as OH and H2O2. These activepecies decompose various organic compounds and bacteria. Toate, numerous studies have examined the application of TiO2 forater and air purification [3,4].

Nevertheless, practical application of the TiO2 for environ-ental purification remains limited. One weak point posed by

iO2 photocatalysis is its poor electron–hole separation efficiency.ost electron–hole pairs generated in TiO2 under UV illumination

isappear by recombination; only a few can move to the surfacend react with adsorbates [5,6]. Inhibition of this electron–holeecombination is an effective approach to improving TiO2 photoca-alytic activity [4]. Various investigations of this concept have beenonducted, such as Pt loading, hybridization of other oxides, and a-scheme design [7–11].

Keggin-type heteropolyacid ion ([XM12O40]n−, M = W, Mo, V, Nb,

r, and Ta, hereinafter denoted as XM12)/TiO2 hybrid systems, haveeen investigated as Z-scheme photocatalytic systems in aqueousedia [12–14]. The reduction potentials of Keggin-type heteropo-

yacid tungstate ions such as PW12O403− (PW12) and SiW12O40

4−

∗ Corresponding author. Tel.: +81 357 342 524/5; fax: +81 357 343 355.E-mail address: anakajim@ceram.titech.ac.jp (A. Nakajima).

926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.apcata.2011.03.030

(SiW12) are lower than the conduction band level of TiO2 [15,16].Therefore, electron transfer from TiO2 conduction band to XW12is expected. Reduced XW12 (XW12

–) absorbs visible light to forman excited state XW12

–(XW12–*), which can reduce organic com-

pounds such as methyl viologen or methyl orange in water [13,14].Recently, we prepared transparent PW12/TiO2 hybrid thin films

using layer-by-layer (LBL) method. We then investigated theirphotocatalytic decomposition activity using gaseous 2-propanol(IPA) [17]. The hybrid films’ photocatalytic activity ((PW12/TiO2)n

(n = times of bilayer deposition)) was increased by the illuminationof UV with visible light. Results suggest that second step photoex-citation using visible light increased the reduction rate of PW12 andincreased the charge separation efficiency. Moreover, we investiga-ted effects of heteroatoms and polyatoms on these hybrid thin filmsusing various tungstate, molybdate, and tungsto molybdate mate-rials. Results revealed that PW12/TiO2 is the best combination forthe improvement of photocatalytic activity [18,19]. However, thestacking-order dependence of photocatalytic activity on the hybridfilms was not well investigated. Moreover, all of those earlier stu-dies were conducted using anatase-type nanoparticles as a TiO2source.

Three TiO2 polymorphs exist: rutile, anatase, and brookite.Although anatase has been commonly used as a photocatalyticmaterial, several reports describe the high photocatalytic activity

of brookite on the decomposition of IPA and acetaldehyde [20,21]and photoinduced hydrophilicity [22]. Using the same method forthis study, we hybridized PW12 with brookite-type TiO2 nanopar-ticles. Then, their photocatalytic activities were evaluated usinggaseous IPA decomposition. Moreover, the effect of stacking order

Cataly

op

2

2

aKdtplF(wi1

Jf3wstsrilsfCadop

e1I7c�o

oTdIfi

depJwRwcauLdt

K. Pruethiarenun et al. / Applied

n the entire photocatalytic activity was examined using counterolymers.

. Experimental procedure

.1. Preparation and characterization of (PW12/TiO2)n films

For (PW12/TiO2)3 film preparation, a commercial brookitequeous suspension (NTB 01, 15 wt% TiO2, pH 4; Showa Denko.K., Japan) was used as the starting material. This suspension wasiluted to concentrations of 0.1 g/L, and the pH value was adjus-ed to 1.5 using a HNO3 aqueous solution because PW12 requiresH lower than 2 to avoid a hydrolysis reaction and TiO2 needs

ow pH to maintain its well dispersion state in the suspension.or these experiments, 12 tungsto(VI) phosphoric acid n-hydrateH3(PW12O40)nH2O; Wako Pure Chemical Industries, Ltd., Japan)as used with no purification. This chemical was also dissolved

nto diluted HNO3 aqueous solution to reach the concentration ofmM.

A quartz glass plate (30 mm × 60 mm × 1 mm; Tosoh Corp.,apan) was washed and soaked in HCl and methanol mixture (1:1)or 30 min and then soaked into conc. (95%) H2SO4 for another0 min. The substrate was washed using distilled water. Then itas immersed into prepared PW12 solution for 20 min. The sub-

trates’ surface is positively charged in acid solution. Therefore,he first layer of PW12 molecules was adsorbed onto the surface ofubstrates by Coulombic potential. Subsequently, the sample wasinsed twice in pH 1.5 HNO3 aqueous solutions and dried at 60 ◦Cn the oven under air atmosphere until dry. To grow the next TiO2ayer on PW12, the (PW12)1 film was immersed again into brookiteuspension. The brookite particles have a sufficient positive chargeor maintaining a well-dispersed state and for LBL coating usingolumbic attraction with a polyacid ion. Rinsing twice in HNO3nd drying at 60 ◦C in an oven were also conducted, which pro-uced (PW12/TiO2)1 film. These procedures were repeated twice tobtain three bilayers (PW12/TiO2)3. Identical procedures were alsoerformed for preparation of other films.

The control (PW12)3 films and (TiO2)3 films were prepared usingither poly diallyldimethylammonium chloride (PDDA, typical Mw00,000–200,000, 20 wt% aqueous solution; Aldrich Chemical Co.,

nc.) or poly sodium 4-styrene sulfonate (PSS, average Mw ca.0,000, 30 wt% aqueous solution; Aldrich Chemical Co., Inc.) asounter polymers. After coating, UV illumination (66 mW/cm2 at= 365 nm) was conducted for the photocatalytic decompositionf counter polymers using a Hg–Xe lamp for 6 h.

Using these polymers, three films with different stackingrder (PW12 → TiO2 → TiO2 (PTT), TiO2 → PW12 → TiO2 (TPT), andiO2 → TiO2 → PW12 (TTP)) were also prepared. Detailed proce-ures for the control films were described in an earlier report [17].

n the present study, we controlled UV absorbance of TiO2 in alllms prepared to almost identical values.

The starting TiO2 suspension was dried at 60 ◦C in air for 1ay, which produced the dry powder. The crystalline phase wasvaluated using X-ray diffraction (XRD; XRD-6100; Shimadzu Cor-oration, Japan) and Raman spectrophotometry (Ramanor T64000;

obin Yvon Co., France). The chemical composition of the materialsas evaluated using X-ray fluorescence analysis (XRF, RIX2000;igaku Corp., Tokyo, Japan). The specific surface area of the powderas evaluated using BET method with N2 (Autosorb-1; Quanta-

hrome Instruments, Boynton Beach, FL, USA). The morphology

nd powder size in the starting TiO2 suspension were observedsing a transmission electron microscope (H-9000UHR; Hitachi,td., Japan). The band gap of the dry powder was evaluated from theiffuse reflectance spectra [23] using a UV–vis-NIR scanning spec-rophotometer (UV-2450; Shimadzu Corporation, Tokyo, Japan).

sis A: General 399 (2011) 22–27 23

The UV–vis spectra were also measured at each coating stepusing a different UV–vis-NIR scanning spectrophotometer (V-630;Jasco Corp., Tokyo, Japan). X-ray photoelectron spectroscopy (XPS,PHI Quantera SXM; PHI Co., USA) with an Al K� X-ray source(1468 eV) was used to evaluate the Ti and W concentration ratioon the film surface at different LBL step. The (PW12/TiO2)3 filmsmorphology and thickness were evaluated through direct observa-tion of the top surface and crosscut section using a field emissionscanning electron microscope (FE-SEM, S4500; Hitachi, Ltd., Tokyo,Japan). Surface roughness of the film was evaluated in 500-nm2

area by atomic force microscopy (AFM, JSPM-5200; JEOL, Tokyo,Japan) with a Si cantilever (NSC36-b; Mikromasch, Narva Mtn.,Estonia).

2.2. Evaluation of photocatalytic activity

Before measurement, prepared films were exposed to UV light(1 mW/cm2, 365 nm) for 30 min to remove organic compoundsadsorbed onto the surface. A Pyrex glass vessel (500 mL volume)with a quartz glass lid was used as a batch reactor. A sample filmwas set at the center of the vessel; then the atmosphere in the reac-tor was replaced by a gas flow (1.0 L/min, 5 min) of air at 20 ◦C and80% relative humidity. Subsequently, the vessel was sealed. ThenIPA gas was injected into it. The injected gas amount was equiva-lent to that for 500 ppm concentration. The vessel was then storedin the dark for 5 h.

After the adsorption equilibrium was confirmed, UV illumina-tion was conducted using a UV illuminator (LA-410UV-1; HayashiWatch Works Co., Ltd., Japan) equipped with a Hg–Xe lamp. Thislight source includes several peaks in the 280–450 nm range, withthe strongest peak at 365 nm [17]. The illumination intensity atthe sample surface was adjusted to 1 mW/cm2 at 365 nm (evenwhen filters were employed.). The photocatalytic decompositionpathway of IPA has been studied extensively [24–27]. The mole-cule of gaseous IPA is initially decomposed into acetone, and finallydecomposed to H2O and CO2. The photocatalytic decompositionrate of acetone by TiO2 photocatalyst is much slower than that ofIPA because of displacement by water vapor from the TiO2 surface[26] and because of the reaction rate difference from •OH radical[27]. In this study, the IPA, acetone, and CO2 concentrations wereevaluated every 30 min or 1 h for 5 h using a gas chromatograph(GC-14A; Shimadzu Corporation, Tokyo, Japan) equipped with aflame ionization detector (FID), a methanizer, and a Sunpak-Acolumn (Shimadzu Corp.). The carrier gas was nitrogen. The respec-tive temperatures for the detector and injection port were 230 ◦Cand 200 ◦C. The initial column temperature was 50 ◦C, which wasmaintained for 2.5 min. The temperature was increased 20 ◦C/minto 130 ◦C to achieve peak separation.

Two colored glass filters were used to limit the wavelengthrange of light illumination: UV-33 (absorbed UV < 330 nm; AsahiGlass Co., Ltd., Japan) and UV-D33S (absorbed visible light > 400 nm;Asahi Glass Co., Ltd.). In both the UV-limited and vis-limitedcase, the illumination intensity was 1 mW/cm2 at 365 nm. Otherexperimental conditions were identical to those of the all-lightillumination cases.

3. Results and discussion

3.1. Characterization of brookite and (PW12/TiO2)3 thin films

The specific surface area of dried brookite powder was 174 m2/g.Fig. 1 shows the XRD pattern and Raman spectra of the powder.Although the peaks of brookite by XRD pattern resemble those ofanatase, the obtained Raman spectra coincided well with thosedescribed in previous reports on brookite [21] (such as 128, 153,

24 K. Pruethiarenun et al. / Applied Catalysis A: General 399 (2011) 22–27

2tpmTs9da

0

0.5

1

1.5

TiO2

PW12

0

1.5

0.5

1

0 1 2 3

200 400 600 800

Wavelength (nm)

Abs

orba

nce

a

b

Abs

orba

nce

Layer number

Fig. 1. (a) XRD pattern, and (b) Raman spectra of dried brookite powder.

47, 322 and 636 cm−1, whereas the spectra of commercial ana-ase powder (ST-21; Ishihara Sangyo Kaisha, Ltd., Japan) providedeaks at 144, 397, 515, and 639 cm−1 under the same measure-ent condition.). This result confirmed that crystalline phase of the

iO2 in the suspension employed for this study was brookite. Sub-

equent XRF analysis revealed that the powder had greater than9% purity of TiO2. Fig. 2 presents TEM image of the TiO2 pow-er. The particle size ranges around 5–20 nm; its morphology wassquare plate-like shape, as reported by Shibata et al. [22]. The

Fig. 2. TEM micrograph of dried brookite powder.

Fig. 3. (a) UV–vis absorbance spectra increment per layer of PW12 and TiO2 of(PW12/TiO2)3, and (b) UV absorbance spectra increment at 249 nm per layer of PW12

and TiO2 of (PW12/TiO2)3.

band-gap of the powder was calculated as 3.18 eV from the diffusereflectance spectra by assuming direct transition (see Supportinginformation). This value was close to the previously reportedvalue [28].

Fig. 3 shows UV–vis spectra of (PW12/TiO2)3. Both PW12 andTiO2 deposition increase the absorbance in UV range. The steadybilayer-step increment indicated that the thickness of the bilayerfilm was almost identical. The results of (TiO2)3 and (PW12)3resembled that of (PW12/TiO2)3. Fig. 4 shows the surface atomicratio of W/Ti for (PW12/TiO2)1 + PW12 (top surface is PW12) and(PW12/TiO2)1 films (top surface is TiO2) by XPS measurement. Theratio dependence on the coating cycle suggests alternative coa-ting using this method. The spectra for (PW12/TiO2)3 hybrid filmsimply that transmittance in the visible area (� = 400–780 nm) washigher than 80%, and that the Ra values of the coatings obtained byAFM at five points on the top surface of (PW12/TiO2)3 and (TiO2)3films were, respectively, 5–10 nm and 4–7 nm. Fig. 5 shows top andcrosscut section views of (PW12/TiO2)3 and (TiO2)3 by SEM. Thethickness of these films was 70–80 nm.

3.2. Photocatalytic properties

Fig. 6 portrays wavelength spectra of the light source (Hg–Xelamp) with (white region) and without (gray and white region)filters. The UV-33 filter cuts the wavelength of the light for theHOMO–LUMO excitation of PW12, and the UVD-33S filter cuts the

K. Pruethiarenun et al. / Applied Catalysis A: General 399 (2011) 22–27 25

Fa

wt

(tr5Tsdnkp[

When most UV light was cut by UV33 (Fig. 7(b)), the IPA decom-

ig. 4. Surface atomic ratio of W/Ti for (PW12/TiO2) + PW12 (top surface is PW12),nd (PW12/TiO2) films (top surface is TiO2) according to XPS measurements.

avelength of the light for the excitation of reduced PW12 (PW12−)

o PW12−* [17].

Fig. 7(a) shows the IPA concentration change for thePW12/TiO2)3, (TiO2)3 and (PW12)3 films under all-light (no fil-ers) UV illumination. The C represents IPA concentration and C0epresents the initial concentration after storage in the dark forh. To avoid complexity, Figs. 7 and 9 plot only the change of IPA.he practical C0 values and the changes of acetone and CO2 arehown in Supporting information. Results show that (PW12/TiO2)3ecomposed IPA faster than (TiO2)3, although (PW12)3 showed

o IPA decomposition activity. The combination of PW12 to broo-ite increased photocatalytic activity. This result corresponds withrevious studies conducted using anatase nanoparticles as TiO217,18], and suggests that PW12 enhances electron–hole separation,

Fig. 5. SEM images of (PW12/TiO2)3 ((a) and (b)), and (TiO2)3 ((c) and (d)) fil

Fig. 6. Wavelength spectra of the light sources with (white region) and without(gray and white region) filters: (a) Hg–Xe lamp and UV33, (b) Hg–Xe lamp andUVD33S.

as discussed in previous studies.

position rate of both (PW12/TiO2)3 and (TiO2)3 decreased greatlycompared with the decomposition rate under all-light UV illumi-nation. However, (PW12/TiO2)3 still showed better photocatalyticactivity than (TiO2)3. This result indicates that PW12 remained able

ms: (a) and (c) present top views; (b) and (d) show crosscut sections.

26 K. Pruethiarenun et al. / Applied Catalysis A: General 399 (2011) 22–27

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

Time (h)

C/C

0

Time (h)

C/C

0

Time (h)

C/C

0

a

b

c

F

(a

t

adUlaivtbtp

awlfit

hmao

ig. 7. Decomposition rate of IPA for (PW12/TiO2)3 ( ), (TiO2)3 ( )) and (PW12)3

) films under (a) all-light UV illumination, (b) UV illumination with a UV33 filter,nd (c) UV illumination with UVD33S filter.

o receive electrons from TiO2.Fig. 7(c) presents the concentration change for (PW12/TiO2)3

nd (TiO2)3 under light illumination through UVD33S. The IPAecomposition rate was slightly lower than that under all-lightV illumination in both films. Because UVD33S cut some of UV

ight with visible light, it is presumed that the rate decrease mightrise from the UV light cut. However, the activity of (PW12/TiO2)3s higher than that of (TiO2)3. This result suggests that the effect ofisible light effects of (PW12/TiO2)3 hybrid film on the overall pho-ocatalytic activity is small, although PW12

− presents the secondaryand level on the visible light region [29]. It was reconfirmed thathe photocatalytic activity of TiO2 plays a major role in the overallhotocatalytic reaction of this binary system.

In previous studies, the activity order between (PW12/TiO2)n

nd (TiO2)n was switched when these two glass filters were usedith a UV light source (namely (PW12/TiO2)n > (TiO2)n under all-

ight UV illumination, and (PW12/TiO2)n < (TiO2)n when these twolters were used.). However, in this study, the activity order washe same, even when two glass filters were used.

Fig. 8 depicts the expected electron transfer scheme of theybrid film of PW12 and TiO2 (brookite). When UV light is illu-inated on the surface of this hybrid film, electrons are excited

nd holes are generated simultaneously at the valence band (VB)f TiO2 and HOMO level of PW12. Generated holes in TiO2 oxidize

Fig. 8. Energy diagram of TiO2 (brookite)-PW12 system.

IPA directly or through radical species, whereas the holes in HOMOlevel of PW12 trap electrons from TiO2 effectively. Consequently,it is reduced to PW12

−, which absorbs visible light and produces astronger reduction power than ground-state PW12

−. Ground-statePW12 was regenerated by O2 reduction [17].

In the case presented herein, the absorbance of TiO2 is 4–5times greater than that described in reports of previous studies[17–19]. The reason remains unclear, but the dispersibility of broo-kite differed from that of anatase. Therefore, it is expected that theamount of electrons excited in TiO2 is larger than that in PW12 inthe present case. Although some electrons in TiO2 are consumedfor healing holes in PW12, excess electrons would transfer directlyto LUMO level of PW12 (Some disappear by recombination withholes in TiO2.). When the excitation efficiency by visible light toform PW12

−* from PW12− is not high, Reduction 1 in Fig. 8 might

become more important than Reduction 2 as the practical reduc-tion path. Experimental results show that PW12 acts effectively asan electron scavenger and that it inhibits the charge recombinationof TiO2, consequently increasing the overall photocatalytic activity.

We prepared a (PW12/TiO2)3 thin film using an anatase suspen-sion (STS-100, 20 wt% concentration; Ishihara Sangyo Kaisha, Ltd.,Mie, Japan) employed in an earlier study [17], and then examinedits photocatalytic activity under the same conditions. The result ispresented in Supporting information. The photocatalytic decompo-sition activity against gaseous IPA for the case of brookite is higherthan that for anatase. However, this result cannot be attributedsimply to the crystalline phase difference. As described earlier, theabsorbance of UV per layer for anatase is less than that for brookite.Moreover, the specific surface area (anatase, 236 m2/g; brookite,174 m2/g) and crystallinity differ. Unless the TiO2 particles havethe same physical and chemical properties, the crystalline phasedependence cannot be discussed.

Additionally, we examined effects of the stacking order on thehybrid films to obtain an optimum film structure using coun-ter polymers described in the experimental section. We preparedPW12/TiO2/TiO2 (PTT), TiO2/PW12/TiO2 (TPT), and TiO2/TiO2/PW12(TTP) films (left-hand-side is next to the substrate and the right-hand-side is the top layer) and compared their photocatalyticactivity.

The results of photocatalytic activity measurements are shownin Fig. 9. Activity order under all-light UV illumination was

PTT < TPT < TTP. For the PTT film, PW12 exists on the bottom layer;thereby, O2 must penetrate through two layers of TiO2. Insuffi-cient O2 on the PW12 surface decreases the reduction rate andthe entire photocatalytic activity of the PTT film. Consequently, theelectron–hole recombination in the top TiO2 coating was not inhi-

K. Pruethiarenun et al. / Applied Cataly

b

<< e-

e-

e-

e-

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

Time (h)

C/C

0

a

Fa

baPToIotoa

asiaopocfi

amah

4

p

[[

[[

[

[[

[

[

[

[

[

[

[[

[25] Y. Ohko, K. Hashimoto, A. Fujishima, J. Phys. Chem. A 101 (1997) 8057–8062.[26] A. Mylonas, A. Hiskia, E. Androulaki, D. Dimotikali, E. Papaconstantinou, Phys.

Chem. Chem. Phys. 1 (1999) 437–440.[27] S.A. Lanson, J.A. Widegren, J.L. Falconer, J. Catal. 157 (1995) 611–625.

PTT TPT TTP

ig. 9. Decomposition rates of IPA for TTP ( ), TPT ( ), and PTT ( ) films underll-light UV illumination.

ited by PW12 in PTT; this film exhibited the lowest photocatalyticbility among the three films. For TPT, more O2 is able to reach theW12 surface. However, O2 must penetrate through the first layer ofiO2. Electrons move in two directions against each other becausef the PW12 position in the film (sandwiched by two TiO2 layers).n this alignment, the bottom TiO2 does not contribute well to theverall photocatalytic activity because holes generated in the bot-om TiO2 do not transfer to the main reaction field: The top surfacef the hybrid film. Therefore, TPT showed moderate photocatalyticctivity.

The best stacking order among these three films was TTP. Itsctivity was almost equivalent to that of (PW12/TiO2)3 despite itsmaller amounts of TiO2 and PW12. Its high photocatalytic activitys attributable to the effective electron transport in one directionnd the abundance of oxygen and water. Although PW12 is locatedn the top surface, the UV absorbance of PW12 is very small com-ared with that of TiO2, as shown in Fig. 3. Therefore, the existencef PW12 does not affect the excitation of TiO2 remarkably. Rather,harge separation was attained on the top surface by virtue of thislm alignment. The overall photocatalytic activity was increased.

This study demonstrated that PW12 improved photocatalyticctivity of brookite as well as anatase. Detailed quantitative esti-ation of the practical efficiency of charge transfer for each scheme

nd investigation of photoinduced wettability conversion of theseybrid films will be addressed in future works.

. Conclusion

Transparent hybrid films of PW12 and TiO2 (brookite) were pre-ared using LBL method. Their photocatalytic activities were then

[[

sis A: General 399 (2011) 22–27 27

evaluated through decomposition of gaseous IPA. The (PW12/TiO2)3hybrid film has higher photocatalytic activity than a pure TiO2film has. It was deduced that electron scavenger effect of PW12plays an important role in this mechanism. The IPA decompo-sition rate of the (TiO2/TiO2/PW12) film was almost equivalentto that of (PW12/TiO2)3 in spite of its smaller material amount.The design of stacking order of PW12 and TiO2 is importantto obtain high photocatalytic activity from these hybrid filmcomposites.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.apcata.2011.03.030.

References

[1] S.N. Flank, A.J. Bird, J. Am. Chem. Soc. 99 (1977) 303–304.[2] M. Fujihira, Y. Satoh, T. Osa, Nature 293 (1981) 206–208.[3] A. Mills, S. LeHunte, J. Photochem. Photobiol. A: Chem. 108 (1997) 1–35.[4] A. Fujishima, X. Zhang, D.A. Tryk, Surf. Sci. Rep. 63 (2008) 515–582.[5] S.Y. Murakami, H. Kominami, Y. Kera, S. Ikeda, H. Noguchi, K. Uosaki, B. Ohtani,

Res. Chem. Intermed. 33 (2007) 285–296.[6] Z. Zhang, J.T. Yates Jr., J. Phys. Chem. C. 114 (2010) 3098–3101.[7] S. Sato, J.M. White, Chem. Lett. 72 (1980) 83–86.[8] M. Anpo, M. Takeuchi, J. Catal. 216 (2003) 505–516.[9] M. Miyauchi, A. Nakajima, K. Hashimoto, T. Watanabe, Adv. Mater. 12 (2000)

1923–1927.10] K. Sayama, K. Mukasa, R. Abe, H. Arakawa, Chem. Commun. (2001) 2416–2417.11] H. Kato, M. Hori, R. Konta, Y. Shimodaira, A. Kudo, Chem. Lett. 33 (2001)

1348–1349.12] H. Park, W. Choi, J. Phys. Chem. B 107 (2003) 3885–3890.13] M. Yoon, J.A. Chang, Y. Kim, J.R. Choi, K. Kim, S.J. Lee, J. Phys. Chem. B 105 (2001)

2539–2545.14] T. Tachikawa, S. Tojo, M. Fujitsuka, T. Majima, Chem. Eur. J. 12 (2006)

3124–3131.15] I.A. Weinstock, Chem. Rev. 98 (1998) 113–170.16] D. Duonghong, J. Ramsden, M. Graetel, J. Am. Chem. Soc. 104 (1982) 2977–

2985.17] S. Yanagida, A. Nakajima, T. Sasaki, Y. Kameshima, K. Okada, Chem. Mater. 20

(2008) 3757–3764.18] S. Yanagida, A. Nakajima, T. Sasaki, T. Isobe, Y. Kameshima, K. Okada, Appl. Catal.

A: Gen. 366 (2009) 148–153.19] A. Nakajima, T. Koike, S. Yanagida, T. Isobe, Y. Kameshima, K. Okada, Appl. Catal.

A: Gen. 385 (2010) 130–135.20] H. Kominami, Y. Ishii, M. Kohno, S. Konishi, Y. Kera, B. Ohtani, Catal. Lett. 91

(2003) 41–47.21] J. Guang, C. Tang, D. Li, H. Haneda, T. Ishigaki, J. Am. Ceram. Soc. 87 (2004)

1358–1361.22] T. Shibata, H. Irie, M. Ohmori, A. Nakajima, T. Watanabe, K. Hashimoto, Phys.

Chem. Chem. Phys. 6 (2004) 1359–1362.23] W. Hu, L. Li, G. Li, C. Tang, L. Sun, Cryst. Growth Des. 9 (2009) 3676–3682.24] P.R. Harvey, R. Rudham, S. Ward, J. Chem. Soc. Faraday Trans. 79 (1983)

1381–1390.

28] M. Gratzel, F.P. Rotzinger, Chem. Phys. Lett. 118 (1985) 474–477.29] T. Yamase, Chem. Rev. 98 (1998) 307–325.