Fuel 89 (2010) 677684

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Variables affecting the in situ transesterication of microalgae lipids

E.A. Ehimen, Z.F. Sun *, C.G. Carrington
Physics Department, University of Otago, 730 Cumberland Street, Dunedin 9016, New Zealand

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

Article history: This paper describes the effect of important reaction variables on the production of biodiesel from non-
Received 17 February 2009 edible microalgae lipids, using the acid-catalysed in situ transesterication process. The specic gravity
Received in revised form 22 September of the biodiesel product was used to monitor the conversion progress. The results indicate that increasing
2009
the reacting alcohol volume and the temperature lead to improved fatty acid methyl ester (FAME) con-
Accepted 13 October 2009
Available online 31 October 2009
versions. With the exception of in situ transesterication carried out at room temperature (23 C), the
equilibrium FAME conversions appear to approach asymptotic limits for reaction times greater than
8 h for all temperatures investigated. Stirring the reaction vessel had a signicant positive inuence on
Keywords:
Microalgae
the rate of biodiesel formation. Increasing the moisture content of the microalgae biomass had a strong
Biodiesel negative inuence on the equilibrium FAME yield, and in situ transesterication was inhibited when the
In situ transesterication biomass water content was greater than 115% w/w (based on oil weight).
2009 Elsevier Ltd. All rights reserved.

1. Introduction acidic (hydrochloric acid, HCl) and alkaline (sodium hydroxide,

The use of microalgae biomass for the production of biodiesel
(fatty acid alkyl esters, FAAE) has been described by Chisti [1] as
one of the most promising biomass feedstocks with the potential
to meet petro-diesel replacement targets without encroaching on
arable land suitable for food production. The production of biodie-
sel from microalgae oil has previously been demonstrated in the
literature using the conventional route [2,3], which involves the
extraction of the lipids from the microalgae biomass followed by
its conversion to FAAE and glycerol. Most of the current research
on the application of the conventional transesterication scheme
on biomass oils has focused on the use of alkaline catalysts and vir-
gin biomass oils with a free fatty acid (FFA) content of <0.5% w/w
(based on oil weight). However, in considering the production of
biodiesel from microalgae, the use of the alkaline catalysed transe-
sterication technology would not be suitable, due to the charac-
teristically high FFA content of microalgae lipids. This is because
the use of alkaline catalysts with high FFA containing oils would
result in a partial saponication reaction, leading to the soap for-
mation [4] and difculties in the biodiesel separation and purica-
tion downstream. The use of inorganic acids, such as sulphuric and
hydrochloric acids, as reaction catalysts have therefore been con-
sidered for microalgae lipid transesterication, due to its insensi-
tivity to the FFA content of this oil feedstock, since both biodiesel
producing transesterication and esterication reactions are facil-
itated via acidic catalysis. Nagle and Lemke [2] examined the use of
the two groups of catalysts for the conversion of microalgae oil:
* Corresponding author. Tel.: +64 3 4797812; fax: +64 3 4790964.
E-mail address: zhifa@physics.otago.ac.nz (Z.F. Sun).
NaOH). These authors showed that the use of acidic catalysts for
the production of fatty acid methyl esters (FAME) from microalgae
oil resulted in higher product yields than alkaline catalysts under
the same reaction conditions. The inuence of the concentration
of the catalyst, the reaction time and the temperature, on the
transesterication reaction was also investigated in Ref. [2], with
an optimal equilibrium yield of biodiesel from microalgae oil ob-
tained using 0.6 N hydrochloric acid in methanol for 0.1 h at a reac-
tion temperature of 70 C. The conventional transesterication of
microalgae lipids was also demonstrated by Miao and Wu [3],
who showed that the use of a 100% acid catalysts (on the basis of
the microalgae oil weight), a molar ratio of methanol to oil of
56:1, a temperature of 30 C and a reaction time of 4 h, gave the
best biodiesel yields within the experimental conditions tested.
The commercial production of biodiesel from microalgae still
faces hurdles however, mainly due to the high costs associated
with the present biomass production and fuel conversion routes.
Various schemes to reduce the cost requirement, in order to im-
prove the competiveness of this biomass source, are currently
being investigated. These include research into improvements in
the biomass productivity and yield, photo-bioreactor optimisation
and alternative microalgae-biodiesel conversion technologies.
One of the options for microalgae lipid conversion to biodiesel is
the supercritical transesterication process, which is facilitated at
high reaction temperatures and pressures (>240 C and >8 MPa,
respectively) without the requirement of catalysts [5]. This tech-
nology has been demonstrated to have the shortest reaction times
to reach equilibrium conversion of the oils to biodiesel (120240 s)
as well as higher fuel product yields compared to the conventional
process, and a simplied fuel purication step [6]. However, this

0016-2361/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fuel.2009.10.011
678 E.A. Ehimen et al. / Fuel 89 (2010) 677684
method is currently disadvantaged due to the adverse process eco-
nomics as well as safety concerns related to the reaction conditions
[7]. The supercritical method is not explored further in this paper,
as the authors are not primarily concerned with the application of
this scheme for reducing microalgae biodiesel production costs.
Another alternative to the conventional process, which is con-
sidered to have potential of reducing the processing units and costs
of the fuel conversion process, is the in situ transesterication
method, the main focus of this paper. The in situ process facilitates
the conversion of the biomass oil to FAAE directly from the oil
bearing biomass, thereby eliminating the solvent extraction step
required to obtain the oil feedstock as in the conventional method.
This biodiesel production scheme could therefore aid in the simpli-
cation of the fuel conversion process, potentially reducing the
overall process cost, hence lowering the nal fuel product costs
[8] as well. This method may be especially advantageous for use
with microalgae, since the extraction of microalgae lipids is usually
accomplished via solvent extraction and not with the use of cheap-
er physical extraction methods (for example, expellers) as utilised
for conventional oil crops. The alcoholysis of the oil in the biomass
directly has been shown to result in increased biodiesel yields,
compared to the conventional route [9]. Process wastes and pollu-
tion could also be reduced [8] by this method. This paper aims to
provide useful preliminary information on the application of the
in situ transesterication process for the production of biodiesel
using microalgae. Due to the FFA content of microalgae lipids, acid
catalysis is considered for this conversion scheme.
The application of the in situ transesterication process using
homogenous acid catalysts for biodiesel production from oil bear-
ing biomass is not a novel one. The method was rst demonstrated
by Harrington and DArcy-Evans [9,10] with sunower seeds as
feedstock. Using the in situ method these authors achieved an in-
crease in biodiesel yields of up to 20% compared to the conven-
tional process. This improvement in the biodiesel yields was
considered by these authors to be attributable to the improved
accessibility of the oil in the biomass by the acidic medium [10].
The in situ transesterication of macerated sunower seeds was
also studied by Siler-Marinkovic and Tomasevic [11] who investi-
gated two temperature levels of 30 C and 64.5 C and a range of
test reaction conditions: the alcohol (methanol) to oil molar ratio
varied from 100:1 to 300:1, the sulphuric acid catalysts concentra-
tion ranged from 16% to 100% (on the basis of the oil) and a reac-
tion time of 14 h. Under the conditions studied, the best FAME
yields (98.2%) based on the oil content of the sunower seeds
was obtained at a molar ratio of methanol to oil of 300:1, an acid
catalyst concentration of 100% and a reaction time of 1 h [11].
The use of the acid in situ transesterication scheme has also been
demonstrated for FAAE production from soybean and high FFA
containing rice bran oil [12,13].
Due to expected process differences arising from biomass mac-
eration, as well as the perceived differences in oil solubility and
conversion, the ndings from these studies cannot be applied to
the in situ process for microalgae biomass. Consequently informa-
tion is required on the effects that important reaction variables
have on the in situ microalgae transesterication process. Since
the main driver for the use of the in situ transesterication process
is its possible application as a cost reducing scheme for microalgae
biodiesel, this paper considers primarily reaction variables that
might potentially inuence the process costs of the in situ process.
Accordingly this investigation includes the following inuences on
the in situ transesterication of microalgae lipids: (i) the reacting
alcohol volume (the greater the alcohol volume, the more the
material costs input), (ii) the temperature (process energy require-
ment increasing with temperature increases), (iii) the reaction
time, and (iv) process mixing (energy requirement for the reaction
agitation). This study has been carried out with the aim of provid-
ing information on the operating conditions that provide the best
yield while also having the least material and energy requirements,
and consequently lowest process costs. In addition, the effect of the
moisture content of the microalgae biomass on the conversion pro-
cess will be studied since this is of importance energetically and
cost-wise, especially in regions where drying cannot be achieved
using low cost sun drying or other solar heating methods. This is
especially important since biomass drying has been demonstrated
to be one of the most important economic steps in the microalgae
production process, reported to account for up to 30% of the total
production costs [14]. With the microalgae biomass containing
up to 80% moisture content after the harvesting process (for exam-
ple, via centrifugation), this study therefore also aims to evaluate
the extent to which moisture containing microalgae biomass can
be utilised for the in situ transesterication process in order to re-
duce the biomass drying requirements.
2. Materials and methods
2.1. Biomass production and microalgae oil characterisation
Local strains of the microalgae species Chlorella obtained from
the algal collection of Landcare Research, Nelson, New Zealand
were used. The microalgae was cultivated as in Ref. [15] using
BG 11 media consisting of 1.49 g l1 NaNO3, 74.9 mg l1 MgSO4,
36.0 mg l1 CaCl22H2O, 6.0 mg l1 citric acid, 11.2 lg l1 Na-EDTA
(pH 8.0, 0.25 M), 2.86 mg l1 H3BO3, 1.81 mg l1 MnCl24H2O,
0.22 mg l1 ZnSO47H2O, 0.39 mg l1 Na2MoO42H2O, 79.0 lg l1
CuSO45H2O, 49.4 lg l1 Co(NO3)26H2O, 1.0 ml l1 Fe ammonium
citrate (1000x), 1.0 ml l1 Na2CO3 (0.19 M), and 1.0 ml l1 K2HPO4
(0.175 M). The Chlorella biomass was grown autotrophically at
25 C (1 C) in 5 l asks in an environmental growth chamber
with a light intensity of 80.15 lmol m2 s1 and a 16 h photope-
riod per day. Aeration and agitation of the cultures investigated
were achieved by bubbling with air. The biomass sample was then
harvested by centrifugation and dried to a constant weight in a
forced draft oven at 80 C. Dried microalgae biomass was ground
in a mortar and its oil content obtained using the extraction meth-
ods described by Zhu et al. [16]. After re-extraction with hexane,
followed by vacuum evaporation and drying with anhydrous so-
dium sulphate, the extracted oil was weighed and the total oil con-
tent per dry mass of algal biomass was determined.
The specic gravity (SG) of the extracted microalgae oil at 25 C
was determined using a pycnometer using the AOCS ofcial meth-
od Cc10a-25 [17]. The fatty acid composition of the Chlorella oil
was determined via gas chromatography (GC) to provide informa-
tion on the concentration of the component fatty acids and aid in
the estimation of the molecular mass of the oil sample. The fatty
acid methyl esters for GC analysis were prepared via saponication
followed by boron triuoride/methanol esterication using meth-
ods described in Ref. [18]. The samples were then run on an Agilent
6890A GC system equipped with a 7683 series auto sampler and a
ame ionisation detector (FID). Separation was performed on a
BPX70 column (0.32 mm ID, 50 m length and 0.25 lm lm thick-
ness, SGE) with hydrogen as the carrier gas (ow rate-2 ml/min).
The temperatures were as follows: injector, 250 C; detector,
250 C; oven, 205 C (programmed to start at 35 C, held at this
temperature for 5 min and heated at a rate of 2.5 C/min to
205 C). The reference FAME standard was FAMQ005 by AccuStan-
dard (USA), diluted 1:10 in dichloromethane. The resulting data
was then analysed using Chemstation software by Agilent.
The acid value as well as the free fatty acid content of the mic-
roalgae oil was titrimetrically determined using the AOCS ofcial
methods [17].
 E.A. Ehimen et al. / Fuel 89 (2010) 677684
2.2. The in situ transesterication set up
The in situ transesterication process was carried out using
dried microalgae, and various investigated volumes of the reacting
alcohol (methanol) containing sulphuric acid as the transesterica-
tion catalyst. A xed sulphuric acid concentration of 0.04 mol,
which relates to a 100% acid catalyst concentration (on the basis
of the microalgae oil content mass), was used throughout this
study. The vessels containing the reaction mixtures were then
heated and maintained at the temperatures of interest for specied
periods. The major in situ transesterication reaction and product
purication steps used are shown in Fig. 1ae. After the transeste-
rication step (Fig. 1a) the reaction vessel was allowed to stand for
1 h to enable its contents to settle. The reaction mixture was l-
tered and the residues washed twice by re-suspension in methanol
(30 ml) for 10 min to recover any traces of FAME product left in the
residues (Fig. 1b). Water (50 ml) was added to the ltrate, to facil-
itate the separation of the hydrophilic components of the extract,
and then transferred to a separating funnel (Fig. 1c). Further
extraction of the FAME product was achieved by extracting three
times for 15 min using 30 ml of hexane (Fig. 1c). The pooled hex-
ane extracts were washed with water (to remove left-over traces
of the acidic catalyst and methanol), separated, and then dried over
anhydrous sodium sulphate (Fig. 1d). The FAME product was then
ltered and evaporated to obtain the FAME yields (Fig. 1e). Dupli-
cate experiments were carried out for each parameter investigated.
The SG of the extracted FAME products was then determined and
compared with that of the microalgae oil to monitor the extent
of the conversion process, as in Ref. [19]. With the forward reaction
resulting in FAME production and the process nearing completion,
the SG of the puried reaction product (after the removal of the
glycerol co-product) is expected to decrease until constant, signify-
ing an equilibrium conversion of the microalgae lipids to the
methyl esters. The terms FAME product and biodiesel are used
interchangeably in this paper to represent the mixture of FAME
and any remaining unreacted glycerides after the glycerol purica-
tion step.
To further elucidate the extent of oil to FAME conversion by
the in situ transesterication process, a more quantitative rela-
tionship between the measured SG and the percentage FAME
conversion was made by subjecting selected puried FAME prod-
ucts, obtained for different experimental levels (as described in
Section 2.3) with dissimilar measured SGs, to the GC analysis
as described in Section 2.1. The concentration of the FAME
species for each sample was acquired and a calibration curve
Fig. 1. Schematic representation of the in situ transesterication steps used in this study.
679
showing a relationship between the recorded SG and FAME con-
version was then obtained.
2.3. Effect of alcohol volume and temperature on in situ
transesterication
HPLC grade methanol (99.9% purity) was used as the reacting
alcohol in this study. Chlorella powder (15 g) was mixed with var-
ious methanol volumes (20.0, 40.0, 60.0, 80.0 and 100.0 ml) con-
taining 2.2 ml of sulphuric acid (98% purity) in screwed cap
reaction vessels as described in Section 2.2. A minimum volume
of 20.0 ml methanol was selected since it was the minimum
amount that facilitated a complete submersion of 15 g of the mic-
roalgae powder. The experiment involved heating the reaction
mixtures in screwed cap vessels for 8 h, with each trial at one of
four different temperatures (23, 30, 60 and 90 C) with continuous
stirring using a hot plate with a magnetic stirrer. The in situ reac-
tion was studied at a local average room temperature of 23 C, i.e.
the process was not heated externally. Due to the expected build
up of pressure in the reaction vessels when investigating temper-
atures greater than the boiling point of methanol (65 C) at atmo-
spheric pressure, the experiments at 90 C were carried out in a
pressurised environment (3 bar) to ensure the reacting species re-
mained in a liquid state. The respective FAME products at different
investigated variable levels were then obtained as in Section 2.2
and their SGs determined.
2.4. Effect of reaction time on in situ transesterication
At each of the four temperature levels, the in situ transesteri-
cation of 15 g microalgae biomass was repeated in duplicate (as
described in Section 2.2) with reaction times of 0.25, 0.5, 1, 1.5,
2, 4, 8 and 12 h with 60 ml methanol containing 2.2 ml sulphuric
acid. This was carried out to provide a greater insight on the pro-
gression of the transesterication process with time with respect
to the various investigated reaction temperatures. The purication
of the FAME product and its SG determination was carried out as
described above.
2.5. Effect of moisture content of microalgae biomass on in situ
transesterication
Chlorella biomass, initially dried in a draft oven at 80 C to a con-
stant mass, was designated to represent a dry (0.0% moisture)
sample, and was used as the basis for calculating the moisture
680 E.A. Ehimen et al. / Fuel 89 (2010) 677684

content of the microalgae samples. To study the effect of moisture ment, transesterication was carried out as before using 15 g bio-
on the in situ transesterication process, freshly harvested Chlorella mass with 60 ml of methanol containing 0.04 mol sulphuric acid
paste was air dried (fan aided) at 26 C (1 C) to obtain samples with a reaction time of 8 h and a temperature of 60 C. The reaction
with biomass moisture contents of 72.5%, 56.0%, 40.9%, 31.7%, products were puried and the SGs determined.
19.5% and 8.7% (based on the dry sample weight). Using a draft
oven at 80 C, samples of 3.2%, 1.4% and 0.7% moisture content
3. Results and discussion
(dry basis) were also obtained. The in situ transesterication pro-
cess was then investigated in duplicate for different moisture lev-
3.1. Lipid content and SG of pure oil
els on a basis of a 15 g dry reacting sample (for example, 25.88,
23.40 and 17.95 g were used for the 72.5%, 56.0% and 19.5% mois-
With the culture conditions used in this study, the Chlorella
ture levels, respectively) at varied reaction times up to 6 h using a
samples were determined to have a total transesteriable lipid
xed reaction temperature of 60 C and a methanol volume of
fraction of 0.276 g of microalgae oil/g of Chlorella biomass. It must
60 ml containing 0.04 mol of sulphuric acid. The purication of
however be noted that the biomass oil content observed in this
the product was carried out and the respective SGs determined.
study is not only inuenced by the microalgae specie, but is also
highly dependent on the specic growth conditions. As mentioned
2.6. Effect of stirring on in situ transesterication
in Section 2.1, increases in the oil content of the biomass, and
hence increases in the biodiesel production potential of the micro-
To investigate the effect of stirring, the reaction vessels used for
algae feedstock, can be achieved by manipulating the microalgae
the in situ transesterication process were subjected to four differ-
culture conditions, such as the culture nutrients and light intensity
ent stirring treatments: (1) continuously stirred, (2) stirred for only
[20].
1 h, (3) stirred intermittently, 1 h on and 1 h off, and (4) not stirred
The SG of the extracted oil, which was used to characterise the
throughout the experiment. The reaction stirring was carried out
reacting oil at the start of the transesterication reaction, was
using a magnetic stirrer system with a rotation speed of 500 rpm
determined to be 0.914 at 25 C (0.1 C). With all the constituent
kept constant throughout the duration for the agitated samples.
microalgal oils converted to their respective FAME using the
This speed was used since it was observed to facilitate a complete
saponication and transesterication methods described in Sec-
suspension of the particles in the reaction vessels. For each treat-
tion 2.1, the results for the percentage principal fatty acids of the
extracted microalgae oil, as elucidated via GC analysis of the result-
ing FAME mixture, are shown in Table 1. This data was used to
Table 1 determine the average molecular mass of the Chlorella oil. Fatty
Fatty acid composition of the Chlorella biomass. acids detected only in trace amounts (<0.1%) were, however, not
Fatty Molecular mass (g/ % in Molecular mass contribution
included in these results. Since the microalgae oil (assumed to be
acid mol) (MMFA) sample (g/mol) (MMc) primarily composed of triglycerides) consists of different fatty
acids, their respective contributions to the overall molecular mass
C16:0 256.42 4.37 11.21
C16:1 254.41 0.44 1.12 of the microalgae lipid is used (as seen in the MMc column in Ta-
C18:1 282.46 61.81 174.59 ble 1) to estimate the mean molecular mass of the constituent lipid
C18:2 280.45 19.94 55.92 fatty acids (MMFA). With the formation of the triglyceride molecule
C18:3 278.43 12.22 34.02
facilitated by the combination of fatty acid molecules and a mole-
C20:1 310.52 1.22 3.79
Average molecular mass of constituent 280.65 cule of glycerol with the condensation of three molecules of water,
fatty acids (MMFA) the average molecular mass of the microalgae oil (MM oil) can be
calculated using Eq. (1).

0.92

0.915

0.91
SG of purified FAME product

0.905

0.9

0.895

0.89

0.885

0.88
0 10 20 30 40 50 60 70 80 90 100
% FAME conversion

Fig. 2. Calibration curve showing the relationship the measured SG and the % FAME conversion obtained using GC.
 E.A. Ehimen et al. / Fuel 89 (2010) 677684 681

MMoil 3MMFA MMglycerol   3MMwater 1 investigated reacting methanol volumes and temperatures, using
a xed reaction time of 8 h and a xed acidic catalyst molar
where, MMglycerol and MMwater represent the molecular masses of
concentration (0.04 mol sulphuric acid), was determined. This is
glycerol and water, respectively. The average molecular weight of shown in Fig 3. Using the SG of the extracted oil (0.914) to indicate
the Chlorella lipid was calculated to be 880 g/mol.
the start of the transesterication reaction, the conversion of the
As described in Section 2.1, following the purication of the constituent microalgae oils to biodiesel can therefore be moni-
FAME products and measurement of the SG for different FAME
tored, with lower asymptotic SG values of the puried transesteri-
yields, the concentration of the FAME species in the fuel product cation products indicating improved equilibrium FAME
were obtained by GC analysis and used to quantify the percentage
production. Eqs. (2)(4) show that the transesterication reaction
of Chlorella oil converted to FAME products. The calibration curve consists of a series of reversible steps in which the oils (repre-
obtained, showing the relationship between the measured SG
sented as triglycerides) are sequentially converted to di- and
and the percentage FAME conversion, is shown in Fig. 2. The curve mono-glycerides, and eventually to glycerol, with one mole of fatty
can therefore be used to predict the corresponding percentage
acid alkyl esters (RCOOR1) liberated at each step.
FAME conversions in this study since it covers a range from a 0%
conversion using the pure extracted oil (SG, 0.914) to 92.22% which Trigyceride R1 OH $ diglyceride RCOOR1 2
was for the lowest measured SG (0.8826).
The acid value of the microalgae oil was determined to be
Diglyceride R1 OH $ monoglyceride RCOOR1 3
10.21 mg KOH/g Chlorella oil. Using the estimated molecular mass
of 280.65 for the constituent fatty acids, the FFA content of the
microalgae oil was determined to be 5.11% (on the basis of the Monogyceride R1 OH $ glycerol RCOOR1 4
oil weight). Due to the high FFA content (>0.5% w/w) of the micro- 1
With methanol as the reacting alcohol (R OH), the equilibrium
algae oil, the choice of acidic over alkaline catalysts for the in situ transesterication products range from a mixture of FAME, glyce-
transesterication process is justied. rides and glycerol (when the reaction is incomplete) to a FAME
and glycerol mixture (on completion). Since the heavier glycerol
3.2. Effect of alcohol volume and temperature product of the reaction is eliminated from the nal product via
purication steps, the SG of the puried FAME product (containing
Using the average molecular mass of the Chlorella oil, as deter- glycerides) as expected will be less than that of the extracted oil,
mined in Section 3.1, and with the knowledge of the transesteri- and would vary with the respective percentage composition of
cation stoichiometry involving 3 mol of reacting alcohol per mole the FAME specie in the puried product. For example, an increase
of oil, the methanol volumes investigated in this study represent in the lighter FAME concentration would result in lower measured
a reacting molar alcohol to oil ratio range of 105:1524:1 (calcu- product SGs.
lated with a methanol density of 0.7918 g/cm3). This range includes The results obtained indicate an improvement of the microalgae
and exceeds that of a similar investigation of the in situ transeste- oil conversion to FAME with increasing temperature and increasing
rication of sunower oil by Siler-Marinkovic and Tomasevic alcohol volume, with the lowest FAME equilibrium conversions ob-
[11], described in Section 1. The SG of the FAME products for the served with the reacting molar ratios of the methanol to oil at

20 ml methanol 40 ml methanol 60 ml methanol 80 ml methanol 100 ml methanol

0.8895

0.8885
Specific gravity (SG) of the extracted FAME product

0.8875

0.8865

0.8855

0.8845

0.8835

0.8825

0.8815
15 25 35 45 55 65 75 85 95

In-situ transesterification temperature (C)

Fig. 3. Effect of the alcohol volume and reaction temperature on the specic gravity of the FAME products after a reaction time of 8 h using 15 g microalgae biomass and 100%
acidic catalyst (on basis of oil).
682
105:1 (that is, a methanol volume of 20 ml) for all the temperature
levels studied. However, with the use of alcohol volumes over
60 ml (i.e. a reacting molar ratio of alcohol to microalgae oil greater
than 315:1) for the in situ transesterication of 15 g of microalgae
biomass, no signicant trends were observed for the measured SGs
determined for the reaction temperature levels of 60 and 90 C.
3.3. Effect of reaction time and temperature
To investigate the inuence of reaction time and temperature, a
methanol volume of 60 ml was used since it was found (Section 3.2)
that no appreciable differences in the equilibrium FAME conver-
sion were obtained with the use of higher alcohol volumes for
reactions carried out at 60 and 90 C. Fig. 4 shows the progression
of the microalgae oil to biodiesel conversion process with time and
at different temperature levels, using the measured SG of the puri-
ed FAME product as a conversion indicator. For the samples
investigated at room temperature (no process heating), asymptotic
FAME conversion values were not reached within the time bound-
aries of this study. When the in situ transesterication process was
carried out at 90 C under the same experimental conditions the
highest equilibrium conversion levels (lowest FAME SG) were
reached within the shortest time, with 70% and 92% of the overall
equilibrium FAME conversion values attained after a reaction time
of 15 min and 1 h, respectively. Using a reacting molar ratio of
methanol to oil of 315:1, and continuously stirring the reaction
at 500 rpm, a faster conversion phase was observed between
15 min and 2 h for the 15 g microalgae samples investigated at
23 and 30 C, as shown in Fig. 4. This phase is considered to be
due to the emulsifying effects of the initial fatty acid esters formed
after the mass transfer controlled phase, as discussed in Ref. [21].
This effect was however not signicantly observed for the 60 and
90 C experiment runs, probably due to the fact that the elevated
temperatures (and pressures) improved the initial miscibility of
the reacting species, leading to a signicant reduction in the reac-
tion times, as observed in Fig. 3. Within the experimental condi-
tions investigated, equilibrium FAME conversions were observed
to reach similar asymptotic values after a reaction time of 2 and
0.915
0.91
Specific gravity (SG) of extracted FAME product
0.905
0.9
0.895
0.89
0.885
0.88
0 2
Fig. 4. Inuence of varying the reaction time on the in situ transesterication of 15 g microalgae at reaction temperatures of 23, 30, 60 and 90 C using 60 ml methanol and
100% acidic catalyst (on basis of oil).
E.A. Ehimen et al. / Fuel 89 (2010) 677684
4 h for temperatures of 60 and 90 C. Although faster conversion
rates can be observed by use of reaction temperatures greater than
the boiling point of the reacting methanol (for example, 90 C), the
process heating and pressure requirements may inhibit the use of
such temperature levels. The use of a reaction temperature of 60 C
may therefore prove more benecial, if we consider the total en-
ergy consumption and operation cost of the whole biodiesel con-
version system.
3.4. Effect of the moisture content of the microalgae biomass
As discussed in Section 1, the inuence of the microalgae mois-
ture levels is important in this study due to its potential impact on
the overall biomass and fuel production costs. The different mois-
ture content levels of the samples investigated were used to repre-
sent biomass samples, ranging from freshly harvested (after
centrifugation) with moisture content of 72.5% to completely
dried. It must be noted, however, that the water content of the
dried sample (0.0%), used as the basis for determining the moisture
content of the microalgae samples investigated, refers to free bio-
mass water, not the more tightly bound water. The measured SG of
the puried FAME yields obtained from Chlorella biomass with dif-
ferent moisture content levels at different reaction time intervals
are shown in Fig. 5. The general trends of the results obtained indi-
cate that increases in the biomass moisture content have a signif-
icant negative effect on the equilibrium conversion of oil to
biodiesel using the acid-catalysed in situ transesterication pro-
cess. The oil to FAME conversion was observed to be similar, and
almost inhibited, for the microalgae samples with moisture con-
tents greater than 31.7% (dry basis) i.e. containing 4.755 g of water
per sample, or 114.85% w/w when compared with the microalgal
oil weight of 4.14 g. Hence, for the conditions investigated, these
results indicate that the process would be inhibited in moisture
containing microalgae samples with water levels greater than
115% of the reacting oil weight for the acid-catalysed in situ transe-
sterication process. Obvious differences were only observed after
a 73% removal of water from the freshly harvested samples, i.e.
with the in situ transesterication of the 19.5% (dry basis) samples.
23C
30C
60C
90C
4 6 8 10 12
In-situ transesterification time (h)

72.5%
0.915
Specific gravity (SG) of the extracted FAME product
0.91
0.905
0.9
0.895
0.89
0.885
0.88
0 1
Fig. 5. Inuence of microalgae moisture content (dry basis) on the FAME conversion using 100% acid catalyst (on basis of oil) and a xed reaction alcohol volume of 60 ml and
temperature of 60 C, respectively.
A reduction of the biomass moisture levels to 0.7% (dry basis),
which relates to a reacting water mass content of 0.1 g, or 2.4%
w/w (reacting oil basis), resulted in equilibrium conversions of
81.7% of the values obtained with completely dried samples after
a reaction time of 6 h. Biomass drying cannot therefore be over-
looked in the in situ transesterication process, with improved
equilibrium conversions observed when using completely dried
samples compared with moisture containing samples.
3.5. Effect of stirring on in situ transesterication
The effect of stirring on the in situ process was investigated as a
potential process cost saving strategy. When the in situ transeste-
rication process was conducted without stirring, the equilibrium
conversion of the microalgae oil content to biodiesel, compared to
that for the continuously stirred sample, was signicantly reduced
(Table 2). This indicates that stirring is required to some extent to
enhance the reaction progress, evidently by aiding the initial mis-
cibility of the reacting species. The stirring of the reaction vessels
for only 1 h (from the start of the reaction) was investigated. This
time was chosen because 87% of the equilibrium FAME conversion
was achieved after 1 h with continuous stirring under the experi-
mental transesterication conditions, using 15 g microalgae bio-
mass with 60 ml methanol (containing 0.04 mol acid catalyst), at
a reaction temperature of 60 C. It was therefore envisaged that,
without further stirring, the equilibrium conversion might ap-
proach that of the continuously stirred system with the produced
Table 2
Inuence of stirring on the in situ transesterication of microalgae lipids with a xed
acid catalyst concentration (100%, on basis of oil), methanol volume of 60 ml and
reaction temperature at 60 C.
No stirring
Stirring for 1 h
Stirred intermittently (1 h off, 1 h on)
Stirred continuously
E.A. Ehimen et al. / Fuel 89 (2010) 677684 683
56.0% 40.9% 31.7% 19.5% 8.7% 3.2% 1.4% 0.7% 0.0%
2 3 4 5 6 7
In-situ transesterification time (h)
methyl esters acting as a dual solvent for the oil and alcohol, there-
by facilitating further FAME production. However, after a reaction
time of 8 h, the samples stirred for 1 h only were observed to
achieve only 91.3% of the equilibrium FAME conversion achieved
by the samples which were continuously stirred. Furthermore,
intermittent stirring at 1 h intervals achieved equilibrium FAME
yields which were close to, but nevertheless lower than, those ob-
served in the vessels stirred continuously.
It should be noted that the FAME product SG for the trial with-
out stirring might have had a higher SG than indicated in Table 2 in
the absence of any stirring at all, since all the samples were initially
swirled by hand before being placed on the hot plate and stirring
system. The purpose of this action was to prevent clumping and
ensure that the biomass samples were adequately exposed to the
acidic methanol mixture. However, it could have also promoted
some initial phase mixing of the reactants in the not stirred sam-
ple. The results in Table 2 may therefore overstate the equilibrium
conversion yields of the unstirred samples.
4. Conclusion
This study investigated the inuence of potentially important
economic reaction factors on the progress of the conversion of mic-
roalgae oil to biodiesel using the acid-catalysed in situ transesteri-
cation process. These variables include the inuence of reacting
alcohol volumes, temperature, reaction time, biomass moisture
content and stirring. The results show that increases in the reaction
temperature and alcohol volume favour the production of biodie-
sel. However, within the experimental conditions, an increase in
the reacting alcohol volume of more than 60 ml did not show sig-
nicant changes in the yield of the FAME products obtained, when
using 15 g of Chlorella biomass. The application of the in situ acid-
SG of extracted FAME product catalysed process has been discussed in the literature [9] to offer
0.9032 conversion rate improvements when compared with the conven-
0.8859 tional process. However, the energetic cost of the recovery of the
0.8845
excess alcohol reactants in this process might, as well as an in-
0.8831
crease in downstream FAME purication requirements, might limit
684 E.A. Ehimen et al. / Fuel 89 (2010) 677684
the achievable cost reduction, one of the main reasons which
would favour the use of the in situ transesterication method.
Biodiesel yields by the in situ transesterication of microalgae
were shown to improve signicantly with process stirring. How-
ever, some savings may be achieved in the stirring energy since
the reactors can be stirred intermittently without a major reduc-
tion in the yield. Biomass drying was observed to play an impor-
tant role, with an increase in the moisture content of the
biomass resulting in signicant reductions of the equilibrium
FAME conversion yields. Therefore extensive biomass drying may
be needed prior to biodiesel production by the in situ process.
Although not covered in this study, continuous removal of the
water product from the reactor, as well as an increase in the acidic
catalyst concentration, are schemes which may be utilised to drive
the forward reaction when microalgae biomass with high moisture
content is used in the in situ transesterication process.
In summary, the production of biodiesel from microalgae is still
in the research and development stages, with laboratory scale
transesterication trials (in situ, in this study) carried out aimed
at providing useful information to aid the design and modelling
of industrial sized biodiesel production processes. Further process
design, optimisation and integration investigations on unit opera-
tions such as biomass drying, ltration, evaporation, extraction,
adsorption, as well as the chemical processing steps are still re-
quired to optimise the use of this biomass feedstock.
Acknowledgements
The authors are most grateful to Assoc Prof. Julian Eaton-Rye
and Jackie Shand of the Photosystem II lab, Biochemistry Depart-
ment, as well as Dr John Birch and Michelle Leus of the Food Sci-
ence Department, University of Otago, Dunedin for the use of
their facilities, instruction and assistance. We wish to also thank
Dr Phil Novis of Landcare Research for providing the Chlorella cul-
ture used in this study and Mr & Mrs P. Ehimen for their kind nan-
cial support. The authors would also like to acknowledge the
anonymous reviewers whose useful comments and criticisms were
most helpful in editing this paper.
References
[1] Chisti Y. Biodiesel from microalgae. Biotechnol Adv 2007;25:294306.
[2] Nagle N, Lemke P. Production of methyl ester fuels from microalgae. Appl
Biochem Biotechnol 1990;24/25:35561.
[3] Miao X, Wu Q. Biodiesel production from heterotrophic microalgal oil.
Bioresour Technol 2006;97:8416.
[4] Al-Zuhair S. Production of biodiesel: possibilities and challenges. Biofuels
Bioprod Biorening 2007;1:5766.
[5] Demibras A. Production of biodiesel from algae oil. Energy Sources Part A
Recov Util Environ Eff 2009;31:16974.
[6] Saka S, Kusidana D. Biodiesel fuel from rapeseed oil as prepared in supercritical
methanol. Fuel 2001;80:22531.
[7] Marchetti JM, Errazu AF. Techno-economic study of supercritical biodiesel
production plant. Energy Conv Manag 2008;49:21604.
[8] Haas MJ, Scott KM, Foglia TA, Marmer WN. The general applicability of in situ
transesterication for the production of fatty acid esters from a variety of
feedstocks. J Am Oil Chem Soc 2007;84:96370.
[9] Harrington KJ, DArcy-Evans C. A comparison of conventional and in situ
methods of transesterication of seed oil from a series of sunower cultivars. J
Am Oil Chem Soc 1985;62:100913.
[10] Harrington KJ, DArcy-Evans C. Transesterication in situ of sunower seed oil.
Ind Eng Chem Prod Res Dev 1985;24:3148.
[11] Siler-Marinkovic S, Tomasevic A. Transesterication of sunower oil in situ.
Fuel 1998;77:138991.
[12] Kildiran G, zgl-Ycel S, Trkay S. In-situ alcoholysis of soybean oil. J Am Oil
Chem Soc 1996;73:2258.
[13] zgl-Ycel S, Trkay S. Variables affecting the yields of the methyl esters
derived from the in-situ esterication of rice bran oil. J Am Oil Chem Soc
2002;79:6114.
[14] Becker EW. Microalgae: biotechnology and microbiology. Cambridge studies in
biotechnology. Cambridge: Cambridge University Press; 1994. 293pp.
[15] Ehimen EA, Carrington CG, Sun ZF, Eaton-Rye J. Biodiesel from microalgae:
investigating reaction conditions for in-situ transesterication. In: Proceedings
of the 2nd international symposium on energy from biomass and waste,
Venice, November 2008.
[16] Zhu M, Zhou PP, Yu LJ. Extraction of lipids from Mortierella alpine and
enrichment of arachidonic acid from the fungal lipids. Bioresour Technol
2002;84:935.
[17] AOCS. The ofcial methods and recommended practices of the AOCS. 5th ed.
American Oil Chemists Society, 2005.
[18] Van Wijngaarden D. Modied rapid preparation of fatty acid esters from lipids
for gas chromatographic analysis. Anal Chem 1967;39:8489.
[19] Al-Widyan MI, Al-Shyoukh AO. Experimental evaluation of the
transesterication of waste palm oil into biodiesel. Bioresour Technol
2002;85:2536.
[20] Qiang H, Sommerfeld M, Jarvis E, Ghiradi M, Posewitz M, Siebert M, Darzins A.
Microalgal triacyglycerols as feedstock for biofuel production; perspectives
and advances. Plant J 2008;54:62139.
[21] Freedman B, Buttereld RO, Pryde EH. Transesterication kinetics of soybean
oil. J Am Oil Chem Soc 1986;63:137580.