Effective extraction of microalgae lipids from wet

biomass for biodiesel production

Hanifa Taher
a
, Sulaiman Al-Zuhair
a,
*, Ali H. Al-Marzouqi
a
, Yousef Haik
b
,
Mohammed Farid
c
a
Chemical and Petroleum Engineering Department, UAE University, Al-Ain, United Arab Emirates
b
Mechanical Engineering Department, UAE University, Al-Ain, United Arab Emirates
c
Chemical and Materials Engineering Department, University of Auckland, New Zealand
a r t i c l e i n f o
Article history:
Received 23 September 2013
Received in revised form
28 January 2014
Accepted 28 February 2014
Available online 20 March 2014
Keywords:
Enzymes
Cell disruption
Wet microalgae
Separation
Nitrogen starvation
Biodiesel
a b s t r a c t
Producing biodiesel from lipid extracted from microalgae is a promising approach for
sustainable fuel production. However, this approach is not yet commercialized due to the
high costs of upstream processes that are associated with the time consuming and/or
energy intensive drying, and lipid extraction processes. In this study, the possibility of
avoiding the drying process, and extracting the lipid directly from the wet concentrated
cells, using enzymatic disruption to enhance the extraction, has been tested. Results
showed that lysozyme and cellulase were both efcient in disrupting cell walls and
enhancing lipid extraction from wet samples, with highest lipid extraction yield of 16.6%
achieved using lysozyme. The applicability of using supercritical CO
2
(SC-CO
2
) in extracting
lipid from wet biomass was also tested and the highest yield of 12.5% was achieved using
lysozyme. In addition, a two-step culturing process was applied, using Scenedesmus sp., to
combine both high biomass growth and lipid content. The strain was able to increase its
biomass productivity in the rst stage, reaching 174 mg l
1
d
1
, with almost constant lipid
content. In the second stage, the lipid content was enhanced by six-fold after three weeks
of nitrogen starvation, but with lower biomass productivity.
2014 Elsevier Ltd. All rights reserved.
1. Introduction
Biodiesel, a series of mono-alkyl fatty acid esters, has received
an increasing attention, especially with the increasing energy
demand and inevitable depletion of fossil fuel reserves [1,2].
Conventionally, biodiesel is produced fromoils extracted from
oil crops. This is highly controversial and competes with their
use as a food stock. In addition, the oil crops require large
arable land development, fertilization, and fresh water
irrigation. These signicant drawbacks limit further industri-
alization and urge to nd new feedstocks. Waste cooking oils
and animal fats have been suggested as alternatives; however
the supply of such feedstocks is not consistent and cannot
satisfy the large demand for biodiesel production. In addition,
poor cold ow properties (cloud and pour points) of oils and
saturated fatty acid contents of animal fats reduce the quality
of produced biodiesel [3]. On the other hand, oil extracted
from microalgae has recently emerged as a potential alter-
native source due to the microalgae high growth rate, high
* Corresponding author. Tel.: 971 35319.
E-mail address: s.alzuhair@uaeu.ac.ae (S. Al-Zuhair).
Available online at www.sciencedirect.com
ScienceDirect
ht t p: / / www. el sevi er. com/ l ocat e/ bi ombi oe
b i o ma s s a nd b i o e ne r gy 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7
http://dx.doi.org/10.1016/j.biombioe.2014.02.034
0961-9534/ 2014 Elsevier Ltd. All rights reserved.
lipid content, and ability to grow in seawater. Furthermore,
they do not require development of agriculture lands. The
lipid productivity of microalgae is reported to be ten times
higher than that of the best oil crop, which makes it a prom-
ising alternative [1,3]. In spite of their obvious advantages over
oil crops, microalgae-biodiesel production processes are not
yet commercialized.
Generally, microalgae can produce both neutral lipids,
composed mainly of triglycerides, and polar lipids such as
phospholipids, which commonly produced in cellular mem-
brane, whereas the former usually accumulated as droplets in
the cytoplasm [4]. The biodiesel production process from
microalgae consists of biomass cultivation, harvesting, dry-
ing, and lipid extraction, followed by the conversion of
extracted lipids to biodiesel, and nally purication of the
produced biodiesel.
Among the main challenges in the biodiesel production
from microalgae are separating the cultivated biomass from
the growth media and extracting lipids from the harvested
biomass. The biomass concentration of Scenedesmus is re-
ported to be low in outdoor cultures, reaching 0.02e0.12 g l
1
[5] and increases to about 0.3e3.4 g l
1
[6] in photobioreactors.
The currently used harvesting methods are centrifugation,
sedimentation and occulation. These methods have been
proven to be effective but are costly, representing 20e35% of
total biomass production cost [7].
On the other hand, a drying step is usually required prior to
lipid extraction from the harvested microalgae cells. Sun
drying is the most commonly used method, as it directly uti-
lizes the solar energy. However, it is a time consuming process
and the drying rate remains the main challenge of such pro-
cess. Other faster processes are energy intensive and/or can
alter the lipid structure and the protein rich leftover biomass,
which affects their quality. It has been reported that the dry-
ing step accounts for 89% of the required energy input [8], and
70e75% of total processing cost [9]. Thus, it is considered as
the major bottleneck of algae based biodiesel production [10].
Therefore, positive net energy from microalgae biodiesel
could be obtained if wet extraction is used. It would be
economically favorable to avoid the drying step while main-
taining an effective lipid extraction from the wet biomass.
Indeed, it has been estimated that the energy required to
produce 1 kg of biodiesel from dried biomass is 4000 times
more than that from wet biomass [11]. This requires expen-
sive and high energy consumption in both the up- and down-
stream processes in order to get high quality biodiesel.
It is vital therefore to develop a cost and energy effective
processes, which is green and environmental friendly, which
can overcome the main technical and economical barriers of
the conventional techniques. In addition, for microalgae uti-
lization process to be feasible, the extraction process should
not negatively affect the de-lipidated leftover biomass to allow
it to be used in useful applications, or its conversion to a
valuable product.
Several extraction methods can be used to extract the lipid
from dried microalgae biomass. Among the common used
techniques are Folch [12], Bligh and Dyer [13], and Soxhlet [14].
However, these techniques are complicated and time
consuming, requiring about 1e3 days for completion. In
addition, an extra solvent evaporation unit is required to
separate the extracted lipid from the solvent. Furthermore,
contamination of the de-lipidated with traces of solvent is
inevitable, which limits their further uses. Recently, super-
critical CO
2
(SC-CO
2
) received considerable attention due to its
tunable solvation power, mild operating conditions, and
environmental benign features. In addition, the process is
fast, and extraction that requires 24 h using conventional
techniques is completed with 30e60 min. This technique also
does not require solvent separation, as CO
2
is separated from
the extracted lipid by simple depressurization. Furthermore,
there is no contamination of the leftover biomass with the
extraction solvent.
On the other hand, the rigid and tough cell walls of
microalgae cells hinder the extraction of the cells lipids. These
cell walls are mainly composed of 24e74% neutral sugars,
1e24% uronic acids, 2e16% protein, and 0e15% glucosamine
[15]. In addition, the presence of water in wet harvested
biomass forms a lmpreventing the solvent fromreaching the
lipid, which further prevents efcient lipid extraction. Thus,
for an efcient extraction of the lipid, the microalgae cells
have to be disrupted to liberate the lipid and allow them to
come into contact with the solvent. This is conventionally
done using wet milling [16], ultrasonication [17,18], bead-
beating [14], microwaves [18], autoclaving at a high tempera-
ture and pressure, and osmotic disruption by treatment with
sodium chloride [19]. All these processes however, are energy
intensive and may affect the properties of the triglycerides
causing downstream difculties in their transesterication to
biodiesel. Inaddition, mechanical disruptionusually results in
excessive heat generation, which requires cooling.
Another technique, which is commonly used is acid
treatment [20]. Usually, this is performed by soaking the
biomass in diluted acid, commonly sulfuric acid, and then
heating it at high temperature for a certain time. In many
cases, this is followed by alkaline addition, commonly sodium
hydroxide, thus leading to pores swelling and decreasing in
cellulose crystallinity. Although this process is efcient in
lignocellulose degradation, sulfuric acid is toxic and corrosive,
which makes this process not recommended, neither for fuel
production nor for the left biomass applications. Freeze drying
has also been suggested [18], wherein harvested cells are
lyophilized, resulting in dried powder, prior to lipid extraction.
Lyophilization, however, is an energy intensive process that is
not justied in energy production processes.
For an-economical and effective lipid production, efcient
cell disruption, mild extraction conditions, and leftover
biomass utilization are essential. Some enzymes have the
potential to facilitate the cell disruption. They can operate at
low temperatures with high selectivity and fewer side prod-
ucts. Utilizing enzymes for cell disruption is expected to
enhance the efciency of lipid extraction at mild conditions.
Operation at mild conditions is less energy intensive and also
has a minimum effect on the triglycerides structure or the
leftover residual biomass that can be utilized in pharmaceu-
tical, food and fuel applications.
Several lysis enzymes can be used, such as cellulase that
can effectively hydrolyze the cellulosic structure of the cell
walls, and lysozyme that can hydrolyze the linkage between
peptidoglycan residues in the cell walls. Specically, it de-
grades polymers containing N-acetylglucosamine, which is a
b i oma s s a nd b i oe ne r g y 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 160
derivative of glucosamine, major component of cell wall. The
enzymes have been tested on Chlorella species and outer cell
wall disruption was observed [21e24]. In this work, the two
enzymes were tested to disrupt cell wall, and liberate the lipid,
which would enhance the extraction process.
To the best of our knowledge, enzymatic disruption com-
bined with SC-CO
2
extraction has not been reported in liter-
ature. The use of such process is less time consuming than
conventional solvent extraction processes, avoids the use of
toxic chemicals, and does not require solvent separation unit.
In addition, the process allows the utilization of the protein
rich leftover biomass.
In additionto the extraction challenges, the lipid content of
the microalgae cells is important. The overall effectiveness of
biodiesel production from microalgae depends on the lipid
productivity, which is a combined effect of biomass produc-
tivity and lipid content. The reported lipid content of micro-
algae is in the range of 20e30% (dry basis) when the
microalgae are grown under controlled conditions with suf-
cient nutrients. However, under stress conditions, microalgae
may accumulate larger lipid content [25e28]. The primary
stress applied to green microalgae is nitrogen deciency,
where accumulations of more than 50% (dry basis) have been
reported [1,29]. This is mainly due to the lack of nitrogen
required for protein synthesis, and the excess carbon from
photosynthesis is then diverted into lipid production pathway
[30].
The objective of this study is to assess the effectiveness of
using enzymes for cell disruption and compare it conven-
tional methods. In addition, investigate the possibility of
extracting lipids fromwet biomass using SC-CO
2
, avoiding the
high organic solvent consumption, the time consuming dry-
ing step and the energy intensive mechanical cell disruption.
In addition, lipid enhancement by nitrogen starvation, com-
bined with efcient biomass production is also studied to
conrm Scenedesmus sp. applicability for biodiesel production.
The success of this work adds great enhancement to the
extraction process, which in turn has a positive effect on the
overall algae biodiesel production process.
2. Materials and methods
2.1. Chemicals and enzymes
n-Hexane, methanol, acetone, sulfuricacid, sodiumhydroxide,
HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid),
14% boron tetraouridemethanol mixture, Nile Red (9-
diethylamino-5-benzo[a] phenoxazinone), dimethyl sulfoxide
(DMSO), were purchased from SigmaeAldrich Inc., USA. Liq-
ueed CO
2
with a purity of 99.95% was supplied by Abu-Dhabi
Oxygen Company, UAE. Ultra high purity helium and zero-air
were supplied by Air Product Company, UAE. Lysozyme from
chicken egg white (activity > 40,000 U mg
1
, according to the
suppliers denition) and cellulase from Trichoderma long-
ibrachiatum (activity 1.0 U mg
1
, according to the suppliers
denition) were purchased fromSigmaeAldrichInc., USA, and
stored below 8

C and above 0

C according to the suppliers
instructions, respectively. Standard of high purity fatty acid
methyl esters (FAMEs), used for gas chromatography (GC)
calibration and fatty acid compositional analysis, comprised
of; 3.9% myristic acid methyl ester (C14:0), 9.9% palmitic acid
methyl ester (C16:0), 6.0%stearic acid methyl ester (C18:0), 10%
elaidic acid methyl ester (C18:1), 24.8%cis-9-oleic methyl ester
(C18:1), 36.1% linoleic acid methyl ester (C18:2n6c), 1.9% lino-
lelaidic acid methyl ester (C18:2), 2.1% arachidic acid methyl
ester (C20:0), and 2.1% behenic acid methyl ester (C22:0) were
obtained from Sigma Aldrich, USA.
2.2. Strains and culture medium
Scenedesmus sp. culture was obtained fromAlgal Oil Company,
Philippines, and cultivated in a modied bassel medium(3N-
BBM) for two weeks, followed by three weeks of stressing in
nitrogen-deprivation medium (N-BBM). The modied me-
dium, 3N-BBM, comprised of (in mM); 8.82 NaNO
3
, 0.17
CaCl
2
$2H
2
O, 0.3 MgSO
4
$7H
2
O, 1.29 KH
2
PO
4
, 0.43 K
2
HPO
4
, 0.43
NaCl, 1 ml l
1
of Vitamine B
12
, and 6 ml l
1
of P-IVsolution that
consisted of 2 Na
2
EDTA$2H
2
O, 0.36 FeCl
3
$6H
2
O, 0.21
MnCl
2
$4H
2
O, 0.37 ZnCl
2
, 0.0084 CoCl
2
$6H
2
O, and 0.017
Na
2
MoO
4
$2H
2
O. In N-BBM, the nitrate source (NaNO
3
) was
removed.
2.3. Growth experiments
For an economical biodiesel production frommicroalgae, high
biomass productivity and lipid content are important. These
two factors are difcult to achieve simultaneously, as condi-
tions favoring high biomass productivity usually result in low
lipid accumulation, and vice versa. To overcome this, a two
stage cultivation approach has been used, wherein the rst
stage, the cells were allowed to grow in a nutrient rich me-
dium (3N-BBM) for two weeks to enhance the biomass pro-
ductivity, and in the second stage cells in their exponential
growth phase were transferred to a nitrogen-decient me-
dium (N-BBM) for three weeks to enhance the lipid accu-
mulation. The cultivations in both stages were done in a 5 l
bubble column photobioreactor with an internal illumination.
All cultivations in this work were autotrophic, with CO
2
naturally present in air bubbled through the system being the
sole carbon source.
Prepared media were sterilized in autoclave (Hirayama HV-
50, Japan) at 121

C for 15 min and cooled to roomtemperature

prior to use. The microalgae was rst grown in 200 ml of 3N-
BBM placed in 500 ml Erlenmeyer asks with ltered air
bubbling at a constant temperature of 25 1

C under light
intensity of 75 mmol m
2
s
1
determined using a light meter
(model 472990, Extech Instruments, Massachusetts) with an
initial cell concentration of 3.2 10
4
cells ml
1
. Details about
cell concentration determination are mentioned later. After 11
days of growth, the culture was transferred to the photo-
bioreactor, where the growth was monitored with time. The
photobioreactor was illuminated with one 50 cm, 60 W, white
uorescent light at a light intensity of 120 mmol m
2
s
1
,
measured using the light meter under 12 h light/dark photo-
period automatically controlled by 24 h timer (S2402, China).
The photobioreactor had an outer diameter of 10 cm, an inner
diameter of 5 cm and 40 cm height.
The cell growth was monitored daily by measuring the
optical density at 680 nm using a spectrophotometer
b i o ma s s a nd b i o e ne r gy 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 161
(Shimadzu UVe1800 UV, Kyoto, Japan). The samples were
measured twice and the average values were considered. The
cell concentration (cells ml
1
) at any given cultivation time
was calculated from a pre-prepared calibration curve of the
optical density at 680 nm versus cell concentration deter-
mined using Neubauer Hemocytometer, placed on a micro-
scope (Eclipse LV100 Pol, Nikon, Japan). The dry weight of algal
biomass was also determined by ltering the algal suspension
using a, prewashed and dried Whatman lter paper, dried
overnight at 105

C in an oven (Memmert, Germany) until
constant weight.
On day 14, while cells were still in their exponential growth
phase, they were allowed to settle at the bottom of the pho-
tobioreactor, and were then concentrated by centrifugation at
3000 r min
1
for 15 min using multispeed centrifuge (IEC CL31,
Thermo Scientic, USA), and washed twice with distilled
water. Concentrated microalgae cells, after centrifugation,
were cultivated back in a similar photobioreactor in N-BBM
medium to enhance the lipid accumulation. Samples were
collected at regular intervals, centrifuged and re-diluted in
distilled water to obtain a 4 ml of cell suspension. For lipid
testing, the accumulations were monitored by staining the
constant concentration samples (1.5 10
6
cells ml
1
) with Nile
Red that emits a yellow uorescent signal in the presence of
the lipid, and the uorescents were visualized using uores-
cence microscope (Olympus). The Nile Red stock solution was
prepared as described by Siaut et al. [31], by dissolving 0.1 mg
of Nile Red in 1 ml acetone, and the solution was stored in the
dark at 4

C. Culture samples (500 ml) were placed in an
eppendorf tube, span in a centrifuge (Sigma 113, Germany) for
30 s at 4000 r min
1
and 410 ml of the supernatant were taken.
DMSO (10 ml) was then added to promote the accessibility of
Nile Red into the cells. The culture was then vortexed and 1 ml
of Nile Red solution was added followed by 20 min incubation
in the dark. The lipid accumulations were then quantied
using Multi-label Plate Reader (PerkineElmer, Boston) with
black 96-well plates. Fluorescences were measured before and
after Nile Red staining and for Nile Red stained N-BBM me-
dium. The intensity was considered after subtracting the
stained medium and sample before staining intensities from
the stained sample intensity at excitation and emission
wavelengths of 485 and 590 nm, respectively.
2.4. Cells disruptions
Wet cell biomass samples (1 g containing 6.8% solids) were
subjected to three different disruption methods, namely: (1)
lyophilization, (ii) enzymatic disruption and (iii) acid treat-
ment. The lyophilization was carried out in a freeze drier
(Telstar, Terrassa, Spain) operated at 54

C and 0.02 mbar for
6 h. The enzymatic pre-treatments were carried as follow;
3.25 ml of 10 mg ml
1
of enzymatic (lysozyme or cellulase)
solution was added to the wet biomass in the presence of
7.5 ml HEPES buffer solution (pH 7.48). This corresponds to
lysozyme and cellulase loadings of 1.92 10
4
and 0.48 U mg
1
dry biomass, respectively, with 6.78 g l
1
biomass loading. The
mixtures were then incubated in SI-300 Shaker at 37

C and
100 r min
1
for 30 min. The acid treatment was done by adding
1 ml of sulfuric acid (1 M) to the wet biomass and heated to
90

C for 30 min in shaking water bath (LabTech, Daihan
LabTech Co., Ltd., Korea), followed by addition of 1 ml sodium
hydroxide (5 M) solution and further incubation at 90

C for
30 min [20].
2.5. Lipid extraction
Microalgae lipids were extracted using two extraction sol-
vents, namely n-hexane and SC-CO
2
. In the static n-hexane
extraction, where no continuous ow or circulation of the
extraction solvent was employed compared to Soxhlet, 30 ml
of n-hexane was added to 1 g (containing 6.8% solids) of the
disrupted biomass (wet basis), incubated in an SI-300 Bench-
top Shaker at 50

C and 100 r min
1
overnight. The total lipid
content of the biomass was determined from lyophilized
biomass, as a base line, by Soxhlet extraction using n-hexane
for 8 h. The wet samples had a water content of 93.2% deter-
mined by drying a pre-weighed sample on a pre-dried What-
man lter paper and drying overnight at 60

C until a constant
weight was reached. For the enzymatic and acidic treated
samples, n-hexane solvent was added directly to the treated
sample, without removing the treating solutions. The mixture
was then centrifuged at 3000 r min
1
for 5 min. The upper n-
hexane layer containing the lipid was collected and the lipid
content was determined gravimetrically after solvent evapo-
ration. Same approach was carried out with lyophilized
samples, where lyophilized cells were extracted directly
without any additional cell disruption step, as lyophilization
can simultaneously dry and disrupt the cell. The extraction
was performed in a duplicate and average values were
considered.
The SC-CO
2
extractions were conducted at a pressure of
500 bar, temperature of 50

C and 3 ml min
1
owof solvent in
ISCO supercritical extraction unit (SFX-220, USA), which are
the optimal operating conditions determined in our previous
study [32]. The high pressure used was to ensure highest lipid
extraction. Unlike in the static n-hexane extraction, in the SC-
CO
2
extraction, separating the wet biomass from the enzyme
solution, prior to extraction, was required. After incubation in
the enzymatic solution for the desired time, the sample was
centrifuged at 3000 r min
1
for 5 min. The supernatant was
removed and residual treated biomass was collected. The
disrupted wet biomass was placed in 10 ml extraction cell
with glass wool placed in bothextractionvessel sides. CO
2
was
pressurized to the operation pressure using a high-pressure
syringe pump (Model 260D, ISCO, USA), and the pressurized
CO
2
was then heated to the operation temperature and
pumped into the extraction vessel. The extract was collected
in a vial after depressurization via micro-metering valve (HIP
15-12AF1-V), and lipid content was determined gravimetri-
cally after solvent evaporation. Further details about SC-CO
2
extraction procedure were reported elsewhere [32,33].
2.6. Fatty acid methyl ester (FAME) prole
To quantify the fatty acids present in the extracted lipid,
which affects the properties of the produced FAMEs and their
applicability to replace petroleumfuel, lipids were esteried to
FAME using 14% boron tetraourideemethanol mixture as
described by Rule [34]. FAME were identied and quantied by
GC (Varian, CP-3800, USA), tted with CP-Sil 88 FAME capillary
b i oma s s a nd b i oe ne r g y 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 162
column (100 m 0.25 mm 0.2 mm, Varian, USA), ame
ionization detector (FID) and equipped with auto-injector (CP
8410. Varian, USA). The oven initial temperature was held at
150

C for 1 min and then increased to 250

C at 4

C min
1
.
Helium and zero air were used as the carrier gases with a split
ratio of 40:1. Both the injector and the detector temperatures
were set at 260

C.
2.7. Statistical analysis
Experiments were conducted with at least duplicate treat-
ments. The data were analyzed using one-way (unstacked)
analysis of variance (ANOVA) followed by Fishers Least Sig-
nicant Differences (LSD). p-Value 5% was considered as
signicant, and data were presented as mean standard
deviation.
3. Results and discussion
3.1. Strain growth and productivity
Scenedesmus sp. cultivation was carried in two stages; in the
rst the cells were allowed to grow in a nutrient rich media
(3N-BBM) for 14 d to enhance the biomass productivity, and
in the second the cells were transferred to nitrogen decient
medium (N-BBM) to enhance the lipid accumulation. Fig. 1
shows the growth and the correspondence time course of
lipid accumulation in the two cultivation stages. The gure
shows that the lag phase was short, and the growth entered
the exponential growth phase almost immediately. As ex-
pected, the lag phase was short as the inoculum into the
photobioreactor was fromthe stock broth of cells grown in the
same medium. In stage 1, the specic growth rate was
0.195 d
1
, determined from the slope of the logarithmic
growth line (not shown here). The lipid content was deter-
mined at the beginning and end of this stage and found to
remain almost constant at 12.6 0.8% (dry basis). The lipid
content was determined after cells lyophilization at 54

C
and 0.02 mbar for 6 h in the freeze drier followed by SC-CO
2
extraction at 50

C, 500 bars and 3 ml min
1
with 100 ml total
CO
2
passed. Multiplying the biomass productivity, determined
fromthe slope of the growth curve, by lipid content resulted in
overall lipid productivity of 19.5 mg l
1
d
1
. Similar lipid
content was also reported in the work of Rodol et al. [35] done
on a strain of a genus similar to the one used in this work. The
exact species of the strainused in this work was not identied.
However, slightly higher biomass productivity was obtained
by Rodol et al. [35], which resulted in higher overall lipid
productivity. This is mainly because the cultivation was done
with 5%CO
2
enrichment, whereas in this work, the cultivation
was autotrophic with CO
2
being the sole carbon source. In
addition, this slightly higher productivity is due to the pref-
erable illumination intensity and duration, which were in this
work 120 mmol m
2
s
1
under 12 h light/dark photoperiod,
respectively, compared to 100 mmol m
2
s
1
under continuous
illumination applied in Ref. [35]. These results are also com-
parable to those found by Ren et al. [36] for the same strain but
grown heterotrophically.
3.2. Effect of nitrogen stressing on lipid productivity
In the second stage, cells were grown in a nitrogen decient
medium (N-BBM), which resulted in a sharp drop in the
specic growth rate to 0.0165 d
1
, as shown in Fig. 1. The slight
drop in biomass concentration observed at the beginning of
the second stage was due to loss of some cells during the
harvesting, after collecting the cells and re-culturing them in
the new medium. However, having grown in stage 1 to a
sufcient concentration, the accumulation of lipid was the
main objective of this stage. This reduces the net energy and
increases biodiesel productivity [8]. Real outdoor cultivations
may have lower growth rates due to self-shading and other
effects. However, even if lower growth rates are, the proposed
technique will still be applicable. The total lipid content in the
rst day of starvation was 12.6% based on dried cell weight.
The lipid accumulation was monitored by staining with Nile
Red. Fig. 2 shows the uorescence microscopic images in
different days after starvation, where the brightness in the
photos indicates the amount of lipids. It is clearly shown that
by day 11, cells started to accumulate the lipid, which
enhanced further until day 23. This has been also proven by
quantitative measurements using micro-plate reader as
shown in Fig. 3. The lipid content was also determined
gravimetrically by Soxhlet extraction from lyophilized sam-
ples and a calibration curve between the uorescence in-
tensity and lipid content was prepared. The results shown in
Fig. 3 further conrmed that lipid accumulation started after
day 11 of starvation and increased exponentially thereafter.
After 3 weeks of starvation, a six-fold increase in lipid content
Fig. 1 e Cell growth (CD3N-BBM and B LN-BBM) and lipid
content (-) curves of Scenedesmus sp. grown in 5 l
photobioreactor
b i o ma s s a nd b i o e ne r gy 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 163
was observed, reaching 73% on a dry weight basis. The pro-
posed two step process results in lowering the harvesting and
lipid extraction costs.
Table 1 shows the FAME compositions of extracted lipid,
after SC-CO
2
extraction, rapid transesterication to FAME, and
analysis using GC-FID, from the growth in nitrogen sufcient
(3N-BBM) and nitrogen decient (N-BBM) media. The re-
sults show that the compositions of the extracted lipid were
almost similar, but the triglycerides content (fatty acid con-
tents) was higher in the lipid extracted from cells grown in
N-BBM conrming that the natural lipid are accumulated by
starvation.
3.3. Pretreatment effect on lipid extraction yield
The use of chemical solvents, such as n-hexane, has several
drawbacks, which include the leftover biomass contamina-
tion with the solvent, long extraction time and the need of
additional separation units. These drawbacks can be over-
come by using SC-CO
2
extraction. It was also assumed that
operating at the supercritical conditions may allow the sol-
vent (CO
2
) to penetrate through the water lm and reach the
lipid. Therefore, SC-CO
2
was tested with both lyophilized and
wet samples. The results, shown in Fig. 4, indicated that the
extraction yield from wet and lyophilized samples was close
to those obtained using n-hexane. The reproducibility of the
results was conrmed by doing the experiments in at least in a
duplicate, and the results presented in Fig. 4 are the average
values with the error bars representing the respective stan-
dard deviations.
As mentioned earlier, drying is a time consuming step and
in many cases an energy intensive process. Avoiding this step
would signicantly enhance the microalgae-biodiesel pro-
duction process. Thus, in this study, at rst, the lipid
extracted from lyophilized cells was compared to that
extracted from wet biomass. Soxhlet extraction from lyoph-
ilized biomass using n-hexane was used to determine the
total lipid content in the biomass. It was found that the total
lipid content was 21.1 1.5%. The lipid extraction was also
performed using static n-hexane, as a solvent and left over-
night in an incubator at 50

C. Fig. 4 shows the lipid extrac-
tion yield expressed as percentage of lipid extracted per dry
biomass weight. As shown in the gure, lipid extraction yield
from wet sample was only 4% compared to 12.6% achieved
from lyophilized sample. The lower lipid extraction yield
from the wet sample is due to the water lm formed over the
lipid, which prevents the solvent from contacting it. In
addition to the drying effect that enhanced the extraction
yield, lyophilization also disrupts algal cell and makes cell
walls more porous [37].
3.4. Lipid extraction from pre-treated wet biomass
Microalgal cell wall rigidity and toughness is the main barrier
to microalgal lipid extraction, and cell walls need to be dis-
rupted in order to enhance the extraction. The main param-
eters that determine disruption method suitability are process
cost, scalability and product contamination with other cell
components. Lyophilization is the most commonly used
technique, and as shown in Fig. 4, the lipid yield increased
from only 4% using wet biomass to 12.6% from lyophilized
sample. However this technique requires high energy to
freeze the samples and its operation and maintenance costs
are relatively high, which is usually not justied in energy
production processes.
Acid treatment has also been suggested to disrupt the cells
and liberate the lipid [20]. Pretreatment using diluted sulfuric
acid was tested and resulted in a lipid extraction yield of 10%.
However, using acids is neither recommended in fuel pro-
duction nor for the leftover biomass applications. In addition,
using acids require special materials of construction, which is
not economical for large-scale applications.
Fig. 2 e Fluorescence microscopy images of Scenedesmus sp. cells stained with Nile Red and showing lipid accumulation
after 1 (a), 11 (b), 14 (c), 20 (d) and 23 (e) days of nitrogen starvation in cultures. The brightness in the photos indicates the
amount of lipids.
Fig. 3 e Relative uorescence intensity (C) determined
using micro-plate reader and lipid content in dry weight
basis (B) in the nitrogen starvation stage.
b i oma s s a nd b i oe ne r g y 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 164
To test enzymatic treatment, two enzymes have been
tested, namely lysozyme, naturally used to disrupt bacterial
cell walls, and cellulase, which catalyzes the hydrolysis of
cellulose. The latter was tested to disrupt the cellulosic
structure of the cell walls. Fig. 4 shows that lipid extraction
yields from wet samples using both enzymes, separately,
resulted in a better yield than even the lyophilized samples,
with values of 16.6 and 15.4% using lysozyme and cellulase,
respectively. This is a signicant nding as it shows that using
either enzyme, with lysozyme showing slightly better results,
lipid can be extracted from wet sample without the need for
the time consuming and energy intensive drying step. The
result of this work clearly shows the superiority of using
lysozyme, or cellulase, in the extraction of lipid from wet
microalgae, while avoiding the drying step. The synergic effect
of both enzymes was then tested using 50% mixture of lyso-
zyme and cellulase; however lower yield of 12% was achieved.
This could be due to the formation of a product from one
enzymatic reaction that inhibits the other enzymatic reaction.
Cellulase, from the same source as the one used in this study,
was also tested by Liang et al. [38] to assist lipid extraction
from pre-ultrasonicated wet cells of Chlorella vulgaris and
compared to extraction from lyophilized cells using Bligh and
Dyer method. It was found that the lipid extraction yield
increased with the addition of the enzyme. However the
highest lipid recovery (lipid extracted to actual lipid content)
was 49%, which was signicantly lower than the one obtained
in this study, which reached 72%. In addition, the solid con-
centration was signicantly higher, where the solids were
concentrated to 18% compared to only 6.8% in this work.
Furthermore, Liang et al. ultra-sonicated the lyophilized
biomass, which was shown in this work to be not required.
The one-way ANOVA proved that there is a signicant
difference in the yield with p-value equal to 0.038 between
treatments and the Fishers (95% condence intervals) multi-
ple comparison demonstrated that the yield of individual
enzyme use is close, while the enzyme mixture yield is not.
Although the use of the enzymes has shownto be effective,
their industrial implementation for low-value products is
often limited due to their high cost. It is worth mentioning
though that the price of lysozyme is relatively not expensive,
compared to other enzymes [39]. In addition, it can be
repeatedly used for many cycles without any signicant lose
of activity when used in an immobilized forms. Lysozyme
immobilized on extrudate-shaped NaY zeolite has been suc-
cessfully used for 12 cycles [40]. Immobilization of the enzyme
also has the advantage of enhancing the recovery and puri-
cation steps.
Due to SC-CO
2
advantages over conventional solvent
extraction, the application of enzymatic treatment with SC-
CO
2
extraction fromwet cells using lysozyme has been tested.
The lipid extraction yield was 12.5%, which is almost 75% of
that extracted using n-hexane. The signicant difference in
the yield was also conrmed by one way ANOVA, where a p-
value of 0.033 was obtained. This relatively lower yield can be
attributed to some free lipid that may have been lost in the
enzymatic aqueous solution during the disruption step after
centrifugation. This did not occur with n-hexane extraction,
as the solvent was added before the centrifugation, which was
not possible in the SC-CO
2
extraction. Nevertheless, due to its
T
a
b
l
e
1
e
F
a
t
t
y
a
c
i
d
c
o
m
p
o
s
i
t
i
o
n
(
%
w
/
w
)
o
f
t
h
e
l
i
p
i
d
e
x
t
r
a
c
t
e
d
u
s
i
n
g
S
C
-
C
O
2
a
n
d
n
-
h
e
x
a
n
e
w
i
t
h
d
i
f
f
e
r
e
n
t
d
i
s
r
u
p
t
i
o
n
m
e
t
h
o
d
s
.
F
a
t
t
y
a
c
i
d
n
-
H
e
x
a
n
e
e
x
t
r
a
c
t
i
o
n

d
S
C
-
C
O
2
e
x
t
r
a
c
t
i
o
n

d
L
y
o
p
h
i
l
i
z
a
t
i
o
n
a
W
e
t
A
c
i
d
t
r
e
a
t
m
e
n
t
L
y
s
o
z
y
m
e
C
e
l
l
u
l
a
s
e
L
y
o
p
h
i
l
i
z
a
t
i
o
n
a
W
e
t
L
y
s
o
z
y
m
e
L
y
o
p
h
i
l
i
z
a
t
i
o
n
b
C
1
4
:
0
3
.
2
0

0
.
4
6
.
1
0

0
.
8
7
.
5
0

0
.
5
7
.
5

0
.
7
6
.
0
0

0
.
4
3
.
5
0

0
.
1
5
.
7
0

0
.
6
7
.
0
0

0
.
2
3
.
2
0

0
.
0
5
C
1
6
:
0
2
1
.
0

0
.
0
9
1
2
.
2

0
.
5
1
5
.
1

0
.
2
1
6
.
4

0
.
0
8
1
1
.
9

0
.
3
1
6
.
5

0
.
4
1
4
.
3

0
.
3
1
8
.
6

0
.
2
1
7
.
4

0
.
7
C
1
8
:
0
8
.
1
0

0
.
5
8
.
2
0

0
.
8
1
1
.
3

0
.
4
1
3
.
4

0
.
1
1
3
.
1

0
.
7
4
.
7
0

0
.
6
5
.
7
0

0
.
9
9
.
3
0

0
.
6
6
.
3
0

0
.
4
C
1
8
:
1
1
2
.
9

0
.
1
2
.
0
0

0
.
9
e
6
.
0

0
.
7
4
.
8
0

0
.
9
1
1
.
8

0
.
5
4
e
4
.
7
0

0
.
1
1
2
.
1

0
.
9
C
1
8
:
1
(
t
r
a
n
s
-
9
)
8
.
1
0

0
.
4
1
2
.
2

0
.
7
9
.
4
0

0
.
7
7
.
5

0
.
9
6
7
.
1
0

0
.
4
9
.
4
0

0
.
1
1
1
.
4

0
.
4
e
8
.
2
0

0
.
5
C
1
8
:
2
1
.
6
0

0
.
6
2
.
0
0

0
.
0
3
1
.
9
0

0
.
2
4
.
5

0
.
8
6
.
0
0

0
.
8
e
2
.
9
0

0
.
5
7
.
0
0

0
.
4
0
.
6
0

0
.
7
C
1
8
:
2
(
n
6
)
4
3
.
5

0
.
9
5
3
.
1

0
.
7
5
0
.
9

0
.
2
4
1
.
8

0
.
1
4
5
.
2

0
.
9
4
4
.
7

0
.
6
5
4
.
3

0
.
1
4
1
.
9

0
.
2
4
0
.
7

0
.
6
C
1
8
:
3
(
n
3
)
e
4
.
1
0

0
.
4
e
e
2
.
4
0

0
.
1
7
.
1
0

0
.
2
2
.
9
0

0
.
3
4
.
7
0

0
.
7
8
.
4
0

0
.
8
C
2
0
:
0
e
e
1
.
9
0

0
.
8
e
3
.
6
0

0
.
4
e
2
.
9
0

0
.
6
4
.
7
0

0
.
3
1
.
2
0

0
.
6
C
2
2
:
0
1
.
6
0

0
.
1
e
1
.
9
0

0
.
1
3
.
0

0
.
6
e
2
.
4
0

0
.
4
e
2
.
3
0

0
1
1
.
9
0

0
.
0
2
T
o
t
a
l
f
a
t
t
y
a
c
i
d
s
8
4
.
3
7
6
.
9
7
9
.
7
8
2
.
2
7
9
.
0
6
9
.
6
7
3
.
5
7
6
.
1
7
8
.
2
U
/
S
c
2
.
0
2
.
8
1
.
7
1
.
5
1
.
9
2
.
7
2
.
5
1
.
4
2
.
3
a
E
x
t
r
a
c
t
i
o
n
f
r
o
m
l
y
o
p
h
i
l
i
z
e
d
c
e
l
l
g
r
o
w
n
i
n

3
N
-
B
B
M
.
b
E
x
t
r
a
c
t
i
o
n
f
r
o
m
l
y
o
p
h
i
l
i
z
e
d
c
e
l
l
g
r
o
w
n
i
n

N
-
B
B
M
.
c
U
n
s
a
t
u
r
a
t
i
o
n
r
a
t
i
o
.
d
T
h
e
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
.
b i o ma s s a nd b i o e ne r gy 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 165
favorable features, the successful use of SC-CO
2
with wet
biomass is still a major nding.
Both operation time and cost are critical parameters in
process commercialization. SC-CO
2
approach has a shorter
extraction time with less solvent consumption compared to
organic solvents extraction. Although the operation cost
might be higher, associated to the pumping cost, this will be
counterbalanced by lower down-streaming cost, as no solvent
recovery unit is required and the solvent can be separated by
simple depressurizing and recycled back.
Using this environmental friendly and recycled approSC-
CO
2
was also tested with lyophilized biomass and compared
to n-hexane. It was found that the difference in yield was
insignicant compared to n-hexane (p-value 0.724).
3.5. Analysis of extracted lipid
Complete analysis of the fatty acid composition of the
extracted lipid by n-hexane and SC-CO
2
was carried out using
GC-FID analysis, as shown in Table 1. Mass fractions are
normalized according to the total fatty acids found in the GC
analysis. It is clearly shown that the extracts were mainly
composed of linoleic acid (42e55%), palmitic acid (12e19%)
and oleic acid (7e12%). The obtained unsaturation ratio was in
the range of 1.5e3. No signicant change in composition was
observed by using different treatments and extraction tech-
niques. However, the total fatty acid of the lipid extracted by
SC-CO
2
was slightly lower in all treatments, which is mainly
due to the solubility of other pigments in SC-CO
2
. When
considering the extract yield, quality, costs and environ-
mental impacts of using chemical solvents, the enzymatic
disruption followed by SC-CO
2
extraction could probably be
the most efcient for biodiesel production from wet biomass.
This is especially true if the leftover, protein rich, biomass
after lipid extraction is used in pharmaceutical or food
industries.
4. Conclusion
The work looked at the possibility of extracting lipid from
microalgae, while avoiding the time consuming and energy
intensive drying step. It was shown that by using enzymatic
cell disruption, using lysozyme or cellulase, the lipid extrac-
tion was achieved from the wet sample that contained 93%
water. The highest lipid extraction yield of 16.6%was obtained
from enzymatically disrupted cells using lysozyme. The SC-
CO
2
extraction was also successful in extracting lipid from
enzymatic disrupted biomass, but with a slightly less yield
compared to n-hexane. In addition, the study also shows that
lipid content was enhanced in nitrogen starvation medium.
Acknowledgment
The authors would like to express their sincere appreciation to
Prof. Kourosh Salehi-Ashtiani, Associate Professor of Biology
at New York University Abu Dhabi, for his valuable assistance
in visualizing the lipid accumulationof the microalgae strains.
r e f e r e n c e s
[1] Chisti Y. Biodiesel from microalgae. Biotechnol Adv 2007;294.
[2] Maggio G, Cacciola G. When will oil, natural gas, and coal
peak? Fuel 2012;111.
[3] Ahmad AL, Yasin NHM, Derek CJC, Lim JK. Microalgae as a
sustainable energy source for biodiesel production: a review.
Renew Sustain Energy Rev 2011;584.
[4] Olmstead ILD, Hill DRA, Dias DA, Jayasinghe NS, Callahan DL,
Kentish SE, et al. A quantitative analysis of microalgal lipids
for optimization of biodiesel and omega-3 production.
Biotechnol Bioeng 2013;2096.
[5] Makareviciene V, Andrulevi ci ut _ e V, Skorupskait _ e V,
Kasperovi cien_ e J. Cultivation of microalgae Chlorella sp. and
Scenedesmus sp. as a potential biofuel feedstock. Environ Res
Eng Manage 2011;21.
[6] Sa nchez JF, Ferna ndez-Sevilla JM, Acie n FG, Cero n MC, Pe rez-
Parra J, Molina-Grima E. Biomass and lutein productivity of
Scenedesmus almeriensis: inuence of irradiance, dilution rate
and temperature. Appl Microbiol Biotechnol 2008;719.
[7] Gudin C, Thepenier C. Bioconversion of solar energy into
organic chemicals by microalgae. Adv Biotechnol Process
1986;73.
[8] Sander K, Murthy G. Life cycle analysis of algae biodiesel. Int
J Life Cycle Assess 2010;704.
[9] Mohn HF. Improved technologies for harvesting and
processing of microalgae and their impact on production
costs. Arch Hydrobiol Beih Ergeb Limnol 1978;228.
[10] Razon LF, Tan RR. Net energy analysis of the production of
biodiesel and biogas from the microalgae: Haematococcus
pluvialis and Nannochloropsis. Appl Energy 2011;3507.
Fig. 4 e Yield of extracted lipid by n-hexane and SC-CO
2
with different treatments.
b i oma s s a nd b i oe ne r g y 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 166
[11] Lardon L, He lias A, Sialve B, Steyer J, Bernard O. Life-cycle
assessment of biodiesel production from microalgae.
Environ Sci Technol 2009;6:475.
[12] Folch J, Lees M, Sloane Stanley GH. A simple method for the
isolation and purication of total lipids from animal tissues. J
Biol Chem 1957;497.
[13] Bligh EG, Dyer WJ. A rapid method of total lipid extraction
and purication. Can J Biochem Physiol 1959;911.
[14] Ryckebosch E, Muylaert K, Foubert I. Optimization of an
analytical procedure for extraction of lipids from microalgae.
J Am Oil Chem Soc 2012;189.
[15] Blumreisinger M, Meindl D, Loos E. Cell wall composition of
chlorococcal algae. Phytochemistry 1983;1603.
[16] ShenY, Pei Z, YuanW, Mao E. Effect of nitrogenandextraction
method on algae lipid yield. Int J Agric Biol Eng 2009;51.
[17] Suarsini E, Subandi S. Utilization ultrasonic to increase the
efciency of oil extraction for microalgae indigenous isolates
from pond gresik, east java. In: Clean Energy and Technology
(CET), 2011 IEEE First Conference on; 2011. p. 275.
[18] Sostaric M, Klinar D, Bricelj M, Golob J, Berovic M, Likozar B.
Growth, lipid extraction and thermal degradation of the
microalga Chlorella vulgaris. Nat Biotechnol 2012;325.
[19] Lee J-Y, Yoo C, Jun S-Y, Ahn C-Y, Oh H-M. Comparison of
several methods for effective lipid extraction from
microalgae. Bioresour Technol 2010;S75.
[20] Sathish A, Sims RC. Biodiesel from mixed culture algae via a
wet lipid extraction procedure. Bioresour Technol 2012;643.
[21] Cho H-S, Oh Y-K, Park S-C, Lee J-W, Park J-Y. Effects of
enzymatic hydrolysis on lipid extraction from Chlorella
vulgaris. Renew Energy 2013;156.
[22] Chuchird N, Hiramatsu S, Sugimoto I, Fujie M, Usami S,
Yamada T. Digestion of Chlorella cells by chlorovirus-encoded
polysaccharide degrading enzyme. Microbes Environ
2001;206.
[23] Fu C-C, Hung T-C, Chen J-Y, Su C-H, Wu W-T. Hydrolysis of
microalgae cell walls for production of reducing sugar and
lipid extraction. Bioresour Technol 2010;8750.
[24] Gerken HG, Henri GG, Bryon D, Eric PK. Enzymatic cell wall
degradation of Chlorella vulgaris and other microalgae for
biofuels production. Planta 2013;239.
[25] Zemke PE, Wood BD, Dye DJ. Considerations for the
maximum production rates of triacylglycerol from
microalgae. Biomass Bioenergy 2010;145.
[26] Liu Z-Y, Wang G-C, Zhou B-C. Effect of iron on growth and
lipid accumulation in Chlorella vulgaris. Bioresour Technol
2008;4717.
[27] Converti A, Casazza AA, Ortiz EY, Perego P, Del Borghi M.
Effect of temperature and nitrogen concentration on the
growth and lipid content of Nannochloropsis oculata and
Chlorella vulgaris for biodiesel production. Chem Eng Process
2009;1146.
[28] Hsieh C-H, Wu W-T. Cultivation of microalgae for oil
production with a cultivation strategy of urea limitation.
Bioresour Technol 2009;3921.
[29] Tornabene TG, Holzer G, Lien S, Burris N. Lipid composition
of the nitrogen starved green alga Neochloris oleoabundans.
Enzyme Microb Technol 1983;435.
[30] Amaro HM, Guedes AC, Malcata FX. Advances and
perspectives in using microalgae to produce biodiesel. Appl
Energy 2011;3402.
[31] Siaut M, Cuine S, Cagnon C, Fessler B, Nguyen M, Carrier P,
et al. Oil accumulation in the model green alga
Chlamydomonas reinhardtii: characterization, variability
between common laboratory strains and relationship with
starch reserves. BMC Biotechnol 2011;7.
[32] Taher H, Al-Zuhair S, AlMarzouqi A, Haik Y, Farid M, Tariq S.
Supercritical carbon dioxide extraction of microalgae lipid:
process optimization and laboratory scale-up. J Supercrit
Fluids; 2014:86.
[33] Taher H, Al-Zuhair S, AlMarzouqui A, Hashim I. Extracted fat
from lamb meat by supercritical CO
2
as feedstock for
biodiesel production. Biochem Eng J 2011;23.
[34] Rule DC. Direct transesterication of total fatty acids of
adipose tissue, and of freeze-dried muscle and liver with
boron-triuoride in methanol. Meat Sci 1997;23.
[35] Rodol L, Chini Zittelli G, Bassi N, Padovani G, Biondi N,
Bonini G, et al. Microalgae for oil: strain selection, induction
of lipid synthesis and outdoor mass cultivation in a low-cost
photobioreactor. Biotechnol Bioeng 2009;100.
[36] Ren H-Y, Liu B-F, Chao Ma LZ, Ren N-Q. A new lipid-rich
microalga Scenedesmus sp. strain R-16 isolated using Nile red
staining: effects of carbon and nitrogen sources and initial
pH on the biomass and lipid production. Biotechnol Biofuels
2013;143.
[37] Wiltshire K, Boersma M, Moller A, Buhtz H. Extraction of
pigments and fatty acids from the green alga Scenedesmus
obliquus (Chlorophyceae). Aquat Ecol 2000;119.
[38] Liang K, Zhang Q, Cong W. Enzyme-assisted aqueous
extraction of lipid from microalgae. J Agric Food Chem
2012;11771.
[39] Kubiak-Ossowska K, Mulheran PA. Mechanism of hen egg
white lysozyme adsorption on a charged solid surface.
Langmuir 2010;15954.
[40] Chang YK, Chu L. A simple method for cell disruption by
immobilization of lysozyme on the extrudate-shaped NaY
zeolite. Biochem Eng J 2007;37.
b i o ma s s a nd b i o e ne r gy 6 6 ( 2 0 1 4 ) 1 5 9 e1 6 7 167