CHE 598 Seminar: Using statistical modeling for increased furan yields from biomass reactions

About the event

Presenter: Stephanie Wettstein, Associate Professor, Chemical and Biological Engineering, Montana State University

New initiatives for reducing adverse environmental impacts in chemical and fuel manufacturing open the door for the development of novel refining strategies using renewable feedstocks. Most research on lignocellulosic biomass focuses on upgrading the sugars and platform chemicals that result after pretreatment; however, a major opportunity exists in improving the pretreatment step and potentially combining pretreatment with upgrading. Furfural has been recognized as an important building block for fuels, solvents, and other value-added chemicals, and numerous recent publications have focused on optimizing furfural production from model sugars and lignocellulosic biomass.

Currently, the most common biomass processing methods are aqueous based, which lead to low yields and low concentrations due to the formation of unwanted by-products known as humins. The use of organic solvents for xylose dehydration reactions has recently gained interest due to faster reaction rates and the higher furfural yields obtained, but solvent selection has been mainly “guess-and-check” versus a systematic selection. This presentation focuses on the experimental work that found correlations between the reaction severity and solvent polarity to both xylose conversion and furfural yield. Using principal component analysis and projection to latent structures, a semi-empirical model was developed that provided estimates of xylose conversion and furfural yield over a range of experimental reaction severity and solvent polarity values.