3.2.4. Relative SSF efficiencyThe e

3.2.4. Relative SSF efficiency
The ethanol yields were 24.7–42.4% and 21.5–51.8% for unwashed and washed alfalfa stems, respectively. It was of interest to know whether the relatively low yields were because of limitations in the SSF procedure. We used similar amount of glucan for the Sigmacell 50 (0.33 g) and alfalfa stems (1 g DM, or ~0.33 g glucan) during the SSF based on our forage analysis results. Thus, the ethanol yields of biomass (alfalfa stems) were compared to those of a pure cellulose model compound (Sigmacell 50); the ratio between ethanol yield from the alfalfa substrate and that from pure cellulose was defined as relative SSF efficiency. The relative SSF efficiency was well distributed between 36% and 97%. The highest relative SSF efficiency approached 100%, indicating the SSF on substrates was very good compared to that on pure cellulose. The well-distributed relative SSF efficiency was an advantage for studying variable effect, because the variable effect might be shielded if most of the experiments showed a high relative SSF efficiency. We also found that our data agreed with the work of Digman et al. (2010) when the substrates were switchgrass with a similar procedure. The highest ethanol yield was only about 50%, probably due to two reasons: (1) the SSF with 72 h was not long enough to reach the maximum ethanol production (Lan et al., 2013; Vincent et al., 2014); and (2) the enzyme load (5 FPU/vial, or _15 FPU/g glucan) was relatively low compared to other studies (Lan et al., 2013; Lu et al., 2013) but likely a more economically feasible level.