The effect of input material pretreatment on product yield and composition of bio-oils from LtL-solvolysis. A continuous process for organosolv fractionation of lignocellulosic biomass and solvolytic conversion of lignin

Abstract

As the world’s population and subsequent energy demand increases, there is a need to supplement existing energy technology with new and alternative approaches. Lignocellulosic biomass represents the vast bulk of terrestrial plant material and possesses both an enormous store of energy and a great potential as a source for biomass derived products. Production of bioethanol from the carbohydrate components of this biomass type is already established, while the remaining 10-25 % of the biomass, comprised by an amorphous phenylpropane copolymer called lignin, also holds unique characteristics. Lignin is the most important source of bio-based aromatics in nature, and lignin derived fuels or platform chemicals are approachable by lignin depolymerisation. Thermochemical conversion of lignin by Lignin-to-Liquid solvolysis depolymerises the lignin copolymer through hydrodeoxygenation and yields an energy rich bio-oil high in alkylphenols. In Lignin-to-Liquid solvolysis, formic acid and a co-solvent (ethanol or water) are added to the lignin, and the reaction mixture is exposed to a high temperature and high pressure as a closed system. A major focus within this thesis was to investigate the impact of initial feedstock species and feedstock fractionation and/or pretreatment method on yields of LtL-oil and LtL-oil composition. All feedstock species and pretreatment methods applied generated lignin rich fractions suitable for LtL-solvolysis. Multiple feedstocks were screened through systematic LtL-experiments with ethanol or water as co-solvent and early results lead to water being chosen as preferred solvent in consecutive experiments due to low cost, availability and its benign nature. Optimal substitution order of the generated phenols within the bio-oils depends on desired utilisation area, and ethanol-system experiments generated phenols with a more complex substitution order than water-system experiments. The produced bio-oils were high in aromatic content and water-system experiments produced phenolic components with similar substitution patterns regardless of feedstock preprocessing. The initial oxygen content of the feedstock used in LtL-solvolysis, e.g. due to carbohydrate residues from biomass fractionation, determined the bio-oil yield due to substantial depletion of oxygen through hydrodeoxygenation. This observation shifted the choice of feedstock towards lignin extracted by organosolv fractionation. Organosolv fractionation treats biomass with an organic solvent or mixtures of organic solvents and water to remove lignin. The lignin obtained is of low molecular weight and of high purity. Lignin extracted by organosolv fractionation provided high yields of biooil after LtL-solvolysis, and the yields also showed a positive correlation with the amount of formic acid in the reaction process. The O/C ratio of the phenolic monomers comprising the bio-oils displayed a reduction with increasing reaction temperature. As organosolv extracted lignin thus proved to be highly suitable for LtLsolvolysis, a process for continuous organosolv fractionation of lignocellulosic biomass and solvolytic conversion of lignin was proposed. A semi-continuous flow-through setup for organosolv fractionation was designed and optimal fractionation conditions were determined for a softwood mixture predominantly containing Norway spruce (and ~ 10 % pine). The extracted and isolated lignin was of high purity, in high yields and proved to be very well suited for LtLsolvolysis in subsequent LtL-experiments. LtL-solvolysis of lignin extracted by semi-continuously fractionated lignocellulose displayed high conversion ratios and yields of bio-oil. The biooils’ structural composition were investigated and quantified to examine the impact of experimental parameters and the bio-oils potential industrial employment. Alkylated phenols are presently being used as fuel additives, while phenols rich in oxygenated substituents are valuable for the chemical and pharmaceutical industry. Solvolysis experiments showed reproducible results with high mass recovery and gave a similar response to the reaction conditions as previously observed, confirming that an increased addition of formic acid input increased the bio-oil yield, and an increased reaction temperature reduced the O/C ratio (oxygen content) within the bio-oils. Quantification of the ten most abundant components identified in the oils showed their concentration to be mainly temperature dependent. Hence, tuning experimental conditions towards desirable bio-oil composition, and the development of methods to separate the bio-oils into series of homologs or similar compounds are both necessary and will strengthen the LtL-oils potential as a future platform chemical.