Greener Hydroformylation with Nanofilterable Rhodium Catalysts in A Stirred Membrane Reactor

Abstract

Hydroformylation for producing industrial chemical intermediates such as aldehydes and alcohols from olefinic substrates is a 15 billion lb/year industry and a potential route for increasing the carbon chain of light olefins such as ethylene and propylene to produce longer chain olefins. Novel, resource-efficient technologies that conserve feedstock and energy while producing the desired product continue to be interest to industry. This dissertation advances a novel continuous reactor concept involving gas-expanded liquids and nanofilterable catalysts with potential for practical viability. It has been shown by previous researchers in our group that the use of simple homogeneous Rh/TPP catalyst for 1-octene hydroformylation in CO2-expanded liquid (CXL) media provides exceptional TOF (~316 h-1) and regioselectivity (n/i ~9) at mild pressure (~40 bar) and temperature (30-60 °C) when compared to conventional media. Systematic phase equilibrium and reaction studies performed in this dissertation show that CXLs provide a way of increasing (i.e., tuning) H2/CO ratio in the liquid phase at fixed syngas feed composition, low syngas partial pressures (i.e. avoiding syngas inhibition) and mild conditions (40-60 °C, tens of bars). A novel process concept for continuous hydroformylation in CXLs was demonstrated using bulky phosphite ligands that are effectively retained in the stirred reactor by a nanofiltration membrane. The reactor was operated at 50 °C with a syngas pressure of 6 bar to avoid CO inhibition of reaction rate and selectivity. The nanofiltration pressure was provided by ~32 bar CO2 that also created CXL phase resulting in enhanced turnover frequency (~340 h-1), aldehydes selectivity (>90%) and high regioselectivity (90%) and high regioselectivity (n/i ~8) at nearly steady operation for up to 50 h (cumulative TON of 17,680). Further, the use of pressurized CO2 also reduced the viscosity of the conventional liquid phase by 30 to 50% thereby improving the mass transfer properties. Constant permeate flux was maintained during the 50 h run with Rh leakage being less than 0.5 ppm. Novel Rh complex catalysts bound to soluble versions of inexpensive functionalized polysiloxanes ligands were also investigated for the hydroformylation of higher olefins (C5+) to enable nanofiltration in a continuous membrane reactor. Polysiloxanes containing two different functional groups (amine- and phosphine-) were used for Rh complexation. Both types of Rh complexes, when tested in batch and continuous hydroformylation experiments, show good activity and chemoselectivity for a variety of olefinic substrates. The phosphine-functionalized ligand exhibited higher activity (TOF = 165 h-1) than the amine-functionalized one (TOF = 17 h-1) at 6 bar and 50 °C. Continuous operation with Rh complexes (with the phosphine-functionalized ligand) lasting several days in a stirred reactor equipped with nanofiltration membrane showed steady hydroformylation activity (TOF = 103 h-1; cumulative TON = 12,240 after 120 hours) and high chemoselectivity ( 91%) towards the aldehydes. The Rh concentrations in the reactor effluent were less than a few ppm, which can be easily recovered by absorption and exceeds the economic viability criterion established by earlier work in our group. This technology concept has potential applications in homogeneous catalytic processes to improve resource utilization and catalyst containment for practical viability.