Luterbacher, Jeremy2013-01-312017-12-202012-08-20bibid: 7959799https://hdl.handle.net/1813/31068Sustainably producing concentrated solutions of carbohydrates is a key bottleneck in the conversion of lignocellulosic biomass to biofuels or bioproducts. Most pretreatment and enzymatic hydrolysis processes used for biomass depolymerization are run at low-solid concentration (<10 wt%) and use chemical catalysts, while high-solids enzymatic hydrolysis reactions are almost always performed with air-dried pretreatment mixtures. Biphasic H2 O-CO2 mixtures are an interesting alternate medium for high-solids (up to 40 wt%) pretreatment. Initial studies were done in a small (25 ml) unstirred reactor using a single temperature stage. More recently, two-temperature stage pretreatment was introduced and optimized in a larger 1 L stirred reactor to take advantage of the biomass depolymerization temperature dependent reaction sequence. Optimally pretreated substrates were then used as feedstock in high-solids (30 wt%) enzymatic hydrolysis reactions that gave glucose yields above 80% for both switchgrass and hardwood after 48 hours of hydrolysis. Therefore, without additional chemical catalysts or any drying, two-temperature stage H2 O-CO2 pretreatment coupled with high-solids enzymatic hydrolysis can produce monosaccharide solutions of 185 g/L and 148 g/L for mixed hardwood and switchgrass, respectively. This suggests that H2 O-CO2 pretreatment is an attractive alternative to chemically catalyzed processes such as dilute acid pretreatment. Parallel to these studies, efforts were undertaken to better understand the relationship between the effects of pretreatment and the enzymatic depolymerization mechanisms of cellulosic substrates. A fluorescence confocal microscopy method was developed for observing and measuring the binding and reaction of cellulase cocktails and their substrates in situ. The Spezyme CP cellulase cocktail was supplemented with a small fraction of fluorescently labeled Trichoderma Reseii Cel7A, which served as a reporter to track cellulase binding onto the internal physical structure of bacterial microcrystalline cellulose. A kinetic model was constructed using the variation in fluorescence intensity of the substrate and the bound enzyme over time and was successfully used to predict reaction yields in solution. Building on this result a reaction and diffusion model was developed to model the enzymatic hydrolysis of pretreated biomass. This model was shown to be able to predict the well-known relationship between accessible surface area in biomass and initial enzymatic hydrolysis rates. The development of this theoretical framework that accurately describes the key relations between pretreatment and enzymatic hydrolysis could be the first step in developing tools that could lead to the rational design of pretreatment technologies.en-USBiomasspretreatmentsupercritical CO2Biofuelsenzymatic hydrolysisporeModelingaccessibilitydissertation or thesis