For biomass derived ethanol to ever compete with traditional sugar or starch based commercial processes it would be extremely beneficial if an ethanol titer of at least 4% (w/v) with >80% yield could be achieved. One promising strategy to try to realize these goals is to process pretreated biomass slurries (combining the hemicellulose and cellulose derived sugars of the biomass feedstock) at high solid consistencies (i.e. high sugar concentrations). However, process parameters that incorporate ways to provide good ethanol production under conditions of high sugar, high inhibitors, high ethanol and low additional nutrient concentrations first need to be established. In earlier work (Schwald et al.
1988, Kumar et al.
2010) it was shown that optimised steam pretreatment conditions not only “opened-up” the cellulosic component of biomass substrates so that enzymatic hydrolysis could be enhanced, under the proper conditions it also provided the recovery of most of the hemicellulose sugars in the water soluble fraction (WSF). However, the hemicellulose rich water soluble fraction also contained most of the inhibitory material, both naturally occurring and process derived, that any fermentative organism would have to deal with (Robinson et al.
2003, Almeida et al.
2007). As noted in earlier work (Robinson et al.,
2003) it is technically very difficult to produce biomass derived sugar at the high, ≥ 200 g/l, concentrations typically used in commercial sugar- and starch-based ethanol processes (Bai et al.,
2011), primarily because of the high solid (pulp) consistencies (~40% w/w) that would be required. The few studies that have reported sugar concentrations in excess of 200 g/l used delignified lignocellulosic substrates and high doses of cellulolytic enzymes (Yang et al.
2010, Zhang et al.
2009). Therefore, to try and at least simulate high gravity (high sugar) concentrations in the presence of most of the anticipated inhibitory materials, supplemental glucose was added to the hemicellulose rich water soluble stream obtained after steam pretreatment of Douglas-fir wood chips. As described earlier we investigated three process parameters pertinent to the high gravity fermentation of the sugar present in glucose supplemented or unsupplemented steam pretreated water soluble fractions. These included the nature of the yeast strain used, the inoculation cell density and the requirement for any additional nutrients.
The strains Lallemand LYCC 6469, Tembec T1, and T2 were best able to ferment the wood derived sugars. The Tembec T1 and T2 strains are natural isolates from spent sulfite liquor (SSL) which is the residual liquid stream obtained after the ammonium bisulfite digestion of wood (Boussaid et al.
2001, Helle et al.
2003). In addition to its high osmotic strength, SSL is rich in numerous fermentation inhibitors including acetic acid, furfural, HMF, lignosulfonates, and sulfate (Helle et al.
2004). Through long-term exposure to SSL, the T1 and T2 strains have been adapted for survival and ethanol production under adverse environmental conditions. Our earlier work had shown that the T1 and T2 strains could ferment most of the sugars present in softwood hydrolysates (Liu
2010) while other SSL-adapted strains of S. cerevisiae (TMB3000 and USM21) have also been shown to be better able to ferment lignocellulosic hydrolysates to ethanol (Laadan et al.
2008, Nilsson et al.
However, when using a low cell density inoculation, even the more robust yeast strains did not grow or ferment the sugars well. In contrast, as has been shown by other workers, (Chung and Lee
1985, Boyer et al.
1992, Palmqvist et al.
1998) high cell density inoculations resulted in a significant and sometimes rapid reduction in some of the recognized fermentation inhibitors. Another beneficial effect is that high cell concentrations tend to down-regulate yeast cell growth, resulting in reduced carbon drain for growth. As a consequence, more carbon can be used for ethanol production (Melzoch et al.
1991, Palmqvist et al.
1998) while a higher cell density, due to the increased number of “working cells”, enables higher volumetric ethanol productivity (Borzani et al.,
1994, Palmqvist et al.
1998, Cardona and Sanchez
2011). In related work, continuous fermentation using spruce enzymatic hydrolysate showed 4.6 times higher productivity values under high cell density conditions (Palmqvist et al.
1998). Similarly, when the hydrolysate from steam pretreated sweet sorghum bagasse was fermented, increasing the cell density from 1 g/l to 3 g/l led to a >3-fold increase in productivity (Shen et al.
2012). Productivity improvements have also been reported when using molasses and sugarcane bagasse hydrolysate (Palmqvist et al.
1998, Canilha et al.,
2010, Canihla et al.,
2012, Nofemele et al.,
The work reported here confirmed the earlier observation by Liu
(2010) that glucose supplementation improved the fermentability of the sugars present in the water soluble fractions of steam pretreated softwoods. It has been suggested that, in the presence of high glucose concentrations, ATP can be more readily regenerated to meet the increased maintenance energy demand under inhibitor stress (Helle and Duff
2004). Transcriptome analyses indicated that HMF tolerant S. cerevisiae strains reprogram metabolic pathways to maintain energy metabolism and redox balance by creating a short cut to the TCA cycle to enhance ATP and NADPH synthesis. As a result, glycolysis is repressed and glucose metabolism is rewired to continue through the pentose phosphate pathway, the main route to NADPH regeneration in yeast (Ma and Liu
2011). NADPH is necessary for the reduction of HMF into HMF alcohol, while ATP is possibly used, among other processes, for pumping toxic products generated by inhibitor damage out of the cells (Liu
2011). It is also possible that the beneficial glucose effect on HMF removal may be an indirect result of changes in gene expression in response to osmotic stress caused by high sugar. Certain stress factors are known to cause changes in expression that enable tolerance to other unrelated stresses (Liu
2011). In the case of the high gravity multi-stress fermentations studied here, the relatively high sugar concentrations used might have induced further expression of the reductases capable of detoxifying HMF.
As well as looking at the influence of high sugar concentrations and inhibitors on the fermentation of wood derived sugars, supplementation of nutrients cocktail, containing 0.5 g/l NH4H2PO4, 0.025 g/l MgSO4.7H2O, and 1.0 g/l yeast extract could significantly improve the ethanol yields of the sugar supplemented water soluble fraction (WSF). Lignocellulosic hydrolysates generally lack some of the key nutrients that are needed for optimal yeast growth and stress tolerance. Other workers have successfully used this cocktail to enhance yeast growth and ethanol production when grown on the sugars present in un-detoxified softwood slurries (Rudolf et al.
2007, Bertilsson et al.
2009, Olofsson et al.
In conclusion, several important process parameters were better defined to help achieve effective high gravity fermentation of softwood derived wood sugars. When the Lallemand LYCC 6469, Tembec T1, and T2 yeast strains were used at a high cell density of at least 150 × 106 cells/ml both the furfural and HMF present in the glucose supplemented, hemicellulose rich water soluble fraction were quickly removed. An ethanol yield of 77% could be achieved after 48 h when strain LYCC 6469 was grown at high cell density with nutrient supplementation. As the water soluble fraction of steam pretreated biomass is known to contain most of the inhibitory material present in the original and processed biomass, it is likely that a similar approach of using inhibitor adapted yeast at high cell density can be effectively used to ferment high consistency slurries of softwood derived cellulose and hemicellulose sugars.