Msa

TRa

b

a

ARRAA

KMMMS2PT

1

aaptpasoCpCnptt

0h

Applied Catalysis A: General 467 (2013) 224–234

Contents lists available at ScienceDirect

Applied Catalysis A: General

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

etathesis of 2-butene to propylene over W-mesoporous molecularieves: A comparative study between tungsten containing MCM-41nd SBA-15

azul Islam Bhuiyana, Palani Arudraa, Muhammad Naseem Akhtara, Abdullah M. Aitania,aed H. Abudawoudb, Mohammed A. Al-Yamia, Sulaiman S. Al-Khattafa,∗

Center of Research Excellence in Petroleum Refining and Petrochemicals, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi ArabiaResearch & Development Center, Saudi Aramco, Dhahran 31311, Saudi Arabia

r t i c l e i n f o

rticle history:eceived 12 February 2013eceived in revised form 8 July 2013ccepted 19 July 2013vailable online 29 July 2013

eywords:esoporous silica

a b s t r a c t

Mesoporous silica (MCM-41 and SBA-15) having tungsten oxide in the framework was synthesized byhydrothermal crystallization process using CTAB and P123 as structure directing agents, respectively.Tungsten oxide was supported on the MCM-41 and SBA-15 by wet impregnation method for a compar-ative study. The synthesized materials were characterized by XRD, N2 adsorption–desorption, NH3-TPD,pyridine-FT-IR, UV–vis DRS and Raman spectroscopy techniques. The catalysts were used to performmetathesis reaction of 2-butene in a fixed-bed reactor at different temperatures and under atmosphericpressure. It was observed that catalysts having tungsten oxide in the framework of mesoporous silica

etathesisCM-41

BA-15-Buteneropyleneungsten oxide

exhibited higher activity as compared to tungsten oxide impregnated catalysts. The results revealed thatWO3-MCM-41(30) catalyst has much higher activity at a low temperature (450 ◦C) as compared to WO3-SBA-15(30). At 550 ◦C, WO3-MCM-41(30) and WO3-SBA-15(30) showed comparable activity with highestpropylene yield of 39 mol% and 37 mol%, respectively. The higher metathesis activity of WO3-MCM-41catalysts can be related to the well-dispersed active tetrahedral tungsten oxide species as evidenced fromXRD, UV–vis DRS and Raman spectroscopy results.

© 2013 Published by Elsevier B.V.

. Introduction

The demand for propylene is increasing for downstreampplications such as polypropylene, acrylonitrile, cumene, oxo-lchohols, propylene oxide, acrylic acid and isopropyl alcohol andolygas chemicals [1]. Several on-purpose propylene productionechnologies are being developed to meet the high demand forropylene which is usually produced from naphtha steam crackingnd refinery fluid catalytic cracking (FCC) units [2]. Olefin metathe-is is an important organic reaction that transforms less desiredlefins to higher desired ones through rearrangement of a pair of

C bonds and currently it is widely used for the production ofropylene from low-value C4 olefins [3,4]. Currently, the Olefinsonversion Technology (OCT) process licensed by Lummus Tech-ology is the only commercial olefin metathesis technology in the

etrochemical industry [5]. The key step in olefin metathesis ishe intermediacy of metal carbene active species and consequentlyhe formation of unstable metallocyclobutane. Both homogeneous

∗ Corresponding author. Tel.: +966 3 860 1429; fax: +966 3 860 4234.E-mail address: skhattaf@kfupm.edu.sa (S.S. Al-Khattaf).

926-860X/$ – see front matter © 2013 Published by Elsevier B.V.ttp://dx.doi.org/10.1016/j.apcata.2013.07.034

and heterogeneous catalysts are used in the metathesis reaction.However, heterogeneous catalysts are widely accepted due to theadvantage of higher thermal stability, easy separation from thereaction products and regeneration [6]. The most successful het-erogeneous catalysts for olefin metathesis are supported rhenium,molybdenum and tungsten oxides [7].

Rhenium oxide catalysts supported on alumina orsilica–alumina are the most active and selective under mildconditions [8]. However, the cost of rhenium compounds are high[5], the catalyst activity with low Re loading (<10 wt%) is poor [9]and the catalysts are easily deactivated [10]. At the same time,molybdenum oxide catalysts supported on alumina [11], siliceousmaterial [12] and silica–alumina [13] have received much attentionbecause of their relatively low price, and their high activity undermild conditions. On the other hand, tungsten oxide supportedon SiO2 [2,14,15], �-Al2O3 [16] and �-Al2O3-HY [1,17,18] requiremuch higher temperature (250–500 ◦C) to achieve acceptablemetathesis activity. Nevertheless, due to long online lifetime,

considerable resistance to poisons and continuous regenerationwithout negative effect on catalyst structure, tungsten based cata-lysts remain the most attractive for commercial use [19]. Catalyticactivity of tungsten oxide catalysts depends on the oxidation

lysis A

s[[

aimtrtdemptdsMadrfpm

msaa

2

2

Taaawp5C

2

s

2

utSrwaisofP

2

s

T.I. Bhuiyan et al. / Applied Cata

tate of tungsten species [19], the content of tungsten oxide2,19,20], the properties of support and the pretreatment condition21].

Mesoporous molecular materials with their large surfacerea and large void volume, exhibit high catalytic activity afterntroducing of heteroatoms into the framework [22]. SBA-15 is a

esoporous silica having large pores (5–20 nm), hexagonal struc-ure, large channel diameter and it is thermally, hydro-thermallyobust due to its thick framework walls (3–6 nm) [23]. Deposition ofungsten into SBA-15 activated the support which allowed higheriffusion rate [24]. Chen et al. reported that, tungsten oxide wasffectively incorporated into the framework of SBA-15 for C4 olefinetathesis exhibiting 92.6% conversion of 1-butene and 27.0%

ropylene yield [25]. Direct tungsten incorporation using hydro-hermal method has advantages such as high tungsten speciesispersibility and strong interaction between tungsten species andilica compared to conventional impregnation method [14,25,26].esoporous silica, MCM-41, shows a highly ordered hexagonal

rray of one-dimensional channel with a very narrow pore sizeistribution [5]. To the best of our knowledge, few publicationseported tungsten distribution in MCM-41 and none were testedor 2-butene metathesis [27]. MCM-41 has been used as a sup-ort for Mo and Ru-based metathesis catalysts for 1-heptene crossetathesis [28].In this study, we have focused on three main effects on 2-butene

etathesis over tungsten supported SBA-15 and MCM-41; (i) theynthesis method and support architecture of tungsten oxide cat-lyst, (ii) the concentration of active WO3 species on the support,nd (iii) effect of reaction temperature.

. Experimental

.1. Materials

Pluronic 123(EO20PO70EO20), tetraethylorthosilicate (98%,EOS), sodium tungstate (99.99%, NaWO4·2H2O), hydrochloriccid (36%), sodium metasilicate (98%, Na2SiO3), cetyltrimethyl-mmonium bromide (98%, CTAB), anhydrous ethyl acetate (99.8%)nd ammonium metatungstate [99%, (NH4)6H2W12O40·xH2O]ere purchased form Sigma–Aldrich and used without any furtherurification. 2-Butene with 99.9% purity (50% cis-2-butene and0% trans-2-butene) was purchased from Saudi Industrial Gasompany, Saudi Arabia.

.2. Support synthesis

Two types of mesoporous supports, SBA-15 and MCM-41 wereynthesized using the hydrothermal method.

.2.1. Mesoporous SBA-15Hexagonally ordered mesoporous SBA-15 was synthesized

sing triblock co-polymer P123 (EO20PO70EO20) as structure direc-ing agent and tetraethylorthosilicate (TEOS) as silica precursor.ynthesis of SBA-15 was performed according to the procedureeported in literature [25]. In a typical synthesis, 6 g of Pluronic 123as dissolved to 210 ml of 2 M HCl and stirred at 40 ◦C to achievetransparent solution. After 3–4 h, 12.8 g TEOS was quickly added

nto the solution. The solution was stirred at 95 ◦C for 2 days. Theolid products obtained were washed with water and dried at 80 ◦Cvernight. The template was removed by calcination in air at 550 ◦Cor 5 h. The molar ratio of the gel composition is 1 SiO2:0.0167123:5.82 HCl:190 H2O.

.2.2. Mesoporous MCM-41Mesoporous support MCM-41 was synthesized by using the

odium metasilicate (Na2SiO3) and cetyltrimethyl-ammonium

: General 467 (2013) 224–234 225

bromide (CTAB) as a source of silica and structure directing agent,respectively. Synthesis was performed according to the methoddescribed in literature [29]. 9.96 g of sodium metasilicate with150 ml H2O was stirred at 85 ◦C for 30 min. Then, 9.8 g of CTAB wasadded to the solution and stirred for 10 min. 30 ml ethyl acetatewas rapidly added into the solution under vigorous stirring. Thestirring was continued for 24 h at 85◦ C. The solid products wererecovered by filtration. After washing with deionized water sev-eral times, the solid was air-dried and then calcined in air at 600 ◦Cfor 4 h to remove the template. In this procedure, cheap and con-venient sodium silicate was used as an inorganic silica precursorinstead of expensive organic TEOS and ethyl acetate was employedas the acid producing precursor instead of a conventional strongacid, such as HCl or H2SO4. The synthesis gel molar composition is1 SiO2:0.335 CTAB:3.8 EtOAc:120H2O.

2.3. Catalyst synthesis

Several catalysts were synthesized by supporting tungsten onmesoporous silica SBA-15 & MCM-41 using hydrothermal crystal-lization and wet impregnation procedures (Table 1). The details ofthe synthesis procedure are given below.

2.3.1. W containing SBA-15 by direct hydrothermal method(WO3-SBA-15)

Sodium tungstate (NaWO4·2H2O) was used as the source oftungsten ion for the synthesis of WO3-SBA-15 by direct hydro-thermal method. In a typical synthesis of WO3-SBA-15, tungstenwas incorporated into the framework of mesoporous supportSBA-15 depending upon the Si/W molar ratio. Pluronic 123 wasdissolved in HCl and TEOS was added with vigorous stirring. Then,calculated amount of sodium tungstate solution was added tothe solution and stirred at 95 ◦C for 3 days under hydrothermalconditions. The resultant solid was filtered, dried and calcined at550 ◦C for 5 h. The catalysts obtained in this way were identifiedas WO3-SBA-15(x), where x represented the molar ratio of sili-con to tungsten (Si/W). The molar ratio of the gel composition is1 SiO2:0.3–0.6 WO3:0.0167 P123:5.82 HCl:190 H2O.

2.3.2. W containing MCM-41 by direct hydrothermal method(WO3-MCM-41)

In a typical synthesis of WO3-MCM-41, the solution of sodiumtungstate was added into the solution of sodium metasilicate andCTAB, depending upon the Si/W molar ratio. Then ethyl acetatewas added as an acid producing precursor and the solution wasstirred for 24 h at 85 ◦C under hydrothermal conditions. After fil-tration, the solid was air-dried and calcined at 600 ◦C for 4 h. Thecatalysts obtained in this way were identified as WO3-MCM41(x),where x represents the molar ratio of silicon to tungsten (Si/W).The synthesis gel molar composition is 1 SiO2:0.3–0.6 WO3:0.335CTAB:3.8 EtOAc:120 H2O.

2.3.3. Impregnation method (WO3/SBA-15 and WO3/MCM-41)WO3/SBA-15 and WO3/MCM-41 catalysts with different tung-

sten loading in the range of 5 and 10 wt% of WO3 were preparedby wet impregnation method. Aqueous solution of ammoniummetatungstate [(NH4)6H2W12O40·xH2O] and the siliceous supportSBA-15 or MCM-41 were mixed together and stirred for 3 h. Then

the impregnated product was dried in an oven at 80 ◦C for 12 h andcalcined at 550 ◦C for 5 h. The catalysts obtained in this way wereidentified as WO3/SBA-15(x) and WO3/MCM-41(x) where x is themolar ratio of silicon to tungsten (Si/W).

226 T.I. Bhuiyan et al. / Applied Catalysis A: General 467 (2013) 224–234

Table 1Physico-chemical properties of synthesized catalysts.

Samples dspacing (nm) % of WO3 (ICP-AES) Surface area (m2/g) Pore diameter (nm) Total pore volume (cc/g)

MCM-41a 3.75 – 1056 4.4 1.169WO3/MCM-41(35)b 3.61 10 419 4.2 0.453WO3-MCM-41(30)a 4.32 7.43 680 4.9 0.785WO3-MCM-41(50)a 4.07 4.91 714 4.7 0.82SBA-15a 5.53 – 833 4.4 0.939WO3/SBA-15(35)b 5.69 10 468 4.3 0.636WO3-SBA-15(30)a 5.58 9.19 501 6.9 0.777WO3-SBA-15(60)a 5.97 4.92 571 6.2 0.988

Si/W molar ratio for each catalyst shown in the parenthesis; the d spacing for MCM-41 and SBA-15 catalyst was calculated using [1 0 0] reflection and [1 1 0] reflectionr

2

2

tfiiaa

2

gBaitcB

2

patsass1m

2

agcptc(assi

C

C

espectively.a Synthesized by hydrothermal method.b Synthesized by impregnation method.

.4. Catalyst characterization

.4.1. X-ray powder diffractionThe samples were characterized by the X-ray powder diffrac-

ion (XRD) method, using a Rigaku Mini-flex II system using nickelltered CuK� radiation (� = 1.5406 A) at a scanning rate of 1◦ min−1

n the 2� range of 1.3–50◦. The X-ray tube was operated at 40 kVnd 30 mA. The amounts of WO3 present in the final samples werenalyzed using ICP-AES (Ultima 2, Horiba Scientific).

.4.2. N2 adsorption–desorptionThe surface areas of the samples were measured by nitro-

en adsorption at −196 ◦C using AUTOSORB-1 (Quanta Chrome).efore adsorption measurements, samples (ca. 0.1 g) were heatedt 220 ◦C for 2 h under nitrogen flow. The nitrogen adsorptionsotherms of calcined materials were measured at liquid nitrogenemperature (−196 ◦C). The surface areas and pore diameters werealculated by the Brunauer–Emmett–Teller (BET) method and thearrett–Joyner–Halenda (BJH) method, respectively.

.4.3. Temperature programmed desorption of ammoniaTemperature-programmed desorption of ammonia (TPD) was

erformed with BEL-CAT-A-200, chemisorption apparatus. Thispparatus consists of a gas mixing unit, quartz micro reactor and ahermal conductivity detector (TCD). Samples were pretreated in atream of He (50 ml/min) at 500 ◦C for 1 h. Then, NH3 was adsorbedt 100 ◦C for 25 min. In order to remove loosely bounded ammonia,amples were purged with He stream for 1 h at 100 ◦C. Then, theamples were heated again from 100 to 700 ◦C at a heating rate of0 ◦C/min in a flow of He (25 ml/min) and the desorbed NH3 wasonitored by a TCD detector.

.4.4. Pyridine adsorption by FT-IRThe acidity of synthesized material was analyzed by pyridine

dsorption followed by FT-IR study. The catalyst sample was finelyround and pressed into a self-supporting wafer. The wafers werealcined under vacuum at 400 ◦C for 1 h, followed by exposure ofyridine vapor. The adsorbed pyridine was desorbed at variousemperatures (150–350 ◦C). The thin wafer was placed in the FT-IRell and the spectrum was recorded on a Nicolet FT-IR spectrometerMagna 500 model). The difference between the spectra of pyridinedsorbed on the samples and that of reference was obtained byubtraction. The amount of Brønsted acid sites (B) and Lewis acidites (L) were calculated using the following equation as reportedn literature [30].

2

(pyridine on B sites) = 1.88 IA(B)R /W.

(pyridine on L sites) = 1.42 IA(L)R2/W.

where C = concentration (mmol g−1 catalyst); IA(B, L) = integratedabsorbance of B or L band (cm−1); R = radius of catalyst disk (cm);W = weight of disk (mg).

2.4.5. Raman spectroscopyRaman spectra were recorded at room temperature on Horiba

Jobin Yvon in situ Raman, iHR 320 with synapse CCD detector (exci-tation line, 532.5 nm of an Ar+ ion laser; source power, 40–60 mW;resolution, 4 cm−1).

2.4.6. UV–vis DRSUV–vis DR spectra were recorded on a JASCO V-670 spectropho-

tometer at 200 nm/min scan rate with BaSO4 as reference.

2.5. Catalysts evaluation

The metathesis reaction of 2-butene (1:1 mixture of trans-2-butene & cis-2-butene) was performed in a fixed-bed tubularreactor (stainless steel tube grade 316 material, 0.312 inchID × 0.562 inch OD × 8 inch length). In a typical experiment, thereactor was charged with 2 ml of catalyst previously sieved to aparticle size of 0.5–1.0 mm diameter. The catalyst sample was firstactivated in a nitrogen stream at 550 ◦C for 1 h. The flow rates of feed(2-butene) and N2 were maintained at 5.0 ml/min and 25 ml/min,respectively, during the reaction. The metathesis reaction was car-ried out at different temperatures, ranging between 450 and 550 ◦C,GHSV (gas hourly space velocity) of 900 h−1, 5 h time-on-stream(TOS) and atmospheric pressure. The quantitative analysis of thereaction products was carried out on line using Varian GC with FID(Varian 450-GC), equipped with a CP-Al2O3/Na2SO4 column (50 mlength × 0.32 mm I.D. × df = 5 �m). The GC was programmed from50 to 90 ◦C at a heating rate of 8 ◦C/min (hold it for 10 min at 90 ◦C)and 90 to 190 ◦C at a heating rate of 10 ◦C/min (hold it for 5 minat 190 ◦C). Helium was used as carrier gas with a constant flow of30 ml/min.

3. Results and discussion

3.1. Catalyst characterization

3.1.1. X-ray powder diffractionThe results of low angle powder X-ray diffraction (XRD) pat-

terns for all samples based on MCM-41 are presented in Fig. 1.The peak at 2� ranging from 1.5◦ to 2.5◦ is indicative of MCM-41 structure. The intense peak at 1.8◦ is due to [1 0 0] reflectionand the weak broad signals between 2.5◦ and 4◦ (2�) are due

to [1 1 0], [2 0 0] and [2 1 0] reflections. The d-spacing values ofthese samples are presented in Table 1. These peaks were indexedby Beck et al. [31] for the hexagonal unit. These XRD patternscoincide with the data already reported in the literature [32].

T.I. Bhuiyan et al. / Applied Catalysis A: General 467 (2013) 224–234 227

F O3-Mw

IrWons[rootsrtMcii

WgwtSnvaa

Fa

ig. 1. XRD patterns of different samples (a) MCM-41, (b) WO3-MCM-41(50), (c) Wide angle.

ncorporation of tungsten into the MCM-41 framework resulted ineduction of [1 0 0] reflection intensity for WO3-MCM-41(30) and

O3-MCM-41(50) samples due to the change in electron densityf the tungsten incorporated materials [33]. In the case of impreg-ated samples, apart from the reduction in peak intensity, a smallhift in the [1 0 0] reflection was also observed. The intensity of the1 0 0] reflection decreased drastically and the [1 1 0] and [2 0 0]eflections are disappeared completely due to the impregnationf WO3 oxide. The d spacing values of 4.32 nm and 4.07 nm werebserved for WO3-MCM-41(30) and WO3-MCM-41(50), respec-ively. In the case of the WO3/MCM-41(35), d spacing value wasmaller than parent material. Very small peaks at 22–30◦ (2�) cor-esponding to the crystalline phase of WO3 were observed forhe WO3-MCM-41(30), whereas no peak was observed for WO3-

CM-41(50). In the case of impregnated samples, strong peaks ofrystalline WO3 were observed in the range 25◦ to 35◦ (2�). Thentensity of crystalline WO3 peaks was increased with an increasen W loading.

The X-ray diffraction patterns of SBA-15, WO3-SBA-15 andO3/SBA-15 samples are shown in Fig. 2 and d spacing values are

iven in Table 1. Two diffraction peaks in the region of 2� = 1.5–2◦,hich can be indexed to the [1 1 0] and [2 0 0] reflections charac-

eristic of the p6mm hexagonal symmetry, were observed for theBA-15 catalysts [34]. In all cases, [1 0 0] reflection (2� < 1.5◦) was

ot observed due to limitation of XRD instrument. The d spacingalues for all the catalysts were calculated using [1 1 0] reflections shown in Table 1. There is no significant difference in [1 1 0]nd [2 0 0] reflections for WO3-SBA-15 catalyst samples, whereas in

ig. 2. XRD patterns of different samples (a) SBA-15, (b) WO3-SBA-15(60), (c) WO3-SBAngle.

CM-41(30), (d) WO3/MCM-41(73), (e) WO3/MCM-41(35) at (A) low angle and (B)

case of WO3/SBA-15 catalysts the peak intensities were decreasedwith increase in amount of WO3 due to poor dispersion. There weresmall peaks assigned to the crystalline phase of WO3 as observedfor WO3-SBA-15 (30 and 60) samples. However, WO3 impregnatedon SBA-15 has strong crystalline WO3 peaks compared with hydro-thermal synthesized WO3-SBA-15 materials. The amount of WO3present in the catalysts was analyzed by ICP-AES and is presentedin Table 1.

3.1.2. N2 adsorption–desorptionTable 1 summarizes the textural properties of the supports

and catalysts calculated from nitrogen adsorption studies. Theisotherms of nitrogen adsorption for the calcined MCM-41 mate-rials are shown in Fig. 3A. The isotherm shows three well-definedstages and they coincide with those already reported in the liter-ature [27]. All MCM-41 samples show a typical type IV adsorptionisotherm with steps of capillary condensation in the primary meso-pores. It is evident from these isotherms that high mesostructuralordering and narrow pore size distribution were attained. Thesurface areas of the catalysts decreased in the order; WO3/MCM-41(35) < WO3-MCM-41(30) < WO3-MCM-41(50) < Si-MCM-41. Thepore volumes also decreased in the same order due to the presenceof textural mesoporosity. The surface area of WO3/MCM-41(35) islower than that of other catalysts probably due to the sintering of

tungsten metal itself and/or tungsten metal and silica support [15].

The isotherms of SBA-15 samples are presented in Fig. 3B.These materials exhibit type IV classification, characteristic ofmesoporous materials. According to the IUPAC classification, the

-15(30), (d) WO3/SBA-15(73), (e) WO3/SBA-15(35) at (A) low angle and (B) wide

228 T.I. Bhuiyan et al. / Applied Catalysis A: General 467 (2013) 224–234

F ) MCMS O3/SB

htAafboWip

3

Fsl(aafwahslaso

Fb

ig. 3. N2 adsorption isotherm of different samples, (A) MCM-41 based samples (aBA-15 based samples (e) SBA-15, (f) WO3-SBA-15(60), (g) WO3-SBA-15(30), (h) W

ysteresis loops of the WO3-SBA-15 can be classified as the H1ype [26]. H1 is associated with a more uniform pore distribution.sharp increase in adsorbed volume at P/P0 of approximately 0.70

rising from capillary condensation of nitrogen within the uni-orm mesoporous structures. The surface area and pore volumeoth shows a monotonic decrease with increasing concentrationf tungsten (Table 1). According to Table 1, WO3-SBA-15(30) andO3-SBA-15(60) materials have high surface area compared to

mpregnated WO3/SBA-15(35) due to the presence of well dis-ersed WO3 in the framework.

.1.3. Temperature programmed desorption of ammoniaNH3-TPD profiles for the supports and catalysts are shown in

ig. 4 and the total acidity (combination of weak and medium acidites) results are presented in Table 2. After Gaussian deconvo-ution, two peaks corresponding to weak (100–300 ◦C) and medium300–500 ◦C) sites were obtained. As shown in Fig. 4, a weekmmonia desorption peak was observed at 140 ◦C for MCM-41nd 180 ◦C for SBA-15. With increasing tungsten amount in theramework, the intensity of ammonia desorption peak at 180 ◦Cas increased along with some medium acid site peaks observed

bove 350–400 ◦C. The total acidity of WO3-SBA-15(30) was slightlyigher than that of WO3-MCM-41(30). In the case of impregnatedamples, the presence of WO3 on the external surface of the cata-

yst was observed and it creates Lewis acid centers of bulky WO3s reported in literature [4,14]. Impregnated samples have verymall amount of tungsten species deposited on the inner surfacef the pores and most of them are incorporated into the walls of

ig. 4. NH3-TPD profiles of different samples, (A) MCM-41 based samples (a) MCM-41, (based samples (e) SBA-15, (f) WO3-SBA-15(60), (g) WO3-SBA-15(30), (h) WO3/SBA-15(35

-41, (b) WO3-MCM-41(50), (c) WO3-MCM-41(30), (d) WO3/MCM-41(35), and (B)A-15(35).

the material and therefore not accessible for ammonia molecules.These results were in accordance with results of XRD. Higher acid-ity of WO3-SBA-15(30) and WO3-MCM-41(30) exists due to thepresence of more Si-O-W bonds, which are predominately formedduring hydrothermal process.

3.1.4. Raman spectroscopyRaman spectra of the supports and catalysts are shown in

Fig. 5. Three intense strong peaks were observed at 265, 707 and797 cm−1 for all the catalyst corresponding to crystalline WO3species. These peaks have been assigned to the symmetric stretch-ing mode of W O, bending mode of W O and the deformationmode of W O W, respectively [26]. It is obvious that pure silicadoes not have WO3 species, therefore no peaks were observed. Asmall peak at 970 cm−1 was assigned to the symmetric stretchingmode of the terminal W O bond of the tetrahedrally coordinatedtungsten oxide species [36–38], which are considered as activesites for metathesis reactions [39]. Another peak at about 900 cm−1

could be related to the incorporation of the transition metal ionsin the silica framework [37]. It is also a typical evidence for theisomorphous substitution of silicon by tungsten in the WO3-SBA-15 catalysts. High intensity crystalline peaks were observed forthe impregnated samples with 10% WO3 loadings compared toWO3-MCM-41 or WO3-SBA-15 framework catalysts. This clearly

indicates that more octahedral tungsten species are present inimpregnated samples due to poor dispersion of tungsten oxides. Onthe other hand, WO3-MCM-41 or WO3-SBA-15 materials show lessintense crystalline peaks with tetrahedrally coordinated tungsten

) WO3-MCM-41(50), (c) WO3-MCM-41(30), (d) WO3/MCM-41(35), and (B) SBA-15).

T.I. Bhuiyan et al. / Applied Catalysis A

F4S

s4W

3

otI

F4

ig. 5. Raman spectra of different samples (A) MCM-41 based samples (a) MCM-1, (b) WO3/MCM-41(35), (c) WO3-MCM-41(30) and (B) SBA-15 based samples (d)BA-15, (e) WO3/SBA-15(35) (f) WO3-SBA-15(30).

pecies. The less intensity of crystalline WO3 peaks in WO3-MCM-1(30) compared to WO3-SBA-15(60) indicates high dispersion ofO3 in the framework.

.1.5. Pyridine adsorption by FT-IR

Pyridine adsorption measured by IR spectroscopy was carried

ut to evaluate the strength and types of acid sites present in theungsten impregnated and framework samples. Fig. 6 shows the FT-R spectra of the all the catalysts recorded after the adsorption of

ig. 6. FT-IR spectra of pyridine adsorbed on different samples at 150 ◦C, (A) MCM-41 based1(35), (B) SBA-15 based samples (e) SBA-15, (f) WO3-SBA-15(60), (g) WO3-SBA-15(30) a

: General 467 (2013) 224–234 229

pyridine and subsequent evacuation at 150 ◦C. The Py-FT-IR spec-trum recorded after adsorption of pyridine at room temperature onthe pure MCM-41 or pure SBA-15 shows two major peaks at 1595and 1445 cm−1, ascribed to pyridine coordinately bonded to weaksurface Lewis acid sites. These peaks disappeared after outgassingat 350 ◦C. In case of WO3-MCM-41 and WO3-SBA-15 samples threeadditional peaks at 1545 and 1635 cm−1 were also observed dueto protonated pyridine bonded to surface Brønsted acid sites, indi-cating the presence of tungsten-incorporated species [35,36]. Thepeak at 1485 cm−1 has been identified as combination of both Lewisand Brøsted acid sites [30]. The intensities of all of these bandswere decreased after outgassing at 350 ◦C (see supporting informa-tion Fig. S1 for SBA-15). Similar spectra are recorded for tungstenimpregnated samples of SBA-15 and MCM-41, the number of theLewis and Brønsted acid sites decreased after outgassing at 350 ◦C.The IR bands near 1445 cm−1 and 1545 cm−1 were used for calcu-lation of Lewis and Brønsted acid sites respectively. The amount ofLewis and Brønsted acid sites was higher for WO3-MCM-41(30) ascompared to WO3-MCM-41(50) due to increase in the tetrahedrallyco-ordinated tungsten oxide species (Table 2). This observation isa good evidence for better catalytic activity of WO3-MCM-41(30)compared to WO3-MCM-41(50). It can be concluded that the pres-ence of Brønsted acid sites in the tungsten containing SBA-15 orMCM-41 samples are related to the presence of tetrahedrally co-ordinated tungsten oxide species.

3.1.6. UV–vis DRSDiffuse reflectance spectra of tungsten supported samples in

the region of 200–600 nm are shown in Fig. 7. This method isvery sensitive for distinguishing the species between incorporatedmetal and extra-framework metal oxides in different heteroatomicmesostructures. The spectrum of sodium tungstate presents aspinel structure at 230 nm which mainly implies the presenceof ligand to metal charge transfer involving isolated transitionmetal sites and is considered a direct proof for the presence offramework tungsten oxide species in tetrahedral co-ordination[27,40]. This band was not observed in case of pure SBA-15 andMCM-41. The WO3-SBA-15(30) and WO3-MCM-41(30) catalystsexhibited a similar band at 230 nm, which could be attributed tothe presence of isolated [WO4]2− tetrahedral species in the cat-alysts. On the other hand, the spectra of impregnated samples,

WO3/SBA-15(35) and WO3/MCM-41(35) have a less intense band at230 nm, indicating small concentration tetrahedral tungsten oxidespecies. The second band at 263 nm may be attributed to iso-lated tungsten species or low-condensed oligomeric tungsten oxide

samples (a) MCM-41, (b) WO3-MCM-41(50), (c) WO3-MCM-41(30), (d) WO3/MCM-nd (h) WO3/SBA-15(35). B = Bronsted acidity; L = Lewis acidity.

230 T.I. Bhuiyan et al. / Applied Catalysis A: General 467 (2013) 224–234

Table 2Acid sites characteristics of tungsten incorporated and impregnated SBA-15 and MCM-41 catalysts.

Samples NH3 TPD (mmol g−1)a FT-IR chemisorbed pyridine (mmol g−1)

TA L.T. H.T. B L B L<300 ◦C 300–550 ◦C 150 ◦C 150 ◦C 250 ◦C 250 ◦C

MCM-41 0.039 0.039 0.00 0.0233 0.3611 0.0033 0.1155WO3/MCM-41(35) 0.160 0.080 0.080 0.0318 0.1889 0.0205 0.0709WO3-MCM-41(30) 0.391 0.210 0.181 0.1527 0.2581 0.0484 0.1685WO3-MCM-41(50) 0.325 0.140 0.185 0.0608 0.2410 0.0241 0.1036SBA-15 0.035 0.035 0.00 0.0265 0.3818 0.0056 0.1397WO3/SBA-15(35) 0.108 0.096 0.013 0.0347 0.1715 0.0152 0.0667WO3-SBA-15(30) 0.458 0.270 0.188 0.1086 0.3521 0.0539 0.1555WO3-SBA-15(60) 0.357 0.180 0.177 0.0433 0.1979 0.0208 0.0952

TA = total acidity; L.T. = low temperature; H.T. = high temperature; B = Bronsted acidity (peaks at 1545 cm−1); L = Lewis acidity (peaks at 1445 cm−1).

snT4oct

a Calculated from the areas of Gaussian deconvolution bands.

pecies. A broad band at 350–400 nm was observed for ammo-ium metatungstate having mainly octahedral tungsten species.he peak at 375 nm was observed for WO3/SBA-15(35), WO3/MCM-1(35) and WO3-SBA-15(30) catalysts indicated the presence of

ctahedral tungsten oxide species. This band was not observed inase of WO3-MCM-41(30) catalyst indicating the presence of moreetrahedral tungsten oxide species present in the catalyst.

Scheme 1. Possible reaction pathways for 2-butene m

3.2. Reaction studies

A systematic study of metathesis reactions of 2-butene overtungsten oxide supported mesoporous molecular sieves (SBA-15

and MCM-41) was performed. 2-Butene can undergo a seriesof different reactions e.g. isomerization, metathesis, oligomeriza-tion, cracking etc. All possible reactions of 2-butene are presented

etathesis over W-based heterogeneous catalyst.

T.I. Bhuiyan et al. / Applied Catalysis A: General 467 (2013) 224–234 231

Fig. 7. UV–vis diffuse reflectance spectra of various samples (a) Na2WO4·2H2O,(4

i1MdIsiua

F12(G

b) WO3-SBA-15(30), (c) WO3-MCM-41(30), (d) WO3/SBA-15(35), (e) WO3/MCM-1(35), (f) ammonium metatungstate.

n Scheme 1 [41]. Metathesis reaction between 2-butene and-butene to produce propylene and 2-pentene is termed asetathesis I. Apart from Metathesis I, propylene can also be pro-

uced by metathesis between 2-butene and ethylene (MetathesisI), as well as 2-butene with isobutylene (Metathesis III). Metathe-is IV is a reaction between isomerized products 1-butene andsobutylene to produce ethylene with hexene. 1-Butene can also

ndergo self metathesis to produce ethylene and hexene, presenteds metathesis V.

ig. 8. Results of metathesis reaction of 2-butene over (a) SBA-15, (b) WO3/SBA-5(73), (c) WO3/SBA-15(35), (d) WO3-SBA-15(60), (e) WO3-SBA-15(30) catalysts ath (A) conversion of 2-butene, propylene yield (mol%), ethylene yield (mol%) and

B) ratio of propylene to pentene (reaction temperature: 550 ◦C, pressure: 1 atm,HSV: 900 h−1).

Fig. 9. Results of metathesis reaction of 2-butene over (a) MCM-41, (b) WO3/MCM-41(73), (c) WO3/MCM-41(35), (d) WO3-MCM-41(50), (e) WO3-MCM-41(30)

catalysts at 2 h (A) conversion of 2-butene, propylene yield (mol%), ethylene yield(mol%) and (B) ratio of propylene to pentene (reaction temperature: 550 ◦C, pres-sure: 1 atm, GHSV: 900 h−1).

3.2.1. Catalysts based on SBA-15The tungsten oxide supported SBA-15 catalysts were prepared

following two different methods. In the first method, two cata-lysts having tungsten oxide with Si/W ratio of 35 and 73 wereprepared following wet incipient impregnation method while inthe second method two catalysts having Si/W ratio of 30 and 60were synthesized following hydrothermal crystallization proce-dure. The results of the metathesis reaction of 2-butene at 550 ◦Cover tungsten oxide supported SBA-15 catalysts are presented inFig. 8 and Table S1. A significant increase in conversion of 2-butenewas observed after incorporation of tungsten oxide. The catalyticconversion of 2-butene increased with an increase in the amountof tungsten oxide in both types of catalysts. The WO3-SBA-15 withSi/W ratio of 30 resulted in high 2-butene conversion of 86.7%.Similarly propylene and ethylene yields were observed to increasewith an increase in the amount of tungsten oxide. Pure SBA-15exhibited 1 mol% of propylene yield and 0 mol% of ethylene yield.The WO3-SBA-15(30) depicted the highest yield of 37.1 mol%, and16.7 mol% of propylene and ethylene, respectively. It was noticedthat C3=/C5= ratio also increased with an increase in the amountof tungsten oxide. The WO3/SBA-15(73) resulted in a low ratioof C3=/C5= (1.3) indicating metathesis I reaction was taking placepredominately, whereas in the case of WO3-SBA-15(30), C3=/C5=

ratio was 2.3 indicating the possibility of other reactions apart frommetathesis I.

It was noticed that catalysts based on hydrothermal synthesisexhibited higher conversion and propylene yield as compared tothat of impregnated ones, indicating the vital role of frameworktungsten oxide sites in the metathesis reaction. The total acidity

232 T.I. Bhuiyan et al. / Applied Catalysis A: General 467 (2013) 224–234

Fig. 10. Results of metathesis reaction of 2-butene at different temperatures over (a) WO3/MCM-41(35), (b) WO3/SBA-15(35), (c) WO3-MCM-41(30), and (d) WO3-SBA-1 hylenp

(H2baw1oWc0S

3

piimpoi

5(30) catalysts at 2 h (A) conversion of 2-butene, (B) propylene yield (mol%), (C) etressure: 1 atm, GHSV: 900 h−1).

Table 2) of WO3-SBA-15(30) was higher than WO3/SBA-15(35).igher acidity of WO3-SBA-15(30) enhanced the isomerization of-butene to 1-butene and thus provided more availability of 1-utene for metathesis reaction with 2-butene. The higher aciditynd availability of more tetrahedrally coordinated tungsten speciesere responsible for superior metathesis activity of WO3-SBA-

5(30) as compared to that of WO3/SBA-15(35). The reproducibilityf results was test by performing three different runs at 550 ◦C usingO3-SBA-15(30) as catalyst. The standard deviation of 2-butene

onversion, propylene yield and ethylene yield was found to be.96, 0.69 and 1.17 respectively (see supporting information Fig.2).

.2.2. Catalysts based on MCM-41The MCM-41 supported tungsten oxide catalysts were also pre-

ared following two different methods. In first case, tungsten wasmpregnated using wet incipient impregnation method, whereasn second case, tungsten was incorporated into the framework of

esoporous MCM-41 support using hydrothermal crystallizationrocedure. The catalytic activity of MCM-41 supported tungstenxide catalysts for 2-butene metathesis at 550 ◦C are presentedn Fig. 9 and Table S2. A significant increase in conversion of

e yield (mol%) and (D) ratio of propylene to pentene (reaction temperature: 550 ◦C,

2-butene was observed after incorporation of tungsten oxide intoMCM-41. The catalytic conversion of butene was increased withan increase in the amount of tungsten oxide in both types of cata-lysts. The WO3-MCM-41(30) exhibited the highest conversion of86.7%. Pure MCM-41 exhibited 5 and 2 mol% yield of propyleneand ethylene, respectively, whereas WO3-MCM-41(30) depictedhighest yield of 36.3 and 16.5 mol% of propylene and ethylene,respectively. It was noticed that C3=/C5= ratio was greater than1.5 over all four catalysts. The higher activity of WO3-MCM-41(30) compared to other catalysts could be related to its higheracidity as well as larger amount of tetrahedrally co-ordinated tung-sten species as observed from pyridine-FT-IR, XRD and Ramanspectroscopy. The lower activity of impregnated MCM-41 sam-ples was probably due to less availability of tetrahedral tungstenspecies and higher intensity of extra framework octahedral WO3species.

3.2.3. Effect of reaction temperature

Metathesis reaction of 2-butenes over WO3/MCM-41(35),

WO3/SBA-15(35), WO3-MCM-41(30) and WO3-SBA-15(30) cata-lysts at three different temperatures from 450, 500 and 550 ◦Cwas carried out and the results are shown in Fig. 10. It was

lysis A

ottaa4Wb(hMp

shlieoitclrAwrn541o

4

•

•

•

•

•

•

•

A

bt

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[[

[[

[

T.I. Bhuiyan et al. / Applied Cata

bserved that conversion of 2-butene increased with increase inemperature for all four catalysts. It was also noticed that at lowemperature, catalysts based on MCM-41 shows higher activitys compared to that of SBA-15 based catalysts. The activity ofll four catalysts at 450 ◦C was: WO3-MCM-41(30) > WO3/MCM-1(35) ∼ WO3-SBA-15(30) > WO3/SBA-15(35). Higher activity ofO3-SBA-15(30) could be correlated to the presence more num-

er of tetrahedrally co-ordinated WO3 species than other catalystsevidenced from pyridine-FT-IR and UV–vis DRS). However, atigher temperature i.e. 550 ◦C the catalytic activity of WO3-CM-41(30) and WO3-SBA-15(30) were comparable with highest

ropylene yield of 39 mol% and 37 mol%, respectively.A summary of various catalysts used so far in butene metathe-

is is presented in Table S3. Harmse et al. [16] have reportedighest yield (33 mol%) of propylene over WO3/Siralox-40 cata-

yst for 1-butene metathesis reaction. However, in our presentnvestigation propylene yield of 39 mol% was achieved. The high-st yield of propylene could be linked to the moderate acidityf WO3-MCM-41 and WO3-SBA-15 catalysts. This moderate acid-ty was sufficient for isomerization of 2-butene to 1-butene andhen metathesis reaction. Higher yield of ethylene at 550 ◦Could be related to metathesis reactions of 2-butene and isobuty-ene as well as 1-butene auto-metathesis (Scheme 1) as theseeactions occur at higher temperature with moderate acidity.t high temperature, C1 and C2 alkanes were not observedhich are considered to be the main products of cracking

eactions. This clearly indicated that metathesis I was predomi-ately taking place at 550 ◦C. The activity of all four catalysts at50 ◦C was: WO3-SBA-15(30) ∼ WO3-MCM-41(30) > WO3/MCM-1(35) > WO3/SBA-15(35). The WO3/MCM-41(35) and WO3/SBA-5(35) catalysts resulted into C3=/C5= ratio ∼1 at 450 ◦C indicatingnly metathesis I reaction taking place at lower temperature.

. Conclusions

The conclusions of this study can be summarized as below:

Catalysts having tungsten oxide in the framework of mesoporoussilica exhibited the higher metathesis activity as compared totungsten oxide impregnated catalysts.The lower activity of impregnated catalysts is due to abundance ofcrystalline WO3, which are shown to be non-active for metathesisreactions.WO3-MCM-41(30) and WO3-SBA-15(30) catalyst showed com-parable activity with highest propylene yield of 39 mol% and37 mol%, respectively at 550 ◦C.The activity order of all catalysts at 450 ◦C was: WO3-MCM-41(30) > WO3/MCM-41(35) ∼ WO3-SBA-15(30) > WO3/SBA-15(35), whereas at 550 ◦C the order was: WO3-SBA-15(30) ∼WO3-MCM-41(30) > WO3/MCM-41(35) > WO3/SBA-15(35).WO3-MCM-41 catalysts exhibited much higher activity at lowtemperature (450 ◦C) as compared to that of WO3-SBA-15.The higher activity of WO3-MCM-41 could be related to the pres-ence of more number of tetrahedrally co-ordinated tungstenoxide species as evident from UV–vis DRS and pyridine-FT-IRspectroscopy.WO3-MCM-41 is a better choice from industrial point of view dueto lower cost of structure directing agent as compared to that ofWO3-SBA-15.

cknowledgments

The authors would like to acknowledge the support providedy Saudi Aramco for funding project# CRP02252. The support fromhe Ministry of Higher Education, Saudi Arabia, in establishment

[[[

: General 467 (2013) 224–234 233

of the Center of Research Excellence in Petroleum Refining &Petrochemicals at King Fahd University of Petroleum & Miner-als (KFUPM) is appreciated. The support of KFUPM is also highlyappreciated.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.apcata.2013.07.034.

References

[1] H.J. Liu, L. Zhang, X.J. Li, S.J. Huang, S.L. Liu, W.J. Xin, S.J. Xie, L.Y. Xu, J. Nat. GasChem. 18 (2009) 331–336.

[2] Q.F. Zhao, S.L. Chen, J.S. Gao, C.M. Xu, Transit. Met. Chem. 34 (2009)621–627.

[3] J.C. Mol, P.W.N.M. van Leeuwen, Handbook of Heterogeneous Catalysis, 2nd ed.,Wiley, Weinheim, 2008, pp. 3240–3256.

[4] D.R. Hua, S.L. Chen, G.M. Yuan, Y.L. Wang, Q.F. Zhao, X.L. Wang, B. Fu, Micro-porous Mesoporous Mater. 143 (2011) 320–325.

[5] P. Topka, H. Balcar, J. Rathousky, N. Zillkova, F. Verpoort, J. Cejka, MicroporousMesoporous Mater. 96 (2006) 44–54.

[6] Z. Cheng, C.S. Lo, ACS Catal. 2 (2012) 341–349.[7] S.G. Huang, S.L. Liu, W.J. Xin, J. Bai, S.J. Xie, Q.X. Wang, L.G. Xu, J. Mol. Catal. A:

Chem. 226 (2005) 61–68.[8] L. Sang, S.L. Chen, G.M. Yuan, M. Zheng, J. You, A.E. Chen, R. Li, L.J. Chen, J. Nat.

Gas Chem. 21 (2012) 105–108.[9] A. Andreni, X. Xiaoding, J.C. Mol, Appl. Catal. 27 (1986) 31–40.10] A. Behr, U. Schuller, K. Bauer, D. Maschmeyer, K.D. Wiese, F. Nierlich, Appl. Catal.

A: Gen. 357 (2009) 34–41.11] J. Handzlik, J. Ogonowski, J. Stoch, M. Mikolajczyk, P. Michorczyk, Appl. Catal.

A: Gen. 312 (2006) 213–219.12] H. Balcar, D. Mishra, E. Marceau, X. Carrier, N. Zilkova, Z. Bastl, Appl. Catal. A:

Gen. 359 (2009) 129–135.13] D.P. Debecker, M. Stoyanova, U. Rodemerck, A. Leonard, B.L. Su, E.M. Gaigneaux,

Catal. Today 169 (2011) 60–68.14] Y.D. Wang, Q.L. Chen, W. Yang, Z. Xie, W. Xu, D. Huang, Appl. Catal. A: Gen. 250

(2003) 25–37.15] S. Chaemchuen, S. Phatanasri, F. Verpoort, N. Sae-ma, K. Suriye, Kinet. Catal. 53

(2012) 247–252.16] L. Harmse, C. van Schalkwyk, E. van Steen, Catal. Lett. 137 (2010)

123–131.17] S.J. Huang, S.L. Liu, W.J. Xin, S.J. Xie, Q.X. Wang, L.Y. Xu, J. Nat. Gas Chem. 15

(2006) 93–99.18] S.J. Huang, S.L. Liu, W.J. Xin, J. Bai, S.J. Xie, Q.X. Wang, L.Y. Xu, J. Mol. Catal. A:

Chem. 226 (2005) 61–68.19] A. Spamer, T.I. Dube, D.J. Moodley, C. van Schalkwyk, J.M. Botha, Appl. Catal. A:

Gen. 255 (2003) 133–142.20] D.R. Hua, S.L. Chen, G.M. Yuan, Y.L. Wang, L. Zhang, Transit. Met. Chem. 36

(2011) 245–248.21] S.J. Huang, F.C. Chen, S.L. Liu, Q.J. Zhu, X.X. Zhu, W.J. Xin, Z.C. Feng, C. Li, Q.X.

Wang, L.Y. Xu, J. Mol. Catal. A: Chem. 267 (2007) 224–233.22] Y. Han, F.S. Xiao, S. Wu, Y.Y. Sun, X.J. Meng, D.S. Li, S. Lin, F. Deng, X.J. Ai, J. Phys.

Chem. B 105 (2003) 7963–7966.23] D.Y. Zhao, Q.E. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120

(1998) 6024–6036.24] N. Rahmat, A.Z. Abdullah, A.R. Mohamed, Am. J. Appl. Sci. 7 (2010)

1579–1586.25] L.F. Chen, J.C. Hu, Y.D. Wang, K. Zhu, R. Richards, W.M. Yang, Z.C. Liu, W. Xu,

Mater. Lett. 60 (2006) 3059–3062.26] J.C. Hu, Y.D. Wang, L.F. Chen, R. Richards, W.M. Yang, Z.C. Liu, W. Xu, Micro-

porous Mesoporous Mater. 93 (2006) 158–163.27] C.Y. Yang, Z. Lin, X.J. Shi, J.F. Kang, X.G. Wang, X.C. Zheng, J. Dispers. Sci. Technol.

32 (2001) 1497–1501.28] H. Balcar, N. Zilkova, J. Sedlacek, J. Zednik, J. Mol. Catal. A: Chem. 232 (2005)

53–58.29] W.L. Dai, H. Chen, Y. Cao, H.I. Li, S.H. Xie, K.N. Fan, Chem. Commun. (2003)

892–893.30] C.A. Emies, J. Catal. 141 (1993) 347–354.31] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992)

710–712.32] Z.R. Zhang, J.H. Suo, X.M. Zhang, S.B. Li, Appl. Catal. A: Gen. 177 (1999) 11–19.33] O. Klepel, W. Bohlmann, E.B. Ivanov, V. Riede, H. Papp, Microporous Mesoporous

Mater. 76 (2004) 105–112.34] J.E. Herrera, J.H. Kwak, J.Z. Hu, Y. Wang, C.H.F. Peden, J. Macht, E. Iglesia, J. Catal.

239 (2006) 200–211.35] C. Martin, P. Malet, G. Solana, V. Rives, J. Phys. Chem. B 102 (1998) 2759–2768.36] X.L. Yang, W.L. Dai, R.H. Gao, K.N. Fan, J. Catal. 249 (2007) 278–288.37] F. Figueras, J. Palomeque, S. Loridant, C. Feche, N. Essayem, G. Gelbard, J. Catal.

226 (2004) 25–31.

2 lysis A

[

[

[

[

34 T.I. Bhuiyan et al. / Applied Cata

38] C.Y. Cheng, K.J. Lin, M.R. Prasad, S.J. Fu, S.Y. Chang, S.G. Shyu, H.S. Sheu, C.H.

Chen, H. Chuang, M.T. Lin, Catal. Commun. 8 (2007) 1060–1064.

39] A. de Lucas, J.L. Valverde, P. Canizares, L. Rodriguez, Appl. Catal. A: Gen. 184(1999) 143–152.

40] E. Briot, J.Y. Piquemal, M. Vennat, J.M. Bregeault, G. Chottard, J.M. Manoli, J.Mater. Chem. 10 (2000) 953–958.

[

[

: General 467 (2013) 224–234

41] D.H. Zhang, X.J. Li, S.L. Huang, X.X. Zhu, F.C. Chen, S.J. Xie, L.Y. Xu, Appl. Catal.

A: Gen. 439–440 (2012) 171–178.

42] E. Mazoyer, K.C. Szeto, S. Norsic, A. Garron, J.M. Basset, C.P. Nicholas, M. Taoufik,ACS Catal. 1 (2011) 1643–1646.

43] N. Liu, S. Ding, Y.I. Cui, N.H. Xue, L.I. Peng, X.F. Guo, W.P. Ding, Chem. Eng. Res.Des. 91 (2013) 573–580.