Bioresource Technology 118 (2012) 628632

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Short Communication

Switchable hydrophilicity solvents for lipid extraction from microalgae for

biofuel production
Alaina R. Boyd a, Pascale Champagne b,, Patrick J. McGinn c, Karen M. MacDougall c, Jeremy E. Melanson c,
Philip G. Jessop a,
a
Department of Chemistry, Queens University, Kingston, Ontario, Canada K7L 3N6
b
Department of Civil Engineering, Queens University, Kingston, Ontario, Canada K7L 3N6
c
National Research Council of Canada, Institute for Marine Biosciences, Halifax, Nova Scotia, Canada B3H 3Z1

h i g h l i g h t s

" We extracted lipids from microalgae using switchable hydrophilicity solvents (SHS).
" SHS extractions at room temperature yielded more intact triacylglycerols than at 80 C.
" Extractions using SHS do not require drying of the solvents prior to use.
" Carbon dioxide is used to separate the SHS from the lipids rather than high energy distillation.

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

Article history: A switchable hydrophilicity solvent (SHS) was studied for its effectiveness at extracting lipids from
Received 9 November 2011 freeze-dried samples of Botryococcus braunii microalgae. The SHS N,N-dimethylcyclohexylamine
Received in revised form 12 April 2012 extracted up to 22 wt.% crude lipid relative to the freeze-dried cell weight. The solvent was removed from
Accepted 18 May 2012
the extract with water saturated with carbon dioxide at atmospheric pressure and recovered from the
Available online 29 May 2012
water upon de-carbonation of the mixture. Liquid chromatographymass spectrometry (LCMS) showed
that the extracted lipids contained high concentrations of long chain tri-, di- and mono-acylglycerols, no
Keywords:
phospholipids, and only 48% of residual solvent. Unlike extractions with conventional organic solvents,
Switchable solvents
Microalgae
this new method requires neither distillation nor the use of volatile, ammable or chlorinated organic
Botryococcus braunii solvents.
Lipids 2012 Elsevier Ltd. All rights reserved.
Biofuel

1. Introduction
Oil derived from microalgae is an attractive alternative to petro-
leum. Microalgae have exceptionally high oil contents compared to
those of terrestrial plant sources such as palm, coconut, castor bean
and sunower seeds (Demirbas, 2010), and can be cultivated on
marginal land using non-potable water. Microalgae can be grown
under a variety of nutritional regimes (autotrophic, heterotrophic,
mixotrophic), and can make use of both inorganic and organic car-
bon sources for growth, in addition to absorbing other nutrients
like nitrogen and phosphorus from wastewater sources (McGinn
et al., 2011).
The economical production of bioenergy from microalgae is
limited in part by the energy costs of the extraction process.
Current extraction methods use volatile solvents that are ineffec-
Corresponding authors. Tel.: +1 613 533 3053; fax: +1 613 533 2128 (P.
Champagne), tel.: +1 613 533 3212; fax: +1 613 533 6669 (P.G. Jessop).
E-mail address: jessop@chem.queensu.ca (P.G. Jessop).
tive with microalgae with high water contents (Cooney et al.,
2009), thus the microalgae must rst be dried. After the extraction,
the organic solvents must be removed by distillation. The high en-
ergy cost associated with drying and the energy required for distil-
lation and recovery of the solvents currently make the use of
microalgae unfeasible. Additionally, the solvents normally selected
are volatile and either ammable or chlorinated, none of which are
desirable characteristics for health and environmental consider-
ations. However, with the application of switchable solvents the
energy barriers are greatly reduced, making the use of microalgae
as a bioenergy source more attractive.
With switchable solvents it is possible to recover extracted
material from the extracting solvent simply by the addition of car-
bon dioxide. Samor et al. (2010) showed the use of one class of
switchable solvents, switchable polarity solvents (SPS), for the
extraction of lipids from microalgae. These solvents make use of
the fact that a low polarity mixture of 1,8-diazabicyclo-[5.4.0]-un-
dec-7-ene (DBU) and an alcohol, becomes more polar when CO2 is
introduced, and returns to its original polarity when the CO2 is

0960-8524/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2012.05.084
 A.R. Boyd et al. / Bioresource Technology 118 (2012) 628632
removed (Jessop et al., 2005;Phan et al., 2008). While these sol-
vents were shown to extract oil from microalgae and be separable
from the oil without distillation, both the SPS and the algae had to
be rigorously dried to prevent the unwanted formation of the
bicarbonate salt of DBU.
Switchable hydrophilicity solvents (SHS) (Jessop et al., 2010,
2011) are a second class of switchable solvents specically de-
signed for use with water and wet samples. These solvents can
be reversibly switched between a hydrophobic form that is poorly
miscible with water, and a hydrophilic form that is miscible with
water. SHS such as certain tertiary amines are hydrophobic enough
to form a biphasic mixture when mixed with water under air, but
with the addition of CO2 at 1 atm they effectively become miscible
with water because they are converted to the bicarbonate salt. The
SHS system has several advantages over the SPS system. The ter-
tiary amines do not need to be dried prior to use as the SHS system
is designed to work in the presence of water. It is anticipated that
this property will improve extraction and separation of lipids from
wet algal cultures, but this is beyond the scope of this paper. The
tertiary amines are also not as expensive and do not suffer from
the hydrolytic instability of amidines. SHSwater miscibility is
fully reversible because, with the removal of CO2 through gentle
means (low heat, bubbling with an inert gas), they return to their
original hydrophobic state.
This preliminary paper aims to show that SHS are more attrac-
tive than conventional solvents or SPS as extraction solvents for
lipids from microalgae.
2. Methods
2.1. Materials
All chemicals were used as received. N,N-dimethylcyclohexyl-
amine 99% from Aldrich was used for the extractions. Carbon diox-
ide was obtained from Praxair: 4.0 grade (99.99%). Cultivation of
Botryococcus braunii (B. braunii) Race A UTEX 572 was conducted
as described by MacDougall et al. (2011). Cells were lyophilized
to a nal moisture content of 2 wt.%. Water used during the extrac-
tion was puried using a Millipore purication system. The amine
solvent was not dried before use. Triacylglycerol (TAG), diacylglyc-
erol (DAG), monoacylglycerol (MAG), and fatty acid standards were
purchased from Nu-Chek Prep (Elysian, MN, USA). The digalacto-
syldiacylglycerol (DGDG) standards DGDG(18:0/16:0) and
DGDG(18:0/16:0) were acquired from Matreya (Pleasant Gap, PA,
USA).
2.2. SHS extraction
Extractions were performed with varying solvent volumes and
at different temperatures, as well as with and without sonication,
with each set of conditions conducted in duplicate. A schematic
depiction of the SHS extraction is shown in Fig. 1. For each extrac-
tion, lyophilized B. braunii was weighed into a 4 dram vial and N,N-
dimethylcyclohexylamine (CyNMe2) was added. Samples 16 used
approximately 1 g of dried microalgae and 3 mL of CyNMe2, and
samples 710 used approximately 0.5 g of dried microalgae and
10 mL of CyNMe2.
The mixtures were either stirred or sonicated using a 130 watt
Fisher Scientic SF30 ultrasonication bath for 30 min, following
which the samples were stirred for 18 h at room temperature,
60 C or 80 C. Each sample was ltered through a Whatman #1 l-
ter paper using vacuum ltration and washed with fresh CyNMe2.
The ltrate was collected and an equal volume of water was added.
The ltrate/water biphasic mixture was subjected to CO2 bubbling
using a gas dispersion tube until the amine and water layers com-
629
bined, leaving the lipid layer oating at the surface. In preparation
for analysis, the lipids were dissolved in chloroform or hexanes and
transferred to a new ask. The solvent was removed at reduced
pressure.
The hydrophilic SHS/water mix was separated into its respec-
tive amine and aqueous layers by bubbling with N2 and heating
until the two layers were resolved.
2.3. Chloroform/methanol extraction
Dry algae (300 mg) were extracted using a 2:1 (v:v) chloro-
form:methanol mixture overnight by Soxhlet extraction at 80 C.
The solvent was removed under reduced pressure.
2.4. Solid phase extraction (SPE) for separation
Samples were fractionated and collected as described by Boden-
nec et al. (2000), using LC-NH2 (500 mg, Supelco) cartridges. Sixty
milligram of sample was loaded onto a column preconditioned
with hexanes, followed by six washes: 15/85 ethyl acetate/hex-
anes, 23/1 chloroform/methanol, 98/2 ethyl ether/acetic acid,
9/1.35 acetone/methanol, 2:1 chloroform/methanol, 30/60/8
chloroform/methanol/3.6 M ammonium acetate. Fractions were
collected using a Preppy Vacuum Manifold, and the solvents
were removed from the fractions under reduced pressure.
2.5. NMR analysis
NMR spectra of samples 16 were collected using CDCl3 on a
Bruker AV-400 spectrometer at 400.3 MHz using a dimethylform-
amide standard for amine content determination. The amine and
water layers of the extraction were collected, the solvent removed
and the residue studied by NMR.
2.6. UPLCMS analysis
Samples 710 were analyzed by ultra-performance liquid chro-
matography (UPLC) on a Waters Acquity UPLC coupled to a Waters
QTOF Premier quadrupole time of ight mass spectrometer using
electrospray ionization (ESI) in positive-ion and negative-ion
modes. Prior to analysis, lipid extracts were diluted to 0.1 mg/mL
in 1:1 methanol:isopropanol, ltered using 0.22 lm Ultra-free-
MC centrifugal lter devices (Amicon Bioseparations). Scanning
was alternated between high (30 eV) and low (5 eV) collision en-
ergy to generate both intact lipid ions and fragment ions to aid
in identication. LC conditions were optimized to analyze both po-
lar lipids and triacylglycerols (TAGs) within 15 min on a Waters
Acquity BEH UPLC C18 column at 55 C, using mobile phases con-
sisting of (A) 40% water with 10 mM ammonium formate/60%
methanol, and (B) 70% isopropanol/30% methanol. Using a ow
rate of 0.4 mL/min, the gradient consisted of a ramp from 40% B
to 100% B over 10 min, and then a hold at 100% B for 30 s. The col-
umn was equilibrated back to the initial starting conditions for
5 min. The injection volume was 3 lL and each extract was ana-
lyzed in triplicate. The MS operating conditions were as follows:
capillary voltage 3.2 kV, sample cone 32 V, extraction cone 2.0 V,
source temperature 120 C, desolvation temperature 400 C, cone
gas 50 L/h, desolvation gas 600 L h, scan time 0.1 s, inter-scan delay
0.02 s, scan range m/z 2001500 (for low collision energy scans)
and m/z 501500 (for high collision energy scans).
2.7. GCMS analysis
Samples were analyzed after SPE separation on a Perkin Elmer
Clarus 680 gas chromatograph attached to a Clarus 600T mass
630 A.R. Boyd et al. / Bioresource Technology 118 (2012) 628632

Fig. 1. The extraction strategy for the removal of lipids from algae cells using SHS.

spectrometer using a 30 m, 0.32 mm, 0.25 mm i.d. column packed After ltration, the solvent/extract mixture was stirred with CO2
with 5% diphenyl/95% dimethyl polysiloxane. The thermal program and water at room temperature, which pulled the majority of the
used started at 130 C and rose to 250 C at a rate of 15 C/min and SHS into the aqueous phase. The NMR spectra indicated residual
was held for 1 min, then rose to 300 C at a rate of 5 C/min and amine content in the extracts of 1824 wt.% when the original
was held for 1 min. extraction was performed at 60 or 80 C, but only 48 wt.% for
those extracted at room temperature. A lower residual amine con-
tent is desired to achieve both higher extract quality and to im-
3. Results and discussion
prove solvent recovery.
After CO2-stripping, 83% of the amine and 75% of the water were
The ability of the SHS CyNMe2 (cyclohexyldimethylamine) to
recovered. For extractions performed at room temperature, the
extract oils from freeze-dried microalgae was tested under a series
NMR spectra of the recovered amine showed residual lipid
of conditions. Temperature, cell disruption and solvent volume
(11 wt.% of the original dry algae sample). Current work in our lab-
were each varied to achieve the best results. As Table 1 shows,
oratory includes testing modications of the method to reduce the
the amount of crude lipids extracted with CyNMe2 and isolated
amount of residual lipid in the recovered amine. Characteristic li-
from the solvent was up to 22 wt.% of the dry cell weight, com-
pid peaks were not observed in the 1H NMR spectrum of the recov-
pared to 52% extracted using chloroform/methanol. Sonication of
ered water.
the SHS extraction samples did not signicantly affect the amount
Extracts 710 were analyzed by UPLCMS and their chromato-
of material extracted using the SHS. As the microalgae had been
grams are shown in Fig. S1 in the supplementary information.
freeze-dried, the cell walls could already have been compromised
Overall, the extracts were composed primarily of tri-, di- or
enough so that the cell-disrupting potential of ultrasonication
mono-acylglycerol, with some glycolipids, formed predominantly
was not necessary to increase the amount of oil extracted from
with oleic acid (18:1) and the long chain fatty acids 28:1 and
the cells.
28:2. These results are consistent with previous lipid proling
studies of B. braunii employing Soxhlet extractions (MacDougall
et al., 2011). Listed in Table 2 are lipid concentrations measured
Table 1 in the extracts by UPLCMS. Due to the variable ionization efcien-
Extraction conditions and results using N,N-dimethylcyclohexylamine.
cies of the individual lipids and the lack of commercially available
Sample Algae CyNMe2 Sonication Temp. wt.% wt.% standards, these values should be considered semi-quantitative.
(g) (mL) (C) Extracted Amine in High free fatty acid levels could be the result of degradation during
extract
the extraction process, which will be investigated in further stud-
1 1 3 No RT 19 2 8 ies. Phospholipids were not observed in any of the samples. The
2 1 3 Yes RT 16 2 4
UPLCMS analysis shows that the room temperature extractions
3 1 3 No 60 20 1 24 recovered 5095% more lipids than those at 80 C, indicating that
4 1 3 Yes 60 22 2 18
heating is not required for extraction and in fact appears to hinder
5 1 3 No 80 18 2 22 the extraction of lipids from the microalgae. It was also determined
6 1 3 Yes 80 20 1 22
that glycolipids were roughly 510 times less abundant than the
7 0.5 10 No RT 12 5 n.d. TAGs present in all of the samples.
8 0.5 10 Yes RT 11 2 n.d.
In order to compare the room temperature SHS process with the
9 0.5 10 No 80 22 1 n.d. more traditional elevated temperature (soxhlet) chloroform/meth-
10 0.5 10 Yes 80 21 4 n.d.
anol method, we determined the neutral lipids content of the
Sonication, if used, was performed prior to stirring for 18 h at the stated temper- resulting extracts by SPE (see Section 2.4 for details). Both the
ature. The amine content of the extracts was determined through NMR analysis SHS method and the chloroform/methanol method extracted the
using a dimethylformamide standard. Error reported is standard deviation of same amount of neutral lipids (5.4 wt.% relative to the mass of
duplicate experiments.
dry algae). However, the chloroform/methanol method extracted
 A.R. Boyd et al. / Bioresource Technology 118 (2012) 628632 631

Table 2
Accurate mass measurement data and lipid concentrations determined by UPLCMS.

Peak # RT (min) m/z (obs) m/z (calc) Error (ppm) Lipid ID Sample 7 Sample 8 Sample 9 Sample 10
conc. (mg/g) conc. (mg/g) conc. (mg/g) conc. (mg/g)
1 1.49 379.2843 379.2819 7.05 MAG(18:1/0:0/0:0) 16 1 17 2 3.1 0.1 5.2 0.3
2 2.37 926.5847 926.5836 1.92 DGDG(18:3/16:3) 1 0.1 0.98 0.1 0.15 0.07 0.28 0.03
3 2.71 954.6154 954.6149 2.22 DGDG(18:3/18:3) 1.6 0.1 1.4 0.1 0.29 0.03 0.42 0.04
4 2.97 956.6275 956.6305 3.17 DGDG(18:3/18:2) 0.5 0.04 0.46 0.04 0.06 0.01 0.1 0.02
5 3.12 932.6314 932.6305 2.36 DGDG(18:3/16:0) 0.4 0.05 0.37 0.04 0.05 0.03 0.10 0.02
6 3.24 958.6477 958.6462 4.22 DGDG(18:3/18:1) 1 0.1 0.8 0.3 0.08 0.08 0.27 0.1
7 3.38 934.6441 934.6462 5.05 DGDG(18:2/16:0) 0.4 0.05 0.31 0.08 0.02 0.02 0.09 0.05
8 3.51 960.6627 960.6618 8.95 DGDG(18:2/18:1) 0.3 0.1 0.23 0.1 0.02 0.02 0.06 0.05
9 3.65 936.6618 936.6618 5.30 DGDG(18:1/16:0) 1.3 0.1 1.2 0.1 0.13 0.1 0.17 0.1
10 3.81 962.6724 962.6775 5.28 DGDG(18:1/18:1) 0.5 0.2 0.54 0.05 0.04 0.04 0.11 0.06
11 4.55 638.5701 638.5718 4.91 DAG(18:1/18:1/0:0) 32 2 31 2 4.7 0.5 7.4 0.6
12 6.09 776.7111 776.7127 3.97 DAG(28:2/18:1/0:0) 12 1 12 1 3.3 1 3.2 1
13 6.23 778.7241 778.7283 5.35 DAG(28:1/18:1/0:0) 14 1 15 1 3.6 0.4 5 0.7
14 6.62 898.7878 898.7858 4.49 TAG(18:3/18:1/18:1) 2.1 0.1 2.3 0.1 0.5 0.03 0.51 0.04
15 6.91 876.8018 876.8015 3.47 TAG(18:1/18:1/16:0) 1 0.06 0.96 0.1 0.34 0.02 0.35 0.03
16 6.98 902.8165 902.8171 4.17 TAG(18:1/18:1/18:1) 15 1 16 1 4.6 1 5.3 1
17 7.71 1038.9453 1038.9423 2.87 TAG(28:1/18:3/18:1) 0.4 0.02 0.43 0.02 0.09 0.02 0.1 0.02
18 7.76 1014.9396 1014.9423 4.37 TAG(26:1/18:1/18:1) 1 0.1 1 0.1 0.22 0.01 0.25 0.02
19 7.86 1040.9584 1040.9580 3.03 TAG(28:2/18:1/18:1) 8 0.4 8.2 0.4 2.1 0.1 2.4 0.2
20 7.94 1042.9735 1042.9736 2.11 TAG(28:1/18:1/18:1) 6.2 0.4 5.8 1 1.5 0.1 1.4 0.1
21 8.01 1068.9942 1068.9893 4.58 TAG(30:2/18:1/18:1) 0.9 0.06 1 0.06 0.17 0.01 0.2 0.02
22 8.52 1179.1007 1179.0988 3.95 TAG(28:2/28:2/18:1) 0.5 0.09 0.59 0.04 0.08 0.01 0.11 0.02
23 8.57 1181.1085 1181.1145 5.21 TAG(28:1/28:2/18:1) 0.3 0.05 0.36 0.07 0.08 0.01 0.12 0.02
24 1.16 277.2160 277.2173 4.58 FFA(18:3) 7.5 0.4 8.3 0.1 1.5 0.02 2.3 0.1
25 1.31 279.2302 279.2330 9.86 FFA(18:2) 2.4 0.1 2.6 0.1 0.45 0.3 0.77 0.04
26 1.42 255.2316 255.2330 5.17 FFA(16:0) 2.9 0.02 3.1 0.1 1.2 0.1 1.4 0.1
27 1.52 281.2486 281.2486 1.77 FFA(18:1) 250 4 260 7 59 1 82 1
28 3.49 419.3912 419.3895 5.92 FFA(28:2) 97 4 93 8 16 4 27 2
29 3.74 421.4040 421.4051 3.32 FFA(28:1) 49 1 51 2 11 0.3 14 0.4
30 4.23 449.4376 449.4364 2.59 FFA(30:1) 3.5 0.1 3.2 0.2 1.7 0.1 1.7 0.1

Calibration for TAGs was achieved using a standard mixture of TAGs of varying degrees of saturation and modeling response factors as a function of retention time. Calibration
for DAG, MAG, and FFA was achieved with a single standard from each lipid class composed of oleic acid (18:1), while DGDG(18:0/16:0) and DGDG(18:0/16:0) were used for
calibration of glycolipids. Abbreviations: triacylglycerol (TAG), diacylglycerol (DAG), monoacylglycerol (MAG), digalactosyldiacylglycerol (DGDG), free fatty acid (FFA). Fatty
acid (FA) notation based on carbon number: number of double bonds. The indicated errors are standard deviations.

a larger amount of other material, so that the concentration of neu-
tral lipids in the extract was small (11 wt.% relative to the total
mass of extract). In contrast, the SHS method was more selective
for neutral lipids and extracted a smaller amount of other material;
the neutral lipid content of the SHS extract was 39 wt.% relative to
the total mass of extract.
While B. braunii species are typically known for high hydrocar-
bon contents, GCMS analysis of the chloroform/methanol extract
showed only minimal C17H34, while the SHS extract showed no
hydrocarbons. Race A is reported to produce C23C33 odd num-
bered n-alkadienes and trienes which are derived from fatty acids
(Metzger et al., 1985). Culturing conditions can strongly inuence
the composition of the lipids in algal cells. The odd-numbered alk-
adiene hydrocarbons are synthesized from fatty acid precursors
and their accumulation depends on the mono-, di- and triacylgly-
cerols present, and may only occur after a prolonged period in
the stationary phase. The algal biomass used in this study was har-
vested before the cells had entered this phase which may account
for their absence or such a low concentration that they were not
detected. However, very long chain fatty acids (C28 and C30) were
found as shown in Table 2, which would be the precursors of C27
and C29 alkadienes.
out any need for distillation. The use of SHS for the extraction of
lipids from microalgae is preferable to the use of conventional sol-
vents because it does not require ammable, highly volatile, or
chlorinated solvents. The SHS extracts just as much neutral lipid
as the standard chloroform/methanol method but is more selective
for those lipids. The use of SHS is preferable to the use of SPS be-
cause drying of the solvent is not required for the extraction with
SHS. However, the amount of lipid left in the recovered amine
needs to be lowered before this technology could be considered
to be practical.
Acknowledgements
A.R.B. would like to thank the Walter C. Sumner Foundation for
funding. P.G.J., P.C. and A.R.B. thank National Science and Engineer-
ing Research Council. P.G.J. thanks the Canada Research Chairs Pro-
gram and P.C. the Ontario Ministry of Research Innovation Early
Researcher Award Program. Contributions to this work conducted
at NRCs Institute for Marine Biosciences were supported by the
National Bioproducts Program in Microalgae Biofuels.
Appendix A. Supplementary data

4. Conclusions Supplementary data associated with this article can be found, in

the online version, at http://dx.doi.org/10.1016/j.biortech.2012.05.
CyNMe2 was shown to be an effective solvent for the extraction 084.
of lipids from freeze-dried samples of B. braunii microalgae. Great-
er amounts of lipids were extracted at lower temperatures and
References
with less contamination by amine than at higher temperatures,
as shown by UPLCMS and NMR spectroscopy. SHS can be removed Cooney, M. et al., 2009. Extraction of bio-oils from microalgae. Sep. Purif. Rev. 38 (3),
from the extract, using carbonated water, and then recovered with- 291325.
632 A.R. Boyd et al. / Bioresource Technology 118 (2012) 628632

Demirbas, A., 2010. Use of algae as biofuel sources. Energy Convers. Manage. 51 McGinn, P.J. et al., 2011. Integration of microalgae cultivation with industrial waste
(12), 27382749. remediation for biofuel and bioenergy production: opportunities and
Jessop, P.G. et al., 2005. Reversible non polar-to-polar solvent. Nature 436, limitations. Photosynth. Res. 109, 231247.
1102. Phan, L. et al., 2008. Switchable solvents consisting of amidine/alcohol or guanidine/
Jessop, P.G. et al., 2010. A solvent having switchable hydrophilicity. Green Chem. 12 alcohol mixtures. Ind. Eng. Chem. Res. 47, 539545.
(5), 809814. Samor, C. et al., 2010. Extractions of hydrocarbons from microalga Botryococcus
Jessop, P.G. et al., 2011. Tertiary amine solvents having switchable hydrophilicity. braunii with switchable solvents. Bioresour. Technol. 101, 32743279.
Green Chem. 13 (3), 619623. Bodennec, J. et al., 2000. A procedure for fractionation of sphingolipid classes by
MacDougall, K.M. et al., 2011. Triacylglycerol proling of microalgae strains for solid-phase extraction on aminopropyl cartridges. J. Lipid Res. 41, 15241531.
biofuel feedstock by liquid chromatographyhigh-resolution mass Metzger, P. et al., 1985. Alkadiene- and botryococcene-producing races of wild
spectrometry. Anal. Bioanal. Chem. 401, 26092616. strains of Botryococcus braunii. Phytochemistry 24, 23052312.