The development of renewable energy technologies has advanced in response to an ever-increasing demand for fossil fuels and global warming caused by CO2 emissions from their combustion [1][2][3]. Biodiesel produced from microalgae is considered a promising substitute for fossil fuels [4][5] because microalgae boast a high growth rate and high lipid content. These features enable a high rate of carbon fixation with the potential for highly efficient biodiesel production. Furthermore, microalgae can be easily grown on non-arable land, which avoids competition with food production [6][7][8].
Biodiesel production from microalgae consists primarily of microalgae cultivation, harvesting, lipid extraction, and transesterification [9]. However, harvesting and lipid extraction are substantial bottlenecks in the development of an energy-efficient and cost-effective process for conversion of microalgae to biodiesel [10][11]. Microalgae typically exhibit small cell size (5–50 µm) and low density (0.5 ~ 5 g/L) in growth media. These factors make it difficult to directly extract lipids without some form of harvesting pretreatment [12][13]. In a typical process, microalgae suspensions are first thickened via gravity sedimentation, flocculation, or flotation to obtain slurry with a biomass content of 3–7%. This concentrated slurry is then mechanically dewatered by filtration or centrifugation to obtain cake with a biomass content of 10–25%. In a final step, the cake is thermally dried to a biomass content of > 90% [5][14][15].
Extraction of lipids is simpler from a completely dry microalgae sample than from a sample that has just been dewatered. A Soxhlet extraction with n-hexane obtained a 45% (dry base) yield of lipid from dried Schizochytrium limacinum [16]. Jie et al. [17] used ethanol to extract 48% (dry base) lipid from dried Synechocystis PCC 6803. Supercritical extraction using CO2 and ethanol has been used to obtain a 34% (dry base) yield of lipid from dried S. limacinum powder [16]; an 18.1% (dry base) lipid yield from dried Chlorella spp. powder was obtained using a mixed extraction solvent of methanol:ethyl acetate at a volume ratio of 2:1 [18].
However, drying is an energy-intensive process. Removal of 1 kg of water from mechanically dewatered microalgae (~ 20% biomass) by thermal drying requires 3,560 kJ of energy input, rendering the net energy balance negative [5]; thus, the energy output from the extracted biodiesel is less than the energy needed to produce the biodiesel [3].
Energy consumption during biodiesel production from thermally dried microalgae is nearly 4,000-fold greater than energy consumption during biodiesel production from merely mechanically dewatered microalgae (~ 20% biomass; wet microalgae)[19]. Hence, a positive net energy balance can be achieved without thermal drying [13][3]. Therefore, several investigations have focused on lipid extraction from wet microalgae. For example, Lakshmikandan [20] used a mixed solvent of hexane and isopropanol (3:2 v/v) to extract lipids from centrifugation-harvested Chlorella vulgaris (biomass content ~ 8.9%) and obtained a maximum lipid yield of 22.5% (dry base). Ethyl acetate has been used to extract lipids from wet Isochrysis galbana (5% biomass) at a yield of 17.6% [21]. Among six evaluated solvent systems, isopropanol:hexane (2:1 v/v) was the most effective in extraction of lipids from wet Scenedesmus obliquus (20% biomass), affording a 7.8% lipid yield (dry base) [3]. A solvent system of chloroform:methanol:sulfuric acid (1:1:0.05 v/v), combined with microwave irradiation, was used to obtain a 19.0% (dry base) yield of lipids from centrifuged Chlorella pyrenoidosa (water content, 80 wt.%) [22].
Because dewatering via centrifugation or filtration requires considerable energy compared to the thickening process [23], direct extraction of lipids from thickened microalgae would significantly improve the net energy gain. Liquid dimethyl ether (DME) is a promising solvent for extraction of lipids from thickened microalgae. Liquid DME is partially miscible with water (7–8 wt.% DME; room temperature) and features a high affinity for organic compounds. Thus, DME is suitable for extraction of lipids from wet biomass samples with simultaneous dewatering. This combined process represents considerable energy savings [24][25]. In addition, DME is a gas under ambient conditions; thus, it can be easily liquefied at 0.51–0.59 MPa and room temperature (20–25 °C). The low boiling point of DME (i.e., -25 °C) allows it to be easily removed via evaporation for recycling/reuse [25][26].
DME has been successfully applied for extraction of lipids and removal of moisture from dewatered biomass, including sludge [25][27], cattle manure [28], microalgae [29] and vegetables [30]; however, its use with merely thickened samples has not been studied. This is largely because its low polarity results in immiscibility with microalgae suspended in water. Although DME can absorb a small fraction of water, as mentioned above, greater proportions of water result in discrete aqueous and organic layers, thereby preventing DME from coming into close contact with microalgae cells. This effect is more noticeable with marine algae, because the polarity of the algal slurry is enhanced by dissolved salts. Although not the most economical approach, one solution involves the use of large volumes of DME [31]; specifically, a 167:1 weight ratio of DME to microalgae (dry base) was necessary for extraction of lipids from a sample containing 9% solids. Another potential solution involves the use of a solvent with infinite miscibility in water for adjustment of DME polarity. Adjustment of the cosolvent ratio allows DME to be completely mixed with water. This type of additive is regarded as an entrainer, and has been shown to improve the efficiency of supercritical fluid extraction [32][33].
Here, liquefied DME was used to extract lipids from AlCl3-flocculated Nannochloropsis oculata (solid content 18.3 g/L). A suitable entrainer for DME was identified among ethanol, dimethyl sulfoxide (DMSO), acetone, and tetrahydrofuran (THF). Extraction performance was evaluated with respect to changes in extraction time, DME dosage, entrainer composition, and entrainer proportion in DME. The performance of our modified DME-based method was compared to the performances of Bligh & Dyer and Soxhlet methods in terms of raw lipid yield, fatty acid yield, and C/H/N composition. For each method, extracted lipids were characterized by thermal gravimetry (TG)/differential thermal analysis (DTA), Fourier-transform infrared spectroscopy (FTIR), and trace elemental analyses.