Quantification and characterisation of fatty acid methyl esters in microalgae: Comparison of pretreatment and purification methods

Accepted Manuscript Quantification and characterisation of fatty acid methyl esters in microalgae: comparison of pretreatment and purification methods Sandra Lage, Francesco G. Gentili PII: DOI: Reference:

S0960-8524(18)30269-4 https://doi.org/10.1016/j.biortech.2018.01.153 BITE 19582

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

26 October 2017 8 January 2018 10 January 2018

Please cite this article as: Lage, S., Gentili, F.G., Quantification and characterisation of fatty acid methyl esters in microalgae: comparison of pretreatment and purification methods, Bioresource Technology (2018), doi: https:// doi.org/10.1016/j.biortech.2018.01.153

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Quantification and characterisation of fatty acid methyl esters in microalgae: comparison of pretreatment and purification methods Sandra Lage, Francesco G. Gentili* Department of Wildlife, Fish, and Environmental Studies, Swedish University of Agricultural Sciences, 901 83 Umeå, Sweden *Corresponding author. E-mail: [email protected] Abstract A systematic qualitative and quantitative analysis of fatty acid methyl esters (FAMEs) is crucial for microalgae species selection for biodiesel production. The aim of this study is to identify the best method to assess microalgae FAMEs composition and content. A single-step method, was tested with and without purification steps—that is, separation of lipid classes by thin-layer chromatography (TLC) or solid-phase extraction (SPE). The efficiency of a direct transesterification method was also evaluated. Additionally, the yield of the FAMEs and the profiles of the microalgae samples with different pretreatments (boiled in isopropanol, freezing, oven-dried and freeze-dried) were compared. The application of a purification step after lipid extraction proved to be essential for an accurate FAMEs characterisation. The purification methods, which included TLC and SPE, provided superior results compared to not purifying the samples. Freeze-dried microalgae produced the lowest FAMEs yield. However, FAMEs profiles were generally equivalent among the pretreatments.

1

Keywords Direct transesterification, lipids, solid-phase extraction, thin-layer chromatography, wastewater 1. Introduction Microalgae are currently being investigated worldwide as a promising sustainable and renewable energy source to meet future energy demands for liquid transportation fuels. In particular, the production of biodiesel coupled with wastewater treatment has been proposed as a cost-effective and feasible alternative to fossil fuels (Pulz & Gross, 2004). The harvest of microalgae biomass and the extraction of lipids (such as for crude oil) is an energy-intensive process because it requires the separation and dewatering of microalgae, biomass pretreatment and the subsequent extraction and purification of lipids. The crude lipids extracted can contain neutral lipids, free fatty acids and polar lipids, which include galacto- and phospholipids (Berge et al., 1995). However, only neutral lipids (namely triglycerides [TAGs]) which can be converted into fatty acid methyl esters (FAMEs) are suitable for biodiesel production. Methods for the accurate determination of both the total quantity and the type of FAMEs in microalgae are needed. At present, there are many available methods, though there is little consensus on the best methods for FAMEs quantification and characterisation. This makes it difficult to do a comparison between microalgae species that have been extracted with different methods because the FAMEs content is dependent not only on microalgae species and culture conditions but also the extraction methods used (Converti et al., 2009). Furthermore, only a limited number of methods have been reported to date with direct application to the biodiesel industry.

2

The first step in lipid analysis is to extract all of the available lipids either directly from the wet microalgae biomass or from pretreated samples. Pretreatment of the harvest biomass is necessary to increase the quality of the stored biomass for later lipid extraction this is necessary because when fresh microalgae is stored, a degradation process (lipolysis) occurs. The lipolysis releases free fatty acids from the glycerol backbone of the microalgae lipids, which is evident by the increase in the amount of free fatty acids and the decrease in the total lipids (Raven et al., 2005). The treatments prior to biomass storage may include boiling the wet biomass with isopropanol and exposing the samples to gaseous nitrogen to kill cells and deactivate lipases (Kates, 1972; Radwan, 1984), freezing the wet biomass or drying it. The drying of the biomass may be achieved by natural sun heat or by freeze-, drum-, oven-, spray- and fluidised bed-drying. Although most studies reveal no effects of freezing or sun-, spray-, ovenand freeze-drying on the microalgae total lipid content (Babarro et al., 2001; Morist et al., 2001; Ryckebosch et al., 2011; Ryckebosch et al., 2012), Esquivel et al. (1993) reported losses of 70% of the lipid content immediately after freeze-drying the microalgae biomass. Crude lipid extraction efficiency is dependent on the polarity of the solvent mixture. At present, various solvents or solvent combinations have been suggested as crude lipid extractants (Halim et al., 2012; Lee et al., 2010; Ramluckan et al., 2014). However, most studies use a chloroform and methanol extraction based on either Folch et al. (1957) or Bligh and Dyer (1959). Although the Bligh and Dyer (1959) method is widely popular (currently cited over 44,000 times), it has often been incorrectly applied or modified, which has resulted in significant underestimation of the total lipid content (2 to 50%) (Iverson et al., 2001; Palmquist & Jenkins, 2003). Furthermore, the Bligh

3

and Dyer (1959) method was originally described for tissue samples with low lipid content (< 2%). A later adaptation reported that an increase in the chloroform-tomethanol ratio could lead to accurate extraction in samples with a lipid content higher than 6% (Lee et al., 1996). This is still unsuitable for microalgae, considering microalgae species studied for biofuel production have an average of 25.5% lipids per dry weight (Hu et al., 2008b). The Folch et al. (1957) method had, instead, produced more reliable results (Axelsson & Gentili, 2014; Halim et al., 2012; Iverson et al., 2001; Ramluckan et al., 2014; Soares et al., 2014) compared to Bligh and Dyer (1959). Recently, a faster and easier procedure based on Folch et al. (1957) was developed that is a single-step method that uses twice the volume of solvents (Axelsson & Gentili, 2014). Crude lipid yields of the single-step and the Folch et al. (1957) methods were identical. In addition, the single-step method excluded a cell disruptive treatment step (Axelsson & Gentili, 2014). Although several studies report higher lipid extractabilities when cell disruption is applied (Lee et al., 2010; Ranjan et al., 2010), the single-step method showed no substantial increase in lipid yield when microwave, Potter–Elvehjem and sonication were performed (Axelsson & Gentili, 2014). Accordingly, Ryckebosch et al. (2012) found no significant improving of cell disruption techniques, e.g. freezedry, freeze-dry and sonication, freeze-dry and liquid nitrogen freeze – thaw cycle and freeze-dry and bead beating, in the amount of total lipids of Phaeodactylum tricornutum extracted a single time with 1:1 chloroform:methanol (v/v). After extraction, crude lipids are quantified gravimetrically or by chromatography (Mansour et al., 2005; Widjaja et al., 2009). Gravimetric methods are the simplest procedure to measure total lipid content because they are a direct measurement of the lipid weight. However, the lipid estimations are not accurate

4

because the crude extract contains not only lipids, but also carbohydrates, proteins and pigments (Palmquist & Jenkins, 2003; Pruvost et al., 2009). In addition, it does not provide sufficient information for biodiesel production process development as it cannot determine the total content or composition of the different FAMEs (Han et al., 2011). Alternatively, gas chromatography (GC) can be used to quantify individual fatty acids and the total concentration of the fatty acids present in a crude lipid extract. Before GC analysis, fatty acids need to be derivatised through a process named transmethylation (that is, the conversion of fatty acid components into corresponding methyl esters, known as FAMEs), which changes the volatility of lipid components and improves peak shape and, thus, provides better separation (Liu, 1994). Most studies, directly analyse FAMEs after derivatisation (Axelsson & Gentili, 2014; Liu, 1994; Mansour et al., 2005; Ramluckan et al., 2014; Widjaja et al., 2009). Although, this is an improvement on gravimetric analysis, it does not account for interfering compounds, that could generate peaks co-eluting with the FAMEs peaks (at the same retention time), resulting in an overestimation of that specific FAME. Thus, an accurate qualitative and quantitative determination cannot be ensured. Moreover, it does not provide information on which lipid class (that is, neutral lipids or polar lipids) from which the FAMEs had originated. To achieve an accurate quantification and characterisation of FAMEs, a preliminary fractionation of the respective lipid classes should be carried out prior to GC analysis. This separation can be performed by thin-layer chromatography (TLC), or solid-phase extraction (SPE) (Christie & Han, 2010). TLC is the older, cheaper and most widely adopted method for the separation of lipid classes. However, it cannot prevent oxidation of polyunsaturated fatty acids during a long exposure to oxygen from the air (Christie & Han, 2010). In addition, it is highly labour-intensive. Alternatively,

5

SPE is considered one of the most powerful techniques currently available for rapid and selective sample preparation (Christie & Han, 2010). It is effective for lipid separation because it concentrates TAGs (Christie & Han, 2010). Although this process of crude lipid extraction, transmethylation and purification with either TLC or SPE has been proven to be effective, it is time and energy consuming and therefore not realistic from an industrial-development standpoint. Thus, some researchers have eliminated extraction completely by a process called direct transesterification (DT). This method converts saponifiable lipids in situ directly to FAMEs using individual or a mixture of acidic and alkaline catalysts or other catalysts without the need of a crude lipid extraction procedure (Carrapiso & García, 2000). This results in a rapid one-step procedure, which has been reported to produce higher FAMEs yields compared to microalgae extracted using the Folch et al. (1957) and the Bligh and Dyer (1959) methods and their adaptations, followed by transesterification and GC (Cavonius et al., 2014; D’Oca et al., 2011; Griffiths et al., 2010). To date, no study has compared the effect of TLC and SPE purification with no purification or with DT for the quantification and the evaluation of the composition of FAME from microalgae. The aims of the present study are the evaluation of: a) the effect of the pretreatment such as boiling with isopropanol, freezing, oven-drying and freeze-drying microalgae biomass in the FAMEs composition and yield; b) the effect of the presence or absence of a purification method during FAMEs extraction and the effect of different purification techniques. Thus, in the present study, FAMEs from two green microalgae species—that is, Scenedesmus dimorphus and Coelastrella sp.—were extracted by either (1) using a crude lipid extraction (the single-step method) and transesterification, or (2) by adding a TLC and SPE separation prior to GC analysis. A DT method with a

6

combination of acidic and alkaline catalysts were also tested. After selection of the most suitable methods, a systematic comparison of the effect of the four pretreatments was carried out. Both comparative exercises provide insight into the best method to obtain FAMEs from microalgae and how to treat microalgae biomass prior to FAMEs extraction.

2. Materials and methods 2.1. Chemicals All chemicals used were of ACS grade unless otherwise stated. Petroleum ether, toluene, isopropanol, chloroform and sulphuric acid were purchased from VWR International AB (Stockholm, Sweden). Anhydrous methanol was purchased from Thermo Fisher Scientific (Hägersten, Sweden). Hexane (GC grade) and dried diethyl ether were purchased from Merck AB (Solna, Sweden). Heptane (GC grade), approximately 10% boron trifluoride-methanol solution (GC grade), 2, 2dimethoxypropane, 25% sodium methoxide solution in methanol, methanol (Highperformance liquid chromatography [HPLC] grade), formic acid and acetic acid were purchased from Sigma-Aldrich (Stockholm, Sweden). FAMEs standards were purchased from Larodan AB (Solna, Sweden).

2.2. Microalgae cultivation and harvest The green microalgae Scenedesmus dimorphus UTEX 417, was purchased from UTEX, The Culture Collection of Algae at the University of Texas, (Austin, Texas, USA), and Coelastrella sp. was isolated from local municipal wastewater (Umeå, Sweden) and genetically characterized at Umeå University (Umeå, Sweden). The

7

microalgae species were grown in 5 L bottles with municipal wastewater (Vakin, Umeå, Sweden) under the following conditions: light intensity photosynthetically active radiation (PAR) ≈ 100 µmol photon s-1 m-2, a 16:8 h light:dark cycle, room temperature (20 ± 2°C) and agitation was provided via a magnetic stirrer (200 rpm). Prior to usage, municipal wastewater was filtered and autoclaved. Batch cultures of S. dimorphus and Coelastrella sp. used for the comparison of extraction and purification methods were harvested after 13 and 6 days of culturing, respectively. The batch cultures of S. dimorphus and Coelastrella sp. used for pretreatment comparisons were cultured for 15 days. Microalgae cells were harvested by centrifugation at 3,250 g for 5 min, rinsed with tap water and re-centrifuged. Samples were harvested in triplicate; each pellet represented 100 mL of culture. 50 µL of a solution containing 0.2 mg mL-1 methyl pentadecanoic acid (C15:0) in methanol was added as an internal standard to each microalgae pellet. 2.3. Pretreatment In the first experiment (a comparison of extraction and purification methods), pellets derived from the same batch culture were immediately boiled at 80°C for 20 min with 2 mL isopropanol and exposed to gaseous nitrogen by flushing the glass tube headspace. After cooling, samples were stored at -20°C until further analysis. In the second experiment (a comparison of pretreatment methods), triplicates of each species derived from the same batch culture were either immediately frozen at -20°C, dried in an oven (Memmert, Schwabach, Germany) at 105°C for 3 h, lyophilised in a freeze dryer (Edwards high vacuum international, Crawley, England) at -45°C overnight or boiled with isopropanol, as previous described. After the different pretreatment conditions, samples were stored at -20°C until further analysis. Dry weight (that is, the

8

biomass content), after being dried in an oven at 105°C for 3h, was measured in triplicate in all cultures. 2.4. Fatty acid methyl esters extraction and purification In the first experiment (a comparison of extraction methods), triplicates of each species culture were either extracted with the single-step method, transmethylated and injected in the GC; extracted with single-step, transmethylated, purified with thin layer chromatography (TLC) and injected in the GC; extracted with single-step, purified with solid phase extraction (SPE), in which three types of SPE cartridges were tested, transmethylated and injected in the GC; or extracted with direct transesterification and injected in the GC. In the second experiment (a comparison of pretreatment methods), triplicates of each species culture and pretreatment were either extracted with the single-step method, purified with SPE, transmethylated and injected in the GC; or extracted with direct transesterification and injected in the GC.

2.4.1. Crude lipids extraction The total lipids were extracted in all samples of the first experiment (except three per species, which were used for direct transesterification) with a single-step method based on the Folch et al. (1957) method, previously developed (Axelsson & Gentili, 2014). In brief, a 2:1 chloroform:methanol (v/v) solution was added to the samples, vortexed for 2 min and 0.73% sodium chloride solution was added to achieve a 2:1:0.8 ratio of chloroform:methanol:water (v/v/v). Phase separation was achieved by centrifugation at 350 g for 2 min. Lipid phase was recovered and washed twice with

9

chloroform and further centrifuged. Subsequently, samples were vacuum dried in a multievaporator (Syncore® Polyvap, Büchi Labortechnik AB, Flawil, Switzerland) at 40°C, 120 rpm and 275 mbar for 3h. Lastly, total lipid content per dry weight was determined gravimetrically and samples were stored at -20°C until further use.

2.4.2. Transmethylation The crude lipid samples (also for further purification with TLC) and the lipid fractions obtained after SPE were dissolved in 1 mL of toluene and 2 mL of 1% sulphuric acid solution in dry methanol. The mixture was vortexed, fluxed with gaseous nitrogen and maintained at 80°C for 2 h in order to convert the lipid fraction to FAMEs (that is, biodiesel). 5 mL of 5% sodium chloride solution was then added and the FAMEs were extracted twice with 5 mL of hexane. The hexane layer was washed with 3 mL of 2% potassium bicarbonate solution and dried over anhydrous sodium sulphate (Christie & Han, 2010). The hexane layer was then recovered and vacuum dried in a multievaporator at 40°C, 120 rpm and 275 mbar overnight. FAMEs dry samples were stored at -20°C.

2.4.3. Thin layer chromatography TLC was used as a purification step to separate FAMEs from other lipid classes after transmethylation. Dry transmethylated samples were dissolved in heptane and placed in TLC silica gel plates (Merck AB, Solna, Sweden). The solvent system used for the elution was an 85:20:2 ratio of petroleum ether:diethyl ether:formic acid (v/v/v). After drying the solvents, the TLC plates were subjected to 2, 7 dichloroflurescein vapor, and the FAMEs spots were identified under ultraviolet light

10

(360 nm). The scraped-off spots were dissolved in methanol, petroleum ether and 0.73% sodium chloride solution. The mixture was vortexed and the petroleum ether phase was recovered by centrifugation at 350 g for 2 min, followed by a double petroleum ether washing and re-centrifugation (Christie & Han, 2010). Samples were vacuum dried overnight in a multievaporator at 40°C, 120 rpm and 275 mbar. Samples were subsequently stored at -20°C, from 12 to 72 h, until GC analysis.

2.4.4. Solid phase extraction SPE was used as a purification step to separate TAGs from other lipid classes prior to transmethylation. Three types of SPE cartridges (500 mg, 3 mL)— HyperSep™ Silica (Thermo Fisher Scientific, Hägersten, Sweden), SupelClean™ Ultra 2400 (Merck AB, Solna, Sweden) and Chromabond® NH₂ (VWR International AB, Stockholm, Sweden)—were tested, each with triplicate samples from the two species cultures of the first experiment. Triplicates of hexane blank samples were also processed to evaluate the potential of artefactual contamination during the SPE protocol. SPE separation was conducted as described by Danielewicz et al. (2011), with minor modifications. SPE cartridges were primed with hexane, and the crude lipid extract was dissolved in hexane before loading in the cartridge. An 80:20:1 mixture of hexane:diethyl ether:acetic acid (v/v/v) was used as the mobile phase for TAGs elution. The elution was vacuum dried overnight in a multievaporator at 40°C, 120 rpm and 275 mbar. The samples were subsequently stored at -20°C. In the second experiment, SPE purification followed the same protocol; however, only one type of cartridge (HyperSep™ Silica) was used.

11

2.4.5. Direct transesterification Direct transesterification was performed according to Griffiths et al. (2010), with minor changes. A combination of base followed by acid catalysis was performed as follows: toluene, 2, 2-dimethoxypropane and sodium methoxide were added to the triplicate samples of the microalgae biomass. The mixture was vigorously vortexed and incubated for 20 min at 80°C; samples were intermittently vortexed during incubation. After cooling at room temperature, boron trifluoride-methanol was added to the mixture and the incubation procedure was repeated. Samples were allowed to cool at room temperature before distilated water and hexane were added. The samples were vortexed and centrifuged at 3,250 g for 1 min. The recovered hexane–toluene layer (that is, the FAMEs extract) was vacuum dried overnight in a multievaporator at 40°C, 120 rpm and 275 mbar. Samples were subsequently stored at -20°C, from 12 to 72 h, until GC analysis.

2.5. Gas chromatography FAME extracts were re-suspended with heptane and injected into a TRACE™ 1310 (Thermo Fisher Scientific, Hägersten, Sweden) GC system equipped with a flame ionisation detector and a 30 m FAMEWAX column (Restek Corporation, Bellefonte, Pennsylvania, USA) with I.D. 0.32 mm and 0.25 µm film thickness. The injection volume was 1 μL, with a split ratio of 11 and split flow of 8 mL min -1. The carrier gas was nitrogen, with a fixed flow of 1.5 mL min-1. The initial temperature was 195°C, and it was increased in increments of 1.8°C min-1 until a temperature of 240°C was reached. The temperature was held at 240°C for 2.8 min. Total runtime was 29 min. A FAMEs mixture of methyl tetradecanoate (C14:0), methyl hexadecanoate (C16:0), methyl

12

palmitoleate (C16:1), methyl heptadecanoate (C17:0), methyl octadecanoate (C18:0), methyl oleate (C18:1), methyl linoleate (C18.2), methyl linolenate (C18:3), methyl eicosanoate (C20:0) and methyl docosanoate (C22:0) was used for identification of the different peaks in microalgae samples. FAMEs were identified by comparison with external FAMEs standards and quantified by the use of the internal standard methyl pentadecanoic acid (C15:0).

2.6. Statistics To investigate the statistical differences between the FAMEs content means of different methods and pretreatments, a one-way analysis of variance (ANOVA) followed by a post-hoc Student's t-test with Bonferroni correction was applied. Analyses were performed with Microsoft Office Excel 2013 Analysis ToolPak.

3. Results and Discussion 3.1. Experiment one – The effect of purification methods The two green microalgae species used—that is, S. dimorphus and Coelastrella sp.—had biomass concentrations of 12.23 ± 0.79 mg dry weight (DW) and 21.77 ± 1.40 mg DW per 100 mL of culture, respectively. The total lipid concentrations extracted with the single-step method (Axelsson & Gentili, 2014) based on Folch et al. (1957) were 36.60% ± 9.93% DW for S. dimorphus and 9.45% ± 3.48% DW for the Coelastrella local isolate. The differences in biomass concentrations and total lipid content can be attributed to species-specific characteristics and the duration of the algae cultivation period because culture conditions were the same for both species.

13

The same strain of S. dimorphus extracted with the same method and grown for four days in a different batch of the same municipal wastewater (Vakin, Umeå) used in the present study, with the addition of flue gases (10% CO2), had a lipid content of about 12% (Axelsson & Gentili, 2014). Again, the same strain grown for 22 days on 50% autoclaved wastewater from a lagoon pond at the Kansas State University animal farm had 38.88% lipid content (Shen & Yuan, 2012). Previous lipid productivities of the Coelastrella local isolate have not been reported. However, another Coelastrella strain grown on 20% aerobically treated swine wastewater for 10 days had a total lipid content of 24.8% (Luo et al., 2016). Several Asian Coelastrella isolates cultured for nine days in BBM media had lipid contents between 13.15% and 36.52% (Minhas et al., 2016). The FAMEs profile of S. dimorphus samples purified with TLC was similar to the samples purified with any of the three types of SPE cartridges (Fig.1 A). This finding suggests that both types of purification have an equivalent performance, and that the selection of either one will be dependent only on the time, solvents and the energy consumption of the method. Although, hexane leaching artifacts with retention times correspondent to methyl hexadecanoate (C16:0) and methyl octadecanoate (C18:0) were detected in blank samples purified with SPE, it had a negligible effect on the FAMEs profile and yield determination (data not shown). Methyl hexadecanoate (C16:0) and methyl oleate (C18:1) are the most abundant FAMEs, constituting 32.14% ± 3.56% and 39.09% ± 6.50% of the total FAME in the SPE 1 profile, respectively. A dominance of this two types of FAMEs samples from the same strain extracted with a modified Folch et al. (1957) method have previously been reported, each with a total FAMEs content of about 35% (Svensson, 2014). Another strain of S. dimorphus extracted with a

14

methanol:acetyl chloride (95:5 v/v) solution instead of the chloroform:methanol (2:1 v/v) solution suggested by Folch et al. (1957) resulted in a dominance of methyl linolenate (C18:3; 20.7% of the total FAMEs content) and methyl hexadecanoate (C16:0; 13.1% of the total FAMEs content) (Islam et al., 2013). Several other strains of Scenedesmus extracted with a modified Folch method show a similar FAMEs profile to the one described here (that is, a dominance of methyl hexadecanoate [C16:0] and methyl oleate [C18:1]) (Minhas et al., 2016). Interestingly, when extracted with direct transesterification a decrease in the content of methyl hexadecanoate (C16:0) and an increase in methyl linolenate (C18:3) was observed (Fig.1 A), suggesting that different extraction systems and solvent mixtures may generate different FAMEs profiles; this finding makes it difficult to determine the best species for biodiesel production by examining results of samples because the samples have been extracted with different protocols. When extracted with the single-step and transmethylated without any purification treatment, the S. dimorphus FAMEs composition was distinct compared to when TLC or SPE purification was performed (Fig. 1 A). The differences can be attributed to the presence of interfering compounds, considering that several small additional peaks were observed in the chromatograms of samples that have not been purified and that were not observed on the purified samples (data not shown). Thus, it can be concluded, that the isolation of neutral lipids by TLC and SPE purification methods and the subsequent removal of interfering compounds, such as pigments, resulted in an overall improvement in the FAMEs determination. Coelastrella sp. had a similar FAMEs profile to S. dimorphus (Fig.1 A and B), which is not unusual because the most common FAME of microalgae are methyl hexadecanoate (C16:0), methyl octadecanoate (C18:0), methyl oleate (C18:1), methyl

15

linoleate (C18.2) and methyl linolenate (C18:3) (Huerlimann et al., 2010). The percentages of each FAMEs were similar between samples purified with TLC and SPE. Also in Coelastrella, the levels of methyl linolenate (C18:3) were higher in the samples extracted with DT. In comparison to S. dimorphus, Coelastrella sp. had slightly higher levels of methyl octadecanoate (C18:0). Several Asian strains of Coelastrella were reported to have mainly methyl hexadecanoate (C16:0), methyl octadecanoate (C18:0), methyl linoleate (C18.2) and methyl linolenate (C18:3). However the dominant FAMEs species, seemed to be strain specific (Minhas et al., 2016). Luo et al. (2016) observed that an increase in the percentage of aerobically treated swine wastewater in which the Coelastrella sp. was cultivated (that is, from 20%, 40%, 60%, 80% to 100%) resulted in a gradual increase in methyl linolenate (C18:3) and a decrease in methyl octadecanoate (C18:0) and methyl hexadecanoate (C16:0). In several microalgae species, the lipid content and FAME composition have previously been demonstrated to be dependent on culture conditions, growth period and environmental situations (Li et al., 2011; Zhu et al., 2013). The total FAMEs concentration extracted in each species was analogous between the different methods (Fig. 1 C and D). The one exception was the use of DT with the S. dimorphus samples, which extracted significantly less FAMEs compared with the other methods. Accordingly, Soares et al. (2014) determined that another DT method produced much lower FAMEs yields—from the microalgae Nannochlorophisis oculata, Chaetoceros muelleri and Chlorella sp.—compared with most of the traditional two-step methods of crude lipid extraction followed by transmethylation. However, the FAMEs yields of Coelastrella sp. extracted using the various methods were not significantly different. Taking into account that the total lipid concentration was lower

16

in Coelastrella sp. compared with S. dimorphus, it can be suggested that the DT method used in the present study has a better performance with samples that contain lower total lipid concentrations. It can also be theorized that a successful FAMEs extraction with DT is dependent on microalgae cell wall thickness. Although the FAMEs yields of the samples extracted with the single-step method with no purification, only transmethylated (that is, T) were not significantly different from the ones purified with TLC and SPE, the standard deviations were much higher, which suggests there were errors in the quantification of FAMEs, perhaps due to the formation of overlapping peaks from interfering compounds (data not shown). The total yields of FAMEs that are derived from microalgae are dependent on several factors, including species, culture conditions, environmental factors and extraction methods. Under conditions of nutrient stress, the total lipid content may increase, and more lipids may be stored in the form of TAGs; this may also change the fatty acids composition (Gouveia & Oliveira, 2009; Rodolfi et al., 2009). Moreover, total lipids, fatty-acid composition and TAGs formation of microalgae have been reported to vary with light and temperature (Hu et al., 2008a; Renaud et al., 1991; Sukenik et al., 1993). FAMEs yields of between 0.6% and 24% DW have been reported (Orr et al., 2016). The FAMEs yields obtained in the current study are within the reported values, although optimisation of culture conditions for FAMEs production was not the aim of the present study. According to the results obtained, the single-step method with SPE purification followed by transmethylation is the best method for FAMEs quantification and characterisation from microalgae among the methods tested because of its FAMEs extraction yields and composition, its reproducibility, and because it is less labour-

17

intensive than TLC. Moreover, its performance is independent of the type of SPE cartridge used. For these reasons, this method was selected to carry out the second experiment to examine the relationship between pretreatment and FAMEs composition and content. Earlier studies have suggested that for quantifying total FAMEs in microalgae, DT methods are more accurate than extraction methods that are based on Folch et al. (1957) and Bligh and Dyer (1959) followed by transmethylation and no purification (Cavonius et al., 2014; D’Oca et al., 2011; Griffiths et al., 2010). DT methods are also less costly and labour intensive. Thus, the performance of DT in microalgae samples was evaluated with different pretreatments (that is, boiled in isopropanol, frozen, oven-dried and freeze-dried).

3.2. Experiment two – The effect of pretreatment S. dimorphus and Coelastrella sp. cultures used in the second experiment had biomass concentrations of 14.47 ± 3.03 and 15.63 ± 0.76 mg DW per 100 mL of culture, respectively. No significant differences in total lipid content were observed for either species for samples that were extracted with the single-step method and that were pretreated by being boiled in isopropanol, oven-dried or freeze-dried (Fig. 2). This finding agrees with the findings of most studies in the published literature (Babarro et al., 2001; Esquivel et al., 1993; Morist et al., 2001; Ryckebosch et al., 2011; Ryckebosch et al., 2012). For instance, no significant difference was detected in the total extracted lipids of Chlorella vulgaris after a freeze-dried or a combination of freeze-dried and boiled in isopropanol pretreatment (Ryckebosch et al., 2012). However, the frozen samples had much higher total lipid content; this is probably a result of the freeze–thaw process, which disrupts microalgae cells due to ice–crystal

18

formation and cell expansion upon thawing and improves crude lipid extraction in comparison to the other pretreatments. However, because the total lipid content was gravimetrically determined, the possibility that the high lipid content in the frozen samples corresponds to an increase in the extraction of proteins, carbohydrates, pigments and other compounds cannot be disregarded. The FAMEs composition of both species was similar after being boiled in isopropanol, oven-dried or freeze-dried, followed by being extracted with the singlestep method, undergoing transmethylation and undergoing SPE purification (Fig. 3 A and B). Accordingly, Ryckebosch et al. (2012) found no effect of freeze-drying on the total lipids, composition of lipid classes and FAMEs profile of P. tricornutum when compared with fresh samples. The percentage of each individual FAMEs was also similar to the observations made from experiment one, where samples were boiled in isopropanol prior to extraction (Fig. 1 A and B). However, the samples that were frozen had a distinct profile. Although in Coelastrella sp. the differences between the pretreatments were minimal, in S. dimorphus, methyl hexadecanoate (C16:0) decreased from 40.44% ± 4.88% of the total FAMEs in the boiled samples to 24.70% ± 0.60% of the total FAMEs in the frozen samples, and methyl linolenate (C18:3) increased from 8.20% ± 0.97% of the total FAMEs in the boiled samples to 36.74% ± 3.33% of the total FAMEs in the frozen samples (Fig. 3A). In contrast, the total FAMEs content was not significantly different between the samples boiled in isopropanol in both species and the samples that were oven-dried for S. dimorphus (Fig. 3C).The freeze-dried samples of both species yielded significantly less FAMEs compared to the other pretreatments. In addition, the total FAMEs content of the oven-dried Coelastrella sp. samples was significantly lower compared to the samples that were frozen (Fig. 3D).

19

Balasubramanian et al. (2013) showed no significant differences to neutral lipids content between freeze-dried, oven-dried and solar-dried microalgae samples extracted with the Folch et al. (1957) method and purified with SPE. However, the type of drying method had a significant effect on the amount of free fatty acids (FFAs); freeze-dried microalgae had the lowest amount of FFAs, followed by oven-dried samples (Balasubramanian et al., 2013; Guldhe et al., 2014). Although the SPE purification method used removes polar lipids, a full removal of FFAs from the samples cannot be ensured. Thus, the lower total FAMEs yield in the freeze-dried samples could be attributed to a lower TAGs extractability, but it is more likely the result of lower concentrations of FFAs in the samples. Furthermore, the similar FAMEs yields from TLC and SPE purifications in experiment one further suggests that the SPE method is extracting FFAs and TAGs together. This is because the TLC method that was used isolates FAMEs, but it is not capable of discerning the FAMEs produced from FFAs in the sample from FAMEs produced from TAGs. When extracted with DT, S. dimorphus and Coelastrella sp. had slight variations on the percent of individual FAMEs among the pretreatments (Fig. 4 A and B). For instance, the oven- and freeze-dried samples of S. dimorphus had more methyl hexadecanoate (C16:0) and methyl oleate (C18:1), but less methyl linolenate (C18:3) compared to when the wet samples were boiled with isopropanol and frozen (Fig. 4A). Both microalgae species extracted with DT had a very distinct FAMEs profiles in comparison to the samples extracted with the single-step method (compare Figs. 3 A and B and 4 A and B). Methyl linolenate (C18:3) was the dominant FAME, varying from 47.97% and 86.90% total FAMEs, with the lowest value in Coelastrella sp. and the highest of S. dimorphus, respectively. All other FAMEs species made up less than

20

25% of the total FAMEs. This result is in agreement with experiment one. Fig. 4 C and D shows no significant difference of FAMEs extractability with DT among the various pretreatments in both species. However, the yield from the oven-dried S. dimorphus sample was four times higher than the boiled and frozen S. dimorphus samples (Fig. 4C). As observed in experiment one with the S. dimorphus, in experiment two the total FAMEs yield of S. dimorphus and Coelastrella sp. were lower in the samples extracted with DT compared to the single-step method, transmetylation and SPE purification. The same hypothesis suggested for the lower FAMEs yields in freeze-dried samples extracted with the single-step method and SPE can be attributed to the samples extracted with DT. In fact, Ryckebosch et al. (2011) registered a significant decrease in the concentrations of FFAs in freeze- and spray-dried Phaeodactylum tricornutum in comparison with fresh samples that were extracted with a DT method. It is known that the presence of FFAs could decrease the efficiency of DT when using alkaline catalysts due to the saponification effect (that is, a high FFAs content will promote soap formation) and the separation of products will be exceedingly difficult. As a result, the yield of biodiesel product will be low (Crabbe et al., 2001; Goodrum, 2002). However, in the current study, a mix of acid- and alkaline-catalysed esterification was used, which is recommended to avoid saponification (Griffiths et al., 2010). Moreover, water is known to interfere with the transesterification reaction (Carrapiso & García, 2000; Liu, 1994). Increasing water content has been suggested to progressively decrease the yield of FAMEs from microalgae extracted with DT (Cao et al., 2013; Wahlen et al., 2011). This might explain the lower FAMEs yields in S. dimorphus samples boiled in isopropanol and frozen. However, the DT method used— which includes the addition of the water scavenger 2, 2-dimethoxypropane with alkaline

21

and acid catalysts—should leave the reaction unaffected by the water content (Griffiths et al., 2010). A drawback of using the water scavenger in all samples, including ovenand freeze-dried samples, in order to improve reproducibility is the production of extra peaks in the chromatogram and, as a result, a change in the efficiency of the protocol.

4. Conclusions TLC and SPE purifications, coupled with single-step crude lipids extraction and transmethylation, produced accurate, precise and reproducible results of FAMEs yields and composition. SPE is preferable due to lower labour intensity. The removal of a purification step strongly affected the method reproducibility. DT had a distinct FAMEs profile in comparison with the other methods. Thus, it should not be used to determine microalgae species suitability for biodiesel production. Overall, pretreatment did not have a strong impact in FAMEs characterisation, suggesting that extreme dewatering techniques, such as oven- and freeze-drying, may not be required. Freeze-dried microalgae produce the lowest FAMEs yields. 5. Acknowledgements This work was supported by funding from the Swedish Energy Agency (project nr. 38239-1), the EU Interreg Botnia-Atlantica (TransAlgae project) and the Kempe Foundation. The authors gratefully acknowledge Prof. Anita Sellstedt and Jean Claude Nzayisenga (Department of Plant Physiology, Umeå University) for technical assistance in GC analysis, and Prof. Christiane Funk and Lorenza Ferro (Department of Chemistry, Umeå University) for providing an axenic sample of the algal strain Coelastrella.

22

6. References Axelsson, M., Gentili, F. 2014. A single-step method for rapid extraction of total lipids from green microalgae. PLoS One, 9(2), e89643. Babarro, J., Reiriz, M., Labarta, U. 2001. Influence of preservation techniques and freezing storage time on biochemical composition and spectrum of fatty acids of Isochrysis galbana clone T‐ ISO. Aquaculture Research, 32(7), 565-572. Balasubramanian, R.K., Doan, T.T.Y., Obbard, J.P. 2013. Factors affecting cellular lipid extraction from marine microalgae. Chemical Engineering Journal, 215, 929-936. Berge, J.-P., Gouygou, J.-P., Dubacq, J.-P., Durand, P. 1995. Reassessment of lipid composition of the diatom, Skeletonema costatum. Phytochemistry, 39(5), 1017-1021. Bligh, E.G., Dyer, W.J. 1959. A rapid method of total lipid extraction and purification. Canadian journal of biochemistry and physiology, 37(8), 911-917. Cao, H., Zhang, Z., Wu, X., Miao, X. 2013. Direct biodiesel production from wet microalgae biomass of Chlorella pyrenoidosa through in situ transesterification. BioMed research international, 2013. Carrapiso, A.I., García, C. 2000. Development in lipid analysis: some new extraction techniques and in situ transesterification. Lipids, 35(11), 1167-1177. Cavonius, L.R., Carlsson, N.-G., Undeland, I. 2014. Quantification of total fatty acids in microalgae: comparison of extraction and transesterification methods. Analytical and bioanalytical chemistry, 406(28), 7313-7322. Christie, W., Han, X. 2010. Lipid Analysis-Isolation, Separation, Identification and Lipidomic Analysis , 446 pages, Oily Press, Bridgwater, UK. Converti, A., Casazza, A.A., Ortiz, E.Y., Perego, P., Del Borghi, M. 2009. Effect of temperature and nitrogen concentration on the growth and lipid content of Nannochloropsis oculata and Chlorella vulgaris for biodiesel production. Chemical Engineering and Processing: Process Intensification, 48(6), 1146-1151. Crabbe, E., Nolasco-Hipolito, C., Kobayashi, G., Sonomoto, K., Ishizaki, A. 2001. Biodiesel production from crude palm oil and evaluation of butanol extraction and fuel properties. Process biochemistry, 37(1), 65-71. D’Oca, M.G.M., Viêgas, C.V., Lemoes, J.S., Miyasaki, E.K., Morón-Villarreyes, J.A., Primel, E.G., Abreu, P.C. 2011. Production of FAMEs from several microalgal lipidic extracts and direct transesterification of the Chlorella pyrenoidosa. biomass and bioenergy, 35(4), 1533-1538. Danielewicz, M.A., Anderson, L.A., Franz, A.K. 2011. Triacylglycerol profiling of marine microalgae by mass spectrometry. Journal of lipid research, 52(11), 2101-2108. Esquivel, B.C., Lobina, D.V., Sandoval, F.C. 1993. The biochemical composition of two diatoms after different preservation techniques. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry, 105(2), 369-373. Folch, J., Lees, M., Sloane Stanley, G. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J biol Chem, 226(1), 497-509. Goodrum, J. 2002. Volatility and boiling points of biodiesel from vegetable oils and tallow. Biomass and Bioenergy, 22(3), 205-211. Gouveia, L., Oliveira, A.C. 2009. Microalgae as a raw material for biofuels production. Journal of industrial microbiology & biotechnology, 36(2), 269-274. Griffiths, M., Van Hille, R., Harrison, S. 2010. Selection of direct transesterification as the preferred method for assay of fatty acid content of microalgae. Lipids, 45(11), 10531060. Guldhe, A., Singh, B., Rawat, I., Ramluckan, K., Bux, F. 2014. Efficacy of drying and cell disruption techniques on lipid recovery from microalgae for biodiesel production. Fuel, 128, 46-52.

23

Halim, R., Danquah, M.K., Webley, P.A. 2012. Extraction of oil from microalgae for biodiesel production: a review. Biotechnology advances, 30(3), 709-732. Han, Y., Wen, Q., Chen, Z., Li, P. 2011. Review of methods used for microalgal lipid-content analysis. Energy Procedia, 12, 944-950. Hu, C., Li, M., Li, J., Zhu, Q., Liu, Z. 2008a. Variation of lipid and fatty acid compositions of the marine microalga Pavlova viridis (Prymnesiophyceae) under laboratory and outdoor culture conditions. World Journal of Microbiology and Biotechnology, 24(7), 12091214. Hu, Q., Sommerfeld, M., Jarvis, E., Ghirardi, M., Posewitz, M., Seibert, M., Darzins, A. 2008b. Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances. The plant journal, 54(4), 621-639. Huerlimann, R., De Nys, R., Heimann, K. 2010. Growth, lipid content, productivity, and fatty acid composition of tropical microalgae for scale‐ up production. Biotechnology and bioengineering, 107(2), 245-257. Islam, M., Magnusson, M., Brown, R., Ayoko, G., Nabi, M., Heimann, K. 2013. Microalgal Species Selection for Biodiesel Production Based on Fuel Properties Derived from Fatty Acid Profiles. Energies, 6(11), 5676. Iverson, S.J., Lang, S.L., Cooper, M.H. 2001. Comparison of the Bligh and Dyer and Folch methods for total lipid determination in a broad range of marine tissue. Lipids, 36(11), 1283-1287. Kates, M. 1972. Techniques of Lipidology: Isolation, Analysis, and Identification of Lipids American Elsevier Publishing Company. Inc., New York, NY. Lee, C.M., Trevino, B., Chaiyawat, M. 1996. A simple and rapid solvent extraction method for determining total lipids in fish tissue. Journal of AOAC International, 79(2), 487-492. Lee, J.-Y., Yoo, C., Jun, S.-Y., Ahn, C.-Y., Oh, H.-M. 2010. Comparison of several methods for effective lipid extraction from microalgae. Bioresource technology, 101(1), S75-S77. Li, Y., Chen, Y.-F., Chen, P., Min, M., Zhou, W., Martinez, B., Zhu, J., Ruan, R. 2011. Characterization of a microalga Chlorella sp. well adapted to highly concentrated municipal wastewater for nutrient removal and biodiesel production. Bioresource Technology, 102(8), 5138-5144. Liu, K.-S. 1994. Preparation of fatty acid methyl esters for gas-chromatographic analysis of lipids in biological materials. Journal of the American Oil Chemists' Society, 71(11), 1179-1187. Luo, L., He, H., Yang, C., Wen, S., Zeng, G., Wu, M., Zhou, Z., Lou, W. 2016. Nutrient removal and lipid production by Coelastrella sp. in anaerobically and aerobically treated swine wastewater. Bioresource technology, 216, 135-141. Mansour, M.P., Frampton, D.M., Nichols, P.D., Volkman, J.K., Blackburn, S.I. 2005. Lipid and fatty acid yield of nine stationary-phase microalgae: applications and unusual C 24–C 28 polyunsaturated fatty acids. Journal of Applied Phycology, 17(4), 287-300. Minhas, A.K., Hodgson, P., Barrow, C.J., Sashidhar, B., Adholeya, A. 2016. The isolation and identification of new microalgal strains producing oil and carotenoid simultaneously with biofuel potential. Bioresource technology, 211, 556-565. Morist, A., Montesinos, J., Cusido, J., Godia, F. 2001. Recovery and treatment of Spirulina platensis cells cultured in a continuous photobioreactor to be used as food. Process Biochemistry, 37(5), 535-547. Orr, V.C., Plechkova, N.V., Seddon, K.R., Rehmann, L. 2016. Disruption and wet extraction of the microalgae Chlorella vulgaris using room-temperature ionic liquids. ACS Sustainable Chemistry & Engineering, 4(2), 591-600. Palmquist, D., Jenkins, T. 2003. Challenges with fats and fatty acid methods. Journal of animal science, 81(12), 3250-3254. Pruvost, J., Van Vooren, G., Cogne, G., Legrand, J. 2009. Investigation of biomass and lipids production with Neochloris oleoabundans in photobioreactor. Bioresource technology, 100(23), 5988-5995.

24

Pulz, O., Gross, W. 2004. Valuable products from biotechnology of microalgae. Appl Microbiol Biotechnol, 65(6), 635-48. Radwan, S. 1984. Glycerophospholipids. Handbook of chromatography. Lipids, 1, 481-508. Ramluckan, K., Moodley, K.G., Bux, F. 2014. An evaluation of the efficacy of using selected solvents for the extraction of lipids from algal biomass by the soxhlet extraction method. Fuel, 116, 103-108. Ranjan, A., Patil, C., Moholkar, V.S. 2010. Mechanistic assessment of microalgal lipid extraction. Industrial & Engineering Chemistry Research, 49(6), 2979-2985. Raven, P.H., Evert, R.F., Eichhorn, S.E. 2005. Biology of plants. Macmillan. Renaud, S., Parry, D., Thinh, L.-V., Kuo, C., Padovan, A., Sammy, N. 1991. Effect of light intensity on the proximate biochemical and fatty acid composition of Isochrysis sp. and Nannochloropsis oculata for use in tropical aquaculture. Journal of Applied Phycology, 3(1), 43-53. Rodolfi, L., Chini Zittelli, G., Bassi, N., Padovani, G., Biondi, N., Bonini, G., Tredici, M.R. 2009. Microalgae for oil: Strain selection, induction of lipid synthesis and outdoor mass cultivation in a low‐ cost photobioreactor. Biotechnology and bioengineering, 102(1), 100-112. Ryckebosch, E., Muylaert, K., Eeckhout, M., Ruyssen, T., Foubert, I. 2011. Influence of drying and storage on lipid and carotenoid stability of the microalga Phaeodactylum tricornutum. Journal of agricultural and food chemistry, 59(20), 11063-11069. Ryckebosch, E., Muylaert, K., Foubert, I. 2012. Optimization of an Analytical Procedure for Extraction of Lipids from Microalgae. Journal of the American Oil Chemists' Society, 89(2), 189-198. Shen, Y., Yuan, W.Q. 2012. Comparison of biomass and oil productivity of selected microalgae in livestock wastewater. Advanced Materials Research. Trans Tech Publ. pp. 655-658. Soares, A.T., Da Costa, D.C., Silva, B.F., Lopes, R.G., Derner, R.B., Antoniosi Filho, N.R. 2014. Comparative analysis of the fatty acid composition of microalgae obtained by different oil extraction methods and direct biomass transesterification. BioEnergy Research, 7(3), 1035-1044. Sukenik, A., Yamaguchi, Y., Livne, A. 1993. Alterations in lipid molecular species of the marine eustigma tophyte Nannochlorosis sp. Journal of phycology, 29(5), 620-626. Svensson, M. 2014. Improved utilization of waste resources from the pulp and paper mill Nordic Paper Bäckhammar and isolation of local freshwater microalgae species. Gothenburg: Chalmers University of Technology. Wahlen, B.D., Willis, R.M., Seefeldt, L.C. 2011. Biodiesel production by simultaneous extraction and conversion of total lipids from microalgae, cyanobacteria, and wild mixed-cultures. Bioresource technology, 102(3), 2724-2730. Widjaja, A., Chien, C.-C., Ju, Y.-H. 2009. Study of increasing lipid production from fresh water microalgae Chlorella vulgaris. Journal of the Taiwan Institute of Chemical Engineers, 40(1), 13-20. Zhu, L., Wang, Z., Shu, Q., Takala, J., Hiltunen, E., Feng, P., Yuan, Z. 2013. Nutrient removal and biodiesel production by integration of freshwater algae cultivation with piggery wastewater treatment. Water Research, 47(13), 4294-4302.

25

Figure Captions Fig. 1 The fatty acid methyl esters (FAMEs) composition (percent of total FAMEs) and total FAMEs content (percent of dry weight) of (A. and C.) Scenedesmus dimorphus and (B. and D.) Coelastrella sp., respectively, as determined by four different methods: extraction with a single-step procedure, transmethylation and purified with thin layer chromatography (TLC); solid phase extraction (SPE) with three types of SPE cartridges (SPE 1, SPE 2 and SPE 3) or not purified (T); and extraction with direct transesterification (DT). Error bars show the standard deviation of the mean (n = 3). The letters above the bars of the same microalgae indicate a significant difference (p < 0.05).

Fig. 2 Total lipid gravimetric yields (percent of dry weight) of Scenedesmus dimorphus and Coelastrella sp. wet samples boiled with isopropanol and frozen, ovendried and freeze-dried samples extracted with a single-step procedure. Error bars show the standard deviation of the mean (n = 3). The letters above the bars of the same microalgae indicate a significant difference (p < 0.05).

Fig. 3 Fatty acid methyl esters (FAMEs) composition (percent of total FAMEs) and FAMEs content (percent of dry weight) of (A. and C.) Scenedesmus dimorphus and (B. and D.) Coelastrella sp. wet samples boiled with isopropanol and frozen, oven-dried and freeze-dried samples extracted with a single-step procedure, transmethylation and purification with SPE. Error bars show the standard deviation of the mean (n = 3). The letters above bars of the same microalgae indicate a significant difference (p < 0.05).

26

Fig. 4 Fatty acid methyl esters (FAMEs) composition (percent of total FAMEs) and total FAMEs content (percent of dry weight) of (A. and C.) Scenedesmus dimorphus and (B. and D.) Coelastrella sp. wet samples boiled with isopropanol and frozen, oven-dried and freeze-dried samples extracted via direct transesterification. Error bars show the standard deviation of the mean (n = 3). The letters above the bars of the same microalgae indicate a significant difference (p < 0.05).

27

Figure 1

28

Figure 2

29

Figure 3

30

Figure 4

Highlights    

Fatty acid methyl esters (FAME) composition was microalgae species specific Thin-layer chromatography and solid-phase extraction improved reproducibility Direct transesterification produced different FAME profiles than single-step method Pretreatment did not have a strong impact in FAME characterisation

31

32