Biodiesel Production From Greenseed Canola Oil†

4652r 2010 American Chemical Society pubs.acs.org/EF

Energy Fuels 2010, 24, 46524658 : DOI:10.1021/ef901202bPublished on Web 02/04/2010

Biodiesel Production from Greenseed Canola Oil

Titipong Issariyakul and Ajay K. Dalai*

Catalysis and Chemical Reaction Engineering Laboratories, Department of Chemical Engineering,University of Saskatchewan, Saskatoon, Saskatchewan S7N 5A9, Canada

Received October 21, 2009. Revised Manuscript Received January 22, 2010

Greenseed canola oil is low-grade oil with a green color. Because of the high level of chlorophyll, this oil isconsidered as a waste product and cannot be used for edible purposes. In this research, biodiesel wasproduced from canola oil and greenseed canola oil via KOH-catalyzed transesterification with methanol,ethanol, and a mixture of methanol and ethanol. The reaction was conducted at 60 C and a stirring speedof 600 rpm for 90 min. Prior to transesterification, greenseed canola oil was bleached to remove pigmentsusing various adsorbents at different conditions. The optimum bleaching material was found to bemontmorillonite K10. The pigment content was reduced from 94 to 0.5 ppm with using 7.5 wt % of thismaterial at 60 C and a stirring speed of 600 rpm for 30 min. Biodiesel derived from the treated greenseedcanola oil showed an improvement in oxidative stability (induction time of 0.7 h) as compared to thatderived from crude greenseed canola oil (induction time of 0.5 h). In addition, it was found that theamounts of unsaturated compounds as well as pigments contained in oil had an adverse effect on theoxidative stability of biodiesel.

Introduction

Biodiesel is an alternative fuel arising from concerns ofdepleting sources of fossil fuels and environmental issues.Biodiesel properties are comparable to those of fossil-baseddiesel fuel, and biodiesels can be produced from animal fats orvegetable oils; thus, they are renewable. Recently, there aremany concerns regarding the use of food crops as feedstockfor fuel production. Using crops for energy and food competewith each other in many ways (agricultural land, skilledlabors, water, fertilizers, etc.).1-3 Moreover, the high priceof biodiesel derived from food-grade vegetable oils makes itdifficult to compete economically with the fossil-based diesel.A less expensive, non food-grade vegetable oil is a potentialfeedstock for biodiesel production.The present work focuses on biodiesel production from

greenseed canola oil. Greenseed canola is an immature canolaseed. The green color is present in greenseed because of thehigh level of chlorophyll. In addition, chlorophyll is retainedin canola seeds if the seeds are exposed to frost during seeddevelopment. According to the Canadian Grain Commission(CGC), canola seeds are graded into three catagories. No. 1canola is the best quality canola seed, which contains lessthan 2% greenseed containing less than 25 ppm chlorophyllcontent and can be sold at C$ 250/ton. No. 2 and No. 3canolas are the lower grade seeds because they contain

26-45 and 46-100 ppm chlorophyll, respectively. The priceof No. 2 and No. 3 canolas are C$ 225/ton and C$ 190/ton,respectively.4,5 As the level of chlorophyll content increases,the selling price of the seed drops and the seed cannot be usedfor edible purposes. Therefore, greenseed canola oil can beconsidered as a nonfood-grade feedstock and can be used forbiodiesel production.Chlorophyll is an effective photoreceptor and can generally

be categorized into two types: chlorophyll A (contains-CH3as its functional group) and chlorophyll B (contains-CHOasits functional group). For plant growth, these two types ofchlorophylls absorb sunlight at slightly different wavelengths,thereby complimenting each other in photosynthesis.6 It isreported that chlorophyll gives an adverse effect on oilstability.5 In addition, chlorophyll can degrade into variouscompounds depending upon the surrounding conditions.7 Inthe presence of weak acids, magnesium ion is removed andchlorophyll degrades to pheophytin. Chlorophyllase can actas a catalyst for the removal of a phytol tail froma chlorophyllmolecule to form chlorophyllide. It is reported that chrolo-phyll derivatives could be converted to compounds that arecapable of being prooxidants, thus giving a deleterious effecton the stability of vegetable oils.8 Ward et al.7 reported thatthe major pigments contained in canola and greenseed canolaare chlorophylls and pheophytins. To remove these pigments,

This paper has been designated for the Bioenergy andGreen Engineeringspecial section.

*To whom correspondence should be addressed: Department ofChemical Engineering, 57 Campus Drive, University of Saskatchewan,Saskatoon, Saskatchewan S7N 5A9, Canada. Telephone: 1-306-966-4771. Fax: 1-306-966-4777. E-mail: [email protected].(1) Torrey, M. Inform 2007, 18 (5), 302306.(2) Monbiot, G. The western appetite for biofuels is causing starva-

tion in the poor world. Guardian website, www.guardian.co.uk/com-mentisfree/2007/nov/06/comment.biofuels (accessed on April 20, 2008).(3) Thailand worries over food shortages amid palm oil debate.

EcoEarth.Info website, www.ecoearth.info/shared/reader/welcome.aspx?linkid=92992 (accessed on April 3, 2008).

(4) Government of Saskatchewan. Frost and greenseed in canola.Government of Saskatchewan website, http://www.agriculture.gov.sk.ca/Default.aspx?DN=bb79745e-e78a-45af-80e7-79e9b7903a2e(accessed on Jan 22, 2008).(5) Kulkarni, M. G.; Dalai, A. K.; Bakshi, N. N. J. Chem. Technol.

Biotechnol. 2006, 81, 18861893.(6) May, P. Chlorophyll. University of Bristol website, http://www.

chm.bris.ac.uk/motm/chlorophyll/chlorophyll_h.htm (accessed on Oct20, 2007).(7) Ward, K.; Scarth, R.; Daun, J. K.; Thorsteinson, C. T. J. Am. Oil

Chem. Soc. 1994, 71 (12), 13271331.(8) Tautorus, C. L.; Low, N. H. J. Am. Oil Chem. Soc. 1994, 71 (10),

11231128.

4653

Energy Fuels 2010, 24, 46524658 : DOI:10.1021/ef901202b Issariyakul and Dalai

various kinds of clay and activated carbon can be used toadsorb chlorophylls and pheophytins.5,9,10

The alcohols commonly used in transesterification areshort-chain alcohols, i.e., methyl, ethyl, propyl, and butylalcohol. It is reported that the use of a mixture of alcohols hascertain advantages. When the mixture of methanol andethanol was used for transesterification, the alcohol-oilsolubility was improved by ethanol and the reaction equilib-riumwas improved bymethanol.11Although there are severalworks in the literature focused on chlorophyll adsorption andbiodiesel production from canola oil, information on bothtechnologies combined is scarce. The objectives of this workare to produce biodiesel from greenseed canola oil and tostudy the effects of pigments on transesterification and esterproperties.

Experimental Section

Materials.Greenseed canola oil was provided byMilliganBio-Tech, Inc., Foam Lake, Saskatchewan, Canada. Commercial-grade canola oil was purchased from a local grocery store.Montmorillonite K10 and KSF clay (adsorbent) were obtainedfromAlfa Aesar,WardHill, MA. Activated carbons (adsorbent)were prepared in our laboratory from biochar obtainedfrom Dynamotive Energy Systems Corp., Vancouver, BritishColumbia,Canada,EnsynCorp.,Wilmington,DE,andAdvancedBiorefinery, Inc., Ontario, Canada, and from char obtained fromLuscar Ltd., Alberta, Canada. Anhydrous methanol (MeOH)(99.8%) and potassium hydroxide (KOH) were purchasedfrom EMD Chemicals, Inc., Darmstadt, Germany. Anhydrousethanol (EtOH) was obtained from Commercial Alcohol, Inc.,Brampton, Ontario, Canada. Reference standard chemicals, in-cluding methyl oleate, triolein, diolein, and monoolein, werepurchased from Sigma-Aldrich, St. Louis, MO. The fatty acidmethyl ester (FAME) mix rapeseed oil reference standard wasobtained from Supelco, Bellefonte, PA.Procedures. Initially, the oil feedstock was analyzed for

pigment content and acid value. The experimental procedurefor this research includes three steps: bleaching of greenseedcanola oil, transesterification, and ester characterization.For bleaching of greenseed canola oil, greenseed canola oil

was treated with eight types of adsorbents. These adsorbentswere composed of three types of clay (montmorillonite K10,montmorillonite KSF, Attapulgus clay) and five types of acti-vated carbons (ACs) [obtained from Dynamotive Energy Sys-tems Corp., Luscar Ltd. (granular and powder), Ensyn Corp.,and Advanced Biorefinery, Inc.]. In the treatment process,100 g of greenseed canola oil was placed in a Parr reactor

(Parr Instrument Company, Moline, IL) and the oil was heatedslowly. When the oil temperature reached 60 C, the adsorbentwas added to the reactor and the stirring speed was keptconstant at 600 rpm. The parameters studied are type ofadsorbent (eight kinds asmentioned above), treatment duration(0.5, 1, 1.5, and 2 h), and adsorbent loading (1, 2.5, 5, 7.5, and10 wt % loading). A UV-260 Shimadzu spectrometer was usedto determine pigment content. The calculation method wasdescribed by Lichtenthaler.12 An attempt has been made toregenerate montmorillonite K10 using a solvent extractiontechnique.13 The solvents used in this step include methanol(MeOH), hexane, tetrahydrofuran (THF), and chloroform.Initially, the used clay was mixed with a solvent at a 30:70clay/solvent weight ratio. The mixture was then stirred at 45 Cand 600 rpm for 30 min. The liquid portion was separated fromthe solid portion by filtration. These steps were repeated twiceprior to drying the clay at 100 C for 24 h.For the transesterification step, a series of methyl ester and

ethyl ester were produced by means of KOH-catalyzed trans-esterification from a 100 g feedstock, which is canola oil, crudegreenseed canola oil, treated greenseed canola oil, and amixtureof canola oil and treated greenseed canola oil (50 g of canola oiland 50 g of treated greenseed canola oil). Esters from themixture of methanol and ethanol were also prepared from thesame set of feedstocks. A 1%KOHbased on the total amount ofoil was used in each case as a catalyst. In the case of using onlyone alcohol (methanol or ethanol), the feedstock was initiallyplaced in a Parr reactor and heated to 60 C. Alcohol(6:1 alcohol/oil molar ratio) and KOH (1 wt % with respect tooil) were then added to the reactor. In the case of ester prepara-tion from themixture of alcohols (methanol ethanol), 3mol ofmethanol and 3 mol of ethanol were used for each mole of oil toset a 6:1 total alcohol/oil molar ratio. The temperature and thestirring speed of the reaction mixture were maintained constantfor 1.5 h at 60 C and 600 rpm, respectively.After the reaction, the transesterification product was allowed

to settle in a separating funnel for glycerol separation. Because ofthe strong emulsion in the case of ethanolysis products, glycerolcould not be separated only by gravity. To separate glycerol fromethyl ester, approximately 10 g of pure glycerol was added to thetransesterification product, the separatory funnel was shakenvigorously, and the product was allowed to settle. The glycerollayer was then separated from the ester layer within 1 h. Distilledwaterwas heated and used in thewashing step.A strong emulsionwas usually formed in the case of ethyl ester preparation. Toavoid the formation of emulsion, tannic acid solution (0.1 wt%)was used in the washing step, thereby neutralizing the excess basecatalyst. Unreacted methanol and water were removed usinga Buchi rotavapor. Biodiesel was finally passed through theanhydrous sodium sulfate, which was previously dried in an ovenat 100 C for 1 h, to remove traces ofmoisture. All of the biodieselsamples produced in this step are shown in Table 1.

Table 1. Biodiesel Samples

sample alcohol used feedstock

CGME methanol crude greenseed canola oilTGME methanol treated greenseed canola oilCME methanol canola oilTGCME methanol treated greenseed canola oil canola oil (50:50)CGEE ethanol crude greenseed canola oilTGEE ethanol treated greenseed canola oilCEE ethanol canola oilTGCEE ethanol treated greenseed canola oil canola oil (50:50)CGMEE MeOH EtOH (50:50) crude greenseed canola oilTGMEE MeOH EtOH (50:50) treated greenseed canola oilCMEE MeOH EtOH (50:50) canola oilTGCMEE MeOH EtOH (50:50) treated greenseed canola oil canola oil (50:50)

(9) Davies, M. E.; Whittle, M. E.; Jones, W.; Mokaya, R. Interna-tional Patent WO 92/19533, April 13, 1992.(10) Mokaya, R.; Jones, W.; Davies, M. E.; Whittle, M. E. J. Mater.

Chem. 1993, 3 (4), 381387.(11) Issariyakul, T.; Kulkarni, M. G.; Dalai, A. K.; Bakhshi, N. N.

Fuel Process. Technol. 2007, 88, 429436.

(12) Lichtenthaler, H.Methods Enzymol. 1987, 148, 350382.(13) Chanrai, N. G.; Burde, S. G. U.S. Patent 20030201228, Oct 30,

2003.

4654


For the analysis part, the ester phase collected from eachexperiment was analyzed for ester and glyceride content using aHewlett-Packard 1100 series [high-performance liquid chromato-graphy (HPLC)] with a refractive index detector and twoPhenogel 5u 100A, 300 7.80 mm, 5 m columns in seriesprotected with a guard column, equipped with ChemStation forLC 3D, Agilent Technologies. THF was used as a mobile phaseat 1 mL/min for 25 min. The operating parameters used were asfollows: injection volume, 5 L; column temperature, 24 C; anddetector temperature, 35 C. Reference standard chemicalsincluding methyl oleate, triolein, diolein, and monoolein wereused for the HPLC calibration. Fatty acid compositions ofesters were determined using an Agilent Technologies 6890NNetwork GC System equipped with GC ChemStation softwarewith a flame ionization detector (FID) and RESTEK 10638Stabilwax column. The injection volume was 2 L, and thetemperature program was started at 160 C, held for 1 min,ramped to 240 C at 4 C/min, and then held for 24min. SupelcoFAMEmix rapeseed oil standard was used as a reference for gaschromatography (GC) calibration. The oxidative stability ofbiodiesel was measured as the induction time using a Metrohm743 Rancimat instrument. A Brookfield DV-I viscometer wasused to measure the viscosity of esters. The sulfur content ofbiodiesel was determined using anAntekN/S analyzer equippedwith an Antek model 9000NS combustion analyzer, Antekmodel 735 controlled rate sample drive, Antek model 740multimatrix sample inlet, and Antek model 738 robotic auto-sampler. The acid and iodine values were determined as per themethods American Oil Chemists Society (AOCS) Te 1a-64 andAOCS Tg 1a-64, respectively.

Results and Discussion

The objective of bleaching of greenseed canola oil is toremove pigments from the oil. The pigments present in the oilare chlorophyll A (ChA), chlorophyll B (ChB), pheophytin A(PhA), and pheophytin B (PhB). Table 2 shows the initialpigment content and acid value of canola and greenseed

canola oils. Canola oil has negligible amounts of pigmentsand acid value, while greenseed canola oil has a total initialpigment content and an acid value of 94 ppm and 3.8 mg ofKOH/g, respectively. To remove these pigments, greenseedcanola oil was bleached according to the process discussed inthe Experimental Section.Table 3 shows bleaching performances of various adsor-

bents selected in this study. Greenseed canola oil was treatedat 60 Cusing 1wt%adsorbent loading anda stirring speedof600 rpm for 30 min. The reproducibility of this experiment iswithin (3 ppm of total pigment content. Different materialshave different pigment adsorption capability. It was foundthat montmorillonite K10 is the most suitable adsorbent forthe bleaching of greenseed canola oil with 29% pigmentadsorbed. This is due to the relatively high surface area ofK10 as compared to KSF (see Table 4). Despite the highsurface area of ACs, these materials were not suitable for theremoval of pigments from greenseed canola oil because of thenarrow pore width of ACs.The effect of the treatment duration on greenseed oil is

shown in Table 5. It was found that the treatment durationof 30 min is sufficient for the bleaching of greenseed canola oilusingmontmorilloniteK10. The increase in treatment durationshowsno significant improvement in thebleachingof greenseedcanola oil. The effect of percent adsorbent loading on thepigment removal from greenseed oil is shown in Table 6. Itwas found that 7.5 wt % montmorillonite K10 loading isrequired for the maximum pigment adsorption from greenseedcanola oil. The optimum conditions for the bleaching of green-seed canola oil is the use of montmorillonite K10 at 7.5 wt %loading, 60 C, a stirring speedof 600 rpm,anda treatment timeof 30 min. After the treatment process, the pigment content ofgreenseed canola oil was reduced from 94.1 to 0.5 ppm. Aminor reduction on the acid value of greenseed canola oil wasobserved (from 3.8 to 3.0; see Table 2).

Table 2. Initial Pigment Content and Acid Value in Canola and Greenseed Canola Oilsa

feedstock ChA (ppm) ChB (ppm) PhA (ppm) PhB (ppm) total (ppm) acid value (mg of KOH/g)

canola oil 0.0 0.1 0.0 0.0 0.1 0crude greenseed canola oil 26.0 2.7 56.6 8.8 94.1 3.8treated greenseed canola oil 0.2 0.0 0.2 0.1 0.5 3.0

aChA, chlorophyll A; ChB, chlorophyll B; PhA, pheophytin A; PhB, pheophytin B.

Table 3. Physical Properties and Performance of Various Adsorbents for Pigment Adsorptiona

adsorbent ChA (ppm) ChB (ppm) PhA (ppm) PhB (ppm) total (ppm)percentadsorbed

BET surfaceareab (m2/g)

average porewidthb (Ao)

crude greenseed canola oil 26.0 2.7 56.6 8.8 94.1 0 n/a n/amontmorillonite K10 18.2 2.2 39.1 7.3 66.7 29.1 250.1 57.8montmorillonite KSF 25.8 3.0 56.5 9.1 94.3 0 1.5 57.4Attapulgus clay 25.4 3.0 55.8 8.8 93.1 1.1 n/a n/aDynamotive Energy Systems AC 21.7 3.1 48.8 6.6 80.1 14.9 454.0 23.0Luscar AC (powder) 18.9 3.3 43.4 5.5 71.2 24.4 312.9 17.1Luscar AC (granular) 24.2 2.8 53.4 8.1 88.6 5.9 n/a n/aEnsyn AC 19.9 12.3 41.6 14.7 88.4 6.1 524.6 21.9Advanced Biorefinery AC 19.0 3.5 43.1 6.2 71.9 23.6 n/a n/a

aTreatment conditions: stirring speed, 600 rpm; treatment temperature, 60 C; treatment duration, 30 min; adsorbent loading, 1% (w/w). AC =activated carbon. bAnalysis was conducted on the bleaching materials.

Table 4. Properties and Performance of Regenerated Montmorillonite K10 for Pigment Adsorption

sample cholophyll content (mol/g) BET surface area (m2/g) average pore diameter (Ao) percent adsorbed

MeOH-treated K10 59.4 0.8 1774.9 13hexane-treated K10 144.4 24.2 133.5 31THF-treated K10 133.8 51.3 93.9 53chloroform-treated K10 82.3 0.2 14.4 12

4655


An attempt was made to regenerate montmorillonite K10(K10). The spent K10 was regenerated on the basis of themethod discussed in the Experimental Section. The regener-ated K10 was analyzed for the Brunauer-Emmett-Teller(BET) surface area and average pore width using BETanalysis and chlorophyll content using inductively coupledplasma (ICP) analysis on Mg content (using the concept ofthat each mole of chlorophyll contains 1 mol of Mg). Theregenerated K10 was then reused to remove pigments fromgreenseed oil using 7.5 wt % loading, a stirring speed of600 rpm, and a treatment temperature of 60 C for 30 min.The performance of regenerated K10 is shown in Table 4.Although greenseed canola oil was removed from the spentK10, chlorophyll was not entirely removed from this bleach-ing material. The outcome was a significant drop in pigmentadsorption capability of these regenerated K10 from the levelthat fresh K10 can provide. Because of its relatively high BETsurface area, the THF-treated K10 exhibited the highestpigment adsorption performance (53%) among all of theregenerated K10.For biodiesel production, all vegetable oils were transester-

ified using the method discussed in the Experimental Section.Figure 1 shows the ester formation during the transesterifica-tion reaction at 50 and 60 C. It is clear that the reaction at60 C is faster than that performed at 50 C and the reactionduration of 90 min is sufficient to complete the reaction. Theerror bars shown in Figure 1 indicate that the reproducibilityof this experiment iswithin 1%.Table 7 shows the percentagesof triglyceride, diglyceride,monoglyceride, and ester aswell asthe percent yield of each ester. Triglyceride, diglyceride,

monoglyceride, and ester percentages were determined usingHPLC, and the percent yield was defined as the ratio of theamounts of the ester phase recovered to the amounts of thefeedstock multiplied by 100.Methyl esters are obtained at a higher ester percentage

compared to ethyl esters (see Table 7). This is because of therelative higher reactivity of themethoxide ion compared to theethoxide ion, leading to a higher amount of methyl esterformation.11,14 The same trend was found in the previouswork.15 The percent yield did not reach 100%mainly becauseof the loss of oil during the washing step. This loss was higherduring the productionof ethyl esters because the emulsionwasstrongly formed in these cases. A comparable ester percentagewas observed in the case of methyl ester and mixed methyl-ethyl ester. This finding indicates that mixed methyl-ethylalcohol is suitable for the production of biodiesel from canolaoil, greenseed canola oil, and themixture of both oils. The lowester percentage in the case of ethanolysis can be improvedby adjusting reaction conditions or using the purificationmethod described in the previous work.15 There was nodistinct difference between the crude greenseed canola oilester percentage and the treated greenseed canola oil esterpercentage. This finding implies that pigments did not play asignificant role in the transesterification reaction.Table 8 shows fatty acid compositions of canola oil methyl

ester (CME), crude greenseed canola oil methyl ester

Table 6. Performance of Montmorillonite K10 at Various Percent Loading for Pigment Adsorptiona

percent adsorbent loading ChA (ppm) ChB (ppm) PhA (ppm) PhB (ppm) total (ppm) percent adsorbed

0 26.0 2.7 56.6 8.8 94.1 01 18.2 2.2 39.1 7.3 66.7 29.12.5 7.7 1.5 14.8 5.1 29.1 69.15 2.0 0.7 3.2 2.1 8.0 91.57.5 0.2 0.0 0.2 0.1 0.5 99.510 0.2 0.0 0.2 0.1 0.5 99.5

aTreatment conditions: stirring speed, 600 rpm; treatment temperature, 60 C; treatment duration, 30 min.

Table 7. Triglyceride, Diglyceride, Monoglyceride, and Ester Per-centages and Percent Yield of Esters (w/w)

biodieseltriglyceride

(%)diglyceride

(%)monoglyceride

(%)ester(%)

yield(%)

CGME 1.3 2.4 1.5 94.8 82.6TGME 0.9 2.2 1.3 95.7 84.7CME 0.0 1.7 1.0 97.3 90.0TGCME 0.0 2.0 1.2 96.8 86.0CGEE 11.7 10.7 7.6 70.0 63.8TGEE 7.0 9.3 9.3 74.4 65.2CEE 0.4 2.4 4.1 93.1 82.8TGCEE 2.8 6.1 7.6 83.4 68.8CGMEE 0.0 2.5 2.4 95.1 79.2TGMEE 0.0 2.2 2.2 95.6 74.0CMEE 0.0 2.2 2.2 95.7 81.4TGCMEE 0.0 2.3 2.5 95.3 83.5

Table 5. Performance of Montmorillonite K10 at Various Bleaching Durations for Pigment Adsorptiona

treatment duration (h) ChA (ppm) ChB (ppm) PhA (ppm) PhB (ppm) total (ppm) percent adsorbed

0 26.0 2.7 56.6 8.8 94.1 00.5 18.2 2.2 39.1 7.3 66.7 29.11 18.2 3.8 39.1 9.1 69.5 26.11.5 17.2 2.2 38.4 4.2 66.9 28.92 17.7 2.5 43.3 7.8 65.7 30.1

aTreatment conditions: stirring speed, 600 rpm; treatment temperature, 60 C; adsorbent loading, 1% (w/w).

Figure 1. Ester formation during transesterification of canola oilusing methanol at 50 and 60 C.

(14) Sridharan, R.; Mathai, I. M. J. Sci. Ind. Res. 1974, 33, 178186.(15) Issariyakul, T.; Kulkarni, M. G.; Meher, L. C.; Dalai, A. K.;

Bakhshi, N. N. Chem. Eng. J. 2008, 140, 7785.

4656


(CGME), treated greenseed canola oil methyl ester (TGME),canola oil ethyl ester (CEE), and treated greenseed canola oilmethyl ethyl ester (TGMEE). Oleic acid was found to be thedominant fatty acid in all esters. The results in this tableare comparable to those reported in the previous work.15

Greenseed oil contains higher unsaturated compounds com-pared to canola oil. The fatty acid compositions ofCGMEaresimilar to those of TGME. This finding indicates that thetreatment process did not alter fatty acid compositions ofgreenseed canola oil. In addition, fatty acid compositions ofCME are comparable to those of CEE. This finding suggeststhat fatty acid compositionof ester prepared from the sameoilremains the same regardless of the type of alcohol used intransesterification. A clear fatty acid composition of biodieselprepared from mixed alcohols was not obtained because ofpeak overlapping between methyl and ethyl esters of variousfatty acids in theGC chromatogram.When the treated green-seed canola oil was transesterified with the mixed methanol/ethanol, methyl esters were formed in higher amounts com-pared to ethyl esters. For example, the amounts of methyloleate and ethyl oleate formed during transesterification were36.20 and 19.95%, respectively. This result confirms theconcept of higher reactivity of methanol toward trans-esterification as compared to ethanol.The additional biodiesel properties, such as acid value,

iodine value, viscosity at 40 C, and sulfur content, arepresented in Table 9. The acid value of CME was 0.2, whichmeets the American Society for Testing and Materials(ASTM) specification (acid value < 0.5). The acid values ofCGME and TGME were both 0.4, which reflect the higher

initial acid value of greenseed canola oil (acid value =3.0-3.8) as compared to that of canola oil (acid value = 0).The acid value of methyl ester prepared from the mixture ofboth oils fell in between that of canola oil methyl ester andgreenseed oil methyl ester. The higher acid values of estersprepared with mixed alcohol and especially ethanol wereprobably because of the tannic acid solution that was usedin the washing step, instead of pure distilled water, as a resultof the formation of strong emulsion in these cases. The iodinevalue of canola oil methyl ester (iodine value = 109.5) waslower than those of greenseedoilmethyl esters (iodine value=111.0 and 111.6). This is because canola oil has less unsatu-rated compounds compared to greenseed canola oil (seeTable 8). The equivalent iodine value of CGME and TGMEis due to their similar fatty acid compositions, as shown inTable 8. The iodine value of methyl ester prepared from themixture of both oils fell in between that of CME and TGME.The lower iodine values of ethyl esters compared to methylesters suggested that ethyl esters have a lower degreeof unsaturation compared to methyl esters. This can beexplained using the concept of the molar concentration ofdouble bonds, as described by Knothe and Dunn.16 If methyland ethyl esters have the same number of double bonds permolecule, ethyl ester, which has a higher molecular weight,would have a lower molar concentration of double bonds,leading to the lower degree of unsaturation. The iodine valuesof esters prepared with mixed methanol-ethanol fell inbetween those prepared with methanol and ethanol, asexpected. Viscosities of esters prepared from methanol andmixed methyl-ethyl alcohol are in the range of 4.8-5.2 cSt,which meet the ASTM specification (viscosity between 1.9and 6.0 cSt). The high viscosities of ethyl esters were due tolower triglyceride conversion. These esters contain higheramounts of glycerides; therefore, the higher viscosities wereobserved. Sulfur contents of all esters were less than 1 ppm,which meet the ASTM specification (sulfur < 15 ppm).To determine the oxidation stability of biodiesel, aRancimat

instrument was used. During the Rancimat test, the samplewas heated to 110 C and the oxygen was supplied. In thepresence of oxygen at high temperatures, the oxidation reac-tion took place and the oxidation derivatives were transferredto the measuring chamber containing Millipore water. Theincrease in conductivity of the water was detected as the

Table 8. Fatty Acid Compositions of Selected Esters

structure compound name CME (%ME) CGME (%ME) TGME (%ME) CEE (%EE) TGMEE (%ME/%EE)

C14:0 myristic ester 0.06 0.07 0.06 0.05 traceC16:0 palmitic ester 4.36 4.60 4.54 4.45 3.08/naa

C16:1 palmitoleic ester 0.16 0.26 0.24 0.26 na/0.10C16:2 hexadecadienoic ester 0.08 0.08 0.08 0.07 traceC16:3 hexadecatrienoic ester 0.09 0.15 0.12 0.11 0.07/traceC18:0 stearic ester 1.96 2.01 1.94 1.95 1.29/naa

C18:1 z9 oleic ester 60.92 55.51 55.05 61.09 36.20/19.95C18:1 z11 vaccenic ester 2.89 3.59 3.47 2.98 na/1.27C18:2 linoleic ester 18.70 20.93 21.16 18.82 13.45/7.60C18:3 linolenic ester 6.79 9.41 9.76 6.85 6.06/3.36C20:0 arachidic ester 0.59 0.66 0.73 trace 0.42/0.23C20:1 eicosenoic ester 1.12 1.34 1.41 1.17 0.88/0.49C22:0 behenic ester 0.22 0.41 0.43 trace 0.21/tracetotal saturated fatty acid 7.19 7.75 7.7 6.45 naa

total monounsaturated fatty acid 65.09 60.7 60.17 65.50 naa

total polyunsaturated fatty acid 25.66 30.57 31.12 25.85 naa

ana = not available because of peak overlap.

Table 9. Acid Value, Iodine Value, Viscosity at 40 C, and SulfurContent of Esters

biodieselacid value

(mg of KOH/g)iodine value

(mg of I2/100 g)viscosity

at 40 C (cSt)

sulfurcontent(ppm)

CGME 0.4 111.6 5.1

4657


oxidation derivatives were transferred into the water. Theinduction time is defined as the time required for the con-ductivity of the water to be increased rapidly and was used asan indication of biodiesel oxidation stability.A typical Rancimat plot is shown in Figure 2, and the

Rancimat results are presented inFigure 3.Table 10 shows thepigment content for all esters. CME exhibited higher stabilitythan those of CGME and TGME. It can be because ofthe higher degree of unsaturation (polyunsaturated com-pounds = 31.12%; iodine value = 111.0) of greenseed oilmethyl ester (GME) as compared to CME (polyunsaturatedcompounds = 25.66%; iodine value = 109.5). It is reportedthat the oil stability increaseswith the decrease in the degree ofunsaturation.17 TGME showed a slightly higher inductiontime as compared to CGME. CGME and TGME are similar

in both fatty acid compositions and iodine value, but thedifference in pigment content was significant (34 ppm forCGME and 1 ppm for TGME; see Table 10). This resultindicates that pigments have adverse effects on biodieselstability. This finding fits well with that reported in theliterature.5,18 It is also observed that the pigment contentwas reduced during transesterification (from 94 to 34 ppm;see Tables 2 and 10). This result is anticipated because it isreported that chlorophyll can be removed during alkali-catalyzed transesterification in the form of water-solublechlorophyllin salt.5 CEE showed a longer induction timecompared to CME. It can be explained by the fact that ethylester has a lower molar concentration of double bonds and,therefore, was more stable than methyl ester. The interpreta-tion of the induction time of esters prepared with mixedmethanol-ethanol requires complete fatty acid composi-tional analysis of these esters. Most of the esters exhibit lowoxidation stability and did not meet the ASTM specification(3 h of induction time). Schober and Mittelbach19 reportedthat the induction times of rapeseed oil methyl ester anddistilled rapeseed oil methyl ester without an addition ofantioxidant were 4.56 and 2.03 h, respectively. This showsthat FAME prepared from the rapeseed family does notpossess high oxidative stability property and an addition ofantioxidant to FAME is required.An attempt was made to combine (a) the bleaching of

greenseed canola oil with (b) the transesterification into asingle step. Crude greenseed canola oil was transesterifiedwith methanol as per the method described in the Experi-mental Section. In addition, 7.5 g ofmontmorilloniteK10wasadded to the reactor at the beginning of the reaction. The esterformed during the reaction and pigment content at the end ofthe reaction are shown inFigure 4. The results suggest that thecombination of bleaching of greenseed canola oil and trans-esterification led to a lower transesterification activity as wellas an impairment of the sorption of pigments. This can beexplained by the sorption phenomenon of potassium inmontmorillonite as described by Muravyov and Sakharov.20

Montmorillonite, by nature, has a high negative potential andsuitable size of interlayer spacing for the sorption of potas-sium. If potassium is trapped in the interlayer of mont-morillonite, less catalyst would be available for transesteri-fication, resulting in a tremendous drop in ester percentage

Figure 2. Rancimat plot of CME.

Figure 3. Induction time of biodiesel prepared from canola andgreenseed canola oils.

Table 10. Pigment Content of Estersa

biodieselChA(ppm)

ChB(ppm)

PhA(ppm)

PhB(ppm)

total(ppm)

CGME 9.1 2.3 16.8 5.9 34.0TGME 0.2 0.2 0.3 0.3 1.0CME 0.2 0.4 0.3 0.4 1.3TGCME 0.1 0.1 0.1 0.1 0.4CGEE 14.4 1.5 29.2 6.0 51.1TGEE 0.2 0.3 0.3 0.5 1.4CEE 0.3 0.6 0.5 0.6 2.0TGCEE 0.2 0.3 0.3 0.4 1.3CGMEE 12.5 1.0 24.7 5.3 43.5TGMEE 0.4 0.7 0.7 0.8 2.7CMEE 0.8 1.3 1.3 1.4 4.7TGCMEE 0.1 0.2 0.1 0.2 0.5

aChA, chlorophyll A; ChB, chlorophyll B; PhA, pheophytin A; PhB,pheophytin B.

Figure 4. Ester formation during transesterification of greenseedcanola oil and pigment content of ester at the end of the reaction.

(17) Neff, W. E.; Ei-Agaimy, M. A.; Mounts, T. L. J. Am. Oil Chem.Soc. 1994, 71 (10), 11111116.

(18) Tautorus, C. L.; Low, N. H. J. Am. Oil Chem. Soc. 1993, 70 (9),843847.(19) Schober, S.; Mittelbach, M. Eur. J. Lipid Sci. Technol. 2004, 106,

382389.(20) Muravyov, V. I.; Sakharov, B. A. Sedimentology 1970, 15, 103

113.

4658


(see Figure 4). In addition, when the potassium cation entersthe region between montmorillonite layers, it will polarize theadjacent layer, which prevents the sorption of other cations.

Conclusions

It was found that pigments have an adverse effect in oilstability but have no effect on the transesterification reaction.These pigments can be removed by the bleaching process. Theoptimum conditions in the bleaching process are the use of7.5wt%montmorilloniteK10 at 60 Cand a stirring speed of600 rpm for 30 min. After the bleaching process, the pigmentcontent of greenseed canola oil was reduced from 94 to0.5 ppm. This bleaching process did not alter fatty acidcompositions of the oil. Montmorillonite K10 is readily

obtainable at low cost, and the chlorophylls could not beentirely removed from the spent K10 during the regenerationprocess. Ethanol has less reactivity toward transesterificationas compared to methanol. Oleic acid was found to be thedominant fatty acid in both canola oil and greenseed canolaoil. An increase in unsaturation compound percentage as wellas pigment content leads to a decrease in biodiesel stability.The combination of bleaching of greenseed canola oil andtransesterification into a single step is not recommended.

Acknowledgment. Theauthors acknowledge theSaskatchewanCanola Development Commission (SCDC), AUTO21 Net-work of Centres of Excellence, and Natural Sciences andEngineering Research Council of Canada (NSERC) for financialassistance.

Documents

Biodiesel Production From Greenseed Canola Oil†