(221c) Development and Validation of a Hemicellulose Depolymerization Kinetic Model for Aspen During Dilute Acid Hydrolysis

Research into ethanol production from lignocellulosic biomass has grown significantly over the last few decades, for it offers a potential solution to replace conventional fossil fuels while not competing with food production. Hemicellulose is a major component of lignocellulosic biomass, accounting for 25-35% of total dry mass. Thus, the efficient conversion of hemicellulose to fermentable sugars is vital to ethanol yield and optimizes the economic performance of the production process. A kinetic mechanism for hemicellulose hydrolysis is highly desired as a tool to understand and improve lignocellulosic biorefining and therefore, it continues to be modeled by researchers around the world.  The most common mechanism is a two-step pseudo first order irreversible reaction where xylan in hemicellulose is hydrolyzed directly to xylose, which is dehydrated subsequently to furfural and eventually tars.  However, oligomers are found to be a significant fraction of the product for dilute acid pretreatment, especially at short times, and oligomers are not taken up by fermenting microorganisms unless the oligomers are hydrolyzed further. In this study, a depolymerization model proposed by Lloyd and Wyman (Lloyd and Wyman 2003) is applied to describe the dilute acid hydrolysis of hemicellulose. Aspen samples were hydrolyzed by 0.5 wt% dilute acid at 150, 160 and 175°C. A HPLC coupled with the Hi-Plex Na column was used to measure the concentrations of xylose, xylooligomers, furfural and hydroxymethyl furfural (HMF) along the reaction time path. The depolymerization model successfully predicts the xylose profiles in dilute acid hydrolysis.  The reaction is faster in higher reaction temperature as indicated by the increasing hydrolysis rate k_h. Simultaneously, xylose is subjected to more severe degradation process, which is indicated by the increasing degradation rate k_d, which is confirmed by the increasing concentrations of the degradation products furfural and HMF. The kinetic model and experimental data agree on many aspects of the reaction trends; i. increasing temperatures lead to decreased time required for peak concentrations on all oligomer species, and ii. the time to reach peak concentrations of each oligomer decreased with degree of polymerization for all temperatures. However, the experimental data of xylooligomers were much lower than the model prediction, which may result from the model itself or the experiment data. New equipment and experiment procedures need to be implemented to obtain more accurate measurements of xylooligomers. The depolymerization model also needs to be modified to include direct degradation of xylobiose and xylotriose. In addition, the assumption that all bonds react at equal rates can be modified to include differences in end bonds, and also that hydrolysis rate depend on the chain length of xylooligomers.

References

Lloyd T, Wyman CE. Application of a depolymerization model for predicting thermochemical hydrolysis of hemicellulose. Applied Biochemistry and Biotechnology 2003;105:53-67