1

This is a revised personal version of the text of the final journal article, which is made available for scholarly purposes only, in accordance with the journal's author permissions. The full citation is:

Olutoye, M.A., Hameed, B.H. Production of biodiesel fuel by transesterification of different vegetable oils with methanol using Al2O3 modified MgZnO catalyst. Bioresour. Technol. 132 (2013) 103-108, http://dx.doi.org/10.1016/j.biortech.2012.12.171 Production of biodiesel fuel by transesterification of different vegetable oils with methanol using Al2O3 modified MgZnO catalyst

M. A. Olutoye, B.H. Hameed* School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia, 14300 Nibong Tebal, Penang, Malaysia *Corresponding author: Fax: +6045941013 e-mail address: chbassim@eng.usm.my (B.H. Hameed)

Abstract

An active heterogeneous Al2O3 modified MgZnO (MgZnAlO) catalyst was prepared and the catalytic activity was investigated for the transesterification of different vegetable oils (refined palm oil, waste cooking palm oil, palm kernel oil and coconut oil) with methanol to produce biodiesel. The catalyst was characterized by using X-ray diffraction, Fourier transform infrared spectra, thermo gravimetric and differential thermal analysis to ascertain its versatility. Effects of important reaction parameters such as methanol to oil molar ratio, catalyst dosage, reaction temperature and reaction time on oil conversion were examined. Within the range of studied variability, the suitable transesterification conditions (methanol/oil ratio 16:1, catalyst loading 3.32 wt%, reaction time 6 h, temperature 182 oC), the oil conversion of 98% could be achieved with reference to coconut oil in a single stage. The catalyst can be easily recovered and reused for five cycles without significant deactivation.

Keywords: Vegetable oils; Transesterification; MgZnAlO catalyst; Biodiesel

2

1. Introduction

The world’s energy utilization has experienced tremendous upswing because of rapid industrialization and increased population (Hoffmann, 2011; Walker, 2010). The anticipation of exhaustibility of the natural supplies and the growing concern for environmental quality are indicators of worldwide rummage for alternative energy sources and raw materials. In this regard, reconnoitering alternative energy resources, such as biodiesel fuel has in recent times been the researchers’ spotlight because of its environmental benefits, biodegradability and renewability (Chouhan & Sarma, 2011; Atadashi et al., 2012). Biodiesel production typically involves the transesterification of a triglyceride feedstock with methanol or other short-chain alcohols.

The usage of heterogeneous catalysts such as metal oxides, carbonates, zeolites and heteropolyacids have been preferred for biodiesel production because they are reusable and easy to separate from reaction products. Also, they are generally much more tolerant to water and free fatty acids (FFAs) in the feedstock and can be designed to give higher activity and longer catalyst lifetimes (Liu et al., 2008). The catalytic activities of most catalysts are carried out in a two-stage process of esterification and transesterification when the FFA in feedstock is higher than 3% (Canakci and Van Gerpen, 2001; Berchmans and Hirata, 2008;) using single type of vegetable oil and in few cases on more than one type of oil (Patil and Deng, 2009). Previously, it was reported the transesetrification of high FFA crude jatropha oil with methanol using MgZnAlO catalyst in a single stage process with FAME yield of 94% at methanol/oil ratio 11:1, catalyst loading, 8.68 wt % and reaction temperature, 182 oC within 6 h reaction time (Olutoye and Hameed, 2011a). The catalyst exhibited bifunctional (acidic and basic) sites, was reusable and easily recovered from reaction products. Based on the encouraging performance of the catalyst, it seems promising to further investigate the applicability of such active heterogeneous catalyst in transesterification of different vegetable oils for prospective industrialist in biorefinery.

Various oils have been used as raw materials for biodiesel production and typical among these is sunflower oil, tobacco seed oil, corn oil and soybean, palm and palm kernel oils, coconut oil, animal fats, or other lipids (Chouhan and Sarma, 2011). In order to hush the intense debate on food for oil, several researchers have investigated other non-edible sources of feedstock to produce biodiesel which includes croton megalocarpus, moringa oleifera and jatropha curcas Linnaeus (Jatropha curcas L.) (Singh and Singh, 2010). The assortment of raw materials, types of catalyst (homogeneous or heterogeneous) and technological routes employed in the biodiesel production has resulted in products with different chemical properties. This non-uniformity in the biodiesel composition from different oils may influence the fuel quality (Pal and Prakash, 2012).

Thus, this paper presents further study on the versatile catalytic activity of as-synthesized active heterogeneous catalyst (MgZnAlO) in transesterification with methanol of various

3

vegetable oils namely: refined palm oil (RPO), waste cooking palm oil (WCPO), palm kernel oil (PKO) and coconut oil (CCO). The choice of feedstock in the investigation of MgZnAlO activity to produce biodiesel was borne out of natural abundance of the plant bearing seeds which thrive in tropical conditions and are mostly found in the regions of Malaysia and vicinities-Indonesia, Thailand, Vietnam and The Philippines (Mukta and Sreevalli, 2010; Murphy et al., 2012). These vegetable oils are extracted from the seeds of their plants, and have found applications in body oils, salves, lotions, soaps, hair tonics, shampoos and pesticides (Khayoon et al., 2012). Thus, this study is focused to establish suitable reaction conditions during transesterification of various oils in biodiesel production using a bifunctional MgZnAlO catalyst. Moreover, the use of these raw materials as a renewable feedstock in an upgraded industrial process would help to successfully develop a self-sustained biodiesel industry in the Southeast Asian countries.

2. Materials and methods 2.1 Materials

Raw materials, refined palm oil (RPO) from Penang, palm kernel oil (PKO) from IOI Oleochemicals Sdn. Bhd., Malaysia and coconut oil (CCO) were purchased from Yasree Store Sdn. Bhd., Selangor, Malaysia. Waste cooking palm oil (WCPO) was obtained from University of Science, Malaysia cafeteria. Analytical reagent grade 99.9% methanol (HPLC) purchased from Merck (Malaysia) was used for the transesterification reactions. The chemicals employed NH4OH (≥ 85%), Mg (NO3)2.6H2O (≥ 99%), Zn(NO3)2.6H2O (≥ 98%) and Al (NO3)3. 9 H2O ( 99%) were analytical grade from Sigma-Aldrich Pty Ltd., Malaysia and used for catalyst synthesis. Methyl heptadecanoate (99.5%), purchased from Sigma-Aldrich (Malaysia), was used as internal standard for gas chromatography (GC) analysis. The solvent for GC analysis, n-hexane (96%), was obtained from Merck (Malaysia). These reagents were used without further purification for catalyst synthesis and the transesterification of various vegetable oil.

2.2 Catalyst preparation

The procedure for the synthesis of MgZnAlO catalyst followed previously reported method (Olutoye and Hameed, 2011a) using co-precipitation from the nitrates of the metals with NH4OH solution. The catalyst was filtered, dried and thermally treated, thereafter employed in transesterification. The empirical formula of the various synthesized catalyst in a defined range for x and y is given as Mg1-xZn1+xAl (2-y)/3O3 i.e. 0.1≤ x ≤ 0.9 and y= 0. The catalyst was characterized and its activity reported on variety of vegetable oils such as refined palm oil (RPO), palm kernel oil (PKO), coconut oil (CCO), and waste cooking palm oil (WCPO).

4

2.3 Transesterification of vegetable oils with methanol using MgZnAlO heterogeneous catalyst

Transesterification of vegetable oils with methanol over the synthesized catalyst was carried out in a batch-type pressure reactor (PARR 4842) from Autoclave Engineers equipped with a mechanical stirrer and a PID. In a particular batch experiment with typical oil (CCO as reference), a ratio of 16:1 methanol to oil molar ratio equivalent to 71 mL oil and 49 mL methanol based on 120 mL total volume and catalyst loading of 3.32 wt% of oil were used. The reactor and its contents were continuously stirred at maximum rpm to avoid mass transfer limitations. At the completion of the reaction, the reactor contents were cooled to room temperature and were discharged, centrifuged at 3000 xg for 10 min for the separation of solid catalyst from reaction products. The FAME obtained was purified, dried and its content determined by gas chromatograph GC-2010plus (Shimadzu, Japan) equipped with a capillary column (Nukol 15 m x 0.53 mm x 0.5 µm); a split/splitless injection unit with FID detector according to the EN 14103 test methods (Munari et al., 2007). In all the experiments, the factors (temperature, time, methanol/oil molar ratio and catalyst loading) considered for the reaction were varied according to the experimental design using the response surface methodology (RSM) provided by Design-Expert software version 6.0.6 (Stat-Ease Inc., USA). The factors considered in the transesterification were chosen based on the design of experiment software as presented in Table 1. Prior to this, the range of values (low and high) had been established based on the preliminary experimental investigation and as reported in the literature (Sharma et al., 2008). 2.4 Catalyst reusability

The reusability of the MgZnAlO solid catalayst, which represents its capacity to perform the same catalytic activity was evaluated by conducting several experimental runs after the first batch at the optimum conditions obtained for the FAME yield. The catalyst after the first reaction was recovered and washed with n-hexane to remove any oil residue adhered to the surface of the catalyst. The washed catalyst was then filtered and dried for 12 h. The catalyst was then used for five consecutive runs to obtain FAME using CCO with all the experiments conducted in one stage process. The prolonged use of the catalyst to evaluate the significance and the feasibility of the catalyst for industrial application was studied. 3. Results and discussion 3.1 Characterization of the MgZnAlO catalyst

The commercial application of catalysts could be assessed based on its stability and activity in the process of choice, in this particular case, transesterification of different vegetable oils. With respect to this, heterogeneous catalysts are potentially more attractive due to their shape, selective behaviour, non-corrosiveness, easy recovery, reusability and environmentally

5

friendly nature. The textural properties which include the BET surface area, pore size and average pore diameter of the synthesized catalysts were determined. The results obtained are given in Table 2 where the surface area of all the synthesized catalyst samples are in the range of 52-61 m2/g which contributed to the observed catalytic performance during transesterification with various vegetable oils. In the particular case of Mg0.7Zn1.3Al0.67O3 where x = 0.3, the BET surface area of the catalyst as previously reported (Olutoye and Hameed, 2011a) was found to have the highest amongst all the other catalysts prepared. Thus, it was chosen for further investigation in the present study. The structural morphology as obtained from the SEM images (Supplementary Fig. 1) gave similar indication of well-ordered pattern of crystal arrangement. The mixed oxides catalyst is believed to form a synergetic network of composite heterogeneous oxides catalyst inter-locked together in one unit as revealed by the SEM images. Further insights into the catalyst performance, obtained from the XRD spectra (Supplementary Fig. 2), revealed the formation of some active components such as characteristic peaks corresponding to ZnO and MgO hexagonal structure, Al2O3 orthorhombic structure, and ZnAlO oxide monoclinic structure (Zn3Al94O144/94Al2O3.6ZnO). The conditions employed during the catalyst preparation at the calcined temperature (461 oC) and time (4.5 h) greatly reflects on its activity. In addition to the surface area of the catalyst, it was also observed that the catalyst possessed high total pore volumes with the value 0.156 cm3/g which are the highest in the series for all the formulations. The MgZnAlO catalyst exhibits bi-functional sites (basic and acidic) which is consequent of performance both in versatility and activity for the conversion to methyl esters of different types of oil. The basic and the acid sites were determined with values of 2190 μmole/ g cat. and 1230 μmole/g cat., respectively. The surface basicity could be seen to approximately double the surface acidity. The composite oxide formed with the metals using the preparatory method and heat treatment at 461 oC contributed to the synergetic oxide which helped in its stability by reducing the leaching of active sites into the reaction medium. 3.2 Effect of vegetable oil types on the transesterification reaction The various parameters affecting the transesterification reaction involving the vegetable oil types used for synthesis of methyl esters by MgZnAlO catalyst was investigated in the present study. The properties and the FAME profiles (%) of the different oils that are obtained by GC are given in Tables 3 and 4. The experiments were designed to determine how the yield and properties of methyl esters are affected by the various parameters such as temperature of reaction, catalyst loading and methanol/oil ratio. The results obtained in the current investigation revealed that the oil properties have contributory effect on the amount of methyl esters produced in addition to other conditions varied. The fatty acid chain composition of the triglyceride in the feedstock, such as the chain length or degree of unsaturation, has significant outcomes on the properties of FAME. Also, the free fatty acid (FFA) content and structure of oils have an influence on the biodiesel production. PKO and CCO have highest degree of saturated lauric acids in the present study with 48% and 47%, respectively. The degree of polyunsaturated fatty

6

acid chains (linoleic) in both oils gave 2% and may be less stable than the saturated types as a result of oxidation during storage. It was found that biodiesel obtained from both oils containing higher concentrations of high melting point saturated long fatty acid chains tends to have relatively poor cold flow properties. Similar observation was reported by Koria and Nithya, (2012) in the analysis of Datura stramonium methyl ester for its fatty composition using Gas chromatography assisted with mass spectrometry. Furthermore, impurities present in the feedstock could affect the quality of the biodiesel. In present study, the WCPO was pretreated to remove impurities before transesterification (Olutoye and Hameed, 2011b). The conversion levels into esters using refined vegetable oils can be higher than using crude vegetable oils under the same condition as obtained for refined CCO and RPO, methyl ester yield of 92% and 91% was obtained higher than crude PKO and WCPO with 81% and 85%, respectively at catalyst loading of 6 wt%, methanol/oil ratio 18:1, temperature of 170 oC during 6 h reaction time. In this study, the FAME obtained (%) of oil rich in oleic acid (18:1), such as CCO was higher than oils rich in linoleic acid (18:2), such as RPO and WCPO with 10% and 13.58% linoleic acid, respectively. Nevertheless, according to the results presented in Table 4, the FAME yields were not found significantly different among all of the oil types, and also between refined and crude oils. Generally, the conversion of oils into biodiesel through the transesterification process resulted approximately similar in comparison with other vegetable oils. Therefore, the readily availability and accessibility of waste cooking oil among other feedstock is a promising venture in Malaysia and vicinities for relatively cheap cost of biodiesel production. 3.3 Effect of catalyst loading and temperature on the transesterification reaction

The effectiveness and versatility of MgZnAlO catalyst and concentration toward

transesterfication of different vegetable oils was investigated. It can be observed that the catalyst performed well based on the FAME yields obtained from the oils. The activity of the present catalyst was compared with some earlier studies reported in literature where PKO and CCO were used as feedstock in transesterification. Though the catalyst type was different but the obtained result for a typical batch experiment proved the high activity and versatility of the catalyst. For example, 86% and 65% FAME from PKO were obtained using ZnO and ZrO2 catalysts, respectively at 200 oC, 50 bar pressure, catalyst loading 3 wt %, and methanol/oil ratio of 6:1in 4 h (Jitputti et al., 2006). In the current study, 94% FAME was obtained from PKO at 170 oC, catalyst loading 6 wt %, and methanol/oil ratio of 14:1in 6 h (no external pressure was applied except for the built up in the reactor due to methanol vapour recorded as autogeneous pressure, 20 bar). Also, 98% FAME was obtained from CCO using 3.32 wt% catalyst at 182 oC, and methanol/oil ratio of 16:1. Srilatha et al., (2010) obtained 92% FAME using 12-Tungstophosphoric acid (TPA, 5-30 wt %) supported on niobia (Nb2O5) in transesterification of used cooking oil at 200 oC, 34 wt%, 18:1 ratio methanol to oil, in 20 h reaction time (Srilatha et al., 2010). This is quite high operating conditions as compared to 89% FAME yield obtained in the present study at 170 oC, catalyst loading 6 wt %, and methanol/oil ratio of 14:1in 6 h. The

7

results showed that FAME yield for all oils depended on temperature, but it was not found significant differences among all of the oil types apparently due to FFA composition. Similarly, increase in catalyst concentration beyond a certain value will cause a decrease in the yield of methyl esters. This is in accordance with the results reported that the formation of metal soaps (emulsion) in the presence of a high amount of catalysts increases the viscosity of the reactants, thus resulting in a lower yield. High catalyst loading was required to neutralized the FFA content and complete the reaction during transesterification of some of the vegetable oils as shown in Table 4. The catalyst shows good tolerance with FFA content of the various oils because no soap formation was observed after the reaction. In the previous literature, many researchers proposed two-step method for transesterification using high FFA feedstock (Jain and Sharma, 2010; Patil and Deng, 2009) but in present study, high yield can be obtained in one-step process despite of the variation in oil properties. 3.4 Optimization of the reaction parameters for different vegetable oil using MgZnAlO catalyst

The optimization experiment was conducted to optimize the MgZnAlO catalyzed

transesterification process with different vegetable oils. The optimized conditions of the transesterification process for each vegetable oil were separately investigated using the design of experiment software. The influence of various process factors like temperature, catalyst concentration and amount of methanol were optimized at fixed reaction time of 6 h with the objective of producing high quality methyl esters with maximum yield. The predicted conditions using the software are presented in Table 5. It can be seen that CCO has the lowest catalyst loading at optimum value (3.33 wt%) while RPO and WCPO have their values at 9 wt%. The optimum temperature for all the oils except PKO is 182 oC with same methanol/oil ratio of 16:1 for all the oils. The observed variation of these parameters is due to the oil composition and quality of the feedstock. The predicted optimum FAME yield of 94% can be obtained for CCO and 85% for PKO. In order to verify this prediction, independent experiments were conducted and the results compared with the predicted values. The deviations obtained showed RPO has the lowest (-3.33%) and the highest -8.51% for CCO from predicted values. The investigation revealed that MgZnAlO is a promising catalyst for the production of biodiesel from different types of vegetable oil. 3.5 Characterization and properties of FAME obtained from vegetable oils

The properties of FAME obtained from the various oils used are summarized in Table 6. The properties presented are few among many required for biodiesel standards (the minimum requirements of ASTM 6751 and EN 14214 biodiesel standards). The conversion of oils to methyl esters (with fuel quality in the limit of minimum standard requirements) using the as synthesized catalyst demonstrates its capability in transesterification of feedstock with high %FFA and high moisture content as the case of PKO and CCO in one stage. This is similar to

8

earlier study reported in the literature where the catalyst was used in transesterification of crude jatropha oil with methanol (Olutoye and Hameed, 2011a). It would be observed that all the tested properties of the methyl esters synthesized in this work are within the range of acceptable values for biodiesel production. The standard specification for the density of biodiesel is in the range 860 and 900 kg m-3 (Balabin et al., 2011). The density plays an important role with regards to flow properties in a fuel injection system. Thus, densities of all samples tested were all within the limits. In a particular case of CCO, the density measured at 870 and the kinematic viscosity was 2.5-4.0 mm2s-1 at 40 oC. The viscosity of the fuel depicts its flow characteristics and also the tendency to deform under stress. Generally, vegetable oils undergo various treatments in order to reduce its viscosity. This is essential because viscosity of the fuel in internal combustion engines require some improvement on the fuel flow properties. It has been reported by (Murugesan et al., 2009) that the viscosity value above the specification limit can be attributed to the incomplete reaction or to the inefficient purification steps of the process, leaving glycerine in the ester phase. High value of viscosity will give rise to poor ignition, incomplete combustion, poor atomization and clogging as a result of carbon deposition. Thus, the viscosity is required to be low. The range of values specified for the viscosity of biodiesel in diesel engines is 3.5 and 5.0 mm2s-1 for good performance and the value for CCO and generally for other oils methyl ester determine in this work falls within range 2.5-6.0 mm2s-1. The samples with lower unconverted triglyceride contents present lower dynamic viscosity, density, and refractive index values. This behaviour could be due to the effect of molecular weight of the FAME produced, polarity, steric hindrance and intermolecular forces. The refractive index values are higher for unconverted oil than for the analyzed purified biodiesel samples with values of 1.47 and 1.40, respectively for RPO. Other oils and their ester products follow a similar trend as in the case of CCO and PKO. The presence of moisture in fuel will cause contamination which will result into engine corrosion; it may also cause a reversion of the produced FAME to fatty acids and lead to filter plugging. In this study, the purified methyl ester was oven dried for 30 min prior to GC analysis. This technique is suitable and requires no extra cost of chemical substance as drying agent. The water content of methyl esters for the various runs is found in the range of 0.01- 0.20 % which is below the specification limits. It was observed that all the esters produced recorded a high flash point. This will make it safer with regard to handling, transport, and storage. The flash point determined for the samples were in the range of 150-170 oC for WCPO, 140-165 oC for RPO and 150 oC for each of CCO and PKO. This value is higher than the minimum requirements of both the ASTM D6751 and EN-14214 biodiesel standards. Similarly, the acid value range of 0.442-0.459 mg of KOH/g was obtained for the esters which as can be seen to fall within the specified standards of EN 14214 (0.5 max) and ASTM D6751 (0.8 max). The pour points of the methyl esters were found to vary between -9 to -11 oC and -10 to -11 oC, respectively for RPO and WCPO. Higher pour point for biodiesel indicates that it tends to gel at high temperatures causing engine problems such as poor fuel atomization, incomplete combustion, and the depositing of carbon on the engine nozzles. The value obtained is characterized by high amounts of the saturated fatty acid alkyl esters present in the feedstock. It should be noted that the unsaturated fatty acid alkyl

9

esters have lower melting points than the saturated fatty acid alkyl esters. Therefore, the properties of methyl esters obtained are quite comparable to those of ASTM biodiesel standards and consistent with what has been reported in the literature.

3.6 Leaching, reusability and regeneration performance of the modified catalyst

The reused activity of the catalyst was checked and it showed very small decrease in

activity after first run. The catalyst was subjected to various experimental tests to check for its stability because deactivation due to coke formation and leaching were reported to be the main causes for reduced yield during the reuse of the catalysts (Molnar, 2011). The catalyst, recovered after each batch reaction by decanting using separation funnel was thoroughly washed with methanol until all traces of adhered oil and glycerol were removed. The slight reduction in the FAME content obtained over five cycles is a proof of the catalyst stability toward leaching with the acid sites retained throughout the cycles of reuse. Also, the observed drop in methyl ester (ME) content of approximately 3% and 5% during reusability from first cycle and after the fifth cycle is attributed to the handling loss during the process of manipulation. The loss will definitely imply reduction in active surface available for conversion of vegetable oil to FAME during subsequent reaction. The catalyst showed good recycling ability without appreciable loss in activity.

4. Conclusions

This study has established the versatility of a heterogeneous catalyst MgZnAlO in

transesterification of RPO, WCPO, PKO and CCO as raw materials. The effects of reaction parameters such as methanol to oil molar ratio, catalyst dosage, reaction temperature and types of oils on production of methyl esters by the solid catalyst were compared. The catalyst with Mg0.7Zn1.3Al0.67O3 where x = 0.3 produced the highest yield in transesterification of 98% with CCO. It was also observed that under optimum conditions, yields in the range of 80-87% could be obtained for the oils. Hence, the use of these raw materials as a renewable feedstock in an upgraded industrial process could help to successfully develop a self-sustained biodiesel industry especially in Malaysia an vicinities where the feedstock are abundantly available. It will also provide low-cost and environmental advantages. Acknowledgement

The authors acknowledge the research grant provided by the Universiti Sains Malaysia, under Research University (RU) grant (project no.: 814126) that resulted in this article.

10

References

[1] Atadashi, I., Aroua, M., Abdul Aziz, A., Sulaiman, N. 2012. The effects of water on biodiesel production and refining technologies: A review. Renewable and Sustainable Energy Reviews. 16(5), 3456-3470.

[2] Balabin, R.M., Lomakina, E.I., Safieva, R.Z. 2011. Neural network (ANN) approach to biodiesel analysis: Analysis of biodiesel density, kinematic viscosity, methanol and water contents using near infrared (NIR) spectroscopy. Fuel. 90(5), 2007-2015.

[3] Berchmans, H. J., Hirata, S., 2008. Biodiesel production from crude Jatropha curcas L. seed oil with a high content of free fatty acids. Bioresource Technology. 99(6), 1716-1721.

[4] Canakci, M., Van Gerpen, J., 2001. Biodiesel production from oils and fats with high free fatty acids. Transactions-American Society of Agricultural Engineers, 44(6), 1429-1436.

[5] Chouhan, A. P. S., Sarma, A. K., 2011. Modern heterogeneous catalysts for biodiesel production: A comprehensive review. Renewable and Sustainable Energy Reviews. 15(9), 4378-4399.

[6] Hoffmann, U., 2011. Some reflections on climate change, green growth illusions and development space. United Nations Conference on Trade and Development (UNCTAD), December. Available at: www.unctad.org/en/PublicationsLibrary/osgdp2011d5_en.pdf

[7] Jain, S., Sharma, M., 2010. Biodiesel production from Jatropha curcas oil. Renewable and Sustainable Energy Reviews. 14(9), 3140-3147.

[8] Jitputti, J., Kitiyanan, B., Rangsunvigit, P., Bunyakiat, K., Attanatho, L., Jenvanitpanjakul, P., 2006. Transesterification of crude palm kernel oil and crude coconut oil by different solid catalysts. Chemical Engineering Journal. 116(1), 61-66.

[9] Khayoon, M. S., Olutoye, M. A., Hameed, B. H., 2012. Utilization of crude karanj (Pongamia pinnata) oil as a potential feedstock for the synthesis of fatty acid methyl esters. Bioresource Technology. 111, 175-179.

[10] Koria, L., Nithya, G. 2012. Analysis of Datura stramonium Linn. biodiesel by gas chromatography-mass spectrometry (gc-ms) and influence of fatty acid composition on the fuel related characteristics. Journal of Phytology. 4(1). 6-9.

[11] Liu, X., Piao, X., Wang, Y., Zhu, S., 2008. Calcium ethoxide as a solid base catalyst for the transesterification of soybean oil to biodiesel. Energy & Fuels. 22(2), 1313-1317.

[12] Molnar, A., 2011. Nafion-silica nanocomposites: A new generation of water-tolerant solid acids of high efficiency- An update. Current Organic Chemistry. 15(23), 3928-3960.

[13] Mukta, N., Sreevalli, Y., 2010. Propagation techniques, evaluation and improvement of the biodiesel plant, Pongamia pinnata (L.) Pierre-A review. Industrial Crops and Products. 31(1), 1-12.

[14] Munari, F., Cavagnino, D., Cadoppi, A., 2007. Determination of Total FAME and Linolenic Acid Methyl Ester in Pure Biodiesel (B100) by GC in Compliance with EN14103. Thermo Fisher Scientific, Milan, Italy.

[15] Murphy, H. T., O’Connell, D. A., Seaton, G., Raison, R. J., Rodriguez, L. C., Braid, A. L., Kriticos, D. J., Jovanovic, T., Abadi, A., Betar, M., 2012. A Common View of the

11

Opportunities, Challenges, and Research Actions for Pongamia in Australia. BioEnergy Research. 1-23. DOI 10.1007/s12155-012-9190-6.

[16] Murugesan, A., Umarani, C., Subramanian, R., Nedunchezhian, N., 2009. Bio-diesel as an alternative fuel for diesel engines-A review. Renewable and Sustainable Energy Reviews. 13(3), 653-662.

[17] Olutoye, M. A., Hameed, B. H., 2011a. Synthesis of fatty acid methyl ester from crude jatropha (Jatropha curcas Linnaeus) oil using aluminium oxide modified Mg-Zn heterogeneous catalyst. Bioresource Technology. 102(11), 6392-6398.

[18] Olutoye, M. A., Hameed, B. H., 2011b. Synthesis of fatty acid methyl ester from used vegetable cooking oil by solid reusable Mg1-xZn1+xO2 catalyst. Bioresource Technology. 102(4), 3819-3826.

[19] Olutoye, M. A., Hameed, B. H., 2010. Transesterification of palm oil on KyMg1-xZn1+xO3 catalyst: Effect of Mg-Zn interaction. Fuel Processing Technology. 91(6), 653-659.

[20] Pal, K. D., Prakash, A., 2012. New cost-effective method for conversion of vegetable oil to biodiesel. Bioresource Technology. 121, 13-18.

[21] Patil, P. D., Deng, S., 2009. Optimization of biodiesel production from edible and non-edible vegetable oils. Fuel. 88(7), 1302-1306.

[22] Sharma, Y., Singh, B., Upadhyay, S. 2008. Advancements in development and characterization of biodiesel: A review. Fuel. 87(12), 2355-2373.

[23] Singh, S., Singh, D., 2010. Biodiesel production through the use of different sources and characterization of oils and their esters as the substitute of diesel: A review. Renewable and Sustainable Energy Reviews. 14(1), 200-216.

[24] Srilatha, K., Issariyakul, T., Lingaiah, N., Sai Prasad, P., Kozinski, J., Dalai, A., 2010. Efficient esterification and transesterification of used cooking oil using 12-Tungstophosphoric Acid (TPA)/Nb2O5 catalyst. Energy & Fuels. 24(9), 4748-4755.

[25] Walker, D. A., 2010. Biofuels-for better or worse? Annals of Applied Biology. 156(3), 319-327.

12

Table 1

Factors and corresponding levels for the response surface design of independent variables used in transesterification. Factor Factor

code

Unit Low (-1) Central (0) High (+1)

Catalyst loading A wt % of oil 1.5 6.0 10.5

Methanol/oil ratio B - 9:1 13.5:1 18:1

Temperature C oC 150 170 190

Reaction time is fixed at 6 h based on previous findings (Olutoye & Hameed, 2010)

13

Table 2

BET surface area, total pore volume and average pore diameter of MgZnAlO catalyst.

No. Catalyst Formulation BET surface

area (m2/g)

Total pore

volume

(cm3/g)

Average pore

diameter (nm)

1. Mg0.9Zn1.1Al0.67O3; x=0.1 55.85 0.090 10.06

2. Mg0.7Zn1.3Al0.67O3; x=0.3* 60.76 0.156 10.24

3. Mg0.5Zn1.5Al0.67O3; x=0.5 52.48 0.076 10.54

4. Mg0.3Zn1.7Al0.67O3; x=0.7 51.64 0.086 6.46

*(Olutoye and Hameed, 2011a)

14

Table 3

Some properties of the vegetable oils used before transesterification with methanol.

Property RPO WCPO PKO CCO Test method

Acid value, KOH, mg g−1 0.73 2.69 8.07 1.23 EN 14104

Water content, % 0.07 0.09 0.42 0.27 EN ISO 12937

Kinematic viscosity (40 °C), m2 s−1 65.42 65.22 45.72 41.17 EN ISO 3104

Density (15 °C), kg m−3 890 891 892 915 EN ISO 3675

Refractive index (30°C) 1.466 1.462 1.461 1.452 Atagorefractometer

model RX-5000α

Phosphorus (mg/kg) 0.549 0.909 0.703 0.4804 D 4951

15

Table 4 Design of experiment and FAME yields for the vegetable oils using MgZnAlO catalyst Expt. No

Factor 1: Catalyst loading (wt %)

Factor 2: Methanol/Oil

ratio (nil)

Factor 3: Temperature

(oC)

Response FAME yield (%)

RPO WCPO PKO CCO

1 8.68 11 158 79 80 83 86 2 6.00 14 170 85 84 76 86 3 6.00 14 170 84 89 71 87 4 6.00 18 170 91 85 81 92 5 3.32 11 182 83 80 71 87 6 6.00 14 150 82 81 63 80 7 3.32 16 158 85 85 82 78 8 6.00 9 170 86 80 72 83 9 6.00 14 170 85 82 80 89 10 8.68 16 182 90 88 80 82 11 1.50 14 170 83 83 84 84 12 8.68 16 158 80 88 70 85 13 3.32 16 182 86 82 75 98 14 3.32 11 158 81 80 67 81 15 8.68 11 182 84 82 71 92 16 6.00 14 170 85 83 94 88 17 10.50 14 170 86 83 72 94 18 6.00 14 190 85 90 72 88 19 6.00 14 170 84 84 72 85 20 6.00 14 170 85 83 81 87

16

Table 5

Process optimum conditions for the various vegetable oils

Parameters Type of vegetable oil

RPO WCPO PKO CCO

Catalyst loading (wt %) 9 9 8 3.33

Methanol/oil molar ratio 16:1 16.1 16:1 16:1

Reaction temperature (°C) 182 182 158 182

Predicted FAME content (%) at theoretical optimum

90 88 85 94

Experimental optimum FAME content (%) validation

87 85 80 86

Deviation from Predicted FAME content, % -3.33 -3.41 -5.88 -8.51

*All experimental conditions at fixed reaction time = 6 h

17

Table 6

FAME properties of different vegetable oils after transesterification with methanol

Property RPO WCPO CCO

methyl ester

PKO methyl ester

EN 14214/ASTM limits

Test methods

Acid value (mg

KOH g-1)

0.459 0.459 0.442 0.450 0.5 max EN

14104

Water content (%) 0.01-0.20 0.01-0.20 0.01-0.10 0.01-0.10 0.05 % max EN ISO

12937

Kinematic

viscosity@ 40 °C

(mm2 s-1)

2.7-3.9 2.7-6.0 2.5-4.0 2.5-5.0 3.50-5.00 EN ISO

3104

Cloud point (oC) -4 to -6 -4 to -8 -4 to-10 -4 to -10 EN

14214

Pour point (oC) -9 to -11 -10 to -11 -8 to -11 -8 to -11 ASTM

D 97

Flash point ( °C) 140-165 150-170 150 150 120 min EN

22719

Density@15°C

(kg m-3)

862 874 870 878 860-900 EN ISO

3675

Refractive index

(30°C)

1.40 1.44 1.45 1.44 Atago

refracto-

meter

model

RX-

5000α

Phosphorus

(mg/kg)

0.01 0.01 0.02 1.02 <4-10 ASTM

D4951

18

Supplementary Data

Supplementary Fig. 1. SEM image for Mg0.7Zn1.3Al0.67O3 where x = 0.3 calcined at temperature 461 oC for 4.5 h

19

Supplementary Fig. 2. XRD spectra for the as-synthesized catalyst (a) Aluminium modified Mg0.7Zn1.3Al0.67O3 where x = 0.3 calcined at temperature 461 oC for 4.5 h and (b) Unmodified MgZnO catalyst

10 30 50 70 90

Rel

ativ

e in

tens

ity (c

ount

s)

2-Theta (degree)

(a)

(b)

ZnO

MgZn

ZnAlO Al2O3 ZnAlO3

ZnO