Ethanol toxicity could be a significant bottleneck in industrial ethanol fermentation of sugars from lignocellulose. To understand ethanol impact on xylose fermentation, batch fermentations were carried out using S. cerevisiae 424A (LNH-ST), an engineered strain capable of co-fermenting glucose and xylose. The fermentation of xylose was carried out in YEP growth media, using largely non-growing cells in the presence of initial ethanol concentrations between 4 - 8% (w/v). The effects of extraneously added ethanol (pure xylose fermentation) and ethanol generated from glucose equivalent (co-fermentation) are compared. This yeast strain was found to cease fermentation of xylose at an extraneously added ethanol concentration of 9% (w/v). However, co-fermentation of glucose and xylose was capable of achieving a final ethanol titer over 11% (w/v). A preliminary unstructured, Monod-type model of these batch fermentations that include ethanol inhibition is presented.

The degradation reaction routes of glucose and fructose under hydrothermal acidic conditions have been studied extensively; in contrast, xylose degradation has received less extensive study under similar conditions. In this study, we investigated the aqueous pH (0.5 – 7.0) impact on xylose degradation, and determined the kinetics of xylose disappearance rates at different pH conditions. The initial buffer system employed in this study was the McIlvaine buffer consisting of phosphate salt and citric acids (except for pH 0.5 – 1.5 buffers, where HCl/NaCl system was employed). It was observed that at pH 2.2, the xylose degradation rate was minimized (e.g. xylose disappearance rate at pH 4.2 is 9-times higher, and at pH 7.0 complete xylose disappearance occurred in 5-min reaction). In addition, the degradation reaction path changed from simple dehydration product (furfural) formation at lower pH range (0.5 – 3.0), to multiple complex liquid and polymerized products formation at higher pH range (4.5 – 7.0). In order to test the effect of buffering salt (phosphate, etc.), experiments at pH 1.0 with equivalent amount phosphate produced identical results to the same condition without phosphate addition. Therefore, the proton concentration in the aqueous solution may be the main controlling factor to which xylose degradation reactions occur. The degree of proton availability in the solution and potential protonation of the sugar –OH groups were analyzed to determine how the pH affects reaction path direction and products formation.

Corn stover is a heterogeneous substrate consisting of different fractions including leaves, stalk fiber and stalk pith. Tissue types and proportions in these fractions are not uniform which result in different cell structures, average cell wall thickness and lignin distribution. These factors may have different impacts on enzyme digestion, since the lignin barrier (content/distribution) and cell wall thickness are believed to be substrate related factors that influence the effectiveness of enzymatic hydrolysis of cellulose in lignocellulosic feedstocks. The hypothesis being tested in this research addresses potential differences of intrinsic reactivity of different parts of stay green corn stover (leaves, stalk fiber and stalk pith). Carbohydrate analysis shows that pith is more readily hydrolyzed than leaves and fiber at cellulase level equivalent to 5 FPU Spezyme CP/g glucan. Hot water pretreatment at 190 C for 15 min removes 40% to 50% hemicellulose in these fractions, respectively, although structural changes in the cell wall are not evident when the residual material is imaged by scanning electron microscopy. Enzyme hydrolysis of pretreated and washed fractions of leaves and pith exhibit much higher glucose conversion than the fractions that have not been pretreated. Pretreated fibere (from the rind) is still resistant to hydrolysis and shows 2/3 lower glucose formation than pretreated leaf or pith at low enzyme loadings equivalent to about 5 FPU Spezyme CP/g glucan.

Pretreatment of lignocellulosic biomass, while improving enzymatic digestibility, can also produce fermentation inhibitors. Two important inhibitors, furfural and HMF, are formed from the degradation of carbohydrates from lignocellulose. Thus, pretreated material may require conditioning to either remove or otherwise detoxify these inhibitors. This paper explores some conditioning methods on hydrolysates obtained from corn stover and poplar pretreated by dilute acid, controlled pH l iquid hot water, SO2 steam explosion, and others. The effects of these conditioning methods on the subsequent fermentation of both glucose and xylose by the recombinant yeast S. cerevisiae 424A(LNH-ST) is presented. Overliming the pretreated corn stover to pH 9 or higher removes 100% of the HMF and furfural present in corn stover hydrolysates. However, the fermentation is negatively affected, producing only 53% of theoretical ethanol yields as opposed to 82% yield from the unconditioned material. Hydrophobic resins (Amberlite XAD2, XAD4, and XAD7) were also examined for their ability to remove HMF and furfural. The resins were able to remove 100% of furfural and approximately 60% or more HMF. The yield from fermentation was 87%; slightly better than the unconditioned corn stover hydrolysate.

Pretreatment of lignocellulosic biomass, while improving enzymataic digestibility, can also produce fermentation inhibitors such as furfural and HMF. Both furfural and HMF can decrease the fermentability and the ethanol yields from sugars derived from lignocellulose. This paper reports a systematic study of the effect of furfural and HMF on the fermentation of both glucose and xylose to ethanol by the recombinant yeast S. cerevisiae 424A(LNH-ST). Fermentations were run with furfural, HMF, or both in a control solution of YEP with glucose and xylose as co-substrates. Inhibitor concentrations were varied and range from 0 to 40 g/L. Further experiments varied inhibitor concentrations in the presence of a single substrate, either glucose or xylose. Batch fermentations were carried out for 48 hours in 300 mL sidearm flasks at 30 C and 200 rpm with periodic sampling for anlaysis by HPLC. Our results show that concentrations of either furfural or HMF below about 5 g/L cause negligible inhibition for yeast cells in early stationary phase. We confirm that furfural is more inhibitory than HMF. Lastly, xylose fermentation to ethanol is more sensitive to these inhibitors than glucose for fermentation to ethanol.

The aqueous pretreatment of corn fiber at a pH of 4 to 7, while being pumped through a hold coil is effective in increasing the rate of enzyme hydrolysis of the cellulose. However, scale-up of the pretreatment process depends on physical properties of the material to be pumped through the system. High concentrations of fermentable sugars require that aqueous biomass streams from which these sugars are derived have a high solids content. Since the corn fiber solids at high loading have characteristics that resemble a shear-thinning fluid, measurement of viscosity in the laboratory is difficult, particularly at temperatures above ambient. Consequently, we carried out measurements in a plant setting. Corn fiber at 150 to 200 g/L were pumped at rates of 1 to 10 gal/minute through sections of jacketed tubing having diameters ranging from 1 to 1.5 inches and a length of 17.25 feet. The temperatures and pressure drops were measured at the inlet and outlet of the tubes and recorded through a LabVIEW programmed data acquisition system. The pressure drop and flow rate enabled calculation of viscosity and determination of correlations that will be useful for scale-up.

Liquid hot water pretreatment of plant biomass produces a liquid stream with dissolved oligosaccharides which are usually converted to fermentable sugars by enzymatic hydrolysis. In previous work, strong cation exchanger, Amberlyst 35W, has shown to hydrolyze cellobiose and oligosaccharides in liquid from corn fiber pretreatment at high conversion rates. This paper reports the effects of particle size, degree of cross-linking, and temperature on hydrolysis of oligosaccharides and degradation of monosaccharides. High temperature and short residence times were required to minimize formation of aldehydes and other fermentation inhibitors formation while achieving high glucose yield. The catalysts, SK104 (4% corrlinked gell type) and Amberlyst 35 (macroreticular sulfonic acid resin) were tested for hydrolysis of maltooligosaccharides at various reaction conditions. Maltooligosaccharides were used as a model oligosaccharide since their activation energy for bond breakage is similar to that of xylo- or cello-oligosaccharides, and since malto-oligosaccharides are more readily obtainable compared to the other types of oligosaccharides. Results show that low percentage cross-linked gel-type cation exchange resins give a higher glucose yield than macroreticular-type resins. The hydrolysis was diffision limited in both resins. A mathematical model that quantifies diffusion and kinetic characteristics of this reaction is presented and potential application of plug flow reactors to hydrolysis of oligosaccharides obtained from pretreatment of cellulose is discussed.