Fuel Processing Technology 102 (2012) 53–60

Contents lists available at SciVerse ScienceDirect

Fuel Processing Technology

journal homepage: www.elsevier.com/locate/fuproc

Direct production of biodiesel from rapeseed by reactive extraction/in

situ transesterification
Rabitah Zakaria a,⁎, Adam P. Harvey b
a
Department of Process and Food Engineering, Faculty of Engineering, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
b
School of Chemical Engineering and Advanced Materials, Newcastle University, NE1 7RU, UK

a r t i c l e i n f o a b s t r a c t

Article history: Biodiesel is a fuel derived from renewable resources such as edible and inedible oil-bearing seed, algae, and
Received 27 October 2011 waste cooking oil. The conventional biodiesel process involves oil extraction, refining and transesterification.
Received in revised form 21 March 2012 Alternatively, transesterification can actually be performed directly from the oil-bearing materials without
Accepted 18 April 2012
prior extraction. This route which is often termed “reactive extraction” or “in situ transesterification” has
Available online 15 May 2012
the advantages of simplifying the biodiesel production process as well as potentially reducing production
Keywords:
cost. In this study, the reactive extraction of rapeseed with methanol has been characterised. The effects of
Biodiesel process parameters on the yield, conversion and reaction rate differ substantially from conventional trans-
Rapeseed esterification due to the dependence on both extraction and reaction. The rate of ester formation is mainly
Canola affected by the catalyst concentration, temperature and particle size while the equilibrium yield largely de-
Reactive extraction pends on the solvent to oil molar ratio. A high yield of ester (>85%) can only be achieved at high solvent
In situ transesterification to oil molar ratios (> 475:1). Parametric studies and light microscope images of reactively extracted seed
suggested that reactive extraction occurs by transesterification of the oil inside the seed, followed by diffu-
sion of the products into the bulk solvent.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction
The exploration of new energy resources has become increasingly im-
portant in recent years due to the dwindling fossil fuel reserve and the
adverse environmental effects of traditional fuel combustion [1]. Conse-
quently, the development of sustainable and “greener” energy resources
has taken centre stage. One of the alternatives of renewable energy is
fuel derived from plant and animal matter. These so-called “biofuels”
can be produced by various processes, such as fermentation, trans-
esterification, pyrolysis, gasification, liquefaction, and hydrotreatment.
Among these processes, transesterification of vegetable oil or animal fat
to produce biodiesel has been successfully commercialised. Biodiesel is a
petroleum diesel substitute and can be readily used in most diesel engines
without any modification. Furthermore, engine performance using bio-
diesel has been shown to be comparable to that of conventional diesel
fuel [2]. Other merits of biodiesel include reduced emission of carbon
monoxides, particulates and hydrocarbon from the engine [2] and en-
hanced engine lubrication [3]. There are abundant types of resources
such as inedible oilseed (i.e. Jatropha, algae, and pongamia) or waste bio-
mass (i.e. waste cooking oil, animal fats, sewage sludges) that can be
utilised to produce biodiesel, directly contributing to the efficient use of
biomass and land resources.
⁎ Corresponding author. Tel.: + 60 389464301; fax: + 60 389464440.
E-mail addresses: rabitah@eng.upm.edu.my (R. Zakaria), adam.harvey@ncl.ac.uk
(A.P. Harvey).
The conventional method for production of biodiesel requires the oil
to be extracted from the biomass before it can be transesterified into
ester. The transesterification reaction occurs in a liquid state. The oil ex-
traction process typically involves the use of a solvent (usually hexane)
in a counter current or percolation extractor, or a mechanical method
such as screw press extractor. However, solvent extraction plant is ex-
pensive, complex and poses health and safety hazard due to the han-
dling of flammable and explosive solvent [4]. On the other hand,
mechanical pressing often yields less oil than solvent extraction
resulting in a high residual oil in the waste biomass. Alternatively, the
need to extract the oil from the biomass prior to transesterification
can be eliminated using a “reactive extraction” method to produce bio-
diesel [5]. In this method, the oil is simultaneously extracted and
transesterified into alkyl ester in -situ in one single process. Hence, sep-
arate extraction and oil refining steps can be eliminated since extraction
is simultaneously carried out by the reactant itself. The use of solid raw
material rather than refined oil could certainly reduce the feedstock
cost. Since the method directly uses oil-bearing materials rather than
pre-extracted oil, it eliminates processes such as crushing, solvent ex-
traction and degumming, and perhaps drying, which are key processes
in conventional biodiesel production. Furthermore, the elimination of
hexane from the overall process provides an added health and environ-
mental benefit as hexane is classified as a hazardous air pollutant and
contributes to the production of smog and global warming [6].
Previous studies have demonstrated that biodiesel production via re-
active extraction (also referred to as in situ transesterification) is capable

0378-3820/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.fuproc.2012.04.026
54 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60
of producing high yields of fatty acid esters from a variety of oil-bearing
raw materials [7–9]. Increase in extraction yield is also possible depending
on the type of oil in the biomass as certain oil contains high amount of
polar components (i.e. phospholipids), which can be extracted by the al-
cohol during reactive extraction, whereas they are non-extractable by
conventional methods [8]. The application of this process to substances
with low oil content, such as distillers dried grains with soluble (DDGS)
and meat and bone meal (MBM) seems to offer economic advantages
due to the higher wastage associated with a conventional two-step pro-
cesses [9]. Furthermore, some plant materials, such as algae, are difficult
to crush: preparing the biodiesel via reactive extraction may overcome
this limitation [10].
Reactive extraction substantially differs from the conventional bio-
diesel process in many ways. Notably, the system is heterogeneous, in-
volving biological material with an array of biological structures and
components. As pointed out by Poirot et al. [11], extraction efficiency
is influenced by the microstructure of the vegetable seed matrix, the
molecular structure and size of the solute, localization and link with
other components. Hence, the efficiency of reactive extraction could
be very different for different biological materials. This is evident in
the differing optimum parameters found in several studies [9,12,13]. Al-
though the method has been shown to be viable by early researchers,
considerable effort needs to be put into fully defining the characteristics
of the process. Most studies reported the yield of ester, but to our
knowledge none provided time profiles of the reaction intermediates
(mono- and diglycerides) as well as triglyceride. This information is es-
sential in order to identify the substance extracted and to fully describe
the process kinetics. This work aims to investigate the factors affecting
reactive extraction of oil- bearing biomass to produce biodiesel. The pa-
rameters that can influence the reactive extraction process were stud-
ied in considerable depth in order to understand their effects on the
equilibrium yield and the kinetics. In depth study of the characteristics
of the process may enable us to understand the process mechanism,
which is critical in predicting the effects of various process parameters
and in providing insight into possible further improvements. The oil-
seed chosen was rapeseed (or “canola”) as it is the biodiesel feedstock
in Europe.
2. Material and methods
2.1. Reagents and material
The rapeseed was kindly provided by a local farm near Newcastle
upon Tyne, UK. The hexane used was of 99.9% purity. The methanol and
ethanol used throughout this study were of 99.97% and 99.95% purity re-
spectively. All solvents were purchased from Fisher Scientific. Methyl
heptadeacanoate and tricaprin internal standards were obtained from
Sigma-Aldrich.
2.2. Oil extraction
The oil seed was first extracted using 4 different solvents namely
methanol, ethanol, hexane and a 50:50 mixture of ethanol and metha-
nol. 20 g of the seed were pulverised using a coffee grinder for 1 min.
The ground seed were transferred into a cylindrical cup, covered with
glass wool, and placed in the Soxhlet extraction unit. 230 ml of solvent
was placed in the round bottom flask and heated. The Soxhlet unit was
allowed to run for a pre-determined time i.e. 1, 3, 6, 8, and 12 h until
constant oil yield has been reached. Solvent was removed from the
extracted oil using a vacuum rotary evaporator.
2.3. Reactive extraction
A known amount of ground and sieved seed was placed in a 500 mL
round-bottom flask equipped with a condenser and an overhead stirrer
and placed in a constant temperature water bath. Alkaline alcohol,
prepared by dissolving a known amount of sodium hydroxide in meth-
anol, was first heated to the desired temperature using a heated circu-
lating water bath. Once heated, the alcohol was transferred to the
round-bottom flask and the reaction was carried out at the desired tem-
perature and reaction periods. The stirrer speed was maintained at
200 rpm for all of the experiment. It was important to ensure that
there was no escape of methanol from the flask as losses of methanol
to the atmosphere would distort the outcome of the parametric study.
After operation for the desired time, a specific amount of glacial acetic
acid was added to the mixture to neutralize the catalyst and prevent fur-
ther reaction. The liquid was then separated from the seed using vacuum
filtration. The solid residue was washed repeatedly with methanol to re-
cover any product that adhered to the seed and the excess methanol was
removed using a rotary evaporator. After evaporation, two layers of liquid
were formed. The upper layer contained the ester phase, while the bot-
tom layer contained the glycerol phase. The layers were separated using
a separating funnel. The masses of the phases were recorded and the
ester content was analysed using a gas chromatograph.
The yield of ester extracted after the reaction completion was calcu-
lated as:
Esteryieldðwt:% Þ ¼ mass of ester phase  0:995  100
ð1Þ
mass of triglycerides in the seed  100
where the factor 0.995 is to correct for the molecular weight difference
between the triglyceride and ester. The concentration of mono-, di-
and triglycerides were quantified using a Gas Chromatograph/Mass
spectrometer.
2.4. Progress of reaction
To obtain the time profile of ester and the reaction intermediates,
0.5 mL samples of the reaction mixture were taken at various time in-
tervals using a syringe and a tube from the sampling pots. 3 μL of glacial
acetic acid was added to stop the reaction. A syringe filter was used to
separate the liquid from the rest of the seed particles. A small portion
of the liquid, which contains methanol and the reaction products, was
analysed using a gas chromatograph in order to quantify the amount
of ester in the bulk methanol phase. The rest of the liquid was evaporat-
ed to remove the methanol until the sample reached constant weight.
The product was then allowed to settle into two layers and the ester
phase was withdrawn from the top and the concentration of mono-,
di- and triglycerides were quantified using a gas chromatograph/mass
spectrometer (GC-MS).
2.5. Effect of water
To study the effect of seed moisture, reactive extraction was con-
ducted using dried seed. Seeds were dried in an oven at 104 °C until
they achieved constant weight. The yield of dry and wet seed was com-
pared by performing the reactive extraction experiment using 25 g of
seed, 250 ml methanol, 0.1 m catalyst concentrations and at 60 °C. The
experiments were carried out in duplicate. The effect of water in meth-
anol was also studied using methanol containing 1, 2 and 4 wt.% water.
The experiment was performed using 25 g of both dried and wet seed,
100 ml methanol, and 0.1 m catalyst concentrations and at 60 °C. The
water concentration in the bulk methanol phase during reactive extrac-
tion was also analysed in duplicate using the Karl Fisher Volumetric
Analysis Method (Metrohm 701 KT Titrino). The sample (about 0.2 g)
was taken from the bulk methanol phase using a clean, dry syringe.
2.6. Ester analysis
Gas chromatography analysis for the determination of individual
and total esters was based on the British Standard BS EN 14103:2003.
The internal standard used was methyl heptadecanoate (MHDN). The
 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60
stock solution of internal standard was prepared by adding 5.00 g of
methyl heptadecanoate to 322.6 g of analytical grade heptane. About
0.625 g of sample and 1.25 g of the stock solution was weighed accu-
rately to 4 decimal places on an AND HR-200 analytical balance. The
GC used was a HP 5890 Series II, with a Varian CP wax column with di-
mensions of 0.53 mm i.d, 10 μm coating thickness and 25 m length. The
carrier gas was helium at a flow rate of 2 mL/min and the column, injec-
tor and detector temperature was maintained at 220 °C. The column
separated the esters and the methyl heptadecanoate (MHDN) into indi-
vidual peaks.
To analyze the ester concentration in the bulk methanol phase
obtained at various reaction times during the experiment, a modified
GC method was employed. The stock solution of the internal standards
was prepared by adding 5 g of methyl heptadecanoate in 322.6 g meth-
anol instead of hexane. A known amount of sample (1–2 g range) was
weighed accurately into a sample vial with 1.25 g of the stock solution.
The injection volume used was 0.2 μl. The gas chromatograph tempera-
ture program used was the same as described above.
2.7. Gas chromatography/mass spectroscopy (GCMS) method for glycer-
ide analysis
A GCMS was used to quantify the monoglycerides, diglycerides and
triglycerides in the ester phase. The method was based on BS EN 14105
with modification to the calibration procedure in order to account for
the higher glyceride contents, which occurred early in the reaction.
Tricaprin (1,2,3-tricaproylglycerol) were used as a standard to calibrate
the glycerides.
For the analysis of monoglyceride, about 20 mg of sample was accu-
rately weighed and placed in a sample vial. 20 μl of tricaprin stock solu-
tion (12.6 mg/mL) and 20 μL of MSTFA was added to the sample vial.
The solution was left to silylate for 15 min after which 2 mL of heptane
was added. For identification of TG and DG about 20 mg of sample was
accurately weighed and place in a sample vial. 100 μL of tricaprin stock
solution (0.5 mg/ml) and 20 μL of MSTFA were added to the sample vial.
The solution was left to silylate for 15 min after which 0.5 mL of heptane
was added. The GCMS was fitted with Perkin Elmer Col-elit column (PE-
5HT) of 15 m length, 0.25 i.d and 0.1 μm film thickness. The detector
type in the mass spectrometer was electron impact positive. The flow
rate of helium was 1 mL/min. The oven temperature protocol was
50 °C hold for 1 min, heat to 180 °C at 15 °C/min, then at 7 °C/min to
230 °C and at 10 °C/min to 370 °C. The temperature was then held at
370 °C for 10 min for a total run time of 31.5 min. The inlet line to the
MS was kept at 270 °C while the MS source temperature was kept at
250 °C. The standard used to calibrate the MS was 1.5,7 Triazabicyclo
(4,4,0) dec-5‐ene purchased from Fluka.
2.8. Light microscope Image
For light microscope observation, the rapeseed particles were rapid-
ly fixed in 1% (wt:vol) osmium tetroxide and then transferred to 2.5%
(vol) glutaraldhyde in 0.1 M sodium cacodylate buffer (pH 7.2) and
left overnight. Subsequently, all the samples were washed twice, each
time for 30 min, in 0.1 M sodium cacodylate or phosphate buffer and
then post-fixed in 1% osmium tetroxide for 2 h. Samples were
dehydrated in graded ethanol serial dilutions for 30 min for each solu-
tion and then finally in 100% ethanol for 30 min 3 times. All samples
were impregnated with Spurr resin and embedded in moulds and poly-
merized at 60 °C. Sections were cut on Reichert Ultracut ultramicro-
tome (Lecia Microsystems Ltd, Milton Keynes, UK), mounted on glass
slides. These procedures were performed by the Newcastle University
Electron Microscopy Research Services. Prior to microscopic examina-
tion, the slide containing the fixed sample was immersed in Sudan
black B for 5 min to stain the lipids. The reagents were then removed
by washing with 90% ethanol and air-dried. Sudan black B and periodic
acid-Schiff reagent was purchased from Sigma Aldrich. The stained
55
samples were examined using an Olympus BX41 light microscope and
digital images were collected using an Altra20 Soft Imaging System.
3.0. Results and discussions
3.1. Extraction with different solvents
The mass of extract from Soxhlet extraction of 20 g of seed for 6 h,
using methanol, ethanol, 50/50 mixtures of ethanol/methanol and
hexane yield different amount of oil can be seen in Table 1.
Methanol extracts some amount of material from the seed but very
little is triglyceride (TG). This was similar to the findings of several pre-
vious studies [14–16]. For example, Ozgul and Turkay [14] found that
methanol dissolves free fatty acid from rice bran oil leaving triglycerides
in the bran. The poor triglyceride solubility in methanol is as expected,
since methanol is a very polar solvent, whereas most triglycerides are
non-polar long chain hydrocarbon molecules. However, various other
compounds in the seed can potentially dissolve in methanol, e.g. phos-
pholipids, sterols, phenols, and vitamins. Ethanol, on the other hand, is a
much better solvent for triglycerides, although it extracts other polar
materials as well. The ability of ethanol to extract considerable amount
of oil has been well-documented [17–19]. A mixture of methanol and
ethanol extracts some oil, but not as much as when using ethanol
alone. Despite the poor solubility of triglycerides in methanol, its use
in simultaneous extraction and transesterification results in extraction
of most of the oil and transesterification into ester [14–16]. The degree
of extraction, however, depends on the process parameters, which is
the subject of the following sections.
3.2. Methanol to oil molar ratio
As can be seen in Fig. 1a and b, increasing the methanol to oil molar
ratio increases the equilibrium ester yield. The yield of ester (as defined
by Eq. (1)) improves greatly as the methanol amount is increased from
0 to 300 molar ratio, but after that, increasing the methanol volume in-
creases the yield only slightly until a maximum of about 82% of the tri-
glycerides content in the seed is reached at about 600 molar ratio.
However, increasing the methanol ratio to more than 900:1 seems to
cause a reduction in the yield of ester. At these molar ratios, the
resulting product consists of a relatively smaller ester phase but a
much larger glycerol phase. When the glycerol phase was extracted
with hexane, it was found that it does contain a significant amount of
ester and the total ester yield is similar to that of the lower molar
ratio. Hence, at higher methanol to oil molar ratio, it appears that the
full separation of ester and glycerol phase becomes increasingly diffi-
cult. It should be noted that at high methanol to oil molar ratio, there
is also a high amount of catalyst in the solvent since the catalyst concen-
tration in the solvent is kept constant. Therefore, a high amount of cat-
alyst may have increased the solubility of the ester in the glycerol phase
due to increasing amount of water that is unavoidably produced upon
dissolution of the catalyst. The time to reach final yield was similar for
high and low methanol to oil molar ratio as shown in Fig. 1b, indicating
that the rate of extraction was not significantly affected by the solvent
amount. Similar finding was observed by the study of Mondala et al.
[20] on in situ transesterification of municipal sludges, who found that
methanol to oil molar ratio had a weak effect on the overall extraction
rate.
Table 1
Soxhlet extraction for 20 g of rapeseed for 6 h using various solvents.
Methanol Ethanol 50/50 ethanol/methanol Hexane
Total extract 3.12 10.21 6.32 9.14
Triglycerides phase 0.10 7.44 1.60 9.14
Other non-triglyceride 3.02 2.77 4.72 0.00
material
56 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60

a increasing ethanol [23]. In another study, the extraction of oil from

100 Jatropha seed in a batch reactor was demonstrated to increase with
90 increasing hexane [24].
80 The high methanol to oil molar ratio requirement has been observed
70 in most studies for reactive extraction involving different oil-bearing
Yield ( wt %)

60 materials [25]. For example Haas et al. [12] found that 226:1 molar
50 ratio is needed to reactively extract soybeans with methanol. A 673:1
40 molar ratio is required to obtain maximum yield for cottonseed [26]
30 while 476:1 is required for sunflower [27]. The high energy required
20 for methanol recovery causes the production cost of biodiesel by this
10 method to be higher than for the conventional method [28].
0 Fig. 2 shows the concentration of ester in the ester phase versus
0 200 400 600 800 1000 1200 1400 methanol to oil molar ratio. The purity of the ester phase indicates the
Molar Ratio extent of conversion of the extracted material into ester in the bulk
ester hexane extracted ester methanol. The data shows that the conversion to ester is similar at
b low or high molar ratio. Hence even at low molar ratio (e.g. at 200:1),
100 the amount of methanol is adequate to convert the extracted material
90 into fatty acid methyl ester. This shows that the low equilibrium ester
80 yield at lower molar ratio is not the result of incomplete conversion of
Yield (wt%)

70
60
50
40
30
20
10
0 In conventional transesterification, the amount of catalyst is usually
0 50 100 150
Time (min)
270 Molar Ratio 475 Molar Ratio 150 Molar Ratio
Fig. 1. a: Effect of methanol to oil molar ratio on the ester yield for 1 h reaction. Reac-
tion conditions: 60 °C, 0.1 m catalyst, 300–1000 μm particle size, 25 g seed. b: Effect of
methanol to oil molar ratio on equilibrium yield. Reaction conditions: 60 °C, 0.1 m cat-
alyst, 300–500 μm particle size, 25 g seed.
The minimum amount of methanol needed for in situ trans-
esterification using batch type reaction is presumably the amount
that is needed to simply cover the seed and to render the solid
phase workable for agitation. For 25 g of seed used in the study, the
amount of methanol needed to adequately cover the seed is 40 ml
which is a methanol to oil ratio of 76:1 for the rapeseed used. This is
much higher than the optimum required for pre-extracted oil trans-
esterification of 6:1 (as suggested by Freedman et al. [21]). The amount
of catalyst used here is 1.1 wt.% of the oil, which is within the range
suggested for conventional transesterification of 0.5 to 1.0 wt.% [22].
Hence these amounts should be able to provide adequate reactants
and catalyst for a significant conversion to be achieved. However, this
is not the case for in situ reaction, as using a methanol to oil ratio of
95:1 and below does not result in any significant biodiesel being
produced.
triglyceride to ester in the bulk methanol phase, but is probably due
to the low concentration gradient between the seed and the bulk liquid
as explained earlier.
3.3. Catalyst concentration
200
0.5 to 1.0 wt.% of the oil. Higher concentrations cause excessive saponifi-
cation [29], while lower concentrations cause low conversion of triglycer-
ides to biodiesel [22]. However, this range may not be applicable to in situ
transesterification due to the physical difference in the reaction phase,
such as the presence of solid and the vast amount of solvent used. The ef-
fect of catalyst concentration ranging from 0.03 m (equivalent to 2.1 wt.%
oil) to 0.1 m is shown in Fig. 3, below. It can be seen that increasing the
catalyst concentration increases both the final ester yield and the rate of
ester formation. Hence, an adequate amount of catalyst is essential in en-
suring complete transesterification of the oil which consequently allows
for the extraction of triglycerides to take place. Increasing the amount of
catalyst causes higher rates of ester extraction, which indicates that the
rate of reaction controls the reactive extraction process.
Fig. 4 shows the interactive effect of molar ratio and catalyst concen-
tration. At a catalyst concentration of 0.05 m, the maximum equilibrium
yield is achieved at a 1200:1 methanol to oil molar ratio whereas for
0.1 m, a similar maximum equilibrium yield is achieved at a 600:1
methanol to oil molar ratio. Increasing the catalyst concentration fur-
ther to 0.15 m produces a similar amount of extracted materials but
the ester yield is lower than at 0.1 m. This is probably due to the adverse
effects of the saponification reaction or dissolution of methyl ester in
Ester concentration in ester phase (wt %)

The fact that a large amount of solvent is needed in in situ trans-

esterification is probably due to the amount required to drive the ex- 100
traction rather than the reaction. Solvent extraction of oil from seed 95
can be explained partly by Fick's Law of diffusion, in which the yield
90
of extraction depends on the concentration gradient between the
bulk liquid and the internal seed, as well as the oil solubility in the 85
solvent. Early in the reaction, the yield increases rapidly due to the 80
high concentration gradient between the inside of the seed and
the bulk liquid. As more products are extracted, the concentration 75
gradient starts to decrease, thereby decreasing the extraction rate. 70
Extraction will stop once the concentration inside the seed is in equi-
65
librium with that in the bulk liquid. Therefore, the higher the amount
of solvent used, the more dilute the bulk liquid will be, so the higher 60
0 200 400 600 800 1000 1200
the concentration gradient, which will allow greater amounts of oil to
Molar Ratio
be extracted. This effect has been observed in several oilseed extrac-
tion studies with various solvents. In the extraction of oil with ethanol Fig. 2. Concentration of ester in the ester phase at different molar ratios. Reaction con-
for example it was found that the equilibrium yield increased with ditions: 1 h, 60 °C, 0.1 m catalyst, 300–1000 μm particle size, 25 g seed.
 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60 57

100 Table 2
90 Ester yield and purity at different catalyst concentrations. Reaction conditions: 1 h,
Ester yield (wt%)

80
70
60
50
40
30
20
10
0
0 100 200 300
Time (min)
0.03 molal 0.05 molal 0.1 molal
Fig. 3. Effect of catalyst concentration on the ester yield. Reaction conditions: 1 h, 60 °C,
475:1 methanol to oil molar ratio, 300–500 μm particle size, 25 g seed.
the glycerol phase associated with excess catalyst. Clearly, increasing
the catalyst concentration reduces the methanol requirement to
achieve maximum yield. However, at 0.05 catalyst concentration and
1200:1 molar ratio, the catalyst to oil ratio is the same as at 0.1 m and
600:1 molar ratio, which explains how at lower catalyst concentration
a similar yield can be achieved by increasing the solvent amount. There-
fore, it appears that it is the catalyst to oil ratio rather than the catalyst
concentration which dictates the amount of triglyceride extracted from
the seed. Low catalyst to oil ratio causes low conversion of the triglycer-
ides to ester and hence low extraction yield. On the contrary, too much
catalyst will reduce the yield presumably due to the excessive saponifi-
cation reaction. The optimum amount of catalyst appears to be in the
region of 7 wt.% of the oil, which is higher than that required for con-
ventional transesterification. Adequate methanol to molar ratio on the
other hand is still important in order to provide the concentration gra-
dient for extraction and even with enough catalyst the yield will still be
low if the methanol to oil molar ratio is low.
Table 2 shows the concentration of ester and the equilibrium yield at
0.1 and 0.05 m catalyst concentration. The data shows even though the
ester yield is lower at 0.05 m, similar concentration of ester is achieved
for both catalyst concentration. This indicates that even at the low cat-
alyst concentration sufficient catalyst is available in the bulk phase to
convert the extracted material into ester. Therefore, it could be inferred
that the lower ester yield obtained with the lower catalyst concentra-
tion is not the result of incomplete reaction in the bulk phase. Converse-
ly, it is possible that there is insufficient conversion inside the seed
which prevents further extraction of the triglycerides.
18
16
14
12
Mass (g)
60 °C, 475:1 methanol to oil molar ratio, 300–500 μm particle size, 25 g seed.
Catalyst concentration (m) Ester concentration (wt.%) Ester yield (%)
0.1 90.3 ± 1.3 88.8 ± 0.1
0.05 89.7 ± 1.4 59.5 ± 0.8
3.4. Particle size
400
The effect of particle size on rate of reactive extraction can be seen
in Fig. 5.
For particle sizes between 300 and 500 μm, the time to reach maxi-
mum ester yield was about 1 hour but it increased to more than 3 h
when a particle size of 1000 to 1400 μm were used. Despite the longer
extraction time, the final yield of the larger particles was essentially
the same as that of the smaller particles. Reducing the particle size
shortens the diffusion path of the oil and solvent, thereby increasing
the extraction rate. It is also possible that the more extensive mechani-
cal grinding needed to reduce the particle size may also cause a higher
degree of distortion of the cell wall causing an increase in the diffusion
coefficient and faster extraction. Although there are several mathemat-
ical formulae in the literature that relate extraction time to particle size
(see: [30,31], the extent of improvement in extraction rate may vary be-
tween particles of the same size since different seed preparation
methods can result in different degrees of cell wall disruption.
The smaller the particle size, the greater the proportion of broken
surface cell walls releasing more oil at the particle surface. The oil at
the surface is not, of course, released to the bulk solvent by internal dif-
fusion, rather it is solubilised or washed off at a much quicker rate. It is
postulated that if most of the oil accumulates at the particle surface, less
solvent is needed for extraction, since the diffusion process is mini-
mized. Comparing the ester yield at different methanol to oil molar ra-
tios for a low particle size of 300 to 500 μm and a particle size of 1000 to
1400 μm, it can be seen from Fig. 6 that similar amounts of solvent are
needed to achieve maximum yield. This indicates that for particle
sizes of 300–500 μm, the extraction of the oil still depends largely on
diffusion rather than solubilisation.
3.5. Temperature
In general, temperature affects both the rate of diffusion and reac-
tion. Examination of Fig. 7 shows that increasing the reaction tempera-
ture from 30 to 60 °C increases the rate of ester formation particularly
early in the reaction, but the final equilibrium yield is the same. A
number of researchers have shown that transesterification can be
adequately carried out at temperatures as low as 20 °C without a
great deal of compromise on the rate and conversion [22,29]. Oil extrac-
tion, on the other hand, proceeds faster at higher temperatures, as the
viscosity of the oil and the solvent is reduced and diffusivity is increased.
90

10 80
Ester yield (wt%)

8 70
60
6
50
4 40
2 30
0 20
0 200 400 600 800 1000 1200 1400 1600 10
Methanol to oil molar ratio 0
0 50 100 150 200 250 300 350
ester yield 0.1 molal ester yield 0.05 molal Time (min)
ester yield 0.15 molal total extract 0.1molal
total extract 0.05 molal total extract 0.15 molal 300-500 micron 800-1000 micron 1000-1400 micron

Fig. 4. Effect of catalyst concentration at different molar ratio on ester yield and total Fig. 5. Effect of particle size at 571:1 methanol to oil molar ratio on ester yield. Reaction
extract. Reaction conditions: 1 h, 60 °C, 300–1000 μm particle size, 25 g seed. conditions: 1 h, 60 °C, 25 g seed, 0.1 m catalyst concentration.
58 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60

100 90
90 80

Ester yield (wt %)

80 70
Ester yield (wt %)

70 60
60 50
50 40
40 30
30 20
10
20
0
10 190 380 571 666
0 Molar Ratio ( Methanol to Oil)
571 380 285 190
Wet Seed Dry Seed ( 0 % Moisture)
Molar Ratio
300-500 micron 1400-1000 micron
Fig. 8. Effect of drying on the yield of ester at different methanol to oil molar ratios. Re-
action conditions: 1 h, 25 g seed, 60 °C, 0.1 m catalyst concentration, 300–500 μm par-
Fig. 6. Effect of particle size on final ester yield at different molar ratios. Reaction con-
ticle size.
ditions: 1 h, 60 °C, 25 g seed, 0.1 m catalyst concentration.

The solubility of the oil also increases at high temperature, which conse-
quently increases the final extraction yield, notably at low solvent to oil
ratio [32]. However, in reactive extraction, the increase in rate is only
significant during the early part of the extraction, as a similar time is
needed to achieve maximum yield for lower and higher temperatures.
Other works in in situ transesterification have also shown that minor
improvement is gained from increasing the temperature in the range
of 30 to 65 °C [12,13]. The extraction and subsequent transesterification
can be efficiently performed at low temperature, which is favourable for
the process economics.
3.6. Drying
As discussed previously, the presence of water can be detrimental to
the transesterification reaction as it causes the formation of hydroxide
ions, which induce saponification of free fatty acids, triglycerides and es-
ters. With respect to extraction, water might further reduce the solubility
of triglycerides in methanol due to its polarity. As the rapeseed contains
6.7% water, the effect of drying the seed to 0% water was studied
(Fig. 8). It can be seen that drying the seed does not increase the ester
yield, nor does it reduce the amount of methanol needed to achieve max-
imum yield. T-test analysis (SPSS 17 Statistical Software) gave a p value of
0.715, which is greater than 0.05, indicating that the difference between
yield obtained from dried and wet seed is not significant. This is in con-
trast to a study by Haas and Scott [33] who found that drying soybeans
to 0% water substantially reduced the amount of methanol needed to
achieve maximum ester yield by 60%. However, soybeans have a lower
oil content of about 22 wt.% and hence for the same amount of soybeans
the amount of methanol used is lower than that for rapeseed. The higher
seed moisture in less methanol will cause a higher concentration of water
in the bulk solvent, which might make the process apparently less toler-
ant to water than in situ transesterification using rapeseed.
Water analysis of the bulk methanol phase (25 g of rapeseed and
250 mL methanol) shows that for the wet seed, the water content of
methanol increases quickly to about 1.3 wt.% after 1 min reaction
and the water level stays practically constant thereafter . The fresh al-
kaline methanol has about 0.6 wt.% water initially, since the methanol
is only 99.9% pure and the dissolution of sodium hydroxide in meth-
anol releases water. For dry seed, the amount of water in the bulk
phase stays at 0.6 wt.% and throughout the reaction.
In order to identify the level of water that affects the reactive extrac-
tion, the reaction was carried out with 25 g seed and 100 mL methanol
that contains 0 to 4 wt.% water. It can be seen from Fig. 9 that when
using completely dried seed, the ester yield was not affected by water
until more than 2 wt.% water is used. This amount of water is equivalent
to 6.3 wt.% of the seed weight, which is approximately the amount of
water that is in the wet rapeseed. Apparently 6.7 wt.% of water in the
seed is not high enough to cause a detrimental effect on the process. If
more than 6.7 wt.% of water in the seed is used, as in the case of meth-
anol with 4% water for dry seed and 1 wt.% for wet seed, the yield of
ester substantially decreases. Hence, as suggested by Haas, a higher
amount of methanol is needed in order to dilute the water concentra-
tion and to achieve similar yield as the dry seed [32].
3.7. Discussion of the process mechanism
It has been shown that methanol was not able to extract triglyceride
from rapeseed to any considerable extent. However the presence of the
alkaline catalyst allows significant production of methyl ester from tri-
glycerides. The mechanism of this process is still not clear. Several stud-
ies [12,13,16] have speculated on the mechanism although they have
not supported their theories with experimental result nor discussed it

100 70
90 60
Ester Yield (wt%)

Ester yield (wt%)

80
70 50
60
40
50
40 30
30
20 20
10 10
0
0 20 40 60 80 100 120 140 0
0 1 2 4
Time (min)
Percent water in methanol (wt%)
30 deg C 40 deg C 60 deg C
Wet seed Dry seed
Fig. 7. Effect of reaction temperature on ester yield. Reaction conditions: 1 h, 25 g seed,
0.1 m catalyst concentration, 475:1 methanol to oil molar ratio, 300–500 μm particle Fig. 9. Effect of water in methanol on the yield of ester at 270:1 methanol to oil molar
size. ratio.
 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60
in details. One possible mechanism is that the initial presence of esters
causes the remaining triglycerides in the oilseed to be more soluble in
the methanol ester mixture. Kildiran et al. [16] suggested the extraction
of triglycerides is possible due to the stepwise dissolution and subse-
quent transesterification of the oil. Haas et al. [12] and Qian et al. [13]
suggested that the alkaline catalyst could destroy the cell walls and in-
tracellular compartmentalization resulting in cellular solubilisation and
subsequent transesterification of triglycerides. However, our light mi-
croscope images of the seed particle stained with Sudan Black B after re-
active extraction (Fig. 10) indicate otherwise [34]. Fig. 10a shows that
the fresh seed consists of cells of about 10–30 μm in diameter and the
cells were surrounded by cell wall membranes 0.5 to 2 μm thick. The
lipid which can be seen as a blue colour was located inside the cells. It
can be seen that most of the lipids have been removed from the cells
after reactive extraction (Fig. 10b). The image also shows that the cells
after reactive extraction consist of a significant amount of fully intact
cells confirming that the addition of sodium hydroxide into the solvent
does not destroy the cell wall structure. The mechanism of extraction
would still require the diffusion of some fraction of the oil or ester out-
ward through the cell membrane and into the bulk liquid. Fig. 10c
shows the image of particle whereby the oil has not been fully extracted
as shown by the blue colour inside the cells. The remaining oil seems to
occur only in the middle of the particle rather than throughout the seed.
This indicates that the lipid does not move radially outward into the
bulk liquid. Rather, it suggests that there was a reaction front that
consumes the lipid similar to a shrinking core process. Hence, the reac-
tion step involves the diffusion of methanol through the plant cell wall
and reaction with the intact oil bodies. Once reacted, the partial glycer-
ides (i.e. mono- and diglyceride) will dissolve in methanol and further
reacted with methanol inside the seed to produce ester. The appearance
of mono-and diglyceride in the bulk solvent depends on the reaction
and diffusion rate,
a b
c
Fig. 10. Light micropscope images stained with Sudan Black B. a) typical seed particle prior to extraction and b) completely extracted seed particle and c) partialy extracted seed
particle after extraction.
59
In order to have a better insight into the process mechanism, the con-
centration of mono-, di-, and triglycerides in the ester phase throughout
the reaction was observed (Fig. 11). As can be seen, there is only a maxi-
mum of about 6 wt.% of mono-, di and triglycerides in total in the ester
phase, which occurs very early in the reaction (1 min reaction time).
During the reactive extraction period, the extracted material is predomi-
nantly ester. It appears that triglycerides and the reaction intermediates
(mono- and diglycerides) do not accumulate to a significant degree in
the bulk solvent. Since the extracted materials are predominantly ester,
it could be inferred that either i) triglycerides are rapidly reacted to
ester once extracted into the bulk solvent or ii) the transesterification re-
action occurred inside the seed and the resulting ester was extracted into
the bulk solvent . The latter mechanism is more likely based on the image
seen in Fig. 10c earlier.
This mechanism also fits well with the behaviour of the process as
observed from the parametric studies. As shown in Section 3.2, the
yield of ester increases with an increase in the amount of methanol
used. If the reaction occurs first in the seed, the ester will diffuse into
the bulk solvent due to the concentration gradient between the ester
in the bulk liquid and in the seed. Increasing the amount of solvent
will cause the concentration gradient to increase, thereby increasing
the yield of ester. On the other hand, if triglycerides were extracted
and reacted immediately as they reach the bulk liquid, the concentra-
tion of triglycerides in the bulk liquid will always be zero or near it.
Therefore, there would always be a maximum triglyceride concentra-
tion gradient inside the seed and the bulk liquid, which does not explain
why higher solvent (methanol) is needed for higher yield. Also, if it is
assumed that triglycerides were extracted since it becomes more solu-
ble in the mixture of ester and methanol, the rate of extraction should
be slower in the beginning of the reaction and should become more
rapid when more ester is produced as the reaction progresses. In
contrast, the rate of extraction is very fast early in the reaction and
100 µm
60 R. Zakaria, A.P. Harvey / Fuel Processing Technology 102 (2012) 53–60
3.0
Concentration (wt %)
2.5
2.0
monoglyceides
1.5 diglycerides
triglycerides
1.0
0.5
0.0
0 10 20 30 40 50 60 70
Time (min)
Fig. 11. Concentration time profile of mono-, di-, and triglycerides in the ester phase.
decreases as the reaction progresses. Furthermore, from Table 2 it was
shown that the ester yield is low at low catalyst concentration, but the
ester concentration in the bulk methanol is similar at higher and
lower concentrations. Hence, inadequate conversion of triglycerides
in the bulk methanol phase is not the cause of the lower extraction
yield. On the other hand, it is likely that at low catalyst concentration,
there is insufficient conversion of triglycerides to ester inside the seed,
which reduces the yield of ester.
4. Conclusions
The maximum yield of reactive extraction of rapeseeds using meth-
anol is determined primarily by the methanol to oil molar ratio. Metha-
nol to oil molar ratios greater than 400:1 were required to obtain ester
yields higher than 80%. This is a much higher amount of methanol than
that needed for conventional transesterification. In order to make the
process economically viable, more research is needed to identify tech-
niques that can reduce the solvent consumption.
The rate of the process, on the other hand, is mainly controlled by
the catalyst concentration and the seed particle size. Decreasing the cat-
alyst concentration and increasing the particle size reduces the overall
rate of extraction. Temperatures within the range of 30 to 60 °C only af-
fect the initial rate of extraction, but the time to reach the maximum
yield is similar. The maximum yield of ester was achieved at a catalyst
concentration of 0.1 mol/kg and 670:1 methanol to oil molar ratio. It ap-
pears that the process is not significantly adversely affected by water
when the seed moisture content is less than about 6.7 wt.%. Hence,
the drying step could be removed if the moisture content does not ex-
ceed this level, which should improve the process economics.
References
[1] S.N. Naik, V.V. Goud, P.K. Rout, A.K. Dalai, Production of first and second genera-
tion biofuels: a comprehensive review, Renewable and Sustainable Energy Re-
views 14 (2009) 578–597.
[2] M. Canakci, J.H. Van Gerpen, Comparison of engine performance and emissions for
petroleum diesel fuel, yellow grease biodiesel, and soybean oil biodiesel, Transac-
tions of the American Society of Agricultural Engineers 46 (2003) 937–944.
[3] M. Balat, Fuel characteristics and the use of biodiesel as a transportation fuel, En-
ergy Sources Part A: Recovery, Utilization and Environmental Effects 28 (2006)
855–864.
[4] A. Kurki, J. Bachmann, H. Hill, Oilseed processing for small-scale producersAvailable at:
http://www.attra.org/attra-pub/oilseed.html2008[Accessed 10 November 2010].
[5] K.J. Harrington, C. D'Arcy-Evans, A comparison of conventional and in situ methods
of transesterification of seed oil from a series of sunflower cultivars, Journal of the
American Oil Chemists' Society 62 (1985) 1009–1013.
[6] U.S EPA, U.S. Environmental Protection Agency, The Clean Air Act Amendments of
1990 List of Hazardous Air Pollutants[online] Available at: http://www.epa.gov/
ttn/atw/orig189.html1990[Accessed 3 May 2010].
[7] S.H. Shuit, K.T. Lee, A.H. Kamaruddin, S. Yusup, Reactive extraction and in situ es-
terification of Jatropha curcas L. seed for the production of biodiesel, Fuel 89
(2010) 527–530.
[8] S. Dufreche, R. Hernandez, T. French, D. Sparks, M. Zappi, E. Alley, Extraction of
lipids from municipal wastewater plant microorganisms for production of biodie-
sel, Journal of the American Oil Chemists' Society 84 (2007) 181–187.
[9] M.J. Haas, K.M. Scott, T.A. Foglia, W.N. Marmer, The general applicability of in situ
transesterification for the production of fatty acid esters from a variety of feed-
stocks, Journal of the American Oil Chemists' Society 84 (2007) 963–970.
[10] M.B. Johnson, Z. Wen, Production of biodiesel fuel from the microalga schizochytrium
limacinum by direct transesterification of algal biomass, Energy and Fuels 23 (2009)
5179–5183.
[11] R. Poirot, L. Prat, C. Gourdon, C. Diard, J.-M. Autret, Fast batch to continuous solid–
liquid extraction from plants in continuous industrial extractor, Chemical Engi-
neering and Technology 30 (2007) 46–51.
[12] M.J. Haas, K.M. Scott, W.N. Marmer, T.A. Foglia, In situ alkaline transesterification:
an effective method for the production of fatty acid esters from vegetable oils,
JAOCS, Journal of the American Oil Chemists' Society 81 (2004) 83–89.
[13] J. Qian, F. Wang, S. Liu, Z. Yun, In situ alkaline transesterification of cottonseed oil
for production of biodiesel and nontoxic cottonseed meal, Bioresource Technolo-
gy 99 (2008) 9009–9012.
[14] S. Özgül, S. Türkay, In situ esterification of rice bran oil with methanol and etha-
nol, Journal of the American Oil Chemists' Society 70 (1993) 145–147.
[15] J. Zeng, X. Wang, B. Zhao, J. Sun, Y. Wang, Rapid in situ transesterification of sun-
flower oil, Industrial and Engineering Chemistry Research 48 (2009) 850–856.
[16] G. Kildiran, S.O. Yücel, S. Türkay, In-situ alcoholysis of soybean oil, Journal of the
American Oil Chemists' Society 73 (1996) 225–232.
[17] D. Franco, J. Sineiro, M. Pinelo, M.J. Núñez, Ethanolic extraction of Rosa rubiginosa
soluble substances: oil solubility equilibria and kinetic studies, Journal of Food
Engineering 79 (2007) 150–157.
[18] L.A. Johnson, E.W. Lusas, Comparison of alternative solvents for oils extraction,
Journal of the American Oil Chemists' Society 60 (1983) 229–242.
[19] J.T. Chien, J.E. Hoff, M.J. Lee, H.M. Lin, Y.J. Chen, L.F. Chen, Oil extraction of dried
ground corn with ethanol, Chemical Engineering Journal 43 (1990) B103–B113.
[20] A. Mondala, K. Liang, H. Toghiani, R. Hernandez, T. French, Biodiesel production
by in situ transesterification of municipal primary and secondary sludges, Bio-
resource Technology 100 (2009) 1203–1210.
[21] B. Freedman, R.O. Butterfield, E.H. Pryde, Transesterification kinetics of soybean
oil 1, Journal of the American Oil Chemists' Society 63 (1986) 1375–1380.
[22] M. Mittelbach, B. Trathnigg, Kinetics of alkaline catalyzed methanolysis of sun-
flower oil, Fat Science. Technology 92 (1990) 145–148.
[23] D. Franco, M. Pinelo, J. Sineiro, M.J. Núñez, Processing of Rosa rubiginosa: extrac-
tion of oil and antioxidant substances, Bioresource Technology 98 (2007)
3506–3512.
[24] S. Sayyar, Z.Z. Abidin, R. Yunus, A. Muhammad, Extraction of oil from Jatropha
seed-optimization and kinetics, American Journal of Applied Sciences 6 (2009)
1390–1395.
[25] F.H. Kasim, A. Harvey, R. Zakaria, Biodiesel production by in situ transesterification,
Biofuels 1 (2010) 355–365.
[26] K.G. Georgogianni, M.G. Kontominas, P.J. Pomonis, D. Avlonitis, V. Gergis, Alkaline
conventional and in situ transesterification of cottonseed oil for the production of
biodiesel, Energy and Fuels 22 (2008) 2110–2115.
[27] K.G. Georgogianni, M.G. Kontominas, P.J. Pomonis, D. Avlonitis, V. Gergis, Conven-
tional and in situ transesterification of sunflower seed oil for the production of
biodiesel, Fuel Processing Technology 89 (2008) 503–509.
[28] M.J. Haas, Simplifying biodiesel manufacture, Industrial Bioprocessing 27 (2005)
5.
[29] G. Vicente, M. Martínez, J. Aracil, A. Esteban, Kinetics of sunflower oil methanolysis,
Industrial and Engineering Chemistry Research 44 (2005) 5447–5454.
[30] H.B. Coats, M.R. Wingard, Solvent extraction. III. The effect of particle size on ex-
traction rate, Journal of the American Oil Chemists' Society 27 (1950) 93–95.
[31] D.F. Othmer, J.C. Argawal, Extraction of soybean: theory and mechanism, Chemi-
cal Engineering Progress 51 (1955) 372.
[32] J. Lajara, Solvent extraction of oil from oilseed: the real basics, Proceedings of the
World Conference on Edible Fats and Oils Processing, 1989, pp. 49–55.
[33] M.J. Haas, K.M. Scott, Moisture removal substantially improves the efficiency of in
situ biodiesel production from soybeans, JAOCS, Journal of the American Oil
Chemists' Society 84 (2007) 197–204.
[34] Y. Ren, A. Harvey, R. Zakaria, Biorefining based on biodiesel production: chemical
and physical characterisation of reactively extracted rapeseed, Biobased Materials
and Bioenergy 4 (2010) 1–8.