Biodiesel production from Nannochloropsis gaditana lipids through transesterification catalyzed by Rhizopus oryzae lipase

Bioresource Technology 203 (2016) 236–244

Contents lists available at ScienceDirect

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

Biodiesel production from Nannochloropsis gaditana lipids through transesterification catalyzed by Rhizopus oryzae lipase Elvira Navarro López, Alfonso Robles Medina ⇑, Pedro Antonio González Moreno, Luis Esteban Cerdán, Lorena Martín Valverde, Emilio Molina Grima Area of Chemical Engineering, University of Almería, 04120 Almería, Spain

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


Saponifiable lipids (SLs) were

extracted from wet biomass using ethanol and hexane.
Intracellular R. oryzae lipase catalyzed the methanolysis of microalgal SLs.
Biodiesel yield and lipase reuse were higher using SLs poorer in polar lipid.
Biodiesel yield and lipase reutilization were higher in tertbutanol than in hexane.
83% of microalgal SLs were transformed to biodiesel in the optimized conditions.

a r t i c l e

i n f o

Article history: Received 23 October 2015 Received in revised form 10 December 2015 Accepted 13 December 2015 Available online 21 December 2015 Keywords: Biodiesel Microalgae Intracellular lipase Rhizopus oryzae Transesterification

Methanol + solvent Optimization of transesterification conditions: 0.7 g RO/g SL, 11/1 methanol/SL (mol/mol); 10 mL tbutanol/g SL, 35ºC, 72 h, 130 rpm.

Lipid extraction ethanolhexane + crystallization in acetone: 95%SLs (49% polar lipids)

N. gaditana wet biomass 25% dry biomass; 12% SLs of dry biomass

Intracellular lipase from Rhizopus oryzae grown in our laboratory

Reuse of R. oryzae lipase. In three reaction cycles biodiesel the conversion decreased a 58%

Homogenization at 1200 bar Fatty acid methyl esters (biodiesel). Conversion: 81% Reuse of lipase. In three reaction cycles the biodiesel conversion decreased a 14% Lipid extraction hexane + crystallization in acetone: 61% SLs (37.5% polar lipids)

Transesterification conditions previously established as optimal

Fatty acid methyl esters (biodiesel). Conversion: 83%

Methanol + solvent

a b s t r a c t Biodiesel (fatty acid methyl esters, FAMEs) was produced from saponifiable lipids (SLs) extracted from wet Nannochloropsis gaditana biomass using methanolysis catalyzed by Rhizopus oryzae intracellular lipase. SLs were firstly extracted with ethanol to obtain 31 wt% pure SLs. But this low SL purity also gave a low biodiesel conversion (58%). This conversion increased up to 80% using SLs purified by crystallization in acetone (95 wt% purity). Polar lipids play an important role in decreasing the reaction velocity – using SLs extracted with hexane, which have lower polar lipid content (37.4% versus 49.0% using ethanol), we obtained higher reaction velocities and less FAME conversion decrease when the same lipase batch was reused. 83% of SLs were transformed to biodiesel using a 70 wt% lipase/SL ratio, 11:1 methanol/SL molar ratio, 10 mL t-butanol/g SLs after 72 h. The FAME conversion decreased to 71% after catalyzing three reactions with the same lipase batch. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Biodiesel is defined as a mixture of mono alkyl esters of longchain fatty acids derived from vegetable oils or animal fats. It is a biodegradable fuel and it can be used without modifications in ⇑ Corresponding author at: Area of Chemical Engineering, Department of Engineering, University of Almería, 04120 Almería, Spain. Tel.: +34 950 015065; fax: +34 950 015491. E-mail address: [email protected] (A. Robles Medina). http://dx.doi.org/10.1016/j.biortech.2015.12.036 0960-8524/Ó 2015 Elsevier Ltd. All rights reserved.

compression ignition engines (Robles-Medina et al., 2009). The most usual method to produce biodiesel is by transesterification of acylglycerols with short-chain alcohol (methanol or ethanol) in the presence of various catalysts (Stamenkovik et al., 2011). The production of enzymatic biodiesel from vegetable oils has been extensively studied over recent years. Nevertheless, the raw material cost and the limited availability of raw vegetable oils are critical issues in biodiesel production (Li et al., 2007). For this reason, using microalgae as the raw material appears to be a suitable alternative energy source. Algae offer higher photosynthetic

237

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

efficiency, biomass productivity and growth velocity than the majority of oleaginous cultures (Chisti, 2007). In the industrial processes so far proposed, alkali catalysts have mostly been used. However, alkali catalyst cannot be used if the oil contains free fatty acids (FFAs > 0.5%), as occurs with microalgal oils; this is because soaps are formed, which lead to a reduction in yield and an increase in the downstream processing of the biodiesel produced (Robles-Medina et al., 2009). Microalgal oils with high FFA contents can be transesterified by acid and enzymatic catalysts, but lipases work at lower temperatures (25–50 °C) and the subsequent separation and purification of biodiesel and glycerol is easier when lipases are used. The use of acid catalysts involves the use of a great amount of water and, consequently, makes it a less environmentally-friendly process (Khor et al., 2010). To overcome these drawbacks, many authors have studied the use of lipases as catalysts for biodiesel production. Some of the extracellular lipases most widely used belong to species such as Candida antarctica (NovozymeÒ 435), Thermomyces lanuginosus (LipozymeÒ TL IM) or Rhizomucor miehei (LipozymeÒ RM IM). However, the use of extracellular lipases requires complicated and expensive recovery purification and immobilization processes for industrial application and there has been considerable interest in the use of whole-cell biocatalysts of intracellular lipase, which have the advantage of avoiding isolation operations, purification and extracellular lipase immobilization (Li et al., 2008; Ban et al., 2002). Consequently, the filamentous fungus from Rhizopus oryzae is one of the most widely used for biodiesel production via enzymatic transesterification. For example, yields of 80% and 90% have been achieved by converting jatropha oil (Tamalampudi et al., 2008) and soybean oil (Hama et al., 2007), respectively, to methyl esters using R. oryzae as the biocatalyst (Table 1). Although most researchers use R. oryzae immobilized within BSPs, some authors have employed the fungus dispersed in the reaction mixture, as has Zeng et al. (2006), who obtained a yield of 86% in the transesterification of vegetables oils (Table 1). Besides their high cost, lipases have other drawbacks for use in biodiesel production such as their rapid deactivation due to methanol. If the methanolysis or ethanolysis are carried out in solvent-free systems, some methanol or ethanol can be non-solubilized due to the limited solubility of these alcohols in the oil, and this non-solved alcohol causes lipase deactivation. To overcome this problem, some authors have

proposed the stepwise addition of methanol (Shimada et al., 1999) with t-butanol as the solvent, which solves methanol and reduces lipase deactivation (Li et al., 2008, 2010). The aim of this work was to produce enzymatic biodiesel by: (i) the extraction of saponifiable lipids (SLs) from wet microalgal Nannochloropsis gaditana biomass and (ii) the transformation of SLs to methyl esters by transesterification using R. oryzae intracellular lipase as the biocatalyst. The objective was to attain high FAME yields while taking lipase stability into account. As can be observed in Table 1, to date, we know of few authors who have studied biodiesel production from microalgal oil using whole cells to catalyze the transesterification reaction. 2. Methods 2.1. Microalgae and chemicals In this study, the microalgae used as an oil-rich substrate was the marine microalgae N. gaditana, cultured in an outdoor tubular photobioreactor in ‘‘Las Palmerillas, Cajamar” research center (El Ejido, Almería, Spain). The wet biomass supplied was harvested and centrifuged (Centrifuge Brand Supelco 4-15, Sartorius, Germany) to 75 wt% moisture content and then stored at 24 °C until use. The fungus R. oryzae IFO 4697 was provided by the NITE Biological Resource Center (NBRC, Chiba, Japan). The chemicals used for the culture of the fungus were peptone, potato dextrose agar (PDA), tween 20 and D-glucose (all of them from Panreac S.A, Barcelona, Spain); NaNO3, KH2PO4, MgSO47H20, H2SO4 (Sigma– Aldrich, Madrid, Spain) and glycerol (JEscuder S.L, Rubí, Barcelona, Spain). For the fungus immobilization we used biomass support particle (BSP) cubes of 6  6  3 mm, made of polyurethane reticulated foam (Filtre TM 25133 (PPI45), Recticel Ibérica S.L., Spain). The chemicals used were analytical grade ethanol (96% v/v), hexane (95% purity, synthesis quality), acetone (analytical quality) (all three from Panreac S.A), methanol (99.9% purity, Carlo Erba Reagents, Rodano, Italy) and tert-butanol (analytical grade, Fluka, Barcelona, Spain). All reagents used in the analytical determinations were of analytical grade. Standards were obtained from Sigma–Aldrich (St. Louis, Mo, USA) and used without further purification.

Table 1 Conditions and results regarding the production of FAMEs by transesterification of oils catalyzed by R. oryzae intracellular lipase. Oil

R. oryzae/oil ratio (w/w)

Methanol/oil ratio

Reaction time (h)

Solvent

FAME conversion (%)

Lipase stability

References

Soybean oil Soybean oil

0.14 mL/g oil 0.14 mL/g

75 50

– –

86 90

0.18 mL/g

20

65

Jatropha oil

4%

3:1 mol/mol

60

2.56 g tbutanol/g oil –

Olive oil Calophyllum inophyllum oil Pistacia chinensis bge seed oil Waste cooking oil

4% 20%

3 72

– –

50 92

4%

3:1 mol/mol 12:1 mol/mol or 5.4 mL/g 5:1 mol/mol

– 11% after 10 uses 15% after 6 uses 10% after 5 uses 5% after 1 use 7% after 6 uses

Zeng et al. (2006) Hama et al. (2007)

Refined Soybean oil

5% BSPs in a packed bed reactor 6%

60

–

92

30%

4:1 mol/mol

35

–

89

Oil from yeast Candida sp. N. gaditana lipids

20,000 UA/mL oil 340 mg/mL oil Whole cell three phase bioreactor 70%

0.24 mL/mL SL

4

41

17 mL ethanol/g FA

96

5 g hexane/g oil –

92

No loss after 5 uses No loss after 40 uses 30% after 6 uses –

11:1 mol/mol (1.6 mL/g SL)

72

10 mL tbutanol/g SL

83

15% in 3 uses

N. gaditana lipids

80

Li et al. (2008) Tamalampudi et al. (2008) Canet et al. (2014) Arumugan and Ponnusami (2014) Li et al. (2012) Chen et al. (2006) Duarte et al. (2015) Araya et al. (2015) This work

238

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

2.2. Extraction and purification of lipids from wet N.gaditana microalgal biomass Lipid extraction from the wet N. gaditana biomass was carried out using two procedures: the ethanol–hexane system and extracting the lipids directly from wet biomass using only hexane. The extraction of lipids with ethanol followed by purification with hexane is described in detail by Navarro López et al. (2015). Using this procedure, an extract with only 31 wt% of SLs was obtained. This purity was increased by crystallization in acetone following the procedure also described by Navarro López et al. (2015). The extraction of SLs with hexane is based on a similar method proposed by Jiménez Callejón et al. (2014). In this method, wet N. gaditana microalgal biomass was firstly homogenized at 1500 bar in a laboratory homogenizer (Panda Plus 2000 S.N. 8983 model, Gea Niro Soavi, Parma, Italy) and then extracted using 10 mL of hexane/g biomass at room temperature for 20 h with a constant agitation of 200 rpm. The hexanic phase was analyzed by gas chromatography (GC, Section 2.5) to determine the SL extraction yield. All the hexanic extracts were evaporated in a rotary evaporator (Buchi, R210, with a vacuum pump V-700, Switzerland) to recover the hexane. These lipids were also subjected to the acetone crystallization treatment to treat to increase the SL purity. 2.3. Microorganism, culture media and biomass support particles (BSPs) All the transesterification reactions were catalyzed using the filamentous fungus R. oryzae IFO 4697, which produces a 1,3positional specific lipase. The fungus R. oryzae was maintained in an agar slant made from 4% potato dextrose agar and 2% agar. The NBRC instructions for recovering the microorganism from the agar slant were followed. The basal medium contained (per liter of distilled water): 70 g of polypeptone, 1 g of KH2PO4, 1 g of NaNO3 and 0.5 g of MgSO47H2O. 30 g of olive oil per liter of medium was added, as the sole carbon source. The pH of the medium was initially adjusted to 5.6 using an aqueous solution of H2SO4 and then it was allowed to follow its natural course. Flasks containing 100 mL of the basal medium were inoculated by aseptically transferring spores (about 106 spores) from an agar slant and then incubating them for around 40 h at 35 °C on a reciprocal shaker (Inkubator 1000, Unimax 1010 Heidolph, Klein, Germany) (130 rpm) (Hama et al., 2007; Zeng et al., 2006). Experiments were carried out using the fungus both with no immobilization and immobilized within BSPs. To prepare the fungus when immobilized within BSPs, flasks containing 100 mL of the basal medium and 150 BSPs were inoculated by aseptically transferring spores from a fresh agar slant and incubating them for 90 h at 35 °C on a reciprocal shaker (150 rpm) (Ban et al., 2002). The amount of immobilized R. oryzae cells was determined as the weight difference between the BSPs after and before the immobilization. In both cases, the fungus, whether immobilized within BSPs or not, was separated from the culture by filtration, washed with distilled water and then lyophilized (Telstar Cryodos 50, Telstar, Spain) for 24 h, and stored at 24 °C until use. 2.4. Transesterification of microalgal saponifiable lipids (SLs) In a typical experiment, 1.4 g of extracted and purified microalgal lipids (95 wt% SL purity) was mixed with 13.3 mL of solvent (hexane or t-butanol) (10 mL solvent/g SLs), 0.93 g of R. oryzae fungus (0.7 g of R. oryzae/g SLs) and 2.1 mL of methanol (methanol/SL molar ratio 11:1). Both the fungus and the methanol were added in several steps; for example, in three steps at the reaction times of 0, 24 and 48 h, adding 0.31 g of R. oryzae at each step and 573 lL of methanol in the first step (3 mol methanol/mol SLs) and 764 lL in the others

(4 mol of methanol/mol SLs per step). SLs are expressed as equivalent fatty acids and, therefore, these molar ratios are expressed as moles of methanol per mole of fatty acid in the microalgal SLs. The methanolysis reaction was carried out in 50 mL Erlenmeyer flasks with silicone-capped stoppers. The mixture was incubated at 35 °C and stirred in an orbital shaking air-bath (Inkubator 1000, Unimax 1010 Heidolph, Klein, Germany) at 130 rpm, for 72 h. This temperature was maintained constant in all the experiments because it is the temperature commonly used with this catalyst (Chisti, 2007; Li et al., 2008; Ban et al., 2002; Zeng et al., 2006). The reactions were stopped by separation of the lipase by filtration (glass plate, pore size range 16–50 lm) and then the reaction mixture was stored at 24 °C until it was analyzed. All reactions were carried out in duplicate and their corresponding analyses in triplicate; thus each value recorded is the arithmetic mean of six experimental data sets (data shown as mean value ± standard deviation). 2.5. Determination of total lipids (TLs) and saponifiable lipids (SLs); analysis of products and fractionations of extracted microalgal lipids TLs comprise both SLs and unsaponifiable lipids. SLs comprise neutral saponifiable lipids (NSLs), such as acylglycerols and free fatty acids; and polar lipids, such as glycolipids (GLs) and phospholipids (PLs). The TL content of the microalgal biomass was determined by the Kochet method (1978). The SL content of N. gaditana microalgal biomass was quantified by direct transesterification of the microalgal biomass to transform all SLs into FAMEs (Jiménez Callejón et al., 2014). FAMEs were then analyzed by gas chromatography (GC) using an Agilent Technologies 6890 gas chromatograph (Avondale, PA, USA). Further details of the analysis method are described in Navarro López et al. (2015). The SL yield (wt%) in the crude lipidic extracts from the microalgal biomass is the percentage of extracted SLs with respect to the total amount of SLs contained in the original biomass, both determined as equivalent fatty acids by GC. To determine the fatty acid amount in the hexane and the ethanol-water phases, a known volume was dried under an N2 stream, and methylation was carried out by direct transesterification with acetyl chloride/methanol (1:20 v/v) (Jiménez Callejón et al., 2014). The SL purity (wt%) is the weight percentage of SLs (determined by GC) with respect to the total amount of extracted lipids, determined by weighing after complete removal of the solvent contained in the lipid extract. In the lipid purification experiments by crystallization in acetone, a known volume of supernatant samples were taken and acetone was removed in the N2 stream. Then 1 mL of hexane and 50 lL of internal standard (prepared weighing 25 mg of nonadecanoic acid, 19:0, in 10 mL of hexane) were added. Samples were also methylated by the procedure described before (Jiménez Callejón et al., 2014). Conversion of SLs to FAMEs using transesterification catalyzed by the intracellular lipase from R. oryzae was determined following the procedure described in Martín et al. (2012). This conversion was calculated by the equation:

FAME conversion ð%Þ ¼

SL amount transformed to FAMEs Total SL amount convertible to FAMEs  100 ð1Þ

To determine the SL amount transformed to FAMEs (the equation numerator (1)), we took a known sample volume containing about 1 mg of fatty acids (around 20 lL) and mixed it with 50 lL of internal standard solution and 1 mL of hexane. This mixture was directly analyzed by GC. Then, these samples were methylated by direct transesterification with acetyl chloride/methanol and analyzed again by GC to obtain the denominator of Eq. (1). The internal standard solution was prepared dissolving 25 mg of 19:0 methyl ester in 10 mL of hexane.

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244 Table 2 Percentages of neutral saponifiable lipids (NSLs) and polar lipids (glycolipids, GLs and phospholipids, PLs) obtained by lipid fractionation (Section 2.5) of microalgal lipids extracted with ethanol–hexane (31% SLs), crystallized in acetone (95% SLs) and extracted with hexane and crystallized in acetone (61% SLs), and percentages of triacylglycerols (TAGs), diacylglycerols (DAGs) and free fatty acids (FFAs) of the NL fractions.b

a b

Lipid species

31% SLs

95% SLs

61% SLs

NSLsa TAGsb DAGsb FFAsb GLsa PLsa

26.9 ± 1.3 74.8 ± 1.3 8.1 ± 1.0 17.1 ± 1.4 61.6 ± 1.7 11.5 ± 0.7

51.0 ± 1.8 64.3 ± 0.2 8.7 ± 0.5 27.1 ± 0.3 43.5 ± 0.9 5.5 ± 1.9

62.6 ± 1.8 – – – 34.9 ± 0.8 2.5 ± 0.3

Percentages with respect to total SLs (NSLs + GLs + PLs). Percentages with respect to NSLs (TAGs + DAGs + FFAs).

The lipid extracts obtained from the wet microalgal biomass by the procedures described in Section 2.2 were fractionated into neutral lipids (NLs), GLs and PLs following the procedure described in Jiménez Callejón et al. (2014), which is based on the elution of lipids in Sep-pack classic cartridges (Waters Corporation, Milford, MA) with chloroform and acetone along with chloroform:methanol 85:15 v/v and methanol to collect the NLs, GLs and PLs from each of these mobile phases, respectively (Kates, 1986). The GC analysis of all fractions gave the percentage of NSLs, GLs and PLs with respect to SLs (Table 2). 3. Results and discussion 3.1. Lipid content and SL extraction from wet N. gaditana microalgae biomass The total lipid (TLs) and SL content of N. gaditana biomass were 31.0 ± 0.1 wt% and 12.0 ± 0.1 wt% in the dry biomass, respectively. This means that the microalgal biomass contains a high percentage of unsaponifiable lipids (19%), which are not convertible into fatty acid methyl esters (FAMEs, biodiesel). This lipid composition is a consequence of the culture conditions used for the microalgae (Jiménez Callejón et al., 2014; San Pedro et al., 2013). The main fatty acids contained in the microalgal SLs are palmitic acid (16:0, 23 wt% of total fatty acids), palmitoleic acid (16:1n7, 25 wt %) and eicosapentaenoic acid (EPA, 20:5n3, 30 wt%) (Navarro López et al., 2015). With the goal of attaining high SL yields, microalgal lipids were firstly extracted using the ethanol (96% v/ v)-hexane solvent system (Section 2.2). Using this procedure, 85.0 ± 2.1 wt% of SLs contained in the microalgal biomass were recovered. However, the SL purity obtained was low (31 wt% SL), which indicates that high amounts of lipids and non-lipid contaminants were extracted from the microalgal biomass. Table 2 shows the percentages of neutral saponifiable lipids (NSLs, TAG, DAG and FFAs) and polar lipids (GLs and PLs) present in the lipids extracted from the microalgal biomass using the ethanol (96%)-hexane system (31 wt% SLs). This table shows the high polar lipid percentages of the SLs extracted from the microalgae (61.6% of GLs and 11.5% of PLs), which contrasts with the low percentages of PLs in vegetable oils; for example, the mass fraction of PLs in crude soybean oil may range from 1.1% to 3.5% (Li et al., 2014). This high polar lipid content has a significant influence on the transformation of the microalgal SLs into FAMEs (Navarro López et al., 2015). 3.2. Synthesis of fatty acid methyl esters (FAMEs) by enzymatic transesterification The aim of this work was to optimize the transesterification conditions to obtain a high conversion of microalgal SLs to FAMEs. Optimization was also carried out taking lipase stability into

239

account to minimize the process cost. In this respect, the transesterification reactions were carried out using R. oryzae intracellular lipase as the biocatalyst, which was grown in our laboratory, and by adding both the lipase and the methanol by steps to reduce the methanol concentration and the contact time between methanol and lipase (Shimada et al., 1999). The experiments were carried out using organic solvents as the reaction media in order to reduce the reaction mixture viscosity, improve mass transfer and also preserve lipase stability. Transesterification was firstly conducted using t-butanol as the reaction medium because it is a solvent with intermediate polarity that dissolves neutral and polar lipids (Table 2), methanol and the reaction products (FAMEs and glycerol). Li et al. (2008) had already determined that R. oryzae stability was poor in a solvent-free system and could be improved considerably using t-butanol as the reaction medium; this difference was due to the accumulation of glycerol and FAMEs in the lipase cells, which is much greater in a solvent-free system and diminishes the reaction velocity by mass transfer limitation and product inhibition. Furthermore, the solubilization of methanol in t-butanol avoids, or decreases, lipase deactivation since this strongly polar reactant removes water from the active site of the enzyme (Shimada et al., 1999). To establish the R. oryzae/SL ratio, prior experiments were conducted with 95% microalgal SLs (Table 2) using ratios of 0.225, 0.7 and 1 g R. oryzae/g SLs. These experiments showed that by using 0.225 g R. oryzae/g SLs, a lower reaction velocity and FAME conversion (61% after 72 h) were obtained; while around 80% conversion was attained after this time with 0.7 g R. oryzae/g SL. No difference was observed between the reaction velocities and conversions obtained using 0.7 and 1 g R. oryzae/g SL (80% and 79.6%, respectively). For this reason, a R. oryzae/SL ratio (w/w) of 0.7 was chosen for future experiments. 3.2.1. Influence of the microalgal SL purity The microalgal lipids extracted using the ethanol–hexane system were only 31% pure (Table 2), although they were extracted with 85% yield. In order to determine whether the microalgal extract’s SL purity influenced the conversion of SLs to FAMEs, a crystallization operation was carried out in acetone (Section 2.2) to increase the SL purity. In this procedure, gums (phospholipids, PLs) and waxes are precipitated out because of their insolubility in acetone at low temperature and therefore SL purity increases because these lipids remain mostly solubilized in acetone (Rajam et al., 2005). In this way, approximately 90–95% of SLs were recovered and their purity increased from 31% to 95%. Table 2 shows that, using this procedure, the NSL content increased from 26.9% to 51.0%, while the polar lipid content (GLs and PLs) decreased from 73.1% to 49.0%, which is still a very high polar lipid percentage compared to the polar lipid content of vegetable seed oils. For example, Watanabe et al. (2002) found that the conversion to FAMEs from crude soybean oil (with only 0.08% phosphatidyl choline) was significantly lower (a mere 10% conversion) than that obtained from the degummed (28% conversion) and refined oil (30% conversion). In any case, these polar lipids (GLs and PLs) are also SLs that can be transformed to biodiesel and so no further purification steps were carried out to avoid too great a decrease in the FAME yield with respect to the total microalgal SLs. Fig. 1A shows that both the reaction velocity and the conversions are lower if lipids are not previously purified. In a previous work (Navarro López et al., 2015), a similar result was obtained using Novozym 435 as the catalyst and may be because 31% SLs have a higher polar lipid content than 95% SLs (Table 2). These polar lipids increase the reaction medium viscosity and might even be adsorbed on the lipase, increasing the lipase inhibition effect. This result is similar to that obtained by Wang et al. (2014), since these authors, using also t-butanol as the reaction medium, and

240

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

(A)

100

(B)

80

Conversion (%wt)

80

Conversion (%wt)

100

60

40

60

40 Methanol/SL 3:1 Methanol/SL 6:1 Methanol/SL 11:1 Methanol/SL 16:1 Methanol/SL 23:1

31% SL 95% SL

20

20

0

0 0

20

40

60

80

Reaction time(h)

0

20

40

60

80

Reaction time (h)

Fig. 1. Influence of microalgal SL purity (31% and 95%) (A) and methanol/SL molar ratio (B) on the FAME conversion. Conditions: (A) 1.4 g lipids, 0.7 w/w R. oryzae/SL ratio, 11:1 methanol/SL molar ratio; (B) 1.4 g lipids (95% pure SLs), 1 w/w R. oryzae/SL ratio; (A and B) 10 mL t-butanol/g SLs, three additions of methanol and fungus at 0, 24 and 48 h, 130 rpm, 35 °C, 72 h.

Novozym 435 as the catalyst, obtained lower efficiencies in the transformation of crude microalgal oils to biodiesel, the higher their polar lipid content was. On the other hand, Table 2 shows that the FFA content increases after the acetone crystallization treatment (from 17% to 27%), and esterification of FFAs is a much faster reaction than transesterification of acylglycerols (Castillo López et al., 2015). This might also explain the higher reaction velocity observed using 95% SLs (Fig. 1A). Therefore, one can conclude that purifying SLs from 31% to 95% improves both the reaction velocity and the conversion, as well as increasing the final biodiesel purity. 3.2.2. Influence of R. oryzae cell immobilization within BSPs In order to ascertain whether R. oryzae immobilization within BSPs could improve both the conversions attained and the lipase stability, experiments were carried out to compare the results obtained using both non-immobilized and immobilized R. oryzae cells. Fig. 2A shows that higher conversions were obtained in the two experiments carried out using the non-immobilized lipase; in fact, a higher final conversion was even achieved using 0.7 g of non-immobilized R. oryzae/g SL than using 1 g of immobilized R. oryzae/g SL; although in this case the reaction velocities were similar until around 48 h. In any case, the minimum reaction velocity and conversions were obtained using 0.7 g of immobilized R. oryzae/g SL. Fig. 2A also shows that small differences exist between the use of 0.7 or 1 g of non-immobilized R. oryzae/SLs at 96 h (80.1 ± 0.2 and 81.8 ± 0.6, respectively) and for this reason, the lower lipase amount was chosen. Zeng et al. (2006) obtained a FAME yield slightly higher (86%) in the methanolysis of soybean oil catalyzed by R. oryzae whole cells without immobilization; although in this case, with soybean oil, a 5 wt% R. oryzae/oil ratio was sufficient (Table 1). In contrast to these results, practically all the researchers that use R. oryzae cells use the immobilized form; although almost everybody uses vegetable oils containing no polar lipids. This seems to indicate that polar lipids affect immobilized lipase more than non-immobilized lipase. This inhibitory effect may result from the binding of polar lipids with the immobilized lipase/carrier. To prove this hypothesis, the BSPs used as catalysts were separated from the reaction mixture at a reaction time of 7 h (see conditions in the Fig. 2A caption) and extracted three times with 90 mL t-butanol/g BSPs per step. This experiment showed that around 0.20 g SLs per gram of BSPs were retained. The

fractionation of these retained SLs into NLs, GLs and PLs (Section 2.5) gave a lipidic profile similar to the initial SL profile (Table 2; 95% SLs). Although the lipid nature changes through the course of the reaction and a more detailed study regarding this lipid adsorption process is required, this result shows that a significant retention of lipids (including polar lipids) occurs on the BSPs – having the influence of lower reaction velocities and FAME conversion when R. oryzae immobilized in BSPs are used as the catalyst. Similarly, Watanabe et al. (2002) showed that PLs bound to the immobilized preparation of C. antarctica lipase and interfered in the interaction of the lipase molecule with substrates; this implied that, in the methanolysis of soybean oil, the lower the FAME yield, the higher the PL content. Although the reaction velocities and conversions obtained using the immobilized catalyst were smaller than with the nonimmobilized catalyst, experiments were carried out to determine the lipase stability of the immobilized cells within BSPs. A lipase batch was used to catalyze two successive transesterification reactions, each of which lasted 96 h. Fig. 2B shows that, after the first use of the immobilized lipase, an appreciable diminution of reaction velocity and transesterification conversion occurred (from around 70% to 43% at 96 h in only two reaction cycles). In this respect, Li et al. (2008) observed that a considerable amount of reaction products (FAMEs and glycerol) accumulated inside the cell during the repeated use of BSPs even using t-butanol as the reaction medium, decreasing the enzymatic activity of the whole cell immobilized within BSPs with the number of uses. Hama et al. (2007) also observed that the FAME conversion decreased sharply (from 90% to 10% in the tenth cycle) with the increasing numbers of batch cycles in the methanolysis of soybean oil catalyzed by immobilized R. oryzae within BSPs in a shaken bottle. Therefore, the accumulation of polar lipids along with the accumulation of glycerol and FAMEs on the BSP surface could be the reason for the low reaction velocities, conversions and lipase stability observed in this work using microalgal oil (Fig. 2). Given that the conversion to FAMEs was higher using the nonimmobilized whole-cell catalyst (81.8%) rather than the immobilized one (70.3%) (Fig. 2A) and that the stability of the immobilized whole-cell lipase was not as high as could be expected (Fig. 2B), free R. oryzae cells were selected as the catalyst in order to obtain higher conversions. This choice also reduces the process costs by eliminating the immobilization of the R. oryzae cells.

241

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

(a)

100

(b)

80

80

Conversion (%wt)

Conversion (%wt)

60

60

40

0.7g RO/g SL

40

20

Cycle 1

1g RO/g SL

20

Cycle 2

0.7g BSPs/g SL 1g BSPs/g SL

0

0 0

20

40

60

80

100

Reaction time (h)

0

20

40

60

80

100

Reaction time (h)

Fig. 2. Influence of the immobilization of R. oryzae cells within BSPs and lipase/SL ratio (A) and the influence of the number of uses of a BSPs batch on FAME conversion (B). Conditions: 1.4 g lipids (95% SLs), 10 mL t-butanol/g SL, 11:1 methanol/SL molar ratio, 35 °C, 96 h, 4 additions of methanol and cells at 0, 24, 48 and 72 h. (A) (d) 0.7 g of R. oryzae cells/g SLs, (s) 1 g of R. oryzae cells/g SLs, (.) 0.7 g of immobilized R. oryzae cells within BSPs/g SLs, (4) 1 g of immobilized R. oryzae cells within BSPs/g SL. (B) Reuse of a batch of R. oryzae immobilized cells to catalyze two reactions using 1 g of immobilized R. oryzae cells.

3.2.3. Influence of methanol/SL molar ratio on the conversion to FAMEs Fig. 1B shows that, as the methanol/SL ratio increases, so does both the reaction velocity and the FAME conversion, up to an 11:1 methanol/SL molar ratio, since the FAME conversions obtained at 72 h with 16:1 and 23:1 methanol/SL molar ratios were lower than with an 11:1 molar ratio. The lowest methanol/SL ratios used (3:1 and 6:1) are not enough to carry out a more complete equilibrium displacement towards the products; and the highest ratios (16:1 and 23:1) could deactivate the R. oryzae lipase due to the significant excess of methanol in the reaction mixture, even in the presence of t-butanol. The stoichiometric methanol to oil ratio is 3:1 but an excess of methanol is needed to shift the reaction towards the products. A similar optimal methanol/oil molar ratio (12:1) was obtained by Arumugan and Ponnusami (2014) in the production of biodiesel by the methanolysis of Calophyllum inophyllum oil catalyzed by R. oryzae in a solvent-free medium; a maximum FAME yield of 92% was attained using a 12:1 methanol/oil molar ratio (Table 1) but decreased when a 16:1 molar ratio was used. In general, the methanol/SL molar ratio used depends on whether the reaction is carried out in a solvent-free media or if a solvent is used. In solvent free-media, low methanol/SL molar ratios are usually employed to avoid or decrease lipase deactivation by methanol (Table 1). So, for example, Tamalampudi et al. (2008) found that, above a 1:1 methanol/oil molar ratio, the activity of R. oryzae decreased sharply in the methanolysis of jatropha oil. However, with t-butanol, higher methanol/SL ratios can be used to displace the reaction equilibrium towards FAME formation. Thus, for example, Wang et al. (2014) achieved a 99.1% conversion efficiency using a 12:1 methanol/oil molar ratio in the transesterification of a crude algal oil catalyzed by C. antarctica lipase in tbutanol medium. 3.2.4. Influence of substrate concentration and solvent type Experiments at different substrate concentrations or solvent/SL ratios (between 5 and 20 mL solvent/g SL) were carried out to try to increase the conversion. Fig. 3A shows that the highest reaction velocity and conversion were obtained at the lowest t-butanol/SL ratios (5 and 10 mL/g) or the highest substrate concentrations. Using 5 mL/g, a slightly higher reaction velocity was obtained but finally the maximal conversion was achieved using 10 mL/g. There was no difference between the conversions attained with 15 and

20 mL t-butanol/g SL. This is the result of the two ways in which the substrate concentration affects the reaction velocity: on the one hand, an increase in this variable increases the velocity of the enzymatic reaction itself but at a high substrate concentration, the reaction mixture viscosity increases and the lower mass transfer velocity can control the observable reaction velocity. On the other hand, due to the lipase deactivation observed (Fig. 2B) hexane was also tested as the reaction medium. Some authors used hexane as the solvent (Duarte et al., 2015; Fjerbaek et al., 2009) since it is non-polar (log P = 3.5), thus trapping the water around the enzyme and helping to maintain its active conformation and preserve the catalytic activity (Chakravorty et al., 2012). Fig. 3B shows that when using hexane the initial reaction velocities increased only slightly with the substrate concentration (except for 5 mL/g) and similar conversions were attained using 10, 15 and 20 mL hexane/g SL. The use of 5 mL hexane/g SL leads to low FAME conversions (35% after 72 h), probably due to the high mass transfer resistance and higher lipase deactivation caused by the non-solubilized methanol. Comparing the conversions achieved with both solvents, Fig. 3A and B indicate that, although the initial reaction velocities are higher using hexane (44% and 36% at 10 h, using hexane and t-butanol, respectively, with 10 mL solvent/g SL), the final conversion to FAMEs is much higher using t-butanol (51.6 ± 1.5% and 81 ± 1.8% using hexane and t-butanol, respectively). The higher reaction velocity obtained with hexane might be related to the greater lipase activity in the non-polar solvents (high log P values). Laane et al. (1987) demonstrated that polar solvents (t-butanol and hexane have log P values of 0.35 and 3.9, respectively) distort the essential water layer that lipase requires to maintain its activity, decreasing both the lipase activity and the reaction velocity. Furthermore, Li et al. (2010) proved that the acyl-migration velocity is faster in hexane than in t-butanol, and acyl-migration is required to transform 1,2-DAG to 1,3-DAG and 2-MAG to 1(3)-MAG to attain high FAME yields due to the 1,3-specificity of the R. oryzae lipase. The rapid reaction velocity decrease in hexane is in accordance with the fact that hexane is a worse solvent of glycerol, methanol and polar lipids than t-butanol. The low solubility of glycerol and polar lipids in hexane prevents these compounds being removed from the lipase surface, causing a decrease in the reaction velocity over long reaction times due to mass transfer limitations. In the

242

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

(A)

100

Conversion (%wt)

80

60

40 5mL t-butanol/g SL 10mL t-butanol/g SL 15mL t-butanol/g SL 20mL t-butanol/g SL

20

0 0

20

40

60

80

Reaction time (h)

(B)

70

Conversion (%wt)

60 50 40 30 5 mL hexane/g SL 10 mL hexane/g SL 15 mL hexane/g SL 20 mL hexane/g SL

20 10 0 0

20

40

60

80

Reaction time (h)

(C)

100

80

Conversion (% wt)

microbial oil because of the higher polar lipid and unsaponifiable matter contents of the latter. As a consequence of the previous results, experiments were also carried out using a mixture of tbutanol/hexane 1:1 v/v as the reaction medium. In this solvent mixture (Fig. 3C), the FAME conversion presents intermediate behavior, i.e. the initial reaction velocity is similar (or a little lower) to that achieved in hexane, and the final conversion is higher than that obtained in hexane but lower than that obtained with tbutanol. Therefore, the optimal conditions to carry out the transesterification of N. gaditana SLs, obtaining the maximal FAME conversion (81.4%) were: 95% pure microalgal SLs, 10 mL t-butanol/g SL, 0.7 g R. oryzae/g SL, 11:1 methanol/SL molar ratio; at 35 °C, 130 rpm, over 72 h and with three additions of methanol and lipase at 0, 24 and 48 h. Other authors obtained similar FAME conversions using R. oryzae as the biocatalyst. For example, Arumugan and Ponnusami (2014) attained a 92% yield by the methanolysis of C. inophyllum oil (an oil with a high free fatty acid content) using a 12:1 methanol:oil molar ratio, 20 wt% of R. oryzae cells over 72 h, adding the methanol at regular time intervals of 24 h (Table 1). However, these authors used vegetable oils as the feedstock and few works exist regarding biodiesel production using microalgal lipids and R. oryzae as the biocatalyst (Table 1).

60

40 10 mL hexane/g SL 10 mL t-butanol/g SL 10 mL hexane:t-butanol/g SL (1:1)

20

0 0

20

40

60

80

Reaction time (h) Fig. 3. Influence of the solvent type and solvent/SL ratio on the FAME conversion. (A) t-butanol, (B) hexane, (C) comparison between t-butanol, hexane and the mixture t-butanol:hexane 1:1 v/v. Conditions: 1.4 g lipids (95% SLs), 0.7 w/w R. oryzae/SL ratio, 11/1 methanol/SL molar ratio, 72 h, 3 additions of methanol and lipase at 0, 24 and 48 h.

case of methanol, its low solubility in hexane determines that the lipase can be partially deactivated during the reaction. In this respect, Duarte et al. (2015) obtained higher FAME yields in hexane than in t-butanol in the transesterification of crude microbial oil from Candida sp. LEB-M3, also catalyzed by the R. oryzae lipase – but this oil only contains 2.1% polar lipids (much lower than the 49% present in the microalgal oil used in this work, Table 2) and hexane is enough to avoid this low polar lipid content from creating mass transfer limitations. These authors also obtained higher reaction velocities and FAME yields with olive oil than with crude

3.2.5. Reuse of lipase under the optimal conditions Fig. 4 shows the conversions achieved in three consecutive transesterification reactions catalyzed by the same R. oryzae batch under the conditions previously established as optimal, using hexane and t-butanol as solvents, in order to check the lipase stability in the two solvents. After each reaction cycle, the R. oryzae cells were filtrated and washed with t-butanol before subsequent use, using 50 mL of t-butanol/g lipase. This figure shows that both the FAME conversion and the reaction velocity decreased with the number of uses of the same R. oryzae batch. Using t-butanol as the solvent (Fig. 4A), the final conversion achieved after 72 h decreased by about 20% in each reaction cycle (maximum conversions in each cycle: 81%, 56% and 34%, respectively). The activity loss was more evident when hexane was used as the reaction medium (Fig. 4B) since the final conversion decreased by almost 40% in the second catalysis cycle with respect to the first (52% in cycle 1 and 10% in cycle 2) and activity completely disappears in the third catalysis cycle. These results show that, under the conditions established as optimal, it seems that R. oryzae lipase deactivation occurs and this activity loss is much more pronounced if hexane is used as the solvent instead of t-butanol. In addition, Duarte et al. (2015) obtained a 30% reduction in yield after 24 h, carrying out the reaction in hexane, using microbial oil as the feedstock and recombinant R. oryzae as the catalyst. Li et al. (2008) observed that the FAME yield decreased by 15% (from 65% to 55%) after six uses of the same R. oryzae batch in the methanolysis of soybean oil in tbutanol medium (Table 1). These authors showed that glycerol and FAMEs accumulated on the immobilized R. oryzae surface even in the presence of t-butanol, and while the accumulated glycerol influenced whole-cell stability, solely through mass transfer limitation, the accumulated FAMEs influenced whole-cell stability through both mass transfer limitation and product inhibition. In addition, Li et al. (2014), in the methanolysis of soybean oil, showed that the stability of the non-immobilized lipase NS81006 (from the genetically-modified Aspergillus niger) decreased as the phospholipid content increased. The microalgal lipids used in the present work contain much higher concentrations of polar lipids and therefore these substances, along with the reaction products (FAMEs and glycerol), contribute to the diminution in FAME conversion observed during the repeated use of the R. oryzae cells as catalysts.

243

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

(A)

100

(B)

80

Conversion (% wt)

80

Conversion (% wt)

100

60

40

60

40 Cycle 1 Cycle 2 Cycle 3

20

20

Cycle 1 Cycle 2 Cycle 3

0

0 0

20

40

60

80

0

20

40

Reaction time (h)

60

80

Reaction time (h)

Fig. 4. Influence of the solvent type and the number of uses of a batch of R. oryzae lipase on the FAME conversion. Conditions: 1.4 g lipids, 10 mL solvent/g SL, 0.7 w/w R. oryzae/SL ratio, 11/1 methanol/SL molar ratio, 72 h, 3 additions of methanol and lipase at 0, 24 and 48 h. (A) t-butanol, (B) hexane.

(A)

100

(B)

80

Conversion (wt%)

Conversion (%wt)

80

100

60

40

60

40

Ethanol + hexane extraction

Cycle 1

Hexane extraction

20

20

Cycle 2 Cycle 3

0

0 0

20

40

60

80

Reaction time (h)

0

20

40

60

80

Reaction time (h)

Fig. 5. Influence of the lipid extraction method on the FAME conversion (A) and the influence of the number of uses of a batch of R. oryzae lipase on the reaction conversion (B). Conditions: 1.4 g lipids, 10 mL t-butanol/g SL, 0.7 w/w R. oryzae/SL ratio, 11/1 methanol/SL molar ratio, 72 h, 3 additions of methanol and lipase at 0, 24 and 48 h.

3.2.6. Transesterification reaction of SLs with lower polar lipid content We observed in this and in a previous work (Navarro López et al., 2015) that polar lipids affect both the reaction velocity and the lipase stability. To prove this influence, microalgal lipids with a lower polar lipid content were obtained. To achieve this, microalgal lipids were extracted using only hexane, with previous homogenization of the wet biomass at 1700 bar (Section 2.2, Jiménez Callejón et al., 2014). Using this procedure, a lower SL extraction yield (57%) was achieved to that using the ethanol–hexane system (85%) – but the SLs are purer (56% SLs) than those obtained using the latter (31%). The acetone crystallization treatment was also applied to these lipids, but in this case, a low precipitate amount was formed and the SL purity only increased to 61% (below the

95% SL purity attained from the lipids extracted with ethanol–hexane). This result indicates that the lipids extracted with hexane contain a low, or negligible, amount of waxes and gums (PLs) susceptible to precipitate in acetone at low temperature. Thus, 61% pure SLs were used in the transesterification reaction catalyzed by R. oryzae. Table 2 shows that these lipids are richer in neutral lipids (NLs) (62.6% versus 51%) and poorer in polar lipids (34.9% GLs and 2.5% PLs versus 43.5% and 5.5%, respectively); this is because hexane is a non-polar solvent that is more selective towards the neutral lipids than towards polar lipids. These lipids were transesterified with methanol under the previously obtained optimal conditions. Fig. 5A shows that the reaction velocity and FAME conversions were higher with the SLs

244

E. Navarro López et al. / Bioresource Technology 203 (2016) 236–244

extracted using hexane, although these SLs had a lower purity (61% versus 95%). Therefore, this result was clearly due to the lower percentage of polar lipids (and higher proportion of neutral saponifiable lipids) of the hexane extracted SLs (Table 2). Fig. 5B shows the conversions achieved using the hexane extracted SLs in three reaction cycles of 72 h each, catalyzed by the same R. oryzae lipase batch. After each reaction cycle, the R. oryzae batch was filtrated and washed with t-butanol. This figure shows that FAME conversion decreased from 83% to 71% (14% with respect to the first FAME conversion) after three consecutive uses of the lipase batch, each of 72 h. However, this conversion decrease is lower than that obtained using ethanol–hexane extracted lipids (from 81% to 34%; i.e. a decrease of 58%, Fig. 4A). Therefore, a lower conversion decrease was obtained using the hexane-extracted SLs, despite these SLs having lower SL purity (61%), than the ethanol–hexane extracted SLs (95% SL), which demonstrates that the polar lipids are also responsible for this conversion decrease following the repeated use of lipase, as previously discussed in Section 3.2.5. Table 1 compares the conditions and results obtained in this and other studies. The table shows that there is extensive research regarding biodiesel production from vegetable oils but little using other oils. Although comparison is difficult due to the different ways of expressing the conditions, Table 1 shows that, under similar conditions, the FAME conversion obtained in this work is in the range of that obtained by other researchers using vegetable, yeast or microalgal oils. However, the oil used in the present work has much higher polar lipid percentages than those used in the other studies, which determines that no solvents were used by those authors. However, as previously mentioned, solvent is important when the polar lipid content is relatively high, as demonstrated by Li et al. (2008) and as has been shown in this work. 4. Conclusions An 83% FAME conversion was obtained by transesterification of SLs extracted from wet biomass with hexane using a 70% w/w R. oryzae/SL ratio, 11:1 methanol/SL molar ratio, 10 mL t-butanol/g SLs, over 72 h. The microalgal lipid composition has a strong influence on the FAME yield. Thus, from SLs extracted with ethanol, lower reaction velocities and conversions were attained with 31% pure SLs (58% conversion) than with 95% pure SLs (80%). Both the reaction velocities and conversions, obtained using the same R. oryzae batch to catalyze successive reactions, were higher for hexane extracted SLs and, therefore, had lower polar-lipid contents. Acknowledgements This research was supported by grants from the Ministerio de Economía y Competitividad (Spain), Project CTQ2010-16931. This project was co-funded by the FEDER (European Fund for Regional Development). References Araya, K., Ugarte, A., Azócar, L., Valerio, O., Wick, L.Y., Ciudad, G., 2015. Whole cell three phase bioreactors allow for effective production of fatty acid alkyl esters derived from microalgae lipids. Fuel, 144–152. Arumugan, A., Ponnusami, V., 2014. Biodiesel production from Calophyllum inophyllum oil using lipase producing Rhizopus oryzae cells immobilized within reticulated foams. Renewable Energy 64, 276–282. Ban, K., Hama, S., Nishizuka, K., Kaieda, M., Matsumoto, T., Kondo, A., 2002. Repeated use of whole-cell biocatalysts immobilized within biomass support particles for biodiesel fuel production. J. Mol. Catal. B Enzym. 17, 157–165. Canet, A., Benaiges, M.D., Valero, F., 2014. Biodiesel synthesis in a solvent-free system by recombinant Rhizopus oryzae lipase. Study of the catalytic reaction progress. J. Am. Oil Chem. Soc. 91, 1499–1506.

Castillo López, B., Esteban Cerdán, L., Robles Medina, A., Navarro López, E., Martín Valverde, L., Hita Peña, E., González Moreno, P.A., Molina Grima, E., 2015. Production of biodiesel from vegetables oils and microalgae by fatty acid extraction and enzymatic esterification. J. Biosci. Bioeng. 119, 706–711. Chakravorty, D., Parameswaran, S., Dubey, V.K., Patra, S., 2012. Unraveling the rationale behind organic solvent stability of lipases. Appl. Biochem. Biotechnol. 167, 439–461. Chen, G., Ying, M., Weizhun, L., 2006. Enzymatic conversion of waste cooking oils into alternative fuel-biodiesel. Appl. Biochem. Biotechnol. 129–132, 911–921. Chisti, Y., 2007. Biodiesel from microalgae. Biotechnol. Adv. 25, 294–306. Duarte, S.H., del Peso Hernández, G.L., Canet, A., Benaiges, M.D., Maugeri, F., Valero, F., 2015. Enzymatic biodiesel synthesis from yeast oil using immobilized recombinant Rhizopus oryzae lipase. Bioresour. Technol. 183, 175–180. Fjerbaek, L., Christensen, K.V., Norddahl, B., 2009. A review of the current state of biodiesel production using enzymatic transesterification. Biotechnol. Bioeng. 102, 1298–1315. Hama, S., Yamaji, H., Fukumizu, T., Numata, T., Tamalampudi, S., Kondo, A., 2007. Biodiesel-fuel production in a packed-bed reactor using lipase-producing Rhizopus oryzae cells immobilized within biomass support particles. Biochem. Eng. J. 34, 273–278. Jiménez Callejón, M.J., Robles Medina, A., Macías Sanchez, M.D., Hita Peña, E., Esteban Cerdán, L., González Moreno, P.A., Molina Grima, E., 2014. Extraction of SL from wet microalgal biomass for biodiesel production. Bioresour. Technol. 169, 198–205. Kates, M., 1986. Definition and classification of lipids. In: Burdon, R.H., van Knippenber, P.H. (Eds.), Techniques of Lipidology, Isolation, Analysis and Identification of Lipids. Elsevier Science Publishers, Amsterdam. Khor, G.K., Sim, J.H., Kamaruddin, A.H., Uzir, M.H., 2010. Thermodynamics and inhibition studies of lipozyme TL IM in biodiesel production via enzymatic transesterification. Bioresour. Technol. 101, 6558–6561. Kochet, G., 1978. Quantitation of the macromolecular components of microalgae. In: Hellebust, J., Cragie, S. (Eds.), Handbook of Phycological Methods. Physiological and Biochemical Method. Cambridge University Press, London. Laane, C., Boeren, S., Vos, K., Veeger, C., 1987. Rules for optimization of biocatalysis in organic solvents. Biotechnol. Bioeng. 30, 81–87. Li, W., Du, W., Liu, D., 2007. Rhizopus oryzae IFO 4697 whole cell catalyzed methanolysis of crude and acidified rapeseed oils for biodiesel production in tert-butanol system. Process Biochem. 42, 1481–1485. Li, W., Du, W., Liu, D., Yao, Y., 2008. Study on factors influencing stability of whole cell during biodiesel production in solvent free and t-butanol system. Biochem. Eng. J. 41, 111–115. Li, W., Du, W., Li, Q., Li, R.-W., Liu, D., 2010. Dependence of the properties of organic solvent: study on acyl migration kinetics of partial glycerides. Bioresour. Technol. 101, 5737–5742. Li, X., He, X.-Y., Li, Z.-L., Wang, Y.-D., Wang, C.-Y., Shi, H., Wang, F., 2012. Enzymatic production of biodiesel from Pistacia chinensis bge seed oil using immobilized lipase. Fuel 92, 89–93. Li, Y., Du, W., Dai, L., Liu, D., 2014. Kinetic study on free lipase NS81006-catalyzed biodiesel production from soybean oil. J. Mol. Catal. B Enzym. 121, 22–27. Martín, L., González, P., Rodríguez, A., Hita, E., Jiménez, M., Esteban, L., Molina, E., Robles, A., 2012. Concentration of docosahexaenoic acid (DHA) by selective alcoholysis catalyzed by lipases. J. Am. Oil Chem. Soc. 89, 1633–1645. Navarro López, E., Robles Medina, A., González Moreno, P.A., Jiménez Callejón, M.J., Esteban Cerdán, L., Martín Valverde, L., Castillo López, B., Molina Grima, E., 2015. Bioresour. Technol. 187, 346–353. Rajam, L., Kumar, D.R.S., Sundaresan, A., Arumughan, C., 2005. A novel process for physically refining rice bran oil through simultaneous degumming and dewaxing. J. Am. Oil Chem. Soc. 82, 213–220. Robles-Medina, A., Gónzalez-Moreno, P.A., Esteban-Cerdán, L., Molina-Grima, E., 2009. Biocatalysis: towards ever greener biodiesel production. Biotechnol. Adv. 27, 398–408. San Pedro, A., González, C., Acién, G., Molina, E., 2013. Marine microalgae selection and culture conditions optimization for biodiesel production. Bioresour. Technol. 134, 353–361. Shimada, Y., Watanabe, Y., Samukawa, T., Sugihara, A., Noda, H., Fukuda, H., Tominaga, Y., 1999. Conversion of vegetable oil to biodiesel using immobilized Candida antárctica lipase. J. Am. Oil Chem. Soc. 76, 789–793. Stamenkovik, O.S., Velicˇkovic´, A.V., Velijkovic´, V.B., 2011. The production of biodiesel from vegetables oils by ethanolysis: current states and perspectives. Fuel 90, 3141–3155. Tamalampudi, S., Talukder, M.R., Hama, S., Numata, T., Kondo, A., Fukuda, H., 2008. Enzymatic production of biodiesel from Jatropha oil: a comparative study of immobilized-whole cell and commercial lipases as a biocatalyst. Biochem. Eng. J. 39, 185–189. Wang, Y., Liu, J., Gerken, H., Zhang, C., Hu, Q., Li, Y., 2014. Highly-efficient enzymatic conversion of crude algal oils into biodiesel. Bioresour. Technol. 172, 143–149. Watanabe, Y., Shimada, Y., Sugihara, A., Yoshio, T., 2002. Conversion of degummed soybean oil to biodiesel fuel with immobilized Candida antarctica lipase. J. Mol. Catal. B. Enzym. 17, 151–155. Zeng, J., Du, W., Liu, X., Liu, D., Dai, L., 2006. Study on the effect of cultivation parameters and pretreatment on Rhizopus oryzae cell-catalyzed transesterification of vegetable oils for biodiesel production. J. Mol. Catal. B: Enzym. 43, 15–18.