A Review: Microalgae and Their Applications in CO2 Capture and Renewable Energy

Fossil fuels, which are recognized as unsustainable sources of energy, are continuously consumed and decreased with increasing fuel demands. Microalgae have great potential as renewable fuel sources because they possess rapid growth rate and the ability to store high-quality lipids and carbohydrates inside their cells for biofuel production. Microalgae can be cultivated on opened or closed systems and require nutrients and CO2 that may be supplied from wastewater and fossil fuel combustion. In addition, CO2 capture via photosynthesis to directly fix carbon into microalgae has also attracted the attention of researchers. The conversion of CO2 into chemical and fuel (energy) products without pollution via this approach is a promising way to not only reduce CO2 emissions but also generate more economic value. The harvested microalgal biomass can be converted into biofuel products, such as biohydrogen, biodiesel, biomethanol, bioethanol, biobutanol and biohydrocarbons. Thus, microalgal cultivation can contribute to CO2 fixation and can be a source of biofuels. This article reviews the literature on microalgae that were cultivated using captured CO2, technologies related to the production of biofuels from microalgae and the possible commercialization of microalgae-based biofuels to demonstrate the potential of microalgae. In this respect, a number of relevant topics are addressed: the nature of microalgae (e.g., species and composition); CO2 capture via microalgae; the techniques for microalgal cultivation, harvesting and pretreatment; and the techniques for lipid extraction and biofuel production. The strategies for biofuel commercialization are proposed as well.


INTRODUCTION
Currently, global energy exigency is increasing greatly with the increasing world population and improving quality of life. In general, fossil fuels have been the major source of energy. Nevertheless, decreasing fossil fuels with increasing fuel demand is unavoidable. To replace fossil fuels, various biomass feedstocks, including both terrestrial plants and aquatic algae, have been discovered to generate renewable fuels (Bahadar and Khan, 2013). Aquatic microalgae are ideal for producing liquid fuels because their rapid growth, high biomass yields, product variety and simple harvest from ponds or closed bioreactor systems allow them to be potentially consumed as sustainable environmentally friendly carbon-neutral fuel sources (Gao et al., 2012;Sing et al., 2013).
Microalgae are microscopic organisms that typically grow suspended in water and are driven by the same photosynthetic process as that of higher plants (Hanelt et al., 2007).
Microalgae can comprise bacteria (cyanobacteria), diatoms (e.g., Chromalveolata), other protists (e.g., Chromista), and unicellular plants (e.g., Chlorophyta) (Bahadar and Khan, 2013). However, unlike higher plants, microalgae do not require a vascular system for nutrient transport, as every cell is photoautotrophic with directly absorbing nutrients. Microalgal cells are sunlight-driven cell factories that can convert carbon dioxide (CO 2 ) into raw materials for producing biofuels (e.g., biohydrogen, biodiesel, and bioethanol), animal food chemical feedstocks and high-value bioactive compounds (e.g., Docosahexaenoic acid (DHA)) (Spolaore et al., 2006;Milledge, 2011;Razzak et al., 2013). In particular, the ability of these cells to absorb CO 2 suggests microalgae cultivation as an attractive alternative for CO 2 sequestration that can be applied to fossil fuel power plant gas effluents to facilitate the reduction of greenhouse gas emissions (Yun et al., 1997).
CO 2 fixation via microalgae is a potential and promising method for CO 2 capture and storage (Masakazu and Masahiro, 1997;Naoto and Masahiro, 1997;Razzak et al., 2013;Zhao and Su, 2014). CO 2 fixation and storage via microalgae are essentially photosynthesis, which can transform water and CO 2 to organic compounds without extra energy addition or consumption and without secondary pollution. Compared to other carbon capture and storage (CCS) methods, CO 2 fixation via microalgae has many benefits, such as a high photosynthesis rate (e.g., 6.9 × 10 4 cells/mL/h (Suali and Sarbatly, 2012)), a rapid growth rate (0.7-3.2 day -1 (Maeda et al., 1995;Ryu et al., 2009)), good environmental adaptability and low cost of operation. As a special advantage, biomass from microalgae for energy consumption is provided after CO 2 capture. The performance of CO 2 fixation via microalgae and biomass production depends on the cultivation conditions (e.g., temperature, light, pH, and nutrient availability), species of microalgae, CO 2 concentration and toxic pollutants in the flue gas (Zhao and Su, 2014).
The high CO 2 and nutrient requirements of microalgae can be supplied by flue gases and waste water from other industrial processes, providing ecological advantages while lowering the cost of biomass production. Microalgae can be used as bioremediation agents to remove inorganic nutrients from wastewaters (cost of wastewater treatment about 0.15-6.0 USD/m 3 (Fu et al., 2008;Yuan et al., 2010)) and to improve water quality due to their high capacity for nutrient uptake (Razzak et al., 2013). Microalgae that grow in the bioremediation of waste are further processed into a wide spectrum of fuel products, including (1) hydrogen (H 2 ) via direct and indirect biophotolysis, (2) biodiesel through transesterification, (3) biomethane via anaerobic digestion, (4) bioethanol and biobutanol via fermentation, (5) bio-oil via thermochemical conversion and (6) green diesel and gasoline through direct catalytic hydrothermal liquefaction (Demirbas, 2009a, b;Damartzis and Zabaniotou, 2011;Nigam and Singh, 2011;Huang and Tan, 2014). An integration of microalgal cultivation with CO 2 utilization and microalgal application as biofuel is presented in Fig. 1. Microalgae grow by consuming nutrients that are dissolved in water or wastewater and CO 2 from ambient air or the exhaust gases from combustion process under solar energy. Here, the microalgae biomass is obtained, and O 2 is released into the atmosphere. Finally, the biomass is converted into energy and food supplementation for humans and animals through a process including extraction (e.g., organic solvent extraction, supercritical fluid (SCF) extraction, etc.), fermentation, transesterification, etc. Therefore, the cultivation of microalgae provides three important benefits, including (1) CO 2 capture from a fossil fuel-based power plant, (2) wastewater treatment and (3) renewable energy sources (Razzak et al., 2013).
Recently, studies are being developed to establish processes and technologies for application of microalgae at the industrial scale. However, the transition from pilotscale to industrial-scale operations often exposes microalgae cells to hostile circumstances, resulting in reduction of product yields. The recovery of microalgae from highly dilute suspensions requires steps to lyse cells and reduce the extract yield. Therefore, the integration of the best microalgae cells and bioprocessing engineering methods to ensure economic and environmental feasibility and to minimize the number of full-scale tails is still challenging. Conditions for technically and economically viable biofuel resources should be competitive or cost less than petroleum fuels, have low to no additional land use, enable air quality improvement (e.g., CO 2 sequestration), and have minimal water use (Bahadar and Khan, 2013). Production costs associated with microalgae-based biofuels are the major barrier in the commercialization of biofuel, but fuels from microalgae are still promising in the view of competition with petroleum fuels.
This article reviews the literature on microalgae that were cultivated using captured CO 2 , technologies related to the production of biofuels from microalgae and the idea of the commercialization of microalgae-based biofuels to demonstrate the potential of microalgae. In this respect, a number of relevant topics are addressed: the nature of microalgae (e.g., species and composition); CO 2 capture via microalgae; the techniques for microalgae cultivation, harvesting and pretreatment; and the techniques for lipid extraction and biofuel production. The strategies for biofuel commercialization are proposed as well.

MICROALGAE SPECIES AND THEIR COMPOSITIONS
Microalgae are made up of eukaryotic cells. Microalgae cells consist of cell wall, plasmatic membrane, cytoplasm, nucleus and organelles, such as mitochrondria, lysosomes and golgi (Taher et al., 2011). Microalgae also have plastids, the bodies with chlorophyll that carry out photosynthesis. However, various strains of microalgae have different combinations of chlorophyll molecules -some have only Chlorophyll A, some A and B, while other strain, A and C (Um and Kim, 2009). The biomass of microalgae contains three main components: proteins, carbohydrates and lipids. The biomass composition of various algae is shown in Table 1. To achieve the maximum benefits from microalgae cultivation, it is essential to pay attention to the selection of suitable species. Microalgae cultivation is composed of a single specific strain that is precisely selected for producing the desired product and the most beneficial outcome of the cultivation process. The cultivation conditions, including (1) water media with adequate pH and temperature, (2) necessary contained nutrients and (3) CO 2 dosed in a controlled manner in the presence of sunlight, are also required for microalgae cultivation. The nitrogen source (e.g., ammonia and nitrates), other minerals and vitamins are the nutrients that must be provided sufficiently to ensure the proper growth of microalgae.

CULTIVATION TECHNIQUES
A wide range of microalgae cultivation techniques has been reported in the literature (Wang et al., 2008;Suali and Sarbatly, 2012;Bahadar and Khan, 2013;Zhao and Su, 2014). There are many types of microalgae cultivation techniques depending on (1) the investment cost, (2) the desired products, (3) the source of nutrients and (4) CO 2 capture. The cultivation systems are categorized into open and closed systems. The open systems are outdoor facilities consisting of ponds, lagoons, deep channels, shallow circulating units and others. In contrast, the closed systems are vessels or tubes with walls that are made of transparent materials and that are located in outdoors under sunlight irradiation or artificial irradiation (Razzak et al., 2013).

Open System
Until now, open ponds have been used for large-scale microalgae cultivation considering their simple construction and easy operation. The cultivation systems can be classified as (1) natural waters (e.g., lakes, lagoons, ponds, etc.) and (2) artificial water systems (e.g., artificial pond, tanks, and container). The shapes, sizes and types (agitated, inclined, etc.) of open systems depend on the applications. There are various types of ponds, including unstirred, raceway and circular ponds. Unstirred ponds ( Fig. 2(a)) are the most economical due to their simple management and construction. Commercial unstirred ponds are built in natural water ponds of less than half a meter in depth. Unstirred ponds are commercially used for some microalgae species, such as Dunaliella salina (Borowitzka and Borowitzka, 1990). However, this type of pond is very limited in its applications, given that microalgae are not able to grow under frequently poor growth conditions and competitive growth with contaminating protozoa, bacteria and viruses (Chaumont, 1993).
Raceway ponds (or stirred paddle wheel open ponds) ( Fig. 2(b)) are the most famous open system in current use. Raceway ponds are usually shallow, between 15 and 25 cm in depth. These ponds are normally constructed as either a single channel or groups of channels that are built by connecting individual raceway ponds. The productivity of the biomass Table 1. Compositions of microalgae based on dry matter (Um and Kim, 2009;Sydney et al. 2010;Singh et al., 2012 (Chen et al., 2009), (d) tubular photobioreactor (Carvalho et al., 2006), (e) plastic bag photobioreactor (Richmond, 2008), (f) air-lift loop reactor (Barbosa et al., 2003) and (g) flat plate photobioreactor (Carvalho et al., 2006).
in the raceway pond is 60-100 mg dry weight/L/day (Razzak et al., 2013). Raceway ponds are mostly used for the commercial culturing of four species of microalgae: Chlorella sp., Spirulina platensis, Haematococcus sp. and Dunaliella salina (Moheimani and Borowitzka, 2006). The circulation of the cultivation media in the raceway pond is induced by paddle. The optimal circulation velocity is required for water flow without the deposition of sedimentation or the aggregation of cells (Brindley et al., 2002). However, problems with solid deposition in stagnant areas are difficult to overcome. Circular ponds (central pivot) (Fig. 2(c)) are the oldest large-scale algae-cultivation open ponds. The depth of these ponds is approximately 25-30 cm. Microalgae are usually grown in concrete circular ponds of up to 45 m in diameter with agitation by rotating paddles. A 20-to-30cm-thick layer of inorganic nutrient solution with algae is exposed to sunlight and CO 2 bubble under the continuous movement of paddle wheels (Lee, 2001).
Although there are many advantages of the open systems, as described above, their limitations are as follows: (1) poor light consumption by cells, (2) evaporative losses, (3) diffusion limitation of CO 2 from the atmosphere, (4) large land area requirement and (5) easy contamination by unwanted algae, mold, and bacteria (Razzak et al., 2013). These limitations can be overcome using translucent plastic covers or greenhouses over the open ponds. However, this proposal cannot solve the contamination issues; in addition, the high capital cost, maintenance and overheating make open ponds that are covered by translucent plastics impractical due to the large land area of the pond.

Closed System
Microalgae can be grown in a closed system under controlled conditions, such as light utilization, area required and percentage of carbon dioxide. The closed system can address some of the problems that are associated with open systems. Photobioreactors are a type of closed pond system that are used for microalgae cultivation, as they can reduce contamination risk from unwanted algae, mold, and bacteria; control temperature; minimize water evaporation; and remove carbon dioxide losses. However, it should be noted that although photobioreactors significantly reduce the growth of competitive algal weeds, they cannot completely eliminate the growth of contaminants. Two disadvantages of this system are that it is difficult to construct and operate and is costly. The various designs of photobioreactors include flat plate, tubular, etc. (Borowitzka, 2007).
Tubular photobioreactors (Fig. 2(d)) are made with transparent materials and are placed in outdoor facilities under sunlight irradiation. Gas exchange vessels to supply CO 2 , air and nutrients and to remove O 2 are connected to the main reactor (Richmond, 2004;Chisti, 2007). For the design of these cultivation vessels, a large surface area per unit volume is required to maximize the exposure of microalgae to sunlight. The tube sizes are generally less than 10 cm in diameter to maintain sunlight permeability. In a typical tubular microalgae cultivation system, the medium is circulated through the tubes, where it is exposed to sunlight for photosynthesis. The medium is circulated back to a reservoir using a mechanical or an airlift pump. The pump also helps to maintain a highly turbulent flow within the reactor to prevent the flocculation of microalgal biomass (Chisti, 2007). A fraction of microalgae is usually harvested after it circulates through the solar collector tube, permitting the continuous operation of the system. However, tubular photobioreactors are still studied on the laboratory scale but are not practical. In some photobioreactors, the tubes are a coiled spiral, forming helical-tubular photobioreactors. These types of reactors are suitable for the cultivation of microalgal species in the presence of sunlight. An artificial light sometime replaces natural light to enhance microalgae growth. However, the use of artificial light adds to the investment costs, leading to the helical-tubular photobioreactors only being adequate for the manufacture of high-value added products (Morita et al., 2001;Briassoulisa et al., 2010).
Microalgae can be cultivated in transparent polyethylene bags, called plastic bag photobioreactors (Fig. 2(e)). These bags are hung or placed in a cage under sunlight irradiation, while the microalgae are mixed with air in the bottom of the bags (Razzak et al., 2013;Xia et al., 2013). Transparent polyethylene sleeves that are sealed at the bottom in a conical shape are used to prevent cell settling.
Airlift photobioreactors ( Fig. 2(f)) are simple and costeffective reactors for the mass cultivation of various types of microalgae. Acrylic glass is used as a material to construct airlift photobioreactors because it is inexpensive and available. There are two zones in airlift photobioreactors, including dark (called rinser) and irradiated zones. The airlift photobioreactors are believed to meet the desired criteria for the new generation photobioreactors of high light penetration and biomass production, ease of maintenance, and minimal contamination (Barbosa et al., 2003). However, airlift photobioreactors are sometimes difficult to scale-up given their complex flow pattern (Mirón et al., 2000). Vertical bubble columns and airlift cylinders can attain a substantially increased radial movement of fluid, with a high cycling of medium between the irradiated and dark zones. The advantages of these units include (1) high mass transfer, (2) good mixing with low shear stress, (3) low energy consumption, (4) relatively easy operation under sterile conditions, (5) good for the immobilization of algae on moving particles and (6) less photo inhibition and oxidation. However, the limitations of these units include (1) high manufacturing and maintenance costs, (2) smaller irradiation per unit surface area, (3) more sophisticated construction materials, (4) higher shear stress on microalgal cultivations, and (5) larger number of units needed to construct a commercial plant (Razzak et al., 2013).
Flat plate photobioreactors ( Fig. 2(g)) are very effective for the biomass cultivation of microalgae. These photobioreactors provide a high surface area to volume ratio for illumination and have a convenient modular design for scale-up (Barbosa et al., 2005). The biomass productivity of microalgae rapidly increases with the mixing rate, which can provide a proper amount of CO 2 to the cultivation while removing excess oxygen and increasing the flashing effect. Flat plate photobioreactors are suitable for outdoor and indoor cultivations, are good for algae immobilization, and are relatively cheap and easy to clean (Ugwu et al., 2008). Vertical flat plates can be accommodated in 1000-2000 L volume capacity units that are successfully operated for a long time period. Therefore, these are fully scalable    Trotta, 1981. Klinthong et al., Aerosol andAir Quality Research, 15: 712-742, 2015 718 photobioreactor units (Richmond, 2004). The ideal strain, advantages and disadvantages of each photobioreactor that has been used to cultivate microalgae are summarized in Table 2.

CO 2 CAPTURE VIA MICROALGAE
Biotechnological techniques for CO 2 fixation are conducted to reduce CO 2 emission. These processes are based on the use of reactors to generate photosynthetic reactions in which microalgae are used as biocatalysts in a series of biochemical reactions that are responsible for the conversion of CO 2 into photosynthetic metabolic products (Jacob-Lopes et al., 2010). Microalgal biomass contains approximately 50% carbon by dry weight (Mirón et al., 2003). This carbon is typically obtained from carbon dioxide; roughly, 100 tons of microalgal biomass fixes 183 tons of CO 2 (Huang and Tan, 2014). Microalgae are capable of using three different inorganic carbon assimilation pathways: (1) direct carbon dioxide assimilation via the plasmatic membrane; (2) the use of bicarbonate by inducing the enzyme carbonic anhydrase, which converts HCO 3 to CO 2 ; and (3) the direct transport of bicarbonate via the plasmatic membrane. However, there are a few differences in inorganic carbon assimilation pathways for different microalgae as demonstrated in Table 3. CO 2 must be fed continuously during daylight hours. The control of CO 2 feeding can be evaluated by pH measurements to minimize the loss of CO 2 . As a result, CO 2 fixation using microalgae can reduce the CO 2 emission from power plants, which has the positive environmental impact (Inoue et al., 1995;Yun et al., 1997;Abu-Khader, 2006;Brennan and Owende, 2010).

Photosynthesis
A promising pathway to capture CO 2 in the form of microalgal biomass is provided by biological processes via photosynthesis. In microalgae, photosynthesis releases oxygen and can be called "oxygenic photosynthesis". CO 2 is converted into lipids and other hydrocarbons in this process, explaining the designation of "CO 2 fixation process". In oxygenic photosynthesis, water is the electron donor, and oxygen is released after hydrolysis. The equation for photosynthesis can be written as follows: In this reaction, the standard free energy for the synthesis of glucose is 2,870 kJ/mol (Zhao and Su, 2014). The overall reaction can be divided into two pathways: (1) lightdependent reaction and (2) dark or light-independent reaction. The light-dependent reaction involves both photochemical and redox reaction steps. The overall equation for light reaction is as follows (Mukherjee and Moroney, 2011;Razzak et al., 2013): where ADP, P and NADP are adenosine diphosphate, phosphate and nicotinamide adenine dinucleotide phosphate. Light energy is used to synthesize ATP (adenosine triphosphate) and NADPH (nicotinamide adenine dinucleotide phosphate) which are energy storage molecules.
In the light-independent reaction or dark reaction, the enzyme ribulose bisphosphate carboxylase/oxygenase (RuBisCO) captures CO 2 from the atmosphere. This process requires the newly formed NADPH, called the Calvin-Benson Cycle (Peter et al., 2010). To be more specific, carbon fixation produces an intermediate product that is then converted into the final carbohydrate products. The carbon skeletons that are produced by photosynthesis are then used in a variety of ensuing processes, forming other organic compounds. For example, cellulose is the precursor for lipid and amino acid biosynthesis or is a fuel for respiration. The overall equation for the light-independent reaction is as follows (Razzak et al., 2013): The fixation or reduction of CO 2 takes place by combining CO 2 with a five-carbon sugar, ribulose 1,5-bisphosphate (Ru5BP), generating two molecules of a three-carbon compound, glycerate 3-phosphate (GP), also called 3phosphoglycerate (PGA). In the presence of ATP and NADPH (from light-dependent stages), GP is reduced to glyceraldehyde 3-phosphate (G3P). This product is also referred to as 3-phosphoglyceraldehyde (PGAL) or even as triose phosphate. Most of the G3P that is produced is used to regenerate Ru5BP so that the process can continue. Of the six molecules of triose phosphates, one is not "recycled" and often condenses to form hexose phosphate, yielding sucrose, starch and cellulose. The sugars that are produced during carbon metabolism yield carbon skeletons that can be used for other metabolic reactions, such as the production of amino acids and lipids (Razzak et al., 2013).

CO 2 Source
Microalgae use CO 2 as a carbon source. These microorganisms cannot grow without a CO 2 supply. An insufficient CO 2 supply is often the limiting factor of productivity. Thus, reducing the atmospheric CO 2 by microalgal photosynthesis is considered safe and advantageous for the human ecosystem (Mukherjee and Moroney, 2011).
The growth performance of some microalgae, e.g., Chlorella sp., can be negatively affected by CO 2 at a higher than 5% (v/v) concentration (Silva and Pirt, 1984;Lee and Tay, 1991;Cheng et al., 2006). However, some microalgae can grow under a flue gas CO 2 concentration level (10-15%), but the carbon fixation and biomass production rates are less than that under lower CO 2 concentration. Very few microalgal species are able to tolerate extremely high CO 2 levels of up to 70% (e.g., Chlorella sp. KR-1 and Chlorella sp. ZY-1) and 100% (e.g., Chlorella sp. T-1). The optimal CO 2 concentration for most microalgal species is usually recommended to be 0.038-10%, for example, the maximum biomass production was observed at 2.5% CO 2 for microalgae Chlorella sp. (Chiu et al., 2008) and at 6% for Scenedesmus obliquus and Chlorella kessleri (de Morais and Costa, 2007b).

Toxic Pollutants in Combustion Flue Gas SO 2
The presence of SO 2 strongly inhibits microalgae growth. When the SO 2 concentration exceeds 100 ppm, it is almost impossible to grow most microalgae (Hauck et al., 1996). Some microalgal species are able to grow with difficultly under conditions of high SO 2 concentration; however, they have a longer lag phase than in the absence of SO 2 . When the SO 2 concentration increases, the inhibition effect is enhanced, resulting in a sharp reduction in carbon fixation and biomass production, for instance, even chlorella sp. KR-1, which is considered a high-performance microalgal species, could not survive under the conditions of 150 ppm SO 2 with 15% CO 2 (Lee et al., 2000). SO 2 may not directly inhibit the microalgal growth (Matsumoto et al., 1997). The inhibition effects of SO 2 on microalgae growth may be attributed to the increased acidity. The H + release that is generated by the hydrolysis of SO 2 results in an increase in the acidity of the cultivation medium (Du et al., 2010). When the pH is less than 3.0, the microalgae cells are killed (Maeda et al., 1995). However, if the pH of the cultivation medium is artificially maintained constant by the neutralization method, the microalgal growth characteristic is approximately the same as that without SO 2 (Zhao and Su, 2014). These results seem to indirectly demonstrate that the effect of SO 2 on microalgae is transmitted through the pH value in the batch strategy; however, this situation is not seen in all occasions. Other investigations have indicated that the effect of SO 2 on microalgae is related not only to the pH value but also to SO 4 2and HSO 4 obtained from the hydrolysis of SO 2 . SO 4 2and HSO 4 are also inhibition factors of microalgal growth (Chiu et al., 2011).

NO x
In flue gas, the NO x emission level varies from several hundred to several thousand ppm with more than 90-95% NO and 5-10% NO 2 . After the flue gas de-NO x process, NO is still at the level of 50-200 ppm. It is difficult for NO to directly impact the growth of microalgae via pH in the cultivation medium. The NO concentration usually has a two-sided influence on the growth of microalgae. An extremely low concentration of NO may even be absorbed by the cultivation medium and transformed into NO 2 as the source of nitrogen nutrition for microalgae when using inorganic forms (Zhao and Su, 2014). However, this positive influence is quite limited: the increased NO concentration results in at least a decreased growth rate of microalgae for most microalgal species; NO at higher than 300 ppm may decrease microalgae growth (Chaumont, 1993) Typically, SO x and NO x in flue gas can be treated separately by flue gas desulfurization (FGD) and selective catalytic reduction (SCR) processes, respectively, or simultaneously by the combined treatment systems before the treated gas stream enters a microalgae reactor. The FGD commonly uses CaCO 3 to absorb SO 2 to form CaSO 4 . SCR uses ammonia or urea and a catalyst such as titanium oxide, silico-alumino-phosphate, zeolite, Al 2 O 3 , etc. to decompose NO x into N 2 and H 2 O (Jin et al., 2005;Hende et al., 2012). The examples of combined treatment systems are the DeSoNox or SNOX processes, where a catalytic reduction of NO x is combined with a catalytic oxidation of SO 2 (Trozzi et al., 2010).

CO 2 Fixation
Microalgae can eventually produce some CO 2 overnight, as occurs with other plants; however, the net CO 2 uptake is still positive. The CO 2 fixation rate is directly related to the light utilization efficiency and the cell density of microalgae. Microalgal CO 2 fixation involves photoautotrophic growth in which anthropogenically derived CO 2 may be used as a carbon source. Therefore, biomass measurements and growth rate evaluations are critical in assessing the potential of a microalgal cultivation system for direct CO 2 removal (Costa et al., 2004;Cheng et al., 2006). The CO 2 removal efficiency in a photobioreactor with microalgal cultivation can be determined as the difference in the CO 2 concentration of the incoming and outgoing effluents. The removal efficiency (%) can be determined using the following formula (Chiu et al., 2009 The efficiency of CO 2 removal or fixation in a closed cultivation system depends on the (1) microalgal species, (2) CO 2 concentration, (3) photobioreactor design and (4) operating conditions (de Morais and Costa, 2007a;Chiu et al., 2009). Chlorella vulgaris possesses a maximum CO 2 removal efficiency of 55.3% at 0.15% CO 2 in a membrane photobioreactor, and Spirulina sp. and Scenedesmus obliquus possess a maximum CO 2 removal efficiency of 27-38% and 7-13%, respectively, in a three serial tubular photobioreactor (Cheng et al., 2006), while their CO 2 fixation efficiencies were reduced to 7-17% and 4-9% under 12% CO 2 aeration (de Morais and Costa, 2007). In the other words, the CO 2 removal efficiency and fixation depend on the species of microalgae due to the physiological conditions of microalgae, such as the potential for cell growth and CO 2 metabolism.
The CO 2 fixation rate could be determined from the carbon content of the microalgal cell (Yun et al., 1997). The fixation rate was calculated as follows: where R CO 2 and μ L are the fixation rate (g CO 2 /m 3 h) and the volumetric growth rate (g dry weight/m 3 h), respectively, while M CO 2 and M C are the molecular weights of CO 2 and elemental C, respectively. C C is the average carbon content as measured by elemental analysis. The microalgal growth rate is determined in the linear growth regime. Due to climatic, land and water restrictions, there are challenges to collecting and utilizing microalgae directly on site. However, the economics of microalgae utilization can be increased using a two-stage process. In such a process, the CO 2 from a power plant or other source is first scrubbed (e.g., amine scrubber) and concentrated with a conventional process (Rochelle, 2009;Yu et al., 2012). The concentrated CO 2 is then transported to a suitable site for microalgae production. This process can be compared to the economics of other conventional CO 2 processes where CO 2 capture involves a separation process followed by transportation and finally disposal in deep oceans and/or depletion in gas wells (Razzak et al., 2013). One should consider that some microalgae species are tolerant to relatively high temperatures (close to and greater than 30°C). These types of microalgae can be cultivated in conjunction with the use of hightemperature flue gases from industrial neighbor sites (Wang et al., 2008). These thermo-tolerant strains may also simplify species control, as the optimum growth temperature of most microalgal species is in the range of 20-30°C. Several unicellular microalgae strains, for example, Chlorella sp., grow at temperatures of up to 42°C, and their tolerance to both high temperatures and a high CO 2 content makes them potentially appropriate microbial cells for photobioreactors that are involved in CO 2 capture for flue gases (Wang et al., 2008). Table 4 presents the microalgae strains that have been studied for CO 2 fixation.

Improvement and Application of the CO 2 Fixation Process
Until now, CO 2 fixation by microalgae, biomass production and energy consumption have experienced great progress in both laboratory-scale scientific research and pilot-scale applications (Chisti, 2010;Chi et al., 2011;Chisti and Yan, 2011). Process improvement as a middle measurement is necessary to more efficiently guarantee CO 2 fixation and biomass production. Whether in laboratory-scale research or pilot-scale application, the performance of CO 2 fixation and biomass production highly depend on the process conditions and parameters. For laboratory-scale research, the microalgae selection, cultivation and promotion to obtain high performance of microalgae species may be important factors. The evaluations of microalgae performance under drastic process and environmental factors, such as high CO 2 concentration, high temperature and toxic pollutants in flue gas, are also required. For pilot-scale application, microalgal cultivation is mainly affected by the cultivation temperature, light exposure and hydrodynamic conditions. Open cultivation is more influenced by outdoor temperature, light intensity (day and night) and season (adversely influences microalgal growth) compared to closed cultivation (Zhao and Su, 2014).
For closed systems, the improvements of the process parameters for CO 2 fixation and biomass production are mainly physicochemical parameters, including light exposure, nutrition conditions and hydrodynamic parameters, such as mixing and mass transfer rates by increasing the gas-liquid contact area and retention time, even though the economy nd practical use are in contention for their scale-up (Mirón et al., 1999). CO 2 fixation and biomass production can be improved using a high aeration rate to achieve turbulence mixing, a high CO 2 mass transfer rate and a high removal rate of excess oxygen in the cultivation medium (Sierra et al., 2008). Nevertheless, high stress to microalgal cells and high operating costs are still important challenges (Zhao and Su, 2014). As a result, CO 2 fixation through microalgae and biomass production in practical use are influenced by various factors, such as microalgal species and physicochemical and hydrodynamic conditions. To maximize the CO 2 fixation and biomass production performance, the synergistic effect and optimization process parameters are of concern and still need to be solved when using microalgae for CO 2 fixation and mass production on industrial scales.

HARVESTING OF THE MICROALGAE BIOMASS
Harvesting is an expensive and problematic segment of the industrial production of microalgae biomass due to the low cell density of microalgae, being typically in the range of 0.3-0.5 g/L, with exceptional cases reaching 5 g/L. However, the optimal requirement for the industrial scale is a cell sludge containing at least 300-400 g/L. Thus, the effluent microalgal suspension needs to be concentrated at least 100 times, which is an energy-intensive process (Wang et al., 2008).
After microalgae cultivation has reached the stationary phase, it is separated from water and the biomass recovered for downstream processes. However, the microalgae-harvesting process possesses a challenging task because microalgae have a micron size (1-20 µm) and are suspended in liquid. Currently, there are several methods to harvest microalgae, including (1) bulk harvesting to separate microalgae from suspension, such as natural gravity sedimentation, flocculation and floatation, leading to a concentration factor of 100-800 times, and (2) thickening to concentrate the microalgae slurry after bulk harvesting, such as centrifugation and filtration, leading to a concentration factor of 30 times. Flocculation and flotation are widely used for the bulk harvesting. The flocculation mechanism is to neutralize or reduce the negative charge on the microalgal cellular surface to aggregate cells in suspension, which can be collected by adding flocculants such as multivalent cations and cationic polymers (Molina et al., 2003). Flocculation can be performed by adjusting the pH of the cultivation to between 10 and 10.6 using NaOH to neutralize the negative charges on the cell surface, and the non-ionic polymer Magnafloc LT-25 is added. The ensuring flocculate is harvested and neutralized to generate a final concentration factor of between 200-and 800-fold. Here, significant breakthroughs in developing polymeric flocculants are urgently required through intensive research to further strengthen the potential use of flocculation in the harvesting process (Lam and Lee, 2012). Flotation depends on trapping cells by dispersed micro-air bubbles without the addition of any chemical reagents. The flotation can capture particles with a diameter of less than 500 mm by collision between a bubble and a particle and the subsequent adhesion of the bubble and the particle. The flotation process where microalgae float to the surface of medium is prone to harvest in microalgae mass culture, and has been used for specific strain, such as Spirulina platensis . The obtained sludge is very clean. However, the floating method is still challenging at a large scale (Wang et al., 2008). The Centrifugation is an efficient but energy-intensive method (Molina et al., 2003). The efficiency of centrifugation depends on the settling characteristics of the cells, the residence time of the cell slurry, and the settling depth. The settling depth can be kept small by the design of the centrifuge, and the resident time of the slurry can be controlled by the flow rate. The centrifugal recovery of the biomass is feasible for high-value products because it can process large volumes relatively rapidly, and the biomass remains fully contained during recovery (Heasman et al., 2000). Filtration, which operates under pressure or vacuum, is the preferred method for harvesting relatively large filamentous microalgae; however, the small cells of microalgae are not suitable, while membrane microfiltration and ultrafiltration are possible alternatives. However, current large-scale microalgal biomass production facilities do not generally apply membrane filtration because of its cost, to which membrane replacement and pumping are the major contributors (Hung and Liu, 2006). Moreover, microfiltration is more cost-effective than centrifugation (Molina et al., 2003). As a strategy to save biomass-harvesting costs, easyto-harvest microalgal strains for CO 2 fixation and biomass production should be selected (Wang et al., 2008). The comparison of harvesting methods is demonstrated in Table 5.

Drying of Microalgae Biomass
The extensive drying of microalgae biomass is required for biofuel production, as the presence of water will inhibit several downstream processes, such as lipid extraction and transesterification. Nevertheless, the drying step is energyintensive, which adds to the cost complexity of the overall production process. The microalgal slurry moisture content must be reduced to at least 10% by drying and dehydration. There are many types of drying technologies, such as drum drying, oven drying, freeze drying, and spray drying, that have been used with microalgae. The selection of the best drying method depends on the required operation scale and desired product value. Sun drying is an old and cheap drying method that can be performed easily by exposure to solar radiation. However, solar drying is not feasible, taking a long drying time and requiring a large drying area due to limited sunlight and uncertain sunlight time (Taher et al., 2011;Lam and Lee, 2012). Moreover, sun drying does not have any sterilization effect, unlike oven or drum drying. In contrast, spray drying can be used for high-value Table 5. Comparison of microalgal harvesting and drying methods (Taher et al., 2011;Zhang et al., 2014).

Method
Advantage Freeze drying 1) Gentle 2) Low species deterioration 1) Slow process 2) Cost intensive products but has the disadvantages of being expensive and possibly significantly deteriorating the algae. Among the drying technologies that can be applied to microalgal concentrate, freeze drying is favored for its mild operating onditions. Thermal drying, although commonly used in laboratory practice, is not recommended because it degrades thermodegradable lipids, results in the evaporative loss of volatile lipids, and yields powder of a nonuniform particulate size (Pourmortazavi and Hajimirsadeghi, 2007). Moreover, freeze drying has the advantage of breaking up species cells and turning them into fine powder, making homogenization unnecessary (Ahlgren and Merino, 1991). However, freeze drying is a slow process and requires a very high capital investment. Table 5 summarizes the advantages and disadvantages of each drying technique.

Particular Size Reduction
Once dried, microalgal biomass forms powder or agglomeration that can be milled into different particulate sizes. Reducing the particulate size of microalgal powder prior to lipid extraction generally enhances lipid recovery because it increases the interfacial surface area that is available for biomass-solvent contact and shortens the diffusion pathway of the extraction solvent. However, the exceedingly small particulate size of the microalgal powder may lead to a higher tendency of lipid re-adsorption and fluid channeling effects in the extraction vessel (for supercritical carbon dioxide (scCO 2 ) extraction) and in homogeneous lipid extraction (Pourmortazavi and Hajimirsadeghi, 2007). Pourmortazavi and Hajimirsadeghi (2007) also conducted a study on the scCO 2 extraction of oil from biomass and verified that smaller biomass resulted in higher oil recoveries.

Cell Disruption Technique
The cell disruption method aims to destruct the cell wall of microalgae to release the intracellular components and to enhance the efficiency of the extraction. Cell wall disruption methods can be categorized into biological, chemical and physical methods.

1) Biological Method Enzymatic Disruption
Enzymes can be applied in oil extraction from microalgae, as they can mediate the hydrolysis of cell walls, enabling the release of their content. For cell wall degradation, cellulases are the most frequently applied enzymes (Mercer and Armenta 2011;Gonçalves et al., 2013). Despite being expensive, enzymes offer several advantages over other cell wall disruption methods. Enzymes present a higher degradation selectivity than physical methods. Furthermore, microalgal cell walls are more recalcitrant than those of other microorganisms, being very resistant to degradation. Thus, the use of physical methods requires higher energy amounts (Gonçalves et al., 2013). Chen et al. (2013) studied bacteria of Flammeovirga yaeyamensis in co-cultivation with microalgae of Chlorella vulgaris ESP-31. The cell wall of the microalgae was hydrolyzed by hydrolytic enzymes (i.e., amylases, cellulases and xylanase) that were produced by the bacteria, increasing the microalgae oil extraction efficiency by nearly 100%.

2) Chemical Method Ozonation
Ozone is an effective oxidizer of organic compounds and an effective disinfecting agent that is capable of deactivating a variety of organisms. Ozonation has been widely used in water treatment for disinfection and for the oxidative removal of contaminants . Cell rupture by ozonation is due to the oxidative reaction of ozone, which compromises the cell membrane's ability to regulate the permeability of substances, eventually leading to the outflow of the cytoplasm (Erden et al., 2010;Yang et al., 2013). Recently, the pressure-assisted ozonation (PO) technique was developed, which is capable of solubilizing cellular materials at high efficiency via cell wall disruption (Cheng et al., 2012).  examined the performance of ultrasonication, conventional bubbling ozonation (CBO) and PO as cell disruption techniques to obtain microalgae lipids from Chlorella vulgaris BG11. The highest lipid yield was obtained using PO, whereas the lowest yield was obtained using CBO. Using CBO and PO involving ozone, saturated fatty acids (C16:0 and C18:0) were predominant. Thus, PO is an effective cell disruption method for biodiesel production with a high yield of lipids and saturated hydrocarbon products.

Hydrogen Peroxide
Alkaline hydrogen peroxide (H 2 O 2 ) treatment is recognized to decrystallize cellulose and has widely been used for bleaching high-lignin wood pulps. The role of H 2 O 2 is to provide highly reactive radicals. When the pH is above 6, H 2 O 2 is readily decomposed to active radicals such as hydroxyl radical ( • OH) and superoxide anion radicals (O 2 -). These active radicals can degrade and oxidize cell wall of biomass (Gould, 1985;Phan and Tan, 2014). Chlorella vulgaris was pretreated by aqueous solution of H 2 O 2 and H 2 O 2 with FeSO 4 (Fenton's reactant) to disrupt cell wall prior to lipid extraction (Steriti et al., 2014). The use of H 2 O 2 allowed increasing the extraction lipid yield from 6.9 to 9.2%, while that with FeSO 4 led to a corresponding increase of lipid yield to 17.4%. According to the Fenton's reaction, the reaction between H 2 O 2 and Fe 2+ ion (from FeSO 4 ) can produce hydroxyl radical ( • OH) which in turn may attack and degrade the organic compounds constituting the cell wall. Probably such a reaction occurs preferentially in specific zones of the cell wall constituted by organic compounds that are easily oxidized by • OH radicals, leading to the disruption of cell wall and the release of the intracellular material, including lipids (Wu et al., 2010).
Once transferred in the liquid bulk, even lipids, especially unsaturated lipids, might be attacked by hydroxyl radical generating degradation products such as lipid peroxide (González et al., 2012), leading to positive effect for upgrading fatty acid methyl ester composition in biodiesel.

Acid Treatment
The acids that are commonly applied in the cell disruption step are hydrochloric acid and sulfuric acid, which are everyday industrial chemicals carrying minimal toxicity in their applied concentrations. The acid treatment is preferable, as it provides high efficiency in converting polysaccharidecontaining materials. In the case of microalgal biomass, complex carbohydrates are entrapped in the cell wall; thus, cell disruption through acid hydrolysis is preferred. Harun and Danquaha (2011) explored the influence of acid exposure as a microalgal pretreatment strategy for bioethanol production. Using a 1% sulfuric acid solution at 140°C for 30 min as a pretreatment to hydrolyze the cell wall of Chlorococcum humicola enhanced the bioethanol yield from 16% to 52% (g bioethanol/g microalgae) due to the release and conversion of the carbohydrates that were entrapped in the microalgae cell wall into simple sugars prior to the fermentation process. Recently, acid-catalyzed hot water was used to treat wet microalgae of Chlorella vulgaris before lipid extraction for the production of biodiesel. The use of hot water weakened or disrupted the crystal structure of cellulose and acid (i.e., sulfuric acid), aimed to improve the efficiency of hot water through acid hydrolysis. The extracted-lipid yield was enhanced through pretreatment by acid-catalyzed hot water .

3) Physical Method Electrical Method
The application of electrotechnologies, such as pulsed electric field (PEF) and high voltage electric discharges, were promising for intracellular extraction from bio-suspensions (Lebovka et al., 2011). PEF seems to be a potential alternative for oil extraction from microalgae. This technique applies brief pulses of a strong electric field to cells in the range of nanoseconds to microseconds, inducing the non-thermal permeabilization of the cell membrane and improving mass transport across the cell membranes. Permeabilization depends on the field strength and pulse number (Guderjan et al., 2005;Taher et al., 2011). Under sufficient conditions, irreversible damage of the membrane is reached, and the hardness of the cell is lost; thus, PEF can lead to the complete disruption of cells into fragments. Moreover, PEF requires less time and energy than other applied methods (Guderjan et al., 2005), and its use as a pretreatment for organic solvent extraction requires fewer organic solvents (usually presenting high toxicity) than the conventional organic solvent extraction (Guderjan et al., 2007). Nevertheless, few studies have applied PEF to pretreat microalgae. Grimi et al. (2014) investigated the intracellular extraction from the microalgae Nannochloropsis sp. disrupted by PEF, which permitted the selective extraction of water-soluble ionic components and microelements, low-molecular-weight organic compounds (amino acids) and water-soluble proteins.
However, this electrically based disruption technique was ineffective for the extraction of pigments, which required subsequent application of such methods as ultrasonication and high-pressure homogenization, but its power consumption was always noticeably smaller.

Ultrasonication
Another method that can be used to promote the cell wall disruption of microalgal cells is the application of ultrasonication. In ultrasonic-assisted pretreatment, cavitation emerges in the solvent. When cavitation bubbles suddenly burst, the extreme high pressure and temperature destroy the adjacent tissue. In this case, the cell walls are damaged by the repeated bursting of cavitation bubbles (Adam et al., 2012). The application of this ultrasound-assisted method to microalgal biomass can improve extraction efficiencies by reducing the extraction times and increasing the oil recovery yields. The experiments that were performed by Cravotto et al. (2008) with the microalgae Crypthecodinium cohnii showed cell disruption after applying ultrasonic pretreatment prior to Soxhlet extraction with n-hexane, resulting in an increase in the oil extraction yields from 4.8 to 25.9% compared to Soxhlet extraction without any pretreatment. Additionally, Ranjan et al. (2010) compared the oil extraction yields from Scenedesmus sp. using the following methods: (1) Bligh and Dyer's method and (2) ultrasound-assisted extraction followed by Bligh and Dyer's method. The obtained results demonstrated that method (2) generated a greater yield of oil extraction than method (1) due to the disruption of the cell wall and the release of the cell components to the solvent after pretreatment.

Microwave
Microwave pretreatment is frequently used in biochemical extraction from plants. In the microwave process, the oscillation of polar species and ionic conduction causes the temperature to rapidly increase, thereby improving the extraction efficiency (Ganzler et al., 1990). When microwaves irradiate the interior of a cell, the oscillation of polar substances causes rapid heating, moisture evaporation and increasing pressure, which lead to cell rupture followed by the release of the cell contents. Cravotto et al. (2008) applied organic solvent extraction with n-hexane and microwave-assisted solvent extraction (using the same solvent) to the microalgae Crypthecodinium cohnii, achieving oil recovery yields of 4.8 and 17.8%, respectively. Recently, Balasubramanian et al. (2011) promoted the cell wall disruption of Scenedesmus obliquus using the microwave-assisted method for oil extraction. Oil analysis indicated that microwave-extracted oil contained higher percentages of unsaturated and essential fatty acids (indicating higher quality) than that extracted without pretreatment. This study also validated for the first time the efficiency of a continuous microwave system for the extraction of lipids from microalgae. Higher oil yields, faster extraction rates and superior oil quality demonstrated that this system is feasible for oil extraction. Ma et al. (2014) used dichloromethane to extract oil from microalgae that were pretreated with microwave and ultrasound. The microwave pretreatment was more effective than the ultrasonic pretreatment for cell disruption because the microwave process provided pressure change through the rapid heating of the microalgae cells, whereas in the ultrasonic process, this rupture was caused by a shock wave outside of the microalgae cells.

Mechanical Method
The mechanical method is the most traditional method of disruption. For example, bead milling is a common way to grind materials into fine powder by shear force, which is caused by the mechanical force. Zheng et al. (2011) used bead milling as a pretreatment for microalgae, obtaining an approximately 10% lipid yield. However, a greater lipid yield could be obtained with manual grinding in liquid nitrogen. A blender was also used to disrupt the cell of Nannochloropsis oculata. The results in Table 6 show that the blender has a relatively high efficiency in breaking the microalgae cell wall and is the most energy efficient method (McMillan et al., 2013).

Laser
In an application of laser for cell disruption, it is suspected that the cell is disrupted by thermal effects similar to those associated with the thermolysis technique. It may be possible that cavitation bubbles induce shear damage, as well. McMillan et al. (2013) applied a laser to a microalgae slurry. The typical damage of perforated cells showed that lipid matter was released, and dried lipid matter was observed on the underside of the cover slip that was in contact with the sample and had risen to the top of the treated solution. Heavily deformed cells displaced lipid material, demonstrating that high temperatures were a major influence on disruption, even more so than bubble cavitation. Large air bubbles were also observed in the sample that encapsulated the lipid matter and pushed away disrupted cells and debris in their path. Cell disruption was achieved at 96.53% though laser pretreatment (Table 6). Use of a laser is the most efficient disruption method, but only slightly better than the microwave. However, laser lysis was highly energy consuming and spatially limited to processing very small volumes at a time.
As above, there are many works focused on cell disruptions with biological, chemical and physical methods. However, only merits of physical method were reported (McMillan et al., 2013). Thus, the merits of biological and chemical methods are suggested to be further investigated.

Extraction
Lipids are one of the main components of microalgae, depending on the species and growth conditions. Microalgae contain lipids at approximately 2-60% of the total cell dry weight. Lipid oils that are derived from microalgae are interesting because these oils contain fatty acids (mostly 16 carbons to 22 carbons in chain length) and triglyceride compounds that can be transesterified into biodiesel. Technology for the production of biodiesel from microalgae must have a high specificity for lipids to minimize contaminants, i.e., carbohydrates, proteins, etc. (Banerjee et al., 2002). Purification technology should also favor the production of acylglycerols over other lipids, such as ketones, chlorophylls, sterols, polar lipids, and carotenes, which are not readily converted into biodiesel (Mojaat et al., 2008). Moreover, technology should have low operating and capital costs, require little energy and time, be safe, and show no sign of reaction with lipids. Table 7 demonstrates the methods, including organic solvent extraction, Soxhlet solvent extraction, and supercritical fluid extraction, that have been used to extract lipids from microalgae.

Organic Solvent Extraction
Lipid-solvent systems are governed by the principle of likedissolves-like in which lipids are extracted using non-polar organic solvents, such as chloroform or hexane (Geciova et al., 2002). This extraction can be classified into five steps: (1) the microalgae are exposed to the solvents, which penetrate the cell membrane and enter the cytoplasm; (2) the solvent interacts with neutral lipids via van der Waal's forces to form a solvent-lipid complex; (3) this complex diffuses across the cell membrane, such that the neutral lipids enter the organic phase, while water and solvent-contaminant complexes (with carbohydrates or proteins) enter the aqueous phase; (4) the organic phase is further separated; and (5) crude lipids are transesterified to generate biodiesel (Lewis et al., 2000;Halim et al., 2012;Bahadar and Khan, 2013). However, the disadvantages of using chemical solvents are mostly related to their high toxicity to humans and the surrounding environment. Chemical solvents, such as n-hexane, methanol, and ethanol, and mixed polar/nonpolar solvents, such as methanol/chloroform (2:1 v/v, Bligh and Dyer method) or isopropanol/hexane, are effective at extracting microalgae lipid, but the extraction efficiency is highly dependent on the microalgae strains. One study found that using isopropanol as a co-solvent (0.068 g of lipid/g of microalgal biomass) increased the lipid yield from Chlorococum sp. by up to 300% more than using pure hexane (0.015 g of lipid/g of microalgal biomass) (Halim et al., 2011). Another study reported that beadbeaten Botryococcus braunii was exposed to five different organic solvents; chloroform/methanol yielded the highest lipid content (0.29 g of lipid/g of microalgal biomass). Nonpolar solvent n-hexane, although widely used to extract oil from various seed crops, is insufficient to extract microalgae lipids because microalgal lipids are composed of a high concentration of unsaturated fatty acids; thus, the selectivity of lipids toward the solvent is largely reduced (Ranjan et al., 2010;Lam and Lee, 2012). n-Hexane, methanol and chloroform are highly toxic compounds that can cause safety and health hazards if proper precautionary steps are not taken. In contrast, ethanol emerged as a greener solvent because it has low toxicity level and can be generated from renewable sources. However, ethanol always generates a low extraction efficiency due to an azeotrope mixture (5% water), and the presence of water will reduce the extraction efficiency (Lam and Lee, 2012).
Organic solvent extraction is usually carried out in a non-continuous batch process, limiting lipid mass transfer equilibrium (Medina et al., 1998). Soxhlet solvent extraction continuously evaporates and condenses the solvent, avoiding the lipid mass transfer limitation and reducing solvent consumption (Wang and Weller, 2006). Soxhlet lipid extraction is more effective than batch extraction (Halim et al., 2011). However, the continuous distillation is an energy-consuming process. For extraction using organic solvents, diffusion is always the rate limiting factor in the overall mechanism. This factor becomes more serious in microalgae because the cell wall further resists solvent penetration into the inner cell. Thus, the cell disruption method (as describe above) can be used to increase solvent diffusion efficiency and results in the improvement of the microalgae lipid recovery rate. Although a higher microalgae lipid yield can be achieved after cell disruption, additional energy is required (Lam and Lee, 2012).

Direct Liquefaction
Oil can be directly obtained from dried or wet microalgae by liquefaction. The high moisture content of microalgae requires much energy to remove water. Liquefaction directly converts the biomass into oil via reaction with water and carbon monoxide/hydrogen in the presence of sodium carbonate. These processes require high temperatures and pressures. In the liquefaction process, biomass is decomposed first into small molecules. These small molecules are unstable and reactive and can repolymerize into oily compounds with a wide range of molecular weight distributions. Liquefaction can be attained directly or indirectly. Direct liquefaction involves rapid pyrolysis to produce liquid tars and oils and/or condensable organic vapors. Indirect liquefaction involves the use of catalysts to convert non-condensable gaseous products of pyrolysis or gasification into liquid products. Minowa et al. (1995) used direct liquefaction at 300°C and 10 MPa to produce oil with a yield of 78.4% from Dunaliella tertiolecta with a moisture content of 78.4%. For high-moisture Botryococcus braunii treated with or without a catalyst of 5% Na 2 CO 3 at 300°C, more than 95% hydrocarbons was recovered (Bahadar and Khan, 2013).

Supercritical Fluid Extraction
Supercritical carbon dioxide (scCO 2 ) can be used as a solvent for lipid extraction from microalgae. scCO 2 extraction is a green technology that promises to replace organic solvent extraction. The basic principal of this technology is achieving a supercritical phase that is beyond the critical point of the fluid, at which separation of the liquid and vapor phase disappears and a single homogeneous phase appears. scCO 2 extraction is very efficient for lipid extraction for several reasons: (1) the crude lipid products are solvent free, (2) the rapid penetration of scCO 2 into microalgae cells promotes a high lipid yield, (3) the properties of scCO 2 are tunable through the adjustment of the operating temperature and pressure, and (4) degumming is not required because scCO 2 is not able to solubilize polar phospholipids (Mendes et al., 2003;Herrero et al., 2006;Sahena et al., 2009).
The applications of scCO 2 to extract microalgae lipids for biodiesel production have been explored recently. Lipid from wet paste Chlorococum sp. biomass was extracted using scCO 2 with a yield of 7.1% at a temperature of 60°C, a pressure of 30 MPa and an extraction time of 80 min. Moreover, the lipid yield obtained from the wet paste was greater than that obtained from the dried paste (5.8%), indicating that the energy consumption for the drying process can be reduced. The presence of water in the system acts as a natural polar co-solvent, facilitating the extraction of polar lipids and improving the total lipid yield (Halim et al., 2011). However, scCO 2 extraction and organic solvent extraction were compared by extracting lipid from Crypthecodinium cohnii. The lipid yield attained from organic solvent extraction was approximately twice that from scCO 2 extraction, indicating that microalgae strains and cultivation conditions play a significant role in determining the appropriate lipid extraction methods (Couto et al., 2010). Lipid extraction from Chlorococcum sp. using scCO 2 yielded 0.058 g of lipid/g of microalgae with a resident time of 80 min, while that using Soxhlet solvent extraction yielded 0.032 g of lipid/g of microalgae after 5.5 h (Halim et al., 2011). One study showed that 90% of lipids was recovered from Spirulina platensis using scCO 2 at 70 MPa and 55°C for 15 min, while the same amount was obtained using Soxhlet hexane extraction for 6 h (Bahadar and Khan, 2013). Chen et al. (2013) used a continuous high-pressure CO 2 extraction process to extract lipid from microalgal biomass with an extraction yield of 90.56%. Recently, scCO 2 (pressure of 50 MPa at 50°C and a flow rate of 3 mL/min) was applied for lipid extraction from Scenedesmus sp. and compared with static n-hexane extraction (shaker at 50°C and 100 rpm, overnight) (Taher et al., 2014a). The lipid extraction yield obtained from the former was 75% greater than that obtained from the latter. Taher et al. (2014b) also optimized the conditions for scCO 2 extraction. scCO 2 extraction was superior to other extraction techniques (Soxhlet, Bligh and Dyer lipid extraction, n-hexane, nhexane/isopropanol). The best operating conditions were 53°C, 50 MPa and a flow rate of 1.9 g/min, producing a lipid extraction yield of 7.41%. Li et al. (2014) extracted lipid from Tetraselmis sp. (strain M8) using different extraction methods, including Bligh and Dyer lipid extraction, organic solvent extraction, direct saponification and scCO 2 extraction. scCO 2 extraction technique (15 MPa, 40°C, 12 h for soaking and a flow rate of 5 mL/min for 30 min for flushing) resulted in the most effective extraction of microalgal lipids, especially for long-chain unsaturated fatty acids.

CO 2 -Expanded Liquid Extraction
CO 2 -expanded liquids (CXLs) have been applied in extraction, reaction and separation (Chen and Tan, 2007;Golmakani et al., 2012;Yang et al., 2012;Lin et al., 2013;Wei et al., 2013). CXLs can be continuously tuned from the neat organic solvent to scCO 2 through changing the CO 2 composition in liquid by adjusting operating pressure (Jessop and Subramaniam, 2007). CXLs have the benefit of requiring mild operating pressure and temperature, leading to the reduction of energy consumption hence the cost of the process (Sih et al., 2008). Similar to supercritical fluids, CXLs have shown to improve mass transfer by decreasing interfacial tension, reduce viscosity and increasing diffusivity (Eckert et al., 2007;Lin and Tan, 2008a, b;Herrero et al., 2013). Recenty, the CO 2 -expanded ethanol (CXE) has been applied for extraction of valuable bioactive from natural sources, since ethanol emerges as a greener solvent than other organic solvents. One work studied the extraction of gamma-linolenic acid from Arthrospira platensis (Spirulina) using CXE showed its extraction yield similar to pressurized extraction but greater than scCO 2 extraction (Golmakani et al., 2012). CXE was also applied to extract polar antioxidant of astaxanthin from Haematococcus pluvialis. Temperature and ethanol content had significant influences on astaxanthin yield and antioxidant activity (Reyes et al., 2014). Our group successfully applied CXE as the solvent to extract docosahexaenoic acid (DHA)-rich lipid from Schizochytrium sp. in continuous extraction process (Wang, 2014a). The lipid yield extracted via CXE in continuous extraction for 30 min was obtained approximately 87% of total lipid content determined by 48-h Soxhlet extraction, much better than ethanol and pressurized ethanol regarding yield and extraction time. The obtained results indicated that the CXE was appropriate for fast and efficient lipid extraction resulting from high diffusion, low viscosity and easy penetration into microalgae cell.

Biofuel Products Biohydrogen
Renewable biohydrogen production is of increasing interest because fossil fuel supplies are being depleted. Hydrogen gas has excellent potential as a renewable energy source because it produces only water when combusted, unlike the carbon pollution of fossil fuels. Biohydrogen has become a viable source given the current energy demand and environmental issues. The main objective is to improve hydrogen yield to make this process more economically available. However, the low yield and rate of hydrogen production are barriers for the commercialization of biohydrogen production. Cheap raw materials, efficient production techniques, and pilot tests of photofermentation plants should make biohydrogen a commercially available source of energy in the near future (Bahadar and Khan, 2013). Biohydrogen can be produced by photolyzing water using solar energy and hydrogenase and/or nitrogenase enzymes. Microalgae use solar energy to transfer electrons to NADPH and ferredoxin to generate hydrogen. Photolysis can be classified into several subcategories: direct photolysis involves splitting water into hydrogen and oxygen using sunlight energy as follows: Hydrogen is produced in two steps. First, photosystem II adsorbs light and generates electrons that are transferred to ferredoxin using light energy, which is absorbed by the photosystem. Hydrogenase then accepts the electron from ferredoxin to generate hydrogen. Indirect photolysis is well suited for nitrogenase-based systems. The reaction is as follows: 12H 2 O + 6CO 2 + light energy → C 6 H 12 O 6 + O 2 (7) C 6 H 12 O 6 + 12H 2 O + light energy → 12H 2 + 6CO 2 (8) However, hydrogen production from a nitrogenase-based system is 1000-fold lower than that from a hydrogenasebased system (Bahadar and Khan, 2013). Nitrogenase is prohibited by oxygen; thus, oxygen must be removed from these systems. A CO 2 concentration between 4 and 18% is optimal, resulting in higher cell densities and more hydrogen production (Das and Veziroglu, 2008). Chlamydomonas reinhardtii, Calothrix sp., Oscillatoria sp., Synechococcus sp. and Gloeobacter sp. can produce hydrogen via both hydrogenase and nitrogenase (Abed et al., 2009;Hallenbeck, 2009). Photosynthetic hydrogen production is the best for converting solar energy into hydrogen; therefore, the current focus of research is to increase the light-use efficiency and to design better reactors for hydrogen production (Bahadar and Khan, 2013).

Biodiesel
The extracted lipid from microalgae is further converted into biodiesel. Biodiesel is a promising renewable energy that does not require engine modification and reduces CO (by 50%) and CO 2 (by 78%) emissions (Chisti, 2007). A main problem of microalgae-based biofuels is their high viscosity. Such high-viscosity fuels are difficult to combust and leave deposits on the fuel injector of diesel engines. Thus, biodiesel is usually blended with conventional diesel. There are several techniques to overcome this problem, including pyrolysis, micro-emulsification, dilution and transesterification (Marchetti et al., 2007). Among these techniques, transesterification offers the most promise for lowering the viscosity.

1) Transesterification
The chemistry of biodiesel synthesis is relatively simple and occurs via transesterification, in which a triglyceride (from the lipid) reacts with a mono-alcohol (most commonly ethanol or methanol) in the presence of a catalyst as follows: The reaction converts three moles of alcohol and one mole of triglyceride into one mole of glycerol and three moles of methyl esters. Although this reaction requires much alcohol, the methyl ester yield is greater than 98% (Chisti, 2007). A homogeneous base catalyst (e.g., KOH or NaOH) is usually used to accelerate the reaction. However, the base catalyst can react with free fatty acids in microalgae lipids to form soap, leading to a lower biodiesel yield and increasing the difficulty of separating biodiesel from glycerol. Acid catalysts (e.g., H 2 SO 4 ) are alternative options because a catalyst is not sensitive to the free fatty acid level in oil; therefore, esterification (in which free fatty acid is converted into alkyl ester) and transesterification can occur simultaneously (Lam and Lee, 2012). However, base-catalyzed reactions are 4000 times faster than acidcatalyzed ones (Bahadar and Khan, 2013). Commercially, alkoxides of sodium and potassium are used at 1% per weight of lipid formed because they have a greater catalytic activity than simple alcohols. Base catalyst reactions are optimized at 60°C under atmospheric pressure for 90 min. At higher temperatures and pressures, although the reaction is faster, the operating cost is greater. The reaction generates two layers of excess methanol and oil. The biodiesel is then separated from contaminants, for instance, glycerol and solids, in a flask separator (Hossain et al., 2008). The biodiesel may be then washed with water to remove contaminated free fatty acids, which cause saponification. Transesterification consumes 4.3 MJ/L biodiesel (Demirbas, 2007).
Heterogeneous catalysts (base or acid) have also been explored extensively for transesterification to produce biodiesel. Unlike homogeneous catalysts, the heterogeneous catalyst can be recycled, regenerated and reused for a subsequent transesterification reaction. Furthermore, this catalyst can be easily separated through filtration after the reaction is finished, which can minimize the product contamination and the number of water-washing cycles for purification. Until now, the application of a heterogeneous catalyst in microalgae biodiesel production is still restricted because it is a relatively new feedstock and is not commercially available in the market. Therefore, additional breakthrough findings are required to address the feasibility of heterogeneous catalysts in the microalgae biodiesel industry (Lam and Lee, 2012). CaO supported by Al 2 O 3 was tested for very a short time in the transesterification of lipid from Nannochloropsis oculata. The obtained yield was 97.5% at a reaction temperature of 50°C, with a methanol-to-lipid ratio of 30:1, a catalyst loading of 2% and a reaction time of 4 h (Umdu et al., 2009). Fu et al. (2013) synthesized a sulfonated carbon-based solid acid catalyst through the in situ partial carbonization and sulfonation of microalgae residues with sulfuric acid. The obtained catalyst possessing a high acid density (5.3 mmol/g) was used in the transesterification of triolein with methanol for biodiesel production. The yield of methyl ester was 24% for a 12-h reaction time. Chlorella vulgaris ESP-31 containing 22.7% lipid was used as the oil source for biodiesel production via transesterification catalyzed by SrO/SiO 2 . The solid catalyst SrO/SiO 2 worked well with water-removed, centrifuged Chlorella vulgaris ESP-31 and a biodiesel conversion of 80% (Tran et al., 2013).

2) In Situ Transesterification
In situ transesterification is a simple process that permits extraction and transesterification to occur in one step, in which the lipid-bearing biomass is directly in contact with the chemical solvent in the presence of a catalyst. There are two roles of the chemical solvent, including (1) a solvent for the extraction of lipid from biomass and (2) a reactant for the transesterification reaction. The advantages of in situ transesterification over conventional transesterification include a minimized solvent separation step, a reduced processing time and, consequently, a reduced overall biodiesel production cost (Shuit et al., 2010).
Many studies have investigated the in situ transesterification of microalgae. Chlorella sp. was in situ transesterified under a reaction temperature of 60°C, with methanol to lipid ratio of 315:1, a H 2 SO 4 concentration of 0.04 mol and a reaction time of 4 h to produce a 90% biodiesel yield (Ehimen et al., 2010). As a result, high methanol consumption was required. However, methanol consumption can be reduced by adding a co-solvent during in situ transesterification. Chlorella pyrenoidosa with an oil content of 56.2% was subjected to in situ transesterification, and a 95% biodiesel yield was attained with a hexane (cosolvent) to lipid molar ratio of 76:1, a methanol to lipid molar ratio of 165:1, a reaction temperature of 90°C, 0.5 M H 2 SO 4 and a reaction time of 2 h (Miao et al., 2011). Other solvents, such as toluene, chloroform and dichloromethane, were also suggested for use during in situ transesterification (Xu and Mi, 2011). However, using wet paste, microalgae biomass has an adverse effect on in situ transesterification; thus, the wet microalgae biomass must be initially dried to ensure efficient and optimum performance during in situ transesterification. Recently, the microalgae biomass with a water content greater than 31.7% caused the inhibition of in situ transesterification, leading to negligible biodiesel conversion (Ehimen et al., 2010). Sathish et al. (2013) reported that microalgal biomass with a moisture content greater than 20% by mass led to statistically significant reductions in biodiesel discovery. Three explanations have been commonly presented to describe this inhibitory effect: (1) the formation of free fatty acid methyl esters is a reversible reaction, and therefore, water can hydrolyze biodiesel back to methanol and free fatty acids; (2) water contained within the biomass can shield lipids from the extracting solvent, preventing lipids from being brought into the reaction; and (3) the acid catalyst can be deactivated due to water competing for available protons in the reaction. To avoid adverse effects, water must be removed from the recovered methanol or ethanol before the methanol or ethanol is used for subsequent in situ transesterification; otherwise, the subsequent reaction might not proceed under optimum conditions due to the presence of water. Therefore, extensive biomass drying is also required prior to reaction to avoid the unrequired side reaction and to simplify the subsequent separation processes (Lam and Lee, 2012). Im et al. (2014) proposed the method for the in situ transesterification of wet Nannochloropsis oceanica (65% moisture content) using chloroform, methanol and sulfuric acid in a one-pot reaction. The addition of chloroform can improve the performance of transesterification with methanol by reducing the diffusion limit of the reactant to the liquid reaction phase. This study provided a yield of greater than 90% without the addition of extra energy inputs for drying microalgae prior to in situ transesterification. Thus, this method can reduce the production cost of biodiesel from microalgae.

3) Supercritical Fluid Transesterification
Supercritical methanol transesterification is now being tested for the production of biodiesel. In the supercritical state, depending on pressure and temperature, intermolecular hydrogen bonding in the methanol molecule will be significantly decreased. As a result, the polarity and dielectric constant of methanol are reduced, allowing it to act as a free monomer. Subsequently, methanol under supercritical conditions can solvate the non-polar triglycerides to form a single-phase lipid/methanol mixture and yield fatty acid methyl esters and diglycerides (Saka and Kusdiana, 2001). These advantages make the supercritical methanol process more beneficial than conventional transesterification methods, in which homogeneous and heterogeneous catalyst processes often face difficulties due to the free fatty acid content and water content in the feedstock, resulting in lower yields and requiring purification of the final product to meet the biodiesel standards (Tan et al., 2011). Patil et al. (2011 used a one-step process for the direct liquefaction and conversion of wet microalgal Nannochloropsis sp. biomass containing approximately 90% water to biodiesel under supercritical methanol conditions. This one-step process enables the simultaneous extraction and transesterification of wet microalgal biomass. Optimal conditions for this process are reported as follows: wet algae to methanol (w/v) ratio of approximately 1:9 and a reaction temperature and a time of approximately 255°C and 25 min, respectively. A more than 84% yield of fatty acid methyl esters was obtained. Moreover, the direct conversion of microalgal Nannochloropsis sp. biomass into biodiesel using supercritical methanol compared with microwave-assisted transesterification methods was investigated by Patil et al. (2012). Wet algal biomass was used as feedstock in the supercritical methanol process, and dry algal biomass was used for the microwave-assisted transesterification. A higher yield of fatty acid methyl esters was obtained using the supercritical methanol process.
A concurrent extraction and transesterification of lipids can also be achieved under supercritical ethanol conditions. Similar to methanol, the dielectric constant of ethanol decreases into the range of organic solvents with increasing temperature, and the exothermic hydrogen bonding in ethanol shifts toward the free monomer above critical conditions, permitting the extraction of lipids from wet microalgae and ethanol to perform the transesterification of triglycerides to fatty acid ethyl esters. Recently, supercritical ethanol was used for the simultaneous extraction and transesterification of lipids in algae to produce fatty acid ethyl esters (Reddy et al., 2014). A maximum yield of approximately 67% of fatty acid ethyl esters was obtained at 265°C with 20 min of reaction time and a 1:9 dry algae to ethanol (w/v) ratio. Furthermore, the obtained fatty acid ethyl esters possessed high oxidative stability and a calorific value of 43 MJ/kg, comparable to that of regular diesel fuel. Significant energy savings were possible by eliminating the microalgae drying step through the simultaneous extraction and conversion of wet algae to biodiesel. Thus, these green conversion processes have the potential to provide an energy-efficient and economical route for the production of renewable biodiesel.

Biomethanol
Microalgae can be used to produce biomethanol as a renewable fuel. Methanol was synthesized theoretically by the following reaction: To convert hydrocarbons into methanol, the reforming of hydrocarbons in the syngas is carried out by a water-gas shift reaction, indicated by Eq. (10). Then, to achieve the reaction indicated by Eq. (11), the ratio of H 2 /CO is adjusted to the ideal value of 2 by the next step of watergas shift reaction indicated by Eq. (12): Spirulina sp. is converted to methanol by gasification (Hirano et al., 1998). The Spirulina biomass was partially oxidized at temperatures of up to 1000°C to produce H 2 , CO, CO 2 and hydrocarbon. The maximum yield was 0.64 g of methanol per gram of microalgal biomass at 1000°C. The carbon dioxide produced by the reaction was removed by absorption using alkanolamine.

Bioethanol
Generally, two methods are normally modified for the production of bioethanol from biomass. The first method is a biochemical process, i.e., fermentation, and the other method is a thermo-chemical process or gasification. The recent attempts at producing ethanol have focused on microalgae as a feedstock for the fermentation process. Microalgae are rich in carbohydrates (Table 1), which are entrapped within their cell wall; an economical physical pretreatment process, such as extrusion or mechanical shear, is required to break down the cell wall so that the carbohydrates can be released and used as carbon sources for fermentation. Bacteria, yeast and fungi are used to ferment carbohydrates to produce ethanol under anaerobic conditions. The products include not only ethanol but also CO 2 and H 2 O, as demonstrated in the following equation: According to this equation, the stoichiometric yields are 0.51 kg of ethanol and 0.49 kg of CO 2 per kg of carbon sugar, i.e., glucose.
The most recent studies of bioethanol production by fermentation have been reported. Chlorella vulgaris and Chlorococcum sp. are widely used for bioethanol production because of their high starch contents. Harun et al. (2010) used Saccharomyces bayanus for the fermentation of Chlorococcum sp. with 3.83 g/L ethanol produced from 10 g/L lipid-extracted microalgae. Chlorella vulgaris cake was enzymatically hydrolyzed and fermented with Saccharomyces cerevisiae to produce the reducing sugars and ethanol. The yields obtained for reducing sugars and ethanol were 0.55 and 0.17 g/g of cake, respectively (Moncada et al., 2013). Schizochytrium sp. was separated into sugars (mainly Dglucose and L-galactose), lipids and proteins. The separated sugars were then converted to ethanol by Escherichia coli KO11, resulting in 11.8 g of ethanol/L being produced from 25.7 g/L of glucose (Kim et al., 2012).
The bioethanol fermentation from microalgae involves less intake of energy, and the process is much simpler compared to the biodiesel production system. In addition, the undesired CO 2 byproduct can be recycled as a carbon source to cultivate additional microalgae, resulting in the reduction of greenhouse gas emissions. However, the commercial production of bioethanol from microalgae is still being investigated (Singh and Gu, 2010;Bahadar and Khan, 2013).

Biobutanol
Biobutanol has been identified as a potential fuel from renewable sources. Comparing with ethanol, butanol provides several benefits as a biofuel such as higher energy contents, lower vapor pressure and lower hygroscopy. Moreover, it can be partially or totally blended with gasoline or diesels fuels (Ventura and Jahng, 2013). In general, a fermentation of sugars with Clostridium sp. can produce acetone, butanol and ethanol (ABE) in the ratio of 3:6:1 acetone:butanol:ethanol, which can be called ABE fermentation (Potts et al., 2012). ABE fermentation is typically featured by anaerobic bacterial metabolisms, acidogenesis and solventogenesis. The acetic acid and butyric acid are the primary acid produced during acidogenesis. Clostridia species absorb these acids to produce ABE. In contrast to the ABE fermentation, Ramey (2006) proposed to accomplish the fermentation in two steps to improve yield, in which sugar was converted to butyric acid in the first step by bacterium such as Clostridium tryobutyricum in acidogenesis phase, and butyric acid to butanol in the second step by solventogenesis bacterium such as Clostridium beijerinckii Ellis et al., 2012;Potts et al., 2012). However, in order to convert butyric acid to butanol, additional energy in the form of glucose is required, adversely affecting the selection of twostep fermentation methods. Microalgae are carbohydratesrich biomass which can be regarded as a source for butanol production. Ellis et al. (2012) used Clostridium saccharoperbutylacetonium N1-4 for ABE fermentation of acid-pretreated wastewater microalgae with 2.26 g/L butanol produced from 8.92 g/L total soluble sugar. Acidpretreated microalga Scenedesmus obliquus was fermented by Clostridium saccharoperbutylacetonium to produce ABE. The obtained butanol yield was 1.91 g/L (Anthony et al., 2013). An isolated Chlorella vulgaris was used in butanol fermentation with a Clostridium acetobutylicum via ABE fermentation. 3.37 g/L butanol was produced from 111 g of acid-pretreated biomass of Chlorella vulgaris (Wang et al., 2014).

Bio-Oil and Biohydrocarbon
Hydrothermal liquefaction can be an alternative way to produce bio-oil from microalgae through the aqueousconversion method, in which freshly harvested wet microalgae biomass are directly processed without drying. During hydrothermal liquefaction, water is heated to a subcritical condition between 200 and 350°C under pressurization to reduce its dielectric constant. The dielectric constant can even drop to a value similar to that of ethanol; therefore, sub-critical water is able to solubilize a less polar compound (Duan and Savage, 2011;Kumar et al., 2011). In other words, water at the sub-critical condition can serve as an effective solvent with significantly less corrosion than other chemical solvents. In an experiment, 43% yield of bio-oil was obtained by hydrothermal liquefaction of Nannochloropsis sp. (initial water content of 79%) at 350°C and 35 MPa for 60 min. Major bio-oil constituents included phenol and its alkylated derivatives, heterocyclic N-containing compounds, long-chain fatty acids, alkanes and alkenes, and derivatives of phytol and cholesterol. CO 2 was always the most abundant gas product. However, the bio-oil required deoxygenation and denitrogenation to N and O-containing compounds to upgrade the bio-oil (Brown et al., 2010).
Alkanes were also produced by hydrogenation and decarbonylation of microalgae oil. The reaction was catalyzed by a Ni/ZrO 2 to produce ketone intermediates, which were then hydrogenated to aldehydes in the presence of Ni catalysts. The reaction was performed in an autoclave. Propane (3.6%) and methane (4.6%) were the main products in the vapor phase, which were formed by the hydrogenolysis of triglyceride and the methanation of CO/CO 2 with H 2 , respectively (Fig. 3). The total liquid yield contained a 70% yield of n-heptadecane and a 75% yield of total liquid alkanes (Peng et al., 2012a). The alkanes in the range of C15 to C18 were synthesized through hydrodeoxygenation of microalgae oil in the presence of 10% Ni on Hbeta zeolite (Ni/Hbeta, Si/Al = 180) in an autoclave at 260°C in the presence of hydrogen at 4 MPa. The obtained products yielded 78% liquid alkanes (60% C18, octadecane), 3.6% propane and 0.6% methane. A Ni/Hbeta catalyst increased the hydrogenation rate and produced propane and fatty acids from saturated triglyceride (Peng et al., 2012b).
A methane fermentation technology can be applied to microalgae to produce methane. The gross stoichiometry of the methane fermentation from glucose as a substrate can be approximated by the following (Klass, 1984): The methane from anaerobic digestion by anaerobic microorganisms can be used as fuel gas and be converted to generate electricity (Holm-Nielsen et al., 2009). The residual biomass from anaerobic digestion is further processed to make fertilizers. Due to the absence of lignin and lower cellulose, microalgae exhibits good process stability and high conversion efficiency for anaerobic digestion (Singh and Gu, 2010). Frigon et al. (2013) screened microalgal strains for the production of methane, which was obtained from anaerobic digestion of microalgae. The strains Scenedesmus sp., Isochrysis sp. and Scenedesmus dimorphus displayed the best methane yield, with 410, 408 and 397 mL of CH 4 /g of total volatile solid, respectively (Table 8). Thermal pretreatment using an autoclave at 120°C for 40 min Fig. 3. Main reaction pathways for microalgae oil transformation to alkanes over Ni/ZrO 2 catalyst (Peng et al., 2012a). was conducted to treat Chlorella vulgaris and Scenedesmus sp. to enhance carbohydrate solubilization, resulting in the enhancement of anaerobic digestibility and 21-50% methane production compared with untreated microalgae (Mendez et al., 2014). Alzate et al. (2014) reported that the anaerobic digestion of lipid-extracted Nannochloropsis sp. exhibited higher CH 4 production rates than its non-extracted counterpart. Moreover, thermal pretreatment supported a CH 4 productivity enhancement of 40% for the nonextracted Nannochloropsis sp. and 15% for the lipidextracted Nannochloropsis sp. As a result, it appears that the pretreatment and lipid removal are important processes to enhance the yield of methane production.
As indicated above, microalgae are a promising source for biofuel production. They are also used for the generation of other products. The cost of producing microalgae-based biofuel can be compensated for by earnings received from other co-products of the microalgal biomass. The integrated production of microalgae-based biofuels can recover energy consumed in bioenergy production processes, meaning that the final products in the form of biodiesel, ethanol and methane can compensate for their energy input or loss (Zhu, 2014). As a result, a hybrid microalgae refinery is more profitable compare to only a product-based or energy-based biorefinery. The challenge lies in the identification of suitable species and in minimizing the costs of production (Singh and Gu, 2010). The use of microalgae presents numerous routes to the integration of raw materials, processes and products to create a hybrid biorefinery, as presented in Fig. 4.

STRATEGIES FOR COMMERCIALIZATION OF MICROALGAE-BASED BIOFUELS
There are five key strategies for successful microalgae biofuel commercialization, including (1) faster, (2) fatter, (3) cheaper, (4) easier and (5) fractionation marketing approaches to help producers reduce costs and accelerate the commercialization of microalgae-based biofuels (Thurmond, 2009). First, faster is a primary strategy for most microalgaebased biofuel producers, who search for microalgae species possessing a high oil content and a quick growth rate for biofuel production. Microalgae with a high oil content, such as Botryococcus braunii, grow slowly and can be harvested only a few times a week, while algae with a lower oil content, such as Dunaliella sp. or Nannochloropsis sp. (20-40% oil content), will grow more quickly and can be harvested daily or a few times a day. For this reason, most algae research and development projects and pre-commercial projects consume algal strains with 20-40% oil content. Second, for fatter, microalgae producers are especially interested in using microalgal species with a high triglyceride oil content for biofuel production, resulting in a significant reduction of capital and operating costs and savings for systems twice their size using species with a lower oil ontent. As a result, signification innovation and improvement are the challenges for microalgae producers to lower costs to enter biofuel markets. Next, in view of economic processes for the commercialization of microalgae-based biofuel and ease of implementation, cheaper and easier processes are the next two important strategies. The estimated costs to produce microalgae oils and microalgae biodiesel range from $9 to $25 per gallon in ponds and from $15 to $40 in photobioreactors. Because several sub-sets of systems (i.e., harvesting, drying, pretreatment, extraction, etc.) are required, a reduction in the number of steps in microalgae-based biofuel production is necessary to provide easier, better and lower cost systems. A crucial economic challenge for microalgae producers is to discover low cost oil extraction and harvesting methods. Extraction systems have estimates of up to $15 per gallon of oil produced, depending on the extraction method, which can be less than cost effective (Singh and Gu, 2010). For example, the combination of harvesting, pretreatment and extraction system technology into a single process can reduce system complexity and costs for microalgae producers. Finally, the co-production of some more valuable fractions and their marketing is also important to success. This biomass fraction of microalgae contains valuable proteins, chemicals or molecular compounds that can be used to produce animal feeds, green plastics, detergents, cleaners and biodegradable polymers and can be sold at a premium price over traditional petroleum-based products. These biomass co-product marketing strategies will be crucial to the success of biofuel producers. As a result, a hybrid biofuel refinery concept (Fig. 4) can be implemented profitably for microalgae-based biofuels. CO 2 and nutrients can be recycled for microalgae cultivation and thus help in carbon sequestration. The biofuel part and other valuable products can be co-generated to make the commercialization process a profitable venture.
Currently, a commercial-scale microalgae-based biofuel industry is still not available due to the high cost deriving from start-up and running operations, maintenance and management (Zhu and Ketola, 2012). However, this cost might be reduced if microalgal productivity and/or lipid concentration is substantial. This industry might become more attractive if substantial subsidies or tax breaks are made available. Thus, the future of microalgae-based biofuels looks bright and promising. Additionally, a certain amount of research and innovation is still demanded, which can include research and development programs focusing on solutions for future energy concerns and on making economically practical biofuels.

CONCLUSIONS
Microalgae have recently received growing attention due to their potential for CO 2 capture and utilization in renewable energy. The use of microalgae exhibits a number of advantages over the use of other plant feedstock, including (1) high photosynthetic conversion, (2) rapid production, (3) high capacity to produce a wide variety of biofuel feedstock, (4) high capability for environmental bioremediation, such as CO 2 fixation from the atmosphere/flue gas and water purification, and (5) non-competitiveness for land with crops and food. In addition, the net CO 2 emission is assumed to be essentially zero if CO 2 released from the microalgaebased biofuel can be recycled and reused for the cultivation of microalgae. As a result, these advantages and potentials make microalgae suitable candidates to solve CO 2 reduction and energy issues.
For the cultivation of microalgae, the cultivation conditions, including water media at the adequate pH and temperature, nutrients and CO 2 dosed in a controlled manner in the presence of sunlight, are a concern. The cultivation in closed photobioreactor systems is more promising than that in open ponds or raceways for meeting the need of biofuel industries.
Pretreatment, especially cell disruption, prior to solvent extraction is necessary to effectively recover the lipid. However, it should be noted that some of the cell disruption methods require a large energy input that may lead to a negative energy balance. For extraction, the selected technology should be efficient (both in terms of time and energy), non-reactive with the lipid, relatively cheap (both in terms of capital cost and operating cost) and safe. Notably, the lipid extraction yield is influenced by the techniques and operating conditions of cell disruption and extraction, which is also related to the microalgae strain. As a result, combination techniques, instead of a single method, are proposed by researchers.
The biofuels obtained from microalgae are not only biodiesel but also biohydrogen, biomethanol, bioethanol, biobutanol and biomethane. These products are conventionally produced using transesterification, fermentations, liquefaction, etc. A hybrid biofuel refinery concept, rather than an individually product-based or energy-based biorefinery, can be implemented profitably for microalgae-based biofuels through recycling of CO 2 and nutrients. The biofuel part and other valuable products can be co-generated to make the commercialization process a profitable venture.