Search

..................................

Select Level of Detail

At a Glance

General

Technical

............................

+ Ethanol Production - Ethanol Dry Grind Process

............................

Access BioWeb Content

Search

Explore By Topic

Browse Index

............................

BioWeb Glossary

Search

Alphabetical Listing

............................

Contributors Log in

bioweb.sungrant.org » Technical » Biofuels » Technologies » Ethanol Production » Ethanol Dry Grind Process

Ethanol—Dry Grind Process	Printer Friendly

In 2006, more than 5 billion gallons of fuel ethanol were produced in the U.S. Most (> 90%) was produced from corn, and used some 2.15 billion bushels, about 20% of the corn harvest in 2006 (RFA, 2006; USDA, 2006). Corn is largely composed of starch which can be hydrolyzed (broken down) into sugars and fermented to ethanol using yeast. U.S. commercial production of corn ethanol uses either a dry-grinding process or a wet milling process. These processes differ with respect to their complexity and associated capital costs, the numbers and types of co-products produced, and the flexibility to produce different kinds of primary products. Dry-grind is the most prevalent process and much of the current expansion of the industry uses this technology as it is a simpler process with lower capital costs relative to wet milling and produces high ethanol yields (2.7 to 2.8 gal/bu corn).   Ethanol is produced by fermenting plant carbohydrates with yeast. Plant carbohydrates are grouped as soluble sugars (such as sucrose from sugarcane), storage carbohydrates (such as starch from grains and tubers), and structural carbohydrates which make up the plant cell wall, such as cellulose, hemicellulose, and pectin. The principal carbohydrate contained in corn grain and other grain crops such as wheat, barley, rice, and oat, is starch (table 1).                          Two major forms of starch exist - amylose and amylopectin (BeMiller, 1996). Amylose is a straight chain polymer of glucose molecules joined by alpha (1-4) glycosidic bonds (figure 1). This primary structure results in the long polymers coiling into a helical conformation (BeMiller, 1996).                           Amylopectin is also primarily a straight chain of glucose but also includes branches occurring every 24 to 30 glucose units consisting of alpha (1-6) bonds (figure 2).                Starch is semi-crystalline and transitions to an amorphous state (a gel) at 60-70°C, through a gelatinization process where water molecules disrupt the hydrogen bonds within and between starch molecules. Starch, especially gelatinized starch, can be easily hydrolyzed to yield the individual glucose molecules.   Producing fuel ethanol from grains using the dry-grind process involves four major steps—preparation of the grain (grinding, liquefaction, and saccharifaction), fermentation of the sugars, recovery of the ethanol, and handling (centrifugation, evaporation, and drying) of the coproducts (figure 3). While a number of grains can be used to produce fuel ethanol, the discussion will focus on the use of corn, as it is the predominant grain used in the U.S.

Grain Preparation
Preparation of the corn grain involves cleaning and conditioning steps, as well as generating an aqueous solution high in simple sugars. Generally, enough grain is stored on site in bins to meet facility needs for 8-12 days of operation (Kwiatkowski, 2006). Broken corn kernels and foreign materials (metal, dirt, cobs, etc.) are removed by blowers and screens. The cleaned corn is then ground in hammer mills fitted with screens with openings ranging between 3.2 and 4.8 mm in diameter, which provides grain particles of a more uniform size: more than 90% of the ground corn, by weight, has a diameter between 0.5 to 2 mm (Rausch, 2005). Grinding serves to break the tough outer coating of the corn kernel and increases the surface area of exposed starch. Liquefaction involves combining the ground corn with process water to form a slurry which is approximately 30% solids by weight (Kwiatkowski, 2006). Ammonia and lime are added at this step to adjust the pH of the slurry to 6.5. The ammonia, which contains nitrogen, also serves as a nutrient for the yeast in the subsequent fermentation step. The slurry is heated to 88°C by direct steam injection using a “jet-cooker”. A thermostable enzyme (alpha-amylase) is added to cleave the starch molecules at random points along the middle of the polymer chain and to break the starch into smaller water soluble fragments called dextrins. After approximately one hour, the output from the first step of liquefaction is combined with “backset”, which is recycled water from the end of the ethanol distillation process. The backset accounts for approximately 15% of the final volume of the corn mash (McAloon, 2000). Critical nutrients for the yeast are also carried in the backset. As the liquefied slurry is cooled to 60°C, the heat is recovered and used to heat new, incoming slurry going to the jet-cooker (Kwiatkowski, 2006). A new enzyme technology developed by Genencor allows for the rapid hydrolysis of granular starch and eliminates the need for gelatination of the starch slurry by jet-cooking, thus significantly lowering the energy requirements for ethanol production from corn (Shetty, 2005). Following liquefaction, sulfuric acid is added to the slurry to lower the pH to 4.5. An additional enzyme, glucoamylase (also called beta-amylase) is added to break the starch and dextrins into glucose via a stepwise hydrolysis of glucose from the end of the molecules. The slurry is held at 60°C for 5-6 hours as the glucoamylase hydrolyzes the dextrins to fermentable glucose (Schenck, 2002). Most of the dextrins are converted to glucose during this step, however the glucoamylase remains active throughout the fermentation step and will continue to hydrolyze any residual dextrins during fermentation. After saccharification, the slurry (now called mash) is cooled to 32°C with the heat recovered and transferred to other process streams. The cooled mash then enters the fermentation tanks. A popular alternative to mash-presaccharification is to add glucoamylase during the filling of the fermentor and to saccharify and ferment the starch simultaneously (SSF, Simultaneous Saccharification and Fermentation). An additional advantage to this approach is that reversion reactions (re-polymerization of glucose) are much less likely to occur (Power, 2003).

Sugar Fermentation
Fermentation uses microorganisms (the yeast Saccharomyces cerevisiae, which is also used for brewing beer and baking bread) to convert sugars to ethanol, and results in the production of ethanol and carbon dioxide, as well as additional yeast cells (from cell division) and heat. One molecule of glucose yields 2 molecules of ethanol and 2 molecules of carbon dioxide. On a mass basis, one kilogram of glucose will theoretically produce 0.51 kilogram of ethanol and 0.49 kilogram of carbon dioxide. However, glucose consumed to generate additional yeast cells (cell mass) does not result in the production of ethanol. Most industrial fermentation processes operate at 90 to 95% of the theoretical yield. In the fermentation step, yeast grown in seed tanks is added to the corn mash to ferment the simple sugars (glucose) to ethanol.  The other components of the corn kernel (protein, oil, etc.) remain largely unchanged during the fermentation process, though the corn oil helps to prevent foaming during the fermentation. In most dry-grind ethanol plants, the fermentation process occurs in batches. A fermentation tank is filled and ferments completely before being drained and refilled with a new batch. The up-stream processes (grinding, liquefaction, and saccharification) and downstream processes (distillation and recovery) occur continuously. Thus, dry-grind facilities using batch processing usually have three or more fermentors, with one fermentor filling, one fermenting (for approximately 46-68 hours), and one emptying and resetting for the next batch all at the same time. While the exact size of each fermentor varies between plant designs, common fermentor sizes range between 300,000–500,000 gallons (1-2 million liters) each (Kwiatkowski, 2006). A major advantage of batch fermernations is that there are fewer opportunities for contamination, provided they are properly sanitized between runs. Bacteria, especially species of Lactobacillus, can infect yeast fermentations and produce organic acids that lower ethanol yields and interfere with the Saccharomyces (Graves, 2006). While batch fermentation is more common in dry-grind facilities, continuous fermentation processes are used in some facilities. In this process, fermentation occurs through a series of cascading tanks where the liquid continuously flows through the process. New fermentation media is continuously added at the front end and fermented product is continuously removed from the back end. While continuous fermentation has greater reactor productivity because it is continuously operating with high yeast loads, much more care needs to be exercised to prevent contamination (Bayrock, 2001). In addition to ethanol, carbon dioxide is also produced during fermentation. Usually, the carbon dioxide is not recovered as a sellable product. If recovered, this carbon dioxide can be cleaned, compressed and sold for carbonation of soft drinks or frozen into dry-ice for cold product storage and transportation. If the carbon dioxide is not recovered, it passes through a water scrubber to remove evaporated ethanol and other volatile organic compounds (VOCs) carried in the gas. The water from the scrubber, containing the recovered ethanol, is sent to the distillation system. The cleaned carbon dioxide is vented to the atmosphere. Heat is generated during fermentation: approximately 12000 kJ per kilogram of ethanol; 516 BTU per pound (Kwiatkowski, 2006). This heat must be continuously removed from the fermentors and is accomplished either by passing water through a cooling coil contained within the fermentor, or by continuously pumping the fermenting mash through a large heat exchanger where the heat is transferred to cooling water before the mash is returned to the fermentor. Failure to remove the heat causes a rise in temperature sufficient to kill the yeast. The optimal temperature for ethanol fermentation is between 27 and 32°C. After the fermentation is nearly complete, the fermented corn mash (now called beer) produced in individual batches is emptied from the fermentor into a beer well where it is stored enabling a continuous stream to be supplied to the ethanol recovery system. The beer contains 8-10% ethanol by weight.

Ethanol Recovery
Separation and recovery of the ethanol is accomplished through a continuous process involving several steps (figure 4). In the first step, the beer is processed through a beer column where steam is used to strip off almost all of the ethanol, along with some water, from the slurry. The ethanol and water vapor exit the top of the beer column and the whole stillage (containing less than 0.1% ethanol by weight) exits from the bottom. The overhead vapor flows to a rectifier column where the ethanol is concentrated from 45% to 91% through fractional distillation. The bottoms from the rectifier pass through a stripping column to remove residual ethanol.  Liquid exiting the bottom of the stripper has less than 0.1% ethanol by weight, and is recycled as process water for slurrying the ground corn.  The overhead vapor from the rectifier (91% ethanol by weight) is superheated and passes through molecular sieves. The final product from the molecular sieve system is ethanol vapor that is at least 99.6% pure. This vapor is condensed and mixed with a denaturant (e.g. gasoline) to render it as non-potable fuel ethanol. Generally, 8 to 12 days worth of denatured fuel ethanol production is stored on site (Kwiatkowski, 2006). All industrial fuel ethanol production uses continuous-feed distillation column systems. Distillation is a common chemical separation method that is based upon differences in volatility (Wankat, 1988). If a mixture of ethanol and water are placed in a container at a given temperature and pressure, after time, the mixture will reach equilibrium. At equilibrium, some of the ethanol will be a vapor in the gas above the liquid and some will be in the liquid phase. Similarly, some of the water will be in the vapor phase and some will be in the liquid phase. Because ethanol is more volatile than water (boils at a lower temperature), the ratio of ethanol to water in the vapor phase is greater than the ratio of ethanol to water in the liquid phase. This characteristic allows for the separation of the ethanol from the water. Through subsequent vaporization of the mixture, condensation, re-vaporization, and re-condensation, the mixture becomes higher and higher in ethanol content because the vapor at each vaporization step has higher ethanol concentration than the liquid from which it was vaporized. Thus, multiple fractionation steps can be used to purify ethanol from water. However, the basic principle by which this occurs – difference in boiling points between water and ethanol – ceases to exist when the mixture is 95.6 wt% ethanol (4.4% water). At this point (the azeotrope), the ethanol and water both vaporize to the same degree and cannot be further fractionated by distillation. Either a third solvent can be introduced to break the azeotrope (e.g. beneze) or an alternate seperation method can be used such as absorption of water using molecular sieves. Molecular sieves used to dry ethanol are crystalline metal zeolites (aluminosilicates) with a 3-dimensional porous structure of silica and alumina tetrahedra. Zeolites strongly and preferentially adsorb water from vapor/gas mixtures. The adsorbed water can be removed by increasing the temperature of the zeolite and passing dry gas over the particles, thus allowing this rather expensive desiccant (~$10/lb) to be reused (Al-Asheh, 2004). While this drying property was discovered with naturally occurring zeolites, commercial molecular sieves are synthetically produced to have highly uniform pores within a tight size distribution. Industrial molecular sieve drying systems consist of multiple columns each filled with a bed of uniform sized zeolite sieves. The ethanol/water vapor mixture leaving the fractional distillation system is super-heated and forced through the molecular sieve bed. The water vapor is selectively adsorbed to the particles while ethanol passes through the column, where it is recovered and condensed to liquid at high purity. After the capacity of the zeolites to adsorb water from the ethanol vapor is reached, feed vapor is stopped and the flow through the column is reversed. Dry gas (usually CO2 produced by the fermentation process) passes over the zeolites while the system is placed under a slight vacuum to drive the desorption of the water from the solid particles. The water vapor and residual ethanol vapor exiting the column is condensed and returned to the stripper in the distillation system to re-vaporize the residual ethanol which improves the overall efficiency of ethanol recovery by the plant (Ladisch, 1979). To achieve continuous processing, molecular sieve dehydration systems consist of pairs of beds. As the first bed in the pair processes wet ethanol, a second molecular sieve bed undergoes regeneration to remove the adsorbed water. When the capacity of the first column to remove water is filled, the duties of the columns are switched so that the wet column begins regeneration and the fresh column continues to process wet ethanol vapor. An alternative to zeolite molecular sieves is using ground corn (corn grits) in a packed bed, similar in design to conventional molecular sieve beds (Chang, 2006; Neuman, 1986; Westgate, 1992). In this system, currently in commercial use by Archer Daniel Midlands, water is selectively adsorbed to corn starch from a water-ethanol vapor mixture (Beery, 1998; Beery, 2001; Ladisch, 1979). Similar in design to the zeolite molecular sieve beds, the corn grit system operates in pairs of beds with one drying ethanol vapor while the other(s) are undergoing regeneration to remove the adsorbed water. The major advantages of the bio-based adsorbents that they are easily available, less expensive than molecular sieves, mechanically stable, and easily disposable (Ladisch, 1997). Industrial fractional distillation to produce fuel ethanol is one of the major energy inputs for the production of fuel ethanol. Process improvements that capture and recycle energy from the process have greatly reduced the cost (Gulati, 1996; Ladisch, 1979).

Co-product production and handling
The principal co-product from ethanol production using the dry-grind process is distiller’s dried grains and solubles (DDGS). The whole stillage leaving the bottom of the beer column contains approximately 15% solids. Centrifugation of the whole stillage removes approximately 83% of the water and results in wet distillers grains (also called wet cake) which is 35 to 40% solids (figure 3). The liquid stream, called thin stillage, is partially recycled as backset to the second stage of the liquefaction process. The remaining thin stillage passes to a surge tank which supplies a steady feed to the evaporators, where it is concentrated. Thin stillage passes through a multiple effect evaporator which removes a significant amount of the water as steam, which is used to vaporize the ethanol during the recovery process (in the reflux of the rectifier). The steam, now condensed to liquid, is mixed with other condensates and added to the ground corn at the beginning of the grain preparation phase. The concentrated product of the evaporators is a syrup containing 55% solids by weight (McAloon, 2000) which is mixed with wet distillers grains and sent to a large rotary drum dryer where the mixture is dried from 64 to 9-10% moisture to form the DDGS. The hot gas from the dryers is processed prior to venting to the atmosphere to remove volatile organic compounds (VOCs) released during drying. Thermal oxidation is commonly used to convert VOCs to carbon dioxide and water (Vij, 2003). Distillers dried grains and solubles are currently used as livestock feed as they contain relatively high quantities of protein (table 2). They have long been used in cattle feed rations at rates up to 40% (Ham, 1994; Peter, 2000), but not as widely used in feed rations for non-ruminants (swine and poultry). Recommended inclusion rates (based on amino acid content) for swine are up to 20% of the ration (Whitney, 2006), but are less than 10% for laying hens (Lumpkins, 2005). As fuel ethanol production from corn dry-grind technology continues to increase, efficient use and capturing the value of the co-products becomes increasingly important (Rausch, 2006; Belyea, 2004).      In addition to protein, DDGS also contains substantial quantities of fiber, which is largely cellulose and hemicellulose (M. Ladisch et al., 2006), Integrating cellulose conversion technologies in dry-grind facilities might further increase the value of DDGS by lowering its fiber content and increasing its relative protein content, as well as increasing the ethanol yield per bushel of corn (Mosier, 2005). Good handling and bulk flowablity properties are critical for efficient transportation of DDGS to markets by barge, rail, or truck (Ganesan, 2006; MCGA, 2005) and improvements are needed. Poor flowability properties mean that the DDGS tends to cake or bridge and not flow easily by gravity through a chute or opening. Improved methods for measuring nutritional quality are currently under development to help insure feed quality for DDGS in the marketplace (American Feed Industry, 2007). Using the dry-grind process, 2.7 – 2.8 gallons of ethanol and 17 pounds of DDGS are typically obtained per bushel of corn. References Al-Asheh, S.; Banat, F. and Al-Lagtah, N. (2004). Separation of ethanol-water mixtures using molecular sieves and biobased adsorbents. Chemical Engineering Research and Design, 82(A7), 855-864. American Feed Industry Association. Evaluation of Analytical Methods for Analysis of Dried Distillers Grains with Solubles. (2007). Bayrock, D. and Ingledew, W. M. (2001). Changes in steady state on introduction of a Lactobacillus contaminant to a continuous culture ethanol fermentation. Journal of Industrial Microbiology and Biotechnology, 27(1), 39-45. Beery, K. E.; Gulati, M.; Kvam, E. P. and Ladisch, M. R. (1998). Effect of enzyme modification of corn grits on their properties as an adsorbent in a skarstrom pressure swing cycle dryer. Adsorption-Journal of the International Adsorption Society, 4(3-4), 321-335. Beery, K. E. and Ladisch, M. A. (2001). Chemistry and properties of starch based desiccants. Enzyme and Microbial Technology, 28(7-8), 573-581. Belyea, R. L.; Rausch, K. D. and Tumbleson, M. E. (2004). Composition of corn and distillers dried grains with solubles from dry grind ethanol processing. Bioresource Technology, 94(3), 293-298. BeMiller, J. N., and Whistler, R. L. (1996). Carbohydrates. In Food Chemistry (3rd ed., pp. 157-223.). New York, New York: Marcel Deker. Chang, H.; Yuan, X. G.; Tian, H. and Zeng, A. W. (2006). Experimental investigation and modeling of adsorption of water and ethanol on cornmeal in an ethanol-water binary vapor system. Chemical Engineering and Technology, 29(4), 454-461. Ganesan, V., Muthukumarappan, K. and Rosentrater, K. A. (2006). Effect of Flow Agent Addition on the Physical Properties of DDG with Varying Moisture Content Soluble Percentages. Paper presented at the 2006 ASABE Annual International Meeting, Portland, OR. Graves, T.; Narendranath, N. V.; Dawson, K. and Power, R. (2006). Effect of pH and lactic or acetic acid on ethanol productivity by Saccharomyces cerevisiae in corn mash. Journal of Industrial Microbiology and Biotechnology, 33(6), 469-474. Gulati, M.; Westgate, P. J.; Brewer, M.; Hendrickson, R. and Ladisch, M. R. (1996). Sorptive recovery of dilute ethanol from distillation column bottoms stream. Applied Biochemistry and Biotechnology, 57-8, 103-119. Ham, G. A.; Stock, R. A.; Klopfenstein, T. J.; Larson, E. M.; Shain, D. H. and Huffman, R. P. (1994). Wet Corn Distillers by-Products Compared with Dried Corn Distillers Grains with Solubles as a Source of Protein and Energy for Ruminants. Journal of Animal Science, 72(12), 3246-3257. Heitz, M.; Capekmenard, E.; Koeberle, P. G.; Gagne, J.; Chornet, E.; Overend, R. P. et al. (1991). Fractionation of Populus-Tremuloides at the Pilot-Plant Scale - Optimization of Steam Pretreatment Conditions Using the Stake-Ii Technology. Bioresource Technology, 35(1), 23-32. Kwiatkowski, J. R.; Mcaloon, A. J.; Taylor, F. and Johnston, D. B. (2006). Modeling the Process and Costs of Fuel Ethanol Production by the Corn Dry-Grind Process. Industrial Crops and Products, 23, 288-296. Ladisch, M.; Tyner, W.; Mosier, N.; Cotta, M.; Dien, B.; Blaschek, H. et al. (2006). Research challenges and opportunities for cellulose conversion technology in a dry mill pathway. Paper presented at the 28th Symposium on Biotechnology for Fuels and Chemicals, Nashville, TN. Ladisch, M. R. (1997). Biobased adsorbents for drying of gases. Enzyme and Microbial Technology, 20(3), 162-164. Ladisch, M. R. and Dyck, K. (1979). Dehydration of Ethanol: New Approach Gives Positive Energy Balance. Science, 205, 898-900. Lumpkins, B., Batal, A. and Dale, N. (2005). Use of distillers dried grains plus solubles in laying hen diets. Journal of Applied Poultry Research, 14(1), 25-31. McAloon, A.; Tayor, F.; Yee, W.; Ibsen, K. and Wooley, R. (2000). Determining the Cost of Producing Ethanol from Corn Starch and Lignocellulosic Feedstocks. NREL/TP-580-28893. MCGA, A. a. (2005). Methods to improve the flowability and pelleting of distillers dried grains with solubles (DDGS) In AURI and Minnesota Corn Growers Distillers Dried Grain Flowability Study Summary. Mosier, N. S.; Hendrickson, R.; Brewer, M.; Ho, N.; Sedlak, M.; Dreshel, R. et al. (2005). Industrial scale-up of pH-controlled liquid hot water pretreatment of corn fiber for fuel ethanol production. Applied Biochemistry and Biotechnology, 125(2), 77-97. Neuman, R.; Voloch, M.; Bienkowski, P. and Ladisch, M. R. (1986). Water Sorption Properties of a Polysaccharide Adsorbent. Industrial and Engineering Chemistry Fundamentals, 25, 422-425. Peter, C. M.; Faulkner, D. B.; Merchen, N. R.; Parrett, D. F.; Nash, T. G. and Dahlquist, J. M. (2000). The effects of corn milling coproducts on growth performance and diet digestibility by beef cattle. Journal of Animal Science, 78(1), 1-6. Power, R. F. (2003). Enzymatic conversion of starch to fermentable sugars. In Jacques, K. A.; Lyons, T. P. and Kelsall, D. R.  (Eds.), The Alcohol Textbook, 4th Edition (pp. 23-32). Bath, England: Nottingham University Press. Rausch, K. D. and Belyea, R. L. (2006). The future of coproducts from corn processing. Applied Biochemistry and Biotechnology, 128(1), 47-86. Rausch, K. D.; Belyea, R. L.; Ellersieck, M. R.; Singh, V.; Johnston, D. B. and Tumbleson, M. E. (2005). Particle size distributions of ground corn and DDGS from dry grind processing. Transactions of the Asae, 48(1), 273-277. Renewable Fuels Association. 2006. Ethanol Industry Outlook. Renewable Fuels Association. Schenck, F. W. (2002). Starch hydrolysates - An overview. International Sugar Journal, 104(1238), 82-+. Shetty, J. K.; Lantero, O. J. and Dunn-Coleman, N. (2005). Technological advances in ethanol production. International Sugar Journal, 107(1283), 605-+. U.S. Department of Agriculture. U.S. Grains Supply and Distribution: Wheat, Corn, Sorghum, Barley, Oats, Rye, and Rice. (2006). Vij, A. (2003, November). A Perspective on VOC Control Technology in the Industry. Ethanol Producer Magazine. Wankat, P. C. (1988). Introduction to Column Distillation. In Equilibrium-Staged Separations. New York, NY: Prentice Hall PTR. Watson, S. A. (1987). Structure and Composition. In Watson, S. A.  and Ramstad, Paul E. (Ed.), Corn: Chemistry and Technology (pp. 53-82). Minneapolis, MN: American Association of Cereal Chemists. Westgate, P.; Lee, J. Y. and Ladisch, M. R. (1992). Modeling of Equilibrium Sorption of Water-Vapor on Starch Materials. Transactions of the Asae, 35(1), 213-219. Whitney, M. H.; Shurson, G. C.; Johnston, L. J.; Wulf, D. M. and Shanks, B. C. (2006). Growth performance and carcass characteristics of grower-finisher pigs fed high-quality corn distillers dried grain with solubles originating from a modern Midwestern ethanol plant. Journal of Animal Science, 84(12), 3356-3363. Wood, T. M. and Saddler, J. N. (1988). Increasing the Availability of Cellulose in Biomass Materials. Methods in Enzymology, 160, 3-11. Xue, X. D. and Ho, N. W. Y. (1990). Xylulokinase Activity in Various Yeasts Including Saccharomyces-Cerevisiae Containing the Cloned Xylulokinase Gene. Applied Biochemistry and Biotechnology, 24-5, 193-199. Zaldivar, J., Nielsen, J., and Olsson, L. (2001). Fuel ethanol production from lignocellulose: a challenge for metabolic engineering and process integration. Applied Microbiology and Biotechnology, 56(1-2), 17-34. Zhang, M., Franden, M. A., Newman, M., McMillan, J., Finkelstein, M., and Picataggio, S. (1995). Promising Ethanologens for Xylose Fermentation - Scientific Note. Applied Biochemistry and Biotechnology, 51-2, 527-536.