EXPLICIT REPRESENTATIONS OF BIOMASS PYROLYSIS OILS CHEMISTRY

Anika Afrin

Fast pyrolysis of lignocellulosic materials leads to the production of bio-oil, a complex chemical mixture with hundreds to thousands of organic compounds. The chemical composition of the oligomeric fractions is still poorly understood even though these oligomers play a significant role in coke formation during bio-oil upgrading. Separating these oils into fractions with better-defined chemical compositions and assigning structure is critical to developing explicit representations of their chemical structures. This study explores an integrated strategy where the experimental part explores method development for bio-oil separation based on chromatographic analysis and the modelling part develops a methodology to use the analytical data collected on each resulting fraction to assign representative chemical structures. BTG bio-oil was initially subjected to cold water precipitation to separate the water-soluble and water-insoluble fractions. The water-insoluble fractions were separated into CH2Cl2-soluble (also known as low molecular weight PL) and CH2Cl2-insoluble (known as low molecular weight PL) fractions. The separated subfractions were then characterized by HESI-FT-Orbitrap MS. The orbitrap MS data show that the WS subfractions were dominated by monomers and dimers, with molecular weights ranging from 100-300 Da. The WS subfractions show highly conjugated aromatic compounds eluted with THF and Methanol, while water eluted smaller, less conjugated sugar-like compounds. The HESI-MS data of WIS DCM-IS subfraction were dominated by monomers and dimers (≤300 Da), the presence of trimers (300-450 Da) and oligomers (>450 Da) can be observed, especially in the EA and IPA subfractions. This subfraction shows the presence of large, conjugated aromatic ring systems in Acetone and Methanol subfractions. Candidate structures for compounds for each fraction (open column and HPLC subfractions) were proposed. Subsequent steps involved column chromatography using solvents of increased polarity. A semi-preparative HPLC was utilized to fractionate the oil further. Preparatory HPLC results also show a substantial difference in the residence time of the fractions obtained depending on the solvent used.Using the data of structural assignments, a database of bio-oil compounds was developed. An NMR spectra database was developed by simulation, and the theoretical spectra were compared to the experimental NMR spectra. The model prediction lacked phenolics, or aromatic-rich lignin-derived compounds, compared to the experimental spectrum. A database of thermophysical properties was also developed. Based on the information from the database, an initial objective function (with concentration value assigned to the initial fit) has been derived to develop an algorithm for the explicit representation of bio-oil.

EXPLICIT REPRESENTATIONS OF BIOMASS PYROLYSIS OILS CHEMISTRY

Files and links (1)

Abstract

Metrics

Details