According to conventional knowledge, in this anaerobic food web ~30% of the methane is generated via hydrogen (with CO2 as the electron acceptor) by hydrogenotrophic methano — gens and ~70% is generated via acetate by acetoclastic methanogens. The latter process is referred to as acetate cleavage (Smith and Mah 1980), and occurs under close-to-optimum environmental conditions without reactor upsets (Mechanism I in Table 4.1). When the envi­ronmental conditions are not as favorable (e. g., stressed conditions), an alternative mechanism exists to convert acetate into methane, reducing the percentage of methane that is generated from acetate through acetate cleavage. This is referred to as acetate oxidation as part of a two-step process in which acetate is first oxidized to H2 and CO2 by homoacetogenic bacteria (in reverse) and H2/CO2 are then converted to CH4 by hydrogenotrophic methanogens (Mechanism II in Table 4.1; Barker 1936; Zinder and Koch 1984; Zinder 1994).

The first step ofbacterial acetate oxidation has a positive standard free energy (Mechanisms II[a] in Table 4.1), resulting in a thermodynamically unfavorable reaction at biological stan­dard conditions. At homoacetogenic nonequilibrium cases with the presence of methanogens in the anaerobic food web, the reaction sum becomes thermodynamically favorable provided

that an extremely low hydrogen partial pressure is maintained by the second step (Mechanisms II[b] in Table 4.1). Therefore, the relationship between an acetate-oxidizing bacteria and a hydrogenotrophic methanogen is a syntrophic relationship. Even though thermodynamically Mechanism II becomes as favorable as Mechanism I, the overall energy yield has to be shared between the two syntrophic members. Syntrophic acetate oxidation associations have been observed in a variety of anaerobic systems (Petersen and Ahring 1991; Schnhrer et al. 1994, 1999; Shigematsu et al. 2004; Schwarz et al. 2007). In such systems, acetate-oxidizing bacteria play an extremely important role guaranteeing the proper functioning of the anaero­bic digestion by preventing acetate accumulation. This alternative pathway is, therefore, important to maintain stability in engineered systems.

The major bottleneck in the production of soluble sugars from cellulosic biomass involves the high cost of the enzymes necessary for this process. In theory, the solution should be simple: either to find cheaper ways of producing the enzymes or to find more active enzymes!

Actually, to improve the catalytic efficiency of cellulases may pose a formidable task, since these enzymes are, arguably, already of the most efficient in nature. Even if we were to improve the efficiency or thermal tolerance of one enzyme (acting alone), the question remains whether this improved enzyme will now work better in a synergistic mixture with the other enzymes. An exception to this consideration seems to be the case of the cellobiohy — drolase I and II enzymes (GH7 and GH6, respectively) from fungi. In the free cellulose system, these enzymes are the star performers producing the majority of the soluble sugars from cellulose, aided primarily by only a small addition of endoglucanase activity.

Nevertheless, current strategies focus on the identification of new and improved enzymes, with the hope that those of highest activities will work best together. Since the improvements in producing industrial quantities of enzymes are more of a technical, engineering feat, we will concentrate here on strategies designed for improving enzyme activities or combinations thereof rather than improvements of processes for their production. In order to assemble improved enzyme systems for biomass conversion, a series of methodologies is required (Figure 5.4) . New enzymes are identified by mining established enzyme databases, newly sequenced microbial genomes, and/or relevant cellulosic microenvironments. The newly identified enzymes are evaluated by genomic, proteomic, or metabolic profiling, and their activities are assessed. The properties of these enzymes can then presumably be improved by rational mutagenesis or directed evolution, leading to improved enzyme cocktails for enhanced conversion of cellulosic biomass to soluble sugars.

The traditional way of handling biomass transport cost is to consider a constant cost compo­nent and a variable cost component (Kumar and Sokhansanj 2007) for the transport equip-

ment. For truck transport, the constant cost component is the cost of loading and unloading. The variable cost component is the “per km and per t” cost of trucking, accounting for fuel, depreciation, maintenance, and labor. The constant cost in case of rail transport includes the capital cost of rail siding, rail cars, and equipment for loading and unloading biomass. The variable cost includes the charges of the rail company that include capital recovery and maintenance for track and engines and fuel and operating costs. Table 7.6 summarizes the cost of transporting biomass using three modes of transport: truck, rail, and pipeline. The cost equation for pipeline was developed based on the data of Kumar et al. (2004).

Figure 7.7 compares the cost of transporting biomass using three modes of transport. For pipeline the annual capacity is assumed 1 million dry t. In this model, the transport cost in $/t for truck and rail does not change with capacity (in real situation, the size of contracts with transport companies affects the prices). Pipeline has the steepest cost curve because of the increased capital cost with distance.

Truck and rail costs intersect at about 110 km for the cost figures used in this analysis. It should be mentioned that the cost structures for rail are much more complicated than what is given in this analysis. In cases where a multi-mode transport is required, the cost structures will be a blend of two or three of these modes. At this point we would like to caution against over-generalization of equations in Table 7.6 and graphs in Figure 7.7. The cost of trucking,

rail, and even pipeline much depends upon available infrastructure, custom rates, road travel regulations and size of contracts.

Table 7.6 also lists estimates for energy consumption by truck, rail, and pipeline. The energy input for truck and for rail is 1.3 and 0.68MJ/t/km, respectively (Borjesson 1996; Kumar et al. 2006). The energy input for rail transport is 0.68MJ/t/km. It is assumed that diesel fuel is used for both truck and rail. The electrical power is assumed to be produced from a coal power plant; we assumed an electricity price of $0.06/kWh to convert from the cost ($) to energy (MJ) consumption for the pipeline.

Table 7.7 lists the minimum and maximum costs involved in biomass collection, prepro­cessing (pelleting), and transport. The delivered cost varies from a minimum of $46/t to more than $78/t. This cost does not include payment to farmers that might be around $10/t. The total energy input to the system ranges from a low of 1 to 1.5GJ/t. This amount of energy input is roughly from 6% to 10% of the total energy content of biomass (estimated at 16GJ/t).

Agricultural residues are only available for a limited time frame each year, but processing facilities must be run around the clock for virtually the entire year to have any hope of real­izing reasonable returns on the huge investments required. This requirement is particularly critical for production of commodity products as margins are thin. Thus, a vital consideration is how to ensure a year-round supply, and only two options are apparent: (1) store enough material during the harvest season to last the entire year or (2) have access to a number of feedstocks that can be harvested year round. The latter may be a viable strategy but is very regional in nature. Thus, our focus will be to discuss considerations in storing agricultural residues.

Given the low margins and high capital costs, it is highly desirable to store agricultural residues in the simplest and least costly manner possible, and simply piling up biomass is about as simple a way as there is. For example, sugarcane bagasse has been stored in piles that are properly maintained to prevent moisture accumulation. Tractor trailer rigs drove on to large piles of bagasse at a furfural plant in Belle Glade, Florida over a period of about 30 years to dump their loads, and bulldozers then spread the material around so the surface slopped downward. Driving large equipment on the piles compacted the bagasse, thereby making it nearly impenetrable to water and also difficult for the pile to catch fire. Slopping the piles not only allowed trucks and bulldozers to travel on the pile but ensured that water would run off, thereby keeping the moisture levels reasonably constant at about 50%. The result was little degradation of the hemicellulose or cellulose in bagasse.

In a systematic study, samples taken from various depths in a pile of sugarcane bagasse over a period of 3-26 weeks were analyzed by analytical pyrolysis (Agblevor et al. 1994). It was found that samples taken from the center of the piles after 3.25, 6.5, 13, and 26 weeks of storage had very small changes in the distribution of hemicellulose, cellulose, and lignin, while hemicellulose and cellulose contents dropped while lignin increased for samples taken from the regions near the outer surface layer. They also found that pentosans were degraded more rapidly than hexosans in the outer regions in the pile where microbial activity was high, consistent with experience that hemicellulose is more readily decomposed than cellulose which is in turn more susceptible to attack than lignin. Thus, these results demonstrate that deterioration only occurred in a very shallow layer near the surface of the piles but not in the bulk of the material, consistent with industrial experience with large piles. If such an approach would work as well for corn stover and other agricultural residues, it would greatly simplify storage.

It has been shown that wrapped bales of corn stover can be stored for extended times with little degradation, consistent with experience with sugarcane bagasse. All bales were stored in the open, with high moisture bales and uncovered square bales processed first while the normal moisture round bales, protected from the weather by their plastic wrap, could be processed later. Open-field storage of wrapped round bales is common practice for straw, hay, and similar materials, with open fields of such bales a common sight in Vermont and New Hampshire, as well as Nebraska, Iowa, and other parts of the Midwest. Corncobs have been compacted for long-term storage for furfural production over more than a year, similar to the approach taken for bagasse described above. However, it is unclear if the lower bulk packing density of corn stover and other agricultural residues will allow this approach.

Another consideration is whether to store biomass near the source or near the conversion facility. The former simplifies gathering operations and avoids moving material that could degrade during storage. However, it is simpler to manage a few piles of biomass than to try to ensure numerous piles are properly cared for to minimize deterioration.

Pretreatment of lignocellulosic biomass using dilute sulfuric acid treatment is carried out at high temperature (121°C for 1 hour). During this process some of the sugars released from biomass react and form chemicals that are inhibitory for cell growth or fermentation, or both. Pichia stipitis, a natural pentose-fermenting yeast, is inhibited by compounds produced during pretreatment and hydrolysis of lignocellulosic biomass (Slininger et al. 2009). Examples of some of these inhibitors include furfural, hydroxymethyl furfural, acids (acetic, ferulic, gluc­uronic, vanillic, syringic, and p-coumaric), and other chemicals such as vanillin and syrin — galdehyde (Tran and Chambers 1985; Grohmann and Bothast 1997; Ezeji et al. 2007a). To be able to ferment toxic hydrolysates to ethanol or butanol a number of detoxifying methods exist such as treatment with Ca(OH)2 (also called overliming), use of XAD resins (Sigma Chemicals, St. Louis, MO), or use of cultures that can metabolize the inhibitors. Studies on developing such inhibitor-metabolizing strains (Coniochaeta ligniaria) have been successful (Nichols et al. 2008). However, inhibitor utilization has to be carried out before the produc­tion of ethanol when using hexose — and pentose-fermenting strains. Another alternate could be the development of cultures that can tolerate and utilize inhibitors and still produce ethanol or butanol. In an interesting approach, acidic and alkaline-electrolyzed water was used to hydrolyze agricultural biomass and produce biofuel without any significant inhibitory effects on the microbial cultures (Wang et al. 2009). However, a much lower sugar yield was obtained when employing these approaches than when using dilute H2SO4.

For the agricultural wastes that are characterized with high protein and/or urea levels, ammonia toxicity to the bacterial and methanogenic populations in the undefined mixed cultures in anaerobic digestions can hamper reactor performance. (Ammonium is the final digestion product for nitrogen-containing polymers, such as proteins.) Ammonia toxicity in anaerobic digestion, therefore, has been studied in depth over the last 30 years (Velsen 1979; Braun et al. 1981; De Baere et al. 1984; Hashimoto 1986; Koster and Koomen 1988; Koster and Lettinga 1988; Bhattacharya and Parkin 1989; Robbins et al. 1989; Poggi-Varaldo et al. 1991- Troyer 1997- Lay et al. 1998- Lu et al. 2008- . Some of these papers documented maximum concentrations of total ammonium-N (NH — and NH-+) and free ammonia (NH3), which is the inhibiting species, for pure cultures of methanogenic populations (Sprott and Patel 1986; Koster and Koomen 1988; Steinhaus et al. 2007). The general consensus in the literature seems to indicate that hydrogenotrophic methanogens are less sensitive to high free ammonia concentrations compared with acetoclastic methanogens (Sprott and Patel 1986; Angelidaki and Ahring 1993). It is, therefore, not surprising that under high ammonia con­centration an alternative pathway that includes acetate oxidation by bacteria and subsequent hydrogen utilization by methanogens emerges in farm-based digesters without the need for acetoclastic methanogens (Schnhrer et al. 1994, 1996; Angenent et al. 2002b). Angenent et al. (2002b) showed that the relative 16S rRNA levels for hydrogenotrophic methanogens of the order Methanomicrobiales increased from 2.3% to 7% with increasing total ammonia-N concentrations of 2000 to 3600 mg/L, whereas the relative 16S rRNA levels for acetoclastic methanogens in the genus Methanosarcina decreased from 3.8% to 1.2%. Others have observed that acetoclastic methanogens from the family Methanosarcinaceae tend to group themselves in a cluster shape when a relatively high total ammonium-N concentration is present to protect themselves from high free ammonia concentrations (Calli et al. 2005a, b). Thus, several mechanisms to sustain an efficient microbial community function under high free ammonia concentrations may be present, resulting in a diverse methanogenic population even when the total ammonium or free ammonia concentration in the farm-based digester has exceeded the maximum levels reported in the literature.

In a recent study, we have operated four identical ASBR systems-fed swine waste at a VS concentration of 20 g/L for close to a 1000-day operating period (Garcia and Angenent 2009). Several operating changes, including ammonia concentration and temperature increases, were made to study the effect on the interactions between temperature and ammonia inhibition on reactor performance. During period 1 (day 0-378), the methane yield was 0.31L CH4/g VS for all digesters (with no statistical differences among them) at a temperature and total ammonium-N levels of 25°C and ~1200mg ammonium-N/L, respectively. During period 2 (day 379-745), the methane yield at 25°C decreased by 45% when total ammonium-N and free ammonia-N were increased in two of the four ASBRs to levels >4000 mg/L and >80 mg/L, respectively. During period 3 (day 746-988), this relative inhibition was reduced from 45% to 13% compared with the low-ammonia control reactors when the operating temperature was increased from 25°C to 35°C (while the free ammonia levels increased from ~100 to ~250 mg ammonia-N/L because of the temperature increase—chemical equilibrium change). The 10°C increase in temperature doubled the kinetic constant for methanogenesis, which overwhelmed the elevated toxicity effects caused by the increasing concentration of free ammonia, resulting in better reactor performances. This was an unexpected result (for the authors) because the literature predicts that methanogenesis would become severely inhibited with free ammonia concentrations of ~250mg-N/L at a higher (35°C) operating temperature (Koster and Lettinga 1984; Sprott and Patel 1986; Koster and Koomen 1988). Based on the chemical equilibrium change, we had predicted pronounced methanogenic inhibition at tem­peratures higher than 25°C. However, based on a long-term study with methanogenic food webs, it is clear that the farmer/operator may alleviate ammonia toxicity by increasing the operating temperature within the mesophilic range. Even a temperature increase from 30- 35°C to 38-39°C may increase digester performance at elevated ammonia-N concentrations (Garcia and Angenent 2009 ).

Hydrogen is produced during fermentative metabolism of glycerol via conversion of pyruvate to formate, a reaction catalyzed by the enzyme pyruvate formate lyase (PFL). The formate is converted to carbon dioxide and hydrogen by the action of formate hydrogen lyase (FHL). This is probably essential for maintaining both the CO2 supply needed for cell growth (Dharmadi et al. 2006) and the production of a proton motive force (PMF) that is essential for cell viability (Bagramyan and Trchounian 2003; Hakobyan et al. 2005). Under certain environmental conditions, engineered E. coli was found to produce hydrogen at the rate of 4.64mmol/L/h with a specific yield of 0.96 mol H-/mol glycerol (Yazdani and Gonzalez 2008) . Biohydrogen production from glycerol containing waste generated from a biodiesel plant was studied using a natural isolate E. aerogenes. The rate of hydrogen production was determined to be 63 mmol/L/h with a specific yield of 0.85 mol H—mol glycerol (Ito et al. 2005; Nishio and Nakashimada 2007- . Production of biohydrogen by E. aerogenes using crude glycerol as substrate affords biodiesel industries the opportunity to add value to crude glycerol and potentially increase their revenue base.

Harvesting

Cotton is harvested by removing the fiber from the other plant parts. Two different harvester designs are used. These machines are representative of a machine that collects material, sepa­rates the desired portion, and drops the rest back onto the ground in the field. Grain combines, of course, have the same functionality.

The spindle harvester has a picking head with a series of cone — shaped spindles with rows of serrated ridges parallel to the axis. These spindles rotate and wind the cotton fiber around the spindle, thus pulling it from the bole. The picking head rotates as the harvester moves forward and doffers sweep the cotton from the rows of spindles. The cotton is then pneumatically conveyed to the storage basket on the harvester.

The stripper harvester has a mechanism to aggressively comb through the standing cotton stalks and collect the fiber. It tends to collect more plant parts (pieces of stem, leaf, and outer hull of the bole) than the spindle harvester.

Raw cotton collected from the field includes the seed, therefore it is referred to as “seed cotton.” A cotton harvester is actually a mobile solid-solid separator technology. The seed cotton (a solid) is separated from the other plant parts (solids).

Infield hauling is done with side-dump wagons sometimes referred to as “bole buggies.” When the harvester basket is full, the harvester stops, the side-dump wagon pulls alongside, and the basket is dumped (Figure 7.12). The wagon then proceeds to a location where the module builder is parked, typically close to a public road to provide ready access for the module hauler.

The wagons cycle continuously between the harvester and module builder until the harvest of a particular field is complete. Harvesting cost ($/t) is lowest when harvester wait time is minimized. Ideally, the module-building location is chosen to minimize cycle time of the wagons. However, in reality, it must be chosen such that road trucks can get into position to pick up the modules without getting stuck, thus there is a trade — off between wagon cycle efficiency and road truck access.

I f biomass were grown for energy to an amount equal to that consumed during their any given production period, there would be no net buildup of COo in the atmosphere (Gao and Mickinley 19940. Microalgae are particularly promising biomass species because of the high growth rate and high CO2 fixation ability compared to plants (Tsukahara and Sawayama 2005 ).

The earliest idea focused on producing methane gas from microalgae. The concept of producing fuel by using microalgae as a source was reported by Meier (1955). Golueke and Oswald (1959) presented the concept of using microalgae as a substrate for anaerobic diges­tion, and the reuse of the digester effluent as a source of nutrients. They realized these con­cepts by using a large pond (40 pa) to grow microalgae, then the microalgae was digested to methane gas for producing electricity. The gas production by the digester averaged about 10 ft3 per lb of volatile matter introduced. The methane content of the gas varied from 68% to 74%. The maximal efficiency attained by the algal culture was 3%, whereas the maximal overall efficiency of the entire conversion unit was approximately 2% (Golueke and Oswald 1959). From the 1970s, the National Science Foundation-Research Applied to National Needs Program (NSF-RANN) started to support laboratory studies of microalgae fermentations to methane gas (Uziel et al. 1975). Six microalgae species were studied, approximately 60% of microalgae biomass energy could be converted to methane gas. It was found that the rate of biogenic methane gas production by the marine strain methanogenic bacteria at 50% wet algal thalli amendment was greater by 33.4% in comparison with results of the freshwater cattle manure strain methanogenic bacteria under similar experimental conditions. The pro­portion of methane gas content in this biofuel gas was 58%, while the remaining gases are CO2 (major portion), H2S, NH3, N2, and O2 (Silvalingam 1982).

Many microalgae, in particular species classified as “green algae,” produce hydrogen after a period of anaerobic conditions in the dark, during which the hydrogenase enzyme was activated and synthesized, and small amounts of hydrogen production were observed (Das and Veziro lu 2001). Green algae are probably better for hydrogen production than cyano­bacteria (blue-green algae) whereas the latter uses more energyiintensive enzymes, ATPi requiring nitrogenase for the production of H2 (Lee and Greenbaum 1997).

Methane production from microalgae became the basis and motivation of the U. S. DOE’s program to develop renewable transportation fuels from microalgae, which started in 1978 and ended in 1998. This program mainly focused on the production of biodiesel from high lipid-content microalgae grown in ponds, utilizing waste CO2 from coal fired power plants. Biodiesel is an alternative fuel produced from triglycerides and fatty acids present in naturally occurring fats and oils. Traditional oil crops such as corn, soybeans, canola, coconut, and oil palm cannot adequately contribute to replacing petroleum derived from liquid fuels in the foreseeable future due to their relatively low oil yield per hectare compared with microalgae. For example, 30% oil (by wt) of microalgae has an oil yield of 58,700 l/ha and 70% oil (by wt) of microalgae has an oil yield of 136,900 l/ha. By comparison, corn and soybeans have only an oil yield of 172 l/ha and 446 l/ha, respectively (Chisti 2007). Hu et al. (2008) reported that based upon the photosynthetic efficiency and growth potential of microalgae, theoretical calculations indicated that annual oil production of larger than 30,000 L or about 200 barrels of algal oil per hectare of land may be achievable in mass culture of oleaginous algae (algal species have been found to grow rapidly and produce substantial amounts of triacylglycerols or oil). This value was 100-fold greater than that of soybeans, a major feedstock currently being used for biodiesel in the United States.

Another unique benefit of using microalgae to produce biodiesel is that it will not com­promise production of food, fiber, and other products derived from crops. Xu et al. (2006) used n-hexane to extract large amounts of microalgal oil from Chlorella, which the crude lipid content is about 55.2%. Then the microalgal oil was transformed into biodiesel by acidic transesterification. The biodiesel was characterized by a high heating value of 41MJ/kg, a density of 0.864 kg/L, and a viscosity of 5.2 x 10-4Pa s at 40°C (Xu et al. 2006).

Pyrolysis with different solvents and re-agents were conducted on algae-protein (Goldman et al. 1980). The reactions yielded rather low conversions in the presence of water in spite of the existence of carbonates and catalysts, for example, nickel sulfate. The presence of benzene improves the yield and the presence of a mixture of K-Mg-Mn salts was beneficial for such a reaction. The nitrogen content of liquid oil decreased in the presence of carbonates and other catalysts. The maximum amount of protein converted into liquid oil was 27% by weight for algae-proteins containing 5.7 wt % nitrogen.

Lipid content in microalgae was considered as the most important component for yielding biofuel, such as biodiesel and other forms of oil. Peng et al. (2001a) studied the pyrolytic characteristics of Chlorella protothecoides. Chlorella protothecoides were pyrolyzed at the heating rates of 15, 40, 60, and 80°C/min up to 800°C. The pyrolysis reactions mainly took place between 160-520°C with a volatile yield of about 80%. The devolatilization stage consisted of two main temperature zones (I and II) with a transition at 300-320°C. The researcher found that crude lipid in cells decomposed at Zone II while other main components at Zone I, which might indicate that more energy input for lipid pyrolysis seems needed in comparison with other main components (Peng et al. 2001b). In another study, two kinds of high protein and lipid contents microalgae, Cyanobacterium Spirulina platensis (SP) and green alga C. protothecoides (CP) were pyrolyzed at the heating rates of 15, 40, 60, and 80°C/min up to 800°C in the thermogravimetric analyzer to investigate their pyrolytic char­acteristics. The results showed the value of activation energy for CP pyrolysis was lower than that of SP, and the char in final residue of CP was 2%-3% less than that of SP, which indi­cated CP was preferable for pyrolysis over SP (Peng et al. 2001).

Fast pyrolysis (a sweep gas [N2] flow rate of 0.4m3/h, and a vapor residence time of 2-3 s) was used to treat C. protothecoides to produce bio-oil. The highest yield of bio-oil is 57.9%, at an operating temperature of 450°C. This yield is 3.4 times higher than that from autotrophic cells also treated by fast pyrolysis. After reaction, the total liquid products were composed of an aqueous and an oil phase. The oil was fractioned using column liquid chromatography, and separated into n-hexane soluble and n-hexane insoluble compounds. The bio-oil was characterized by a much lower oxygen content, with a higher heating value (41MJ/kg), a lower density (0.92kg/L), and lower viscosity (0.02Pa • s) compared to those of bio — oil from autotrophic cells and wood. These properties are comparable to fossil oil (Miao and Wu 2004).

Algal lipid or even the whole algae could be pyrolized to a similar, high-octane, aromatic gasoline product slate when passed over HZSM — 5, a medium-pore, shaper/ selective, acid catalyst. The first of these results was reported by Milne and Evans (1987). The same authors carried out exploratory studies of the pyrolysis and zeolite conversion of whole algae and their major components. Four species of microalgae were pyrolized: Chaetoceros muelleri var. subsalsum, Monoraphidium minutum, Naviculus Saprophilla — and Nannocloropsis sp. However, the results of whole algae were ambiguous due to very high mineral matter content (10%-50%) and unknown proportions of water, lipids, and other organic components of the exact sample used (Milne et al. 1990).

One of the shortcomings of using pyrolysis to convert microalgae is due to the high moisture content of microalgae. The large amount of energy consumed to vaporize the water during the pyrolysis process was considered as a negative effect of this method (Minowa et al. 1995a. Dote et al. (1994) performed a liquefaction of Botryococcus braunii, with and without sodium carbonate as a catalyst. High-quality oil was obtained, which was more than the content of hydrocarbons in B. braunii (50 wt % db), in a yield of 57-64 wt % at 300°C. The oil was equivalent in quality to petroleum oil (Dote et al., 1994). The properties of the oil obtained from B. braunii were clarified by the same research group (Inoue et al. 1994). The oil was fractionated into three fractions by silica gel column chromatography and analyzed to deter­mine its composition. The yields of the three fractions based on organics were 5% of lower molecular weight hydrocarbons (MW = 197-281), 27.2% of botryococcenes (MW = 438­572), and 22.2% of polar substances (MW = 867-2209). The maximum recovery (78%) of botryococcenes in the liquefied oil was achieved at 200°C with the use of a catalyst.

Minowa et al. (1995b) used HTL to convert Dunaliella tertiolecta (78.4% moisture content) into oil at around 300°C and 10 MPa. The D. tertiolecta was cultured batchwise in an open tank of 10-L capacity at continuous light of 20,000 1ux, temperature of 27°C, and bubbling air with 3% CO2. Then the algal cells grown for HTL were harvested by a centrifugal sepa­rator. Some Na2CO3 (0-5 wt % of the dry solid in the algal cells) was used as additive. However, the results showed it had no catalytic effect on either the oil yield or its properties. Nitrogen was introduced to purge the residual air in the autoclave. To prevent water from vaporizing, additional nitrogen was added to 3 MPa. Several reaction temperatures were tried: 250, 300, and 340°C. The retention time (holding time) was from 5 minutes to 60 minutes.

The gas phase was primarily CO2 . Besides the gas phase, the reaction mixture consisted of a tar-like material and a water phase. The tar-like material floated on the surface of the water phase in all experiments and was easily separated. The oil was extracted from the reaction mixture by dichloromethane. Then the dichloromethane was evaporated from the extract at 35°C under reduced pressure, yielding a dark-brown viscous material, which was referred to as the oil. The oil was obtained in the range of 31%-44% (average 37%) on an organic basis. The oil yield exceeded the algal cells crude fat content (20.5%). The reaction param­eters, including reaction temperature, holding time, and sodium carbonate addition, had no significant effect on the oil yield. However, the author thought the properties of oil strongly depended on the reaction temperature. The viscosity decreased and the heating value increased slightly with a rise in temperature. The carbon and hydrogen content tended to increase with temperature increases. The oil obtained at a reaction temperature of 340°C and holding time of 60 minutes had a viscosity of 150-330mPa • s and a calorific value of 36MJ/kg. These values were comparable to those of No. 3 fuel oil in JIS (50-1000mPa • s, about 40MJ/kg). The results of the energy consumption ratio (ECR) indicated the liquefaction was a net energy producer (Minowa et al. 1995b).

Metal catalysts had been used in microalgae liquefaction. Matsui et al. (1997) investigated the liquefaction of Spirulina, a high-protein microalgae in various organic solvents or water under hydrogen, nitrogen, or carbon monoxide in the temperature range 300-425°C, using Fe(CO)5-S catalyst. Among the solvents of tetralin, 1-methylnaphthalene, toluene, and water, it seemed more favorable for liquefaction of Spirulina to take place in water. After reaction, tetrahydrofuran (THF) was used to extract the production. Then the THF-soluble fraction was further separated into hexane-insoluble and hexane — soluble fractions by precipitation into hexane. The hexane — soluble fraction was denoted as oil in this research. Liquefaction of Spirulina at 300-425°C under hydrogen gave more than 90 wt % conversion and 60 wt % oil yield. Addition of the Fe(CO)5-S catalyst increased the oil yield from 52.3 wt % to 66.9 wt % at 350°C for 60 minutes in tetralin. Liquefaction in water gave an oil yield as high as 78.3 wt % at 350°C even under nitrogen without a catalyst. Liquefaction in toluene gave oil fractions having a heating value of 32-33 MJ/kg, but products obtained in water, containing large amounts of oxygen, were estimated to have a lower heating value of 26 MJ/kg (Matsui et al. 1997).

The microalgae and coal were co-liquefied with the presence of coal liquefaction catalysts (Ikenaga et al. 2001- . Dried samples of Chlorella, Spirulina. and Littorale were used as microalgae and Australian Yallourn brown coal and Illinois No. 6 bituminous coal were employed. Commercial iron pentacarbonyl (Fe[CO]5), trisruthenium dodecacarbonyl (Ru3[CO]i2- , and molybdenum hexacarbonyl (Mo[CO]— were employed as catalysts. The co-liquefaction was carried out under pressurized H2 in 1-methylnaphthalene at 350-400°C for 60 minutes. Co-l iquefaction of Chlorella with Yallourn coal was successfully achieved with excess sulfur to iron (S/Fe = 4), The conversion and the yield of the hexane-soluble fraction were close to the values calculated from the additivity of the product yields of the respective homo reactions. All three catalysts were effective for the co-liquefaction of micro­algae with coal. Some 99.8% of conversion and 65.5% of hexane soluble fraction were obtained at 400°C with Fe(CO)5 at S/Fe = 4, when the 1:1 Chlorella and Yallourn coal were co-liquefied. In the co—iquefaction of Chlorella with Illinois No. 6 coal, the oil yield was close to the additivity of the respective reaction with Fe(CO)5-S, even at S/Fe = 2. Mo(CO)6 catalyst (S/Mo = 4) was the most effective for the respective homo-liquefactions of Chlorella and Yallourn coal. When Littorale and Spirulina were used as microalgae, a similar tendency was observed with the iron catalyst.

Microalgae culturing conditions and the fixation of CO2 during the microalgae growth were studied in order to find the optimal condition to harvest large amounts of algae. It was also useful to investigate the energy balance and CO2 mitigating effect of a liquid fuel production process from microalgae using HTL. Kishimoto et al. 2 1994) studied the CO2 fixation during the microalgae growth and oil production of microalgae by using thermo­chemical liquefaction. The green microalga D. tertiolecta AHTP 30929, which contained 10% glycerol, was used throughout the study. The effects of saline, light intensity, and CO2 concentration on microalgae growth were investigated. The details of the liquefaction were not mentioned in this paper, but the authors concluded that the heavy oil yield from the liquefaction was 35.6%. The contents of the chemical elements were as follows: carbon 73%, hydrogen 9%, nitrogen 5%, and oxygen 13%. The heating value of the heavy oil was 34.7 MJ/kg, which was almost the same as C heavy oil. The viscosity of the oil was 860 mPa • s, the same as that of castor oil. The nitrogen content was higher than that of ordinary petroleum.

Continuous culturing of the B. bruunii Berkeley strain in secondarily treated sewage (STS) was conducted and then liquefied by Sawayama et al. (1995) • B. bruunii grew continuously for a period of over 1 month at a growth rate of 200 mg dry weight per liter water per week and the algal cells fed on STS containing 49% oil as the hexane soluble fraction. Liquefaction of microalgae cells was conducted in a 300 mL autoclave. Wet cells (30g wet weight, 92% moisture content) were charged to the autoclave with 5 wt % or without sodium carbonate. N2 was used as the initial gas for preventing water from vaporizing. The autoclave was heated to 200, 300, and 340°C, with a retention time of 1 hour. Oil was extracted with dichloro — methane from the reaction mixture. The maximum yield of oil obtained by liquefaction was 64 wt % on a dry basis at 300°C with a sodium carbonate catalyst. The yield of the hexane soluble fraction was 97 wt % compared with that in the feedstock algal cells. The heating value of the liquefied oil obtained from this reaction was 49 MJ/kg and the viscosity was 64mPa s at 50°C.

Different microalgae are not the same in producing oil through liquefaction. The data of liquefaction of microalgae B. braunii and D. tertiolecta were collected and compared, and then the energy balance of the reactions was calculated by Sawayama et al. (1995). Liquefaction was performed using a stainless steel autoclave with 100 or 300 mL capacity using 0-5 wt % and Na2CO3 as a catalyst. After purging with nitrogen, the autoclave was charged with nitro­gen at 2-3 MPa. The reaction temperature was 300°C. The reaction mixture included a gas mainly composed of CO2, a tar-like material that sank to the bottom, a water phase, and an oil-like material that floated on the surface of the water phase. Solvent extraction with CH2C12 or acetone was performed to separate the oil from the reaction mixture. The yield of oil was determined as a percentage by weight of the organics in the original material. Based on the energy calculation and comparison, the yield of liquid fuel produced from B. braunii and its lower heating value were high compared with those of D. tertiolecta; therefore, the energy inputs for cultivation and separation of B. braunii were calculated to be smaller than those of D. tertiolecta. The energy input for fertilizers of B. braunii was also smaller than that of D. tertiolecta. Therefore, the hydrocarbon-rich microalga, B. braunii, could be more suitable for liquid fuel production using thermochemical liquefaction compared with D. tertiolecta. The authors also concluded: if a 100-MW thermal plant using coal would be replaced by liquid fuel produced from B. braunii, the quantity of CO2 mitigation could be 1.5 x 105t/yr and 8.4 x 103 ha of microalgal cultivation area could be necessary (Sawayama et al. 1999).

To find the most suitable operational conditions of HTL of microalgae, Yang et al. (2004) used the Microcystis viridis strain as feedstock. The series of experiments were conducted under the following various conditions: catalyst (Na2 CO3) loading rate of 0 and 5 wt %; reaction temperature of 300 and 340°C; and holding times of 30 and 60 minutes. The initial operational pressure was designed at 3 MPa of nitrogen, and the maximum pressure of the autoclave was 10-20MPa in order to decrease water evaporation. The reaction mixture was extracted with chloroform to recover the oil by evaporating chloroform at 40°C. The aqueous phase and insoluble mixture were separated by filtration. The chloroform insoluble fraction remaining on the filter paper was dried at 105°C for one day to obtain a solid residue. The evolved gas was transferred to a sampling bag and composition was determined by gas chromatography. The oil yield was defined as the ratio of the weight of oil products after liquefaction to the weight of organic matter in feedstock. The energy yield was defined as the ratio of the weight of C and H in oil products after liquefaction to the weight of C and H in feedstock. After investigating the effects of the reaction parameters, such as retention time, reaction temperature, and load of the catalyst, the maximum 33% oil yield (energy yield of around 40%) was obtained with the 30-minute holding time, reaction temperature of 340°C, and alkali catalyst dosage of 5 wt %. The heating value of the obtained oil was 31 MJ/kg, less than that of heavy oil (40 Mj/kg). The elemental composition of liquefied oil was composed of 62% carbon, 8% hydrogen, 8% nitrogen, and 2% sulfur. The liquefied oil contained n-alkane of C17-C18 hydrocarbon as a main component of the saturated compounds, so typical aromatic compounds of heavy oil, such as n-naphthalene and n-dibenzothiophene were found in liquefied oil, and it was considered that the liquefied oil should be classified as heavy oil. The gas consisted primarily of CO) and methane. The total nitrogen in the aqueous phase ranged from 998 to 1157mg/L, and half of the total nitrogen was detected as ammonia nitrogen. The total phosphate in the aqueous phase ranged from 2.47 to 5.38mg/l.

After microalgae liquefaction, other forms of biofuel—such as biodiesel—could be obtained by further extraction. Based on the efficiency of biodiesel extraction, Aresta et al. (2005) compared HTL of algae with the extraction using supercritical carbon dioxide (sc-CO2). Green alga Chaetomorpha linum was used as the feedstock. Almost 20 g of fresh, washed thalli of C. linum were placed into the glass reactor which was then put into the autoclave. The latter was closed under N2 atmosphere to purge the residual air and 3.0 MPa of N) . The autoclave was then heated to the desired temperature (250, 300, 350, and 395°C) for 1 hour. The reaction mixture was recovered and treated with CH2 Cl2. Then an organic and an aqueous/solid suspension were separated. From the organic solu­tion, after evaporation of the solvent under controlled conditions, an amber-yellow oily liquid was obtained. The aqueous solution was separated and a solid was recovered by centrifugation. For the sc-CO2 extraction, the thalli of alga were dried for 5-8 days at room temperature, 3-5 g of dried algae were milled in liquid nitrogen (5 mL) to break the cellular wall in order to increase the extraction yield. A 0.5-1 mL of methanol was used as a co-solvent to improve the efficiency of the extraction. In both liquefaction and sc-CO2 extraction, the oils obtained were treated to convert all fatty acid (FA) components into the mono-methyl-esters (Biodiesel). Then the extracted material was analyzed quantitatively by GC and qualitatively by GC-MS. For liquefaction, increasing the temperature for the total amount of oil extracted from the algae increased reaching a plateau between 350- 395°C (about 80 mg/g dry wt). Increasing the temperature affected the amount of fatty acid extracted. However, at higher temperatures, some oil decomposed, resulting in a decrease of the recovered amount of extracted fatty acid. After comparing the results from the sc — CO2 extraction, the authors implied that HTL seemed to be more efficient from the quantitative point of view.

Alkali pretreatment is a process in which alkaline solutions like NaOH or KOH are used to remove and solubilize lignin component of lignocellulosic biomass and efficiently increase the accessibility of enzyme to the cellulose component. Using alkali to treat lignocellulosic biomass has been known for years to improve cellulose digestibility. For example, lime has been used to pretreat poplar wood (150°C for 6 hours with 14-atm oxygen; Chang et al. 2001), switchgrass (100°C for 2 hours; Chang et al. 1997), wheat straw (85°C for 3 hours; Chang et al. 1998), corn stover (100°C for 13 hours; Karr and Holtzapple 1998, 2000), and sugarcane bagasse (ambient conditions up to 192 hours; Playne 1984). During dilute NaOH pretreatment of lignocellulosic biomass, the mechanisms of reaction are by solvation and saponification (Hendriks and Zeeman 2009). The intermolecular ester bonds cross—inking

xylan hemicellulose, other hemicellulose, and lignin (Figure 3.2a) are saponified. During solvation and saponification reactions, the lignocellulosic material is swollen, leading to an increase in surface area, decrease in crystallinity of material, disruption of lignin structure, and separation of structural linkages between lignin and carbohydrates (Fan et al. 1987; Sun and Cheng 2002) . Alkaline pretreatment of chopped rice straw with 2% NaOH and 20% solid loading at 85°C for 1 hour decreased the lignin by 36% and substantially increased enzymatic cellulose hydrolysis (Zhang and Cai 2008).