Fuel 155 (2015) 144–154

Contents lists available at ScienceDirect

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

A route to produce renewable diesel from algae: Synthesis and

characterization of biodiesel via in situ transesterification of Chlorella
alga and its catalytic deoxygenation to renewable diesel
Carolina Vieira Viêgas a, Imane Hachemi b, Suely Pereira Freitas a, Päivi Mäki-Arvela b, Atte Aho b,
Jarl Hemming b, Annika Smeds b, Ivo Heinmaa c, Filipe Batista Fontes a, Débora Cristina da Silva Pereira a,
Narendra Kumar b, Donato Alexandre Gomes Aranda a, Dmitry Yu. Murzin b,⇑
a
Federal University of Rio de Janeiro, Brazil
b
Åbo Akademi University, Turku, Finland
c
National Institute of Chemical Physics and Biophysics, Tallinn, Estonia

h i g h l i g h t s

 In situ transesterification of Chlorella algae with sulfuric acid as a catalyst.

 FAME yield is 96–98% from the theoretical one.
 Characterization of purified biodiesel.
 Residual biomass after transesterification contained sugars and proteins.
 Successful hydrodeoxygenation of Chlorella based biodiesel over Ni–H-Y-80 zeolite catalyst.

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

Article history: In situ transesterification of Chlorella alga was performed using 5–20 wt% sulfuric acid as a catalyst at
Received 23 June 2014 either 60 or 100 °C. The maximum ester yield in the range of 96–98% is comparative to the specification
Received in revised form 23 December 2014 of ester content in biodiesel, 96%. A high excess of methanol was used in transesterification ensured also a
Accepted 25 March 2015
high ester yield. The FAME was purified via adsorption of chlorophyll and carotenoids onto a clay.
Available online 14 April 2015
Properties of the purified biodiesel were investigated with several methods. The results showed that
the Chlorella based biodiesel exhibits slightly lower oxidative and thermal stability compared to soybean
Keywords:
based biodiesel due to the presence of polyunsaturated FAMEs. In addition to biodiesel, also the residual
Algae
Biodiesel
biomass was characterized showing that it contained sugars and proteins. An additional hydrogenation
Transesterification would increase the oxidative stability. Hydrodeoxygenation of Chlorella based biodiesel was also demon-
Deoxygenation strated over 5 wt% Ni–HY-80 zeolite with SiO2/Al2O3 ratio of 80 and with 5 wt% Pd/C at 300 °C and 30 bar
Fuel quality in dodecane as a solvent. Ni–HY-80 was superior to Pd/C catalyst giving more than 95% yield of
hydrocarbons.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction hydrodeoxygenation or deoxygenation of triglycerides and fats,

Energy demand in the world is expected to increase by 59% by
the year 2035 [1]. In addition, liquid transportation fuels from
renewable resources are urgently needed due to the depletion of
fossil resources. During the last 10 years intensive research efforts
have been devoted for production of liquid biofuels, such as etha-
nol, butanol, biodiesel and green diesel. Among these biofuels only
the green diesel, which has been synthesized either via catalytic
⇑ Corresponding author. Tel.: +358 22154985; fax: +358 2215 4479.
E-mail address: dmurzin@abo.fi (D.Yu. Murzin).
exhibits the same fuel properties as conventional diesel, since it
has a similar composition. On the other hand, ethanol and butanol
exhibit lower heating values and can suffer from incomplete burn-
ing thus forming oxygenated, harmful, gaseous byproducts.
Chemically biodiesel is methyl ester mixture of fatty acid contain-
ing oxygen. Thus it does not have the same composition as fossil
diesel and it causes also corrosion to engines [2].
The raw materials for production of green diesel have tradition-
ally been triglycerides from vegetable oils. This approach faces,
however, the ethical challenges, such as competition in the raw
material supply with the food supply, thus alternative resources

http://dx.doi.org/10.1016/j.fuel.2015.03.064
0016-2361/Ó 2015 Elsevier Ltd. All rights reserved.
 C.V. Viêgas et al. / Fuel 155 (2015) 144–154
should be found. Algae have beneficial properties, such as high
growth rate and lipid content [1]. Furthermore, their lipid content
is possible to tune via a proper nutrient supply. The biggest chal-
lenge in the commercial utilization of algae as a lipid source for
fuel is its costly drying step.
The next step after drying is extraction of lipids from algae.
Algae extraction has already been investigated for more than two
decades by applying several methods and different solvents.
Mechanical or chemical rupture of hydrophilic cell walls in order
to enhance the extraction capacity was also applied utilizing ultra-
sound [3], microwaves, or supercritical CO2 [4,5]. The algae oil
extracted from dry algae can be used for synthesis of biodiesel.
This process is not, however, economically very feasible [6].
Furthermore, lipid extraction is not very selective, especially with
polar solvents [7–9], since polar phospho- and glycolipids are also
then extracted, whereas higher extraction selectivities toward tri-
acylglycerides have been obtained with non-polar solvents, such as
hexane [3]. On the other hand, hexane has difficulties to open the
cell wall and typically low extraction yields have been achieved.
Algae oil transesterification has been recently demonstrated
already by a few authors [10,11]. For example in [10] alkali cat-
alyzed transesterification resulted in over 95% FAME yields within
60 min using KOH as a catalyst. In addition, three different algae
oils, originating Chlorella, Nannochloropsis and from a third one,
which name was not revealed, were transesterified with the alka-
line catalyst, NaOH [11]. In [11] the main aim was to characterize
the algae oil based FAME.
In several methods the utilization of wet algae has been pro-
posed, for example hydrothermal treatment and transesterification
of wet algae. The results showed that in the two step process com-
posed of a hydrolysis step followed by in situ supercritical water
transesterification, maximally 79 wt% FAME was achieved with
the crude biodiesel yield of 68.3% at 275 °C during 120 min. It
has also been reported that the transesterification of wet algal
biomass is not very efficient resulting in 20% less FAME yields
compared to the process starting from dry biomass [4].
For process intensification in algae extraction and transesterifi-
cation, a few different approaches have been described, for exam-
ple in situ transesterification [12–16], in situ transesterification
under supercritical ethanol [17] or methanol [18,19], a two-step
method comprising in situ hydrolysis of wet algae followed by
supercritical transesterification of wet fatty acid rich solid fraction
[6] and thermo-chemical conversion of algae to biofuels [20]. One-
step in situ transesterification of dry algae over acidic [11–14] and
basic catalysts [15] has been also proposed. Already after 4 h trans-
esterification a nearly equilibrium conversion was achieved at
60 °C and at 90 °C with Chlorella algae using 0.04 mol sulfuric acid
as a catalyst with the molar ratio of methanol to oil 315:1 [14]
indicating that 60 °C is already an adequate temperature for this
reaction. In addition the effect of water was studied in [14] and
the results revealed that water has a detrimental effect on conver-
sion, since already with the water level of 0.7 wt% only 81.7% of the
equilibrium conversion was achieved. On the other hand, in base
catalyzed transesterification of algae the highest yield of FAME
from Chlorella vulgaris was 77.6 wt% after 75 min reaction time
using the molar ratio of methanol to oil of 600:1 at 60 °C [15]. In
addition, product characterization was done for biodiesel from
Chlorella alga synthesized via in situ transesterification using
90:1:3.8M ratio of methanol, oil and sulfuric acid [21]. In that work
the main emphasis was on studying the effect of growing
conditions on the lipid yield and on biodiesel quality.
Transesterification of marine microalga, Schizochytrium mangrovei
PQ6 was also demonstrated using HCl as a catalyst, a co-solvent,
dichloromethane together with methanol at 60 °C during
3 h [22]. The FAME yield based on algal oil was 88%.
Thermochemical conversion of Chlorella pyrenoidosa at 350 °C
145
during more than 30 min in the presence of ethanol resulted in
the production of solid reside, bio-oil with the heating capacity
of 36.45 MJ/kg and gas showing also the potential of this thermo-
chemical approach [20].
Biodiesel has also been synthesized under supercritical ethanol
[17] or methanol [18,19]. In the former work [17] the FAME yield
achieved during 30 min was 67% at 275 °C starting from wet alga
Nannochloropsis Salina and using 1:9 dry algae to ethanol (wt%/
vol) ratio [17]. In the in situ supercritical methanolysis of algae,
N. Salina, under microwave heating maximally about 22 % FAME
was obtained at 130 °C and 4 MPa using the algae to MeOH ratio
of 15 wt%/vol% during 10 min with 3 wt% KOH as a catalyst.
In the two step method, in situ hydrolysis of wet algae followed
by transesterification of wet acid rich residue has been demon-
strated by Levine et al. [7].
Recently catalytic hydrotreatment of crude algal oil has been
studied either over Ni–H-Beta catalysts [23–25], Pd/C [26] or using
supercritical water at 430 °C over Pt/C, Mo2C and H-ZSM-5 cata-
lysts [27]. The results showed that about 78% of liquid alkanes
were achieved with mildly acidic Ni–H-Beta catalyst at 260 °C
under 40 bar hydrogen. The theoretical liquid hydrocarbon yield
was calculated to be 84%. Noteworthy is that these alkanes were
mainly octadecane (59%) together with 19% heptadecane [24]. In
the work of Robota et al. [26] hydrotreatment of algal oil in a con-
tinuous mode at 350 °C under 5.5 MPa producing 85% of alkanes
was performed. The ratio between C18 to the sum of C18 and
C17 together with the ratio C16 to C16 + C15 increased from 0.3
to 0.6 during time on stream. In addition, 76 wt% of carbon was
retained in the feedstock after algal bio-oil treatment at 430 °C in
supercritical water using Mo2C as a catalyst [27].
Biodiesel contains oxygen and thus it is not directly compatible
with diesel motors causing corrosion and incomplete burning.
Catalytic deoxygenation of biodiesel is thus an efficient method
to transfer biodiesel to so-called green or renewable diesel.
Catalytic deoxygenation of fatty acids and triglycerides has been
already demonstrated both in semibatch [28,29] and continuous
mode [30] over Pd/C catalysts. Catalytic deoxygenation of biodiesel
has been earlier demonstrated using Pd/SBA-15 [31] and Pt/Al2O3
[32] as catalysts. In the former work [30] 70% conversion of techni-
cal methyl oleate was achieved at 270 °C, 60 bar hydrogen during
6 h over Pd/SBA-15, whereas deoxygenation of methyl stearate
resulted in 40% yield of heptadecane over 1 wt% Pt/Al2O3 in
6.8 bar helium at 325 °C within 6 h at 44% conversion [32]. The
above mentioned short summary of the state-of-the art in algae
extraction and biofuel production shows that the algae process
technology is not yet fully established and more research is
required.
The aim of this work was to synthesize biodiesel from Chlorella
algae using sulfuric acid as a catalyst, to determine the composi-
tion of biodiesel and its fuel properties. In addition, catalytic
hydrodeoxygenation of algae-derived biodiesel was demonstrated
both over Pd/C and Ni–HY-80 zeolite catalysts. Since lipid fraction
is not the main part in alga, the residual fraction after transesteri-
fication has been analyzed and its utilization has been emphasized.
2. Experimental
2.1. Raw material
Dry Chlorella powder from Fuqing King, Drarmsa Spirulina Co.,
Ltd., China was used as a feedstock in the current work.
According to manufacturer it contains 57.3 wt% protein, 17.3 wt%
carbohydrates, 2.3 wt% of chlorophyll and maximally 7 wt% mois-
ture and ash based on dry weight. The particle size of the spray-
dried powder was 125 lm.
146 C.V. Viêgas et al. / Fuel 155 (2015) 144–154
Hexane/water
MeOH/H2SO4
Microalgal trans-
filtration
biomass esterification
Product 1:
residue
Fig. 1. The flowsheet of the transesterification of algae biomass.
2.2. In situ transesterification of Chlorella algae
In situ transesterification of dry Chlorella algae was performed
in a system comprising of an oil bath and 50 ml reactor with a
reflux condenser. Magnetic stirring with 300 rpm was used.
Initially reactions were carried out using 10 g of dry algae and by
varying the concentration of sulfuric acid, namely 5, 10, 15 and
20 wt% relative to the dry weight (g H2SO4/g biomass) and a pro-
portion of 3:1 methanol/biomass (mL g 1). The reaction time and
temperature were 4 h and 60 °C, respectively. At the end of the
reaction, the biomass was separated by vacuum filtration and the
alcohol was evaporated in a rotavapor and recovered. After filtra-
tion in a separator funnel, the organic phase containing the esters
was added to a volume of 100 ml of hexane for the separation of
polar compounds (unconverted). The polar fraction was drained
and the fraction soluble in hexane (the main product) was sepa-
rated. Thereafter hexane was evaporated in a rotary evaporator
and the crude biodiesel was quantified (Fig. 1).
In order to select the most favorable operational conditions for
transesterification with 20 wt% H2SO4 in relation to dry biomass,
two reaction temperatures, namely 60 °C and 100 °C, were used
as well as two reaction durations, 2 h and 4 h, respectively.
2.3. Purification of biodiesel
The methyl esters were purified in a glass counter column, in
which 10 g of kaolin type commercial clay (Argila verde, www.for-
cadaterra.com, Brazil) was used as an adsorbent. About 4 grams of
crude biodiesel was added to the column, which was eluted with
hexane. Compressed air was also fed into the column in order to
increase the flow velocity of the crude biodiesel through the bed.
2.4. Experimental methods for catalyst preparation and
hydrodeoxygenation of biodiesel
5 wt% Ni/H-Y-80 (SiO2/Al2O3 ratio 80) was prepared by evapo-
ration impregnation method using Ni(NO3)2
6H2O as Ni precursor.
5 wt% Pd/C was obtained from Sigma–Aldrich. Catalyst characteri-
zation methods are described in Section 2.5.16.
Catalytic hydrodeoxygenation of biodiesel was performed in an
autoclave using 1.0 g of FAME in 100 ml of dodecane over 5 wt%
Ni–HY-80. Prior to the experiments the catalysts were reduced in
flowing hydrogen (20 ml/min) with the following temperature
program. The temperature increased from room temperature to
the reduction temperature (200 °C in case of Pd and 350 °C in case
of Ni) during 80 min with the rates of 2 °C/min and 4 °C/min
approximately for Pd and Ni respectively. The temperature was
maintained constant for 2 h at the reduction temperature; after-
wards the catalyst was cooled down to the ambient temperature.
70%/30%
Product 2:
extraction adsorption
FAME
Polar phase
extraction EBE
Product 3:
Extracted oil
Thereafter the deoxygenated liquid containing both solvent and
biodiesel was injected into the reactor containing the pre-reduced
catalyst. The reaction mixture was heated up to the desired tem-
perature and the pressure was increased to 30 bar by adding
hydrogen. The reaction was commenced by starting the stirring.
Stirring rate of 1200 rpm was used in order to remove the external
mass transfer limitations and small catalyst particles were used
(below 63 lm) in order to suppress the internal mass transfer lim-
itations. Samples were withdrawn from the reactor and analyzed
by GC (see Section 2.3.5). The reaction mixture was also analyzed
by GC–MS in order to confirm all the reaction products.
2.5. Analytical methods
2.5.1. Quantitative analysis of lipid fraction and fatty acids
The gravimetric analysis for total lipids was performed by
extracting fresh alga using Bligh and Dyer method (1957) [33].
1 g of dry algae was extracted with a solvent mixture (3 ml) con-
taining chloroform and methanol in a weight ratio of 2:1.
Thereafter the lipid fraction was saponified with KOH at 60 °C for
1 h followed by extraction with hexane in order to remove
unsaponified fraction containing carotenoids. The polar fraction
was hydrolyzed with sulfuric acid and extracted with hexane.
The non-polar phase containing free fatty acids was recovered by
evaporating hexane and weighted.
2.5.2. Thermogravimetric analysis
Thermogravimetric analysis of alga and fresh and spent adsor-
bent clay was performed under synthetic air with SDT Q600
(V20.9 Build 20) instrument using the purge gas flow rate of
100 ml/min. The sample was inserted in a platinum pan and heated
with the following temperature programme: 25 °C – 10 °C/min –
1000 °C. In addition, thermogravimetric analyses of biodiesel was
performed in Pyris-TGA, Perkin Elmer under nitrogen flow of
20 ml/min by heating the sample from 20 °C up to 800 °C in
synthetic air.
2.5.3. Scanning electron microscopy analysis and EDXA-analysis
The morphology of fresh alga, solid residue after transesterifica-
tion with 20 wt% H2SO4 and clay was studied by scanning
electron microscope (Zeiss Leo 1530 Gemini) equipped with a
ThermoNORAN vantage X-ray detector. The energy dispersive
X-ray analysis (EDXA) was carried out with the same instrument.
2.5.4. Organic elemental analysis
Organic elemental analysis for quantification of N, C, H and S
was performed for two different fractions, namely for fresh
Chlorella and residual biomass after treatment with 20 wt% H2SO4
using Flash 2000 instrument from Thermo Scientific. For all
 C.V. Viêgas et al. / Fuel 155 (2015) 144–154
samples three replicates were performed. About 2 mg of standards
and samples were weighed into tin cups. He and O2 gases were
used in the experiment. Analysis of nitrogen content was cali-
brated using the commercial cystine as a standard. In addition,
the commercial standard BBOT was analyzed as a sample. The
standards were purchased from Elemental Microanalysis Limited
(UK). Cystine standard showed a good agreement with the certified
values.
2.5.5. Analysis of fatty acids and esters
Fatty acids and esters were determined by GC using ASTM
D6584 and Shimadzu GC-2014 gas chromatograph with a split/
splitless injector equipped with a flame ionization detector (FID).
The identification of the profiles of methyl esters of fatty acids from
microalgae was performed in a gas chromatograph, Shimadzu GC-
2014 coupled to a flame ionization detector (FID) using a split and
a Carbowax 20 M capillary column (30 m length, internal diameter
of 0.32 mm and film thickness of 0.25 lm). The operating parame-
ters for the GC were as follows: 1 L injection volume, split ratio
50:1, injector and detector temperatures were 250 °C and 280 °C,
respectively, carrier gas helium (99.95%), temperature programme:
50 °C (6 min) 5 °C/min – 150 °C (1 min) – 200 °C/min – 5 °C/min –
240 °C (5 min). The injections were performed in duplicate.
The methyl esters of fatty acids were identified using FAME
C8:0–C24:0 standard mixture. The quantification was made using
methyl heptadecanoate as an internal standard with a concentra-
tion of 1.0 mg/mL in heptane.
In addition, another method was applied for analyzing FAMEs
as follows. The sample was diluted in dodecane (Sigma Aldrich,
P99%) with the approximated ratio of 1–10. Thereafter it was sily-
lated with the following procedure: 100 ll of sample was diluted
in 1 ml of pyridine (Sigma Aldrich, P99%). Subsequently 100 ll
of eicosane (Acros Organics, 99%) was added as an internal stan-
dard together with 100 ll of BSTFA (Acros Organics, P98%) and
50 ll of TMCS (Sigma Aldrich, P98%). Silylation of samples was
performed in an oven at 70 °C for an hour before they were injected
in the GC. The samples were analyzed with a gas chromatograph
equipped with a capillary column (ZB-5HT Inferno, length 30 m,
internal diameter 0.32 mm and film thickness 0.25 lm. The detec-
tor and injector temperatures were both 300 °C.
The free fatty acids obtained via extraction of fresh algae using
chloroform–methanol in ratio of 2:1 under magnetic stirring for
two hours at 60 °C were titrated with NaOH as follows: the sample,
about 1 g, was combined with 25 ml ethanol and 1 ml of
phenolphthalein.
2.5.6. Analysis of monoglycerides, diglycerides, triglycerides, free and
total glycerin by GC
For determination of glycerides ASTM D6584 method was used
in a gas chromatograph GC 2010 Shimadzu with on-column injec-
tion. The injection volume was 1 ll. Hydrogen was applied as a car-
rier gas. The GC was equipped with a flame ionization detector and
a CG 2010 capillary column (15 m long  0.32 mm internal diame-
ter  0.1 m film thickness) containing 5% phenyl-polydimethyl-
siloxane (DB5-HT). The following temperature programme was
applied: 50 °C (1 min) – 15 °C/min – 180 °C – 7 °C/min – 230 °C –
30 °C/min – 380 °C (10 min).
2.5.7. Analysis of fresh algae and solid residue after transesterification
with size exclusion chromatography
The solid samples, fresh alga and the residue after transesterifi-
cation of fresh algae with 20 wt% sulfuric acid at 60 °C for 1 h were
added into tetrahydrofuran solution under stirring. After gravimet-
ric determination of the amount of filtrate, it was diluted with
tetrahydrofuran to the concentration 1 mg/ml, filtrated and ana-
lyzed by SEC as follows.
147
The SEC system composed of LT-ELS-detector (Low
Temperature Evaporative Light-Scattering Detector, Sedex 85,
Sedere LT-ELSD) with degasser (DGU-14A), the gradient pump
(FCV-10ALVP), the fraction collector (Pharmacia LKB-Helifrac)
together with the system controller (SCL-10AVP). Three columns,
two Jordi Gel DVB 500A (300 mm  7.8 mm) and one Guard
column (50  7.8 mm) were used for separation and the peaks
obtained in the detector were integrated with Shimadzu
Class- VP 8v.6.12 SP5 integrator.
2.5.8. Total organic carbon (TOC) analysis
Freeze dried fresh alga and residue after treating the alga with
20 wt% sulfuric acid samples in ‘‘in situ’’ transesterification were
methanolysed. The released sugar monomers were silylated and
analyzed by GC according to the method described by Sundberg
et al. [34]. Resorcinol was used as an internal standard instead of
sorbitol described in the original method. The GC (Perkin–Elmer
AutosystemXL) conditions were split injection at 250 °C with the
split ratio of 1:25 using a HP-1 column (Agilent) with the length
of 25 m, internal diameter 0.2 mm and film thickness 0.11 lm.
The detector (FID) temperature was 310 °C. The column tempera-
ture program was 100 °C – 4 °C/min – 175 °C – 12 °C/min –
300 °C (7 min).
2.5.9. FTIR analysis of biodiesel
The infrared spectrometry analysis was performed with
ShimadzuÒ IR-Prestige-21. The spectra were recorded in the range
4000–400 cm 1 using a resolution of 1 cm 1 and 64 scans per
second.
2.5.10. Differential scanning calorimetry of purified biodiesel
The thermal properties of the feedstock and extracted oil were
characterized with the DSC Q1000 v9.9 instrument. Samples of ca.
2 mg in aluminum pans were heated from 90 °C to 85 °C then
cooled to 90 °C and heated again to 85 °C at 10 °C/min.
2.5.11. Determination of the viscosity of biodiesel
The viscosity of the biodiesel formed is determined according to
ASTM D-445 using Anton Parr CT-200. For one measurement 5 mL
of the sample was used and the analysis was carried out at 40 °C.
2.5.12. Determination of biodiesel density
The density of biodiesel formed was determined with Cannon
DMS 4500 apparatus according to ASTM D-4052 method using
3 mL of the sample. The analysis was performed at 20 °C.
2.5.13. Determination of the water content in biodiesel
The amount of water present in the biodiesel was determined
by the colorimetric method of Karl Fisher titration according to
ISO EN 12937 method. The method is based on the reduction of
sulfur dioxide by iodine in the presence of water via potentiomet-
ric titration using Metron apparatus.
2.5.14. Determination of the oxidative stability of biodiesel
The induction time analysis was conducted in Rancimat appara-
tus (Metrohm, 743 Biodiesel RancimatÒ), using about 3.00 g of the
sample under pure oxygen atmosphere of 10 Lh 1 at 110 °C, fol-
lowed by the method EN 14112 established by ANP Resolution
42. In these accelerated tests the minimum acceptable time for
biodiesel oxidation is 3 h.
2.5.15. Specific surface area determination of the adsorbent by
nitrogen adsorption
The specific surface area of the clay was measured using nitro-
gen adsorption with Tristar 300 Surface Area and Porosimetry
Analyzer from Micromeritics apparatus. The clay was evacuated
148 C.V. Viêgas et al. / Fuel 155 (2015) 144–154
at 150 °C for 12 h prior to the measurement. The BET equation was
used for determination of the specific surface area with N2 adsorp-
tion. The volume of micropores was calculated by t-plot method
and the specific volume of pores with a pore diameter obtained
by the BJH method.
2.5.16. Catalyst characterization methods
Both 5 wt% Ni/H-Y-80 and 5 wt% Pd/C were used as catalysts for
biodiesel deoxygenation. These catalysts were characterized by
nitrogen adsorption (Brunauer–Emmett–Teller) and CO pulse
chemisorption (Autochem 2910 apparatus) for measurement of
the specific surface area and metal dispersion, respectively. In CO
chemisorption the stoichiometry Pd/CO was one.
Ni particle size was determined with JOEL JEM-1400 Plus
transmission electron microscope operated at 80 kV acceleration
voltage, since carbon monoxide chemisorption for Ni catalyst
was difficult to perform due to agglomerate formation.
The acidity of the support, H-Y-80 was measured by 27Al MAS
NMR. Single pulse excitation spectra were obtained with Bruker
AVANCE III-800 spectrometer in 18.8 T magnetic field. The
resonance frequency (RF) for 27Al was 208.4 MHz, Bruker MAS
probe for 3.2 mm outer diameter zirconia rotors was used.
Sample spinning frequency of 22.0 kHz was maintained. To keep
quantitative intensities 10° short excitation pulses at RF field
strength 50 kHz and 0.1 s relaxation delay between the accumula-
tions were used. Intensity was normalized dividing the absolute
intensity with the mass of the sample and with the number of
scans. The spectra were referenced to the resonance frequency
of KAl(SiO4)2  12H2O.
3. Results
3.1. Transesterification results
3.1.1. Preliminary experiments using different amounts of sulfuric acid
Preliminary experiments for the transesterification of Chlorella
algae at 60 °C for 4 h were performed with different amounts of
sulfuric acid as a catalyst. Chlorella alga used in the current study
had about 15 wt% the total lipid content according to Bligh and
Dyer method, whereas only about half of it could be saponified.
The total lipids are composed of neutral lipids, such as fatty acids,
phospholipids and glycolipids. The FAME yield calculated from the
1. 2.
30000
3.
25000
20000
Intensity (a.u.)
15000
10000
5000
0 results revealed that the highest yield of FAME was achieved at
0 500 1000 1500
Time (s)
Fig. 2. SEC chromatogram of a) a lipid fraction extracted with tetrahydrofuran (blue
curve). Notation: 1) and 2) triglycerides and diglycerides, 3) free fatty acids and
monoglycerides, b) of the residue achieved from transesterification of Chlorella with
20 wt% H2SO4 during 4 h at 60 °C and extracted with tetrahydrofuran (red curve).
(For interpretation of the references to color in this figure legend, the reader is
referred to the web version of this article.)
total amount of saponifiable lipids decreased as expected in in situ
transesterification of Chlorella alga with sulfuric acid at 60° C for
4 h using the mass ratio methanol to algae of 3.8:1 as follows:
5.6 wt% (5 wt% H2SO4) < 27 wt% (10 wt% H2SO4) < 41 wt% (15 wt%
H2SO4)  98 wt% (20 wt% H2SO4). This result indicates that acid
hydrolysis of cell wall, triglycerides, phospholipids and carotenoid
esters is incomplete, when using less than 20 wt% sulfuric acid. The
results can be explained by the fact that Chlorella alga has thick cell
walls, which is also visible in Fig. S1. Chlorella alga contained also
high amounts of free fatty acids (FFA) confirmed by titration and
SEC-analysis (Fig. 2). The SEC results revealed that about 25% of
the lipid fraction extracted with tetrahydrofuran was free fatty
acids. In addition, SEC results showed that the solid residue after
transesterification with 20 wt% sulfuric acid was nearly free from
lipid components (Fig. 2), which is in accordance with the elemen-
tary analysis results. The amount of free fatty acids was analyzed
both by titration of extracted lipid fraction and using size exclusion
chromatography. In the former method Chlorella alga was
extracted with chloroform–methanol at 60 °C for two hours and
analyzed by titrating the lipid fraction with KOH resulting in free
fatty acid amount of 21 wt%. Due to the presence of a high amount
of free fatty acids it is beneficial to use acid instead of alkali as a
catalyst in order to avoid saponification reaction.
3.2. Optimization of in situ transesterification of Chlorella algae
The FAME profile with 20 wt% sulfuric acid as a catalyst in
in situ transesterification of triglycerides from alga is shown in
Table 1. This profile shows that about 51.84 mol% were methyl
esters of polyunsaturated fatty acids, which also lower the oxida-
tion stability of the biodiesel (see Section 3.3). It contains FAMEs
in the range of C12–C18. The content of saturated, monounsatu-
rated and polyunsaturated FAMEs is 34.21 mol%, 9.85 mol% and
51.84 mol%, respectively. Based on the fatty acid profile achieved
from FAME analysis and given in Table 1, the average molar mass
of oil was calculated according to the formula in [15] to be
812.9 g/mol. The molecular weight in the Chlorella based oil
reported in [14] was 880 g/mol being slightly higher than the
one reported in this work. This deviation is due to the higher
amount of C16 FAMEs in this work. Furthermore, the average
molecular weight of soybean oil is 874 g/mol [35].
Noteworthy is that this Chlorella contains quite much polyun-
saturated FAMEs, especially linoleic and linolenic acids, which
are not very beneficial for the biodiesel due to their lowered oxida-
tion stability (see Section 3.1.3). Very high fractions of polyunsat-
urated FAMEs were reported in Chlorella spp algae extracted
from Kenya, being 65.7 wt% [36]. On the other hand, the amount
of PUFA was 54.9 wt% in Chlorella sp. DRLMA3 [37].
The molar ratio of methanol to oil was calculated to be 801:1
using the molecular weight of the oil determined above. It is gen-
erally known that high molar ratio of methanol to oil facilitates
high equilibrium conversion, as was the case also in this study.
Furthermore, a high initial free fatty acid content is known to
require high catalyst concentrations and high molar ratio of alco-
hol to oil [13,38].
Optimization of the reaction time and temperature for in situ
transesterification was performed with 20 wt% sulfuric acid. The
60 °C during 4 h experiment (Table 2), whereas nearly equally good
results were obtained at 100 °C during 4 h. As a comparison in situ
transesterification of Chlorella was performed at 60 °C at 315:1M
ratio of methanol to oil using sulfuric acid as a catalyst, and several
temperatures were investigated. It was stated by Ehimen et al. [14]
that high equilibrium conversions were achieved both at 60 °C and
90 °C already within 4 h using 0.14 ml H2SO4 per 1 g Chlorella alga,
which is comparable with the result in the current work using
 C.V. Viêgas et al. / Fuel 155 (2015) 144–154 149

Table 1
The amounts of different of methyl esters and acids (mol%) using 20 wt% of sulfuric acid as a catalyst in in situ transesterification of Chlorella algae at 60 °C for 4 h using the mass
ratio methanol to algae of 3.8:1.

Myristic acid Methyl palmitoleate Methyl palmitate Palmetoleic acid Palmitic acid Methyl linoleate Methyl linolenate
0.23 8.62 32.68 2.31 0.37 26.21 25.63

Methyl oleate Methyl stearate Linoleic acid Oleic acid Stearic acid
1.23 1.53 0.55 0.38 0.26

Table 2
FAME yield (% of the apolar fraction, which is given in parenthesis as wt% from the
algae) as a function of temperature and time achieved in the transesterification of
Chlorella alga using 20 wt% sulfuric acid as a catalyst. 10 g algae and 30 ml methanol.

Temperature (°C) FAME yield (%)

2h 4h
60 70.04 (8.36) 98.44 (7.85)
100 67.30 (7.27) 97.98 (7.34)

0.11 ml H2SO4 per 1 g Chlorella. On the other hand, only 86% yield
of FAME was reported from Chlorella at 60 °C for in situ transester-
ification with sulfuric acid within 19 h, when the molar ratios of
methanol to lipids and sulfuric acid to lipids were 600:1 and
0.35:1, respectively [15]. The latter ratio corresponds to 0.02 ml
sulfuric acid per g algae, which is 5.5 times less than that used in
the current case, explaining a lower conversion. From the eco-
nomic point of view it should be stated that in the current case a Fig. 3. SEM image of the clay.
large excess of methanol was used together with 20 wt% sulfuric
acid, which is not feasible in the large scale. As a comparison
wet algae paste (containing 10% solid material) was in situ transes-
Table 4
terified in supercritical methanol at 260 °C during 30 min with Amounts of inorganic elements in fresh clay (wt%).
12 wt%/vol methanol giving 85.75 % FAME [19], which might be
Element Fresh clay
one potential route in the future.
Al 7.8
Si 16.6
3.3. Purification of biodiesel and properties of purified biodiesel P 0.14
K 2.36
Ca 1.0
Crude biodiesel after transesterification with 20 wt% H2SO4 at Ti 0.5
60 °C for 4 h exhibited dark color due to the presence of carote- Fe 4.9
noids and chlorophyll. After purification using clay as an adsor-
bent, the color of the biodiesel changed to yellow. The yield of
the biodiesel was 6.5 wt% after removal of impurities (Table 3). the specific surface area and micropore volume for spent clay were
As a comparison a two-step method, in which algae was first 119 m2/g and 0.043 cm3/g indicating that 21% reduction of surface
extracted with chloroform–methanol mixture, followed by trans- area occurred.
esterification of the lipid fraction resulted in slightly lower lipid Several physico-chemical methods, such as FTIR, TGA, DSC and
yield compared to the one achieved in in situ transesterification other analytical methods were used to characterize biodiesel pre-
of fresh algae in one step. pared by transesterification of dry Chlorella alga with 20 wt%
The adsorbent clay has been characterized by SEM, EDXA, nitro- H2SO4 at 60 °C for 4 h and purified using clay as an adsorbent.
gen adsorption and TGA techniques. SEM image of the clay is Properties of algae biodiesel were compared with the biodiesel
depicted in Fig. 3. The fresh clay contained 14.7% of ash. The specifications according to different standards as well as with
EDXA results revealed that the main components of the clay were FAME synthesized via transesterification using different algae as
Si and Al (Table 4). The specific surface area of the fresh clay a feedstock, such as Chlorella vulgaris [21].
according to nitrogen adsorption was 151 m2/gcat and the microp- The FTIR of both crude and purified Chlorella biodiesel is shown
ore volume 0.053 cm3/gcat measured by Dubinin method, whereas in Fig. 4a and b. The peaks corresponding to esters at 2925 cm 1

Table 3
The amounts of FAME obtained in the optimized conditions (time 4 h, temperature 60 °C, 20 wt% H2SO4) in the transesterification of Chlorella alga and from lipid fraction
extracted with chloroform: MeOH 2:1 followed by transesterification.

Feedstock Amount of Methanol (ml) Amount of crude Amount of purified wt% of purified FAMES
biomass (g) FAME (g) FAME (g) based on the initial biomass
Alga 100 300 7.09 6.57 6.5
Lipid fraction 10.55 30 5.03 n.m. 5a
a
Unpurified.
150 C.V. Viêgas et al. / Fuel 155 (2015) 144–154

(a) contains characteristic bands for proteins [41] with a specific

1.0 amide I band at 1656 cm 1, which is also very strong in Chlorella
alga with the high protein content. In addition to amide I, also
amide II band at 1536 cm 1 is visible [39]. Polysaccharide bands
0.8 appear in the region 1200–900 cm 1 [11] and according to TOC
Transmittance, a.u.

analysis the amount of polysaccharides was about 20.3 wt% (see

Section 3.4). The presence of lipids can be confirmed by the strong
0.6
absorption bands in the region between 3100 and 2800 cm 1.
Peaks at 2860 cm 1 and 2930 cm 1 clearly visible in Fig. 5a are
0.4 close to the CH2 asymmetric and symmetric stretching at 2928
and 2860 cm 1 of a lipid [41].
The ash content in the purified biodiesel was very low according
0.2 to TGA analysis, being only 0.39 wt% (Fig. S2), whereas the ash con-
tent in the fresh Chlorella alga was 5.9 wt%. This value is quite close
to the value reported by being 4.49 wt% and 5.93 wt%, respectively.
0.0
4000 3500 3000 2500 2000 1500 1000 500
The FAME decomposition occurred at 211 °C (Fig. S2). The TGA
differs from the algea FAME prepared from Nannochloropsis salina
Wavenumbers, cm-1
sp [18], in which three separate decomposition temperatures were
obtained, at 190 °C, 290 °C and at 480 °C. According to [18] the first
(b) peak is originated from oxidation and decomposition, whereas the
1.0
subsequent peaks come from polymerization and combustion
reactions. The main decomposition peak in TGA of FAME [18]
0.8 was, however, relatively close to the one achieved for Chlorella
Transmittance, a.u.

based FAME in the current work. The exact FAME composition in

[18] was not reported. It was only stated that the main compo-
0.6 nents were FAME containing C16:1, C18:1, C18:2, C18:3 and
C20:5 acyl chains.
0.4
Differential scanning calorimetry results showed that crystal-
lization and melting temperatures were 4.0 °C and 5.9 °C, respec-
tively (Fig. 5). The crystallization temperature for soy biodiesel,
0.2 4.08 °C was close to that of algae biodiesel, whereas the melting
temperature for soy biodiesel, 1.7 °C, was lower than that for algae
biodiesel [48]. Thus as a conclusion it can be stated that the ther-
0.0
mal properties are slightly worse in algae biodiesel than in soy
4000 3500 3000 2500 2000 1500 1000 500
biodiesel. Furthermore the energies of crystallization and melting
Wavenumbers, cm-1 were 33.2 J/g and 14.5 J/g, respectively.
The ester content should be higher than 96.5 wt% according to
Fig. 4. FTIR from (a) crude biodiesel and (b) purified biodiesel.
EN 14103 and in the current product, the ester amount of 96.2%
is close to the limit (Table 6).
The biodiesel standard has also limits for free glycerol and total
Table 5 glycerin, since these compounds are harmful for engines and they
The interpretation of different bands in FTIR of Chlorella algae and biodiesel.
can plug filters and cause combustion problems in engines. Total
Bands Interpretation Component Reference glycerin is the sum of free and bound glycerol, i.e. mono- and
(cm 1) diglycerides. The biodiesel product from Chlorella contained about
1047 mCAOAC Carbohydrate [40]
1079 mCAO Carbohydrate [40,42]
1079 mP@O Phosphodiester [40] 1.0
1243 mCO Carboxylic acid in algae, nucleic [43]
acid, phosphodiester
1422 mCH3 bending Methyl lipid [42]
1422 mCH2 bending Protein [42] 0.5
1536 mNAH bending, Amide II, protein [39]
Heat flow (W/g)

mCAN stretching
1656 mC@O stretching Amide I, protein [45]
1740 mC@O Esters [44] 0.0
2860 mCH2 [41]
2925 mCH2 Lipid or FAME [40,41,48]
77 mOH, mNH Water, protein [44]

-0.5

and 1740 cm 1 are clearly visible and the purification of biodiesel

with a clay decreases the intensities of the bands in the range of
-1.0
1400–1500 cm 1 as well as in 1100–1200 cm 1 corresponding to
-100 -50 0 50 100
the presence of protein and phosphodiester according to the
Table 5. In comparison, the FTIR results from the fresh Chlorella Temperature (°C)
alga (picture not shown here) displayed intensive bands in the Fig. 5. DSC curve for biodiesel. Notation: exothermic peak (black curve) and
region 3000–3700 cm 1, at 2370 cm 1, 950–1800 cm 1 as well as endothermic peak (red curve). (For interpretation of the references to color in this
500–750 cm 1. The region in the range of 1500–1800 cm 1 figure legend, the reader is referred to the web version of this article.)
 C.V. Viêgas et al. / Fuel 155 (2015) 144–154 151

Table 6 values reported for original biodiesel [11]. Only one algae biodiesel,
Biodiesel specification and the properties of the biodiesel prepared via in situ Eldorado biofuel containing 9 wt% saturated C16 and C18 acyl
transesterification. Conditions: Temperature 60 °C, time 4 h, 20 wt% H2SO4.
groups together with 47 wt% oleate, 36 wt% linoleate and only
Method Specification Biodiesel from 7 wt% linoleate acyl chains showed high enough oxidative stability
for biodiesel* Chlorella algae [49]. Also in the work of Kleinova et al. [11] only one algae oil bio-
(in situ process)
diesel derived from Chlorella, but containing only 7.1 wt% linolenic
Esters (concentration), % wt/wt, (EN 96.5 96.23 acid, passed the oxidative stability test [11]. In the current work
14103)
Monoglycerides, % wt/wt, (ASTM-D 0.8 0.175
the content of linolenoate ester was 23 wt%.
6584, EN 14105) For the use of biodiesel as a fuel there are also feedstock restric-
Diglycerides, % wt/wt, (ASTM-D 6584, 0.2 0.081 tions defined by EN14214 [50]. These restrictions limit the
EN 14105) amounts of FAME from linolenic acid and from acids containing
Triglycerides, % wt/wt, (ASTM-D 0.2 0.371
more than four double bonds to 12% and 1%, respectively [50].
6584, EN 14105)
Oxidation stability at 110 °C, h, (EN 6h 1.30 When comparing the FAME profile obtained in this work, it can
14112) be concluded that the amount of FAME from linolenic acid does
Viscosity, mm/s2, (ASTM-D 445) 3.0–6.0 4.07 not fulfill the criterion. It has been proposed that in such case
Density, kg/m3, (ASTM- D 4052) 850–900 879.7 the biodiesel should be hydrogenated, which, however, decreases
Water, mg/kg, (EN ISO 12937) Max. 500 1019
the cold flow properties [50]. In addition, it is known that for cur-
*
Parameters from Brazilian National Petroleum Agency – regulator of the rent motors FAME can be blended due to its oxygen content and its
parameters for commercialization of biodiesel in Brazil. energy content is lower than that of diesel, being 38 MJ/kg and
43 MJ/kg, respectively [51].
12% of the limit level of free glycerol being 0.02% according to
ASTM D6584 standard. Furthermore, its total glycerol content
3.4. Characterization of the solid residue after transesterification with
was only 26% of the limit value for total glycerol, 0.38 % according
20 wt% sulfuric acid
to ASTM D6584 standard.
Water content in the current biodiesel was twofold higher com-
The elemental compositions of fresh Chlorella alga and its resi-
pared to the limit [5]. This might be originated partially from
due after transesterification with 20 wt% sulfuric acid are shown
washing of biodiesel with water in order to remove residual sulfur.
in Tables 7 and 8 determined by OEC and EDXA, respectively. The
It has been proposed that water can be removed from biodiesel
molar ratio of H/C and O/C calculated from EDXA and OEC results,
with a vacuum drier [14], causing, however, extra production costs.
respectively, were 1.72 and 0.61, which are quite close to the val-
The density of biodiesel was 879.9 kg/m3, which is acceptable
ues reported in [47] being 1.78 and 0.6. The comparison of the ele-
by ASTM-D 4052 according to the standards required by ANP.
mental composition of fresh alga and the residue showed that the
The density of biodiesel is directly linked with the molecular struc-
carbon content was higher in the fresh alga than in the residue.
ture of its molecules. The larger the length of carbon chain of the
Nitrogen content determined by OEC method was close to 9 wt%
alkyl ester, the higher the density, however, this value decreases
(Table 6) corresponding the protein content of 54 wt% if nitrogen
with higher degree of unsaturation in the molecule.
amount is multiplied by six.
The kinematic viscosity observed for the esters of Chlorella sp.
The residue after treatment of the Chlorella alga with 20 wt%
showed a value of 4.07 mm/s2 which according to the standard
sulfuric acid contained about the same amount of nitrogen as
(3.0–6.0 mm/s2) is acceptable being within the required limits.
was found in the fresh algae, since the C/N molar ratios were for
Analogous results were achieved for Chlorella oil derived biodiesel
fresh algae and for the residue 2.7 and 2.6, respectively.
in [11]. The viscosity was expected to be within the standard limit
because the fatty acid profile showed also the presence of large Table 7
amounts of unsaturated fatty acids. The viscosity of biodiesel The OEC analysis of different fractions of Chlorella. The results are average of three
increases with the length of the carbon chain and the degree of sat- replicates.
uration and has influence on the burning process in the combus- Sample Sample Nitrogen Carbon Hydrogen Sulfur
tion chamber of the engine. High viscosity leads to heterogeneity weight
in the biodiesel fuel due to the decrease of the efficiency of (mg) (wt%) (wt%) (wt%) (wt%)
atomization in the combustion chamber, causing deposition of
Fresh Chlorella 1.98 8.98 47.6 6.86 0.56
residues in the inner parts of the residual soaps, as well as unre- Residual biomass after 1.906 6.35 31.79 4.72 3.25
acted glycerides (mono-, di-, and triglycerides). In addition, the H2SO4 20 wt%
oxidative degradation products of biodiesel increase the viscosity. treatment
These contaminants can therefore be monitored indirectly by mea-
suring the kinematic viscosity at 40 °C.
The oxidation stability of the biodiesel in the current case was Table 8
only 1.3 h. This is much lower than the specification for the oxida- The elemental composition of fresh algae, ash and biomass after treatment with
20 wt% H2SO4 determined by EDXA.
tion stability of biodiesel being 6 h, which is the standard for soy-
bean biodiesel. Furthermore, it is known that the oxidative stability Element In fresh algae (wt%) Residue after treatment with 20 wt% H2SO4
for neat methyl oleate is 2.79 h according to the Rancimat test [11]. C 42.68 37.59
This result was expected due to a high amount of unsaturated O 34.66 38.27
esters in the prepared biodiesel and indicates that an antioxidant Fe 0 0
Na 0.1 0
as an additive should be used with the biodiesel [11]. The relative
Mg 0.34 0.15
rates for oxidation stability for C18:2, C18:3 and C20:4 are 41, 98 K 0.92 0.60
and 195 [15]. In addition, it has been proposed that some algae- P 1.77 0.95
derived biodiesel qualities would rather be suitable as a feedstock S 0.71 5.21
for renewable diesel for hydroprocessing [46]. More than tenfold Ca 0.29 0.20
Al 0 0
higher oxidative stability values have been reported for Si 0 0.16
Scenedesmus sp oil based hydrogenated biodiesel compared to the
152 C.V. Viêgas et al. / Fuel 155 (2015) 144–154

Table 9
The amount of different sugars and acids obtained via methanolysis in fresh alga and residue from the in situ transesterification using 20 wt% H2SO4. The amounts are given in
mg/g.

Sugar Arabinose Galactose Galactonic acid Glucose Glucuronic acid Mannose Rhamnose Xylose Sum
Fresh 1.1 47.3 0.8 126.3 5.0 6.1 12.7 4.4 203.6
Residue from 20 wt% H2SO4 1.5 29.7 1.1 143 4.0 4.7 5.9 2.7 192.6

Furthermore, the molar ratios O/C for these materials were 0.6 and
0.8, respectively, based on EDXA analysis. These results indicate
that the residual cake contained more proteins and carbohydrates,
whereas lipids were removed. On the other hand, the amount of
phosphorus decreased and the molar ratio of P/C decreased from
fresh algae to the residue from 0.016 to 0.01 indicating that phos-
phorus, although it is present in the fresh algae as a polar phospho-
lipid, goes more into the lipid fraction than stays in the residue.
Furthermore, according to OEC analysis results, sulfur amount in
the residue has increased substantially, as expected.
Sugar analysis after methanolysis was performed both for fresh
alga and for the residue after transesterification of alga with
20 wt% sulfuric acid. The results showed that the main sugar in
Chlorella was glucose followed by galactose (Table 9). The total
amount of sugars was 20.3 wt% which is relatively close to the
value given by manufacturer, being 17 wt%. In addition also rham-
nose, mannose, xylose and glucuronic acid were present in rela-
tively large amounts. These results agree quite well with those
published by [52]. The residue after in situ transesterification with
20 wt% sulfuric acid contained 19.3 wt% sugars and the enrichment
of glucose was observed, whereas the amounts of galactose, man-
nose and rhamnose decreased. Thus in conclusion it can be stated
that the residue is rich in carbohydrates and proteins and could be
further utilized for production of e.g. furan derivatives from sugars
as well as food and/or feed from proteins thus facilitating complete
utilization of algae.
4. Catalytic hydrodeoxygenation of biodiesel
4.1. Catalyst characterization results
Hydrodeoxygenation of biodiesel was investigated, since bio-
diesel is not fully compatible with the current motors due to its
high oxygen content. Furthermore, the net heating value of
green diesel produced via hydrodeoxygenation is higher than
that of biodiesel (see Section 3.3), being 42–44 MJ/kg [53].
Hydrodeoxygenation of the purified algae biodiesel was performed
using dodecane as a solvent either with 5 wt% Pd/C or 5 wt% Ni/H-
Y-80 as a catalyst at 300 °C and 30 bar hydrogen. TEM images
revealed the mean catalyst particle size to be typically around
4 nm, whereas the metal dispersion for Pd/C was 60% correspond-
ing to 2 nm particle size. The specific surface areas of Ni/H-Y-80
and Pd/C were 784 m2/g and 540 m2/g, respectively.
The amounts of different aluminum species of the zeolite
support, H-Y-80, was investigated by 27Al MAS NMR. The results
indicated that three broad components, corresponding to Al tetra-
hedral sites Al(IV), octahedral sites Al(VI) and five-coordinated
300 °C under a hydrogen atmosphere with total pressure of
30 bar in dodecane. The catalytic HDO of FAME gave both fatty
acids and hydrocarbons as products over both Ni/HY-80 and Pd/C.
Initial reaction rates for transformation of FAMEs were
0.3 mmol/(min gcat) and 0.64 mmol/(min gcat) for 5 wt% Pd/C and
5 wt% Ni–HY-80, respectively. As a result of a fast saturation of
these compounds, the kinetics profile represents two regions; the
first is characterized by a rapid conversion with the rate of
0.3 mol/min gcat while the second is moderately slower with the
rate of 0.02 mol/min gcat. The first rate corresponds to the first
15 min and reflects the time needed for the saturation of methyl
linoleate and methyl linolenate. An analogous trend in the HDO
rates was observed for Ni–HY-80, although the rate was faster with
the nickel catalyst. The main products were fatty acids over Pd/C
catalyst and only about ca. 18 mol% hydrocarbons were formed
within 390 min (Table 10). Only traces of fatty alcohols were
formed. Furthermore, the conversion of FAME with Pd/C was only
40% within 390 min.
FAMEs reacted much more rapidly over Ni–HY-80 than over
Pd/C. Very high yields of hydrocarbons were achieved already after
150 min (not shown in the Table 10) over Ni–HY-80 and the inter-
mediate compounds, fatty acids reacted rapidly to hydrocarbons.
The maximum yield of fatty acids was 25% achieved after 30 min
reaction time. As FAME is a mixture of fatty acids with different
chain lengths; the product is as well a mixture of hydrocarbons
and intermediates. From the mechanistic point of view it is impor-
tant to observe, that unsaturated esters reacted rapidly into satu-
rated ester under HDO, after approximately 15 first minutes of
the reaction only traces of methyl linoleate and methyl linolenoate
which initially were the major compounds in FAME, were found.
Octadecane, hexadecane and heptadecane were the main products
of HDO over Ni/H-Y-80. No stearone was formed over Ni–H-Y-80
indicating that the amount of Brønsted acid sites was suitable for
producing hydrocarbons from biodiesel. Analogously it was
reported that only traces of stearone were formed in the HDO of
stearic acid at 40 bar hydrogen and 260 °C with Si/Al ratio of 75
[25]. In our preliminary experiments with stearic acid HDO using
for example 15 wt% Ni-Beta-25 with lower Si-to-Al ratio about
60% stearone was formed.
During HDO, the unsaturated esters reacted rapidly into satu-
rated ester, after approximately 15 first minutes of the reaction
only traces of methyl linoleate and methyl linolenoate, initially
the major compounds in FAME, were found.
Table 10
Composition of biodiesel products (mol%) after HDO over Pd and Ni.

sites Al(V) of non-framework aluminum were observed. Al(IV) Component Biodiesel 5 wt% Ni/H-Y 80 5 wt% Pd/C
bands appearing close to 60 ppm correspond to Brønsted acid sites, Time, min
whereas the band at 0 ppm reflects Lewis acid sites [54]. The rela-
0 240 370 240 390
tive ratio of Al(IV), Al(V) and Al(VI) were 28%, 24% and 48%, respec-
tively, indicating that the ratio between Brønsted to Lewis acid C 14 0.00 1.68 1.81 0 0
C 15 0.00 14.18 13.7 2.54 5.82
sites was 0.58.
C 16 0.00 29.20 29.1 0.47 0.79
C17 0.00 18.10 19.2 3.94 8.17
4.2. Hydrodeoxygenation results of biodiesel C18 0.00 30.80 31.5 2.79 2.93
C16OHAC18OH 0.00 0.27 0.27 1.08 1.09
Acids 3.83 3.53 3.25 31.35 40.13
The catalytic HDO of FAME over two different catalysts was car-
FAME 96.17 2.24 1.17 57.83 41.07
ried out in a semi-batch reactor. The reactions were performed at
 C.V. Viêgas et al. / Fuel 155 (2015) 144–154
5. Future perspective
In situ transesterification of alga resulted in about 6.5 wt% yield
of FAME calculated per initial weight of biomass, which is about
half of the total saponifiable fatty acids. The large amount of sulfu-
ric acid makes the process unfeasible in large scale. The FAME yield
based on the total amount of biomass is low, and thus it should be
pointed out here, that algae with higher lipid content should be
used for production of FAME. It has, for example, been reported
that Chlorella sp. MP-1 contains 28.8 wt% lipids [47]. This is an
indicative number, since the lipid content varies with the growing
conditions. Moreover some algae species contain even up to 40%
lipids [55]. Life cycle analysis of algae biodiesel made for 1000 MJ
unit showed that the main obstacle in algae based biodiesel tech-
nology is the thermal dewatering of algae requiring large amounts
of fossil fuel [56]. In [56] carbohydrates were already taken as a
co-product having a potential to replace corn as a feedstock for
ethanol production. If considering an additional step of HDO of bio-
diesel, there would be an additional cost of hydrogen. Utilization of
whole algae could make the process economically more feasible
only if all fractions could be utilized for value added products,
carbohydrates could be used for production of fuels, e.g. furan
derivatives and proteins for food or feed and even for chemical
applications, such as for biobased surfactants. In addition new dry-
ing processes should be invented instead of using fossil fuel.
6. Conclusions
In situ transesterification of Chlorella alga was investigated at 60
and 100 °C using sulfuric acid as a catalyst under excess of metha-
nol. The FAME yield, 96–98% of theoretical level was obtained with
20 wt% sulfuric acid within 4 h, whereas lower sulfuric concentra-
tions gave much lower FAME yields. It should, however, be pointed
out, that the large amount of sulfuric acid used in in situ transes-
terification is not economically feasible and more research is
required for further process optimization. Since Chlorella alga con-
tained also substantial amounts of free fatty acids, acid catalyzed
transesterification is preferred instead of base catalysts causing
soap formation. The ester content of biodiesel was adequate for
the biodiesel standard. The biodiesel was purified via adsorption
of chlorophylls and carotenoids onto a clay. The purified biodiesel
was characterized by several methods. The biodiesel contained 54%
polyunsaturated FAMEs as well as 33 wt% saturated and 9.5 wt%
monounsaturated FAMEs. Due to high content of polyunsaturated
FAMEs the oxidative stability was lower than that for soybean bio-
diesel. The residual biomass after transesterification contained
sugars and proteins. The former one could be transformed to fura-
nic compounds.
Due to the fact that biodiesel is not fully compatible with cur-
rent motor technology, the hydrodeoxygenation of biodiesel was
also demonstrated over Pd/C and Ni–HY-80 catalyst. The main
products were fatty acids over Pd/C catalyst giving only 40%
conversion within 390 min. Nickel based catalyst was superior to
produce hydrocarbons compared the Pd/C catalyst, since more
than 95% yield of hydrocarbons was achieved within 250 min at
300 °C under 30 bar.
Acknowledgement
This work is part of the activities of Åbo Akademi Process
Chemistry. The authors are grateful to Mr. A. Holappa at Åbo
Akademi University for DSC analysis, P. Backman for TGA analysis
and Mr. S. Lindroos for organic elemental analysis and Mr. L.
Silvander in PCC for SEM-EDXA analysis.
153
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.fuel.2015.03.064.
References
[1] EIA. International Energy Outlook Early release 2012. <http://www.
iea.doe.gov.> [access 10.6.14].
[2] Haryono I, Suryantoro MT. Effects of biodiesel on engine generator
components. Int J Eng Technol 2014;4:256–60.
[3] Menetrez MY. An overview of algae biofuel production and potential. Environ
Sci Technol 2012;46:7073–85.
[4] Ceron MC, Campos I, Sánchez JF, Acién FG, Molina E, Fernández-Sevilla JM.
Recovery of lutein from microalgae biomass: development of a process for
Scenedesmus almeriensis Biomass. J Agric Food Chem 2008;56:11761–6.
[5] Cheng C-H, Du T-B, Pi H-C, Jang S-M, Lin Y-H, Lee H-T. Comparative study of
lipid extraction from microalgae by organic solvent and supercritical CO2.
Biores Technol 2011;102:10151–3.
[6] Mendes RL, Reis AD, Palavra AF. Supercritical CO2 extraction of c-linolenic acid
and other lipids from Arthrospira (Spirulina) maxima: comparison with
organic solvent extraction. Food Chem 2006;99:57–63.
[7] Levine RB, Pinnarat T, Savage PE. Biodiesel production from wet algal biomass
through in situ lipid hydrolysis and supercritical transesterification. Energy
Fuels 2010;24:5235–43.
[8] Kale A. Methods of producing biofuels, chlorophylls and carotenoids. US patent
2011/0263886 A1; 2011.
[9] Jones J, Manning S, Montoya M, Keller K, Poenie M. Extraction of algal lipids
and their analysis by HPLC and mass spectrometry. J Am Oil Chem Soc
2012;89:1371–81.
[10] Chen L, Liu T, Zhang W, Chen X, Wang J. Biodiesel production from algae oil
high in free fatty acids by two-step catalytic conversion. Biores Technol
2012;111:208–14.
[11] Kleinova A, Cvengrosova Z, Rimarik J, Buzetzki E, Mikulec J, Cvengros J. Biofuels
from Algae. Procedia Eng 2012;42:231–8.
[12] Johnson MC. Tester JW lipid transformation in hydrothermal processing of
whole algal cells. Ind Eng Chem Res 2013;52:10988–95.
[13] D́Oca MGM, Viegas CV, Lemoes JS, Miyasaki EK, Moron-Villarreyes JÁ, Primel
EG, et al. Production of FAMEs from several microalgal lipidic extracts and
direct transesterification of the Chlorella pyrenoidosa. Biomass Bioenergy
2011;35:1538–9.
[14] Ehimen EA, Sun ZF, Carrington CG. Variables affecting the in situ
transesterification of microalgae lipids. Fuel 2010;89:677–84.
[15] Velasquez-Orta SB, Lee JGM, Harvey A. Alkaline in situ transesterification of
Chlorella vulgaris. Fuel 2012;94:544–50.
[16] Velasquez-Orta SB, Lee JMG, Harvey AP. Evaluation of FAME production from
wet marine and fresh water microalgae by in situ transesterification. Biochem
Eng J 2013;76:83–9.
[17] Reddy HK, Muppaneni T, Patil PD, Ponnuswamy S, Cooke P, Schaub T, et al.
Direct conversion of wet algae to crude biodiesel under supercritical ethanol
conditions. Fuel 2014;115:720–6.
[18] Patil P, Reddy H, Muppaneni T, Ponnuswamy S, Sun Y, Dailey P, et al.
Optimization of microwave-enhanced methanolysis of algal biomass to
biodiesel under temperature controlled conditions. Biores Technol
2013;137:278–85.
[19] Patil PD, Gude VG, Mannarswamy A, Deng S, Cooke P, Munson-McGee S, et al.
Optimization of direct conversion of wet algae to biodiesel under supercritical
methanol conditions. Biores Technol 2011;102:118–22.
[20] Duan P, Jin B, Xu Y, Yang Y, Bai X, Wang F, et al. Thermo-chemical converion of
Chlorella pyrenoidosa to liquid biofuels. Biores Technol 2013;133:197–205.
[21] Mallick N, Mandal S, Kumar Singh A, Bishai M, Dash A. Green microalga
Chlorella vulgaris as a potential feedstock for biodiesel. J Chem Technol
Biotechnol 2012;87:137–45.
[22] Hong DD, Mai DTN, Thom LT, Ha NC, Lam BD, Tam LT, et al. Biodiesel
production from Vietnam heterotrophic marine microalga Schizochytrium
mangrovei PQ6. J Biosci Bioeng 2013;116:180–5.
[23] Peng B, Yuan X, Zhao C, Lercher JA. Stabilizing catalytic pathways via
redundance. Selective reduction of microalgae oil to alkanes. J Am Chem Soc
2013;134:9400–5.
[24] Peng B, Yao Y, Zhao C, Lercher JA. Towards quantitative conversion of
microalgae oil to diesel-range alkanes with bifunctional catalysts. Angew
Chem Int Ed 2012;51:2072–5.
[25] Song W, Zhao C, Lercher JA. Importance of size and distribution of Ni
nanoparticles for the hydrodeoxygenation of microalgae oil. Chem A Eur J
2013;19:9833–42.
[26] Robota HJ, Alger JC, Shafer L. Converting algal triglycerides to diesel and HEFA
jet fuel fractions. Energy Fuels 2013;27:985–96.
[27] Duan P, Savage PE. Catalytic treatment of crude algal bio-oil in supercritical
water. Energy Environ Sci 2011;4:1447–56.
[28] Kubickova I, Snåre M, Eränen K, Mäki-Arvela P, Murzin D Yu. Hydrocarbons for
diesel fuel via decarboxylation of vegetable oils. Catal Today 2005;106:
197–200.
154 C.V. Viêgas et al. / Fuel 155 (2015) 144–154
[29] Simakova I, Simakova O, Mäki-Arvela P, Simakov A, Estrada M, Murzin D Yu.
Deoxygenation of stearic acid over supported Pd catalysts: effect of metal
dispersion. Appl Catal A Gen 2009;355:100–8.
[30] Theilgaard Madsen A, Rosmyslowich B, Simakova IL, Kilpiö T, Leino AR, Kordas
K, et al. Step changes and deactivation behavior in the continuous
decarboxylation of stearic acid. Ind Eng Chem Res 2011;50:11049–58.
[31] Lee SP, Ramli A. Methyl oleate deoxygenation for production of diesel fuel
aliphatic hydrocarbons over Pd/SBA-15 catalysts. Chem. Central J
2013;7(149):1–10.
[32] Do PT, Chiappero M, Lobban LL, Resasco DE. Catalytic deoxygenation of
methyl-octanoate and methyl-stearate on Pt/Al2O3. Catal Lett 2009;130:9–18.
[33] Bligh EG, Dyer WJ. Lipid extraction and purification. Can J Biochem Physiol
1959:911–7.
[34] Sundberg A, Sundberg K, Lillandt C, Holmbom B. Determination of
Hemicelluloses and Pectins in wood and pulp fibres by acid methanolysis
and gas chromatography. Nordic Pulp Paper Res J 1996;11(4):216–9.
[35] Ionescu M. Polyols from renewable resources. Chemistry and technology of
polyols for polyurethanes; 2005. p. 435–470, Edited by RAPRA, UK. ISBN:1-
85957-501-3.
[36] Abubakar LU, Mutie AM, Kenya EU, Muhoho A. Characterization of algal oil and
its potential as biofuel in Kenya. J Appl Phytotechnol Environ Sanit
2012;1(4):147–53.
[37] Kaur S, Sarkar M, Srivastava RB, Gogoi HK, Kalita MC. Fatty acid profiling and
molecular characterization of some freshwater microalgae from India with
potential for biodiesel production. New Biotechnol 2012;29:332–44.
[38] Lam MK, Lee KT. Catalytic transesterification of high viscosity crude
microalgae lipid to biodiesel: effect of co-solvent. Fuel Proc Technol
2013;110:242–8.
[39] Dean AP, Martin MC, Sigee DC. Resolution of codominant phytoplankton
species in a eutrophic lake using synchrotron-based Fourier transform infrared
spectroscopy. Phycologia 2007;46:151–9.
[40] Brandenburg K, Seydel U. Fourier transform infrared spectroscopy of cell
surface polysaccharides. In: Mantstsh HH, Chapman D, editors. Infrared
spectroscopy of biomolecules. Wiley: Chichester; 1996. p. 203–89.
[41] Stuart R. Infrared spectroscopy: fundamentals and applications. 1st Ed. West
Sussex: John Wiley & Sons Ltd.; 2004.
[42] Ponnuswamy I, Madhavan S, Shabudeen S. Isolation and characterization of
green microalgae for carbon sequestration, waste water treatment and bio-
fuel production. Int J Biosci biotechnol 2013;5:17–25.
[43] Benning LG, Phoenix VR, Yee N, Tobin MJ. Molecular characterization of
cyanobacterial silicification using synchrotron infrared micro-spectroscopy.
Geochim Cosmochim Acta 2004;68:729–41.
[44] Liu J, Mukherjee J, Hawkens JJ, Wilkinson SJ. Optimization of lipid production
for algal biodiesel in nitrogen stressed cells of Dunaliella salina using FTIR
analysis. J Chem Technol Biotechnol 2013:1807–14.
[45] Dumas P, Miller L. The use of synchrotron infrared microspectroscopy in
biological and biomedical investigations. Vib Spectr 2003;32:3–21.
[46] Tsai WT, Chen HR, Kuo KC. Surface characterization of dead microalgae-based
biomass using methylene blue adsorption. Surf Interf Anal 2011;43:959–63.
[47] Phukan MM, Chutia RS, Konwar BK, Kataki R. Microalgae Chlorella as a
potential bio-energy feedstock. Appl Energy 2011;88:3307–12.
[48] Xu YX, Hanna MA. Synthesis and characterization of hazelnut oil-based
biodiesel. Ind Crops Prod 2009;29(2–3):473–9.
[49] Knothe G. Fuel properties of highly polyunsaturated fatty acid methyl esters.
Prediction of fuel properties of algal biodiesel. Energy Fuels 2008;22:1358–64.
[50] Bucy HB, Baumgardner ME, Marchese AJ. Chemical and physical properties of
algal methyl ester biodiesel containing varying levels of methyl
eicosapentanoate and methyl docosahexanoate. Algal Res 2013;1:57–69.
[51] Kalnes T, Marker T, Shonnard DR. Green diesel: a second generation biofuel. Int
J Chem React Eng 2007;5:A48.
[52] Kapuan E, Loos E, Reisser W. Cell wall composition of virus-sensitive symbiotic
chlorella species. Phytochem 1992;31:3101–4.
[53] Bezergianni S, Dimitriadis A. Comparison between different types of
renewable diesel. Ren Sust Energy Rev 2013;21:110–6.
[54] Pescarmona PP, Janssen KPF, Strobants C, Molle B, Paul JS, Jacobs PA, et al. A
high throughput experimentation study of the synthesis of lactates with solid
acid catalyst. Top Catal 2010;53:77–85.
[55] Demirbas A. Use of algae as biofuel sources. Energy Conv Managam
2010;51:2738–49.
[56] Sander K, Murthy GS. Life cycle analysis of algae biodiesel. Int J Cycle Assess
2010;15:704–14.