First Report on Tertiary Amine as a Co2 Switchable Solvent for Hypersaline Trebouxiophycean Microalgae Towards Greener and Competent Lipid Extraction

Considering the momentous cost drivers in energy ecient algal biorenery processes, a green alternative in the lipid extraction process from microalgae is anticipated. Switchable solvent system using tertiary amines namely DMBA (Dimethylbenzylamine), DMCHA (Dimethylcyclohexylamine), and DIPEA (Diisopropylethylamine) for lipid extraction from wet hypersaline microalgae was investigated in this study. Interestingly, showed that at 1:1 (v/v of fresh DMBA solvent: microalgal biomass), and for 1 h extraction time, the lipid yield was 41.9, 26.6, and 33.3% for Chlorella sp. NITT 05, Chlorella sp. NITT 02, and Picochlorum sp. NITT 04 respectively and for recovered DMBA solvent at 1:1 (v/v) and for 1 hour extraction time, the lipid yield was 40.8, 25.97, and 32%, respectively. Similarly, lipid extraction using DMCHA solvent for Chlorella sp. NITT 05, Chlorella sp. NITT 02, and Picochlorum sp. NITT 04 at 1:1 (v/v of solvent: microalgal biomass) and 1 h extraction time showed 34.28, 24.24 and 23.33% lipids, respectively for fresh solvent and 34.01, 24.24 and 23.18% for recovered solvent respectively; while DIPEA was not competent in lipid extraction from three tested microalgae. FAME prole shows the presence of major saturated fatty acid as C16:0 (~ 30%) and major unsaturated fatty acid as C18:1 (~17%).


Introduction
In the pursuit of gratifying the mainstream of energy mandate, fossil fuel takes the step front always.
However, uncertainties in fossil fuel availability and its augmented greenhouse gases emission exemplify two major disputes viz., steady escalation in fuel price and climatic hitches like global warming 1,2 . Petroleum or fossil fuels are considered as a depleting energy reserve against growing demand due to their non-renewable feature, and unarguably they pose a potential threat to the transportation sector 3 . In this scenario, biodiesel is clearly emerging as an alternative supplement, which can be readily, introduced into the existing transportation infrastructure without engine modi cation 4 . Though, many feedstocks are being used for biodiesel, microalgae hold great promise as a hopeful production platform for biodiesel 5 and carbon dioxide sequestration due their traits like less land requirement, high biomass and lipid productivity 2,6 , accumulation of non-polar glycerolipids (triacylglycerol) 7 , versatility to grow in fresh and seawater. However, microalgal biodiesel production still needs to throw light on resilient strain selection, pertinent lipid extraction method, and inexpensive harvesting technique. The key challenge that needs to be addressed with an emphasis on commercially viable microalgal biodiesel production is to ascertain an opposite, eco-friendly lipid extraction technique 8 . Till date, the supreme barrier for the development of green chemical-based extraction is that speci c solvents are required to solvate speci c lipid component type since each solvent has its own solubility. Reuse of solvent recovered through certain thermal processes (rotary evaporation or distillation) is also essential to obviate energy e ciency, but by doing so, would eventually escalate the energy and cost. Therefore, energy e cient and ecofriendly lipid extraction method is need of the hour as the present toxic solvent system with unproductive extraction e ciency cannot be an ideal choice. In a perspective of ecofriendly and reusability attributes, CO 2 switchable solvent systems for lipid extraction would indeed be a great choice. CO 2 switchable solvent system includes amine which can be primary, secondary, and tertiary amines. Among these, primary and secondary amines were explored for algal lipid extraction in the recent years. In case of tertiary amines, DMCHA has been used for few microalgal lipid extractions 9,10 . To best of our knowledge, this is the rst study on using tertiary amine as a CO 2 switchable solvent for hypersaline microalgal strains as a greener and safe lipid extraction processes. Hence, in the present study tertiary amines such as DMBA, and DIPEA are used for lipid extraction from marine microalgal species with a comparison of DMCHA solvent. The aim of this study is to investigate the use of tertiary amines as a safer alternative for wet extraction of Chlorella sp. NITT 02, Chlorella sp. NITT 05, and Picochlorum sp. NITT 04. Lipid extracted from fresh solvent, the percentage of solvent recovery and the lipid extracted from the recycled solvents are comparatively studied. Further, fatty acid composition of lipids extracted using SPS system was analysed.  1). Of the strains tested, Picochlorum sp. NITT 04 was found to show higher cell density at about 2.23 OD on 27th day (Fig. 1c). The growth of Picochlorum sp. NITT 04 was gradually increased from day 0 till day 27; however after 27th day of its growth, no remarkable trend in growth rise was noticed. Similar to the results reported from this study, the OD value of Picochlorum oklahomensis after 25th day was around 2 and subsequently decreased over time 11 . In the case of Chlorella strains, the initial OD of the Chlorella sp. NITT 02 and NITT 05 culture was 0.61 on day 0 and the OD got increased during the growth and the high culture density (maximal OD) was seen on cultivation day 27 which is 1.45 and 1.62 OD, respectively (Fig. 1a, 1b) [13][14][15][16] . The cell density obtained in this study was higher since the culture was grown in ambient condition without CO 2 purging. It could rise further by cultivating it in high strength medium purged with CO 2 .

Lipid extraction by Switchable solvent system
In the present study, growth metrics of Picochlorum sp. NITT 04, Chlorella sp. NITT 02, and Chlorella sp. NITT 05 disclosed that higher biomass was produced on 27th of cultivation, but, no information on lipid content and yield has been retrived, which is considered to be a important parameter for biodiesel production. Therefore, it is imperative to study lipid content of the strain to evaluate its feasibility for biodiesel and the lipid extraction was done using the biomass harvested on 27th day. Switchability nature of tertiary amine during lipid extraction from marine microalgae is illustrated in Fig. 3. The amount of lipid extracted by the solvent (fresh tertiary amine) indicated as "fresh" and the amount of lipid extracted by the solvent after rst use, indicated as "recovered" in Fig. 4a were calculated gravimetrically. It is apparent that DMBA solvent was able to extract more lipids than DMCHA and DIPEA solvents. Also, Chlorella sp. NITT05 yielded signi cantly more lipids followed by the other two species. It is to be noted that, higher lipid at about 42% was extracted by both fresh and recovered solvent of DMBA from Chlorella sp. NITT05 whereas it was 23% using DMCHA. (1) In case of tertiary amines, bicarbonate salts which are soluble in water are formed when exposed to CO 2 .
The switching of tertiary amine is illustrated by the following reaction ,Eq. (1) 9 . In this regard, it is noteworthy to mention that, DMBA is determined to be an e cient and competent tertiary amine for microalgal lipid extraction in terms of extraction performance, and easy switching (energy e cient processes like heating and stirring not required).

Solvent recovery and reuse
The tertiary amines DMBA, and DMCHA recovered after rst extraction of lipid was subsequently used for extraction of lipids from the wet microalgae for the second cycle. The solvent recovery is presented in Fig. 4b and it is inferred that the overall solvent recovery for both DMBA and DMCHA is less than 30% and varied slightly between the microalgal species. Among the two tertiary amines, DMBA is considered as a potential solvent in terms of both lipid extraction and solvent recovery followed by DMCHA. In concern with DIPEA, it was di cult to recover the amines after rst extraction and therefore, the recovered solvents of DMBA (23-28%) and DMCHA (25-30%) were used for second cycle of extraction. Though the solvent recovered was in minimal quantity, it retains its lipid extracting e cacy as the lipid content extracted from fresh and recovered solvent was similar as evident from Fig. 4a.

TLC for qualitative estimation
Qualitative determination of lipid types present in switchable solvent extracted total lipids was carried out. To ensure the presence of neutral lipids in DMBA, DMCHA, and DIPEA extracted total lipids, TLC was carried out. Further, only one strain i.e., Picochlorum sp. NITT 04 was chosen for TLC experimentation as TLC refers to be qualitative separation and therefore, lipids of all strains were subjected for GC analysis to quantify the fatty acid composition. As shown in Fig. 5, the presence of neutral lipid (Picochlorum sp.) in TLC is encircled in all solvent types used. Based on i) the presence of neutral lipid ii) solvent recovery and iii) lipid content and (iv) switchability phenomenon, DMBA is considered as ideal solvent for lipid extraction.

Fatty acid compositional analysis of microalgae
The predominant fatty acids present in Chlorella sp. NITT02, Chlorella sp. NITT05, and Picochlorum sp. NITT04 was presented in Table 2. Among the fatty acids, the 16 carbon long chain palmitic acid is the found to be predominant at about 24.77, 29.4, and 28.32% in Chlorella sp. NITT 02, Chlorella sp. NITT 05, and Picochlorum sp. NITT 04, respectively. Palmitic acid is one of the major saturated fatty acids which are commonly found in many microalgal species. The difference among the fatty acid dominance for all the three strains was clearly visualized based on color-coding in Heat map (Fig. 6). The second most predominant fatty acid present in tested strain was oleic acid (C18:1), which accounts 14.4, 17.7, and 18.5%, respectively for Chlorella sp. NITT 02, Chlorella sp. NITT 05, and Picochlorum sp. NITT 04 among the total fatty acid composition. In fact, many studies report the same type of fatty acids to be the major component in the total FAME content 15,23−25 . As graphically represented in heat map, of the strains analyzed for FAME, Chlorella sp. NITT 02 possesses high levels of polyunsaturated fatty acids (PUFA) such as linoleic acid, linolenic acid, and Cis-13,16-docosodienoic acid, which contributes 27% in total fatty acid composition of the strain. In contrast, Chlorella sp. NITT 05 accumulates saturated fatty acids (SFAs) and monounsaturated fatty acids (MUFAs) maximally. It is noteworthy to say that the presence of higher amount of SFAs than PUFAs makes the biodiesel less incline towards rancidi cation. In the case of Picochlorum sp. NITT 04, it is interesting to note that, monounsaturated fatty acids were present in higher concentration along with lower levels of PUFA. These results are similar to the studies conducted by El-kassas, (2014) and Yang et al., (2015) where high amounts of MUFAs are observed than the polyunsaturated fatty acids for marine Picochlorum sp. The single double bond containing fatty acid types namely oleic acid and palmitoleic acid accounts at about 18.5 and 8.5%, respectively in Picochlorum sp. of this study and further, linolenic acid content was estimated to be less than 12%. This fatty acid content is almost similar to the fatty acid content of Picochlorum oklahomensis where 13.85%, 8.2%, and 13.52% of oleic acid, palmitoleic acid, and linolenic acid resepectively were observed 11 . From Table 2, it is understood that maximal concentration of SFAs and MUFAs and minimal concentration of PUFAs were observed in both Chlorella sp. NITT 05 and Picochlorum sp. NITT 04. Likewise, 32.8% of SFAs and 62.3% of total monosaturated and disaturated fatty acids were observed in Picochlorum sp.
under ambient conditions 27 . Fatty acid composition decides the fuel properties of biodiesel. The fuel properties to be assessed for using biodiesel as a fuel are oxidative stability, cetane number, viscosity, cold ler plugging point or low temperature property, ash point, pour point, ash content, total glycerol content, calori c value etc 28 . Higher concentration of SFAs is bene cial for biodiesel as they determine the oxidative stability of the fuel and higher SFA content is linearly proportional to oxidatively stable biodiesel. Also, higher cetane number (CN) is obtained by the presence of high level of SFAs where CN determines the ignition potential i.e., higher CN, less time for ignition and vice versa. At the same time, presence of linolenic acid greater than 12% causes the CN to be very low, thus making the quality of biodiesel to be at low standard 29 . The FAME pro le of the tested microalgal species in this study satis es the above-mentioned properties of biodiesel standard. On the other hand, high level of SFAs increases the viscosity as well as raise Cold lter plugging point (CFPP) thus, claiming the biodiesel to be unsuitable to operate under low temperature 30 . Overall, there should be optimal balance of saturated and unsaturated fatty acids to qualify the properties listed for biodiesel standard. Autooxidation of the fuel relies upon the double bond present in the fatty acids. Increase in the double bond numbers (PUFA) in fatty acids makes the biodiesel susceptible to autoxidation. Compared to allylic positions, bis-allylic positions in the fatty acid chain are susceptible to autooxidation. And it is widely known that linoleic acid contains one bis-allylic position and linolenic acid contains two bis-allylic position which is why their concentration in the total FAME content is supposed to be less 31 .

Conclusion
The present study showed the potential of tertiary amine to be used as green solvent in microalgal lipid extraction amidst the conventional solvents available for lipid extraction. Among the tertiary amines used, DMBA exhibited better extraction e ciency than DMCHA. Direct use of microalgal suspension without dewatering and recycling of solvent for subsequent lipid extractions in this study portrays green energy from sustainable source. Moreover, the FAME pro le of individual marine microalgal species elucidated the prospects of being a good biodiesel. Overall, this study imparts the use of tertiary amines as switchable solvent for cost and energy effective lipid extraction from microalgae.

Strain collection and maintenance
Marine microalgal samples collected from three different sites across the South-East coast of India were brought to the laboratory and isolated through conventional puri cation methods. The unialgal and axenic microalgal strains were inoculated in ASN-III medium 32 in a thermostatically controlled environment under the speci ed conditions of 25ºC ± 2ºC with 1500 lux light intensity. All the experimental cultures were maintained at arti cial illumination of 16 h light: 8 h dark photoperiod.

Growth assessment
To harvest excessive amount of biomass from microalgae, growth pro le of the strains needs to be known. Optical density was taken as growth metrics in this study. Biomass density of cultures was determined from 0th day to day 27 spectrophotometrically based on the absorbance (Optical density) at 750 nm as there is no chlorophyll interference at this wavelength.

Switchable solvents for lipid extraction
Switchable solvents used for lipid extraction are tertiary amines, which includes DMCHA (Dimethycyclohexylamine), DMBA (Dimethylbenzylamine) and DIPEA (Diisopropylethylamine). The methodology of tertiary amine as switchable solvent for wet lipid extraction from marine microalgae is illustrated in Fig. 2. In brief, known volume of microalgal suspension taken directly from the culture ask was used for lipid extraction without drying, which was ultrasonicated for rupturing the cells. To this, 1:1 (v/v) of tertiary amines (DMCHA, DMBA and DIPEA) was added which formed two immiscible layers. It was then subjected to constant stirring for 1 hour at 1000 rpm to allow the tertiary amine to extract lipids from ruptured microalgal cells. After this, CO 2 was purged to enable switchability and lipid in top layer was collected and the hydrophilic amines were switched to its native (hydrophobic) state by purging N 2 .
Eventually, the recovered SPS was used for another cycle of lipid extraction. The lipids extracted were dried and expressed in %.

Qualitative analysis of lipids extracted by tertiary amines
The total lipid extracted by each of the tertiary amines (DMCHA), DMBA and DIPEA) was qualitatively analysed for different classes of lipids by Thin Layer Chromatography (TLC). Silica gel coated plate was used as a stationary phase and the solvent mixture in the ratio of 70:30:1 (Hexane: Diethylether: Acetic acid) was used as mobile phase. The solute and TAG standard (Triolein) were loaded and run in parallel. The presence of the solute spot was compared with the standard lipid and the lipid classes were separated and visualized by Iodine vapour and Charring method. For visualization through charring method, TLC plates were sprayed with 10% CuSO 4 in 8% orthophosphoric acid solution and kept for charring at 180ºC for few minutes 33 .

FAME production from Switchable solvent extracted lipids
Conversion of lipid to FAME was carried out based on our previous work 34 . In brief, homogeneous acid catalysis using methanolic sulphuric acid (3.5%) was used for fatty acid methylation; 65 °C and 2.5 h was set as reaction temperature and reaction time, respectively. At the end of the reaction, the crude or unpuri ed FAME contents were allowed for the separation of ester and glycerol and the upper FAME layer was collected and puri ed with water wash. The puri ed FAME was dissolved in hexane for its pro le analysis.
4.4.1. Gas chromatographic conditions for FAME analysis FAME composition of switchable solvent extracted lipids were analysed on gas chromatography (Perkin Elmer Clarus 500) coupled to SP 2560 capillary column (100 m length) and detector-FID. Nitrogen as carrier gas with ow rate -1 mL min − 1 was used. The oven temperature-140 °C -240 °C, injection port and detector temperature-260 °C and 45 min total run time are the operating conditions of the GC. The peak area of the fatty acids was compared with the peak area of Supelco 37 component FAME mix standard to ascertain and quantify each fatty acid analyte.

Statistical analysis
For heat map plotting, "R" language has been used through R studio package. All the experiments were carried out in triplicates to obtain reproducible data with accuracy. Positive and negative deviation of the triplicate experiments was recorded. Declarations