Bioresource Technology 156 (2014) 329–334

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Biodiesel production from transesterification of palm oil with methanol over CaO supported on bimodal meso-macroporous silica catalyst Thongthai Witoon a,b,c,⇑, Sittisut Bumrungsalee a, Peerawut Vathavanichkul a, Supaphorn Palitsakun a, Maythee Saisriyoot a, Kajornsak Faungnawakij d a

Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand Center for Advanced Studies in Nanotechnology and Its Applications in Chemical Food and Agricultural Industries, Kasetsart University, Bangkok 10900, Thailand NANOTEC-KU-Center of Excellence on Nanoscale Materials Design for Green Nanotechnology, Kasetsart University, Bangkok 10900, Thailand d Nanomaterials for Energy and Catalysis Laboratory, National Nanotechnology Center (NANOTEC), National Science and Technology Development Agency (NSTDA), Klong Luang, Pathumthani 12120, Thailand b c

h i g h l i g h t s  A series of CaO-loaded unimodal and bimodal porous silica catalysts were prepared.  Basicity increased with increasing the amount of CaO content.  The presence of macropores enhanced the accessibility of CaO inside the pellet.  Bimodal catalyst exhibited a high %FAME with excellent regeneration efficiency.

a r t i c l e

i n f o

Article history: Received 20 December 2013 Received in revised form 15 January 2014 Accepted 19 January 2014 Available online 27 January 2014 Keywords: Biodiesel Calcium oxide Pellet size Heterogeneous catalyst Bimodal porous silica

a b s t r a c t Calcium oxide-loaded porous materials have shown promise as catalysts in transesterification. However, the slow diffusion of bulky triglycerides through the pores limited the activity of calcium oxide (CaO). In this work, bimodal meso-macroporous silica was used as a support to enhance the accessibility of the CaO dispersed inside the pores. Unimodal porous silica having the identical mesopore diameter was employed for the purpose of comparison. Effects of CaO content and catalyst pellet size on the yield of fatty acid methyl esters (FAME) were investigated. The basic strength was found to increase with increasing the CaO content. The CaO-loaded bimodal porous silica catalyst with the pellet size of 325 lm achieved a high %FAME of 94.15 in the first cycle, and retained an excellent %FAME of 88.87 after five consecutive cycles. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Biodiesel is a biodegradable and non-toxic fuel derived from plant or algal oils and animal fats via transesterification, which is extensively accepted as a viable alternative to current petroleumderived diesel (Luque et al., 2010). Homogeneous basic catalysts such as sodium and potassium hydroxides are commonly used in the industrial production of biodiesel (Vicente et al., 2004). However, using homogeneous catalysts has several disadvantages such as, corrosion problems, inability to reuse the catalysts and generation of large amount of wastewater (Ma and Hanna, 1999), which is ⇑ Corresponding author at: Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand. Tel.: +66 2579 2083; fax: +66 2561 4621. E-mail address: [email protected] (T. Witoon). http://dx.doi.org/10.1016/j.biortech.2014.01.076 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.

certainly to diminish their attractiveness associated with high reaction rate under mild reaction conditions. Heterogeneous catalysts have the potential to address the disadvantage issues created by the conventional caustic homogeneous catalysts as they provide easier separation techniques, leave the product free of catalyst impurities and exclude the requirement for product neutralization and purification steps (Semwal et al., 2011). Furthermore, the lesser consumption of heterogeneous catalysts with reusability could potentially lead to economical production costs of biodiesel (Dossin et al., 2006; Mbaraka and Shanks, 2006). A large variety of different heterogeneous catalysts have been investigated (Madhuvilakku and Piraman, 2013; Obadiah et al., 2012; Vieira et al., 2013). Among them CaO has been received much interest due to its mild reaction condition, relatively cheap and less impact on environment (Correia et al., 2014; Ho et al.,

330

T. Witoon et al. / Bioresource Technology 156 (2014) 329–334

2012; Tang et al., 2013). Verziu et al. (2011) reported 92% yield of methyl esters in the transesterification of sunflower oil after 2 h at 75 °C with a 4:1 methanol to oil ratio and 1.4 wt% catalyst using thermally activated CaO catalyst. Kouzu et al. (2008) studied the CaO in the transesterification of refined soybean oil, using a 12:1 M ratio of methanol to oil, and obtained 93% yield of methyl esters after 1 h reaction time. However, partial dissolution of Ca2+ from the CaO surface does occur under reaction conditions, this contribution leads to the problem with separation of the Ca2+ from the products. Supporting CaO on high surface area materials such as silica, alumina, and zinc oxide is found to be a good method to improve stability against the dissolution of Ca2+ (Albuquerque et al., 2008; Alba-Rubio et al., 2010; Umdu et al., 2009). The high surface area of the supports also allows a good dispersion of the metal oxide and thus increasing the catalytically active surface. Umdu et al. (2009) have investigated pure CaO and CaO supported on alumina in the transesterification of algal lipids to form biodiesel. The supported catalyst was found to be more active than pure CaO, which could be attributed to the much smaller CaO crystallite size of 5 nm formed on the support while the CaO crystallite size of the pure CaO was 164 nm. Albuquerque et al. (2008) firstly reported the use of CaO-loaded mesoporous silica (MCM-41 and SBA-15) as base catalysts for biodiesel production from sunflower oil. After 5 h, the biodiesel yield was found to be 95% by using 14 wt% CaO-loaded SBA-15 catalyst. Samart et al. (2010) used mesoporous silica as CaO support in the transesterification of soybean oil. The 95.2% yield of methyl esters is achieved after 8 h when using 15 wt% CaO-loaded mesoporous silica at 60 °C with 5 wt% catalyst and a 16:1 methanol: oil ratio. Although ordered mesoporous silicas (MCM-41 and SBA-15) derived catalysts have proven popular candidates for biodiesel production, these supports possess a two dimensionally ordered hexagonal arrangement of isolated 2.2–7.8 nm diameter parallel channels, leading to inefficient use of metal oxide dispersed inside the pores, resulting in low catalytic activity (Wilson and Lee, 2012; Witoon, 2012). Georgogianni et al. (2009) studied the transesterification reaction of rapeseed oil with methanol in the presence of either homogeneous or heterogeneous catalyst. The transesterification reaction using homogeneous catalyst was found to be 1–2 orders of magnitude higher when compared to that of the heterogeneous ones, which could be potentially explained by the diffusion constrains of the bulky rapeseed molecules into the relatively small pore of the solid catalysts. Several attempts have been made to tailor catalyst porosity to minimizing mass transfer limitations of these bulky and viscous C16–C18 triglycerides of free fatty acids. In contrast to ordered mesoporous materials templated by surfactant micelles, mesoporous silica materials synthesized without template (silica xerogel) exhibited three dimensionally (3D) interconnected pore networks, which provided a greater accessibility through the open porous structure where active metal oxide could be anchored. In addition to 3D interconnected pore networks, the incorporation of macopores into mesoporous silica materials can significantly improve active site accessibility (Gheorghiu and Coppens, 2004; Witoon et al., 2011, 2013; Woodford et al., 2012). In the present study, bimodal meso-macroporous silica material was synthesized and used as a CaO supported catalyst for the transesterification of palm oil with methanol. Unimodal mesoporous silica material having the identical mesopore diameter to bimodal meso-macroporous silica material has been prepared for the purpose of comparison. The physical properties of silica supports and CaO-loaded silica supports were characterized by means of N2 adsorption–desorption, X-ray diffraction (XRD), scanning electron microscope (SEM), CO2 temperature-programmed desorption (TPD) and X-ray fluorescence (XRF). The influences of the CaO

content and the pellet size of the catalysts on the yield of methyl esters were investigated. 2. Methods 2.1. Materials Chitosan with 80% deacetylation was purchased from Eland Corporation. Calcium nitrate tetrahydrate, acetic acid, hydrochloric acid, and sodium hydroxide were purchased from Sigma–Aldrich Company. Sodium silicate (Na2Si3O7: 30 wt% SiO2, 4 wt% NaOH) was obtained from Thai Silicate Company. All chemicals and reagents are of analytical grade and used without any further purification. 2.2. Preparation of unimodal and bimodal porous silica materials Bimodal (meso–macro) porous silica was prepared via a sol–gel process using sodium silicate as a silica source and chitosan as a natural template. In a typical synthesis, 0.4 g chitosan was dissolved in 100 mL of 1% v/v acetic acid in deionized water at room temperature for 12 h. Then 5.5 g sodium silicate, primarily diluted with 10 mL deionized water, was slowly added to the chitosan solution under vigorous stirring. Afterwards, the pH value of the mixture was quickly adjusted to 6 by the addition of 2 M HCl or 2 M NaOH solution, respectively. The mixture was stirred at 40 °C for 3 h and after that it was poured into a Teflon container and aged in autoclave at 100 °C for 24 h. The obtained product was filtered, washed several times with deionized water, dried at 120 °C for 12 h and calcined at 800 °C for 4 h at a heating rate of 2 °C/min. Unimodal porous silica having equivalent mesopore diameter to bimodal porous silica was synthesized using the similar condition as mentioned above except for the addition of chitosan. 2.3. Preparation of calcium oxide-loaded porous silica materials Series of XCaO/U and XCaO/B catalysts were prepared by incipient wetness impregnation method, in which X was the amount of CaO content. The unimodal and bimodal porous silica supports were impregnated with the desired amount of calcium nitrate tetrahydrate (Ca(NO3)2 4H2O) in aqueous solution. The slurry mixture was stirred at 60 °C for 1 h, dried at 120 °C for 12 h and calcined at 800 °C for 4 h at a heating rate of 2 °C/min. 2.4. Characterization BET surface area and total pore volume of the samples were measured at 196 °C with a Quantachrome Autosorb-1C instrument (USA). Prior to measurements, the samples were degassed at 200 °C for 24 h. Pore size distributions of the samples were determined from the adsorption branch of the isotherms in accordance with the Barrett–Joyner–Hallenda (BJH) method (Gregg and Sing, 1982). The specific BET (SBET) was estimated for P/P0 values to be between 0.05 and 0.30. The total pore volume was measured at the relative pressure (P/P0) of 0.995. X-ray diffraction (XRD) patterns of the catalysts were attained on a diffractometer (Bruker D8 Advance) with Cu-Ka radiation. The measurements were made at temperatures in a range of 15–70 °C on 2h with a step size of 0.05°. The diffraction patterns were analyzed with the employment of the Joint Committee on Powder Diffraction Standards (JCPDS). The CaO crystallite size was calculated by means of the Scherrer Equation from the most intense CaO peak at 2h of 37.4 as shown below (Patterson, 1939):

331

T. Witoon et al. / Bioresource Technology 156 (2014) 329–334 

ð1Þ

where d denotes the mean crystallite diameter, k the X-ray wave length (1.54 Å), and B the full width half maximum (FWHM) of the CaO diffraction peak. X-ray fluorescence (Epsilon 5, PANalytical B.V., Netherlands) was used for the quantitative analysis of the CaO content in the recovered catalysts. The calcined samples were mixed with a binder (ratio of 1:9) and then pressed into tablets (Eze et al., 2013). The operation conditions were an accelerating voltage of 60 kV, a maximum current of 100 mA and a measuring time of 30 min. The spectra were analyzed by means of the Omnian software of PANalytical B.V. All the samples were analyzed in triplicates. The morphologies of the samples were assessed with the application of a field emission scanning electron microscope (FE-SEM S4700, Hitachi, Japan). The SEM was operated at 20 kV of an accelerating voltage. The samples were sputter-coated with platinum prior to analysis. Basicity of the catalysts was measured with a SDT2960 simultaneous DTA-TGA Universal 2000. The sample (20 mg) was loaded into an alumina sample pan. Prior to any CO2 adsorption experiment and to remove pre-adsorbed CO2 and H2O, the sample was first activated by heating it from room temperature in a flow of pure N2 (100 mL/min) at a rate of 20 °C/min until 800 °C was achieved and kept constant for 30 min, then the sample was cooled to 40 °C. When the sample equilibrated at the specified temperature (40 °C), the furnace purge was switched from the N2 gas to the pure CO2. After approximately 10 min the sample reached a constant mass, the furnace purge was switched back to the N2 gas for 30 min to remove the physisorbed CO2. Then, the sample was heated form 40 °C to 800 °C with a ramp rate of 10 °C/min to observe the desorption of CO2. 2.5. Transesterification Transesterification reactions were performed in a batch reactor. A mixture of methanol and catalyst was added to a 250 mL two-necked round-bottom flask fitted with a reflux condenser (Ngamcharussrivichai et al., 2007). The mixture was heated at 60 °C with stirring for 5 min. Then, palm oil (25 mL) preheated at 60 °C was added into the mixture. The transesterification was carried out under a condition of methanol/oil molar ratio of 12:1 (Viriya-empikul et al., 2010) and catalyst amount of 5 wt% (Samart et al., 2010) for the required reaction times. To finish the reaction, the catalyst was separated from the mixture by centrifugation, and then excessive amount of methanol was evaporated before analysis of %FAME. The duplicate experiments were performed and the errors of %FAME contents were found to be within ±2.0 wt%. The compositions of biodiesel were determined by a gas chromatograph (GC-2014, Shimadzu, Japan) equipped with capillary column (DB-WAX) and a flame ionization detector. Methyl heptadecanoate was used as internal standard for quantification, according to EN14103 standard method (Wan and Hameed, 2011). The biodiesel content was represented in term of %FAME as a function of time. 3. Results and discussion 3.1. Characterization of the catalysts N2 adsorption–desorption isotherms and pore size distributions of CaO-loaded porous silica catalysts (XCaO/U and XCaO/B) are shown in Fig. 1. The adsorption–desorption of unimodal porous silica (0CaO/U) is typical IV-type pattern with H2-type hysteresis loop (Fig. 1a), which is a dominant characteristic of mesoporous

1600

Volume of gas adsorbed (cm3/g)

0:89k 180  B cos h p

(a)

1200

800

0CaO/U 10CaO/U 20CaO/U 30CaO/U 40CaO/U 50CaO/U

(b)

0CaO/B 10CaO/B 20CaO/B 30CaO/B 40CaO/B 50CaO/B

Type IV

Type II

Type IV

400

0 0

Adsorption dV/d(log D) (cm3/g)

d¼

12

8

4

0.2 0.4 0.6 0.8 1 0 Relative pressure (P/P0)

0.2 0.4 0.6 0.8 1 Relative pressure (P/P0)

50CaO/U

50CaO/B

40CaO/U

40CaO/B

30CaO/U

30CaO/B

20CaO/U

20CaO/B

10CaO/U

10CaO/B

0CaO/U

(c)

0CaO/B

(d)

0 1

10 100 Pore diamter (nm)

500 1

10 100 Pore diamter (nm)

500

Fig. 1. N2 adsorption–desorption isotherms (a and b) and pore size distributions (c and d) of CaO-loaded unimodal (XCaO/U) and CaO-loaded bimodal (XCaO/B) porous silica catalysts.

structure with ink-bottle shape. With 10–30 wt% CaO addition (Fig. 1a), the hysteresis loop is shifted toward higher relative pressure, an indication of an increase of average pore size. At 40– 50 wt% CaO loading (Fig. 1a), the isotherms are found to be almost parallel to the x-axis, which is a characteristic of non-porous material meaning that the pores of the silica support are entirely filled by CaO particles. The isotherm of the bimodal porous silica (0CaO/B) displays composite type IV–II isotherms (Fig. 1b), indicating that the 0CaO/B sample contains both meso-macroporous structures. Similar to the XCaO/U catalysts, the amount of N2 adsorbed of XCaO/B catalysts is found to decrease with increasing the amount of CaO loading. However, at high CaO content (40–50 wt%), the catalysts show the volume of N2 adsorbed at a relative pressure in excess of 0.90, an indication of a large number of macropores to be present in the structure of the 40CaO/B and 50CaO/B catalysts. The pore size distribution of the 0CaO/U sample reveals a very narrow and unimodal pore size distribution in the range of mesoporous region with mean pore diameter of 12.2 nm (Fig. 1c), while that of the 0CaO/B sample shows the bimodal pore size distribution in the range of mesopore and macropore regions with mean pore diameter of 12.2 and 124 nm, respectively. In comparison to the original porous silica (0CaO/U and 0CaO/B samples), the mean pore size in the range of mesopore region is found to gradually shift toward larger pore diameter with a reduction of peak intensity when 10–30 wt% CaO is added. The peak intensity of the pore size in the range of macropore region of 10CaO/B, 20CaO/B and 30CaO/B catalysts is slightly affected by the CaO addition. With a yet higher CaO loading (40 wt%), the peak intensity in the mesopore region of both 40CaO/U and 40CaO/B catalysts is relatively low, however, the moderate peak intensity in the macropore region of 40CaO/B catalyst is still observed. At 50 wt% CaO loading, the peak intensity in the mesopore region of both 50CaO/U and 50CaO/B catalysts disappear, and broad peak in the macropore region of 50CaO/B catalyst is found. Table 1 shows the physical properties including BET surface area, mesopore volume and macropore volume of the catalysts. A

T. Witoon et al. / Bioresource Technology 156 (2014) 329–334

0CaO/U 10CaO/U 20CaO/U 30CaO/U 40CaO/U 50CaO/U

371 ± 2 215 ± 1 110 ± 1 60 ± 1 36 ± 1 22 ± 2

1.04 ± 0.01 0.87 ± 0.01 0.61 ± 0.01 0.44 ± 0.01 0.24 ± 0.01 0.05 ± 0.01

0.06 ± 0.02 0.02 ± 0.01 0.04 ± 0.01 0.02 ± 0.01 0.04 ± 0.02 0.04 ± 0.02

0CaO/B 10CaO/B 20CaO/B 30CaO/B 40CaO/B 50CaO/B

415 ± 1 224 ± 2 139 ± 2 65 ± 1 45 ± 2 24 ± 1

1.14 ± 0.01 0.87 ± 0.01 0.65 ± 0.01 0.51 ± 0.01 0.26 ± 0.01 0.14 ± 0.01

1.20 ± 0.02 1.04 ± 0.03 0.95 ± 0.02 0.65 ± 0.03 0.45 ± 0.02 0.24 ± 0.02

Mesopore volume measured at pores smaller than 50 nm in diameter. Macopore volume measured at pores larger than 50 nm in diameter.

monotonic decrease of BET surface area and mesopore volume with increasing CaO loading was observed for both series of catalysts (XCaO/U and XCaO/B). The unimodal catalysts contained only small amount of macropore volume (0.02–0.06 cm3/g). When compared at the identical CaO loading the macropore volume of the bimodal catalysts was found to be 6–50 times greater than that of the unimodal catalysts. The addition of CaO more than 20 wt% into the bimodal porous silica support (0CaO/B) led to a significant decrease in the macropore volume. This indicates that the CaO particles begin to deposit into the mesopores, and then into the macropores. The surface features of samples before and after impregnation of CaO were examined with SEM images (see Supplementary material, Fig. S1). The unimodal porous silica support (0CaO/U) shows a dense aggregation of fine silica nanoparticles while the bimodal porous silica support (0CaO/B) exhibits loosely packed aggregates of silica domains with inter-particle voids (macropore volume). At 20–30 wt% CaO loading, there are no CaO particles on the surface of the supports, indicating that CaO nanoparticles are highly dispersed inside the particles of both supports. Once the loading amount of CaO is increased to 40 wt%, an aggregation of CaO particles on the external surface of the 40CaO/U catalyst is found, which suggests the loading amount of CaO to be in excess of the maximum uptake capacity of the unimodal porous silica support. In contrast to 40CaO/U catalyst, no aggregation of CaO particles on the external surface of the 40CaO/B catalyst is observed. At 50 wt% CaO loading, CaO particles are formed on the external surface of both 50CaO/U and 50CaO/B catalysts. However, the CaO particle size (approximately 1–2 lm) of the 50CaO/U catalyst is found to be considerably larger than that of the 50CaO/B catalyst (approximately 100–300 nm). The results demonstrate that the bimodal porous silica is more efficient for the CaO loading as a higher dispersion of CaO nanoparticles can be obtained. The XRD analysis was performed to investigate types of calcium species and average CaO crystallite size of the CaO-supported porous silica catalysts (see Supplementary material, Fig. S2). At 10–30 wt%, the XRD patterns of both series (XCaO/U and XCaO/B) exhibit only broad peak, a characteristic peak of amorphous silica, with no appearance of any crystalline calcium-containing phases, which suggests calcium species to be well dispersed in porous silica materials. At 40–50 wt% CaO loading, XRD patterns show a typical pattern of CaO appeared as the major peaks at the identical 2h angles of 32.3°, 37.4°, 53.9°, 64.2° and 67.4°. The most intense peak centered at 37.4 is used to calculate the average CaO crystallite size with Scherrer Equation (see Supplementary material, Fig. S2). The average CaO crystallite size of the 40CaO/U and 50CaO/U catalysts is found to be 22 and 54 nm, respectively, which is considerably

3.2. Biodiesel production activity The catalytic performance of all catalysts for biodiesel production from transesterification of palm oil with methanol is shown in Fig. 3a and b. The catalyst prepared at low amount of CaO content (10CaO/U) was not very active for the reaction as only 6.4%FAME was obtained at 6 h. The reaction rate and %FAME were found to be enhanced with the increase in CaO content, which corresponded well with the basic strength of the catalysts. However, it was noteworthy that the %FAME of the 40CaO/U catalyst was considerably higher than that of the 30CaO/U catalyst. Based on the results of pore size distribution (Fig. 1) and SEM images (see Supplementary material, Fig. S1), the CaO nanoparticles were highly dispersed inside the mesopore of the 30CaO/U catalyst, while an agglomeration of CaO particles on the external surface was observed for the 40CaO/U catalyst. For gas phase reaction, a high dispersion of metal catalyst usually provided a high active surface which undoubtedly enhanced catalytic activity. However, for liquid phase reaction, the enhancement of %FAME over the catalysts with high dispersion of CaO (10CaO/U, 20CaO/U and 30CaO/ U) was not achieved, which suggested a diffusion limitation of reactants to react with CaO inside the mesopore of the catalysts.

485

Macropore volumeb (cm3/g)

0

50CaO/U

50CaO/B

40CaO/U

40CaO/B

30CaO/U

30CaO/B

20CaO/U 10CaO/U

20CaO/B 10CaO/B

200

400

600

Temperature (oC)

800 0

200

400

578

Mesopore volumea (cm3/g)

630

BET surface area (m2/g)

487

a b

Catalysts

larger than that of the 40CaO/B (14 nm) and 50CaO/B catalysts (36 nm) when compared at identical CaO loading. The XRD results were in good agreement with the SEM images. Apart from a CaO phase, the catalysts also contain small peaks of a Ca(OH)2 phase. Efforts were made to protect the samples from the surrounding atmosphere after calcination to keep the samples as dry as possible but it was not possible to completely remove the risk of Ca(OH)2 formation due to high reactivity of CaO towards moisture. The basic site distributions of the catalysts were examined by CO2 temperature-programmed desorption. The CO2-TPD profiles of the catalysts are shown in Fig. 2. The desorption curve over the 10CaO/U catalyst shows a very broad desorption band extending from 400 to 550 °C which can be assigned as interaction of CO2 with site of medium basic strength. When the CaO loading increases to 20 wt% (20CaO/U), the desorption curve displays a narrower band at higher temperature which can be attributed to the presence of much stronger basic site. The CO2 desorption peak is found to gradually shift toward higher temperature ranging from 540 to 630 °C with further increase of CaO content from 30 to 50 wt%, indicating that the basic strength increases with increasing CaO content. In comparison to the XCaO/U catalysts at identical CaO content, the CO2 desorption peak of the XCaO/B catalysts appears at a lower temperature, which suggests a lower basic strength of the XCaO/B catalysts. This demonstrates that the basic strength of the catalysts is increased with increasing the crystallinity of CaO (see Supplementary material, Fig. S2).

560 605

Table 1 BET surface area, mesopore volume and macropore of the catalysts.

CO2 desorption (a.u.)

332

600

800

Temperature (oC)

Fig. 2. CO2-TPD profiles of CaO-loaded unimodal (XCaO/U) and CaO-loaded bimodal (XCaO/B) porous silica catalysts.

333

T. Witoon et al. / Bioresource Technology 156 (2014) 329–334

(a)

96.54

% FAME

85.35

92.45

50CaO/U 40CaO/U

65.66

20.22 30CaO/U 20CaO/U 10CaO/U

10.15

(b)

95.16

% FAME

65.41 24.42

16.25

30CaO/B

20CaO/B 10CaO/B 50CaO/U-82.5 50CaO/B-82.5 50CaO/B-335 50CaO/B-725

% FAME

(c)

50CaO/U-335 50CaO/U-725

1

2

3 4 Time (h)

5

6

Fig. 3. Effect of unimodal porous silica support (a), bimodal porous silica support (b) and catalyst pellet size (c) on %FAME obtained in transesterification of palm oil with methanol at temperature of 60 °C, methanol/oil molar ratio of 12:1 and catalyst amount of 5 wt%.

The presence of macropores of the series of XCaO/B catalysts (Fig. 3b) was expected to diminish the diffusion limitation but only slight increase of %FAME was observed when compared at identical CaO content (10–30 wt%). At high content of CaO (40–50 wt%), there was no sign of %FAME improvement that could be linked to the presence of macropores. The results indicated that the presence of macropores did not significantly improve the mass transfer of the reactants and once the diffusion limitation did not exist for the reaction, the %FAME was mainly related to the basic strength of the catalyst. 3.3. Effect of catalyst pellet size The catalytic performances presented above were carried out using the catalyst with the pellet size of 82.5 lm. It was troublesome to separate the catalysts with small pellet size from the liquid product. However, the catalyst with large pellet size was inevitably subjected to diffusional limitation phenomena. It was therefore important to determine the optimal pellet size of the catalyst. In order to investigate the effect of pellet size on the biodiesel production from the transesterification, the pellet size of the catalysts were sieved and separated into three fractions in the ranges of 75–90 (82.5) lm, 250–420 (335) lm and 600–850 (725) lm, respectively. The selected catalysts for pelletization were 50CaO/U and 50CaO/B catalysts, representative for the

3.4. Leaching and reusability of the catalysts In order to test catalyst reusability, the used catalyst was recovered from the product mixture by centrifugation at 9000 rpm, washed with petroleum ether and methanol to remove some impurities attaching the catalyst, dried at 120 °C overnight and then calcined at 800 °C for 2 h. The recovered catalyst was used for the next transesterification reaction and the amount of CaO content in the recovered catalyst was further investigated with X-ray fluorescence. Fig. 4a shows %FAME for the reusability of the bimodal catalysts prepared at different pellet sizes through five consecutive cycles. The catalyst with the smallest pellet size (50CaO/B-82.5) was found to be unstable as the %FAME was 95.16, 89.78, 85.45, 65.14 and 57.56 when the catalyst was used for the first to the fifth cycles, respectively. As shown in Fig. 4b, the amount of CaO content in the recovered catalyst (50CaO/ B-82.5) was found to considerably decrease from 51.14 to 41.24 wt% after the first round of transesterification reaction, which was an indication of the dissolution of CaO particles distributed at the external surface of the pellets. However, the %FAME was slightly decreased from 95.16 to 89.78, implying that the CaO particles remained into the macropores were very active in transesterification reaction. When the amount of CaO content was lower than 36 wt% (Fig. 4b), a sharp decay of %FAME was

% FAME

50CaO/B 89.87 40CaO/B

78.87

0

unimodal and bimodal catalysts, respectively. As shown in Fig. 3c, the %FAME of both catalysts was found to decrease with the increase in the pellet size, which could be attributed to internal diffusion in the pellets. In comparison to the %FAME of the catalyst with pellet size of 82.5 lm, the reduction of %FAME of the 50CaO/U catalyst with the pellet size of 335 and 725 lm was found to be 0.61 and 0.40, respectively, while that of the 50CaO/B catalyst was found to be 0.98 and 0.83, respectively. The slight decrease in the %FAME of the 50CaO/B catalyst could be attributed to the presence of macropores of the 50CaO/B catalyst which enhanced the accessibility of CaO distributed inside the pellet.

100 90 80 70 60 50 40 30 20 10 0

(a)

50CaO/B-82.5 50CaO/B-335 50CaO/B-725

(b)

50

CaO (wt%)

100 90 80 70 60 50 40 30 20 10 0 100 90 80 70 60 50 40 30 20 10 0 100 90 80 70 60 50 40 30 20 10 0

40 30

50CaO/B-82.5

20

50CaO/B-335

10

50CaO/B-725

0

0

1

2

3

4

5

Number of cycles Fig. 4. Effect of catalyst particle size on %FAME (a) and the amount of CaO content (wt%) (b) with the number of cycles. %FAME obtained in transesterification of palm oil with methanol at temperature of 60 °C, methanol/oil molar ratio of 12:1, catalyst amount of 5 wt% and reaction time of 6 h.

334

T. Witoon et al. / Bioresource Technology 156 (2014) 329–334

observed, an indication of dissolution of CaO particles deposited at macropores. The remaining CaO dispersed inside the mesopores gave a low %FAME due to diffusion limitation problem as mentioned earlier. For the bimodal catalysts with larger pellet size (50CaO/B-325 and 50CaO/B-725), only a small decrease in %FAME was observed after the fifth cycle. The amount of CaO content of the recovered catalysts after the fifth cycle was found to be higher than 42 wt%, implying that a larger amount of CaO particles was still presented in the macropores. This demonstrated that the catalyst with the large pellet size retarded the dissolution of CaO. The preparation cost of the bimodal catalyst (50CaO/B), including the costs for the preparation of bimodal porous silica, the calcium nitrate tetrahydrate and the consumed solvent, was estimated in comparison to that of the unimodal catalyst (50CaO/U). The price for starting materials was investigated using the alibaba.com search engine. Offers of materials high purity were excluded from the search, to make valid comparisons between materials of industrial grade purity. The estimated preparation cost of the bimodal catalyst is about 8.5$/kg, while that of the unimodal catalyst is about 7.0$/kg. The preparation cost of the former is slightly higher than of the latter. This is because only small amount of chitosan is used for the preparation of the bimodal catalyst. 4. Conclusion The purpose of this work is to delineate the effect of bimodal meso-macroporous silica support and catalyst pellet size on %FAME obtained from transesterification. The incorporation of macropores in the structure of the catalyst was found to be enhanced diffusion of reactants to the active sites when the catalyst pellet size was larger than 82.5 lm, leading to a significant enhancement in %FAME compared to the unimodal catalyst. The 50CaO/B-325 catalyst exhibited a slight decrease in %FAME obtained after five consecutive cycles which was a proof of the catalyst stability, making it promising as a heterogeneous catalyst for biodiesel production. Acknowledgements This work was financially supported from Thailand Research Fund to T.W. and K.F., and the National Research University Project of Thailand (NRU). The authors would like to thank the partial support from the Nanotechnology Center (NANOTEC), NSTDA, Ministry of Science and Technology, Thailand through its program of Center of Excellence Network, and the Kasetsart University Research and Development Institute (KURDI). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2014.01. 076. References Alba-Rubio, A.C., Santamaría-González, J., Mérida-Robles, J.M., Moreno-Tost, R., Martín-Alonso, D., Jiménez-López, A., Maireles-Torres, P., 2010. Heterogeneous transesterificaion processes by using CaO supported on zinc oxide as basic catalysts. Catal. Today 149, 281–287. Albuquerque, M.C.G., Jiménez-Urbistondo, I., Santamaría-González, J., MéridaRobles, J.M., Moreno-Tost, R., Rodíguez-Castellón, E., Jiménez-López, A., Azevedo, D.C.S., Cavalcante Jr., C.L., Maireles-Torres, P., 2008. CaO supported on mesoporous silicas as basic catalysts for transesterification reactions. Appl. Catal. A 334, 35–43.

Correia, L.M., Saboya, R.M.A., Campelo, N.D.S., Cecilia, J.A., Rodríguez-Castellón, E., Cavalcante Jr., C.L., Vieira, R.S., 2014. Characterization of calcium oxide catalysts from natural sources and their application in the transesterification of sunflower oil. Bioresour. Technol. 151, 207–213. Dossin, T.F., Reyniers, M.F., Berger, R.J., Marin, G.B., 2006. Simulation of heterogeneously MgO-catalyzed transesterification for fine-chemical and biodiesel industrial production. Appl. Catal. B: Environ. 67, 136–148. Eze, C.P., Nyale, S.M., Akinyeye, R.O., Gitari, W.M., Akinyemi, S.A., Fatoba, O.O., Petrik, L.F., 2013. Chemical, mineralogical and morphological changes in weathered coal fly ash: a case study of a brine impacted wet ash dump. J. Environ. Manage. 129, 479–492. Georgogianni, K.G., Katsoulidis, A.K., Pomonis, P.J., Manos, G., Kontominas, M.G., 2009. Transesterification of rapeseed oil for the production of biodiesel using homogeneous and heterogeneous catalysis. Fuel Process. Technol. 90, 1016– 1022. Gheorghiu, S., Coppens, M.O., 2004. Optimal bimodal pore networks for heterogeneous catalysis. AIChE J. 50, 812–820. Gregg, S.J., Sing, K.S.W., 1982. Adsorption, Surface Area and Porosity, second ed. Academic Press, London. Ho, W.W.S., Ng, H.K., Gan, S., 2012. Development and characterization of novel heterogeneous palm oil mill boiler ash-based catalysts for biodiesel production. Bioresour. Technol. 125, 158–164. Kouzu, M., Kasuno, T., Tajika, M., Sugimoto, Y., Yamanaka, S., Hidaka, J., 2008. Calcium oxide as a solid base catalyst for transesterification of soybean oil and its application to biodiesel production. Fuel 87, 2798–2806. Luque, R., Lovett, J.C., Datta, B., Clancy, J., Campelo, J.M., Romero, A.A., 2010. Biodiesel as feasible petrol fuel replacement: a multidisciplinary overview. Energy Environ. Sci. 3, 1706–1721. Ma, F., Hanna, M.A., 1999. Biodiesel production: a review. Bioresour. Technol. 70, 1– 15. Madhuvilakku, R., Piraman, S., 2013. Biodiesel synthesis by TiO2–ZnO mixed oxide nanocatalyst catalyzed palm oil transesterification process. Bioresour. Technol. 150, 55–59. Mbaraka, I.K., Shanks, B.H., 2006. Conversion of oils and fats using advanced mesoporous heterogeneous catalysts. J. Am. Oil Chem. Soc. 83, 79–91. Ngamcharussrivichai, C., Wiwatnimit, W., Wangnoi, S., 2007. Modified dolomites as catalysts for palm kernel oil transesterification. J. Mol. Catal. A Chem. 276, 24– 33. Obadiah, A., Swaroopa, G.A., Kumar, S.V., Jeganathan, K.R., Ramasubbu, A., 2012. Biodiesel production from palm oil using calcined waste animal bone as catalyst. Bioresour. Technol. 116, 512–516. Patterson, A.L., 1939. The Scherrer formula for X-ray particle size determination. Phys. Rev. 56, 978–982. Samart, C., Chaiya, C., Reubroycharoen, P., 2010. Biodiesel production by methanolysis of soybean oil using calcium supported on mesoporous silica catalyst. Energy Convers. Manage. 51, 1428–1431. Semwal, S., Arora, A.K., Badoni, R.P., Tuli, D.K., 2011. Biodiesel production using heterogeneous catalysts. Bioresour. Technol. 102, 2151–2161. Tang, Y., Xu, J., Zhang, J., Lu, Y., 2013. Biodiesel production from vegetable oil by using modified CaO as solid basic catalysts. J. Clean Prod. 42, 198–203. Umdu, E.S., Tuncer, M., Seker, E., 2009. Transesterification of Nannochloropsis oculata microalga’s lipid to biodiesel on Al2O3 supported CaO and MgO catalysts. Bioresour. Technol. 100, 2828–2831. Verziu, M., Coman, S.M., Richards, R., Parvulescu, V.I., 2011. Transesterification of vegetable oils over CaO catalysts. Catal. Today 167, 64–70. Vicente, G., Martínez, M., Aracil, J., 2004. Integrated biodiesel production: a comparison of different homogenous catalysts systems. Bioresour. Technol. 92, 297–305. Vieira, S.S., Magriotis, Z.M., Santos, N.A.V., Saczk, A.A., Hori, C., Arroyo, P.D., 2013. Biodiesel production by free fatty acid esterification using lanthanum (La3+) and HZSM-5 based catalysts. Bioresour. Technol. 133, 248–255. Viriya-empikul, N., Krasae, P., Puttasawat, B., Yoosuk, B., Chollacoop, N., Faungnawakij, N., 2010. Waste shells of mollusk and egg as biodiesel production catalysts. Bioresour. Technol. 101, 3765–3767. Wan, Z., Hameed, B.H., 2011. Transesterifcation of palm oil to methyl ester on activated carbon supported calcium oxide catalyst. Bioresour. Technol. 102, 2659–2664. Wilson, K., Lee, A.F., 2012. Rational design of heterogeneous catalysts for biodiesel synthesis. Catal. Sci. Technol. 2, 884–897. Witoon, T., 2012. Polyethyleneimine-loaded bimodal porous silica as low-cost and high-capacity sorbent for CO2 capture. Mater. Chem. Phys. 137, 235–245. Witoon, T., Chareonpanich, M., Limtrakul, J., 2011. Effect of hierarchical mesomacroporous silica supports on Fischer–Tropsch synthesis using cobalt catalyst. Fuel Process. Technol. 92, 1498–1505. Witoon, T., Chareonpanich, M., Limtrakul, J., 2013. Effect of bimodal porous silica on particle size and reducibility of cobalt oxide. J. Porous Mater. 20, 481–488. Woodford, J.J., Dacquin, J.-P., Wilson, K., Lee, A.F., 2012. Better by design: nanoengineered macroporous hydrotalcites for enhanced catalytic biodiesel production. Energy Environ. Sci. 5, 6145–6150.