Production of isobutanol by fermentation is looking attractive owing to two main reasons: (1) the higher tolerance of producing microorganisms to isobutanol, usually 36.9—51.9 g/l as compared to n-butanol (called butanol in the later sections of this chapter), which is 20—30 g/l in selected hyper-producing strains and (2) having a lower boiling point (108 °C vs 118 °C) than butanol, which comparatively may be economical to recover. The above titer values for the isobutanol are without simultaneous product recovery (Baez et al., 2011). However, in this report, it was mentioned that in situ recovery by gas stripping improves isobutanol production. To the authors’ knowledge, this is the only report where isobutanol fermentation and recovery were performed simultaneously.

Fermentative production of isobutanol or butanol can be economically achieved in two ways: (1) by developing the high-tolerant or high-producer strain, which also offers some benefits during the recovery process, and (2) using energy-efficient process engi­neering techniques to simultaneously remove the toxic product. Interestingly, the first approach has been reported for isobutanol fermentation (Baez et al., 2011) with great success. For butanol producing strains, numerous attempts have been made to improve perfor­mance; however, success has been limited, with maximum titer stagnating around 21 g/l (total acetone butanol ethanol (ABE) 32.6 g/l) (Chen and Blaschek,

1999) . Nevertheless, butanol has drawn significant amount of attention from the process engineering point of view. One of the main focuses has been the develop­ment of integrated process technologies where fermen­tation and simultaneous product recovery have been integrated. The reader is directed to a couple of following reports, where much higher production of ABE than in a batch system has been achieved. Employing such inte­grated systems (gas stripping and perstraction), cumula­tive ABE production from 461 g/l to 698 g/l (Ezeji et al., 2013; Jeon and Lee, 1989) has been achieved compared to 21 g/l butanol or 51.9 g/l isobutanol. It should also be noted that several simultaneous product recovery systems such as adsorption, liquid—liquid extraction, pervaporation, ionic liquid extraction, and reverse osmosis have been investigated (Qureshi et al., 2013). Also, other advances have been made for butanol pro­duction from agricultural residues such as wheat straw (Qureshi et al., 2007; Qureshi et al., 2008a), barley straw (Qureshi et al., 2010a), corn stover (Parekh et al., 1988; Qureshi et al., 2010b), switchgrass (Qureshi et al., 2010b), and distillers dry grains and soluble (Ezeji and Blaschek, 2008).

CONCLUSION AND
FUTURE PERSPECTIVE

Given the higher microbial tolerance of isobutanol and its greater volatility in comparison to butanol, it is likely that simultaneous product recovery using gas stripping, perstraction, and/or pervaporation would achieve even higher production levels than reported for butanol, thus benefiting economics of isobutanol production process. To make this biofuel even more attractive, recent advances in fermentation of lignocellu — losic biomass such as separate hydrolysis of ligno — cellulosic biomass, combined with fermentation, and product recovery separate hydrolysis, fermentation, and recovery (SHFR), and simultaneous saccharification, fermentation, and recovery (SSFR) should be applied (Qureshi et al., 2013). In a process where wheat straw was used to produce butanol by SSFR, 192 g/l ABE was produced from 430 g/l of lignocellulosic sugars (Qureshi et al., 2008b). At this stage, engineering producing micro­organisms by applying similar method already reported by Higashide et al. (2011) to utilize pentose sugars such as arabinose and xylose as substrates for the production of isobutanol is looking promising.

Owing to the fact that butanol producing cultures have the potential to be tolerant to isobutanol, it is reasonable to express isobutanol-producing genes in sol — ventogenic Clostridium species. Such an undertaking would have two advantages: (1) ability of the developed strain to utilize pentose sugars that accounts for about 30—40% of carbohydrates present in lignocellulosic biomass and (2) the developed strain may produce higher titers of isobutanol than yeast or E. coli.

Though only a limited amount of work has been done to date, it is clear that hydrothermal catalytic conversion of algae can produce hydrocarbons for liquid biofuels. Thus, there is tremendous potential for this field and the outlook is bright. The majority of the work to date on producing liquid fuels from hydrothermal conver­sion of aquatic biomass has focused on homogeneous catalysis by metal salts or alkali. More recent studies, however, are beginning to examine heterogeneous cata­lysts due to advantages in separation and selectivity of the catalyst. More work is needed to identify better het­erogeneous catalysts for these applications. In particular, the development of nonprecious metal-based catalysts would provide a major advance.

CONCLUSIONS

Microalgae are a promising source of clean, renew­able biofuel. Not only can it be grown and produced on a large scale, it can be grown in virtually every part of the world including locations that are considered to be otherwise unsuitable to agricultural production and thus lie dormant. However, several challenges remain to its full execution: (1) the successful production of
feedstock on a large scale; (2) the development of pro­cessing methods that are cost-effective and leave intact the desired molecular end-products; and (3) a richer un­derstanding of microalgal chemistry and product accu­mulation during both growth and processing phases. Whether employing open ponds or PBRs for biomass production, cultures must be carefully monitored to maintain the desired composition of the culture. Factors such as nutrient loads, mixing and light source, and con­taminants all drive the production of biomass and thus biofuel precursors. There is a growing trend toward pro­cessing microalgae directly from the aqueous stream, eliminating costly drying steps and conserving water. As such, HTL is an emerging process that converts biomass to biocrude in hot, compressed water, thereby eliminating the need for drying or organic solvents. Further, all organic components serve as the feedstock for the HTL process rather than discreet components, such as lipids for biodiesel or ethanol for bioethanol. Biohydrogen is another provocative fuel derived from microalgae. Whatever the feedstock and biofuel process, additional improvements to each of the technologies are required to make the production of renewable fuels from microalgae cost-effective. These improvements can only result from systems using real-time analytical feedback to inform growth and processing and from innovations derived from a multidisciplinary approach.

Family-Size Biogas Plant

Straws and most other biomass have a tendency to float on the water and form a scum layer. Tests in India have shown that these types of materials will generate biogas when they have been in contact with digester fluid for at least three days. This can be achieved by filling the digester with a quantity of wet compost. The compost will float. New material is added from the bottom of the digester. This material will push the compost upward but will stay under the digester liquid, as the compost being wet is relatively heavy. The result­ing digested material forms a thick mat. The digested material can be removed once per year after opening the plant (Oosterkamp, 2003).

A plant suitable to convert maize stalks and other biomass into biogas is of the water jacket floating drum plant type. The water jacket reduces the emission of methane. In fixed dome plants 10% of the methane production is lost. The main modification is the use of a large 0.3 m diameter inlet pipe. Through this pipe the waste biomass can be pushed down into the digester proper. The floating drum can be removed and digested material taken out from the top.

Wet Digesters

Most digester systems in Europe are manure based (wet) and/or use maize silage for their feed. Some of these wet systems add about 10% straws (VS). Two or more sequentially linked digesters give about 15% more gas yield than a single continuously stirred digester (Angelidaki et al., 2005). The wet systems have the disad­vantage that solids (10—15%) need to be kept in suspen­sion by an impeller. With a high percentage of grasses and straw scum layers are formed that need to be removed mechanically after opening the digester.

In addition to xylan, hardwoods contain 2—5% of a GM, which is composed of b-glucopyranose and b-mannopyranose units linked by b-(1 / 4) bonds (Table 17.2). However, the mannose/glucose monomer ratio may vary depending on the original source of GM. The ratio of glucose to mannose varies between 1:2 and 1:1. Galactose is not present in hardwood mannan. The mannosic bonds between the mannose units are more rapidly hydrolyzed by acid than the cor­responding glycosidic bonds, and GM is easily depoly — merized under acidic conditions. There may be certain short side branches at the C3 position of the mannoses and acetyl groups randomly present at the C6 position of a sugar unit. The acetyl groups frequently range from 1 per 9 to 1 per 20 sugar units (Peng et al., 2012).

Xyloglucans

Besides xylan and GM, xyloglucans (XGs) are also present in the primary cell walls of some higher plants (mainly in hardwoods, and less in softwoods) (Peng et al., 2012). They can also appear in small amounts (2—5%) in grasses. XGs consist of b-1,4-linked D-glucose (cellulosic) backbone with 75% of these residues substituted at O-6 with D-xylose. L-Arabinose and D-galactose residues can be attached to the xylose resi­dues forming di — or triglycosyl side chains. Also L-fucose has been detected attached to galactose residues. In addition, XGs can contain O-linked acetyl groups. XGs interact with cellulose microfibrils by the formation of hydrogen bonds, thus contributing to the structural integrity of the cellulose network (Girio et al., 2010).

Galactoglucomannans

The major hemicelluloses in softwoods (gymno — sperms) are acetylated GGMs, accounting up to 20—25% of their dry mass (Girio et al., 2010). GGMs consist of a linear backbone of b-D-glucopyranosyl and b-D-mannopyranosyl units, linked by b-(1,4) glycosidic bonds, partially acetylated at C2 or C3 and substituted by a-D-galactopyranosyl units attached to glucose and mannose by a-(1,6) bonds. Acetyl groups content of GGM is around 6%, corresponding, on average, to one acetyl group per three to four hexose units (Girio et al.,

2010) (Table 17.2). Some GGMs are water soluble, presenting in that case higher galactose content than the insoluble GGMs. There are two main types of acetylgalactoglucomannans in softwoods, one being galactose-poor (5—8% of dry wood) and the other galactose-rich (10—15% of dry wood). The ratio of galactose:glucose:mannose is approximately 0.1:1:3 and 1:1:3, for the two woods, respectively (Peng et al.,

2012) . GGMs have an approximate DP between 100 and 150, which is equivalent to a molecular weight around 16,000—24,000 Da. GGMs are easily depolymer — ized by acids, especially the bonds between galactose and the main chain. The acetyl groups are much more easily cleaved by alkali and acid (Peng et al., 2012). GMs occur in minor amounts in the secondary wall of hardwoods (<5% of the dry wood mass) (Girio et al.,

2010) . As GGMs, they have a linear backbone of b-D-glucopyranosyl (Glcp) and b-D-mannopyranosyl (Manp) units but the ratio Glcp:Manp is lower. In GGMs and GMs the extent of galactosylation governs their association tendency to the cellulose microfibrils and hence their extractability from the cell wall matrix (Ebringerova et al., 2005).

In spite of the large number of studies dedicated to the utilization of technical lignins, the correlation be­tween lignin chemical structure and its properties and functions has not been established yet for most indus­trial applications. Lignin functional properties are most often correlated with physical properties, such as glass transition point (Tg), solubility and, sometimes, with molecular mass distribution (Glasser, 2000) as well as with such compositional features as the ash, carbohy­drate, sulfur, and carbon content. The effect of the chem­ical structure of lignins on lignin performance in specific applications is usually anticipated based on funda­mental knowledge rather than on experimentally estab­lished correlations. For example, the behavior of lignin in polyurethane production is correlated with the amount of hydroxyl groups predominantly. The utiliza­tion of lignin in phenol—formaldehyde (PF) resins re­quires the unsubstituted 5-position of the aromatic ring (o-position to the phenolic hydroxyl group) and therefore higher proportion of G-units is desirable, in contrast to S-units, which cannot participate in this reac­tion. The comparison of lignin reactivity is typically limited to the comparison of lignins originated from different technical processes and different feedstock or­igins (Tejado, 2007; Mansouri and Salvado, 2006; Evtu — guin et al. 1998; Rials and Glasser, 1986). An attempt to correlate lignin structure with its performance includes, for instance, the observation that the presence of ethyl groups in Alcell lignin act as an internal plasticizer and improve the lignin performance in poly(ethylene oxide) blends (Kubo and Kadla, 2004). Another attempt was the correlation of the quantity of aliphatic, phenolic hydroxyl groups and methoxyl groups as well as the lignin Mw with antioxidant lignin properties (Pan et al., 2006). However, the correlations between the amount of phenolic hydroxyls, Mn and the Radical Scav­enging Index observed in this study were rather poor or inexistent (R2 = 0.53 or lower) implying that the depen­dence is more complex and it requires comprehensive lignin structural elucidation.

In summary, it would appear that the main reason for the lack of clear structure—functional performance corre­lations is the high heterogeneity and variability of tech­nical lignins and the absence of widely accepted and understood, quantitative, fast, and simple analytical tech­niques (Glasser, 2000). In the past few years, leveraging all the recent advances made in the development of new lignin analytical techniques, a very comprehensive work on correlation between process parameters and the structure and properties of the produced technical lig­nins was undertaken at the R&D Department of Lignol Innovations, Ltd (Vancouver, Canada) (Balakshin, Berlin et al., 2013). Three categories of feedstock (softwoods, hardwoods, and annual fibers) including various biomass species sourced from different continents were processed under at least 30 different combinations of pro­cess conditions (time, temperature, ethanol concentra­tion, pH, and L:S ratio). The extracted lignins were analyzed using advanced rapid and comprehensive 13C high-resolution NMR spectroscopy coupled with a Cryo — Probe technology for lignin structural characterization, as described above, along with traditional methods for lignin composition analysis, Mw, thermal properties (Tg), antioxidant activity, and other properties generating results for over 50 different characteristics for each lignin sample. This effort generated a very comprehensive

database including several thousands of data points covering a wide diversity of OS lignins (Balakshin and Berlin, 2010). This unique database has been playing a very important role in developing Lignol’s lignin applica­tions helping in the selection of best lignin candidates for specific customer needs. Moreover, very accurate models correlating process parameters and lignin characteristics have been also developed on the basis of these studies (Balakshin, Berlin et al., 2013).

The bidirectional hydrogenases can either produce or consume H2 according to the cellular redox environ­ment. The enzyme functions in dark fermentation and under specific conditions in photoproduction of H2.

In cyanobacteria the bidirectional hydrogenase con­sists of two structural moieties: the hydrogenase (encoded by hoxYH) and the diaphorase unit (encoded by hoxUFE) capable of oxidation of NAD(P)H. Over the last several years, significant progress has been achieved in the identification of the transcription factors, such as LexA and AbrB-like proteins, which are members of the complex signal cascade that directs the expression of the bidirectional hydrogenase genes (Oliveira and Lindblad, 2009).

On the basis of high sequence similarity, it has been hy­pothesized that the diaphorase subunit of the bidirec­tional hydrogenase also serves as the three missing activity subunits of cyanobacterial respiratory NDH-1 complex (Appel and Schulz, 1996). However, more recent results have not supported this hypothesis since the mu­tants lacking the diaphorase subunits do not show mal­function of the respiratory activity (Boison et al., 1999).

Bidirectional hydrogenase has been found in all non-N2-fixing and some N2-fixing cyanobacteria. Thus, many filamentous N2-fixing cyanobacteria contain both the bidirectional and the uptake hydrogenase. However, a few species have only the uptake hydrogenase. In cyanobacteria, the bidirectional hydrogenase is constitu­tively expressed under both aerobic and anaerobic condi­tions but is active only in the dark, anoxic conditions or during the transition from dark to light (Cournac et al., 2004; Schutz et al., 2004). The biological function of bidi­rectional hydrogenase in filamentous cyanobacteria is not well understood (Tamagnini et al., 2007). Mutational studies with hox-defective mutants suggested that the bidirectional hydrogenase in N2-fixing cyanobacteria does not support N2 fixation (Masukawa et al., 2002). In non-N2-fixing cyanobacteria the bidirectional hydroge — nase is the main H2-producing enzyme and it is thought to interact with photosynthetic pathways (Ludwig et al.,

2006) . However, H2 production catalyzed by bidirectional hydrogenases is only transient (less than 30 s in light) since it is quickly inhibited by increasing photosynthetic O2 evolution (Cournac et al., 2004). In line with this, no transient H2 evolution was detected in different Hox deletion mutants studied by Aubert-Jousset et al. (2011). It is hypothesized that the bidirectional hydrogenase functions as a safety electron sink thereby removing excess reducing equivalents during the dark, anaerobic to light transition in unicellular Synechocystis cells (Appel et al., 2000; McIntosh et al., 2011). This hypothesis is inter­esting due to the natural environment of cyanobacteria being highly dynamic, with rapid fluctuations in light in­tensity. Such fluctuations might strongly unbalance the function of the photosynthetic complexes, resulting in production of reactive oxygen species and destroying photosynthetic apparatus. Cyanobacteria have unique flavodiiron proteins, Flv1 and Flv3, functioning as a strong electron sink at the end of light reactions by direct­ing excess electrons to O2 without production of reactive oxygen species, thus maintaining the redox balance of the electron transport chain (Helman et al., 2003; Allahver — diyeva et al., 2011, 2012). A bidirectional hydrogenase possibly takes over the role of a strong electron sink upon dark to light transitions during anaerobiosis, the condition created in cyanobacterial mats and blooms, and where the Flv1 and Flv3 pathway is not functional (Gutthann et al., 2007).

In Synechocystis, both NADPH and NADH can act as electron donors for the bidirectional hydrogenase. Recent studies showed that NADH is a preferential sub­strate of the diaphorase moiety, whereas NADPH is an efficient activator of the bidirectional hydrogenase (Aubert-Jousset et al., 2011). These results are in line with the observed dynamics of H2 production during dark—light transition. In the dark anaerobic conditions, H2 is produced by oxidation of NADH, the major prod­uct of glycolysis assimilation. Sudden exposure to light produces NADPH by photosynthetic electron transfer chain, which functions as an activator of the hydroge — nase and begins consumption of H2. Although there is no strong evidence for direct electron donation from reduced Fd (E0 = —0.42 V), which is a stronger reductant than NADH (E0 = —0.315 V), to the bidirectional hydrogenase, such an electron transfer cannot be completely excluded (McNeely et al., 2011). Direct linkage of Fd to the bidirectional enzyme in mutants lacking the diaphorase domain could be employed to improve cyanobacterial H2 production.

Accounting for the high affinity of the bidirectional hydrogenase to H2, it has been suggested that the enzyme can function in utilization of H2 under physio­logical conditions. However, it should be kept in mind that the bidirectional hydrogenase reversibly evolves H2 under dark, anaerobic conditions as a result of fermentation of photosynthetically stored carbon inter­mediates in cyanobacteria. Recent electrochemical investigations of the bidirectional enzyme from Synecho­cystis PCC 6803 have revealed unexpected properties. The rate of H2 production at low pH and low H2 pres­sure was shown to be about 1.4 times faster than the rate of H2 consumption at high pH and high H2 pressure (McIntosh et al., 2011).

Both edible (e. g. rapeseed, soybean or palm) and nonedible (e. g. Jatropha) oilseeds can be used for bio­diesel production. Biodiesel production is mainly achieved from soybean in the USA, rapeseed (or sun­flower in lower quantities) in Europe and palm oil in South-East Asian countries. Biodiesel production could be also achieved using nonconventional resources including microbial oil produced by yeast, fungi and algae (Meng et al., 2009). The continuous growth of bio­diesel production coincides with proportional produc­tion of by-products streams. The main by-product is glycerol that is generated during transesterification of triglycerides with predominantly methanol leading to the production of fatty acid methyl esters and glycerol (10%, w/w). It has been estimated that by 2021, the share of biodiesel production from vegetable oils will increase and the worldwide biodiesel production from such oils is projected to reach approximately 30 x 106 t (Anonymous, 2012c). This means that approximately 3 x 106 t of glycerol will be available for chemical and biopolymer production. Crude glycerol streams produced from biodiesel plants have purities in the range of 77—90% (w/w) (Mothes et al., 2007). The main impurities in crude glycerol are water, methanol, residual fatty acids and corresponding esters, and salts (NaCl or K2SO4) in varying proportions depending on the extent of glycerol purification (Mothes et al., 2007). Purification methods for glycerol have been proposed (Chatzifragkou and Papanikolaou, 2012) for the removal of impurities and salts after biodiesel production but they seem to be rather unprofitable, especially for small industries. Novel uses of glycerol involve both green chemical conversions and microbial bioconversions. Glycerol represents an easily assimilated carbon source for many microorganisms. Crude glycerol has been eval­uated as carbon source for various microbial bioconver­sions, such as 1,3-propanediol, citric acid, ethanol, succinic acid, propionic acid, microbial oil and PHAs (Koutinas et al., 2007a; da Silva et al., 2009; Koutinas and Papanikolaou, 2011; Sarris et al., 2011).

Biodiesel production from oilseeds leads to the pro­duction of oilseed meals as a valuable by-product stream. Oilseed meal is the protein — and carbohydrate — rich fraction that remains after the extraction of oil. The main conventional commercial outlet for oilseed meals is as animal feed. In the period 2012—2021, bio­diesel production from edible vegetable oils will still rely mainly on rapeseed and sunflower. However, bio­diesel production from palm oil is projected to increase twofold (Anonymous, 2012c). Based on recent estimates, approximately 315 x 106 t of oilseed meals are expected to be produced by 2021, corresponding to an increase up to 23% based on current production capacities (Anony­mous, 2012c). Future biodiesel industries could be con­verted into novel biorefineries through valorization of crude glycerol and oilseed meal streams leading to the production of biodiesel, chemicals, food and feed ingre­dients and biopolymers such as PHAs.

Ashby et al. (2004, 2011) evaluated the production and properties of PHAs accumulated by the bacterial strains Pseudomonas oleovorans NRRL B-14682 and Pseu­domonas corrugata 388 cultivated on crude glycerol. Ashby et al. (2011) reported that the molecular weight of PHB was decreased with increasing methanol concen­tration in crude glycerol. Mothes et al. (2007) and Garcia et al. (2013) evaluated the effect of NaCl and K2SO4 on PHA production during fermentation with the bacterial strains Paracoccus denitrificans, C. necator JMP 134 and C. necator DSMZ 545. These salts are present in crude glycerol depending on the catalyst (NaOH or KOH) employed during transesterification of triglycerides. The inhibition caused by NaCl on PHA production is more pronounced at significantly lower concentrations than K2SO4. Mothes et al. (2007) reported that bioreactor fermentations with C. necator JMP 134 cultivated on crude glycerol and inorganic chemicals as additional nu­trients could lead to the production of PHB contents up to 70% (w/w). Crude glycerol has also been employed in bioreactor fermentations for the production of PHB us­ing the bacterial strain C. necator DSM 545 leading to 50% (w/w) PHB content and 1.1 g/l h PHB productivity (Cavalheiro et al., 2009). Tanadchangsaeng and Yu (2012) stressed that the productivity (around 0.92 g/l) of glyc­erol fermentation to PHB synthesis is lower than the one achieved from glucose. Crude glycerol could be also combined with other carbon sources that could be used as precursors for the production of PHA co­polymers (Cavalheiro et al., 2012). The production of P(3HB-co-4HB-co-3HV) was reported when C. necator DSM 545 was cultivated on crude glycerol, propionic acid (stimulator of 4HB accumulation and 3HV precur­sor) and g-butyrolactone (4HB precursor). In all studies presented above, inorganic chemicals were used as nutrient supplements.

Apart from fermentation efficiency of PHA produc­tion, it is also crucial to assess the properties of the poly­mer produced and the associated production cost. Tanadchangsaeng and Yu (2012) reported that although the thermal and physical properties of the PHB pro­duced from glycerol is similar to the one produced from glucose, the molecular weight of the glycerol — derived homopolymer is lower than the molecular weight of the PHB produced from glucose. Posada et al. (2011) presented a comparative technoeconomic evaluation of PHB production from crude glycerol using two different bacterial strains, C. necator and B. megate — rium, and three different downstream separation strate­gies. Fermentation of C. necator resulted in the production of 81.6 g/l of which 57.1 g/l was PHB, significantly higher than B. megaterium. The most cost — competitive process involved PHB production in fed — batch fermentations with C. necator followed by PHB separation and purification with heat pretreatment, enzymatic alkaline digestion, centrifugation, washing, evaporation, and spray drying. Posada et al. (2011) re­ported also that glycerol purification to 98% (w/w) con­tributes approximately 6% of the overall PHB production cost, thereby slightly affecting the total cost. In this study, it was concluded that the PHB production cost from crude glycerol could be as low as US$2/kg depending on the downstream process utilized.

PHA production from crude glycerol could be com­bined with the valorization of oilseed meals or residues remaining after extraction of microbial oil. For instance, rapeseed meal could be utilized for the production of various value-added fractions including protein iso­lates, carbohydrates, hulls, phenolic compounds and glucosinolates with various applications such as animal feed, pesticidal agent, bioactive proteins, glues and ad­hesives, paper coatings and ingredients for cosmetics among others (Anonymous, 2011; Egues et al., 2010). Another alternative application of oilseed meals is based on the production of complex nutrient supplements for fermentation processes including PHA production. In this way, commercial inorganic chemicals will be replaced improving the sustainability of the whole bio­refinery concept. Oilseed meals contain significant quantities of protein, minerals and other necessary nu­trients for microbial growth. Enzymatic hydrolysis of protein to amino acids and peptides, and phytic acid to phosphorus could provide a hydrolysate suitable for PHA production. Crude enzymes could be produced via solid-state fermentation employing appropriate fungal strains and oilseed meals as substrates (Wang et al., 2010; Kachrimanidou et al. 2013). Wang et al.

(2010) reported the production of a nutrient-rich hydro­lysate from rapeseed meal with a free amino nitrogen content of 2016.2 mg/l and inorganic phosphorus (IP) of 304 mg/l that was subsequently used successfully as nutrient supplement combined with glucose as car­bon source for the cultivation of Saccharomyces cerevisiae. Garcia et al. (2013) investigated the generation of a mi­crobial feedstock through hydrolysis of rapeseed meal, which was combined with crude glycerol as the sole fermentation medium for PHA production. Fed-batch fermentations resulted in a production of 10.9 g/l P(3HB-co-3HV) without addition of any precursor. The properties of the biopolymer produced were also exam­ined, leading to the conclusion that this bioprocess could be incorporated in rapeseed-based biodiesel plants contributing to the sustainability of biodiesel bio­refineries. Kachrimanidou et al. (2013) reported the uti­lization of sunflower meal for the production of nutrient-rich hydrolysates that could be subsequently supplemented with crude glycerol for the production of 9.9 g/l P(3HB-co-3HV) with a PHA content of 50% (w/w) in shake-flask fermentations using the microbial strain C. necator DSM 545 without addition of any precursor.

Figure 24.2 presents a biorefinery concept in which sunflower (or other oilseed) meal is utilized only for the production of fermentation feedstock involving pro­duction of crude enzymes via solid-state fermentation followed by hydrolysate production via enzymatic hy­drolysis. Preliminary bioreactor fermentations carried out in fed-batch mode at the Agricultural University of Athens with the microbial strain C. necator DSM 7237 cultivated on sunflower meal hydrolysate and crude glycerol lead to the production of more than 20 g/l

PHB with a PHB content of approximately 70% (w/w). However, this processing scheme does not take advan­tage of the full potential of sunflower meal that contains value-added ingredients that could be isolated contrib­uting in the development of a more sustainable bio­refinery approach.

Figure 24.3 presents a sunflower-based biorefinery where besides fermentation feedstock, sunflower meal is also used for the production of an antioxidant-rich stream and a protein isolate product. The sunflower seed is covered by the hull that could be removed before oil separation by mechanical pressing and solvent extraction in biodiesel production processes. The protein content in sunflower meals can be increased via dehul — ling and complete oil extraction. The composition of

sunflower meal is variable and is highly dependent on cultivation conditions, sunflower variety and the indus­trial process used for biodiesel production. Dehulling or partial dehulling of sunflower seeds could provide a by­product that could be used for the production of energy, hemicelluloses, organic amendment for the soils, and biomaterial (Anonymous, 2011). The sunflower meal that remains as a by-product after (partial) dehulling and complete oil extraction could be fractionated in three different fractions (i. e. a protein-rich fraction, a lignocellulose-rich fraction and a liquid fraction) by a simple sedimentation/flotation process based on the formation of an aqueous suspension (Bautista et al., 1990; Parrado et al., 1991). This separation is based on the different densities of major components in sunflower

meal. Subsequently, antioxidants can be removed from the protein-rich fraction, as well as from the lignocellu — losic fraction. The most important of the phenolic com­pounds found in sunflower is chlorogenic acid.

The protein isolate extracted from the protein-rich fraction, after treatment with acid and alkaline solutions, can be utilized for the production of biopolymers and edible films. Yust et al. (2003) improved protein extrac­tion from sunflower meal through treatment with alka — lase. Remaining fractions can be used as substrate in solid-state fermentation with a fungal strain of Asper­gillus oryzae, an efficient producer mostly of proteolytic enzymes. The solids at the end of solid-state fermenta­tion can be used as enzyme-rich medium for hydrolysis of macromolecules contained in remaining sunflower. The liquid fraction from sunflower meal fractionation can be used as suspension liquid in enzymatic hydroly­sis, aiming at the generation of a nutrient-rich supple­ment. At the end of hydrolysis, remaining solids are separated from the hydrolysate by centrifugation, and can be possibly used for combustion to generate heat or as a carbohydrate-rich resource for the production of hydrolysates for other fermentations. The nutrient — rich supplement has been used successfully for enhanced production of PHB. The advanced biorefinery concept results in the production of three products (an­tioxidants, protein isolate and PHB) from the same raw material presenting a high potential of improved pro­cess economics.

Fermentation, the third step of bioconversion, con­verts the hydrolysates, mainly glucose, xylose, arabinose

and mannose to bioethanol using microorganisms. In addition to bioethanol, fermentation can be used to generate other useful end products, e. g. biobutanol, fatty acids, lactic acid, bioplastics or other biochemicals. The hydrolysates are often detoxified before fermentation due to the production of inhibitory compounds, such as phenolics and furan derivatives, in the pretreatment and hydrolysis steps (Dashtban et al., 2009; Ong, 2004). Saccharomyces cerevisiae is the most commonly used microorganism as it has a high fermentation rate and the application of recombinant DNA techniques has enabled the bioengineering of strains, capable of converting arabinose and xylose, as well as glucose, to bioethanol (Dashtban et al., 2009). This allows utilization of a larger amount of the hydrolysates, thus giving a higher percentage yield of bioethanol.

Even though the process of wood decay by many common white — and rot fungi has been well character­ized, other types of decay caused by soft-rot fungi or bacteria are still not well understood (Blanchette et al., 2002, 2004). Soft rot is caused by fungi taxonomically classified in the phylum Ascomycota, including related asexual taxa. The term soft rot is used because it was first identified from soft, decayed wood surfaces in contact with excessive moisture (Findlay, 1984). Soft rot can also occur in dry environments (Blanchette, 2000) and seems to predominate in extreme environments such as excessively wet or dry sites, where white — and brown-rot fungi growth is inhibited, and in substrates that do not favor the growth and development of other types of fungi (Blanchette, 1995; Blanchette et al.,

2004) . Soft-rot fungi attack the lignocellulose matrix in wood by formation of cavities (type I) or cell wall erosion (type II). Cellulases and hemicellulases, but not ligninases, are involved in soft-rot attack leading to extensive loss of the carbohydrate polymers; high amounts of lignin remain even in advanced stages of soft rot (Blanchette, 1995; Eriksson et al., 1990; Nilsson et al, 1989). The most studied and applied soft-rot fungus, Trichoderma reesei, and its mutants, are mainly used for large-scale commercial production of cellulases and hemicellulases (Durand et al., 1988; Esterbauer et al., 1991; Tomme et al., 1988).

Bacteria

Bacteria degrade plant cell walls through three main morphological forms: tunneling, erosion, and cavitation (Blanchette, 1995; Daniel et al., 1987; Singh and Butcher, 1991, 1985; Singh et al., 1990). An early study has confirmed that the Gram-positive filamentous bacte­rium Streptomyces viridosporus degrades softwood lignin into low molecular weight fragments (Crawford et al., 1982). Furthermore, enzymes similar to the fungal sys­tem such as peroxidases, ligninases and manganese per­oxidases have been implicated in bacterial biomass delignification (Glenn and Gold, 1983; Kirk et al., 1986). Interestingly, some bacteria can attack high lignin-containing hard wood that is considered durable and resistant to fungal decay (Nilsson et al., 1992; Singh and Butcher, 1991). However, compared to fungi, bacte­ria are not as efficient for lignocellulosic biomass pre­treatment, as shown by a recent work comparing eight microorganisms including fungi and bacteria, for pre­treatment of sugarcane waste (Singh et al., 2008).

How the electrons released by organic carbon oxida­tion in the bacterial anaerobic cytoplasm are transferred by the biofilm to the anode surface is an important factor in MFC performance (Neto et al., 2010). Major advances

have been made between 2000 and 2010 in understand­ing the electron transfer mechanisms by electrogens. There are two primary mechanisms: one is the direct electron transfer (DET) and the other, mediated electron transfer (MET).

DET for Anodic Biofilms

DET occurs via a direct physical contact between the microbial cell wall and the anode surface, or via a pilus that links the two. Gene expression studies (Holmes et al., 2008) and electrochemical analysis (Busalmen et al., 2008) have demonstrated that there are active sites for cytochrome proteins on the outer cell surface (Franks and Nevin, 2010; Zhou et al., 2012). When the microbes contact the anode surface, the cytochromes can transfer the electrons from the inside of the microbial cell wall to the outer cell wall and then to the anode surface (Rinaldi et al., 2008) (Figure 9.2(a)). Shewanella putrefaciens (Kim et al., 2002), R. ferrireducens (Chaudhuri and Lovley, 2003) and G. sulfurreducens (Bond and Lovley, 2003) use such cytochromes to achieve electron transfer. One major disadvantage for this mechanism is that only a monolayer of sessile cells in a biofilm can transfer elec­trons to the anode surface. This explains why the power and current densities of MFCs relying on this kind of DET are lower, sometimes by several orders of magni­tude, than that of MFCs with MET (Schroder, 2007) because MET can utilize more than one monolayer.

Recently, some researchers have observed that some microbial strains (such as Shewanella oneidensis and G. sulfurreducens; Logan and Regan, 2006; Torres et al.,

2010) can produce pili (conductive nanowires) to form physical conductive connections between the cell wall and the anode surface while the microbial cell wall is at a short distance from the anode (Reguera et al., 2005; Rinaldi et al., 2008). An extensive pilus network would allow several layers of sessile cells to donate elec­trons to the anode, thus multiplying the MFC power output. Summers et al. found a whole cell aggregate

(b)

FIGURE 9.2 Different DET methods: (a) direct cell wall-electrode contact (b) conductive pili or filament linkage. (For color version of this figure, the reader is referred to the online version of this book.)
consisting of G. sulfurreducens and Geobacter metalliredu — cens is conductive when the coculture was grown on ethanol. The c-type cytochrome OmcS of G. sulfurredu — cens was suspected to play a key role in accepting elec­trons from G. metallireducens (Summers et al., 2010). This overcomes the inability of G. sulfurreducens to use H2 for interspecies electron transfer. The ability of inter­species electron exchange suggests that a nonelectro­genic species in a synergistic biofilm consortium may contribute to electricity generation as long as its elec­trons can be taken up by an electrogenic species through interspecies electron transfer. In addition to H2, other molecules such as formate may act as an electron shuttle for biofilm communities (Morita et al., 2011).

An exciting new discovery by Pfeffer et al. (2012) provides new hope for greatly enhancing electron trans­fer in microorganisms. They found that some filamen­tous bacteria in marine sediments are capable of transferring electrons over centimeter-long distances via conductive filaments that are 200 nM or wider in diameter. This distance is far greater than the much thinner pili could achieve. This indicates that potentially many more layers of sessile cells could be networked via the conductive filaments than pili could achieve. Figure 9.2 is a schematic illustration of DET via direct cell wall—electrode contact and via a pili or conductive filament linkage.