The Simultaneous Saccharification and Fermentation of Pretreated Woody Crops to Ethanol

Four promising woody crops (Populus maximowiczii x nigra (NE388), P. trichocarpa x deltoides (Nll), P. tremuloides, and Sweetgum Liquidambar styraciflua) were pretreated by dilute sulfuric acid and evaluated in the simultaneous saccharification and fermentation (SSF) process for ethanol production. The yeast Saccharomyces cerevisiae was used in the fermentations alone, and in mixed cultures with/~-glucosidase producing Brettanomyces clausenii. Commercial Genencor 150L cellulase enyme was either employed alone or supplemented with (~glucosidase. All SSFs were run at 37~ for 8 d and compared to saccharifications at 45~ under the same enzyme loadings. S. cerevisiae alone achieved the highest ethanol yields and rates of hydrolysis at the higher enzyme loadings, whereas the mixed culture performed better at the lower enzyme loadings without ~-glucosidase supplementation. The best overall rates of fermentation (3 d) and final theoretical ethanol yields (86-90%) were achieved with P. maximowiczii x nigra (NE388) and Sweetgum Liquidambar styraciflua, followed by P. tremuloides and P. trichocarpa x deltoides (Nll) with slightly slower rates and lower yields. Although there were some differences in SSF performance, all these pretreated woody crops show promise as substrates for ethanol production. Index Entries: Simultaneous saccharification and fermentation (SSF); dilute acid pretreatment; wouuy crops; cellulase; ~-glucosidase. *Author to whom all correspondence and reprint requests should be addressed. Aoplied Biochemistry and Biotechnology 773 Vol. 28/29, 1991 774 Spindler, Wyman, and Grohmann INTRODUCTION The simultaneous saccharification and fermentation (SSF) process for conversion of cellulose into ethanol was first studied by Takagi et al. (1,2) more than 10 years ago, and the process still dhows great potential for economic production of ethanol. Ethanol is a clean-burning, high octane fuel, and in light of our current concerns (i.e., urban air pollution, global warming, strategic vulnerability, and the trade deficit), the conversion of biomass to ethanol becomes an attactive alternative to fossil fuels. The SSF process employs a fermentative microorganism, in combination with cellulase enzyme, to minimize accumulation of sugars in the fermenter. As a result, inhibition of the enzyme by its product sugars is reduced, and higher hydrolysis rates and yields are possible than for straight saccharification (2). However, to produce ethanol from the SSF process that is competitive in price with petroleum-derived fuels, hydrolysis yields must be further increased, enzyme costs must be reduced, and ethanol production rates must be improved. SSF modeling, integration, and process engineering studies are presently underway to address some of these challenges. A recent economic analysis (3) of the SSF process with xylose fermentation estimates the selling price at $1.35/gallon. As continued research is conducted in the area of biomass-to-ethanol, further reduction of cost will be realized, with the goal to replacing petroleum fuels. Yeast selection for SSF has been described in several publications (4-9). Some of this work involved the selection of thermotolerant yeast (7-9), with the goal of selecting a yeast that can ferment at a temperature close to the optimal hydrolysis temperature for the cellulase enzyme, 45~ (7). However, although an increase in temperature can speed up the hydrolysis, loss of cell viability counters these gains, and 37 to 40 ~ still appears to be the best temperature for the SSF process (7,9). Saddler et al. (10) also found that a temperature of 37~ maintained fermentation of released sugars by Zymomonas and S. cerevisiae strains. CeUobiose-fermenting yeast have also been studied because additional /~-glucosidase activity can speed up the SSF reaction (5,11-13). The importance of end product inhibition of the ceUulase enzymes during cellulose hydrolysis has been modeled by Howell (14). Some publications discuss the advantage of the ceUobiose-fermenting yeasts in decreasing end product inhibition of cellobiose to the ceUulase enzyme (15,16). In general, S. cerevisiae, a strong glucose fermenter with a fast rate of fermentation, has been found to perform well if the enzyme preparation is high in fl-glucosidase, whereas a mixed culture of B. clausenii and S. cerevisiae provides better yields, rates, and concentrations if the enzyme is lower in fl-glucosidase. Another important element in the SSF process is the choice of substrate. Several cellulosic substrates have been evaluated in the SSF process, including sugar cane bagasse, rice straw, wheat straw, wood fractions, and paper mill byproducts (16-22). Although these substrates are all potenApplied Biochemistry and Biotechnology Vol. 28/29, 1991 SSF of Woody Crops to Ethanol 775 tially important, fast growing trees may prove economically attractive as substrates for ethanol production, and an important consideration is the acceptability of these fast growing woody crops for biological conversion to ethanol. Therefore, this project was undertaken to evaluate the most promising of these woody crops as substrates for the SSF process. Because high//-glucosidase activity has been shown for high yields by Spindler et al. (18), the cellulase enzyme was used alone and with/J-glucosidase supplementation to establish the highest possible cellulose conversions. MATERIALS AND METHODS Materials Four woody crops were employed in this study, Populus maximowiczii x nigra (Hybrid NE388), from Pennsylvania State University, trichocarpa x deltoides (Hybrid Nl l ) , from the University of Washington, Washington State, tremuloides (Aspen), from Colorado, and a native strain of Sweetgum Liquidambar styraciflua, from North Carolina State University. P. maximowiczii, P. trichocarpa, and S. liquidambar were supplied to us through the coordination of the biomass production laboratory of Oak Ridge National Laboratory (ORNL), Oak Ridge, TN. The fermentation yeasts used were S. cerevisiae (DsA), a SERI strain genetically derived from Red Star Brewers Yeast, and B. clausenff Y-1414, obtained from the U.S. Department of Agriculture (USDA) Northern Regional Research Laboratory (NRRL), Peoria, IL. Chemicals were purchased from the Sigma Chemical Company, St. Louis, MO, and yeast extract and peptone growth media were ordered from Difco, Detroit, MI. Cellulase enzyme came from Genencor Inc., San Francisco, CA, and fl-glucosidase (Novozyme-188) from NOVO Laboratories, Inc., Wilton, CT. Shaker flasks, 250-mL Pyrex graduated vessels, and Braun Biostat V fermentation vessels were used for the fermentations. Methods Shaker flask SSFs were carried out in 250-mL flasks outfitted with stoppers constructed to vent CO2 through a water trap. These flasks contained 100 mL of fermentation broth, and were agitated at 150 rpm in a shaker incubator at 37~ A 1% yeast extract and 2% peptone (w/v) media was used with a substrate loading of 7.5% (w/v) cellulose. A lipid mixture of ergosterol (5 mg/L) and oleic acid (30 mg/L) was added to the media for improved ethanol yield (23). Also, penicillin and streptomycin at 10 mg/L were used to minimize bacterial contamination. The inocula were grown in a shaker flask with (YP) media and 2% (w/v) glucose at 37~ and a 1/10 (v/v) yeast culture to total volume of media was added to the fermentation. The substrate was autoclaved in fermentation flasks, and sterile media, lipids, antibiotics, and enzyme were added before the inoculum. Applied Biochemistry and Biolechnology Vol. 28/29. 1991 776 Spindler, Wyman, and Grohmann Ethanol concentrations in the supernatant were measured by gas chromatography, using a Porapak Q80/100 column. The internal staildard was 4% isopropanol. For the larger scale 3 L SSFs, residual sugars (glucose and cellobiose) were determined as glucose by incubation of the sample with 2 mg/mL almond extract fi-glucosidase from Sigma for I h at 37~ and total sugars were measured on the model 27 glucose analyzer from Yellow Springs Instruments, Yellow Springs, OH. Viable cell densities were measured as colony forming units (CFU) by plating serial dilutions on YPD or YPC plates. Cellulase enzyme loadings of 7, 13, 19, and 26 IU/g cellulose substrate were used in the shake flask screening experiments to span the range of activity previously shown to be important for SSFs. In this paper, IU stands for international units of filter paper activity in micromoles of glucose/minute (24). fi-glucosidase enzyme was employed in this study at ratios of 1, 2, and 8 parts to I part of cellulase, as measured by IU of fi-glucosidase/IU of cellulase. The fi-glucosidase activity was determined by .Nitrophenyl-~-glucoside (PnPGU) assay at a temperature of 37~ because this is the temperature for the SSFs. The activity of cellulase increases with increasing temperatures to an optimum at 45~ (7,9), the temperature selected for saccharification without fermentation studies. The IUPAC revisions of measured cellulase activities indicate that the level of/~-glucosidase in an enzyme preparation may affect the results of the cellulase assay in filter paper units (24). Wood crops were completely debarked, and two 500 g batches of Wiley milled (2 mm screen) wood were pretreated with dilute sulfuric acid (0.45% v/v) in a 2 g Par Reactor. All woods were pretreated at 140~ for I h, with stirring at 185 rpm. After reaction, the wood slurries were washed several times with hot water in a large Buchner funnel lined with a linen sheet to bring the pH of 1.3 up to 4.5. These batches were combined, immediately placed in freezer storage bags, and stored at -20~ Approx 70% of the pretreated woods dry wt was found to be cellulose, 29% lignin and acid insoluble ash, and 1% xylan. Before pretreatment, approx 22% of the wood consists of lignin, 50% cellulose, and 14% xylan. Exact values are listed in Table 1 and will be addressed in the discussion. Shaker flask results are reported as percent of maximum theoretical ethanol yields, and do not account for substrate used in cell growth. Thus, the maximum expected ethanol yield is about 95%, assuming about 5% of the substrate is needed for cell growth. These calculations are based on the measured ethanol concentrations and a 56.7% theoretical ethanol yield conversion of cellulose to ethanol only. However, the saccharification with cellulose is reported on the basis of percent of the maximum amount of sugars possible. Thus, comparison of the SSF and straight saccharification results must consider that subsequent fermentation of the sugars produced in the latter will also result in about a 5% loss to cell growth and maintenance. Applied Biochemistry and Biotechnology VoL 28/29, 1991 SSF of Woody Crops to Ethanol 777