For the purposes of this chapter, the high temperature, dilute acid pretreatment and subsequent enzymatic saccharification method will be examined in more detail as a conversion technology for ethanol production from alfalfa stem fractions. The high temperature, dilute acid pretreatment is designed to remove noncellulosic cell wall polysaccharides and lignin, because these constituents will interfere with the cellulase enzyme cocktails used for hydrolysis of the cellulose. One design goal of this pretreatment is to reduce the pH of the feedstock reaction mixture to 1.3-1.5 prior to heating (National Renewal Energy Laboratory, Golden, CO; Laboratory Analytical Procedure-007, May 17, 1995). The amount of sulfuric acid required to reach this pH target for alfalfa stems was 8.1 mmol g-1 biomass DM in a 1% solids slurry, compared to 6.4 mmol for switchgrass and corn stover (Jung, unpublished). Maturity of alfalfa stems and switchgrass did not influence the acid requirement. Dien et al. (2005) observed that the sulfuric acid loading required to maximize release of nonglucose sugars from alfalfa stems when heated at 121°C for 1 h was 2.5% (wt/vol), whereas 1.5% was sufficient for switchgrass. The higher acid require­ment for alfalfa stems is most likely due to the greater pectin content of alfalfa cell

Immature Mature

walls compared to grasses; however, the hemicellulose content is lower and lignin content is similar in alfalfa stems compared to the grasses (Dien et al., 2005). Torget et al. (1990, 1992) also observed that legume feedstocks are more recalcitrant to acid pretreatment than grasses.

The efficiency of glucose release by acid pretreatment, followed by enzymatic saccharification from cell wall polysaccharides (cellulose and xyloglucans), declined as alfalfa stems became more mature (Figure 5.2). While efficiency of glucose conversion declined with maturity, the total yield of glucose was not altered (Figure 5.2), because cellulose content increased in more mature alfalfa stems. Similar declines in efficiency with maturity were observed for switchgrass and reed canary grass (Dien et al., 2005), but the efficiency of glucose release from the grasses was greater than from alfalfa stems. This may reflect the higher concentration of lignin in the alfalfa stems because across all three species, efficiency of glucose release was negatively correlated with lignin content of the feedstock. Increasing the temperature of the acid pretreatment resulted in improved efficiency of glucose release from alfalfa stems (Dien, personal com­munication). While efficiency of glucose release was lower for alfalfa than for grasses, total yield of glucose was very similar between the feedstocks. This again reflects the interaction of efficiency with glucose content of the feedstocks.