Lignin valorization via thermal deconstruction and catalytic hydrodeoxygenation

Abstract

Lignocellulosic biomass contains three main biopolymers – namely, cellulose, hemicellulose and lignin. In the current biorefinery scheme, cellulose is converted to ethanol, hemicellulose is separated during the pretreatment step that allows for further processing into chemical building blocks, and lignin is typically used as a low-value fuel for heating boilers due to its recalcitrance toward biological and chemical transformations. Lignin valorization, therefore, could be an important venue to improve the economics of biorefinery units. One strategy to convert lignin into value-added chemicals is to thermally deconstruct it and then deoxygenate the vapors via chemical catalysis routes. Low-pressure hydrodeoxygenation (HDO) of lignin pyrolysis vapors appear to be an attractive and effective approach for deoxygenating these vapors before condensation. An approach for converting the oxygenated species produced during the pyrolysis of lignin is provided herein via low-pressure catalytic HDO over a bulk MoO3 catalyst. This catalyst was found to completely remove oxygen from the lignin pyrolysis vapors without hydrogenating the aromatic ring. Various lignin samples with different yields of oxygenated species in the pyrolysis vapors were used in this work. Nevertheless, MoO3 was able to funnel the complex mixture into a much simpler chemical stream containing aromatics, alkenes, and alkanes. Ideally, alkenes should be produced instead of alkanes since these unsaturated species are more valuable than alkanes and would require less hydrogen during the reaction. In addition, a stream of aromatics and alkenes is compatible with the current petrochemical infrastructure avoiding additional processing steps and resulting in lower costs of separation. Therefore, a series of γ-Al2O3-supported MoOx catalysts were synthesized and the effect of their properties, including acidity and MoOx speciation, was investigated on product selectivity. The formation of bulk-like MoO3 on γ-Al2O3 was found to be responsible for the majority of excessive hydrogenation of alkenes to alkanes. It was found that catalysts with lower MoO3 loading and higher acidity were more selective to alkenes (as opposed to alkanes) but at the same time side reactions such as alkylation and condensation made catalysts with high amounts of acid sites unsuitable for these reactions. While aromatics and alkenes are most compatible with the current infrastructure as drop-in chemicals, product diversification from lignin-derived species could help lignin valorization strategies through exploring other markets such as pharmaceuticals and specialty chemicals. Therefore, a section of this thesis evaluates such venues by designing bimetallic catalysts to modulate the selectivity of products from multi-functional molecules such as those found in lignin-derived species. Supported monometallic and bimetallic Pd catalysts were found to be active for hydrogenation of C=O and C=C aromatic bonds and hydrodeoxygenation of C-OH bonds to form a variety of products from acetophenone as a model compound. The addition of Fe and Ga drove the reaction toward C=O hydro(deoxy)genation to selectively produce ethylbenzene, which similar to previous projects could be used as a drop-in chemical. The addition of Li, however, considerably enhanced the hydrogenation of the phenyl group leading to the formation of new products with potential applications as specialty chemicals. The findings of this dissertation as well as the recommendations within would help in guiding future catalyst designs for converting lignin into value-added chemicals. The hope is to communicate the significance of large-scale lignin valorization strategies while keeping in mind the potential of implementing smaller scale transformations to maximize the value creation from lignin and help biorefineries become more competitive in the near future.