Pretreatment categories, process alternatives and material characteristics in enzymatic hydrolysis of lignocellulose:Dissertation

Fractionation of lignocellulose materials to sugars is a major strategy for the production of renewable fuels and chemicals. This study compares the potential of two major pretreatment categories, hydrothermal treatment and delignification, and contributes to scientific understanding of the phenomena behind enzymatic hydrolysability of wheat straw. Delignification was found to allow higher sugar yields. Since enzyme consumption is a key cost of the fractionation process, the optimal yield target depends on enzyme price. To allow yield optimization, a novel empirical model was developed for the process sugar yield as a function of enzyme consumption and hydrolysis time. The usability of the model was demonstrated by comparing the feasibility of different process alternatives for fractionation. The changes in the material properties of lignocellulose by pretreatments were correlated to cellulose hydrolysability, and for the first time, the importance of the different properties was determined statistically. In the order of importance, the hydrolysis yield depended on cellulose surface area, pore accessibility, lignin content, lignin surface chemistry, cellulose crystallinity and hemicellulose content. During enzymatic hydrolysis, the surface area of cellulose correlated linearly with the total cellulose content, but contrary to expectations, hydrolysis did not reveal fresh lignin surfaces. Different rate constraining mechanisms were incorporated in a Michaelis-Menten type kinetic model, and it was found that permanent hydrolysis-dependent enzyme inactivation should be included with the previously well-established effects of product inhibition and reduction of hydrolysability. For improving fractionation processes, different technological solutions were studied. A flow through process was found to improve fractionation by delignification, but no additional improvement was achieved by counter-current operation. By studying and simulating the packing density and flow properties of a packed straw bed, a flow-through process was found to be possible without clogging the straw bed by compaction. The height of an industrial scale column is restricted by the applicable flow rate. With the simulation model, it was possible to determine the maximum volumetric throughput as a function of column height. Recycling of the solid residue during enzymatic hydrolysis was found to be inefficient for enzyme recycling, but efficient for product removal, with similar benefits as sequential hydrolysis. Both processes significantly improved the volumetric productivity of hydrolysis by increasing the solids concentration without reducing yield. Alternatively, this benefit could be redirected into increasing the yield by maintaining reaction volume with additional water, leading to dilution of the hydrolysis conditions