Improving enzymatic conversion of lignocellulose to platform sugars:Dissertation

Increasing demand and uncertain availability of fossil fuels urge us to find alternative resources available in large quantities especially for the petrol-based transportation sector. Lignocellulosic biomass, available worldwide in plant cell walls, is a promising alternative feedstock. It can be depolymerised to sugar monomers, which provide potential raw material for sugar platform-based production of fuels and chemicals. However, the enzymatic saccharification of lignocellulose to platform sugars is hindered primarily by the complexity of lignocellulosic substrates as well as by the performance of the hydrolytic enzymes involved. This study focuses on various rate limiting factors such as the decrease in the reactivity and accessibility of the substrates which slow down the hydrolysis, on auxiliary enzymes needed for the efficient solubilisation of cellulose, as well as on the adsorption of enzymes. Consequently, solutions to these limitations were sought to improve the efficiency of biomass conversion processes. Following the morphological and structural changes in the substrate during hydrolysis revealed that the average crystal size and crystallinity of cellulose remained constant while particle size generally decreased (Paper I). In particular, cellulose microfibrils were proposed to be hydrolysed one-by-one in fibre aggregates by peeling off cellulose chains layer-by-layer from the outer crystals of microfibril aggregates. Microscopic observation showed that almost intact particles remained in the residue even after 60% conversion. Lignocellulose is a complex network of lignin and polysaccharides. Lignin was found to impede the hydrolysis of cellulose, and its extensive removal doubled the conversion yields of softwood (Paper II). On the other hand, accumulation of lignin during hydrolysis did not affect hydrolysability by commercial cellulase preparations. Residual hemicelluloses, especially glucomannan, were resistant to enzymatic hydrolysis but could be removed together with lignin during delignification. This suggests that especially glucomannans are bound to lignin as lignin-carbohydrate complexes. In addition, cellulose, xylan and glucomannan were shown to be structurally interlinked in softwood (Paper IV). The hydrolysis yield of these polysaccharides remained below 50% without the simultaneous hydrolysis of all polysaccharides. Synergism between the solubilisation of cellulose and hemicelluloses was found, and the release of glucose, xylose and mannose was in linear correlation. The adsorption and desorption of enzymes were followed during hydrolysis (Paper III). After a quick initial adsorption, slow desorption and re-adsorption of enzymes was observed in alkaline delignified spruce. On the other hand, unproductive adsorption to lignin as well as enzyme inactivation was predicted to play a primary role in the irreversible adsorption of cellulases in steam pretreated spruce or Avicel during hydrolysis when no desorption of cellulases could be detected. This study showed for the first time that increasing substrate concentration could compensate for the absence of carbohydrate binding modules (CBMs) in hydrolytic enzymes (Paper V). The performance of cellulases lacking CBMs was comparable to that of cellulases comprising CBM at 20% substrate concentration. At the same time, over 60% of the enzymes without CBMs could be recovered at the end of the hydrolysis. Thus, the major part of hydrolytic enzymes without CBMs could potentially be recovered in industrial high consistency processes