1638
S.J. CLARK, L. WAGNER, M.D. SCHROCK AND P.G. PIENNAAR
ACKNOWLEDGMENTS
This research was funded by USDA Grant 59-2202-1-6-059-0; the North Central USDA Energy Laboratory, Peoria, Illinois is the monitoring agency for the project.
REFERENCES
1. Energy Information Administration, Projection of Energy Supply and Demand and Their Impacts, Annual Report to Con- gress, Vol. II, Executive Summary, Department of Energy/
EIA-O036/2, Washington, DC, 1977.
2. Bruwer, J.J., B.v.D. Boshoff, F.J.C. Hugo, J. Fuls, C. Hawkins, A.N.v.d. Walt, A. Engelbrecht and L.M. du Plessis, in Agricul- tural Energy, Vol. 2, Biomass Energy/Crop Production, ASAE Publication 4-81, American Society of Agricultural Engineers, St. Joseph, MI, 1981, pp. 385-390.
3. Kaufman, K.R., in Fuels and Chemicals from Oilseeds: Tech-
nology and Policy Options, edited by E.B. Shulz, Jr., and R.P.
Morgan, Weswiew Press/AAAS Selected Symposium Series Volume, Boulder, CO, 1983, chapt. 8.
4. Alternate Fuel Council, Engine Manufacturers Association 200- Hour Screening Test for Alternate Fuels in Energy Notes for September 1, 1982, Northern Agricultural Energy Center/
Northern Regional Research Center, ARS/USDA, Peoria, IL.
5. Fort E.F., and P.N. Blumberg, in Vegetable Oil Fuels, Proceed- ings of the International Conference on Plant and Vegetable Oils as Fuels, August, 1982, Fargo, ND, ASAE Publication 4-82, American Society of Agricultural Engineers, St. Joseph, MI, 1982, pp. 374-383.
6. Diesel Vehicle Fuel, Lubricant, and Equipment Research Com- mittee, Diesel Engine Rating Manual, Coordinating Research Committee (CRC), Inc., New York, NY, 1959.
[Received March 14, 1984]
, Variables Affecting the Yields of Fatty Esters from Transesterified Vegetable Oils 1
B. FREEDMAN, E.H. PRYDE and T.L. MOUNTS, Northern Regional Research Center, Agricultural Research Service, U.S. Department of Agriculture, Peoria, I L 61604
ABSTRACT
Transesterification reaction variables that affect yield and purity of the product esters from cottonseed, peanut, soybean and sunflower oils include molar ratio of alcohol to vegetable oil, type of catalyst (alkaline vs acidic), temperature and degree of refinement of the vegetable oil. With alkaline catalysts (either sodium hydroxide or methoxide), temperatures of 60 C or higher, molar ratios of at least 6 to 1 and with fully refined oils, conversion to methyl, ethyl and butyl esters was essentially complete in 1 hr. At moderate temperatures (32 C), vegetable oils were 99% transesterified in ca.
4 hr with an alkaline catalyst. Transesterification by acid catalysis was much slower than by alkali catalysis. Although the crude oils could be transesterified, ester yields were reduced because of gums and extraneous material present in the crude oils.
INTRODUCTION
F a t t y esters f r o m vegetable oils have shown promise as alternative diesel fuels as a result of i m p r o v e d viscosity and volatility relative to the triglyceride (1-3) and this has s t i m u l a t e d interest in o p t i m i z a t i o n o f transesterification reaction c o n d i t i o n s (4-7). Transesterification (also called alcoholysis) has been studied intensively by n u m e r o u s investigators (8-12). O f the m o r e than a d o z e n US p a t e n t s on transesterification, the p a t e n t o f Bradshaw and Meuly (13) is cited f r e q u e n t l y .
O u r objectives were to develop basic i n f o r m a t i o n con- cerning the transesterification reaction of crude and refined oils to resolve inconsistencies in the literature, such as m o l a r ratios, and to delineate o p t i m u m reaction c o n d i t i o n s for m a x i m u m conversion to f a t t y esters with b o t h alkali- and acid-catalyzed reactions. T h e effects of such variables were evaluated q u a n t i t a t i v e l y with an Iatroscan T L C / F I D (thin layer c h r o m a t o g r a p h y / f l a m e ionization d e t e c t o r ) a n a l y z e r n o t available to earlier investigators.
EXPERIMENTAL PROCEDURES Materials
All refined vegetable oils were edible grade. R e f i n e d soy- bean oil was o b t a i n e d f r o m Central Soya, C h a t t a n o o g a , aPresented at the American Oil Chemists' Society annual meeting, Chicago, May 1983.
T e n n e s s e e ; s u n f l o w e r oil f r o m PVO I n t e r n a t i o n a l Inc., R i c h m o n d , California; p e a n u t oil f r o m Hain Pure F o o d Co., Los Angeles, California. Crude vegetable oils were o b t a i n e d f r o m the following sources: soybean and safflower oil f r o m A r r o w h e a d Mills, Inc., H e r e f o r d , T e x a s ; peanut oil f r o m Birdsong Corp., Suffolk, Virginia; c o t t o n s e e d oil f r o m Texas A&M University; s u n f l o w e r oil f r o m North D a k o t a State University.
T h e crude oils were processed by mechanical pressing, were n o t further refined and c o n t a i n e d s e d i m e n t b u t no added preservatives. T h e y were used as received; u n i f o r m samples were o b t a i n e d by t h o r o u g h m i x i n g of oil and any solids present. T h e edible-grade oils were d e g u m m e d , alkali-refined, bleached, filtered and d e o d o r i z e d .
S o d i u m m e t h o x i d e , as an a n h y d r o u s powder, was pur- chased f r o m Aldrich C h e m i c a l Co., Milwaukee, Wisconsin, s o d i u m h y d r o x i d e ( A C S grade) f r o m J.T. Baker Co., Phil- lipsburg, N e w J e r s e y ; c o n c e n t r a t e d sulfuric acid from Mallinckrodt, Inc., St. Louis, Missouri). Methanol and b u t a n o l were MCB O m n i s o l v (spectrograde) and were stored over m o l e c u l a r sieves 4A. E t h a n o l was distilled b e f o r e use and stored over m o l e c u l a r sieves 4A.
Transesterification Reaction Procedures
A 100 m L 3-necked flask e q u i p p e d with m e c h a n i c a l stirrer, t h e r m o m e t e r and c o n d e n s e r (to which a drying tube was a t t a c h e d ) was h e a t e d to e x p e l m o i s t u r e . On cooling, 60 g (0.0682 mol, assuming a m o l e c u l a r weight of 879.5 for s u n f l o w e r oil) edible-grade s u n f l o w e r oil was added to the flask, f o l l o w e d by 16.572 m L (13.115g, 0.4093 tool) m e t h a n o l . T h e m i x t u r e was stirred and h e a t e d in a silicone oil bath to 60 C, at which p o i n t 0.3 g s o d i u m m e t h o x i d e (0.5%, by weight of oil) was added rapidly. The tempera- ture rose to 63.3 C in 1 min, and the reaction m i x t u r e b e c o m e less turbid. Heating c o n t i n u e d f o r 1 hr at 60-63 C, at which p o i n t the r e a c t i o n was 98% c o m p l e t e d as deter- m i n e d b y latroscan analysis. A f t e r the r e a c t i o n m i x t u r e was allowed to cool to r o o m t e m p e r a t u r e , the ester and glycerol layers were separated in a separatory funnel. Excess me- t h a n o l in the ester layer was r e m o v e d on a rotary evapora- tor at reduced pressure or by distillation at atmospheric pressure. T h e ester was purified f u r t h e r by dissolving in JAOCS, Vol. 61, no. 10 (October 1984)
FATTY E S T E R S F R O M
petroleum ether (35-60 C), adding water, adding glacial acetic acid or phosphoric acid drop by drop until a pH o f ca. 7 was obtained, washing 3 times with water (vigorous shaking causes difficult emulsions), drying the petroleum ether solution over anhydrous magnesium sulfate and filter- ing and removing solvent on a rotary evaporator or by distillation at atmospheric pressure.
An acid-catalyzed butanolysis of soybean oil was con- ducted by heating 20 g (0.0253 m o b refined soybean oil, 69.47 mL (56.25 g, 0.759 tool) butanol and 0.2 g concen- trated sulfuric acid at 117 C for 4 hr. Samples were with- drawn after 0.1 hr, 0.2 hr, 0.5 hr, 1 hr, 2 hr, 3 hr and 4 hr.
On removal of excess butanol on a rotary evaporator at 50 C under reduced pressure and cooling to room tempera- ture, the glycerol and ester separated into 2 phases.
Sample Preparation and latroscan T L C / F I D Analysis F o r alkali-catalyzed studies, 0.5 m L of reaction mixture was withdrawn after 1 min, 2 min, 4 rain 30 min and 60 rain. F o r acid-catalyzed studies, samples were taken at longer intervals. Samples were quenched in 0.5 mL water.
A drop of the oil layer was weighed on an analytical ba- lance, and chloroform was added to give a concentration of 3.0 mg of oil per mL of solution. A chloroform solution of stearyl alcohol (as an internal standard) of known con- centration was added to the solution of reaction mixture so t h a t the weight ratio o f oil to stearyl alcohol was 10:1.
Analyses were performed with an latroscan TH-IO TLC Analyzer MK. III (Iatron Laboratories, T o k y o , J a p a n ; worldwide distributor, Newman-Howells Assoc., Winchester, Hants, United Kingdom) as previously described (4,14).
The FID was operated with hydrogen and air flow rates of 160 and 2000 mL/min. Scan speed (4) was 0.42 cm/sec.
The strip chart recorder was a Linear Instrument Corp.
(Irvine, California) Model 561 and was used at 1 mV full-scale deflection at a chart speed of 20 era/rain.
Solutions to be analyzed were spotted on Chromarod-S rods with a Hamilton 10 ttL syringe in amounts of 1-2 ttL o f solution per rod. Three-to-four replicates were used for each solution. The rods were developed in a glass t a n k containing petroleum ether (b.p. 35-60 C)/diethyl e t h e r / acetic acid (90:10:1) for 20 min, air-dried for 5 min, oven dried at 110-130 C for 5 min and then placed on the Iatroscan. The signals produced were sent to the recorder and were also integrated by a computer-assisted program.
A second computer program printed the percentages of composition of the reaction mixture.
R E S U L T S A N D D I S C U S S I O N
E f f e c t of Moisture and Free Fatty A c i d
Wright et al. (15) have noted t h a t the starting materials used for alkali-catalyzed aleoholyses of triglycerides should meet certain specifications. The glyceride should have an acid value less than 1, and all materials should be substan- tially anhydrous. The addition of more sodium h y d r o x i d e catalyst compensated for higher acidity, but the resulting soaps caused an increase in viscosity or f o r m a t i o n of gels and interfered with separation of glycerol. As little as 0.3%
water in the reaction mixture reduced glycerol yields b y consuming catalyst. Other investigators have also stressed the importance of a nearly dry oil substantially free (<0.5%) of fatty acids (13,16,17).
As we show later, when the reaction conditions do n o t meet the above requirements, ester yields are significantly reduced. Both sodium methoxide and sodium h y d r o x i d e should be maintained in an anhydrous state. Prolonged c o n t a c t With air will diminish the effectiveness of these catalysts through interaction with moisture and carbon
1 6 3 9 VEGETABLE O1LS
dioxide. Although transesterifications have been c o n d u c t e d in a nitrogen atmosphere to exclude moisture and carbon dioxide and to prevent oxidation of the oil (17,18), we found this precaution to be unnecessary.
Effect of M o l a r Ratios
One of the most i m p o r t a n t variables affecting the yield of ester is the molar ratio of alcohol to vegetable oil em- ployed. The stoichiometry of this reaction requires 3 tool alcohol per tool triglyceride to yield 3 tool f a t t y ester and 1 tool glycerol. Bradshaw (13,16) stated that a 4.8: t molar ratio of methanol to vegetable oil led to a 98% conversion, and if the alcohol was added in 3 or 4 portions, this ratio could be reduced from 4.8:1 to 3.3:1. He noted that ratios greater than 5.25:1 interfered with gravity separation of the glycerol and added useless .expense to the separation. In contrast, other investigators (17-19) obtained high ester conversions with a 6:1 molar ratio. In the ethanolysis o f peanut oil, a 6:1 molar ratio liberated Significantly more glycerol than did a 3:1 molar ratio (17). These workers also found t h a t glycerol yield increased from 77% to 95% as the sodium h y d r o x i d e catalyst increased from 0.2% to 0.8%
at the 6:1 molar ratio. Other investigators have used higher molar ratios, up to 45:1, particularly when the triglyceride contained large amounts of free fatty acids (20).
We studied the effect of molar ratio on ester yield, both because of the variation in the molar ratio reported as necessary by various investigators and because no cor- relation has been found of a systematic variation in molar ratio with ester yield. At the same time, the changes in concentrations of tri-, di- and monoglyceride were also monitored. The results obtained from the methanotysis of sunflower oil, in which the molar ratio was varied from 6:1 to 1:1, are shown in Figure 1. A t a 6:1 molar ratio, a 98%
100 I.,~'@
. I °
9O i o /
Triglyceride
.~."
,o .S
/ Methyl Sunflower Esters
/
70 /'
=~ ,®
-
/
I /
- Reaction Conditions
~t / 60 g Sunflower 0il (edible grade)
sQ \ i 6ooc
- 40
V
0.5 % Catalyst (Sodium meth0xide),0 ,.\
, I g ,
1; " \ ~ ,.J
10 Monoglycedde "~'~e ~
. . . . , . . x . . . . , . . . _
0|' 1 I I J " ",=,m====,~nn.
0 1 2 3 4 5 6
Molar Ratio Me0H/Sunflower 0il
I I 1 I I I I
- 1 0 0 - 6 6 - 3 3 0 33 66 100
Excess Me0H, wt. %
FIG. t. Product composition in transesterified sunflower oil at various molar ratios.
JAOCS, Voi. 61, no. 10 (October 1984)
B. F R E E D M A N E . i t . P R Y D E A N D T . L . M O U N T S
100
S yb r~ O i l ~ / ~ / ~k\_Cottonseed Oi
/ / /
~- 7 0 -
t .//¢/
Cor",d i £ 1 o,~8 7 1~ / f o ° ~ : " t ~ l " al . CEcirbl" °rad'e) 5o
~- J / 0.5% NoOCH s Ca+,~lysJ¢
4"0 ~ / ~ Sun1 e ] o w e r 0 i d
3 0 - _ _ _ _ ~ L - I I J I _. I I
- - I 0 0 - 6 6 - 3 3 0 3 3 6 6 I 0 0 t
E x c e s s M e O H , w£ . ~ I
1
_ _ I , I I I t ~ 1 I
0 I 2 3 4 5 6
Mo l ar Rat i o M e O H / V e g e t a b [ e 0 i i
F I G . 2 . Ester content in lxansesterified vegetable o i l s - e f f e c t o f
reactant ratios.
~00 s-,e 9o
80~
1
:' 601
Reoct[o~ C o ~ d r t r o n = 60g V e g e t o b l e O , i ( E d i b l e Grade) 60°0
N o ] ~ Ro£1o-
N e t h o n o l / V e g e t o b l e 0 ] ] ( 6 : 1 ) 0 5 % NczrOON 3 C ~ L ~ y s £
O i l
: / i /
S L ' S ~ r O . . . Oi I~ ~ " / L
~ CoLLomseed 0 i ;
5o'[ . . . -i . . . I ~ _ ~ ~ ... I ... 1_ A _ _ I ~ . , _ ~
I 2 3 a 5 10 PO 30 ; 0 ,bO IOO
7 me m in
F I G . 3 . Effect of time on the alkali-catalyzed transesterification of
vegetable oils at molar ratio CH 3 OH/oil of 6:1, sodium m e t h o x i d e catalyst (0.5%).
conversion to the ester was obtained. As the molar ratio decreased to the theoretical ratio of 3 : 1, the percentage of ester decreased to 82%. Tri-, di- and monoglycerides, which were at low levels at the 6:1 ratio, also increased with decreasing molar ratio, indicating that conversion to the ester was n o t complete. Our results suggest that for maxi- mum conversion to the ester, a 6:1 ratio should be used.
Ratios greater than 6:1 do n o t increase yields (already at 98-99%), complicate ester and glycerol recovery, and add costs to alcohol recovery. The results o f varying the molar ratio of methanol to vegetable oil, when extended to re- fined soybean, sunflower, peanut and cottonseed oils, showed that these oils behaved similarly, and that highest conversions to esters (93-98%) were observed at the 6:1 molar ratio (Fig. 2).
E f f e c t of R e a c t i o n T i m e on the A l k a l i - C a t a l y z e d A l c o h o l y s i s
Several investigators have followed fatty ester formation with time. In the first 5 min o f the ethanolysis o f p e a n u t oil, ca. 80% o f the ester obtained after 2 hr was formed ( I 7 ) . The alcoholysis o f menhaden oil, using a variety o f alcohols, has been m o n i t o r e d by thin layer chromatography
(TLC) (18,19). Depending on the alcohol used, reactions at 60-100 C resulted in 80-99% conversions in 2-60 rain.
We investigated the methanolysis of the 4 refined vegetable oils, following the conversion to ester with time.
Molai" ratios of methanol to oil of 6:1 (Fig. 3) and 3:1 were employed. Samples were taken from the reaction mixture after adding catalyst. F o r soybean and sunflower oils, ca.
80% ester was observed after 1 min, confirming the rapidity of alkali-catalyzed methanolysis. F o r unknown reasons, ester conversions after 1 rain with p e a n u t and cottonseed oils were distinctly lower. After 1 hr, however, the per- centage of methyl ester for all 4 oils ranged from 93-98%.
Thus, given sufficient time, these 4 oils underwent con- version to about the same extent. As expected for a 3:1 molar ratio, ester conversion for all 4 oils was considerably lower, ranging from 89-74% for soybean oil and cottonseed oil esters, respectively, after 1 hr.
Our study was e x t e n d e d to the alcoholysis of sunflower oil with ethanol and butanol as well as methanol at molar ratios of 6:1 and 3:1. Each reaction was conducted a few degrees below the boiling point of the alcohol and was catalyzed with 0.5% sodium methoxide based on the weight of the oil. At both the 6:1 and 3:1 molar ratios, a higher ester conversion occurred initially with butanol than with ethanol or methanol because of the higher reaction tem- perature employed (Fig. 4). After 1 hr at the 6:1 ratio, the percentage of ester from all 3 alcohols was nearly identical (96-98%). After 1 hr at the 3:1 ratio, conver- sion to butyl esters was 88% compared with 81% and 82%
for ethyl and methyl esters. These results confirm the importance of using an excess of alcohol compared with the stoichiometric amount for maximum conversion to the ester.
Change in G l y c e r i d e C o m p o s i t i o n D u r i n g Transesterification
Most investigators have not studied changes in glyceride composition during the aleoholysis of fats and oils. How- ever, Grtin et al. (21) reported that the alcoholysis is a step by step process and that mono-, di- and triglycerides are all present in the final product. Feuge and Gros (17) followed glyceride distribution during ethanolysis of peanut oil. The latter investigators used the more complicated classical chemical methods for glyceride analyses in contrast to the more convenient and faster T L C / F I D method des- cribed here.
We determined differences in glyceride composition between reactions employing a 6:1 vs a 3:1 molar ratio of ethanol to sunflower oil (Fig. 5). As anticipated, total gly-
I
Molar Ratio o Butyl, 114%l ROH/0il & Methyt, 60~C
100 ~ ~ o Ethyl, 75~C
A ~ A ss ~°
i l J J . . . .= -%"-'--"
76~,,-- o . . . . - : : - - "
68 L . - = " ' ; ; x ~ =~" . . . . "~':"""
60 I°" t l .... I I
1 2 4 30 60
Time, rain
F I G . 4 . Transesterification of sunflower oil to m e t h y l , ethyl a n d
butyl esters.
J A O C S , Vol 61, no 10 ( O c t o b e r 1 9 8 4 )
1641
FATTY ESTERS FROM VEGETABLE OILS
14'1"-J'- " . . . . o - . .
" " - - a ---3:1
I'% Oiglyceride - ~ /
10 ~ '%% Triglyceride ~
8
~ - ~ o ~ . . M0n0glycer,de. e... °
i ,'
. . . " % - . . " . . ..
N Di~llvceride ~ - " " o . . . 0
o.
... ~
~A~L.~.. M0nonlvcefide2 4 30 60
hme. rain
FIG. 5~ G l y c e r i d e c o m p o s i t i o n d u r i n g t r a n s e s t e r i f i c a t i o n o f s u n - f l o w e r oil t o e t h y l esters.
cerides during the 6:1 run were generally lower than those during the 3:I run, indicating that the former reaction was more complete. In both cases triglyceride concentration was lower than di- and monoglyceride concentration after 60 min, showing that the triglyceride had reacted to form intermediates and product. The combination of glycerides and ester accounted for all major components of the reaction mixture; only trace amounts of fatty acids and u n k n o w n s were detected, Even under our most suitable conditions for high conversions to the ester, traces of glycerides were still present because of an equilibrium be- tween products and reactants.
Glycerol Recovery
Although the fatty ester is the major product desired, recovery of glycerol, the by-product, is also important.
Glycrol has many industrial uses. An estimate has been made that the cost of the transesterification processing could be paid for if suitable markets can be found for the glycerol (A. Aguiar, personal communication). Bradshaw has noted that glycerol produced by aleoholysis of refined fats is anhydrous and can be used directly in many pro- cesses without further refining (16).
We have determined the yields of glycerol formed during the transesterification of refined sunflower, peanut and cottonseed oils. Crude glycerol yields were about the same for all 3 oils and depended on the molar ratio of alcohol to oil employed. With sunflower, for example, the molar ratios and resulting yields of crude glycerol were: 6:1,
100%; 5:1, 100%; 4:1, 79%; 3:1, 56%; 2:1, 18%; 1:1, 0%.
Thus, for maximum yields of both ester and glycerol, a 6 : 1 ratio was best.
Methanolysis of Crude and Refined Vegetable Oils
Although a dry, neutral oil is preferred as a starting ma- terial, some investigators have reported the use of crude fatty materials for alcoholysis. Bradshaw and Meuly (13,16) stated that their process could accommodate unrefined oils containing small a m o u n t s of moisture and free fatty acids.
With such oils, reaction required larger quantities of alkali, the soaps formed needed to be acidified before glycerol would separate and glycerol separation was slower. Sprules and Price (20) stated that the use of a large excess of alco- hol coupled with alkaline and acid catalysis can also accom- modate low quality and unrefined fats and oils.
We compared the transesterification of crude and refined oils. A 6:1 molar ratio of methanol to oil was employed.
Free fatty acid and phosphorus contents, the former cer- tainly and the latter possibly leading to catalyst destruc- tion, were determined in the crude oils by standard AOCS methods; both values were higher in the crude than in the refined oils as expected (Table I). The a m o u n t of catalyst was adjusted for the acid values of the crude oils. In every case substantial conversion to ester was obtained with the crude oils, b u t these yields were always noticeably lower than those obtained with refined oils. Part of the reduced yields with the crude oils could be ascribed to the presence of solid, extraneous material present in these oils. For maximum yields the refined oils should be used. However, if the esters are to be used as a diesel fuel, using a crude degummed and filtered oil as starting material for the transesterification may suffice.
Effect of Temperature on Ester Formation
Although alkaline alcoholysis of vegetable oils is normally conducted near the boiling point of the alcohol, researchers have reported that the reaction may be carried out at room temperature (8,15,22). In methanolysis of castor oil to methyl rieinoleate, the reaction proceeded most satis- factorily at 20-35 C at molar ratios of methanol to oil of 6:1 to 12:1 (23). South African scientists have reported a 90% yield of ethyl esters of sunflower oil by conducting a large-scale alkali ethanolysis at room temperature (24).
We studied the methanolysis of refined soybean oil at 60 C, 45 C and 32 C using conditions shown in Figure 6.
On adding catalyst to the solution at ambient temperature (28 C), the temperature rose to 32.6 C (because of the T A B L E I
T r a n s e s t e r i f i c a t i o n o f Crude a n d R e f i n e d V e g e t a b l e Oils with Methanol
Phosphorus b Methyl ester
Oil Type Acid value a (ppm) yield (wt. %)
Peanut Crude 6.66 264 67
Refined 0.08 5 95
Crude 1.67 953 83
Soybean Refined 0.12 1 98
Safflower Crude 0.44 4 86
Refined -- -- -
Crude 0.28 <1 84
Cottonseed Refined 0. 06 0.5 93
Sunflower Crude 1.64 85 81
Refined 0.08 0.7 97
aAOCS Official Method Cd 3a-63.
bAOCS Official Method Ca 12-55.
JAOCS, Vol. 61, no. 10 (October 1984)
1642
B. FREEDMAN, E.H. PRYDE AND T.L. MOUNTS 100
90
8O
~ 7o 6O
. •
Reaction Conditions_._/yc 0o,: 0,,
/ Me0HIS0ybean 0il (6:1)
1% Catalyst (Sodium Hydroxide)
p
50 I 1 t ... I 1 1 1
0.1 0,2 0.5 t 2 4 10 20
Time, hr
30
FIG. 6. Effect .of temperature and time on methyl ester formation, sodium hydroxide catalyst (1%).
,ooL ,, 0;;,6:,, .__.=,
t_ " ,
; 00/- o j O
20 ~ " ~ a ~ I
0t t I _ ,I J ... J J
1 2 4 8 16 30 60 120
TimB, mW
FIG. 7. Comparison of sodium methoxide vs sodium hydroxide catalysts.
I O0
8 0
~ 40
2 0 g S o y b e o n O i f / E d l b ; e G r o c t e )
~!o ~ a r R ~ t ~o-
Aleoho~/Sc, ybe~n 0 [ ~ (30 [}
1~ H2SO 4 0 a t o ] y ~ +.
~ V ~ / Ethy, /'/
"i-- "%"
/ /
65°C / /
/ / j /
t 2 3 4 5 ~0
[ i m e . ½r-
o [ - _ . L _ ~ t ~ L / _ , . ~ ~ t ,a L . t . . t J a ~
0 . ' 0.2 0.,5 20 "30 50 100
FIG. 8. Transesterification of soybean oil to methyl, ethyl and butyl esters with 1% Hz SO 4 catalyst.
e x o t h e r m i c i t y of the reaction) and remained at 32-+1 C during the run. A f t e r 0.1 hr, the esters present in the 60 C, 45 C and 32 C runs were 94%, 87% and 64%, showing the influence of temperature on ester conversion. However, after 1 hr, ester formation was identical for the 60 C and 45 C runs, and only slightly lower for the 32 C run. After 4 hr the ester conversion for the 32 C run slightly exceeded that of the other runs.
C o m p a r i s o n o f S o d i u m M e t h o x i d e vs. S o d i u m H y d r o x i d e C a t a l y s t s
Alkali metal alkoxides are the most effective alcoholysis catalysts (8,11,12) and are also effective in interesterifica- tion (11,25). The lower cost of sodium h y d r o x i d e has
attracted its wide use in large-scale transesterifications. On a laboratory scale, Hartman (26) compared the methanol- ysis of several triglycerides with various alkali metal ca- talysts including sodium m e t h o x i d e and sodium hydroxide.
At the 0.5% level for each catalyst, the glycerol set free by the methoxide was ca. 99% compared with ca. 98% with the h y d r o x i d e catalyst. Although geometric and positional isomerization might be anticipated as possible side reactions, under the mild conditions employed in alkali-catalyzed alcoholysis such isomerizations are not significant (18).
We compared these 2 catalysts at a 6:1 and 3:1 molar ratio of methanol to vegetable oil (Fig. 7). Ester conver- sions at the 6:1 ratio for both catalysts were very similar, and after 60 min were essentially identical. Hence at the 6:1 molar ratio, 0.5% soidum methoxide was as effective as 1% soidum hydroxide. At the 3:1 ratio, however, the m e t h o x i d e catalyst was d e a r l y superior to the h y d r o x i d e catalyst during the entire 60-rnin reaction time.
A c i d - C a t a l y z e d A l c o h o l y s i s
Alkali catalysis is considerably faster than acid catalysis for the atcoholysis of esters (27-28). Even at ambient temperature, the alkali-catalyzed reaction proceeds rapidly, whereas acid-catalyzed reactions c o m m o n l y require tem- peratures above 100 C (8,12). With acid-catalysis, reaction times o f 3-48 hr have been reported, except when reactions were conducted under high temperature and pressure (29, 30). Keim (28) has noted that acid-catalyzed alcoholysis can be used when the starting materials are low-grade fats or have a high free fatty acid content; the fatty acids would deactivate an alkaline catalyst.
We compared the transesterification of soybean oil with methanol, ethanol and butanol using 1% concentrated sulfuric acid based on the weight of oil. In preliminary experiments with 6:1 and 20:1 molar ratios at 3 and 18 hr, respectively, conversions to ester were unsatisfactory. A molar ratio of 30:1, however, resulted in a high conversion to the methyl ester. This molar ratio was used during our studies o f methanol, ethanol and butanol (Fig. 8). Each aleoholysis was conducted near the bioling point of the alcohol. The number of hours needed to obtain high con- versions to the ester were 3, 22 and 69, respectively, for the butyl, ethyl and methyl esters. Reaction temperature rather than the type of alcohol appeared to control the time to completion. To confirm this, alcoholysis was conducted at 65 C for all 3 alcohols. At this temperature, rates of conver- sion and final conversion to ester after 69 hr were similar for all 3 alcohols.
R e c o m m e n d e d C o n d i t i o n s f o r Ester F o r m a t i o n
To obtain maximum ester formation by transesterification of vegetable oils, refined oils, substantially anhydrous with a free fatty acid content of less than 0.5% (acid value less than i ) , should be used. The alcohol should also be mois- ture free. A molar ratio of alcohol to oil of 6:1 gives o p t i m u m conversion to the ester. Either 0.5% sodium methoxide (for laboratory use) or 1% sodium hydroxide (for larger-scale reactions) are effective catalysts. These catalysts should be stored under anhydrous conditions free from air. Ester conversions of 96-98% are obtained by transesterifying refined oils with methanol, ethanol and butanol at 60 C, 75 C and 114 C for 1 hr with 0.5% soidum methoxide.
ACKNOWLEDGMENT
Doug Williams assisted in conducting reactions and analyzing the products.
J A O C S , Vo(. 6 1 , n o . 1 0 ( O c t o b e r 1 9 8 4 )
FATTY ESTERS FROM VEGETABLE OILS
1643
REFERENCES
I. Hawkins, C.S., and J. Fuls, in Vegetabl Oil Fuels-Proceedings of the International Conference on Plant and Vegetable Oils as Fuels, American Society of Agricultural Engineers, St.
Joseph, MI, 1982, p. 184.
2. Pischinger, G.H., A.M. Falcon, R.W. Siekman and F.R. Fer- nandes, Ibid. p. 198.
3. Clark, S.J., L. Wagner, M.D. Schrock and P.G. Piennaar, JAOCS 60:736 (1983).
4. Freedman, B., and E.H. Pryde, in Vegetable Oil Fuels-Pro- ceedings of the International Conference on Plant and Vege- table Oils as Fuels, American Society of Agricultural Engineers, St. Joseph, MI, 1982, p. 117.
5. Hassett, D.J., and R.A. ttasan, Ibid. p. 123.
6. Kusy, P.F., Ibid. p. 127.
7. Freedman, B., and E.H. Pryde, Vegetable Oil as Diesel F u e l - Seminar II, Northern Agricultural Energy Center, Peoria, IL, October 1981, Appendix 8.
8. Markley, K.S., in Fatty Acids, Their Chemistry, Properties, Production and Uses, edited by K.S Markley, 2rid edn,, Interscience Publishers, New York, 1961, Part 2, p. 861.
9. Markley, K.S., Ibid. 1968, Part 5, p. 3536.
10. Sonntag, N.O.V., in Bailey's Industrial Oil and Fat Products, edited by D. Swern, Vol. 2, 4th edn., J. Wiley and Sons, New York, 1982, p. 130.
11. Formo, M,W., JAOCS 31:548 (1954).
12. Sridharan, R., and I.M. Mathai, J. Sci. Ind. Res. 33:178 (1974).
13. Bradshaw, G.B., and W.C. Meuly, US Patent 2,360,844 (1944).
14. Freedman, B., E.H. Pryde and W.F. Kwolek, JAOCS 61:1215 (1984).
15. Wright, H.J., J.B. Segur, tt.V. Clark, S.K. Coburn, E.E. Lang- don and R.N. DuPuis, Oil & Soap 21:145 (1944).
16. Bradshaw, G.B., Soap Sanit. Chem. 18:23 (1942).
17. Feuge, R.O., and T. Gros, JAOCS 26:97 (1949).
18. Gauglitz, E.J., and L.W. Lehman, Ibid. 40:I97 (1963).
19. Lehman, L.W., and E.J. Gauglitz, Ibid. 43:383 (1966).
20. Sprules, F.J., and D. Price, US Patent 2,494,366 (1950).
21. Grfin, .A', F. Wittka and E. Kunze, Chem. Umschau Gebiete Fette Ole Wasche u. Harze 24:15 (1917).
22. Eckey, E.W., JAOCS 33:575 (1956).
23. Smith, M.K., US Patent 2,486,444 (1949).
24. Fuls, J., and FJ.C. Hugo, Beyond the Energy Crisis--Oppor- tunity and Challenge, Volume llI, Third International Con- ference on Energy Use Management, West Berlin, October 26-30, 1981, Pergamon Press, Oxford, p. 1595.
25. Sreenivasan, B., JAOCS 55:796 (1978).
26. Hartman, L. Ibid. 33:129 (1956).
27. Reid, E.E., Am. Chem. J. 45:506 (1911).
28. Keim,G.l., US Patent 2,383,601 (1945).
29. Taylor, G.R., and H J . Clarke, J. Am. Chem. Soc. 49:2829 (1927).
30. Allen, H.W., and W.A. Kline, US patent 2,383,579 (1945).
[Received March 14, 1984]
JAOCS, Vol. 61, no. 10 (October 1984)
