The shift from an oil based chemical industry to one based on renewable resources, such as biomass, is bound to ultimately happen. However, the environmental problems associated with the burning of fossil resources means that it is desirable to accelerate this shift. In order to compete with the available technologies for petrochemical refineries, efficient and feasible processes for converting biomass to commodity chemicals are needed. Conversion of biomass by pyrolysis processes is a desirable route due to the high volumetric production rates and the ability of processing a wide range of biomass feedstocks. Pyrolysis of aqueous sugars, as a component of biomass feedstock, in a fluidized-bed reactor is a promising way to produce potential platform chemicals such as glycolaldehyde and pyruvaldehyde. Through this method a wide range of substrates could be converted with a high selectivity towards C1-C3 oxygenates via retro-aldol condensations. These platform chemicals could further be used as precursors for the production of value-added chemicals such as ethylene glycol and acetic acid.
Current processes of converting carbohydrates, however, do not have full selectivity towards the platform chemicals, and give rise to a few percent of bigger C3-Cx molecules. The identity of these so called “by-products” are not fully known but is required for further optimizations of reactor operating conditions to maximize selectivity. Furthermore, the existence of by-products causes problems with the downstream processes for conversion of biomass to commodity chemicals. The oxygenated mixtures, obtained as the result of carbohydrate conversion processes, require suitable chemical characterization techniques that can screen their complex composition.
This PhD has tried to address these challenges by developing and validating a set of analysis techniques for identification of chemical constituents of these carbohydrate fragmentation products. First a liquid injection gas chromatography mass spectrometry method was developed and validated for a fast and an accurate quantification of the main product of carbohydrate conversion processes, glycolaldehyde. Through this method, glycolaldehyde, can be analyzed in under 6 minutes, quantified with a precision of above 96 % and an accuracy of above 90 %. This method can also be used for accurate quantification of other major components of the carbohydrate conversion, such as acetol and acetic acid.
Next, four sample preparation methods, namely, liquid-liquid extraction, static headspace, solid-phase microextraction and derivatization, interfaced with gas chromatography mass spectrometry, to inspect the full chemical composition of the sugar fragmentation mixtures, were developed and validated. These four methods, along with the non-preparative liquid injection method, were able to detect hundreds of peaks, 39 of which were fully identified using their commercial standards. An identification workflow was developed for identifying, categorizing and documenting all detected analytes. A wide range of compounds, ranging from C1-C12 oxygenates, with varying number of oxygens and functionalities were identified. Furanics, benzenes, acids and phenols were amongst the majority of the by-products observed, based on the number of peaks. This study also confirmed the existence of starting feed material – sugars – in the mixture. Production pathways to some of these by-products were also suggested.
Finally, this study aimed at using one of the developed methods, solid-phase microextraction, which could detect the most number of constituents, to investigate the different operating condition’s relationship with the relative abundances of the by-products observed in the mixture. These operating conditions were the fluidized-bed reactor temperature, feed type and concentration. This study could help the sugar conversion process optimizations by understanding which reaction parameter stimulates the production of which groups of low-concentration analytes. It was observed that sucrose gives rise to the highest relative abundance of by-products (mostly ring-membered compounds), contrary to xylose with the least relative abundance of by-products (mostly consisting of acids and other aliphatic C1-C3 oxygenates). High glucose feed concentrations and bed temperatures, lead to an increase in the relative abundance of furanics, phenols and benzenes (amongst the heavier by-products) in the headspace. In contrast, high fructose-fed bed temperatures lead to a decrease in the relative abundance of the heavier by-products.
Overall, this study has managed to peel back the first layer, in unveiling and understanding the carbohydrate fragmentation products. The glycolaldehyde quantification method can help with accurate tuning of the operating conditions to maximize the yield of this platform chemical. The library of constituents and their abundance variation analysis will help further sugar conversion process optimizations, as well as help avoid problems later in the downstream processes. More studies need to be carried out to fully apprehend the carbohydrate conversion mechanism to commodity chemicals, and the aforementioned studies are a small step towards this goal.