Three-phase
Hydrogenation Reactions using a Ruthenium Coated Polymeric Membrane Reactor
John
Stanford1,2, Peter Pfromm2,
and Mary Rezac2*
1IGERT
Trainee in Biorefining
2Department of Chemical Engineering, Kansas State University
Manhattan, Kansas, United States
*rezac@ksu.edu
Catalytic membrane reactors afford an alternative and potentially more efficient method for performing three-phase heterogeneous chemical reactions. Traditional three-phase reactors often present mass transfer limitations, namely relatively large diffusional distances to reach catalytic sites exacerbated by low gas solubility in the liquid phase. Membrane reactors can alleviate the inherent mass transfer limitations by directly and abundantly supplying gas to the catalytic sites located on the membrane surface, which acts as a gas/liquid phase contactor, and thus lessening the necessity for higher gas phase pressures. The reactions investigated in this work include the hydrogenation of 5-hydroxymethylfurfural (HMF) in an alcohol solvent and the hydrogenation of levulinic acid (LA) in an aqueous solvent.
Polyimide polymers are used in this work for membrane synthesis because of their good chemical and high temperature resistance. The polymer solutions are cast as asymmetric integrally-skinned flat sheet membranes and are then coated with a ruthenium catalyst on the non-porous surface. The completed membrane is positioned in a flow over configuration maintaining liquid contact on the metal coated surface while allowing hydrogen gas to permeate from the porous support side to the catalytic sites on the non-porous surface. The multi-functionality of the membrane reactor/contactor has allowed several areas to be investigated, including catalytic activity and reaction kinetics, membrane performance and characterization, and solvent/polymer interactions. Quantitative hydrogenation product formation with reaction kinetics similar to or better than more traditional three-phase reactors has been achieved. The membrane reactors are shown to be stable and catalytically active for several days of continuous operation. This work has also demonstrated that unless the dense separating layer of the polymeric membrane exhibits and maintains a high degree of ‘defect free' quality, then the penetrating liquid reaction solvent reintroduces the mass transfer limitations for the gas phase that the membrane was intended to eliminate. Continued efforts to improve membrane reactor performance such as increasing hydrogen permeance, increasing catalytic site availability, and decreasing liquid phase permeance should yield an increasingly favorable comparison to traditional reactor systems.