Three methods are possible for hydrolyzing cellulose into glucose (C6 sugar for fermentation): 1. dilute acid hydrolysis (<1% H2SO4, 215°C, 3 min with 50-70% glucose yield) which is no longer a viable candidate; 2. concentrated acid (30-70% H2SO4, 40°C, a few hours, >90% glucose yield), which has been used in Japan and will be evaluated in a DOE-funded pilot facility (Table 1); and 3. enzymatic hydrolysis (cellulase mixture, ~50°C several days, 75-95% glucose yield).

The efficient enzymatic hydrolysis of cellulose by cellulases requires a coor­dinated and synergistic action of three groups of cellulases: endoglucanase (EG, E. C. 3.2.1.4), exoglucanases like cellodextrinase (E. C. 3.2.1.74) and cellobiohy — drolase (CBH, E. C. 3.2.1.91), and в-glucosidase (BG, E. C. 3.2.1.21). EGs and CBHs act on insoluble cellulose molecules [39]. EGs randomly act internally on the amorphous regions of a cellulose polymer chain and generate oligosaccharides of various lengths and additional free ends (reducing and non-reducing ends) for CBH action. CBHs usually hydrolyze both amorphous and crystalline cellulose and cellooligosaccharide chains from the non-reducing ends in a sequential way with cellobiose as the major product, but some CBHs can hydrolyze cellulose chains from both reducing and non-reducing ends [40-42]. The hydrolysis products of these two groups of enzymes include cellodextrins, cellotriose, cellobiose, and glucose. в — glucosidases hydrolyze soluble cellodextrins and cellobiose into glucose from the non-reducing end and remove the product feedback inhibitory effect of cellobiose on EG and CBH (Fig. 5 ).

Factors impacting the activity of cellulases include enzyme source (e. g. organ­isms and producing conditions), concentration, and combinations. The normal enzyme dose for cellulose hydrolysis study is 10-60 FPU per gram of dry cellulose or glucan; glucanases to в-glucosidase ratio is approximately 1.75-2.0 IU of в — glucosidase for each FPU of glucanase used [29]. Most commercial glucanases are produced by Trichoderma resell and the в-glucosidase is typically from Aspergillus niger [43].

Under research conditions, the reported digestibility or the conversion yield of cellulose from pretreated lignocellulose can be high (Table 4). However, actual glu­cose yield may vary greatly depending on the type of biomass, method/condition of pretreatment, cellulases (composition, source, and dose), solid to liquid ratio of the hydrolysis mixture, and other unspecified factors. The cellulose digestibil­ity of corn stover and corn fiber can reach >90% following dilute acid or liquid hot water pretreatment [44], while the digestibility of rice hulls after similar pre­treatment was about 50% [45]. Similar low digestibility results were obtained on dilute acid pretreated sorghum stubble in our lab (unpublished data). The vari­able digestibility of different biomass sources following dilute acid pretreatment may be an indication that this particular pretreatment is not universally effec­tive. Currently, all the reported results for AFEX [44] and alkaline peroxide [44, 46] treated biomass sources showed consistently high cellulose recovery, and high digestibility, even at lower enzyme concentrations and shorter incubation time (48 h vs normal 96 h) [47] .

Digestibility, or glucose yield, is high when cellulose load is low (1-3% cellu­lose load) in the hydrolysis system. Glucose yield from pretreated biomass typically increases as enzyme load increases [47,48, 49], while digestibility decreases as the cellulose load increases [48, 50]. We are unaware of any reports of >20% cellu­lose load with high digestibility. Starch-based ethanol production involves starch loadings of 20-25% or higher, that results in finished beers with ethanol concen­tration around 10-12% (w/v). Most lignocellulosic ethanol fermentation studies have used hydrolysates with 3-10% cellulose load, which resulted in a finished mash with ~3-5% (V/V) ethanol. Additional research is required to improve the lignocellulose situation. Some non-cellulolytic enzymes (e. g. ferulic acid esterases and various xylanases) have been studied as pretreatment agents and showed promising results in increasing glucose yield from lignocellulose [51].

Since enzyme cost is a large contributor to the total production cost for lignocel — lulosic ethanol [30, 44], considerable research has been undertaken in attempts to increase the efficiency and reduce the cost of enzymes. Addition of protein (bovine albumen) and other additives (Tween 20 or 80, polyethylene glycerol, etc.) that reduce the affinity between cellulases and lignin all improve the efficiency of cel­lulose hydrolysis [27]. A recycling process using an ultrafiltration membrane to separate hydrolyzed glucose showed that cellulases could be re-used up to 3 times for pretreated low lignin biomass, or until ~50% of the cellulases were bound on accumulated lignin [48].

To help lower enzyme costs and possibly improve effectiveness, a research strategy has been developed to genetically-engineer biomass to express transgenic endocellulases. Microbial cellulose transgenes have been expressed in several crops: tobacco, potato, tomato, alfalfa, rice, maize, and barley [52-54]. Endoglucanase 1 (E1) concentration in some transgenic experiments has reached 1% (corn stover) [55] to 5% (rice straw) [54] of total soluble proteins. In some cases, both treated and non-treated E1 engineered biomass showed higher digestibility than biomass of their wild counterparts. Whether transgenic expression of appropriate enzymes is a viable long-term strategy when used for large-scale production remains under investigation.