Ultrasonics Sonochemistry 12 (2005) 367–372

www.elsevier.com/locate/ultsonch

Fatty acids methyl esters from vegetable oil

by means of ultrasonic energy
Carmen Stavarache a, M. Vinatoru a, R. Nishimura b, Y. Maeda b,*

a
Costin D. Nenitzescu Institute of Organic Chemistry, 71141 Bucharest, P.O. Box 15-254, Romania
b
Department of Applied Material Science, College of Engineering, Osaka Prefecture University, Gakuen-cho 1-1, Sakai, Osaka 599-8531, Japan
Received 1 February 2004; received in revised form 29 March 2004; accepted 14 April 2004
Available online 7 June 2004

Abstract

The transesterification of vegetable oil with short-chain alcohols, in the presence of base-catalyst, by means of low frequency
ultrasound (28 and 40 kHz) in order to obtain biodiesel fuel was studied. By using ultrasounds the reaction time is much shorter (10–
40 min) than for mechanical stirring. The quantity of required catalyst is 2 or 3 times lower. The molar ratio of alcohol/oil used is
only 6:1. Normal chain alcohols react fast, while secondary and tertiary alcohols show some or no conversion after 60 min of
reaction. Surprisingly, 40 kHz ultrasounds are much more effective in the reduction of the reaction time (10–20 min). Twenty eight
kilohertz give slightly better yields (98–99%), but longer reaction time, while higher frequencies are not useful at all for the
transesterification of fatty acids.

2004 Elsevier B.V. All rights reserved.

Keywords: Vegetable oil; Transesterification; Ultrasound; Base catalysis; Biodiesel

1. Introduction to be the most promising alternative. Today, methyl or

The idea of using vegetable oil as a substitute for
diesel fuel was demonstrated by the inventor of the
diesel engine, Rudolph Diesel, around the year 1900,
when vegetable oil was proposed as fuel for engines.
Vegetable oils have good heating power and provide
exhaust gas with almost no sulfur and aromatic poly-
cyclic compounds. Due to the fact that vegetable oils are
produced from plants, their burning leads to a complete
recyclable CO2 . The oil use as diesel fuel was limited due
to its high viscosity (near 10 times of the gas oil). In
order to adapt the fuel to the existing engines the veg-
etable oils had to be modified.
Various derivatives such as microemulsions or blends
of various vegetable oils with conventional fuel have
been proposed as alternative fuels for diesel engines
[1,2]. The esters of vegetable oils or animal fats appear
*
Corresponding author. Tel./fax: +81-72-254-9321.
E-mail address: maeda@ams.osakafu-u.ac.jp (Y. Maeda).
ethyl esters of fatty acids are used as substitute to
petroleum-based diesel fuels under the name of ‘‘bio-
diesel’’ [3].
Bio-diesel fuels have many advantages over petro-
leum diesel fuel: produce less smoke and particles, have
higher cetane number, produce lower carbon monoxide
and hydrocarbon emissions, are renewable, biodegrad-
able and non-toxic. When ethyl esters are used as fuel
the advantage of totally recyclable CO2 cycle is obtained
since ethyl alcohol could be of vegetal origin [4].
Transesterification is not a new process. The scientists
Duffy and Patrick conducted it as early as 1853 [5].
Transesterification involves the reaction between an
alcohol and a vegetable oil or animal fat that are mix-
tures of triglyceride (esters of glycerin with long chain
fatty acids).
To complete the reaction stoichiometrically a 3:1
molar ratio of alcohol to triglyceride is necessary. Due
to the fact that the transesterification is an equilibrium
reaction, an excess of alcohol is used to displace the
reaction towards esters formation. Fats and alcohols are
not totally miscible, so their reaction takes place at the

1350-4177/$ - see front matter
2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ultsonch.2004.04.001
368 C. Stavarache et al. / Ultrasonics Sonochemistry 12 (2005) 367–372
interface and it is a very slow process [6]. The reaction is
catalyzed by alkali, acid or enzyme. Enzymes-catalyzed
procedures, using lipase as catalyst, do not produce side
reactions [7–9], but the lipases are very expensive for
industrial scale production and a three-step process was
required to achieve a 95% conversion [10]. Acid-cata-
lyzed process is useful when a high amount of free acids
are present in the vegetable oil [11,12], but the reaction
time is very long (48–96 h), even at the boiling point of
the alcohol, and a high molar ratio of alcohol was
needed (20:1 wt/wt to the oil) [13].
In the base-catalyzed procedure, some soap is formed
and it acts as phase transfer catalyst, thus helping
the mixing of the reactants. Base-catalyzed process is
strongly affected by the mixing of the reactants and/or
by efficient heating [4,14] that produces tiny droplets,
thus increasing the reaction area [14]. Today, mixing/
heating is the process of choice used in industrial
application in over 85 biodiesel plants worldwide [3].
Low frequency ultrasonic irradiation is a useful tool
for emulsification of immiscible liquids. The collapse of
the cavitation bubbles disrupts the phase boundary and
causes emulsification, by ultrasonic jets that impinge one
liquid to another [15].
We were interested in finding a more efficient alter-
native for biodiesel fabrication by the help of ultra-
sound. The aim of this study was to see the effects of
low frequency ultrasonic irradiation on the transeste-
rification of vegetable oil under base-catalysis condi-
tions.
The results of the investigation of the transesterifi-
cation of neat vegetable oil with short-chain alcohols
(C1–C4) in the presence of ultrasound to give alkyl es-
ters of fatty acids are given. Reaction conditions that
affect the yield include the type and concentration of the
catalyst, reaction time and ultrasound frequency.
2. Method
2.1. Reagents and materials
Sodium hydroxide (>96%) and potassium hydroxide
(>85%) were purchased from Wako Chemicals and used
after milling, to facilitate the dissolution in alcohol. The
alcohols employed in the reactions (methanol, ethanol,
n-propanol, iso-propanol, n-butanol, iso-butanol, ter-
tiary-butanol), from Wako Chemicals had more than
99.5% purity and were used as received. The vegetable
oil was of commercial edible grade and it had the fol-
lowing basic properties:
Acid value (i.e. the percentage of free fatty acids in
the fat) ¼ 1.247,
Iodine value (i.e. gram of iodine absorbed by 100 g of
oil) ¼ 70,
Saponification value (i.e. milligram of KOH required
to neutralize the free fatty acids and to saponify the
esters in 1 g of oil) ¼ 195.
2.2. Procedures
The influence of low frequency ultrasound (28 and 40
kHz) versus traditional stirring on the bio-diesel pro-
duction from neat vegetable oil, at room temperature,
has been investigated. The reaction mixtures consisted in
vegetable oil, alcohol and alkaline catalyst. The molar
ratio of alcohol to oil was 6:1, and the quantity of catalyst
was 0.5%, 1.0% and 1.5% (wt/wt) to the oil respectively.
The hydroxide was dissolved into the alcohol prior to the
addition of vegetable oil. The silent reactions were per-
formed using a Matsushita Electric Ind., Model
SCV35W stirrer at 1800 rot/min. The ultrasonic reactions
were performed using Honda Electronics Ultrasonic
Cleaners WS 1200-28 and WS 1200-40, with a total power
of 1200 W, working power being set at 60%.
After the complete conversion of the vegetable oil, as
determined by TLC analysis, the reaction was stopped
and the reaction mixture was allowed to stand for phase
separations: the ester mixture formed the upper layer
and glycerin formed the lower layer. The residual cata-
lyst and non-reacted alcohol were distributed between
the two phases: some traces into the ester layer and a
high amount in the glycerin layer (due to a high misci-
bility of the alcohol with glycerin). After the phases
separation the traces of catalyst and alcohol were wa-
shed out with water from the esters mixture and the
esters were dried over anhydrous calcium chloride. All
experiments were performed in an Erlenmeyer type flask
having 100 ml volume.
2.3. Sampling and analysis
Samples were taken at 10 min time intervals and
analyzed by TLC (eluent chloroform/petroleum ether,
1:3 ratio) to check the conversion to biodiesel. TLC was
chosen as a rapid analytical method and it gives quite
accurate indication of oil and biodiesel content in the
mixture After purification the biodiesel was analyzed
using a Shimadzu GC-MS Model QP-2010, equipped
with a DB-5 capillary column (0.53 mm · 30 m) J&W
Scientific.
The chemical composition of the vegetable oil was
determined on the basis of GC-MS analysis of the me-
thyl esters. The mean molecular weight of the esters
mixtures was calculated averaging the individual
molecular weights of each constituent ester. For methyl
esters the mean molecular weight was 293.65. The mean
molecular weight of the oil was calculated averaging the
individual molecular weights of each constituent tri-
glyceride, according to the fatty esters analysis (Table 1).
The average molecular weight of the oil was 878.23.
 C. Stavarache et al. / Ultrasonics Sonochemistry 12 (2005) 367–372 369

Table 1 wards formation of the products, the alcohol must be

The composition of the fatty acid esters
Fatty acid % case, the complete conversion is difficult to be reached.
Hexadecanoic 10.12
Octadecadienoic 47.18
Octadecenoic 38.30
Octatetranoic 0.29
Octadecanoic 3.82
Eicosanoic 0.27
3. Results and discussions
Base-catalyzed procedures are simpler and much
more rapid. They do not require a high excess of alco-
hol. Their only disadvantage is the side-reaction between
triglyceride and catalyst, the formation of soap, espe-
cially when the catalyst is added in amounts higher than
1.4–1.5% [16]. However, some soap formation is neces-
sary because is acting as a phase transfer catalyst and
increases the mixing of reactants.
The reaction mechanism was formulated as a three-
step reaction [2,17,18]. The first step involves the
reaction between the triglyceride and one molecule of
alcohol, leading to diglycerides and one molecule of
ester. The second step is the reaction of the diglycerides
with a second molecule of alcohol, leading to mono-
glycerides and another molecule of ester, and the third
one the reaction of monoglycerides with the third mol-
ecule of alcohol leading to glycerin and ester. The
kinetics of this mechanism is very complicated. Even if
we consider the reaction between a triglyceride having
the same fatty residue (R1 ) and alcohol, the overall
process involves at least twelve equilibrium reactions
(Scheme 1):
All the reactions involved in the transesterification
process are equilibrium reactions, making very difficult
the kinetic study. In order to displace the reaction to-
present in excess in the reaction mixture. Even in this
Since base-catalyzed procedure is the major technique
used for industrial processes for biodiesel production,
we were interested in finding a less expensive and time-
costing method, and ultrasonic irradiation proves to be
a good choice. These advantages reflect in shortening
the reaction time and the quantity of catalyst used with
respect to the yield of isolated product [19].
3.1. The influence of the alcohol type
A set of experiments was carried out to see the effect
of alcohol type to the transesterification of the trigly-
cerides. In Table 2 are presented the results obtained for
the transesterification of vegetable oil with secondary
and tertiary alcohols after 60 min of reaction.
The secondary alcohols show some conversion under
ultrasonic irradiation, while under stirring the trans-
esterification does not take place. Tertiary alcohols have
no conversion in both types of activation. We assume
that this behavior is due to the sterical hindrance that
impedes the access towards the reaction center, or in
some case the access is limited, that makes the reaction
rate very slow compared with normal-chain alcohols.
Thus, from two competing reactions, i.e. saponification
and transesterification, the first one takes place more
Table 2
The influence of alcohol on the transesterification time
Alcohol After 60 min reaction time
Stirring Ultrasonic irradiation
Iso-propanol No conversion Some conversion
Iso-butanol No conversion Some conversion
Tertiary-butanol No conversion No conversion

Ester + MG2

Ester + DG1 Gly + Ester

Ester + MG1
TG + alcohol

Ester + DG2

Where:
TG DG1 DG2 MG1 MG2 Gly Ester Alcohol

H2C OCOR1 H2C OCOR1 H2C OCOR1 H2C OCOR1 H2C OH H2C OH R2OCOR1 R2OH
HC OCOR1 HC OH HC OCOR1 HC OH HC OCOR1 HC OH
H2C OCOR1 H2C OCOR1 H2C OH H2C OH H2C OH H2C OH

Scheme 1. The mechanism of transesterification.

370 C. Stavarache et al. / Ultrasonics Sonochemistry 12 (2005) 367–372

rapidly, consuming the oil and catalyst and leading to
soap formation.
In Table 3 the yields of isolated methyl esters for
different types of activation in the presence of NaOH as
catalyst are given.
From the results we can point out that under ultrasonic
irradiation at 40 kHz frequency the reaction was complete
faster. The best yields were obtained when the catalyst
was used in small concentration, i.e. 0.5% wt/wt of oil.
By increasing the amount of catalyst, emulsions are
formed in the washing step thus hindering the purifica-
tion. During washing the soap present in the esters
phase has the tendency to accumulate at the interfacial
region between the liquids. Due to its structure (one
hydrophilic end and a long hydrophobic chain) the soap
molecules are oriented perpendicular on the interfacial
region between the two immiscible phases: water and
esters. The soap molecules collect inside esters and water
molecules forming emulsion, thus making the separation
difficult. When the concentration of catalyst is increased,
soap is formed in higher amount, leading to formation
of gels. In the case of 1.5% (wt/wt) NaOH the yields of
isolated esters was very low due to the fact that the soap
production in high quantity led to formation of gels in
high quantity, and a high amount of esters remains
trapped in the glycerin layer.
In the case of ethyl esters the yields of isolated
product were lower (Table 4). Ethanol dissolves the oil
as soon as esters and soap are formed. The solubility of
the two phases increases and the reaction time is short.
Since the yields of isolated products in sonolysis are
higher than stirring, we suppose that ultrasounds pro-
duce less soap.
Indeed, after less than one minute of mixing or son-
ication the mixture became homogeneous. For 28 kHz,
increasing the catalyst concentration decreases the time
to complete the reaction with almost the same yield. In
the experiments carried out under mechanical stirring by
increasing the catalyst concentration the yields are
lowered. The same tendency is observed in the case of
sonication at 40 kHz. Since in both types of activation
the yields of isolated product are decreased, the reason
cannot be the mixing efficiency, but isolation of the
desired product. Due to increased miscibility between
ethanol, glycerin and esters the separation and purifi-
cation of esters mixture is difficult. This causes low
yields of isolated product.
Increasing the chain length, the miscibility between
oil and alcohol increases thus decreasing the reaction
time (10–20 min) but also making the separation of
esters difficult (Table 5).
As one can see in Table 5 the yields of isolated n-
propyl esters are lower than methyl esters but there are
only small differences in comparison with ethyl esters
(Table 4). Even if the reaction is completed very fast (as
observed by TLC analysis), the washing and purification
is complicated due to the fact that the separation of the
two phases does not take place in this case.
The miscibility of the oil is related to the solubility of
glycerides and is affected mainly by unsaturation and

Table 3
The yields of isolated methyl esters
Method 0.5% (wt/wt) NaOH 1.0% (wt/wt) NaOH 1.5% (wt/wt) NaOH
Time (min) Yield (%) Time (min) Yield (%) Time (min) Yield (%)
Mechanical stirring 60 80 10 91 10 35
Ultrasonic irradiation 28 kHz 40 98 10 95 10 75
Ultrasonic irradiation 40 kHz 20 98 10 91 10 68

Table 4
The yields of isolated ethyl esters
Method 0.5% (wt/wt) NaOH 1.0% (wt/wt) NaOH 1.5% (wt/wt) NaOH
Time (min) Yield (%) Time (min) Yield (%) Time (min) Yield (%)
Mechanical stirring 20 79 20 40 10 47
Ultrasonic irradiation 28 kHz 50 74 20 75 20 75
Ultrasonic irradiation 40 kHz 20 88 10 48 10 30

Table 5
The yields of isolated n-propyl esters
Method 0.5% (wt/wt) NaOH 1.0% (wt/wt) NaOH 1.5% (wt/wt) NaOH
Time (min) Yield (%) Time (min) Yield (%) Time (min) Yield (%)
Mechanical stirring 10 78 10 47 10 79
Ultrasonic irradiation 28 kHz 20 92 20 83 20 75
Ultrasonic irradiation 40 kHz 10 88 10 56 10 78
 C. Stavarache et al. / Ultrasonics Sonochemistry 12 (2005) 367–372
chain length of the fatty acids constituents. In the
present study, by using the same molar ratios of 6:1
alcohol to oil, 1 mole of oil dissolved 1.88 moles of
methanol, 2.77 moles of ethanol, 3.92 moles of n-pro-
panol and 6 moles of n-butanol. These results clearly
show that the longer the alcohol chain the soluble the
alcohol into oil.
In the case of n-butanol good yields were obtained for
both types of activation (Table 6), but in long reaction
time. After 10 min of reaction, some oil still persists, as
determined by TLC analysis. According to Freedman
et al. [18] the transesterification of oil with MeOH is
different from BuOH due to a higher polarity of meth-
oxy anion compared with butoxy anion. Also, the
smaller size of methanol molecule could facilitate the
simultaneous attack on all three chains of triglyceride,
while butanol should follow the all pathway from TG to
DG to MG in order to form the esters (Scheme 1).
By examining the results presented in Tables 2–6 one
can see that there are notable differences in the conver-
sion of neat vegetable oil in the presence of various
alcohols. The normal chain alcohols react quite rapidly
under ultrasonic irradiation. We assume that this
behavior is due to an increased mass transfer in the
presence of ultrasound. The velocity of an ultrasonic
wave through a material depends on its physical prop-
erties. Hence, the ultrasonic velocity decreases with
increasing density.
The ultrasonic properties of an emulsion vary sig-
nificantly. The droplets of more dense oil move upwards
and form a cream layer, while the alcohol moves
downwards [20], facilitating the mixing and increasing
the contact surface between alcohol and oil.
3.2. The influence of ultrasonic frequency
As could be seen in the tables, the reactions per-
formed under the ultrasonic irradiation at 40 kHz were
completed faster than under 28 kHz when a low quan-
tity of catalyst was used. When the catalyst was added in
high amount, the yields were higher for 28 kHz. Usually
there are no important differences in the formation of
the cavitation bubbles at 28 and 40 kHz, so we did not
expect any differences. However, in our case, at 40 kHz
the reaction time was shorter, but the transesterification
yields are slightly lower than at 28 kHz. The differences
Table 6
The yields of isolated n-butyl esters
Method 0.5% (wt/wt) NaOH
Time (min) Yield (%)
Mechanical stirring >60 83
Ultrasonic irradiation 28 kHz 60 99
Ultrasonic irradiation 40 kHz >60 92
371
in the product yield are mainly due to difficulties
through washing. At 40 kHz soap is formed in higher
amount and acts as phase transfer catalyst leading to
formation of esters much rapidly than at 28 kHz. But
during washing the soap hinders the separation and
some ester is trapped into the soap micelles and thus the
yield in isolated product is decreased.
Increasing the frequency the production of cavitation
in liquid decreases, thus the irradiation power has to be
increased in order to maintain the same amount of
cavitation in the system. Other frequencies (up to 100
kHz) were tested, but no remarkable results were ob-
tained. At higher frequencies the collapse of cavitation
bubbles are not strong enough to impinge one liquid to
the other. The mixing between the two immiscible layers
(alcohol and oil) is very poor hence emulsification does
not occur, and the transesterification takes place mainly
at the boundary between the two layers. We can con-
clude that only low frequency ultrasonic irradiation is
useful for transesterification reactions, at least for tri-
glycerides.
3.3. The type and quantity of catalyst
The next step was to establish the influence of catalyst
type and concentration on the yield of isolated esters.
Taking into account the economical factor the most
convenient base catalysts were sodium or potassium
hydroxides. There were no great differences in the time
to complete conversion between the two types of cata-
lyst. The use of ultrasonic irradiation proved more
proficient than mechanical stirring in the terms of iso-
lated methyl esters (Table 7).
When potassium hydroxide was used, high yields
were obtained even for 1.5% catalyst concentration.
Potassium soap is softer, more soluble in water and does
not make as much foam as sodium soap. The washing of
esters when using potassium hydroxide is easier and the
yields of isolated product are higher.
From Table 8 the same tendency of decreasing yields
for ethyl, n-propyl and n-butyl esters when catalyst is
used in high amount is also observed for potassium
hydroxide. Also, in the case of sonication at 40 kHz the
yields are high when catalyst concentration is small,
while for 28 kHz the best results are obtained for 1.5%
catalyst, except for ethanol.
1.0% (wt/wt) NaOH 1.5% (wt/wt) NaOH
Time (min) Yield (%) Time (min) Yield (%)
40 90 20 89
40 90 40 87
40 99 20 90
372 C. Stavarache et al. / Ultrasonics Sonochemistry 12 (2005) 367–372

Table 7
The yields of isolated methyl esters
Method 0.5% (wt/wt) KOH 1.0% (wt/wt) KOH 1.5% (wt/wt) KOH
Time (min) Yield (%) Time (min) Yield (%) Time (min) Yield (%)
Mechanical stirring 60 86 40 85 20 83
Ultrasonic irradiation 28 kHz 40 95 40 93 40 93
Ultrasonic irradiation 40 kHz 20 96 40 92 20 90

Table 8
The yields of isolated esters

Method 0.5% (wt/wt) KOH 1.0% (wt/wt) KOH 1.5% (wt/wt) KOH
EtOH PrOH BuOH EtOH PrOH BuOH EtOH PrOH BuOH
Mechanical stirring 55 76 91 57 60 91 44 29 97
Ultrasonic irradiation 28 kHz 81 78 92 71 68 94 49 81 98
Ultrasonic irradiation 40 kHz 55 76 91 65 55 95 38 53 98
*
The reaction time is the same as in the case of sodium hydroxide.

We can conclude that low frequency ultrasound is an
efficient, time saving and economically functional,
offering a lot of advantages over the classical procedure.
Ultrasounds can be a valuable tool for the transesteri-
fication of fatty acids, aiming to prepare the bio-diesel
fuel at industrial scale.
Acknowledgements
Vinatoru thanks the Japan Society for the Promotion
of Science (JSPS) for a long-term research fellowship at
Osaka Prefecture University.
References
[1] E.G Shay, Biomass and Bioenergy 4 (1993) 227.
[2] F. Ma, M.A. Hanna, Bioresource Technology 70 (1999) 1.
[3] Y.M. Choo, A.N. Ma, Chemistry & Industry (2000) 530.
[4] H. Noureddini, D. Harkey, V. Medikonduru, JAOCS 75 (1998)
1775.
[5] J. Metzer, Pulse of the Planet (1996).
[6] F. Ma, L.D. Clements, M.A. Hanna, Bioresource Technology 69
(1999) 289.
[7] M. Iso, B. Chen, M. Eguchi, T. Kudo, S. Shrestha, Journal of
Molecular Catalysis B 16 (2001) 53.
[8] V.D. Athawale, S.C. Rathi, M.D. Bhabhe, Separation and
Purification Technology 18 (2000) 209.
[9] S.Y. Shimada, Y. Watanabe, T. Samukawa, A. Sugihara, H.
Noda, H. Fukuda, Y. Tominaga, JAOCS 76 (1999) 789.
[10] Y. Watanabe, Y. Shimada, A. Sugihara, H. Noda, H. Fukuda, Y.
Tominaga, JAOCS 77 (2000) 355.
[11] E.W. Eckey, JAOCS 33 (1956) 575.
[12] E. Crabbe, C. Nolasco-Hipolito, G. Kobayashi, K. Sonomoto, A.
Ishizaki, Process Biochemistry 37 (2001) 65.
[13] M. Canakci, J. Van Gerpen, ASAE Transactions 42 (1999) 1203.
[14] R. Alcantara, J. Amores, L. Canoira, E. Fidalgo, M.J. Franco, A.
Navarro, Biomass and Bioenergy 18 (2000) 515.
[15] T.J. Mason, Practical Sonochemistry, Ellis Horwood Ltd, 1991.
[16] D.G. Boocock, S.K. Konar, V. Mao, H. Sidi, Biomass and
Bioenergy 11 (1996) 43.
[17] D. Darnoko, M. Cheryan, JAOCS 77 (2000) 1263.
[18] B. Freedman, R.O. Butterfield, E.H. Pryde, JAOCS 63 (1986)
1375.
[19] C. Stavarache, M. Vinatoru, R. Nishimura, Y. Maeda, Chemistry
Letters 32 (2003) 716.
[20] R. Daniels, Skin care forum online 28 (2002).