Reaction Engineering Implications of Using Water for the Conversion of Lignocellulosic Biomass
Conversion of lignocellulosic biomass via hydrolysis of cellulose to simple sugars has failed to achieve economic competitiveness to produce renewable fuels and chemicals partly due to the inherent recalcitrance of the substrate and partly due to the use of non-recyclable catalysts. Solid acids have been proposed for cellulose hydrolysis as a recyclable alternative to enzymes and homogeneous acids. However, their catalytic mechanism has not been elucidated partly due to incomplete structural characterization. We focused on elucidating the structure of chloromethyl polystyrene based catalysts which exhibit remarkable activity towards hydrolyzing cellulose. By carrying out spatially resolved analysis of CMP-SO3H-0.3, a catalyst decorated with benzyl chloride and benzyl sulfonic acid groups, we discovered that the external surface of the catalyst is devoid of any chloride groups, which were hypothesized to interact with cellulose. Despite apparent greater reactivity than sulfonated-only catalysts, we found the CMP-SO3H-0.3 reacts with water at the reaction conditions used for cellulose hydrolysis, resulting in leaching of homogeneous hydrochloric acid, which in turn is responsible for the observed cellulose hydrolysis. Building on these results we investigated whether catalysts from various structural classes are stable in the hydrothermal environment or leach homogeneous acid. Surprisingly, we discovered that materials commonly used for cellulose hydrolysis are hydrothermally unstable and the leached homogeneous acid they produced was responsible for their apparent catalytic activity. On the other hand, hydrothermally stable materials did not exhibit greater hydrolysis activity than water. Cellulose crystallinity has been theorized for decades as a structural parameter determining the reactivity of cellulose, which motivated decrystallization pretreatment processes. However, water-induced recrystallization had not been accounted for in hydrolysis models, albeit being a well-documented phenomenon, and all hydrolysis processes use water as a reaction medium. By carrying out detailed structure-reactivity analysis we concluded that decrystallized cellulose undergoes a rapid transformation to an active crystalline cellulose, characterized by allomorphs I and II and greater content of surface polymer chains. Water-induced recrystallization reduced the reactivity of cellulose and prevented conversion of highly reactive amorphous regions. To circumvent the recrystallization pathway, we used ethanolysis as a means for rapid and selective depolymerization of amorphous cellulose. Ethanolysis of ball-milled cellulose for 30 minutes at 410 K resulted in 38% conversion, while hydrolysis at the same conditions in only 15%. Scission-relaxation caused recrystallization and limited conversion via ethanolysis. By using co-solvents capable of swelling cellulose, we were able to increase cellulose conversion to 48%. The results presented in those studies can guide future development of catalysts and depolymerization processes that circumvent the inhibiting effects caused by the use of water.