This document describes a process to extract free fatty acids from wet biomass of the microalga Nannochloropsis gaditana for biodiesel production. The process involves five steps: 1) direct saponification of the wet biomass using KOH-ethanol solution to extract fatty acids as potassium salts, 2) separation of unsaponifiable lipids from the solution with hexane, 3) acidification and extraction of purified free fatty acids with hexane, 4) esterification of the free fatty acids into biodiesel using methanol catalyzed by sulfuric acid, and 5) purification of the biodiesel by washing with hot water and adsorption with bentonite. The final biodiesel purity was

1. Extraction of free fatty acids from wet Nannochloropsis gaditana
biomass for biodiesel production
Estrella Hita Pe~na, Alfonso Robles Medina*
, María J. Jimenez Callejon,
María D. Macías Sanchez, Luis Esteban Cerdan, Pedro A. Gonzalez Moreno,
Emilio Molina Grima
Area of Chemical Engineering, University of Almería, 04120 Almería, Spain
a r t i c l e i n f o
Article history:
Received 21 January 2014
Accepted 7 October 2014
Available online 25 October 2014
Keywords:
Biodiesel
Microalga
Nannochloropsis gaditana
Fatty acid
Esterification
a b s t r a c t
The objective of this work is to develop a process for producing biodiesel from the saponifiable lipid (SL)
fraction of the wet microalgal biomass Nannochloropsis gaditana. The method consists of five steps.
Firstly, crude fatty acid salt extraction was carried out using a KOH-ethanol (96%) solution, which allows
one to extract the SLs as potassium salts. This transformation permits better separation of the unsapo-
nifiable lipids (the second step) and finally produces purer biodiesel. The unsaponifiable lipids were then
separated with hexane, after establishing the ethanol-water solution water content at 30% w/w. Some
unsaponifiable lipids (carotenoids and phytosterols) are products of interest that might be purified from
this fraction thus helping to improve the process's profitability. Thirdly, free fatty acids (FFAs) were
purified by acidification of the ethanol-water solution to pH 5 and were then extracted with hexane.
Fourthly, the FFAs were transformed to biodiesel by esterification with excess of methanol catalyzed
using sulphuric acid, removing the excess by washing with hot water. Under these conditions the bio-
diesel purity and yield were 74.8% and 82% w/w, respectively. Finally, the biodiesel was clarified/purified
up to 96.5% purity by adsorption with bentonite. The final biodiesel yield was 80.9%.
© 2014 Elsevier Ltd. All rights reserved.
1. Introduction
The search for renewable transportation biofuels has been
stimulated by the negative environmental impact of fossil fuels and
concerns about petroleum supplies. Microalgae have recently been
considered one of the best alternative sources of biodiesel [1]. The
lipid composition of microalgae is qualitatively different from that
of common vegetable oilseeds and the conventional technologies
for processing them may be unsuitable for microalgae. Further-
more, the particular microalgae species and the time of harvesting
greatly affect lipid composition. Microalgae biomass harvested in
the exponential growth phase will contain more polar lipids than
biomass harvested in the late stationary growth phase. Only
saponifiable lipids (SLs) are transformable into methyl esters (bio-
diesel) but algae contain a high content of other lipids, as well as
non-lipid components. Thus, only around 24e27 % of total lipids in
Isochrysis galbana and Phaeodactylum tricornutum microalgae are
neutral lipids (mainly acylglycerols, although neutral lipids also
contain unsaponifiable wax ester components, hydrocarbons and
sterols), while the remainder are polar lipids (phospholipids and
glycolipids) [2]. All the unsaponifiable constituents create diffi-
culties in the crude lipid refining process necessary for biodiesel
production.
The extraction of lipids from microalgae has been repeatedly
attempted using both physical and chemical methods, as well as a
combination of both. A comprehensive review of lipid and fatty acid
extraction procedures can be seen elsewhere [3]. Once the lipids
have been extracted, the crude extract obtained must be purified
before commencing the conversion of lipids to fatty acid methyl
esters. Processes for refining crude algal lipid extracts, similar to
those existing for terrestrial plants, have not been developed to
date.
An alternative to lipid extraction for biodiesel is the direct
extraction of fatty acids from biomass by means of direct saponi-
fication, which enables fatty acids to be obtained as potassium or
sodium salts instead of as crude lipids [4e6]. Direct saponification
is faster and cheaper than lipid extraction although more intensive
operating conditions are necessary [7]. In a previous paper, a
* Corresponding author. Area of Chemical Engineering, Department of Engi-
neering, University of Almería, 04120 Almería, Spain. Tel.: þ34 950 015065;
fax: þ34 950 015484.
E-mail address: arobles@ual.es (A. Robles Medina).
Contents lists available at ScienceDirect
Renewable Energy
journal homepage: www.elsevier.com/locate/renene
http://dx.doi.org/10.1016/j.renene.2014.10.016
0960-1481/© 2014 Elsevier Ltd. All rights reserved.
Renewable Energy 75 (2015) 366e373

2. comprehensive study was carried out for optimizing fatty acid
extraction from wet I. galbana biomass followed by its purification
[6]. Fatty acid extraction was performed using a three-step
method: direct saponification of the wet biomass, followed by
the extraction of unsaponifiable constituents, and finally the
extraction of purified fatty acids. Even though the high water
content of the wet microalgal biomass (about 80% w/w) affects the
saponification yield, it is preferable using wet rather than dry
biomass because biomass drying involves high costs both in terms
of equipment and energy.
The goal of this work was to optimize the variables involved in
the above-mentioned three-step process and to adapt it for the
extraction and purification of fatty acids from the wet biomass of
the microalga Nannochloropsis gaditana. The free fatty acids
extracted were transformed into methyl esters (biodiesel) by acid-
catalyzed esterification. Finally, these methyl esters were washed
with hot water and decoloured/purified by adsorption with
bentonite.
2. Materials and methods
2.1. Microalgal biomass and chemicals
Wet paste biomass from the marine microalga N. gaditana was
used as an oil-rich substrate. Cells were grown in an outdoor
tubular photobioreactor, centrifuged at 7000 rpm for 10 min, and
then stored at À20 C until the time of use. This wet biomass
contained 20.4 ± 0.2% w/w of dry biomass and 23.1 ± 0.2% w/w of
total lipid to biomass dry weight. The total fatty acid content (or
saponifiable lipids as equivalent fatty acids) in the biomass was
10.8 ± 0.5% w/w of dry biomass. Table 1 shows the fatty acid
composition of the wet paste biomass from N. gaditana used in this
study. The chemicals used were analytical grade hexane (95% pu-
rity), ethanol (96% v/v) (aqueous ethanol), HCl (37%), H2SO4 (96%
purity) and diethyl ether from Panreac S.A. (Barcelona, Spain),
methanol (99.9% purity, Carlo Erba Reagents, Rodano, Italy), KOH
(85% purity, J.T. Baker, Deventer, Holland), bentonite (Guinama,
Alboraya, Spain) and distilled water. All reagents used in the
analytical determinations were also of analytical grade. These re-
agents were acetone, ethyl ether (both from Panreac S.A., Barce-
lona, Spain), nonadecanoic acid (19:0) and its methyl ester (used as
internal standards for gas chromatography (GC) analyses (Fluka
Analytical, SigmaeAldrich, St. Louis, MO, USA) and acetyl chloride,
used as the sample methylation catalyst for analysis by GC (Fluka
Analytical). For the determination of carotenoids by HPLC and
spectrophotometry, b-carotene and lutein standards were pro-
vided by Sigma Chemical Co. (St. Louis, MO); violaxanthin and
zeaxanthin were obtained from DHI LAB (Horsholm, Denmark) and
neoxanthin was provided by ChromaDex LGC Standards (Barce-
lona, Spain).
2.2. Fatty acid extraction from wet biomass
2.2.1. Direct saponification of wet biomass
Fatty acid extraction was performed using the three-step
method shown in Fig. 1. In the first step, potassium salts were
formed from the fatty acids contained in the saponifiable lipid (SL)
fraction of the microalgal biomass. In a typical experiment (Fig. 1),
24.5 g of wet biomass (equivalent to 5 g of dry biomass) were
treated with 150 mL of aqueous ethanol, containing 1.0 g of KOH, in
a 1 L reactor that was jacketed for temperature control. Saponifi-
cation was carried out at 60 C for 1 h with constant magnetic
agitation in an argon atmosphere. The mixture obtained was then
filtered through a 100e160 mm microporous glass filter (Pobel,
Madrid, Spain) and the biomass residue was washed with 70 mL
aqueous ethanol. In this step, the KOH to biomass ratio ranged
from 0.2 to 1.6 g KOH per gram of dry biomass, and the aqueous
ethanol to biomass ratio ranged from 5 to 76 mL aqueous ethanol
(96%) per gram of dry biomass. This ratio was modified in the
biomass residue washing step. Up to three washings were tested
with 70 mL of aqueous ethanol each (i.e. 14 mL aqueous ethanol/g
of dry biomass).
2.2.2. Extraction of unsaponifiable lipids
The fatty acids dissolved in the ethanol-water solution were
purified extracting the unsaponifiable lipids with hexane. In a
typical experiment, 36 mL of water was added to 200 mL of the
fatty acid salt solution (14.5% w/w water and 27.4 mg/L of carot-
enoids) to obtain a solution with 30% w/w water; and then the
unsaponifiable lipids were extracted at 20 C by adding hexane and
shaking (Fig. 1). The two phases were subsequently separated by
decantation and aliquots of each phase were taken to determine
carotenoids (hexane phase) and fatty acids (ethanol-water phase).
Several experiments were carried out with different percentages of
water in the ethanol-water phase (14.5e70 % w/w) with a view to
maximizing the recovery of unsaponifiable lipids. A hexane to
ethanol-water solution ratio of 1:1 (v/v) was used in this extraction.
The optimal water content of the ethanol-water phase proved to be
30% w/w. At this ethanol-water solution, and on this small scale, up
to three extractions of unsaponifiable lipids from the ethanol-water
phase were tested, with a hexane/ethanol-water solution ratio of
1:1 v/v per step, in order to reduce the amount of carotenoids in the
ethanol-water phase.
2.2.3. Extraction of purified fatty acids
In this third step, the pH of the 70:30 w/w ethanol-water so-
lution was adjusted to values of between 1 and 6.5 using 37% HCl.
Extractions of fatty acids were then performed in an argon atmo-
sphere, at 20 C, shaking for 10 min and using a hexane/ethanol-
water solution ratio of 1:1 v/v. The phases were then separated
and an aliquot of the hexane phase was taken for fatty acid
determination.
2.3. Scaling-up of fatty acid extraction from wet microalgal biomass
The scheme shown in Fig. 1 was followed. 500 g of wet biomass
was treated with 3075 mL of aqueous ethanol, containing 20.6 g of
KOH in a 5 L reactor, which was jacketed for temperature control
Table 1
Fatty acid composition (percentage with respect to the total
fatty acids of the biomass) of the wet paste biomass used from
Nannochloropsis gaditana.
Fatty acids wt %
14:0 3.0 ± 0.0
16:0 16.0 ± 0.1
16:1n7 17.1 ± 0.1
16:3n4 4.8 ± 0.0
18:1n9 3.4 ± 0.0
18:1n7 0.5 ± 0.0
18:2n6 8.0 ± 0.1
18:3n3 9.0 ± 0.0
20:4n6 4.7 ± 0.0
20:5n3 20.5 ± 0.1
Others 13.0 ± 0.1
P
Saturated 19.4 ± 0.0
P
Monounsaturated 21.0 ± 0.0
P
PUFAs 47.3 ± 0.0
Total fatty acids (*)
10.8 ± 0.5
Total lipidsa
23.1 ± 0.2
a
Percentage of biomass dry weight.
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373 367

3. and equipped with a discharge valve at the bottom. Saponification
was carried out at 60 C for 1 h with constant agitation (250 rpm)
using a propeller stirrer (Eurostar digital, IKA Staufen, Germany)
with two blades at different heights. Then, the biomass residue was
separated from the ethanol-water solution by filtration using a
12 cm-diameter porous glass plate (100 mm pore diameter, Pobel,
Madrid, Spain). In this case only one biomass residue washing step
was tested using 1.4 L of aqueous ethanol (14 mL aqueous ethanol/g
dry biomass), and this washing solution was added to the ethanol-
water fatty acid solution from the main extraction. The ethanol-
water phase, which contains the soaps, was placed in a 10 L glass
extractor equipped with a propeller stirrer with two blades. Water
(493 mL) was added to reach 30% of the solution weight and
unsaponifiable lipids were extracted with 3970 mL of hexane. This
extraction was carried out by stirring at 250 rpm for 10 min at room
temperature. The solution was then allowed to settle for 30 min,
presenting an orange upper phase (a hexane phase containing the
unsaponifiable lipids) and a green lower hydroalcoholic phase that
contained the fatty acids as potassium salts. The unsaponifiable
solution was removed, leaving the lower phase in the extractor.
Concentrated HCl was then added to the ethanol-water phase until
it reached pH ¼ 5, to form the free fatty acids. Hexane was added
(3970 mL) and this mixture was shaken at 250 rpm for 10 min to
extract the free fatty acids. This mixture was allowed to settle for
30 min and two phases were obtained, a lower one, which was
discarded, and the upper hexane phase containing the fatty acids,
which was kept. This fatty acid solution was evaporated in a rotary
evaporator (Buchi R210, with a V-700 vacuum pump and a V-850
controller, Switzerland) to remove the solvent and recover the free
fatty acids.
2.4. Methylation of free fatty acids and washing of methyl esters
Free fatty acids from N. gaditana biomass were transformed to
methyl ester by acid-catalyzed esterification. This reaction was
carried out in 50 mL Erlenmeyer flasks placed in an orbital shaking
air-bath (Inkubator 1000, Unimax 1010 Heidolph, Klein, Germany),
at 60 C, stirring at 200 rpm for 2 h and with a 10:1 methanol/free
fatty acid molar ratio and 1.8% w/w sulphuric acid (with respect to
the fatty acids) [8]. Then the methyl esters (biodiesel) were washed
with hot distilled water in the orbital shaking air-bath, at 50 C,
stirring at 300 rpm for 15 min to remove the methanol and
Fig. 1. Optimized process to extract and purify fatty acids from the microalga Nannochloropsis gaditana.
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373368

4. sulphuric acid residues. Three washings were carried out using a
0.5:1 v/v water/biodiesel ratio in for each washing. The two phases
formed were separated by centrifugation at 3900 rpm for 10 min
(Mixtasel JP, Selecta, Barcelona, Spain). In the first and second
washings, the aqueous phase was brownish-green, then colourless
washing water was obtained after the third washing.
2.5. Clarification/purification of methyl esters by adsorption with
bentonite
Methyl esters were clarified and purified by adsorption with
bentonite. In all these experiments, 1 g of crude methyl esters was
dissolved in 10 mL of hexane-diethyl ether 99:1 v/v [9] and 2 g of
bentonite (bentonite/crude biodiesel ratio 2:1 w/w) were added.
This mixture was agitated at 250 rpm in the orbital shaking air-
bath, at 30 and 40 C over different periods (from 1 to 24 h). The
bentonite was separated from the biodiesel solution by centrifu-
gation at 3900 rpm for 10 min and washed with 5 mL of hexane-
diethyl ether 99:1 v/v to recover adsorbed methyl esters. This
second liquid phase was separated from bentonite again by
centrifugation and then added to the first one.
2.6. Analysis
2.6.1. Total lipids, fatty acid content and recovery yields
The total lipid content of N. gaditana biomass was determined
by the method described by Richardson et al. [10]. Fatty acids in
the biomass, in the hexane phase and in the ethanol-water phase
were analyzed by capillary gas chromatography (GC). To deter-
mine fatty acids in both the hexane and the ethanol-water phases,
a known volume was dried under a N2 stream, and methylation
was carried out by direct transesterification with acetyl chloride/
methanol (1:20 v/v) following the method of Rodríguez Ruiz et al.
[11]. The methyl esters were analyzed with an Agilent Technology
6890 gas chromatograph (Avondale, PA, USA), using a capillary
column of fused silica Omegawax™ (0.25 mm  30 m, 0.25 mm
standard film, Supelco, Bellefonte, PA, USA) and a flame-ionization
detector (FID). A detailed methodology analysis has been
described elsewhere [12]. The total fatty acid content in the
biomass was determined by direct fatty acid transesterification
[13].
The saponifiable lipid (SL) content in the microalgal biomass
was expressed as the fatty acid weight percentage although these
fatty acids are present as acyl groups in the saponifiable lipids.
Recovery yields were expressed as the fatty acids recovered in any
of the processes with respect to the total of fatty acids contained in
the initial microalgal biomass.
2.6.2. Conversion of the methylation reaction, yield and biodiesel
purity
Samples from the methylation reaction (i.e. crude biodiesel,
Section 2.4) with 10 mL (0.125 mg) of internal standard (non-
adecanoic acid methyl ester, 19:0) in hexane (concentration 1 mg/
mL) were analyzed directly by GC. Using this procedure, the only
fatty acids determined were those transformed to methyl esters
(biodiesel) by the methylation of free fatty acids (Section 2.4). In
addition, these samples were also methylated by direct trans-
esterification with acetyl chloride/methanol (1:20 v/v) and
analyzed again by GC; all the fatty acids were determined using this
procedure. The conversion was calculated by Eq (1) [14].
Conversion ¼
fatty acid amount in methyl esters
total amount of fatty acids
100 (1)
Recovery yields were also expressed as the fatty acids trans-
formed to biodiesel with respect to the total fatty acids contained in
the initial microalgal biomass. Methyl ester or biodiesel purity was
calculated by Eq. (2)
where the amount of methyl ester (the numerator) was quantita-
tively determined by GC and the total amount of sample (the de-
nominator) was determined by weighing (i.e. crude biodiesel
determined by weighing).
2.6.3. Determination of carotenoids by HPLC and spectrophotometry
The different types of carotenoids present in the samples were
determined by HPLC. Prior to HPLC analysis, the samples were
extracted in acetone and saponified as described in Ref. [15]. The
carotenoids in the extracts thus obtained were analyzed by the
chromatographic method described by Mínguez-Mosquera et al.
[16] (modified by Del Campo et al. [17], and Ceron et al. [18]) using a
Shimadzu SPD-M10AV high-performance liquid chromatograph
(Columbia, USA). Elution order, retention times and peak identifi-
cation were compared with those previously reported [19]. b-
carotene, lutein, violaxanthin, zeaxanthin and neoxanthin were
used as standards (Section 2.1).
Carotenoids were also determined as a measure of the unsapo-
nifiable lipid extraction yield using a modified version of the method
employed by Whyte [20]. To determine carotenoids in the ethanol-
water solution, from the saponification reaction, a known volume of
this solution was dried under a N2 stream. The residue was resus-
pended in an aqueous solution with 60% KOH, and the carotenoids
were extracted with ethyl ether. The ethyl-ether phase was then
dried, and carotenoids were resuspended in a known volume of
acetone. To determine carotenoids in the hexane phase, a known
volume of this solution was dried and the residue was resuspended
in a known volume of acetone. The optical densities (OD) of these
suspensions were determined at 444 nm using acetone as the
reference. Carotenoid concentration (C) was determined by Eq. (3):
C
À
mg LÀ1
Á
¼ 4:32 OD À 0:0439 (3)
This equation was obtained using b-carotene and acetone as
solute and solvent, respectively, with a correlation coefficient (r2
) of
0.9998. These determinations were carried out using a Minolta CM-
3500d Spectrophotometer connected to a computer with Spec-
traMagic 3.6.1 software (Osaka, Japan).
3. Results and discussion
3.1. Direct saponification of wet biomass. Optimization of KOH/
biomass and aqueous ethanol/biomass ratios
In a previous work [7], direct saponification was conducted on
different extraction solvents (n-butanol, ethanol-water, hexane-
Methyl ester or biodiesel purity ¼
methyl ester amont in the final biodiesel
total amount of sample
100 (2)
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373 369

5. isopropanol, hexane-aqueous ethanol and ethanol (96%)) - the
highest yield was obtained with ethanol (96%) (aqueous ethanol).
Consequently, this solvent was chosen to carry out the direct
saponification of saponifiable lipids (SLs) in the N. gaditana
biomass. The aqueous ethanol/biomass ratio used in that work [7]
(76 mL/g) was chosen from the lipid extraction methods of Bligh-
Dyer [21] and Kates [22]. In these extractions, 1.6 g of KOH and
20 mL of aqueous ethanol/g dry biomass were also used for
washing the biomass residue. The experiments presented here
were carried out in order to reduce these amounts and adapt the
process to the microalga N. gaditana.
Saponification experiments were performed with a constant
amount of wet biomass (24.5 g; i.e. 5 g of dry biomass). KOH acts by
saponifying the SLs and breaking the cell wall, which is rich in
phospholipids [23]. To begin with, the amount of KOH/dry biomass
ratio was varied between 0.2 and 1.6 g/g. The stoichiometric KOH/
dry biomass ratio required to saponify all the fatty acids present in
the wet biomass is 0.024 g/g (the average molecular weight of the
fatty acids contained in the wet N. gaditana biomass is 274.3 g/mol,
which was calculated according to the fatty acid composition
shown in Table 1). Therefore, in all cases, an excessive amount of
KOH was used to ensure complete fatty acid saponification. Table 2a
shows that, with this biomass, the KOH/biomass ratio had little
effect on the fatty acid yield within the range tested. Consequently,
the smallest amount of KOH can be used, i.e. only 0.2 g KOH/g dry
biomass. This low amount of KOH will reduce the alkalinity of the
medium and will also save HCl for the subsequent transformation
of fatty acid salts into free fatty acids.
Table 2b shows the fatty acid extraction yields at increasing
aqueous ethanol/biomass ratios (between 5 and 76 mL/g). In the
range tested, no significant differences were observed between
yields. With aqueous ethanol/dry biomass ratios equal to, or higher
than, 30 mL/g, extraction yields were about 90%; while below this
ratio the yields were not too low (generally 80e87 %). However,
with these low ethanol/biomass ratios, the subsequent filtration
took too long. Therefore, this ratio can be reduced to 30 mL aqueous
ethanol/g dry biomass. This aqueous ethanol/biomass ratio was
used in the large-scale experiment (with 500 g of wet biomass), and
a fatty acid recovery yield of about 87.6% was attained, which is
similar to the result in the small-scale experiment (90.6%).
The biomass residue from saponification (Fig. 1) still contains
some fatty acids that can be recovered to increase the fatty acid
yield. For this reason, trials were carried out to optimize the
amount of aqueous ethanol used for biomass washing and the
number of washings. At the small scale, up to three washings were
performed with 14 mL of aqueous ethanol/g dry biomass for each
wash. Fig. 2 shows that only a small amount of free fatty acids was
recovered after three biomass residue washings. Only 4, 0.5 and 0%
of fatty acids present in the initial biomass were recovered in the
first, second and third washings, respectively. Therefore it seems
that a single washing may be sufficient (14 mL ethanol/g dry
biomass) to drag the residual fatty acids from the biomass residue.
With these new KOH and aqueous ethanol amounts, and carrying
out only one biomass residue washing step, the global fatty acid yield
in this first step was about 93e94% at the small scale. As a result, in
the large-scale experiment, a single biomass residue washing step
was tested. In this case, the fatty acid recovery yield only increased
marginally from 87.6% (in the extraction of fatty acid salts from the
original biomass; see the last line of Table 2b and Fig. 2) to 89.1%
(after one biomass residue washing), i.e. the fatty acid recovery yield
only increased 1.5% (Fig. 2); and, in this case, 1.4 L of hexane was
needed (14 mL hexane/g dry biomass). For this reason, the biomass
residue was not washed in the optimized process, as Fig. 1 shows.
3.2. Extraction of unsaponifiable lipids
The crude fatty acid extract obtained when using aqueous
ethanol/KOH in the saponification step contains fatty acid potassium
salts, and also pigments, such as carotenoids and chlorophylls, as
well as proteins and other lipid and non-lipid contaminants. This
ethanol-water solution contains 27.3 mg/L carotenoids, which are
made up of: neoxanthin (43.1%), violaxanthin (7.3%), astaxanthin
(1.8%), zeaxanthin (45.8%) and b-carotene (1.8%). The unsaponifiable
lipids, such as these carotenoids, can be extracted by treating the
ethanol-water crude lipid extract with solvents such as chloroform
or hexane, in which the fatty acid potassium salts are not soluble.
Here, hexane was used because it is cheaper and less toxic than
chloroform; and it is widely accepted for use in food processing [15].
The ethanol-water fatty acid solution only contains the water
added with the wet biomass and the aqueous ethanol used. When
30 mL of aqueous ethanol/g dry biomass was used, the water
Table 2
Fatty acid yields obtained by direct saponification of N. gaditana wet biomass:
optimization of the KOH/biomass ratio (A) and the aqueous ethanol/biomass ratio
(B).
A) KOH/biomass ratio (g/g dry biomass) Recovery yield (wt %)a
1.6 90.9 ± 1.6
1.2 87.8 ± 0.3
0.8 83.7 ± 0.8
0.4 89.4 ± 1.1
0.2 87.3 ± 0.4
B) Aqueous ethanol/biomass ratio
(mL/g dry biomass)
Recovery yield (wt %)a
10 80.0 ± 1.8
14 85.3 ± 4.5
22 87.5 ± 0.5
30 90.6 ± 0.7
38 90.8 ± 0.8
58 90.9 ± 0.1
76 87.3 ± 0.7
30b
87.6 ± 1.5
Operational conditions: A) 24.5 g wet biomass (5 g dry biomass), 380 mL aqueous
ethanol (76 mL/g dry biomass), 1 h, 60
C. B) 24.5 g wet biomass, 1 g KOH (85%)
(0.2 g/g dry biomass), 1 h, 60
C.
a
Percentage of fatty acids extracted with respect to the fatty acids present in the
initial biomass.
b
Experiment with 500 g of wet biomass.
Fig. 2. Washing of the biomass residue after direct saponification and filtration: fatty
acid recovery yields (wt % of fatty acid with respect to the fatty acids in the initial
biomass) in three washings (small scale), or a single washing (large scale), of the
biomass residue compared to the fatty acid yield in the filtrates after saponification.
Operational conditions: filtrate obtained after the saponification of 24.5 g (small scale)
or 500 g (large scale) of wet biomass with 0.2 g KOH/g dry biomass and 30 mL aqueous
ethanol/g dry biomass; three washings (small scale) with 14 mL aqueous ethanol/g dry
biomass each or a single washing with the same ratio (large scale).
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373370

6. content of the ethanol-water solution was 14.5% w/w. An increase
in this water content improves the immiscibility of the ethanol-
water solution and hexane, which significantly affects the distri-
bution equilibrium of the unsaponifiable lipids between both the
ethanol-water and hexane phases, and increases the unsaponifiable
lipid extraction yield [6]. Table 3 shows the influence of the
ethanol-water phase water content (from 14.5 to 70 % w/w) on the
carotenoid extraction yield and the fatty acid recovery yield into the
ethanol-water phase, when unsaponifiable lipids were extracted
with a hexane to ethanol-water solution ratio of 1:1 (v/v).
This table shows that the fatty acid recovery yields did not vary
with the addition of water, which indicates that fatty acid salts
remained in the ethanol-water phase and were not extracted to-
wards the hexane phase. Additionally, it was checked that fatty acid
salts were not present in the hexane phase. However, carotenoid
recovery increased with the water content and the maximum
unsaponifiable lipid extraction yield achieved was 69.3% with 50%
water. These extractions are not quantitative but, in a previous
work (see Fig. 2 of [6]), it was observed that the carotenoid equi-
librium distribution between the two phases was displaced to-
wards the ethanol-water phase, at the low concentration range. On
the other hand, the formation of emulsions is more probable at
water contents over 50% [6]. Emulsions make extraction difficult
and they can even decrease fatty acid recovery. The ethanol-water
solution from saponification contains surfactant agents, such as
soaps and proteins, which facilitate and stabilize the ethanol-water
phase/hexane emulsions. An increase in the water content in-
creases the amount of hydrating water that stabilizes the hydro-
philic moieties of the surfactant agents thus stabilizing the
emulsion [6]. Consequently, an optimal water content of 30% was
chosen. With this water content, the carotenoid extraction yield
was 56.3% with only one extraction stage.
Fig. 3 shows the influence of the water content of the ethanol-
water solution on the carotenoid yield extracted into the hexane
phase. This figure shows that the extraction of total carotenoids
measured by HPLC and spectrophotometry are similar. The most
abundant carotenoid was zeaxanthin (45.6% in the initial ethanol-
water phase) instead of violaxanthin (7.3%); this is because
N. gaditana was cultivated under external conditions, and under
certain conditions of light violaxanthin is transformed into zeax-
anthin. Neoxanthin and violaxanthin were not easily extracted
using hexane (Fig. 3), which is in accordance with its higher po-
larity [24]. b-carotene was extracted almost completely into the
hexane phase, while zeaxanthin was extracted to a lower per-
centage (80% with a 30% water content), which is also in line with
the higher polarity of zeaxanthin due to the presence of OH groups
in its molecule [25]. All these carotenoids (non-polar pigments)
were separated in the hexane phase, while chlorophylls (alcohol-
soluble pigments) remained in the ethanol-water phase along with
the fatty acid salts. This hexane phase could then be treated to
recover these carotenoids, since some of them (b-carotene and
zeaxanthin) are high-value products [26].
In order to purify the fatty acid salts and recover the maximum
amount of carotenoids, three hexane extractions were carried out
from the ethanol-water solution which contained a 30% water
content. Table 4 shows the concentration of the different caroten-
oids before and after extraction and the carotenoid recovery yields
in each of the three extractions. This table shows that carotenoid
extraction was almost complete after these three steps. It also
found that fatty acid salts were not present in these hexane solu-
tions. However, carrying out three extraction steps on unsaponifi-
able lipids involves considerable hexane consumption on a larger
scale. Therefore, although hexane can be recovered, at large scale it
is more feasible to carry out a single unsaponifiable lipid extraction.
3.3. Extraction of fatty acids from the ethanol-water phase.
Influence of pH
In a previous work [27], the pH of the ethanol-water solution
was lowered to 1 so as to transform the potassium salts to free fatty
0,0
20,0
40,0
60,0
80,0
100,0
Neoxanthin Violaxanthin Zeaxanthin b-carotene Totals HPLC Totals
Spectrophot.
Carotenoid extracted
dleiynoitcartxedionetoraC
(%)
14,50% 30% 40% 50% 60%
Fig. 3. Extraction of carotenoids contained in ethanol-water solutions with varying water content using hexane: influence of the water content in the ethanol-water phase on
carotenoid extraction yield in the hexane phase determined by HPLC and spectrophotometry.
Table 3
Extraction of unsaponifiable lipids from the ethanol-water solution using hexane:
influence of the ethanol-water solution water content on the carotenoid yields
extracted into the hexane phase and fatty acids into the ethanol-water phase.
Water content
(wt %)
Carotenoid yield into
hexane phase (wt %)
Fatty acid yield after
unsaponifiable lipid
extraction (wt %)a
HPLC Spectrophotometry
Before extraction 100 90.6 ± 0.7
14.5 26.7 ± 0.8 27.0 ± 1.0 91.2 ± 0.1
30.0 56.3 ± 0.4 52.7 ± 0.3 92.5 ± 0.8
40.0 59.8 ± 0.9 56.3 ± 1.1 91.1 ± 1.6
50.0 69.3 ± 0.5 65.2 ± 0.7 90.3 ± 1.5
60.0 67.5 ± 1.0 62.2 ± 0.8 89.3 ± 0.7
70.0 e e
Unsaponifiable lipid extraction was carried out with an ethanol-water solution/
hexane ratio 1:1 (v/v).
a
Yields expressed as percentage of fatty acids in the ethanol-water phase with
respect to the fatty acids present in the initial biomass.
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373 371

7. acids and then extract them with hexane. However, in another
work [6], emulsions formed when the pH was lowered to this value
and hexane was added. Thus, in this work, the influence of pH on
the fatty acid extraction yield was analyzed. Table 5 shows that the
fatty acid extraction yield was higher when the pH was adjusted to
between 3 and 5. At pH 6.5, a low fatty acid extraction yield was
obtained, probably because the fatty acids were still partially in the
form of potassium salts. Therefore, pH 5 was selected because at
this pH, the fatty acid extraction yield was the same as that ob-
tained at pH 3. In this extraction step the chlorophylls, which are
more soluble in ethanol than in hexane, should remain in the
ethanol-water phase in greater proportion.
3.4. Production of methyl esters (biodiesel) from microalgal free
fatty acids
This step involved a modification of the procedure optimized by
Chongkhong et al. [8] for biodiesel production by esterification of
palm fatty acid distillate (PFAD), which contains 93% w/w free fatty
acids, as well as mono, di and triacylglycerols. The optimal condi-
tions obtained by these authors for the continuous esterification
process were a methanol/PFAD molar ratio of 8:1, with 1.8% w/w of
sulphuric acid with respect to the fatty acids, at 70 C and for
60 min. Under these conditions, these authors reduced the amount
of free fatty acids to less than 2%. In the present work, the tem-
perature was reduced to 60 C in order to work under milder
conditions but the reaction time was increased up to 120 min and a
10:1 methanol/fatty acid molar ratio was used to ensure high
conversion. Under these conditions, the conversion of fatty acids to
methyl esters was almost complete (98e99 %).
After the free fatty acid esterification, the methyl esters formed
also contain sulphuric acid and the excess of methanol. To remove
them, three washing steps were necessary with hot water (50 C).
After these washings, the water was neutral and colourless; and the
fatty acid recovery yield with respect to the fatty acid content in the
initial biomass was 82%. Thus, only 18% of fatty acids were lost
throughout this process (extraction, purification, esterification and
washing steps). However, this biodiesel contains 74.8% w/w of
methyl esters and was a dark-brown colour. Therefore, an addi-
tional purification step was carried out in an attempt to eliminate
this 25.2% of impurities, which come from the microalgae.
This process has also been carried out on the fatty acids obtained
after three unsaponifiable lipid extractions using hexane, obtaining
79.8% purity and an 80.4% yield. Therefore, a 5% purity increase was
attained carrying out three unsaponifiable lipid extractions instead
of only one.
Table 6 compares the stages and results obtained by the process
developed in this work as well as that obtained by a three-step
process in which lipids were extracted as saponifiable lipids
(instead of as free fatty acids) from the same wet N. gaditana
microalgal biomass [28]. In this case wet biomass first had to be
homogenized at 1700 bar and then saponifiable lipids were
extracted with hexane for 20 h at room temperature (20e22 C). By
using hexane as the extraction solvent, one was able to obtain
methyl esters of 63% purity; whereas extracting the lipids with
ethanol (96%) reduced the purities below 40%. Also the biodiesel
yield was significantly lower (82% versus 57%, Table 6). Moreover,
unsaponifiable lipids were not separated from the saponifiable ones.
Table 5
Influence of the fatty acid ethanol-water solu-
tion pH on the fatty acid extraction yield using
hexane.
pH Yield (wt %)a
1 81.7 ± 0.9
3 91.1 ± 0.9
4 89.8 ± 0.2
5 90.7 ± 2.3
6 88.1 ± 0.0
6.5 74.3 ± 1.8
Extractions were carried out with a hexane/
ethanol-water solution 1:1 (v/v).
a
Yields expressed as percentage of fatty acids
extracted with respect to total fatty acids in the
initial biomass.
Table 6
Comparison of the results obtained in this work with those obtained in a previous
work [28] where saponifiable lipids were extracted with hexane.
Process step SLsa
extracted as FFA
(this work)
SLsa
extracted with
hexane [28]
Homogenization e 1700 bar
Lipid extraction 30 mL ethanol (KOH)/g dbb
,
1 h, 60
C
10 mL hexane/g dbb
, 20 h,
20-22
Cc
Unsaponifiable
lipid extraction
1:1 hexane/ethanol-water
(v/v), 10 min, 20-22
Cc
e
Fatty acid extraction 1:1 hexane/ethanol-water
(v/v), 30 min, 20-22
Cc
e
Methylation 1.5 mL methanol/0.018 g
H2SO4/g FFA, 60
C, 2 h
1.4 mL hexane/12 mL
methanol/0.22 g H2SO4/g
SLsa
, 80
C, 2 h
Recovery or washing
of biodiesel
Washing with hot
water (50
C)
Extraction with 3/1
hexane/methanol (v/v)
Yield (%) 82 57
Purity (%) 75 63
a
SLs: Saponifiable lipids;
b
db: Dry biomass.
c
Room temperature.
Table 4
Concentrations and recovery yields of carotenoids after three unsaponifiable lipid
extraction steps using hexane.
Carotenoid Carotenoid concentration
in the ethanol-water
phase (mg/L)a
Recovery yields of
carotenoids into hexane
phases (wt %)a
Before
extraction
After three
extractions
First
extraction
Second
extraction
Third
extraction
Neoxanthin 11.8 ± 0.8 2.5 ± 0.6 36.8 ± 0.6 29.2 ± 0.7 16.8 ± 0.5
Violaxanthin 2.0 ± 0.4 0.8 ± 0.2 33.1 ± 0.5 17.2 ± 0.3 12.6 ± 0.6
Zeaxanthin 12.5 ± 0.6 0.4 ± 0.1 79.8 ± 0.7 12.1 ± 0.4 10.2 ± 0.4
b-carotene 0.5 ± 0.1 0.0 ± 0.0 96.1 ± 0.2 0.0 ± 0.0 0.0 ± 0.0
Totals 27.3 ± 0.7 4.0 ± 0.9 56.7 ± 0.4 19.4 ± 0.6 12.8 ± 0.5
a
Yields expressed as percentage of fatty acids in the ethanol-water phase with
respect to the fatty acids present in the initial biomass. Ethanol-water phase with
30% water (w/w).
Table 7
Influence of temperature and adsorption time on the yield and purity of methyl
esters obtained by treatment of the crude methyl esters solution from methylation
and washing (purity 74.8% and 82% yield) by adsorption with bentonite.
Temperature (
C) Time (h) Yield (wt %)a
Purity (wt %)b
30 1 80.9 ± 1.1 75.9 ± 0.1
4 79.2 ± 0.1
8 82.2 ± 0.1
24 87.7 ± 0.1
40 1 84.6 ± 0.1
4 88.4 ± 0.1
8 89.3 ± 0.4
24 96.5 ± 0.5
Operational conditions: 1 g methyl esters, 2 g bentonite, 250 rpm, 10 mL hexane-
diethyl ether 99:1 v/v.
a
Percentage of methyl esters with respect to saponifiable lipids in the original
biomass (both as equivalent fatty acids determined by GC).
b
Percentage of methyl esters in total mass.
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373372

8. 3.5. Purification of biodiesel by adsorption with bentonite
Bentonite is an adsorbent commonly used in the decolouration
of oils. For this reason, to clarify the biodiesel and increase its pu-
rity, an adsorption treatment was carried out with bentonite and
hexane-diethyl ether 99:1 v/v [9,29]. After some previous experi-
ments for the optimization of the adsorption system, the solvent/
bentonite/biodiesel ratio was established at 10/2/1 mL/g/g. Table 7
shows the purities and yields obtained at two different tempera-
tures (30 and 40 C) and by increasing adsorption times. This table
shows that treatment with bentonite slightly decreases the recov-
ery yield to 80.9% although this yield did not depend on time or
temperature. However, purity did increase with time and temper-
ature and, finally, a 96.5% pure biodiesel was obtained after 24 h
treatment at 40 C. In addition, the methyl esters obtained were a
clear yellow colour.
4. Conclusions
In this work a five-step process has been developed to obtain
96.5% pure methyl esters (biodiesel) from the microalga N. gadi-
tana, with an 80.9% yield. This process tests wet biomass, as
opposed to dry or lyophilized biomass, with a view to reducing both
energy and equipment costs. Firstly, the fatty acids contained in the
saponifiable lipids (SLs) were directly saponified in the microalgal
biomass and transformed to fatty acid potassium salts. This trans-
formation allows better separation of the unsaponifiable lipids and
results in a purer biodiesel (fatty acid methyl esters) than would be
obtained if the SLs were extracted without previous saponification;
which is important to meet the specifications of current standards.
In addition, some unsaponifiable lipids, such as certain carotenoids
or phytosterols, are products of interest that could be purified from
the separated unsaponifiable lipid fraction, thus helping to improve
the profitability of the process. The extraction of SLs as free fatty
acids also allows for the use of the fatty acids present in the polar
lipid fraction (phospholipids and glycolipids) [2]. Another advan-
tage of this process is that the esterification reaction velocity for
biodiesel production (free fatty acids þ alcohol) is much greater
than the transesterification reaction velocity
(acylglycerols þ alcohol). Therefore, this paper presents an
economical process for obtaining biodiesel from microalgae with
the possibility of producing other products of interest contained in
the unsaponifiable lipid fraction. If the purity requirements were
not too high, some process steps, such as purification with
bentonite, could be avoided.
Acknowledgement
This research was supported by grants from: (1) ENDESA
(Spain), Projects “Subproyecto: captura, fijacion y valorizacion de
CO2 por medio de planta piloto ubicada en una central termica del
litoral (Algaplane)” and “Energías Renovables y Combustion Limpia
(Novare Valor CO2)”; both projects are co-funded with the “Min-
isterio de Economía y Competitividad” (Spain); and (2) “Ministerio
de Educacion y Ciencia” (Spain), Project CTQ2007-64079; this
project was co-funded by the FEDER (European Fund for Regional
Development).
References
[1] Chisti Y. Biodiesel from microalgae. Biotechnol Adv 2007;25:294e306.
[2] Robles Medina A, Molina Grima E, Gimenez Gimenez A, Iba~nez Gonzalez MJ.
Downstream processing of algal polyunsaturated fatty acids. Biotechnol Adv
1998;16:517e80.
[3] Molina Grima E, Iba~nez Gonzalez MJ, Gimenez Gimenez A. Solvent extraction
for microalgal lipids. In: Borowitzka Michael A, Moheimani Navid R, editors.
Algae for biofuel and energy. Heidelberg/New York/London: Springer; 2013.
pp187e206.
[4] Robles Medina A, Gimenez Gimenez A, García Camacho F, Sanchez Perez JA,
Molina Grima E, Contreras Gomez A. Concentration and purification of
stearidonic, eicosapentaenoic, and docosahexaenoic acids from cod liver oil
and the marine microalga Isochrysis galbana. J Am Oil Chem Soc 1995;72:
575e83.
[5] Gimenez Gimenez A, Iba~nez Gonzalez MJ, Robles Medina A, Molina Grima E,
García Salas S, Esteban Cerdan L. Downstream processing and purification of
eicosapentaenoic (20:5n-3) and arachidonic acids (20:4n-6) from the micro-
alga Porphyridium cruentum. Bioseparation 1998;7:89e99.
[6] Iba~nez Gonzalez MJ, Robles Medina A, Molina Grima E, Gimenez Jimenez A,
Carstens M, Esteban Cerdan L. Optimization of fatty acid extraction from
Phaeodactylum tricornutum UTEX 640 biomass. J Am Oil Chem Soc 1998;75:
1735e40.
[7] Molina Grima E, Robles Medina A, Gimenez Gimenez A, Sanchez Perez JA,
Garcia Camacho F, Garcia Sanchez JL. Comparison between extraction of lipids
and fatty acids from microalgal biomass. J Am Oil Chem Soc 1994;71:955e9.
[8] Chongkhong S, Tongurai C, Chetpattananondh P, Bunyakan C. Biodiesel pro-
duction by esterification of palm fatty acid distillate. Biomass Bioenergy
2007;31:563e8.
[9] €Ozgücel-Yücel S, Türkay S. Purification of FAME by rice hull ash adsorption.
J Am Oil Chem Soc 2003;80:373e6.
[10] Richardson B, Orcutt DM, Schwertner HA, Martinez CL, Wickline HE. Effects of
nitrogen limitation on the growth and composition of unicellular algae in
continuous culture. Appl Microbiol 1969;18:245e50.
[11] Rodríguez-Ruiz J, Belarbi E-H, García Sanchez JL, Lopez Alonso D. Rapid
simultaneous lipid extraction and transesterification for fatty acids analysis.
Biotechnol Tech 1998;12:689e91.
[12] Hita Pe~na E, Robles Medina A, Camacho Paez B, Ramírez Fajardo A, Esteban
Cerdan L, Jimenez Callejon MJ, et al. Production of structured triacylglycerols
(STAG) rich in docosahexaenoic acid (DHA) in position 2 by acidolysis of tuna
oil catalyzed by lipases. Process Biochem 2007;42:415e22.
[13] García Sanchez JL, Molina Grima E, García Camacho F, Sanchez Perez JA,
Gimenez Gimenez A. Cuantificacion de acidos grasos a partir de biomasa
microalgal. Grasas aceites 1993;44:348e53.
[14] Martín Valverde L, Gonzalez Moreno PA, Rodríguez Quevedo A, Hita Pe~na E,
Jimenez Callejon MJ, Esteban Cerdan L, et al. Concentration of docosahexae-
noic acid (DHA) by selective alcoholysis catalyzed by lipases. J Am Oil Chem
Soc 2012;89:1633e45.
[15] Ceron MC, Campos I, Sanchez JF, Acien FG, Molina E, Fernandez-Sevilla JM.
Recovery of lutein from microalgae biomass: development of a process for
Scenedesmus almeriensis biomass. J Agric Food Chem 2008;56:11761e6.
[16] Mínguez-Mosquera MI, Gandul-Rojas B, Gallardo-Guerrero ML. Rapid method
of quantification of chlorophylls and carotenoids in virgin olive oil by high-
performance liquid chromatography. J Agric Food Chem 1992;40:60e3.
[17] Del Campo JA, Moreno J, Rodríguez H, Vargas MA, Rivas J, Guerrero MG.
Carotenoid content of chlorophyceanmicroalgae: factors determining lutein
accumulation in Muriellopsis sp. (Chlorophyta). J Biotechnol 2000;76:51e9.
[18] Ceron MC, García-Malea MC, Rivas J, Acien FG, Fernandez JM, Del Río E, et al.
Antioxidant activity of Haematococcus pluvialis cells grown in continuous
culture as a function of their carotenoid and fatty acid content. Appl Microbiol
Biotechnol 2007;74:1112e9.
[19] Orosa M, Torres E, Fidalgo P, Abalade J. Production and analysis of secondary
carotenoids in green algae. J Appl Phycol 2000;12:553e6.
[20] Whyte JN. Biochemical composition and energy content of six species of
phytoplankton used in mariculture of bivalves. Aquaculture 1987;60:231e41.
[21] Bligh E, Dyer W. A rapid method of total lipid extraction and purification. Can J
Biochem Physiol 1959;37:911e7.
[22] Kates M. 2nd printing. In: Burdon RH, Van Knippenberg PH, editors. Tech-
niques of lipidology: isolation, analysis and identification of lipids. 2nd ed.
Amsterdam, Netherlands: Elsevier Science Publisers B.V.; 1998. p. 168e70.
[23] Thomson GA. Lipids and membrane function in green algae. Biochimica Bio-
physica Acta 1996;1302:7e45.
[24] Lubian LM, Establier R. Estudio comparativo de la composicion de pigmentos
en varias cepas de Nannochloropsis (Eustigmatophyceae). Investig Pesq
1982;46:379e89.
[25] Jeffrey SW, Mantoura RFC, Wright SW. Phytoplankton pigments in oceanog-
raphy: guidelines to modern methods. Paris: UNESCO; 1997. p. 661.
[26] Asker D, Awad TS, Beppu T, Ueda K. Novel zeaxanthin-producing bacteria
isolated from a radioactive hot spring water. Methods Mol Biol 2012;892:
99e131.
[27] Cartens M, Molina Grima E, Robles Medina A, Gimenez Gimenez A, Iba~nez
Gonzalez MJ. Eicosapentaenoic acid (20:5n-3) from the Marine microalga
Phaeodactylum tricornutum. J Am Oil Chem Soc 1996;73:1025e31.
[28] Jimenez Callejon MJ, Robles Medina A, Macías Sanchez MD, Hita Pe~na E,
Esteban Cerdan L, Gonzalez Moreno PA, et al. Extraction of saponifiable lipids
from wet microalgal biomass for biodiesel production. Bioresour Technol
2014;169:198e205.
[29] Berrios M, Martín MA, Chica AF, Martín A. Purification of biodiesel from used
cooking oils. Appl Energy 2011;88:3625e31.
E. Hita Pe~na et al. / Renewable Energy 75 (2015) 366e373 373