Ts

Wa

b

c

d

e

a

ARRAA

KSVTB

1

vgegTtcnitfimoh

T

0d

Applied Catalysis A: General 407 (2011) 224– 230

Contents lists available at SciVerse ScienceDirect

Applied Catalysis A: General

jo u r n al hom epage: www.elsev ier .com/ locate /apcata

oluene decomposition using silver vanadate/SBA-15 photocatalysts: DRIFTStudy of surface chemistry and recyclability

en-Sheng Changa, Yu-Chu M. Lib, Tsair-Wang Chungc, Yung-Sen Lind, Chao-Ming Huange,∗

Green Energy & Environment Research Labs, Industrial Technology Research Institute, Hsinchu, TaiwanDepartment of Mechanical Engineering, Southern Taiwan University, Tainan, TaiwanDepartment of Chemical Engineering/R&D Center for Membrane Technology, Chung Yuan Christian University, Taoyuan, TaiwanDepartment of Chemical Engineering, Feng Chia University, Taichung, TaiwanDepartment of Materials Engineering, Kun Shan University, Tainan, Taiwan

r t i c l e i n f o

rticle history:eceived 13 May 2011eceived in revised form 27 August 2011ccepted 29 August 2011vailable online 3 September 2011

eywords:BA-15

a b s t r a c t

Silver vanadate (SVO) containing SBA-15 visible-light-driven photocatalyst was synthesized using theincipient wetness impregnation procedure. X-ray diffraction (XRD) results reveal that the SVO/SBA-15powders consisted of three kinds of phase: pure Ag4V2O7 or pure �-Ag3VO4 or mixed phases of Ag4V2O7

and �-Ag3VO4. The mass spectra indicate that the main oxidation intermediate of toluene is benzalde-hyde. The sample loaded with 51 wt% SVO (51SVO/SBA-15) exhibited the best photocatalytic activity. Theresults of two consecutive cyclic runs and regeneration indicate that the accumulation of benzaldehydecauses an irreversible deactivation of P25, but no deactivation of 51SVO/SBA-15. The enhanced photo-

isible-light-driven photocatalystoluene photo-oxidationrønsted and Lewis acids

catalytic activity of 51SVO/SBA-15 is attributed to mixed crystalline phases of Ag4V2O7 and �-Ag3VO4,where �-Ag3VO4 is the major component. In situ diffuse reflectance infrared Fourier transform spec-troscopy (DRIFTS) confirms the presence of Brønsted and Lewis acids on the SVO/SBA-15 composites. Afavorable crystalline phase combined with high intensities of Brønsted and Lewis acids is considered themain cause of the enhanced adsorption capacity, outstanding photoactivity, and long term stability ofthe SVO/SBA-15 composites.

. Introduction

Volatile organic compounds (VOCs), extensively used as sol-ents, aerosol propellants, and raw materials, are considered asreat contributors to atmospheric pollution, with some consid-red toxic. Due to increasing eco-awareness, regulations set byovernments regarding the emission of VOCs have grown stricter.he most conventional reduction method of VOCs is adsorptionechnology, which transforms air pollutants into another phase,reating secondary pollution. Since the adsorption of VOCs doesot actually destroy the pollutants, the regeneration of adsorbents

s a problem after saturation. Among the available alternativeso adsorption, photocatalytic oxidation (PCO) is quite promisingor VOC reduction as it has the potential to completely mineral-ze VOCs to CO2 and H2O, which may meet the requirements of

ore stringent VOC emission control in the future. A vast numberf studies related to titanium dioxide, a UV-active photocatalyst,ave been conducted. Recently, a lot of research effort has been

∗ Corresponding author at: 949 Da-Wan Rd., Yung-Kang Dist., Tainan 71003,aiwan. Tel.: +886 6 2050530; fax: +886 6 2050493.

E-mail address: charming@mail.ksu.edu.tw (C.-M. Huang).

926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.apcata.2011.08.043

© 2011 Elsevier B.V. All rights reserved.

devoted to loading TiO2 species onto or incorporating them into themesoporous silica called Santa Barbara Amorphous-15 (SBA-15) tocombine adsorption and photocatalysis for the rapid removal ofVOCs [1–4]. However, the use of titania-SBA-15 composites is hin-dered by some major shortcomings. First, TiO2 has low efficiencyof degrading VOCs under visible-light illumination due to its wideband gap, which is in the range of 3.0–3.2 eV. Second, the mostgeneral method for synthesizing TiO2 is the sol–gel method, whichrequires high-temperature calcination (673 K or higher) to obtaingood crystallinity. Moreover, the reaction intermediates can be insome cases more toxic and/or stable than the parent VOC duringthe PCO process. Therefore, the development of visible-light-activephotocatalysts that can minimize the amount of intermediates dur-ing the photocatalytic process is desirable.

The present work demonstrates the preparation of silvervanadate/SBA-15 composites (SVO/SBA-15) via hydrothermal syn-thesis using a post-synthesis step without high-temperaturecalcination. The photodecomposition of toluene was selected as amodel reaction because toluene is a common atmospheric indoor

and industrial air pollutant. Toluene has usually shown high ini-tial conversions followed by very low steady-state conversions,which has been attributed to catalyst deactivation during the PCOprocess [5]. In order to determine the stability of the SVO/SBA-15

lysis A

cw

2

2

Mto3wT2i(

vt(tNuamwnfaarwwciS

2

sCrtso(tDt(cSpt3tfpaspt(

W.-S. Chang et al. / Applied Cata

omposites, the recyclability test of the SVO/SBA-15 compositesas investigated using DRIFTS.

. Experimental procedure

.1. Preparation of photocatalyst

SBA-15 was synthesized with Pluronic P123 (EO20PO70EO20,av = 5800; Aldrich) and tetraethylorthosilicate (TEOS) according

o a previous report [6]. Briefly, 4.0 g of P123 was dissolved in 30 gf de-ionized water and 120.0 g of HCl solution (2 M) with stirring at13 K for 2 h. Then, 8.5 g of TEOS was added into the P123 solution,hich was stirred for another 22 h until a white gel precipitated.

he gel was transferred to a Teflon bottle and heated at 403 K for4 h. The precipitate was filtered, washed several times with de-

onized water, dried overnight at 373 K, and then calcined at 773 Kheating rate of 1 K/min) for 4 h in air.

The silver-vanadate-loaded SBA-15 (SVO/SBA-15) was preparedia the incipient wetness impregnation procedure. In the prepara-ion process, 0.204 g AgNO3 was dissolved in urea aqueous solution120 g H2O, 0.577 g urea) with stirring at room temperature for 0.5 ho obtain solution A. Solution B was prepared by mixing 0.047 gH4VO3 with 0.3 g SBA-15 in de-ionized water at 343 K for 1 hnder an ultrasonic bath. A suspension formed when solution A wasdded dropwise to solution B under vigorous stirring for 1 h. Theolar composition of the suspension AgNO3/NH4VO3/CO(NH2)2as 3.0/1.0/12.0. The suspension was titrated to pH 7 using ammo-ia solution, followed by additional stirring at room temperature

or 24 h. Finally, the as-obtained suspension was transferred into Teflon-lined autoclave with hydrothermal treatment (temper-ture: 413 K, time: 4 h). After the hydrothermal procedure, theesulting precipitates were collected and washed with de-ionizedater three times, and then dried at 353 K for 12 h. These samplesere denoted xSVO/SBA-15, where x represents the weight per-

entage of silver vanadate (wt%). The sample synthesized underdentical conditions without the addition of SBA-15 was denotedVO.

.2. Sample characterization

The X-ray diffraction (XRD) patterns of the powders were mea-ured using an X-ray diffractometer (PANalytical X’Pert PRO) withu radiation (� = 0.15418 nm) in the 2� range of 20–60◦. High-esolution transmission electron microscopy (HRTEM) images ofhe samples were observed on a Philips Tecnai G2 F20 micro-cope equipped with energy-dispersive X-ray spectroscopy (EDX)perated at an accelerating voltage of 200 kV. PhotoluminescencePL) spectra were recorded by a fluorescence spectrophotome-er (Dongwoo, Optron) under excitation at 325 nm. In situRIFTS measurements were performed using a FTIR spectrome-

er (PerkinElmer, Spectrum GX) and a diffuse reflectance accessoryHarrick Scientific, DRP-PE 9) with a temperature- and atmosphere-ontrolled high-temperature low-pressure reaction cell (Harrickcientific, HVC-DRP 3). Prior to the IR measurements, the sam-les were dehydrated under vacuum from room temperatureo 523 K at 10 K/min in N2 flow (30 ml/min), held at 523 K for0 min, and then cooled to 303 K. DRIFT spectra with a resolu-ion of 4 cm−1 were collected in the interval of 2800–4000 cm−1

or surface hydroxyl functional groups [7]. The surface area andore volume of the as-prepared samples were determined using

volumetric sorption analyzer (Micromeritics, ASAP 2020). The

amples were degassed at 473 K under vacuum conditions for aeriod of at least 4 h prior to measurements. The nitrogen adsorp-ion/desorption isotherms were measured over a relative pressureP/P0) range of approximately 10−3 to 0.995. The surface areas

: General 407 (2011) 224– 230 225

were calculated using the Brunauer–Emmett–Teller (BET) methodin the relative pressure range of 0.06–0.2. The pore size distri-butions were determined from the analysis of the adsorptionisotherm using the Barret–Joyner–Halenda (BJH) algorithm. Thetotal pore volumes were estimated from the adsorbed N2 amountat P/P0 = 0.973.

2.3. Photocatalytic activity evaluation with GC–MS and DRIFTS

The photocatalytic oxidation of toluene was performed in situin an IR cell with ZnSe windows. An LED lamp, with a wave-length ranging from 430 to 620 nm with a photon intensity of4 mW/cm2, was used as the visible-light source [8]. A gaseoustoluene/N2 mixture was generated, corresponding to the targettoluene concentration of 200 ppmv. Prior to the experiments, thesamples were pre-treated by heating and flushing with N2 flow(20 ml/min) from room temperature to 523 K, held at 523 K for30 min, and then cooled to 303 K. The procedure for all PCO experi-ments was as follows: (1) toluene/N2 flow was introduced into thephotoreactor at a constant flow rate of 20 ml/min; (2) when thephotoreactor inlet and outlet toluene concentrations were approx-imately equal, the LED lamp was turned on and the toluene/N2flow was stopped; (3) the photoreactor was flushed with the N2flow at 5 ml/min for 30 min, and then the N2 flow was stopped.The gaseous products (intermediates) were analyzed on-line byGC–MS. GC–MS analyses were carried out on a gas chromato-graph (GC, Perkin-Elmer, Clarus 500) using a 60-m DB-624 capillarycolumn, coupled with a quadrupole mass spectrometer (MS, SRS,QMS 300) at regular intervals. The spectra of the adsorbed specieson the catalyst surface were recorded under both darkness andillumination. The spectra of the catalyst and the reaction intermedi-ates during the reaction were expressed in units of Kubelka–Munk(K–M).

2.4. DRIFT characterization of NH3 species

To determine the types of acid site present on the sam-ples, temperature-programmed desorption of ammonia (NH3) wascarried out using DRIFT. Ammonia usually provides the probemolecules in spectroscopic experiments to determine the type ofacid site in heterogeneous catalysis: Brønsted sites or Lewis sites.Prior to the experiments, the samples were heated in situ fromroom temperature to 523 K at 10 K/min in N2 flow (30 ml/min),held at 523 K for 30 min, and then cooled down to 303 K. The sam-ples were saturated at 303 K with a gas mixture of 5% NH3 in N2(30 ml/min) for 30 min. At the end of the saturation process, thesamples were flushed with N2 flow (30 ml/min). Then, the sam-ples were heated again at a heating rate of 10 K/min from 30 K to523 K and held at 523 K for 60 min. The DRIFT spectra with a reso-lution of 4 cm−1 were collected in the interval of 1200–1700 cm−1

for determining surface acidity.

3. Results and discussion

3.1. X-ray diffraction analysis

The XRD patterns of SVO and SVO-SBA-15 are shown in Fig. 1.Fig. 1a shows the low-angle XRD patterns of SBA-15 and SVO/SBA-15 samples. The XRD pattern of SBA-15 shows three well resolvedpeaks: a sharp peak at 1.0◦ and two weak peaks near 1.5◦ and2.0◦, respectively, which can be indexed to (1 0 0), (1 1 0), and(2 0 0) reflections of an ordered hexagonal P6mm space group,

respectively. For SVO/SBA-15 samples, the XRD peaks decreased inintensity, which was probably caused by the decreasing scatter con-trast between pore walls and the pore space with the introductionof SVO. For the samples with high SVO content, the intensities of the

226 W.-S. Chang et al. / Applied Catalysis A: General 407 (2011) 224– 230

ptcliwbXaT�v1

To better understand the formation of SVO nanocrystals on themesopores and explore the effect of SVO on the pore structure ofSVO/SBA-15 photocatalyst, HRTEM images were recorded. HRTEM

Table 1Specific surface areas and pore properties of P25, SVO, SBA-15, and SVO/SBA-15.

Sample SBET (m2 g−1) Pore volume(cm3 g−1)

Pore size(nm)

P25 56 0.25 17.52SVO 2 0.002 7.30SBA-15 787 1.09 7.6117SVO/SBA-15 193 0.70 14.9034SVO/SBA-15 141 0.54 16.1451SVO/SBA-15 99 0.35 15.32

Fig. 1. (a) Low-angle and (b) wide-angle XRD patterns of SVO and SVO/SBA-15.

eaks are much weaker but can still be discerned, demonstratinghat the uniform porous structure of SBA-15 is retained. The unit-ell parameter (a0) of SVO/SBA-15 catalysts (ca. 13.0–14.0 nm) isarger than that of SVO (11.8 nm). This may be ascribed to the partialncorporation of SVO into the framework of SBA-15. Fig. 1b shows

ide-angle XRD patterns of SVO/SBA-15 samples and SVO. As cane seen, the SVO and SVO-SBA-15 samples have three kinds ofRD patterns, assigned to the pure Ag4V2O7, to the pure �-Ag3VO4,nd to the mixed phases of Ag4V2O7 and �-Ag3VO4, respectively.he pure SVO and 51SVO-SBA-15 samples had mixed phases of

-Ag3VO4 and Ag4V2O7. When the weight percentages of silveranadates were lower than 51%, pure �-Ag3VO4 was obtained for7 and 34SVO/SBA-15 samples. When the weight percentages of

Fig. 2. N2 adsorption–desorption isotherms of SBA-15 and SVO/SBA-15 samples.

silver vanadates were increased to 60 and 70%, all peaks attributedto �-Ag3VO4 disappeared and pure Ag4V2O7 was observed.

3.2. Porosity and surface area characterization

Fig. 2 shows the nitrogen adsorption-desorption isotherms ofSBA-15 and SVO/SBA-15 samples. The N2 adsorption–desorptionisotherms of SBA-15 show type IV adsorption with a H1 hysteresisloop, indicating a mesoporous characteristic. For the SVO/SBA-15samples, a sharp inflection in the relative pressure (P/P0) rangeof 0.65–0.95, corresponding to capillary condensation within uni-form mesopores, was observed. Compared to SBA-15, the capillarycondensation of SVO/SBA-15 shifted to higher relative pressure,indicating an increase of the pore diameter when SVO was loadedinto/onto SBA-15. The textural properties of SBA-15 support andSVO/SBA-15 samples are listed in Table 1. It can be seen thatthe specific surface area and pore volumes of SVO/SBA-15 sam-ples significantly decrease with increasing silver vanadate loading,whereas the pore diameter increases. When the SVO/SBA-15samples were synthesized using a post-synthesis method, thesilver and vanadate species reacted on the surface of SBA-15;thus, the BET surface and pore volumes of SVO/SBA-15 com-posites decreased with increasing amount of SVO. The averagepore diameter increased for SVO/SBA-15, which might be due tothe small pores of SBA-15 being obstructed by silver vanadatenanoparticles.

3.3. Morphology

60SVO/SBA-15 96 0.34 13.6170SVO/SBA-15 85 0.33 15.86

W.-S. Chang et al. / Applied Catalysis A: General 407 (2011) 224– 230 227

) SBA-

iTc7tbSmtmituvptSSisswartaavsm

Fig. 3. HRTEM images of (a

mages of SBA-15 and 51SVO/SBA-15 samples are shown in Fig. 3.he highly ordered mesoporous channel structure of pure SBA-15an be observed in Fig. 3a, which shows a pore diameter of around–8 nm, in agreement with N2 adsorption–desorption results. Dueo the electronic density contrast of TEM, clear and gray stripes cane observed between empty channels and silica walls. When theVO loading was increased to 51 wt% (Fig. 3b), the ordered array ofesopores remained intact. Silica walls remained white whereas

he SVO particles, with a larger electronic density, turned black. Theorphology results indicate that nanocrystalline SVO was inserted

nto the mesoporous channels of SBA-15 when urea was used ashe chelating agent, which indicates a strong interaction betweenrea and metal ions that prevents the precipitation of silver andanadium ions before the formation of silver vanadate–silica com-osites. EDX elemental analysis of a selected point of SBA-15 andhe SBA-15 supported catalyst (Fig. 3) confirms the presence ofVO particles in the center of a SBA-15 channel. Fig. 4 showsEM images for SBA-15 support and SVO/SBA-15 composites. SEMmage (Fig. 4a) of parent SBA-15 sample presents well-definedausage-like particles of approximately 1.5 �m in length. Theseausage-like particles cluster together to form rope-like aggregates,hich are parallel straight mesochannel arraying along the long

xis. The SEM images in Fig. 4b–d demonstrate that the orderedope-like domains are preserved after silver-vanadate impregna-ion for 17, 34, and 51SVO/SBA-15 samples. On the other hand, 60nd 70SVO/SBA-15 samples show the presence of an overgrowth

s the irregular ragged particles when high amounts of silver-anadate nanoparticles are deposited (Fig. 4e and f). However, thetacked nanoparticles do not lead to any significant change in theorphology of silica.

15 and (b) 51SVO/SBA-15.

3.4. Adsorption and photocatalytic oxidation of toluene

The adsorption and photodegradation curves of gaseous tolueneover samples were shown in Fig. 5. Prior to photocatalytic exper-iments, the concentration of gaseous toluene rapidly decreasedduring the initial 10 min, which was due to the adsorption oftoluene on catalyst surface. After 30 min, the concentration ofgaseous toluene is increased to the initial state, indicating that theadsorption of toluene reaches equilibrium. In the absence of irradi-ation, the as-prepared SVO/SBA-15 composites except 17SVO/SBA-15 exhibit a much higher adsorption capability of toluene thanthose of SVO and P-25. The toluene adsorption capacity is thusaffected by the silver vanadate content of SVO/SBA-15 composites.It is well known that photocatalytic oxidation of organic pollu-tants in the initial stage follows Langmuir–Hinshelwood kinetics;the L–H model can be simplified to a pseudo-first-order expres-sion: ln(Ce/C) = kt (where Ce and C are the equilibrium concentrationof adsorption and the concentration of VOC at the exposuretime, t, respectively, and k is the apparent rate constant). Thecalculated kapp of toluene decreased in the order: 51SVO/SBA-15 (0.032 min−1) > 34SVO/SBA-15 (0.024 min−1) > 60SVO/SBA-15(0.020 min−1) > 70SVO/SBA-15 (0.019 min−1) > SVO (0.018 min−1)> 17SVO/SBA-15 (0.016 min−1) > P25 (0.015 min−1).

The evolution of gaseous benzaldehyde as a function of irra-diation time was plotted in Fig. 6. During the first 10 min oftoluene photodegradation, the amounts of benzaldehyde pro-

duced for all SVO/SBA-15 samples were higher than those ofSVO and P25. However, at longer reaction times, the amountof benzaldehyde generated for 17SVO/SBA-15 is lower thanthat of SVO. The amount of benzaldehyde decreased in the

228 W.-S. Chang et al. / Applied Catalysis A: General 407 (2011) 224– 230

-15, (

f1p1csAtwoa5oap

Fig. 4. SEM images of various SBA-15 samples: (a) parent SBA-15, (b) 17SVO/SBA

ollowing sequence: 51SVO/SBA-15 > 34SVO/SBA-15 > 60SVO/SBA-5 > 70SVO/SBA-15 ∼ SVO > 17SVO/SBA-15 > P25. A big degree ofhotodegradation of toluene takes place in case of 51SVO/SBA-5. Generally speaking, photocatalytic activity is related to therystalline phase. Konta et al. [9] reported that �-Ag3VO4 showedtronger photocatalytic activity than those of �-AgVO3 andg4V2O7 for O2 evolution from an aqueous silver nitrate solu-

ion under visible-light irradiation. For the SVO/SBA-15 series, itas observed that the 51SVO/SBA-15 sample, with mixed phases

f �-Ag3VO4 and Ag4V2O7, exhibits the highest photocatalyticctivity, whereas 34SVO/SBA-15 shows lower activity than that of

1SVO/SBA-15 even though 34SVO/SBA-15 has a high crystallinityf �-Ag3VO4. An increasing number of studies have indicated that

bicrystalline framework of anatase and rutile show much betterhotocatalytic activity than that of pure anatase TiO2 [10–12].

c) 34SVO/SBA-15, (d) 51SVO/SBA-15, (e) 60SVO/SBA-15, and (f) 70SVO/SBA-15.

It is well known that the adsorption capacity is mainly deter-mined by the specific surface area and nature of the surface ofthe photocatalyst. Studies have shown that the much higher spe-cific surface area and pore volume of TiO2-containg mesoporoussilica composites compared to those of pure titania are benefi-cial for the adsorption of organic pollutants [13–15]. As shownin Fig. 5, 51SVO/SBA-15 exhibited the highest adsorption capac-ity of toluene instead of 17SVO/SBA-15. However, 17SVO/SBA-15had much higher specific surface area and pore volume than thoseof 51SVO/SBA-15; therefore, the toluene adsorption capacity isthus greatly influenced by the silver vanadate content, rather

than the specific surface area. To further characterize the sur-face attributes that affect toluene adsorption, the surface acidity,the existence of Brønsted and Lewis acid sites, of the sampleswas examined using DRIFTS to detect ammonia adsorption on

W.-S. Chang et al. / Applied Catalysis A: General 407 (2011) 224– 230 229

Fig. 5. Time-dependent concentrations of gaseous toluene in the dark and undervisible light.

Fp

tshatt3btso

F

ig. 6. Time-dependent concentrations of gaseous benzaldehyde during toluenehotocatalytic degradation with P25, SVO, and SVO/SBA-15.

he photocatalysts surface. A literature survey indicates that IRpectroscopic studies of ammonia adsorbed on solid surfacesave made it possible to distinguish between Brønsted and Lewiscid sites of a catalyst [16–21]. Before the DRIFT measurements,he samples were saturated with NH3/N2, flushed with N2 flowo remove physically adsorbed ammonia, and then heated from03 K to 523 K. The intensity of chemisorption was determinedased on the irreversible adsorption of ammonia. Fig. 7 shows

he IR spectra of ammonia adsorbed onto SVO and SVO/SBA-15amples which were heated at 523 K for 30 min. The existencef NH4

+ ions adsorbed onto Brønsted acid sites of the silver

ig. 7. IR spectra of NH3 adsorbed onto SVO and SVO/SBA-15 samples at 523 K.

Fig. 8. Normalized PL intensity of the samples measured at 300 K.

vanadate surfaces is supported by the presence of a band at1425 cm−1 due to the asymmetric deformation mode [16–18]. Theband at 1604 cm−1 is assigned to NH3 coordinately bonded toLewis acid sites [19–21]. It can be concluded that both Brønstedand Lewis acid sites exist on the surfaces of SVO and SVO/SBA-15.The intensities of Brønsted and Lewis acid sites, detected at 1425and 1604 cm−1, respectively, follow the sequence: 51SVO/SBA-15 > 34SVO/SBA-15 > 17SVO/SBA-15 ∼ SVO. The trend of Brønstedand Lewis acid sites is in agreement with that of the adsorp-tion capacity. That is, higher intensities of Brønsted and Lewisacids of silver vanadate indicate a larger adsorption capacity fortoluene.

3.5. Photoluminescence spectra

PL spectra have been widely used to disclose the migra-tion, transfer, and recombination processes of the photogeneratedelectron-hole pairs in the semiconductor particles [22]. Fig. 8shows the PL spectra with 325 nm excitation wavelength in therange of 400–600 nm for P25, SVO, and SVO loaded compos-ites at room temperature. The PL intensity of these samplesdecreases in the order of P25 > SVO > 17SVO/SBA-15 > 60SVO/SBA-15 > 70SVO/SBA-15 > 34SVO/SBA-15 > 51SVO/SBA-15. The PL ofSVO/SBA-15 samples show obvious decrease in the intensity of PLspectra as compared to SVO, indicating the recombination of pho-toelectrons and holes is efficiently suppressed in the compositesemiconductors. The intensity of photoluminescence spectra cor-responds to the recombination rates of the holes formed in the O2pband and the electron in the V3d band. The slower recombinationprocess of photogenerated charges (the less the PL intensity) canfacilitate the enhancement of photocatalytic activity of SVO/SBA-15composite.

3.6. DRIFTS study for recyclability of photocatalyst

During the photocatalytic reaction, organic pollutants aresupposed to be adsorbed or concentrated in situ and then photo-degraded in vitro under factory process; therefore, it is importantto evaluate the catalyst recyclability in situ. Since 51SVO/SBA-15 had the highest apparent rate constant, it was chosen for therecyclability test. A complete cycle was composed of adsorptionunder darkness for 30 min and then an illumination reaction for60 min. After the 1st cycle, the catalyst was regenerated in situby further irradiating the LED light for 1 h. Then, the catalyst wasreused in situ without any washing or thermal treatment for the

2nd cycle. The reliability test was stopped when the catalyst wasdeactivated as ascertained from DRIFTS and color analyses. The cat-alyst reliability results were expressed as IR spectra. The resultsof P25 and 51SVO/SBA-15 are shown in Fig. 9. After exposing P25

230 W.-S. Chang et al. / Applied Catalysis A

Fss

taatctaotptDthmbttha5T2ba

[[

[

[[

[[[

[

[[[[

[

ig. 9. DRIFT spectra of reliability test of (a) P25 and (b) 51SVO/SBA-15. VIS 1 denotespectrum obtained under visible-light irradiation in the 1st cycle, and VIS 2 denotespectrum obtained under visible-light irradiation in the 2nd cycle.

o LED light for 60 min, a decrease of the 1635, 1606, 1496, 1447,nd 1378 cm−1 bands was observed, with new bands appearingt 1684 and 1582 cm−1 (Fig. 9a). This suggests that a fraction ofhe adsorbed toluene was photo-oxidized to benzaldehyde. Theolor of P25 became ocher after the 1st photocatalytic run. Afterhe 2nd cycle, the intensities of the bands of toluene at darkdsorption for 30 min (Dark 2) were almost the same with thosef illumination reaction for 60 min (VIS 2). Since the intensity ofoluene was not reduced after illumination, the P25 had lost itshotocatalytic activity. Several studies have observed TiO2 deac-ivation during the photocatalytic oxidation of toluene [23–26].irect evidence comes from the color change of TiO2 from white

o yellow or yellow–brown after several hours of irradiation. Itas been reported that the deactivation results from the accu-ulation of strongly adsorbed species, such as benzaldehyde and

enzoic acid, on the TiO2 surface [21,23–25]. However, deactiva-ion does not occur for 51SVO/SBA-15. An obvious decrease ofhe toluene bands and an intense appearance of the benzalde-yde bands were clearly observed for Dark 1 vs. VIS 1 (1st cycle)nd Dark 2 vs. VIS 2 (2nd cycle), respectively (Fig. 9b). After1SVO/SBA-15 was exposed to LED light for 1 h, it regenerated.

he color of 51SVO/SBA-15 remained light orange color after thend cycle. This proves that the 51SVO/SBA-15 photocatalyst is sta-le and can be regenerated and recycled without losing its originalctivity.

[

[[

: General 407 (2011) 224– 230

4. Conclusion

Environmentally friendly SVO/SBA-15 photocatalysts weresynthesized using hydrothermal synthesis without a high-temperature calcination process. XRD and HRTEM results indicatethat SVO/SBA-15 composites had three kinds of phase and that SVOdispersed well in the channels or on the surface of SBA-15 withoutaffecting the SBA-15 mesoporous structure, respectively. The massspectra indicate that the main oxidation product of toluene is ben-zaldehyde, which is strongly adsorbed on the surface of the catalyst.The SVO/SBA-15 photocatalyst has a much higher photodegrada-tion ability for toluene than does P25. Based on the recyclability test,the accumulation of benzaldehyde on the P25 surface appears to beresponsible for its deactivation; no deactivation was observed for51SVO/SBA-15, which was still stable after 2nd cycle as evidencedby DRIFTS analysis. The existence of the acidity of the SVO/SBA-15composites was confirmed using DRIFTS. A bicrystalline frameworkof �-Ag3VO4 and Ag4V2O7 and high intensities of Brønsted andLewis acids are responsible for the enhanced adsorption capac-ity, superior photoactivity, and long term stability of SVO/SBA-15composites.

Acknowledgements

The authors are grateful to the National Science Council ofTaiwan (grant no. NSC 98-2221-E-168-008) and the Bureau ofEnergy, Ministry of Economic Affairs (grant no. 7455VH7200), forsupporting this study.

References

[1] K. Su, Z. Li, B. Cheng, K. Liao, D. Shen, Y. Wang, J. Mol. Catal. A 315 (2010) 60–68.[2] S. Zhang, D. Jiang, T. Tang, J. Li., Y. Xu, W. Shen, J. Xu, F. Deng, Catal. Today 158

(2010) 329–335.[3] T. Klimov, O. Gutierrez, L. Lizama, J. Amezcua, Microporous Mesoporous Mater.

133 (2010) 91–99.[4] R. Zukerman, L. Vradman, L. Titelman, L. Zeiri, N. Perkas, A. Gedanken, M.V.

Landau, M. Herskowitz, Mater. Chem. Phys. 122 (2010) 53–59.[5] K. Demeestere, J. Dewulf, H. Van Langenhove, Crit. Rev. Environ. Sci. Technol.

37 (2007) 489–538.[6] H. Sun, J. Han, Y. Ding, W. Li, J. Duan, P. Chen, H. Lou, X. Zheng, Appl. Catal. A:

Gen. 390 (2010) 26–34.[7] C.M. Huang, G.T. Pan, Y.C.M. Li, M.H. Li, T.C.K. Yang, Appl. Catal. A: Gen. 358

(2009) 164–172.[8] C.M. Huang, G.T. Pan, P.Y. Peng, T.C.K. Yang, J. Mol. Catal. A: Chem. 327 (2010)

38–44.[9] R. Konta, H. Kato, H. Kobayoshi, A. Kudo, Phys. Chem. Chem. Phys. 5 (2003)

3061–3065.10] Z. Ding, G.Q. Lu, P.F. Greenfield, J. Phys. Chem. B 104 (2000) 4815–4820.11] D. Gumy, S.A. Giraldo, J. Rengifo, C. Pulgarin, Appl. Catal. B: Environ. 78 (2008)

19–29.12] G. Liu, Z. Chen, C. Dong, Y. Zhao, F. Li, G.Q. Lu, H.M. Cheng, J. Phys. Chem. B 110

(2006) 20823–20828.13] Y. Chen, Y. Huang, J. Xiu, X. Han, X. Bao, Appl. Catal. A: Gen. 273 (2004) 185–191.14] T. Hoshikawa, T. Ikebe, M. Yamada, R. Kikuchi, K. Eguchi, J. Photochem. Photo-

biol. A: Chem. 184 (2006) 78–85.15] J. Mo, Y. Zhang, Q. Xu, R. Yang, J. Hazard. Mater. 168 (2009) 276–281.16] G. Ramis, L. Yi, G. Busca, Catal. Today 28 (1996) 373–380.17] M.A. Centeno, I. Carrizosa, J.A. Odriozola, Appl. Catal. B: Environ. 19 (1998)

67–73.18] L.C. Chen, G.T. Pan, T.C.K. Yang, T.W. Chung, C.M. Huang, J. Hazard. Mater. 178

(2010) 644–651.19] J.R. Sohn, W.C. Park, Appl. Catal. A: Gen. 239 (2003) 269–278.20] J.R. Sohn, S.H. Lee, Appl. Catal. A: Gen. 266 (2004) 89–97.21] J.R. Sohn, J.S. Han, J. Ind. Eng. Chem. 11 (2005) 439–448.22] H. Yamashita, Y. Ichihashi, S.G. Zhang, Y. Matsumura, Y. Souma, T. Tatsumi, M.

Anpo, Appl. Surf. Sci. 121 (1997) 305.23] G. Martra, S. Coluccia, L. Marchese, V. Augugliaro, V. Loddo, L. Palmisano, M.

Schiavello, Catal. Today 53 (1999) 695–702.24] L. Cao, Z. Gao, S.L. Suib, T.N. Obee, S.O. Hay, J.D. Freihaut, J. Catal. 196 (2000)

253–261.25] M.C. Blount, J.L. Falconer, Appl. Catal. B: Environ. 39 (2002) 39–50.26] M. Lewandowski, D.F. Ollis, Appl. Catal. B: Environ. 43 (2003) 309–327.