Polymer-modified catalyst supports as tailored local solvation environments to control reaction selectivity in liquid-phase chemical processes

Catalytic processing of lignocellulosic biomass is a promising strategy toward partially displacing petroleum-derived chemicals and energy with lower-carbon alternatives. To this end, biomass and other renewable oxygenates are often processed in water or organic solvents, where products are made at the catalyst support's liquid-solid interface. Under these conditions, the selectivity of acid-catalyzed biomass conversion processes is particularly challenging to control due to the reactive nature of biogenic oxygenates, which can form multiple products in series and parallel. According to our central hypothesis, surface-grafted polymer moieties should modulate the solvation free energy of reactant and product states near active sites, and potentially enhance the rates of desired reaction steps over undesired ones, similar to how liquid solvent mixtures control reactivity over mineral acid catalyst. Here, we utilized readily tabulated polymer solubility descriptors (Hansen Solubility Parameters) to pre-select polyimides (PIM) as HMF-selective silica modifiers from a subset of candidate polymer systems. We utilize solvent-based methods and coprecipitation techniques to generate monodispersed silica nanoparticles (110-13,000 nm) with well-defined acid densities of 0.25-2.35 mmol g-1. We synthesized polyimide moieties from 4,4’-oxidianiline (ODA) and 4,4’-(hexafluoroisopropylidene)diphthalic anhydride (6FDA) precursors, and used THF/water mixtures to tune the solubility of the as-synthesized polyimide film and recover chains of well-defined molecular weights. We graft these polyimide structures on acid-functionalized silica nanoparticles (NP’s); and utilize electron microscopy, laser light diffraction and thermogravimetric analysis to confirm their successful grafting on the silica surface (~4-16 wt%). We report reaction kinetics measurements, as well as modeling results accounting for reaction/diffusion dynamics to demonstrate how polyimide-decorated silicas preferentially stabilize the product state in acid-catalyzed dehydration of fructose into 5-hydroxymethyl furfural (HMF). As a result, rates and selectivity towards HMF are enhanced relative to the same catalyst in pure water. Although the polymers do create a strong diffusion boundary which limits their performance, we introduced a molecular dynamics-enabled strategy to screen materials and design polymer modified surfaces that should balance diffusion resistance with enhanced reactivity. In a forthcoming publication, we will present a general strategy to extend this concept to the design of arbitrary surfaces for a given reaction sequence.