| Improving the efficiency of lignocellulosic ethanol production is of the utmost importance if cellulosic bioethanol is to be competitive with fossil fuels and first generation bioethanol from starch and sucrose. Improvements in individual processes (pretreatment, saccharification, fermentation) have been ongoing, but few researchers have considered the effect that the incoming raw biomass can have on the process. It is important to understand how biomass can be altered to provide the maximum yield of hydrolysable and fermentable sugars from whatever is available. Since the moisture content is highly variable and easily altered, the effect of drying and rewetting on bioconversion was studied on switchgrass, sugarcane bagasse and hybrid poplar. For switchgrass and sugarcane bagasse, the ethanol yield after simultaneous saccharification and fermentation was improved 18-24% by increasing the moisture content by soaking prior to pretreatment. It was also found that soaking had no effect when the samples were not catalyzed with SO2 confirming that the effect of moisture content is directly related to SO2 uptake and diffusion into the biomass. In hybrid poplar, the results were similar to herbaceous biomass for chips with less than 2% absorbed SO2. However, when the SO2 uptake was increased to 3% even the air dried chips exhibited high digestibility, indicating that increased SO2 uptake can overcome the poor diffusion in dried biomass.;Alongside controlling the biomass moisture content, improving knowledge and control of the processes can also increase efficiency and product yields. By monitoring reactions continuously with accurate, robust, on-line sensors, operators can detect when reactions deviate from the norm, and when they are complete. Avoiding process upsets and contamination could be the difference between an economically viable biorefinery and one that struggles to compete. Real time, continuous Raman spectroscopy was used to continuously monitor both a synthetic glucose and a lignocellulosic hydrolysate fermentation and measure glucose and ethanol. Models developed using offline HPLC validation samples had extremely high correlation between predicted and observed values for ethanol in both fermentations (R2 = 0.98 and 0.94 for synthetic and hydrolysate, respectively) while glucose proved more difficult to detect in the hydrolysate fermentation (R2 = 0.92 and 0.51). This work showed that it is possible to monitor the ethanol and glucose in a hydrolysate with a high fluorescent background. |
