(79d) Design of Active and Stable Materials for Gamma-Valerolactone Decarboxylation

Bond, J. Q., Syracuse University
Kellicutt, A. B., Syracuse University
Chatzidimitriou, A., Syracuse University
Salary, R., Syracuse University

γ-valerolactone (GVL) is a versatile bio-based chemical that can be prepared in good yields from cellulose (~70% of theoretical) through catalytic processes via intermediate formation of levulinic acid.  As a platform, GVL offers many processing options to various fuels and chemicals and stands to be a key intermediate in lignocellulose refining.  A potentially interesting application for GVL is as a butene precursor, which could be of interest in both fuel and chemical markets.  Specifically, catalytic decarboxylation of GVL over solid acids delivers butene monomers selectively, and nearly quantitative yields are possible.  GVL decarboxylation is proposed to be mediated by Brønsted acidity and occurs via a carboxylic acid intermediate that has a carbenium ion in the β position. 

Amorphous silica alumina (ASA) has been previously demonstrated to be effective for GVL decarboxylation at temperatures between 573K and 673K.  Under these conditions, ASA additionally undergoes pronounced deactivation via coke deposition, losing anywhere from 50 – 90% of its initial activity during the first 24 hours on stream.  Deactivation can be managed to some extent via a steam co-feed, which dilutes GVL and mitigates coke deposition; however, this increases downstream complexity and decreases the energy efficiency of the process.  A preferable alternative is to employ materials that, by design, facilitate decarboxylation and minimize coke deposition, but this outcome requires a description of the intrinsic activity and stability of an acid site during GVL decarboxylation.  Here, we consider the performance of several families of solid acids such that perturbations in catalyst morphology as well as strength and surface density of Brønsted sites can be correlated with the activity and stability of the material.  This effort is supported by comprehensive characterization of the materials to permit comparison of catalytic activity between various analogs on a per-site basis.

Influences of Brønsted and Lewis acidity were decoupled through consideration of amorphous SiO2-Al2O3 samples having varied silicon-to-aluminum ratios.  Amorphous samples were subsequently compared to high-silica zeolites (MFI and BEA) to reveal the effects of catalyst morphology and increasingly well-defined Brønsted acidity.  Finally, MFI samples were prepared at decreasing H+:Na+ ratios to determine whether Brønsted site density impacts the stability of the material. Temperature Programmed Desorption of ammonia and isopropylamine were employed to quantify the total and Brønsted acid site densities, and FTIR spectra of adsorbed probe molecules, such as pyridine and benzene, revealed insight into the strength of acidic centers.  We have observed that catalyst morphology, Brønsted:Lewis ratios, and acid site strength all influence the activity and stability of materials employed for GVL decarboxylation.  Results indicate that well-defined, strongly acidic Brønsted sites in MFI zeolites are among the most active for GVL decarboxylation, while the best stability is observed over amorphous catalysts with relatively low Si:Al ratios.