A FEASIBILITY ANALYSIS OF A NOVEL-APPROACH FOR THE CONVERSION OF XYLOSE TO ETHANOL

Economic production of ethanol from plant biomass could be significantly increased if the feedstock for the fermentation is more completely utilized. Currently, simple sugars (mostly D-glucose and D-xylose) can be recovered from lignocellulose by enzymatic or acid hydrolysis. However, while glucose can be readily converted to ethanol by yeasts, the xylose is not fermentable by many of the same species of yeasts that are able to convert glucose into ethanol. Nevertheless, xylose can be converted to its ketose isomer, xylulose, by the enzyme xylose isomerase and this isomer can be converted to ethanol. A major obstacle, however, in converting the xylose to xylulose and then simultaneously converting the xylulose to ethanol is that the pH at which xylose isomerase displays its optimal activity (pH of 7.0-8.0) is much different from the pH at which the fermentation of the xylulose and glucose is best carried out (pH of 4.0-5.0). Herein we propose a novel scheme to provide a means by which the isomerization and the fermentation can both take place simultaneously. The xylose isomerase is immobilized in a porous polymer pellet and this pellet is then coated with an additional layer in which the enzyme urease is immobilized. These bilayered pellets are dispersed in a fermentation broth which contains a prespecified concentration of urea in addition to all the other necessary ingredients. The xylose isomerase, immobilized in the core region of the pellet, catalyses the conversion of xylose to xylulose, and this latter substrate at the same time also becomes available for the fermentation reaction immediately. It is also possible to sustain the necessary spatial pH gradient between the bulk liquid phase and the core region of the pellet, because as hydrogen ions diffuse into the pellet they are neutralized by the ammonia produced in the outer layer of the pellet by the hydrolysis of urea by urease. A mathematical model is developed to demonstrate that the required pH gradient can be achieved. The model evaluates the effect of pellet design parameters (enzyme loading and thickness of the outer layer) and reactor conditions (bulk concentration of urea, and the effect of weak acids) on the ability to generate the required pH gradient across the outer layer of the pellet. © 1992, Taylor & Francis Group, LLC. All rights reserved.