Experimental and computational analysis of acid catalysts for biomass conversion
Courtney, Timothy D.
University of Delaware
The Earth's growing population and limited supply of fossil fuels implore us to develop alternatives in order to maintain our quality of life far into the future. This need is only exacerbated by mounting evidence that sustained fossil fuel use is changing our climate with trillions of dollars in potential damage. Wind, solar, and nuclear power can help to address our energy need, but fossil fuels are also an invaluable feedstock for commodity chemicals and the production of consumer goods. A future driven by renewable energy therefore requires renewable routes to replace these chemicals. Biomass feedstocks are highly oxygenated, enabling a rich and diverse chemistry with immense potential, but also making it difficult to control reaction selectivity. Chemical engineers have spent 100 years learning to control the reaction network of fossil fuels, and must now extend that control to biomass. Catalysis is critical to the economic feasibility of these processes not only to accelerate reactions, but to selectively accelerate only those reactions most desirable. In this thesis, Density Functional Theory (DFT) calculations, experimental reactor kinetics, and catalyst characterization are collectively employed to investigate a survey of open questions in biomass processing. Particular emphasis is placed on understanding the catalytic systems holistically and developing models that include other salient features of the system beyond the simple catalyst that fits on an organic chemistry reaction arrow. Glycerol is an immensely promising biomass feedstock available in surplus as a waste product of biodiesel formation. With these depressed feedstock costs, it can be made economically viable to convert glycerol to acrolein. This reaction has been studied exhaustively in the gas phase catalyzed with a gaseous proton. More realistic reactor systems however often employ liquid-phase reactions for this chemistry in the interest of process intensification. Also of interest are analogous reactions with sorbitol or sugars which are challenging to vaporize at all. A DFT model was therefore constructed to explore glycerol dehydration in the aqueous phase and the impact of explicit solvent molecules on the chemistry. Participation of solvent molecules in the chemistry permitted some reactions which were impossible in the gas phase while inhibiting some that are known to be less significant. The end result better predicted the dominant reaction pathway observed in experimental systems. This study was followed by experimental work studying the dehydration of pure, liquid propylene glycol, in contrast to existing studies employed dilute, gas-phase feed streams. Using experiment and DFT models, Brønsted-acidic zeolites were compared to the catalytically-simpler mesoporous sulfonic-acid resin Amberlyst 36Dry in order to better understand the impact of the zeolite micropores and their oxide character. The zeolite was found to amplify the already high selectivity towards dehydration of the secondary carbocation, and also to inhibit a sequential reaction leading to undesired side products. Lastly, work with Lewis-acidic Sn-BEA zeolite over the last decade has shown remarkable but inconsistent catalytic activity for a range of promising chemistries, with various catalyst pretreatments shown to improve or nearly completely deactivate the catalyst. Chief among these is the generation of open sites through calcination in a humid environment. Sn-BEA was studied with an array of characterization techniques in conjunction with DFT calculations in order to better understand the nature of the active site, improve characterization protocols, and expound upon the interaction between Sn-BEA and water. It was determined that the 960 cm-1 peak commonly associated with Sn embedded in the zeolite was in fact the consequence of a silanol defect which forms readily on Sn-BEA in water. The silanol groups account for 4-5 water molecules per Sn atom in the zeolite and condense and desorb at a temperature between 400 and 700 °C.