419780 Effective Aqueous Phase Hydrogenation: GVL Production Kinetics and Stability

Monday, November 9, 2015: 1:10 PM
355C (Salt Palace Convention Center)
Omar A. Abdelrahman1, Helen Luo2, Andreas Heyden3, Yuriy Román-Leshkov2 and Jesse Q. Bond1, (1)Biomedical and Chemical Engineering, Syracuse University, Syracuse, NY, (2)Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA, (3)Department of Chemical Engineering, University of South Carolina, Columbia, SC

γ-valerolactone (GVL) is a lignocellulosic platform chemical that offers tremendous flexibility in down-stream applications and upgrading.  The primary route envisioned for the production of lignocellulosic GVL is the aqueous-phase hydrogenation of levulinic acid (LA) 1. This reaction is generally carried out over supported metals, most commonly Ru given its high activity for ketone hydrogenation in water. For the process to be feasible, catalysts must deliver high intrinsic activity towards LA hydrogenation and maintain that activity for extended times on stream. Through an investigation of the kinetics and catalyst stability during aqueous-phase LA hydrogenation over supported Ru nanaoparticles, we identify parameters allowing the design of efficient catalysts for this reaction.

We first decouple 4-hydroxypentanoic acid (HPA) and angelicalactone (AL)-mediated hydrogenation pathways to illustrate that, at low temperatures and in water, GVL formation occurs primarily through ketone hydrogenation followed by intramolecular esterification (Figure 1,a). This transformation appears independent of the nature of the Ru support, and occurs with an average turnover frequency of 0.11 s-1 at 323K, 0.5M LA, and 24 bar hydrogen on Ru/C, Ru/SiO2,Ru/TiO2, and Ru/ γ -Al2O3.  In the formation of GVL, the rate of ring closure of HPA was found to be the slowest step, and its apparent barrier (70 kJ mol-1) is larger than that of the Ru catalyzed hydrogenation (48 kJ mol-1). Rates of low temperature GVL production can be improved by coupling Ru/C with Amberlyst-15, which accelerates the rate of hydroxy-acid ring closure2.

Despite Ru/C possessing a high intrinsic activity for the hydrogenation of LA in the aqueous phase, it deactivates rapidly once on stream, losing ~50% of initial activity within 5 hrs. Partial regeneration of the catalyst was possible by reducing the catalyst at 673K in a H2 atmosphere; however, a substantial portion of the deactivation was irreversible and attributed to particle sintering. By comparing sintering kinetics across silica, titania, γ -alumina and carbon, we find that the extent of sintering is inversely related to the mean electronegativity of the support3. The source of reversible deactivation in this system could not be conclusively identified; however, the extent of reversible deactivation appears to correlate with support point of zero charge (PZC) and/or the prevailing surface charge of the catalytic material in water. The above insights establish a framework for designing stable supported metal catalysts for use in aqueous phase hydrogenation and hydrodeoxygenation, thus addressing a central need in biomass processing.  


1-     Wright, W. R. H., Palkovits, R., Development of heterogeneous catalysts for the conversion of levulinic acid to y-valerolactone, ChemSusChem, 5, 1657-1667 (2012).

2-     O. A. Abdelrahman, A. Heyden and J. Q. Bond, Analysis of Kinetics and Reaction Pathways in the Aqueous-Phase Hydrogenation of Levulinic Acid To Form γ-Valerolactone over Ru/C, ACS Catal., 2014, 4, 1171–1181.

3-     O.A. Abdelrahman et al., Toward rational design of stable, supported metal catalyst for aqueous phase processing: Insights from the hydrogenation of levulinic acid, J.Catal. (2015), http://dx.doi.org/10.1016/j.jcat.2015.04.026


Extended Abstract: File Not Uploaded