The role of the counterion in the catalytic activity of vanadium(V)(salen) complexes

Abstract

Over the last ten years, the North group has developed VO(salen)X complexes as an efficient catalytic system for the asymmetric addition of trimethylsilyl cyanide to aldehydes. It was found that the nature of the counterion X has a significant influence on the catalytic activity, but not on the enantioselectivity of the reaction. Complexes with the most coordinating counterions displayed the highest levels of catalytic activity. Kinetic studies revealed that the monometallic VO(salen)X complexes exist in equilibrium with bimetallic complexes, and both are catalytically active. This was supported by mass spectrometry which detected both [VO(salen)]+ and [VO(salen)]2+ ions, with the latter involving both V(V) and V(IV) ions. In this project, electron paramagnetic resonance spectroscopy (EPR) was used to monitor vanadium(IV) formation, revealing that the rate of formation is directly related to the catalytic activity of the complex. Also using EPR, cyanide was found to be the reducing agent and to be oxidized to cyanogen via a non-radical mechanism. Oxovanadium complexes bearing highly coordinating counterions were most rapidly reduced to vanadium(IV), thus favouring the formation of highly reactive bimetallic species. In contrast, less coordinating counterions resulted in the formation of much lower amounts of dinuclear species. The potential of the counterion to display Lewis-base catalysis became increasingly clear during this project. A Hammett plot based on a series of para- and meta-substituted benzaldehydes, was used to determine the relative importance of Lewis-acid and Lewis-base catalysis within VO(salen)X complexes. As expected, the vanadium catalysts studied gave a positive reaction constant indicating that there is an increase in electron density at the benzylic carbon during the transition state. However, a less positive reaction constant (ρ = 1.2) was found for VO(salen)NCS which possessed a strongly coordinating counterion, compared to that of VO(salen) EtOSO3 (ρ = 1.9) which possessed an ionic counterion, which indicates a possible Lewis base influence from the thiocyanate counterion. These complexes were also compared to metal(salen) complexes of titanium and aluminium. The latter required the presence of triphenylphosphine oxide as an achiral Lewis-base cocatalyst, and exhibited predominantly Lewis base catalysis with a reaction constant of 0.7, whereas the titanium catalyst was found to function almost entirely as a Lewis-acid catalyst with a reaction constant ρ = 2.4. Thiocyanate was also found to be an excellent Lewis base ii catalyst for racemic cyanohydrin synthesis, for which a mechanism involving activation of the trimethylsilyl cyanide through a hypervalent silicon bond was suggested. The use of propylene carbonate as an alternative solvent to dichloromethane was shown to affect the rate of cyanohydrin synthesis when VO(salen)NCS was used as the catalyst. Thus, a mechanistic study was undertaken. The reaction was found to obey second order kinetics in both propylene carbonate and dichloromethane. However, when the order with respect to the catalyst was determined, it became evident that propylene carbonate altered the monomer-dimer equilibrium towards the monomer. The monomer was the most abundant species in solution and hence was responsible for most of the catalytic activity. 51V-NMR experiments provided evidence for propylene carbonate coordination to VO(salen)NCS, blocking the sixth coordination site, and hence inhibiting both dimer formation and aldehyde coordination. Further evidence for this effect was provided by a Hammett analysis, which showed that Lewis base catalysis was more pronounced when propylene carbonate was the solvent.