(624b) Rational Design of Mixed Solvent Environments for Acid-Catalyzed Biomass Conversion Reactions: A Combined Approach Using Experiments and Molecular Simulations

Walker, T. - Presenter, University of Wisconsin - Madison
Chew, A., University of Wisconsin
Demir, B., University of Wisconsin-Madison
Li, H., Dalian Institute of Chemical Physics
Zhang, Z. C., Dalian Institute of Chemical Physics
Huber, G., University of Wisconsin-Madison
Van Lehn, R., University of Wisconsin-Madison
Dumesic, J., University of Wisconsin-Madison
Recently, mixtures of water with organic cosolvents (i.e., mixed solvent environments) have garnered interest for their ability to improve the yields of important acid-catalyzed biomass conversion reactions. However, the mechanistic details underlying these improved yields are not understood in way that generalizes broadly across different reaction classes and solvent systems. As such, insights to guide the rational design of mixed solvent environments are limited, and optimizing the composition of the liquid phase for new processes often represents a time-consuming, trail-and-error exercise.

Here, we will present results for the Brønsted-acid-catalyzed reactions of ethyl tert-butyl ether, tert-butanol, levoglucosan, 1,2-propanediol, fructose, cellobiose, and xylitol in mixtures of water with [gamma]-valerolactone, tetrahydrofuran and 1,4-dioxane in varying compositions. We demonstrate how reactants containing more hydroxyl groups exhibit greater catalytic turnover rates for both dehydration and hydrolysis reactions as the water content of the mixed solvent environment decreases. We present classical molecular dynamics (MD) simulations to probe the nature of these solvent effects by quantifying the extent of water enrichment in the local domain about the reactants, and the hydrogen bonding strength between water molecules and the reactants, both as a function of solvent composition. By correlating the experimental reaction kinetics data with the behavior of these simulation-derived observables, we develop a model that accurately predicts the rates of all seven acid-catalyzed reactions as a function of the composition of the liquid phase. The development of this modeling tool represents an important step toward a general understanding of solvent effects in acid-catalyzed biomass conversion processes, and toward the rational design of mixed solvent environments for new liquid phase processes.