9.3.1 Pretreatment

Pretreatment is necessary to overcome the recalcitrant nature of native lignocellulose. In contrast to the polysaccharides starch and glycogen, the function of cellulose is not to serve as an energy store, but rather to serve as a construction material. Protection of the cellulose by lignin and hemicellulose reinforces that structure, and the degradation rate in nature is therefore lower than desirable for technical application (but well suited for a functioning eco-system). A pretreatment of the material makes the fibers more accessible to enzymatic attacks. The treatment aims to open up the structure of fibers, solubilize parts of the material (i.e., lignin or hemicelluloses), reduce particle size and degree of polymerization (DP), and increase the surface area of the material.

The pretreatment in many ways determines the bioethanol process, since the properties of the material after pretreatment will have an impact on all the subsequent steps in the production. In general, it is important to design the pretreatment so that not too much degradation products are formed, since these will impact primarily the fermentation, but also the hydrolysis (Almeida et al., 2007).

Pretreatment methods can be divided into physical or chemical methods, although a combination of the two is often used. One can say that there are two principal strategies of the pretreatment; either you aim to remove mainly the lignin or you aim to remove mainly the hemicellulose fraction from the material. No method will completely succeed with any of these two aims, but hemicellulose is mostly removed with dilute acid catalyzed steam pretreatment, and lignin is to a larger extent removed by alkaline pretreatment, and oxidation improves the efficiency of lignin removal (Alvira et al., 2010; Galbe and Zacchi, 2007). Two common pretreatment abbreviations used are STEX for steam explosion and AFEX for ammonia fiber explosion. A compilation of some of the most common technologies is given in Table 9.3.

9.3.2 Hydrolysis

The pretreatment will leave most of the cellulose in polymeric form, and, depending on the pretreatment, some hemicelluloses in polymeric or oligomeric form. Enzymatic hydrolysis of cellulose (and hemicelluloses) into monomeric sugars is thus needed since most microorganisms utilize only monomeric (or possibly dimeric) sugars during fermentation. The enzymatic hydrolysis was long regarded as the principal bottleneck in bioethanol production from lignocellulose due to the rather slow action of cellulases and the need for large amounts of expensive enzymes. However, impressive progress in the past decade has resulted from major R&D efforts by enzyme developers, such as Novozymes, Genencor/DuPont, DSM and others, to reduce both enzyme loadings and production costs. The US Department of Energy (DOE) set a target that enzyme costs should not exceed US$0.12/gallon ethanol by 2012 in their grants to some major enzyme companies. A benchmarking of enzymes from these companies was recently conducted by the National Renewable Energy Laboratory (NREL) (McMillan et al., 2011).

Traditionally, enzymatic hydrolysis has been regarded as a synergetic reaction between three major classes of enzymes, i.e. endo-1,4-β-glucanases, exo-1,4-β-glucanases and β-glucosidases (Van Dyk and Pletschke, 2012). Endo-1,4-β-glucanases randomly cleave internal bonds in the cellulose polymer which results in the formation of two new chain ends. Exo-1,4-β-glucanases (mainly cellobiohydrolases, CBH) are the most abundant component in both natural and commercial enzyme mixtures and they mainly work in a processive manner, cutting (mainly) cellobiose units from either the reducing or non-reducing ends of the glucan chain. β-Glucosidases catalyze the hydrolytic splitting of cellobiose into glucose. The enzymes are end-product inhibited by glucose and especially cellobiose. Therefore it is important to have an enzyme blend with sufficient β-glucosidase activity in order to avoid cellobiose inhibition of the cellobiohydrolases and to be able to cope with the activity loss due to the accumulated high amount of glucose.

The synergistic effects between the different cellulase components have been extensively studied throughout the years and more recently also the synergism between cellulases and other auxiliary enzymes, e.g. xylanases, xylosidases, mannanases and esterases have been studied (Van Dyk and Pletschke, 2012). In particular it has been shown that xylobiose, and larger xylo-oligomers, exhibit a very strong inhibition on the cellulases, which can be reduced/minimized by adequate supplementation of hemicellulases (Qing and Wyman, 2011). This xylo-oligomer inhibition can be quite significant when working with xylan-rich lignocellulosic substrates (e.g., agricultural residues and hardwood; Table 9.2), especially when a rather mild pretreatment method has been used, which gives solubilization of xylan into oligomeric rather than monomeric xylose.

Another complication when hydrolyzing lignocellulosic materials is the presence of large amounts of lignin in the material (if a lignin dissolving pretreatment method has not been used). Lignin could act as a physical barrier on the cellulose surface, limiting the accessibility of the enzymes, and it has been shown to unproductively bind enzymes which lower the free amount of enzymes able to hydrolytically degrade the cellulose/cellobiose (Palonen et al., 2004).

Recently a new type of enzyme and enzyme action was identified as an important component for efficient enzymatic hydrolysis of cellulosic material (Horn et al., 2012). The enzyme belongs to the GH61 family (the abbreviation GH stands for glycoside hydrolases) but the enzyme is in fact an oxidizing enzyme that oxidizes the C1/C4 carbon on the glucan chain. Hence it has a completely different way of cleaving intermolecular glucose bonds compared to the hydrolytic cellulases mentioned earlier. Unlike the cellobiohydrolases, the GH61 enzyme does not work in a processive manner, and does not appear to need a specific binding site on the polymer chains. Therefore it can attack and cut the cellulose chain at any location, creating two new free ends for the cellobiohydrolases to act on. As such, it has the potential to break crystalline cellulose structures and enhance the hydrolysis rate by creating more reactive sites for the CBH enzymes.

9.3.3 Fermentation

Once the fermentable sugars have been obtained in the hydrolysis, the next issue is the conversion of these sugars (all sugars, or only a selected stream) by fermentation. The ideal reaction for conversion of glucose (the main hexose sugar) to ethanol is:

${\mathrm{C}}_6{\mathrm{H}}_{12}{\mathrm{O}}_2\to 2{\mathrm{C}}_2{\mathrm{H}}_5\mathrm{OH}+2{\mathrm{CO}}_2$

i.e., one mole of glucose gives a maximum of two moles of ethanol, which corresponds to 0.51 g ethanol/g glucose. In practice, one should not count on more than about 90% efficiency in the fermentation, i.e. about 0.46 g/g (Öhgren et al., 2007). The other main hexose sugars are mannose and galactose (see Table 9.2). Out of these sugars, mannose is normally easily fermented, whereas the fermentation of galactose will depend on the microorganism used. The fermentation does not need to be done after the hydrolysis, but can be done concomitant with the enzymatic hydrolysis – a process option called SSF (simultaneous saccharification and fermentation; see further discussion later on). The microorganism used for the fermentation must be able to work in the process streams with as little conditioning of the stream as possible. The microorganism should therefore be:

There are many microorganisms that are able to ferment sugars, including yeasts, e.g. Saccharomyces cerevisiae, Scheffersomyces stipitis,² Kluyveromyces marxianus, Dekkera bruxellensis (Blomqvist et al., 2010), bacteria such as Zymomonas mobilis (Rogers et al., 1982), Escherichia coli (reviewed by, e.g., Jarboe et al., 2007) and filamentous fungi such as Mucor indicus, and Rhizopus oryzae (Abedinifar et al., 2009). The organisms have different pros and cons in terms of ethanol tolerance, fermentation rate, sugar utilization range, by-product formation pattern, and tolerance to inhibitors. Genetic engineering of the organism is often needed to improve properties. For instance E. coli produces only little ethanol in its native state, but can be transformed into an ethanologenic organism (Gonzalez et al., 2003).

The process design, or the entire refinery concept, will influence the choice of organism. The organism will need to be able to use all sugars if a maximum ethanol production is desired (which is not the case in all biorefinery concepts). Alternatively, if the SSF configuration is chosen, a high thermo-tolerance of the microorganism will be an advantage (apart from the above four) as the temperatures optima for enzymatic hydrolysis (typically 45–50°C) and fermentation (typically 30–35°C) differ significantly. The thermo-tolerant yeast Kluyveromyces marxianus may be of interest in these processes (Pessani et al., 2011). However, the prime choice in the current sugar and starch-based fermentation industry has been the common Baker’s yeast, Saccharomyces cerevisiae and there is good cause to believe that this will be the case also in lignocellulose-based biorefinery approaches. There is well-established process technology for large-scale production of the yeast (Østergaard et al., 2000), and it is fully accepted in the fermentation industry today and enjoys GRAS status (i.e., it is generally regarded as safe). S. cerevisiae has an excellent ethanol tolerance, some strains tolerate up to 20% (v/v) (Verduyn et al., 1990; Casey and Ingledew, 1986). Not least important is that the organism is relatively robust to inhibitors from the pretreatment step of lignocellulose (Hahn-Hägerdal et al., 1994; Olsson and Hahn-Hägerdal, 1993).

9.3.4 Pentose utilization

In biorefinery concepts aiming at a maximum ethanol yield, all sugar streams should be fermented to ethanol, which requires the conversion also of pentoses. However, wild-type S. cerevisiae is not able to ferment pentoses, and genetic engineering of the organism is needed to enable pentose conversion. The main pentose sugar in most biomass is xylose (D-xylose) (Table 9.2) and metabolic engineering of the yeast has therefore aimed at enabling xylose fermentation in S. cerevisiae (reviewed by, e.g., Almeida et al., 2011; Van Vleet and Jeffries, 2009; Matsushika et al., 2009; Hahn-Hägerdal et al., 2007). There are two main options to introduce the enzymatic pathway which will enable conversion of the pentose xylose. One option is the two-step conversion via the enzymes xylose reductase (XR) and xylitol dehydrogenase (XDH). The other option is to use the one-step isomerization catalyzed by xylose isomerase (XI). In both cases the sugar xylulose is formed, which later is phosphorylated by the native enzyme xylulokinase (XK) to xylulose-5-phosphate, and enters into the main glycolysis via the pentose phosphate pathway (PPP) (see Fig. 9.4). The mere introduction of the pathway is, however, not sufficient for obtaining efficient yeast strains and additional improvements are necessary (as discussed in, e.g., Almeida et al., 2011). In the XR/XDH pathway, a main issue has been the co-factor dependencies of the two reactions, which may result in excretion of the by-product xylitol. This co-factor problem is avoided when using the XI pathway, but a main challenge has been to find an isomerase showing sufficient activity at allowable temperature for the yeast. Successful results were not really obtained until a XI from the anaerobic fungus Piromyces was expressed in S. cerevisiae (Kuyper et al., 2003, 2005). More recently also other useful XIs, from the anaerobic bacterium Clostridium phytofermentans (Brat et al., 2009), the fungus Orpinomyces (Madhavan et al., 2009), and from Burkholderia cenocepacia (de Figueiredo Vilela et al., 2013), have been found and expressed in S. cerevisiae.

In addition to the direct genetic engineering, i.e. introduction of the ‘missing’ heterologous genes, it has been shown to be very important to evolve the strains on the xylose-rich substrate. In particular, this was crucial in obtaining the first successful XI strain (Kuyper et al., 2003). During the evolution, the relative expression levels are ‘tuned’ for optimal performance by the yeast itself through spontaneous mutation and selection mechanisms (Zhou et al., 2012; Lee et al., 2012). Some of the currently available S. cerevisiae strains designed for utilization of xylose are given in Table 9.4.

A second pentose found at relatively high levels in some materials is arabinose (L-arabinose). The problem of introducing arabinose utilization into yeast is similar to that for xylose utilization as both pathways end with the formation of xylulose (Fig. 9.4; Almeida et al., 2011). Also similarly there is an isomerase pathway and a reduction/oxidation pathway. Arabinose utilization on its own is less interesting, and work has therefore focused on combining utilization of L-arabinose and D-xylose (Bettiga et al., 2009; Sanchez et al., 2010). Whereas there are currently several xylose fermenting yeast strains available, this is not yet quite true for arabinose co-utilizing yeasts.

9.3.5 Inhibitor tolerance

The challenge in lignocellulose fermentation is not only to develop organisms which use more sugars, but to have them do that in a complex medium containing several inhibitors (Fig. 9.5). The pretreatment process generates several compounds, which affect the performance of the fermenting microorganism (Klinke et al., 2004; Palmqvist and Hahn-Hägerdal, 2000). These inhibitors can be divided into:

The furans are formed in the degradation of monosaccharides, and levulinic and formic acid come from further degradation of the furans. Acetic acid is different in the sense that it is already present in the material itself in terms of acetyl groups on hemicellulose.

One can handle the inhibition problem in different ways (Fig. 9.6). Obviously, the best situation would be not to have any inhibitors in the medium. That would require either a process that does not generate any inhibitors, or the removal of inhibitors. Alternatively, one can screen for strains that are tolerant, or develop more tolerant strains. Also the fermentation process in itself may be important in avoiding problems with inhibition. The reason is that several of the inhibitors, notably several aldehyde compounds such as furfural and HMF, are converted by the yeast itself into less toxic alcohols (Almeida et al., 2007, 2009) and an in-situ detoxification can be obtained during the process if suitably tuned.

By understanding this particular mechanism of inhibition, several genetic targets have been identified. Overexpression of the gene encoding for alcohol dehydrogenase (ADH6) was, for instance, found to increase the ability of S. cerevisiae to convert HMF (Petersson et al., 2006) and also increased the ethanol formation from a hydrolysate. Alriksson et al. (2010) observed that overexpression of three native genes (ATR1, FLR1, YAP1) involving multidrug resistance and stress response in S. cerevisiae resulted in enhanced resistance to coniferyl aldehyde, HMF, and spruce hydrolysate. More targets for improved resistance are continuously reported. Very promising results have also recently been reported from adding sulfite or dithionite to the medium, which gave detoxifying effects on spruce hydrolysates (Alriksson et al., 2011). The mechanism of detoxification may partly be reduction, but also sulfonation of inhibitors (Cavka et al., 2011).

Acetic acid is a different story in comparison to many other inhibitors in that it is linked to the material composition per se since the hemicellulose is acetylated. Thus acetic acid is very difficult to avoid if the hemicellulose is to be hydrolyzed. The effects of weak acids are pH dependent, since it is the undissociated form of the acids which diffuses across the plasma membrane. There it dissociates and releases a proton due to a higher intracellular pH. To maintain the intracellular pH, the cell has to export the proton which costs ATP. The pKa value of the acid will determine at what pH the undissociated form of the acid dominates. Since many carboxylic acids have pKa in the range 4.5–5, large changes in toxicity are likely to occur for pH changes in this range. Xylose utilization appears to be extra sensitive to acetic acid (Bellissimi et al., 2009; Casey et al., 2010).

9.3.6 Thermo-tolerance

Thermo-tolerant S. cerevisiae would be an advantage in the SSF process since the temperature optima for hydrolysis (45–50°C) and fermentation (30–35°C) are different. Exposure of S. cerevisiae cells to high temperatures induces expression of several heat stress–response genes, including those encoding heat shock proteins (HSPs) and enzymes involved in trehalose and glycogen metabolism (Morano et al., 1998; Singer and Lindquist, 1998). A high-temperature growth phenotype (Htg +) was categorized in thermo-tolerance of S. cerevisiae (Steinmetz et al., 2002). Results of classical genetic analysis suggested that the Htg phenotype is dominant and approximately six genes, designated HTG1 to HTG6, are responsible for conferring this phenotype. HTG6, one of the six genes was recently identified to be the gene RSP5, which encodes a ubiquitin ligase (Shahsavarani et al., 2012) and overexpression of RSP5 ubiquitin ligase improved thermo-tolerance in S. cerevisiae. Zhang et al. (2012) isolated thermo-tolerant S. cerevisiae with a high-energy pulse electron (HEPE) beam, and obtained a strain which could produce more than 80 g/L of ethanol at as high a temperature as 43°C.

9.4.1 Separate hydrolysis and fermentation (SHF) or simultaneous saccharification and fermentation (SSF)?

The two basic choices have different pros and cons (Table 9.5). The fact that hydrolysis and fermentation are carried out in separate vessels in SHF introduces a degree of freedom to operate each process at its respective optimal conditions, however, at the cost of one more vessel. The extra capital cost may be quite substantial as indicated by Wingren et al. (2003). If the solid fraction (mainly lignin) is removed after the enzymatic hydrolysis, it is possibly to recycle the yeast in an SHF process. This is not feasible in SSF, since the yeast cannot readily be separated from the remaining lignin. However, in the removal of the lignin fraction in SHF, there may be non-negligible sugar losses, especially if a high solids loading is used in the enzymatic hydrolysis. In contrast, if the entire slurry is sent to fermentation, all monosaccharides liberated in the pretreatment can potentially be fermented. The removal of sugars as they are released in the hydrolysis in SSF is furthermore an important advantage since end-product inhibition of cellulases can be minimized. An increased overall yield for these two reasons was the main benefit stated by the original inventors of SSF (Gauss et al., 1976). The enzymes are to some extent inhibited by the ethanol produced, but this inhibition is relatively low compared to that of cellobiose and glucose (Wu and Lee, 1997). Decreased end-product inhibition of cellulases, in particular cellobiose inhibition on CBH, is, however, a major target in development of more efficient enzymes, which may tilt the balance in favor of SHF processes.

Co-fermentation of pentoses and hexoses in SSF is sometimes dignified with the abbreviation SSCF, standing for simultaneous saccharification and co-fermentation. SSCF may offer an advantage over SHF in these cases since the ratio between xylose and glucose can be higher than in SHF (for a more thorough discussion, see Olofsson et al., 2008). This is true for the cases in which the pretreatment results in the release of monosaccharides from hemicellulose, e.g. in acid catalyzed steam pretreatment, but not in cases where the pretreatment does not release monosaccharides.

Since the yeast cannot be reused in the SSF process, it has to be produced for each batch. Efficient aerobic cultivation of S. cerevisiae has to be performed in a fed-batch mode to avoid glucose repression which would give a low biomass yield. The fed-batch cultivation of the yeast should use lignocellulose hydrolysate (derived from the pretreatment stage) to adapt the yeast since this improves the performance significantly (Alkasrawi et al., 2006). The cultivation must be designed not only to avoid glucose repression but also inhibition by the medium. RQ (respiratory quotient), DOT (dissolved oxygen tension), or ethanol concentration can be used as measured variables for feedback control during the propagation of the yeast (as discussed by Rudolf, 2007).

The choice between SHF and SSF will depend on both feedstock and the desired product spectrum in the biorefinery. There are many different variants of both these concepts (Fig. 9.7), in particular with respect to feeding of substrates and enzymes. Feeding will be a means to minimize effects of inhibition, increase xylose conversion, or handle high viscosities (Olofsson et al., 2008).

9.5.1 Hybrid processes and novel concepts

The desire to increase the biomass content in the process has led to different ‘hybrid’ process configurations which employ a combination of the SHF and SSF approach – aiming at getting the best of both worlds. One such option is the inclusion of a so-called viscosity reduction step, which in essence is an enzymatic hydrolysis step operated at optimal hydrolysis temperature for a short time, but not to complete hydrolysis. A high initial hydrolysis rate can be achieved, and the process can continue as an SSF process, much like in corn-based ethanol production. One can also combine this with various fed-batch strategies in which material, enzymes and/or yeast are fed to the fermentation vessel during the process. Another option is to apply a changing temperature during the process. This is of particular interest in the SSF process, where a non-isothermal operation may give overall process improvements (Kang et al., 2012; Mutturi and Lidén, 2013). The relative rates of enzymatic hydrolysis, glucose consumption, yeast growth and decay are affected during such non-isothermal operation. Process benefits will depend on the recalcitrance of the pretreated material, and optimal and permissible temperature ranges of enzymes and yeast.

Another process concept is the so-called consolidated bioprocessing (CBP). In a CBP process, four steps (i.e., production of hydrolytic enzymes (cellulases and hemicellulases), hydrolysis of polymeric carbohydrates in the pretreated substrate, fermentation of hexoses, and fermentation of pentoses) are carried out in a single reactor (Lynd et al., 2005). One can either use a wild-type organism, such as Clostridium phytofermentans (Jin et al., 2011) able to both produce cellulases and ferment, or use genetic engineering to introduce genes encoding one or several cellulases into a fermenting microorganism, or alternatively engineer a good cellulase producing organism into an ethanol producer. The main approach has been to introduce cellulase encoding genes into a fermenting organism, e.g. S. cerevisiae. CBP is slowly gaining recognition as a potential low cost biomass processing methodology (van Zyl et al., 2007). Also here one can foresee various hybrid concepts, in which perhaps a ‘base mixture’ of enzymes is added, complemented by some specific enzyme activities that are expressed by tailored yeasts.

9.7.1 Industrialization and process development

The development of lignocellulose-based ethanol biorefineries has gone through several ups and downs in the past decades. Steadily, however, research has come closer to industrial realization and there are today at least 10 demonstration-scale facilities in Europe and the US alone (see Table 9.6). A very significant step forward is the current completion of the first commercial (or semi-commercial) scale ethanol biorefinery plants. The plant in Crescentino, Italy, built by Beta Renewables leads the way, closely followed by plants built by Abengoa and POET in the US. The completion of commercial plants shows that the technology has reached a sufficient level of maturity for implementation. This, however, does not mean that there are no development needs, but rather that these now shift into more directly process-related issues and optimization. Some of the main issues worked on are summarized in Table 9.8. The number of biofuel-related patent filings is rising. Some recent innovations (Table 9.9) give a flavor of the current development efforts on process efficiency, pretreatment improvement, novel enzymes and fermenting strains as well as the ‘biorefinery’ issues on valorization of by-products.

9.7.2 A future transition of paper industries?

One may argue that paper industries are in fact full-scale biorefineries. This is true in a sense, but the product focus has been very strong on pulp and paper only. The development of integrated forest biorefineries with also other products may open up for a new kind of business (Marinova et al., 2009). The major advantage of such integrated biorefineries would be to use the existing infrastructure for transport and material handling. Some options for the transition of a traditional pulp and paper industry into an integrated forest biorefinery are:

The black liquor in traditional pulp mills is used in recovery boilers for steam generation. However, alternative technologies are being developed for black liquor gasification, wherein the lignin present in the black liquor is converted to synthesis gas (Pettersson and Harvey, 2012). The synthesis gas (or syngas) is in turn used for production of dimethyl ether, methanol or other fuel compounds. A schematic representation of an integrated forest biorefinery based on a Kraft mill is shown in Fig. 9.11

9.7.3 Lessons from LCA studies

The future success of bioethanol refineries will be determined not only by the technical and economic performance of the plants, but also the environmental performance of the plants. Lowering of net greenhouse gas emissions (GHG) is, as mentioned in the introduction, particularly critical. Life cycle assessment (LCA) is necessary to monitor the above mentioned benefits when compared to that of fossil fuels, and LCA is now an accepted tool for guiding decision-making towards sustainability. LCA involves a holistic cradle-to-grave assessment approach to evaluate environmental performance by considering the potential impacts from all stages of manufacture, product use (including maintenance and recycling), and end-of-life management (von Blottnitz and Curran, 2007). Moreover, the International Organization for Standardization (ISO), a worldwide federation of national standards bodies, has standardized this framework within the ISO 14040 series on LCA.

The results from recent studies on LCA of biofuel production plants have established that global warming emissions and fossil energy consumption can be reduced significantly when conventional diesel and gasoline are substituted with biofuels (Punter et al., 2004; Kim and Dale, 2005; von Blottnitz and Curran, 2007; Liska et al., 2009; Cherubini and Jungmeier, 2010). However, in terms of other environmental impacts related to, for example, land usage, eutrophication of surface water, acidification and water pollution by pesticides, negative effects can result and must be carefully considered (Cherubini and Jungmeier, 2010). The environmental impact of a biorefinery depends on factors such as biomass feedstocks used, conversion routes, process configurations, and end-use applications. The LCA analysis can be used to identify further steps to be taken to, for example, reduce GHG emissions and improve overall energy efficiency of biorefineries. Examples include use of thermo-compressors for steam condensation, waste heat recovery and reuse of heat, and combustion of volatile organics using thermal oxidizers (Liska et al., 2009).

9.7.4 Future crops

The availability of feedstocks is a strategic question, and development on future dedicated feedstocks for biofuel production has also taken a prominent place in the agenda of bioenergy research. It is not possible to point out one single new feedstock which should be used for bioethanol production, as local factors related to regio-specific agricultural practices, market forces, political directives, social and biological issues will define the suitability (Sticklen, 2008). However, factors to consider include:

Technologies for the conversion of lignocellulosic feedstocks, such as agriculture residues from corn, rice, sugarcane, perennial grasses (switchgrass and giant miscanthus) or short rotation woody crops (fast-growing poplar and shrub willow), which contain higher cellulose fractions, should be given more prominence (Ravindranath et al., 2009; Sticklen, 2008). Newer and emerging tools in plant genetic engineering offer great potential in reducing the processing costs of these feedstocks to bioethanol. Techniques such as production of cellulases and hemicellulases within the crop biomass, genetic manipulation to modify lignin content, upregulation of cellulose and hemicellulose biosynthesis, and expression of drought tolerant traits are some of the potential research areas for improving biomass feedstocks for bioethanol production (Sticklen, 2008).