Chemical Engineering Journal 223 (2013) 785–794

Contents lists available at SciVerse ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier .com/locate /cej

Synthesis and characterization of the acidic properties and pore textureof Al-SBA-15 supports for the canola oil transesterification

1385-8947/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.cej.2013.03.065

⇑ Corresponding author at: School of Occupational Safety and Health, Chung ShanMedical University, Taichung 402, Taiwan, ROC. Tel.: +886 4 24730022; fax: +886 423248194.

E-mail address: hhtseng@csmu.edu.tw (H.-H. Tseng).

Chenju Liang a, Ming-Chi Wei b, Hui-Hsin Tseng c,d,⇑, En-Chin Shu c,d

a Department of Environmental Engineering, National Chung Hsing University, 250 Kuo-Kuang Rd., Taichung 402, Taiwan, ROCb Department of Food Science, Central Taiwan University of Sciences and Technology, Taichung 402, Taiwan, ROCc School of Occupational Safety and Health, Chung Shan Medical University, Taichung 402, Taiwan, ROCd Department of Occupational Medicine, Chung Shan Medical University Hospital, Taichung 402, Taiwan, ROC

h i g h l i g h t s

�Mesopore and surface acidity of Na/SBA-15 improve the transesterification activity.� Pore structure of Na/SBA-15 were altered using different aging temperature.� Surface acidity of Na/SBA-15 were modified by incorporation of heteroatom Al.� The FAME yield of Na/Al-SBA-15 was comparability with commercial support.

a r t i c l e i n f o

Article history:Received 5 October 2012Received in revised form 11 March 2013Accepted 13 March 2013Available online 25 March 2013

Keywords:BiodieselTransesterificationCanola oilSolid catalystSBA-15

a b s t r a c t

Ordered mesoporous SBA-15 was used as a support for the transesterification of canola oil to producebiodiesel. The pore size and surface acidity of SBA-15 were modified by synthesis conditions using differ-ent aging temperatures and the addition of heterogeneous Al atoms. The results of the characterizationstudy indicated that a large amount of mesopore and surface acidity can greatly improve the transeste-rification reaction at a high aging temperature and with the incorporation of aluminum into the SBA-15framework. This improvement in transesterification activity is due to the formation of more stable„SiAOAAl„ bonds on the surface of the framework. The transesterification reaction was also comparedwith synthesis involving traditional supports such as acidic and basic c-Al2O3, hydrotalcite, and SiO2. Theactivities of the catalysts were related to their acidic strength.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

The biodiesel fatty acid methyl ester (FAME) is a renewable andclean energy source that can replace oil and has the potential tohelp reduce oil dependence and global warming [1]. Biodiesel issynthesized by the transesterification of vegetable oils or animalfats with a short-chain alcohol in the presence of catalysts. In thesynthesis, the exchange of the organic group R00 of an ester withthe organic group R0 of an alcohol occurs [2,3]. Transesterificationis usually catalyzed by acid or base catalysts that donate a protonto the carbonyl group or remove a proton from the alcohol, therebymaking them both more reactive [4,5]. Most biodiesels are cur-rently produced in the presence of homogeneous basic catalysts,which offer advantages such as higher reaction rates and a low

temperature requirement to obtain high biodiesel yields in a shortperiod of time [6]. However, the main drawback of using homoge-neous basic catalysts is the occurrence of saponification as a sidereaction (i.e., RACOOH + NaOH ? RACOONa + H2O), thereby mak-ing the process inefficient [7].

Although the reaction rate of syntheses that involve the use ofheterogeneous catalysts is slower compared to the reaction rateof syntheses that involve homogeneous catalysts, the heteroge-neous catalysts still offer several advantages, which include beingnoncorrosive, able to regenerate, and causing fewer disposal prob-lems [8]. The process is thus continuous. Heterogeneous catalystscan be designed to exhibit higher activity, selectivity, and longerlifetimes by modifying the characterization of the catalysts.

Several types of heterogeneous catalysts have been investigatedfor transesterification. These catalysts include alkaline earth metaloxides [4,8,9] or metal oxides supported on alumina [2,5,10], zeo-lite [1], or hydrotalcite [11,12]. Zeolites were initially considered asa promising material because of their unique porous structure,high surface area, and high thermal stability. However, reactions

786 C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794

involving zeolites were unsuccessful because larger reactive mole-cules are involved [7]. The acidity of the support is also essential intransesterification because solid acid catalysts can simultaneouslycatalyze the transesterification of triglycerides and the esterifica-tion of free fatty acids when waste cooking oil is used to producesecond-generation biofuels [13]. Therefore, aside from the physicaltexture, the surface acidity of the support is also essential inimproving the biodiesel yield rate. Based on previous studies, thefollowing strategies have been recommended to improve the cata-lyst performance: (i) developing the pore structure with a largepore diameter and surface area to improve the mass transfer andmetallic dispersion; and (ii) increasing the support surface acid/ba-sic strength for proton donation/removal [14]. These strategiesmay be accomplished by using supports with appropriate mor-phologies and acidic surface properties.

Amorphous silicas with ordered mesoporous structures (likeMCM-41 and SBA-15) have been widely used as supports for vari-ous metals [15–18] because of their controllable molecular-sizedpores, strong surface acidity and large number of active sites[17]. These supports are particularly attractive applied in heteroge-neous reactions involving large organic molecules where micropo-rous zeolites cannot be used [19]. Among different supports withan ordered mesopore structure, SBA-15 seem to be more suitedto react in transesterification because SBA-15 has larger pore sizes(4.6–30 nm), thicker pore walls (3.1–6.4 nm) and higher surfacearea (up to 1000 m2/g). SBA-15 also shows a greater opportunityto improve the hydrothermal stability [17] than the conventionalMCM-41 support. Samart et al. [1] applied mesoporous silica as asupport in transesterification, and they evaluated the reaction con-ditions for the reaction of soybean oil with methanol. They showedthat the methyl ester yield was close to the yield produced by theconventional homogeneous catalysis method. However, the char-acteristics of the silica, which are important in determining theactivity of the catalysts, were not evaluated. Saravanamuruganet al. [20] controlled the SBA-15 morphology with an amino groupand applied this support to the transesterification of diethyl mal-onate with various aliphatic and aromatic alcohols. They foundthat functionalized catalysts exhibited enhanced catalytic activitystrongly dependent on the morphologies of the support.

However, the electrically neutral framework of silica providesSBA-15 materials with a weak acidity due to the silanol groups lo-cated on the pore walls. Several efforts have been made to synthesizeSBA-15 with various heteroatoms such as aluminum [15,17,18,21],iron [15], and zirconium [16,22] to enhance its acidity by creatingLewis acid sites. Al-SBA-15 materials with a large number of moder-ately acidic sites are one kind of very promising catalysts and sup-ports involving catalytic reaction. Li et al. [17] have showed thatAl-SBA-15 materials exhibit high hydrothermal stability due to theformation of more stable „SiAOAAl„ bonds on the surface of theframework. Szczodrowski et al. [23] also indicated that the dopingprocedure enhanced SBA-15 acidity, especially as observed withAl-SBA-15 compared to Zr-SBA-15 or Ti-SBA-15.

This study aims to improve the activity of NaOH/SBA-15 catalystsby altering the physicochemical properties of the support. The prop-erties of the support can be altered by changing its pore structureand surface acidity. The introduction of Al into the silica frameworkto improve the acidity and to modify the pore texture by changingthe aging temperature was systematically investigated.

2. Experimental

2.1. Materials

Commercial chemicals with the highest purity available (>98%),namely, NaOH (RDH, Germany), aluminum oxide (Sigma–Aldrich,

France), hydrotalcites (Sigma–Aldrich, France), silicon dioxide (Sig-ma–Aldrich, France), and methanol (Fluka, France) were used with-out further purification. Canola oil (purity >99.9%) was obtainedfrom Tatung Chang Chi Food Stuff Factory Co., Ltd. (Taiwan).

Tetraethyl orthosilicate (TEOS), non-ionic triblock copolymerPEO–PPO–PEO ((poly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide)) (pluronic P123, MW: 5800 g/mol), hydro-chloric acid (HCl, 37%), and aluminum isopropoxide (AIP) werepurchased from Sigma–Aldrich (France). All chemicals were usedfor the preparation of pure or Al-doped SBA-15 mesoporousmaterials.

2.2. SBA-15 and Al-SBA-15 synthesis

The SBA-15 support was synthesized using pluronics P123 as astructure-directing agent and TEOS as a silica source. In this typicalprocess, 4 g of P123 was dissolved in 150 mL of aqueous HCl at pH1.5. Subsequently, 5 mL of TEOS was added dropwise to the solu-tion. After stirring at 40 �C for 24 h, the mixture was aged at90 �C, 100 �C, 110 �C, or 120 �C for 24 h. The white precipitatewas then collected by filtration and dried at 110 �C overnight.The product was obtained by removing the template at 550 �Cfor 6 h at a heating rate of 1 �C/min. The samples obtained weredesignated as S90, S100, S110, and S120.

Al-SBA-15 was synthesized through a ‘‘direct synthesis’’ meth-od [24] by using the same procedure and by using AIP as an Alsource. In a typical synthesis, 5 mL of TEOS and a calculatedamount of AIP were added to 10 mol of aqueous HCl at pH 1.5 toobtain Si/Al molar ratios of 10, 30, and 50. The solution was stirredfor 3 h and then added to a second solution containing 4 g of P123in 140 mL of 2 M HCl solution at 40 �C. The mixture was stirred for24 h at the same temperature and then aged at 110 �C for 24 h inclosed Teflon bottles. The solid product obtained was filtered, driedat 110 �C and finally calcined in air at 550 �C for 6 h at a heatingrate of 1 �C/min. The sample was designated as Al-S110(X), whereX is the theoretical Si/Al molar ratio.

Catalyst preparationThe impregnation of the supports with the Na precursor salts

was conducted using a pore volume impregnation technique. Anappropriate concentration of NaOH ethanol solution (20 g sup-port/150 mL ethanol) was used to impregnate the support with5 wt% loading weight of Na. After 24 h, the solution was heatedand constantly stirred until the liquid was entirely evaporated.NaOH loaded on basic c-Al2O3, acid c-Al2O3, silica dioxide, andhydrotalcite was also prepared for comparison with other catalystsusing the impregnation method from an aqueous solution. The pre-pared catalysts were dried in an oven at 105 �C for 2 h and calcinedat 400 �C for 5 h in air.

2.3. Catalyst characterization

The crystal structure of the catalyst was determined by a smallangle X-ray diffraction system (SAXRD; Mac Science Co., M18XHF),with a Cu Ka X-ray source. The diffraction patterns were acquiredwith a 2h angle, which ranged from 0.5� to 8� and a scanning stepof 0.02�.

The morphology of the supporting materials and the supportedcatalysts was examined using a field emission scanning electronmicroscope (FESEM, JEOL JSM-6700F).

Fourier transform infrared (FT-IR) spectra were obtained inthe 4000–400 cm�1 region with a PerkinElmer 1730 FT-IRspectrometer.

High-resolution transmission electron micrograph (HRTEM)images were obtained using a JEOL 2010 microscope. The solidswere ultrasonically dispersed in heptane, and the suspension wascollected on carbon-coated grids.

C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794 787

The molar ratio of silicon to alumina and the actual amount ofsodium loading weight in the materials obtained after the last cal-cination step were evaluated using an inductively coupled plasmamass spectrometer (PerkinElmer, SCIEX ELAN 5000). The sampleswere pretreated by microwave digestion before the analysis.

The specific surface areas and pore size distributions of the cat-alysts were determined by the multipoint Brunauer–Emmet–Teller(BET) method using the PMI Automated BET Sorptometer(201AEL). The samples were outgassed overnight down to 10�3 -Torr at 378 K to eliminate the impurities adsorbed on the samplesurface prior to the measurements.

Pyridine adsorption was performed on the samples by using anFT-IR spectrometer at 25 �C to evaluate and analyze the strengthand type of the acid sites. The sample was heated to 400 �C in avacuum and cooled to room temperature before pyridine adsorp-tion. After the pyridine was adsorbed on the sample at room tem-perature, evacuated, and heated at various temperatures, the IRspectra of the samples were obtained.

Fig. 1. SAXRD patterns of SBA-15 supports synthesized under (a) 9

Table 1Textural and structural characteristics of SBA-15 and Al-SBA-15(X) supports.

Sample Si/Al molar ratioa d100 a0b (nm)

S90 – 9.10 10.51S100 – 9.30 10.74S110 – 9.40 10.85S120 – 9.80 11.32Al-S110(50) 68.2 (70.3)g 9.70 11.20Al-S110(30) 39.1 (39.5) 9.60 11.10Al-S110(10) 17.6 (18.1) 9.60 11.10

a Determined by ICP analysis.b a0 = 2d100/

p3 is the hexagonal lattice parameter derived from the XRD data.

c Dpore is the mean pore diameter derived from N2 desorption data based on the BJHd W = a0 � D is the mean pore wall thickness.e S is the specific surface area.f V is the total pore volume.g The values in the parentheses were the Si/Al molar ratio of spent catalysts.

2.4. Transesterification reaction

Transesterification experiments were carried out in a thermo-stated glass tank reactor equipped with a stirrer, a temperatureprobe, a tube for sample withdrawal, and a condenser with coolingwater. Methanol and canola oil at a 6:1 M ratio (total reactionweight of 100 g) were added to the reactor with 10 g of the cata-lyst. The temperature was then increased to 65 ± 3 �C with con-stant stirring at 600 rpm. The reaction time for transesterificationwas 6 h. The samples were removed from the reaction mixture,and the biodiesel portions were separated by centrifugation. Threephases were formed. The top layer was biodiesel, the bottom layerwas the catalyst, and a small amount of glycerol was observed.

The biodiesel concentrations were analyzed using an Agilent6890 N gas chromatograph equipped with a flame ionizationdetector and an AB-WAX 2025–3002 column (30 m � 0.25 mm �0.25 lm). The separation was achieved using the isothermal anal-ysis method. The oven was heated in helium gas at a rate of 10 �C/

0 �C, (b) 100 �C, (c) 110 �C, and (d) 120 �C aging temperatures.

Dporec (Å) Wd (nm) SBET

e (m2 g�1) Vtotalf (m3 g�1)

37.9 6.72 738.74 0.7045.1 6.23 898.21 1.0149.1 5.94 684.18 0.8449.8 6.34 585.15 0.7360.4 5.16 790.40 1.1959.1 5.19 711.36 1.0558.8 5.22 691.55 1.02

method.

788 C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794

min from 150 �C to 230 �C and was maintained at 230 �C for20 min. A temperature of 250 �C was used for the injection portand the detector.

The yield for conversion of oil to biodiesel was calculated fromthe methyl ester content using the following equation [23]:

Yield ð%Þ ¼ mactual

mtheoretical� 100% � Cester � Vester

moil� 100%

� Cester � Vmoil

� 100% � Cester

qoil� 100%

where both mactual (g) and mtheoretical (g) are the methyl ester mass,moil (g) is the canola oil mass used in the reaction, Cester (g/mL) is themass concentration of methyl ester acquired by GC, qoil (g/mL) isthe canola oil density, and Vester (mL) and Voil (mL) are the crude es-ter layer and canola oil volumes, respectively.

3. Results and discussion

3.1. Textural properties of SBA-15

3.1.1. Effect of aging temperatureThe SAXRD patterns of the SBA-15 obtained using different

aging temperatures are shown in Fig. 1. The results exhibit an in-tense peak lower than 1.1� at 2h, along with two other weak peaksbetween 1.5� and 2.0�, which were assigned to the (100), (110),and (200) planes in a hexagonal arrangement. All SBA-15 supportsshow the optimal ordering of the two-dimensional hexagonalp6mm structure [25], which indicates that a good preservation ofthe ordered mesoporous SBA-15 silica can be achieved even atthe highest aging temperature of 120 �C. The peaks shifted towardthe left side after alter-aging at different temperatures. These datacan be indexed to a hexagonal lattice with a d100 spacing ranging

(a)

(c)

Fig. 2. TEM images of SBA-15 synthesized under (a) 90 �C, (b

from 9.1 nm to 9.8 nm and an average unit cell parameter a0 of10.51 nm to 11.32 nm according to the formula a0 ¼ 2d100=

ffiffiffi

3p

.The lattice parameters increased with increasing aging tempera-ture (Table 1). Table 1 also shows that the average pore size grad-ually increased with increasing aging temperature. However,differences were observed in the S110 and S120 samples. The sur-face area and total pore volume of S110 and S120 decreased whenthe aging temperature was increased to 120 �C. This change wasdue to the decreased fraction of microporosity present in the pores,thereby resulting in a small surface area and pore volume [26].

The TEM images also confirmed the well-ordered hexagonalstructure of the four SBA-15 samples (Fig. 2). Fig. 3 shows the FES-EM images of the SBA-15 supports, which possess long rod-shapedstructures. The S90 had a length of about 2 lm and a diameter of0.5 lm. S100, S110, and S120 became longer with increasing agingtemperature. After aging at a high temperature, the length in-creased and the diameter was extended from 0.5 lm to 2 lm.The lattice parameters and particle size expansion can be attrib-uted to the temperature-dependent hydrophilicity of PEO. Accord-ing to a previous study [27], the segments of the triblockcopolymer PEO–PPO–PEO are in a hydrophilic state at low temper-atures, and the PEO blocks become hydrophobic at a high temper-ature. Moreover, the PEO segment becomes longer at hightemperatures; thus, the hydrophobic volume increases and thepore wall becomes thinner. As a result, the lattice parametersand particle size increase rapidly with increasing aging tempera-ture. Moreover, the pore size and pore volume also increase.

3.1.2. Effect of Si/Al ratioThe chemical compositions of the synthesized Al-SBA-15 sup-

ports are presented in Table 1. The Si/Al molar ratio of all samplessynthesized using the direct synthesis method was in close agree-ment with the expected X value (50, 30, or 10). This result indicates

(b)

(d)

) 100 �C, (c) 110 �C, and (d) 120 �C aging temperatures.

(a1) (a2)

(b2)

(c2)

(d2)

(b1)

(c1)

(d1)

Fig. 3. FESEM images of SBA-15 aged at (a) 90 �C, (b) 100 �C, (c) 110 �C, and (d) 120 �C. (The left side is 10 K, and the right side is 100 K.)

C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794 789

that Al was mostly incorporated into the SBA-15 framework by theprocedure performed in this study. The SAXRD pattern of the Al-S110(X) supports is shown in Fig. 4. All Al-S110(X) supports withdifferent Si/Al ratios consist of three well-resolved peaks that cor-respond to the reflections (100), (110), and (200) associated withthe p6mm hexagonal symmetry of SBA-15. The intensities of these

signals did not change after Al incorporation; however, their posi-tions shifted slightly to the low-angle region compared with thepattern of the S110 sample. These results demonstrate that the ori-ginal long-range periodicity order of the S110 sample was whollypreserved after Al incorporation, and the pore size increase wasdue to the expansion of the aluminosilicate framework. Table 1

Fig. 4. SAXRD patterns of (a) S110 and Al-S110(X) supports with various Si/Alratios: (b) 50, (c) 30, and (d) 10.

790 C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794

shows that the distance between the (100) plane (d100) shifted tohigher values with increasing Al content. The distance changedfrom 9.4 nm for S110 to 9.7 nm for Al-S110(50). All results in theliterature are in agreement that the initial introduction of Al intothe SBA-15 structure increased the d spacing [28,29], which is re-lated to the longer AlAO bond compared with the SiAO bond [28].However, the d spacing did not show any significant increase whenthe Al content was further increased. Similar results were alsofound by Vinu et al. [29]; however, the reasons for such an obser-vation are not yet clear.

The textural characteristics of Al-containing SBA-15 are alsoshown in Table 1. The results show that a high surface area anda high pore volume of the hexagonal mesoporous Al-SBA-15 wasobtained from Al-S110(50). The pore size, surface area, and porevolume decreased with increasing Al content, which is in goodagreement with the data reported in the literature [28]. The FESEMimages of Al-SBA-15 supports with various Si/Al ratios (Fig. 5)show a well-organized hexagonal mesopore structure, in agree-ment with the results of the SAXRD patterns and nitrogen sorptionisotherms. Some ‘‘impurities’’ were observed on the outer surfacewhen the Al content was increased to Si/Al = 10, which indicatesthat further polymerization and organization of the aluminosili-cates occurred after the Al content was increased [28].

3.2. Surface acidity of SBA-15 and Al-SBA-15

The surface acidity of SBA-15 aged at different temperaturesand with Al incorporation was determined by the FT-IR spectrumof the pyridine adsorption technique. Fig. 6 shows that differentsignals were observed in the region between 1700 and1400 cm�1. The bands at 1440 and 1596 cm�1 can be assigned tohydrogen-bonded pyridine, bands at 1446 and 1621 cm�1 to thestrong Lewis acid-bound pyridine, and a band at 1580 cm�1 tothe weak Lewis acid-bound pyridine. The bands corresponding tothe vibration of the pyridinium ion ring associated with Brønstedacid sites were observed at 1546 and 1639 cm�1, and a band at1492 cm�1 was assigned to the pyridine associated with bothBrønsted and Lewis sites [30].

All pure siliceous SBA-15 supports aged at different tempera-tures exhibited signals because of the pyridine adsorption of the

hydrogen bonds with the silanol groups (a shoulder at 1440 cm�1

and a band at 1596 cm�1) and pyridine adsorbed on Lewis acidsites (absorption peaks at 1446, 1485, 1580, and 1639 cm�1). TheBrønsted acid sites observed in the spectrum of the SBA-15 sup-ports were weak.

Compared with the S110 sample, the intensity of the bands cor-responding to the hydrogen and Lewis acid sites of the Al-incorpo-rated S110 sample decreased remarkably. This result indicates thatAl incorporation in the S110 leaves fewer free silanol groups andLewis acid sites. At the expense of these bands, new signals at1546 and 1639 cm�1 were observed in the spectra of the Al-S110(X) samples, which suggests that Brønsted acid sites weregenerated in these samples. These results indicate that the incor-poration of Al created Brønsted acid sites and enhanced the acidstrength of both the Lewis and Brønsted acid sites.

The Al3+ charge was lower than the charge of the Si4+ ions. Thenet negative charge of AlAOASi clearly differed from SiAOASiwhen the higher charged Si4+ was replaced by the lower chargedAl3+ ion in the silica framework. This observation led to changesin the electron density around Si [19], which is one of the possiblereasons for the increase in the generation of Brønsted acid sites onthe Al-S110(X) solid.

3.3. Transesterification activity

The catalytic activity of the Na catalyst supported on SBA-15and Al-S110(X) materials with different textural structures andacid strengths was assessed in the transesterification of canolaoil by using methanol as a short chain alcohol. Fig. 7 shows thatNa/S90 exhibited FAME yields of 16% within 6 h. Compared withNa/S90, catalysts Na/S100, Na/S110, and Na/S120 exhibited largermean pore diameters, thereby improving FAME yields from 16%to 28%. Although the FAME yields of biodiesel with Na as the cata-lyst and a series of SBA-15 supports showed low and similar results(from 16% to 28%), the conversion rate increased slightly withincreasing aging temperature.

In this study, the mean pore sizes of SBA-15 were enlarged byaging at a higher temperature because of the increase in PEO vol-ume (Table 1). The surface area and the total pore volume werealso reduced significantly; however, this reduction did not exhibitany significant influence on FAME yields. Na/S120, which has thelowest surface area and pore volume, showed the highest activityin FAME yields compared with Na/S90. The mean pore size of theSBA-15 support is essential for canola oil to enter the catalytic cen-ters of the catalysts. Therefore, the mesoporous SBA-15 materialsconsisting of large mesopores can significantly minimize the diffu-sion resistance for reactants to obtain access to the catalytic cen-ters of the catalyst [31].

The FAME yields of Na/Al-S110(X) catalysts, which are ex-pressed as the Si/Al ratio, are also presented in Fig. 7. The transe-sterification activity improved significantly when Al-S110(X) wasused as the supporting material. The FAME yields increased signif-icantly from 25% to approximately 99% when the Si/Al ratio de-creased from 1 to 10. The highest yields were obtained after theSi/Al ratio reached 30. The incorporation of alumina into theSBA-15 framework accelerated the transesterification reaction sig-nificantly compared with pure silica SBA-15 supports. The acidityof the catalysts strongly influences the catalytic activity of transe-sterification. Jiménez-Morales et al. [14] observed a similar behav-ior with respect to the Si/Al ratio for Al-SBA-15. They studied threeAl-SBA-15 materials with Si/Al ratios of 10, 20, and 30, which wereprepared by post-synthesis alumination for the methanolysis ofsunflower oil. In their study, the catalytic activity increased withincreasing Si/Al ratio but not in proportion to the Al content in

)2b()1b(

)2c()1c(

(d2)(d1)

(a2)(a1)

Fig. 5. FE-SEM image of (a) SBA-15 aged at 110 �C and Al-S110(X) supports with various Si/Al ratios: (b) 50, (c) 30, and (d) 10.

C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794 791

the structure, which can be attributed to the presence of the Lewisacid site.

The increase in the formation of biodiesel can also be related tothe higher hydrophobic surface characteristic, which facilitates theentry of hydrophobic triglyceride molecules to the active catalyticsites. The surface hydrophobic characteristic is essential in masstransfer for the application of Al-SBA-15(X) porous support in the

transesterification of canola oil to biodiesel and in facilitating thediffusion of high molecular weight and hydrophobic triglyceridemolecules involved in this reaction to the catalytic centers. There-fore, studying the hydrophobic characteristics of the Al-S110(X)surface in the canola oil/methanol binary liquid phase is essential.Szczodrowski et al. [23] used a liquid-flow calorimeter to measurethe integral enthalpy of the competitive adsorption of butanol from

0 200 400 600 800 1000

20

40

60

80

100

35 40 45 50 55 60 65 70 750

20

40

60

80

100

0.0 0.2 0.4 0.6 0.8 1.0 1.2

20

40

60

80

100FA

ME

yie

ld (

%)

BET surface area m2/g

R2=-0.36618

FAM

E y

ield

(%

)

Average pore diameter (A)

R2=0.85764

FAM

E y

ield

(%

)

Total pore volume cm3/g

R2=-0.08294

Fig. 8. Correlation between the total pore volume, average pore diameter, BETsurface area, and the FAME yield of supported Na catalysts.

(a) (b) (c)

Fig. 6. IR spectra of pyridine adsorbed on (a) Na/SBA-15 catalysts aged at different temperatures and (b) Na/Al-S110 (X) catalysts with various Si/Al ratio, and (c) Na/acidicAl2O3, Na/basic Al2O3, Na/SiO2, and Na/hydrotalcite. H: hydrogen-bonded pyridine; B: Brönsted-bonded pyridine; L: Lewis-bound pyridine.

a b d d e f g h i j k0

20

40

60

80

100

Yie

ld o

f FA

ME

(%

)

Kind of catalysts

SBA-15 Al-S110(X) Commercial materials

Fig. 7. Yield of FAME transesterificated from canola oil over Na catalysts supportedon (a) S90, (b) S100, (c) S110, (d) S120, (e) Al-S110(50), (f) Al-S110, (30), (g) Al-S110,(10), (h) acidic Al2O3, (i) basic Al2O3, (j) SiO2, and (k) hydrotalcite.

792 C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794

n-heptane solution, i.e., adsorption of a bipolar solute from a binaryliquid solution in an apolar solvent. Their results indicated that thedensity of the surface ‘‘polar centers’’ in Al-SBA-15 was enhancedbecause of the incorporation of Al, which created additional surfacehydroxyl groups related to ‘‘bridging’’ the Si(OH)Al hydroxyl struc-tures with strong Brønsted acidity.

Considering all the information obtained from the catalyst char-acterization, we can conclude that the Na/Al-S110(30) and Na/Al-S110(10) catalysts exhibit a higher formation of FAME. This highlevel of FAME formation may be attributed to the surface acidityand the hydrophobic characteristics, where triglyceride moleculesare assumed to be chemisorbed [23,32]. Surface area and porestructure did not have a significant effect on the formation of FAMEbecause the development of the Al-S110(10) pore structure wasnot better than the development of the Al-S110(50) pore structure.

3.4. Comparison with other supports

In investigating the influence of the support characteristic onthe transesterification activity with SBA-15 or Al-SBA-15 as the

support, a comparison with several commercial and widely usedmaterials can be helpful. Thus, the surface characteristics of acidicand basic c-Al2O3, SiO2, and hydrotalcite and the transesterifica-tion activity of canola oil were investigated. Fig. 8 shows thatAl2O3, SiO2, and hydrotalcite have intermediate activity comparedwith SBA-15 and Al-SBA-15. The pore structure and surface charac-

Table 2Pore structure of acidic Al2O3, basic Al2O3, SiO2, and hydrotalcite materials.

Support SBETa (m2/g) Db (Å) Vtotal

c BJH adsorption cumulative pore volumes of pores between 8.5 and 1500 Å radius (cm3/g)

Vmicro Vmeso Vmacro

Al2O3-acidic 134 72 0.2408 0.0517 0.1801 0.0090Al2O3-basic 150 69 0.2569 0.0567 0.1906 0.0096SiO2 7 208 0.0389 0.0025 0.0322 0.0043Hydrotalcite –d – – – – –

a SBET is the surface area measured by the BET method.b D is the average pore diameter (4 V/S) determined by BET adsorption method.c Vtotal is the total pore volume.d The pore structure of hydrotalcite cannot be analyzed because of the moisture release continuously during the analysis.

Table 3Representative biodiesel conversion of various catalysts.

Catalysts Optimum reaction conditions Conversions Ref.

Oil source Methanol/oil molarratio

Temperature(�C)

Time(h)

Catalyst amount (wt.% ofoil)

Metal loading weight(wt.%)

KI/mesoporoussilica

Soybean oil 16 70 10 5 15 94 [1]

SZr-SBA-15 Triacetin oil 12 60 12 5.3 – 72.9 [16]Cs/SBA-15 Canola oil 40 135 5 – 2 25.35 [7]20Al-SBA-15 Sunflower 12 200 6 5 – 92 [14]Ca/SBA-15 Sunflower

oil27 200 8 5 0.5 (atomic ratio) 99 [26]

Na/Al-SBA-15 Canola oil 6 65 6 5 99 Presentwork

KNO3/Al2O3 Jatropha oil 12 70 8 6 35 85 [5]KF/c-Al2O3 Vegetable

oil15 (alcohol) 65 8 3 15 97.7 [10]

C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794 793

teristics of acid and basic c-Al2O3, SiO2, and hydrotalcite supportswere noticeably distinct from a series of SBA-15 and Al-SBA-15supports. These supports possess low surface area, large porediameter, low pore volume (Table 2), and stronger acidic strength(Fig. 6) compared with SBA-15 and Al-SBA-15 supports. The corre-lation between total pore volume, average pore diameter, BET sur-face area, and activity with Na as the catalyst in all supportmaterials is plotted in Fig. 8. The most influential parameters aretotal acidity and mean pore diameter. A linear relationship was ob-tained for the average pore diameter with a correlation coefficientof R2 = 0.85764. Moreover, Al-SBA-15 showed a higher acidity andtransesterification activity compared with other mesoporousmaterials.

The results reported in Table 3 represent various catalysts pre-pared and characterized by different investigators for transesterifi-cation of vegetable oil to biodiesel reported in the literature[1,5,7,10,14,16,26]. The present work reported here shows superiorperformance over Al-incorporated SBA-15 supported Na catalysts.As shown, a biodiesel yield of as high as 99% was achieved onthe Na/Al-SBA-15 catalyst at a relative lower reaction temperatureof 65 �C and lower reaction time of 6 h compared with other SBA-15 supported catalysts. In the meantime, no deactivation wasfound after three cycles of Na/Al-SBA-15 catalysts prepared in thisstudy in the transesterification reaction, which showed a betterstability of chemical structure.

4. Conclusions

Mesoporous SBA-15 and Al-SBA-15 supports were synthesizedthrough hydrothermal reaction. These pure siliceous and alu-mina-silica SBA-15s were examined by XRD, TEM, SEM, and BETmethods. The results indicate that a hexagonal mesoporous struc-ture was obtained for all Si/Al ratios. The pyridine adsorption anal-

ysis indicates that these materials possess a variety of Brønstedand Lewis acid sites that vary with aging temperature and Si/Al ra-tio of the materials. The canola oil transesterification activity ofthese materials that support Na catalysts indicates that the cata-lytic activity depends strongly on surface acidity rather than onpore structure. Other traditional mesoporous materials, namely,acidic and basic c-Al2O3, SiO2 and hydrotalcite were also com-pared. The correlation between surface area, mean pore diameter,pore volume, acidity, and transesterification activity suggests thatthe total acidity, especially on the Brønsted acid site, is responsiblefor the canola oil transesterification activity. Na/Al-S110(10),which exhibits high acid strength, provides the highest catalyticactivity, even when its pore structure is not well developed.

References

[1] C. Samart, P. Sreetongkittikul, C. Sookman, Heterogeneous catalysis oftransesterification of soybean oil using KI/mesoporous silica, Fuel Process.Technol. 90 (2009) 922–925.

[2] H.J. Kim, B.S. Kang, M.J. Kim, Y.M. Park, D.K. Kim, J.S. Lee, K.Y. Lee,Transesterification of vegetable oil to biodiesel using heterogeneous basecatalyst, Catal. Today 93–5 (2004) 315–320.

[3] S.P. Singh, D. Singh, Biodiesel production through the use of different sourcesand characterization of oils and their esters as the substitute of diesel: areview, Renew. Sustain. Energy Rev. 14 (2010) 200–216.

[4] R. Sree, N. Seshu Babu, P.S. Sai Prasad, N. Lingaiah, P.S. Sai Prasad,Transesterification of edible and non-edible oils over basic solid Mg/Zrcatalysts, Fuel Process. Technol. 90 (2009) 152–157.

[5] A.P. Vyas, N. Subrahmanyam, P.A. Patel, Production of biodiesel throughtransesterification of Jatropha oil using KNO3/Al2O3 solid catalyst, Fuel 88(2009) 625–628.

[6] I. Jimenez-Morales, J. Santamaria-Gonzalez, P. Maireles-Torres, A. Jimenez-Lopez, Methanolysis of sunflower oil catalyzed by acidic Ta2O5 supported onSBA-15, Appl. Catal. A: General 405 (2011) 93–100.

[7] H. Kazemian, B. Turowec, M.N. Siddiquee, S. Rohani, Biodiesel production usingcesium modified mesoporous ordered silica as heterogeneous base catalyst,Fuel (2012), http://dx.doi.org/10.1016/j.fuel.2012.07.058.

[8] X. Liu, H. He, Y. Wang, S. Zhu, X. Piao, Transesterification of soybean oil tobiodiesel using CaO as a solid base catalyst, Fuel 87 (2008) 216–221.

794 C. Liang et al. / Chemical Engineering Journal 223 (2013) 785–794

[9] D. Vujicic, C. Comic, A. Zarubica, R. Micic, G. Boskovic, Kinetics of biodieselsynthesis from sunflower oil over CaO heterogeneous catalyst, Fuel 89 (2010)2054–2061.

[10] N. Boz, N. Degirmenbasi, D.M. Kalyon, Conversion of biomass to fuel:transesterification of vegetable oil to biodiesel using KF loaded nano-c-Al2O3

as catalyst, Appl. Catal. B: Environ. 89 (2009) 590–596.[11] N. Barakos, S. Pasias, N. Papayannakos, Transesterification of triglycerides in

high and low quality oil feeds over an HT2 hydrotalcite catalyst, Bioresour.Technol. 99 (2008) 5037–5042.

[12] L. Gao, G. Teng, G. Xiao, R. Wei, Biodiesel from palm oil via loading KF/Ca–Alhydrotalcite catalyst, Biomass Bioenergy 34 (2010) 1283–1288.

[13] A. Demirbas, Biodiesel from waste cooking oil via base-catalytic andsupercritical methanol transesterification, Energy Convers. Manage. 50(2009) 923–927.

[14] I. Jiménez-Morales, J. Santamaria-Gonzalez, P. Maireles-Torres, A. Jimenez-Lopez, Aluminum doped SBA-15 silica as acid catalyst for the methanolysis ofsunflower oil, Appl. Catal. B: Environ. 105 (2011) 199–205.

[15] R.V. Grieken, J.M. Escola, J. Moreno, R. Rodríguez, Direct synthesis ofmesoporous M-SBA-15 (M = Al, Fe, B, Cr) and application to 1-hexeneoligomerization, Chem. Eng. J. 155 (2009) 442–450.

[16] Y. Du, S. Liu, Y. Zhang, F. Nawaz, Y. Ji, F.S. Xiao, Urea-assisted synthesis ofhydrothermally stable Zr-SBA-15 and catalytic properties over their sulfatedsamples, Microporous Mesoporous Mater. 121 (2009) 185–193.

[17] Q. Li, Z. Wu, B. Tu, S.S. Park, C.S. Ha, D. Zhao, Highly hydrothermal stability ofordered mesoporous aluminosilicates Al-SBA-15 with high Si/Al ratio,Microporous Mesoporous Mater. 135 (2010) 95–104.

[18] M. Kumaran, S. Garg, K. Soni, M. Kumar, L.D. Sharma, M. Dhar, K.S. Rama Rao,Effect of Al-SBA-15 support on catalytic functionalities of hydrotreatingcatalysts: I. Effect of variation of Si/Al ratio on catalytic functionalities, Appl.Catal. A: General 305 (2006) 123–129.

[19] T. Klimova, J. Reyes, O. Gutiérrez, L. Lizama, Novel bifunctional NiMo/Al-SBA-15 catalysts for deep hydrodesulfurization: effect of support Si/Al ratio, Appl.Catal. A: General 335 (2008) 159–171.

[20] S. Saravanamurugan, Sujandi, D.S. Han, J.B. Koo, S.E. Park, Transesterificationreactions over morphology controlled amino-functionalized SBA-15 catalysts,Catal. Commun. 9 (2008) 158–163.

[21] M. Kumaran, S. Garg, K. Soni, M. Kumar, J.K. Gupta, L.D. Sharma, K.S. Rama Rao,G. Murali Dhar, Synthesis and characterization of acidic properties of Al-SBA-15 materials with varying Si/Al ratios, Microporous Mesoporous Mater. 114(2008) 103–109.

[22] W. Thitsartarn, S. Kawi, Transesterification of oil by sulfated Zr-supportedmesoporous silica, Ind. Eng. Chem. Res. 50 (2011) 7857–7865.

[23] K. Szczodrowski, B. Prlot, S. Lantenois, J.M. Douillard, J. Zajac, Effect ofheteroatom doping on surface acidity and hydrophilicity of Al, Ti, Zr-dopedmesoporous SBA-15, Microporous Mesoporous Mater. 124 (2009) 84–93.

[24] P. Bhange, D.S. Bhange, S. Pradhan, V. Ramaswamy, Direct synthesis of well-ordered mesoporous Al-SBA-15 and its correlation with the catalytic activity,Appl. Catal. A: General 400 (2011) 176–184.

[25] H.Y. He, T. Wang, S.L. Zhu, Continuous production of biodiesel fuel fromvegetable oil using supercritical methanol process, Fuel 86 (2007) 442–447.

[26] H. Sun, J. Han, Y. Ding, W. Li, J. Duan, P. Chen, H. Lou, X. Zheng, One-potsynthesized mesoporous Ca/SBA-15 solid base for transesterification ofsunflower oil with methanol, Appl. Catal. A: General 390 (2010) 26–34.

[27] A. Katiyar, S. Yadav, P.G. Smirniotis, N.G. Pinto, Synthesis of ordered large poreSBA-15 spherical particles for adsorption of biomolecules, J. Chromatogr. A1122 (2006) 13–20.

[28] T.P.B. Nguyen, J.W. Lee, W.G. Shim, H.Y. Moon, Synthesis of functionalized SBA-15 with ordered large pore size and its adsorption properties of BSA,Microporous Mesoporous Mater. 110 (2008) 560–569.

[29] R. Grieken, J.M. Escola, J. Moreno, R. Rodríguez, Direct synthesis of mesoporousM-SBA-15 (M = Al, Fe, B, Cr) and application to 1-hexene oligomerization,Chem. Eng. J. 155 (2009) 442–450.

[30] A. Vinu, V. Murugesan, W. Bohlmann, M. Hartmann, An optimized procedurefor the synthesis of Alsba-15 with large pore diameter and high aluminumcontent, J. Phys. Chem. B 108 (2004) 11496–11505.

[31] L.F. Chen, L.E. Norea, J. Navarrete, J.A. Wang, Improvement of surface acidityand structural regularity of Zr-modified mesoporous MCM-41, Mater. Chem.Phys. 97 (2006) 236–242.

[32] M.K. Lam, K.T. Lee, A.R. Mohamed, Homogeneous, heterogeneous andenzymatic catalysis for transesterification of high free fatty acid oil (wastecooking oil) to biodiesel: a review, Biotechnol. Adv. 28 (2010) 500–518.