Jennifer M. Schomaker, PhD , University of Wisconsin (Madison)
N2O is a greenhouse gas recently implicated in global warming. It is a potent and selective oxidant, albeit kinetically inert. Although the enzyme nitrous oxide reductase accomplishes the decomposition of N2O at ambient temperatures and pressures, the majority of homogeneous catalysts require high temperatures and pressures. The reasons for this are not clear, and this lack of information has impeded the design of more efficient homogenous catalysts for N2O activation.
Research design. The Groves group has previously demonstrated the stoichiometric epoxidation of trans-β-methylstyrene using a Ru(VI)(TMP)(O)2 complex (TMP=5,10,15,20-tetrakis(2,4,6-trimethylphenyl)porphyrin) formed from Ru(II)(TMP)(THF)2 and N2O. To our knowledge, the only catalytic examples of homogeneous catalysts for oxidation using N2O are also based on Ru(VI)(TMP)(O)2 using 10 atm N2O and 140 °C.
For our kinetic studies, we utilized a new, bulkier Ru(VI)(por)O2 complex supported by a D4-porphyrinato[5,10,15,20-tetrakis[1,2,3,4,5,6,7,8-octahydro-1,4:5,8–dimethano- anthracen–9-yl]porphyrin] to minimize competing dimerization of monomeric Ru(IV)(por)(O) complexes into inactive Ru(IV) µ-oxo dimers. The epoxidation of cholesteryl acetate was chosen as the model reaction.
Initial studies. The impact of several reaction parameters on the efficiency of the epoxidation was explored. As expected, coordinating solvents inhibited the reaction by binding to the metal center. Non-polar, non-coordinating solvents, such as benzene and chlorobenzene, performed best. The yield of epoxide increased when the pressure of N2O was increased from 1 to 12 atm, but remained essentially unchanged at higher pressures. The epoxidation did not proceed effectively at 25-80 oC, even at 17 atm N2O. Temperatures greater than 80 °C were necessary to promote the reaction. The initial catalyst concentration was varied, and as expected, the reaction rate and yield increased as the [Ru(VI)(por)(O)2] was increased. The effect of the olefin concentration on reaction rate and overall yield was examined. The initial concentration of alkene was studied in the range of 5.5×10-3 M to 3.5×10-2 M, with the catalyst concentration fixed at 5.2×10-4 M (3 mol%) and the pressure of N2O at 17 atm. The reaction were run in chlorobenzene at 100 ºC for 22 h. Interestingly, at [alkene] > 1.5×10-2 M, the rate and the yield of epoxide decreased significantly until no product was observed at an [alkene] = 3.5 x 10-2 M. However, [alkene] = 1.2×10-3 M led to 95% yield of epoxide. The effect of [alkene] was further demonstrated by the ability of the catalyst to promote reaction at only 80 oC under dilute conditions.
Kinetic studies. To shed light on the mechanism of epoxidation using N2O, we conducted further kinetic studies on our model system. The initial rates of reaction were measured by 1H NMR spectroscopy. Plots of [alkene] and [epoxide] vs. time were linear up to at least 30% completion of reaction and did not show an initial induction period.
Saturation behavior with respect to pressure of N2O was observed on the initial rates of epoxidation. The initial rates increased as N2O pressure was increased from 1 to 10 atm, and then remained almost unchanged with a further increase in the pressure of N2O (> 10 atm). The reaction appeared to be 1.6 order with respect to the catalyst [Ru(VI)(por)(O)2], although further studies are needed. The initial reaction rates vs. [alkene] showed a curved plot for the epoxidation of cholesteryl acetate. Higher alkene concentrations retarded the initial rates of reaction and eventually inhibited the reaction. Obviously, substrate inhibition is problematic, although studies on product inhibition have yet to be completed.
The effective rate constants for the epoxidation were determined to yield an effective Ea for the epoxidation of cholesteryl acetate with our Ru(VI)(por)(O)2/N2O system of 12.5 kcal/mol. This compares to Ea = 8.1 kcal/mol for the stoichiometric epoxidation of cholesteryl acetate with Ru(VI)(por)(O)2. Not surprisingly, this indicates the rate-determining step is the regeneration of the Ru(VI)(por)(O)2 catalyst.
Proposed mechanism. Based on our kinetic studies, we propose a plausible mechanism whose features may help us develop more effective catalysts for N2O activation. The active species, Ru(VI)(por)(O)2, affords the epoxide and Ru(IV)(por)(O). The Ru(IV)(por)(O) can bind N2O and undergo oxidation to Ru(VI)(por)(O)2, or undergo rapid disproportionation to yield Ru(VI)(por)(O)2 and Ru(II)(por). The poor ligating ability of N2O, as well as literature precedent, led us to propose binding to the Ru(II)(por) species in preference to Ru(IV)(por)(O). Generation of the desired Ru(por)(O) complex from Ru(II)(por)N-N-O is proposed to require another molecule of Ru(II)(por) or Ru(IV)(por)(O). The terminal oxygen of the bound N2O interacts with this second metal center to form a Ru-N-N-O-Ru bridged complex. Any other molecule that competes for the axial coordination site will inhibit the reaction. This explains the need for low [alkene] in order for the reaction to proceed. Weakening of the N-O bond in the bridged complex leads to heterolytic cleavage and the formation of Ru(IV)(por)(O) and Ru(II)(por)N2 complexes. Disproportionation of Ru(IV)(por)(O) as previously described closes the catalytic cycle.
In conclusion, our kinetic studies on the epoxidation of cholesteryl acetate using a Ru(VI)(por)(O)2/N2O catalyst system have led us to a working mechanism to guide future catalyst design. For example, the steric environment around the metal center will be manipulated in order to bind N2O selectively over the substrate and/or product. Our kinetic data suggest that the rate-determining step in the reaction is the breaking of the N-O bond in the intermediate Ru-N-N-O-Ru bridging complex. Thus, bimetallic catalysts that can bind N2O in such a manner will be synthesized and kinetic studies undertaken to compare reaction rates between the bi- and monometallic complexes. Attempts to observe the bridged complex at low temperatures will also be undertaken to provide definitive proof for this proposed intermediate.