Formation of Propargylic Stereocentres using Desymmetrised Alkyne Complexes

Abstract

Addition of Grignard reagents to alkynyl aldehyde complexed bimetallic system proceeds with good to excellent stereocontrol.The creation of chiral centres at propargylic (prop-1-ynyl) positions has received some interest in the past few years. Methods for the preparation of stereocentres next to alkynes include the addition of nucleophiles to propargylic aldehydes using titanium catalysis 1 and boron reagents 2 and a metal catalysed ene reaction. 3 The other main methods have been the hydride reduction of propargylic ketones with chiral borohydrides, 4 in particular the CBS oxazaborolidine catalyst system 5 and by transfer hydrogenation. 6 The importance of this chemistry is perhaps best illustrated when the newly formed stereocentre is used in the preparation of complex natural products, such as macrolactin, 2 isocarbacylin derivatives, 3 methyl nonactate 4a and scopadulcic acid. 4cOur interest in metal alkyne chemistry prompted us to find a new method for the generation of propargylic stereocentres. Of relevance to this are the extensive studies by Nicholas 7 and Hanaoka 8 on the addition of enols and enolates to complexed propargylic aldehydes. The metal group facilitates high diastereocontrol in the reaction, and can be readily removed to leave the organic fragment with the newly created chiral centres. In addition, Nicholas has shown that phosphine substituted cobalt alkyne complexes can be used to form enantio-enriched propargylic stereocentres since the complex itself is rendered chiral by the introduction of the phosphine ligand. 9 Access to products of this type has been achieved, again by Nicholas, by adding the PPh 3 ligand to the propargylic alcohol complexes. 10 However, the diastereoselectivity was highly dependant on the size of the substituents. Remarkably, there have been, to the best of our knowledge, no reports of addition of Grignard or organolithium reagents to the same types of complexes.Starting with 2-butyn-1-al diethyl acetal (Scheme 1), complexation to cobalt was achieved simply by stirring with Co 2 (CO) 8 in DCM. Treatment of the acetal with a catalytic amount of p-toluene sulfonic acid gave the aldehyde. This was then reacted with triphenyl phosphine to produce the desired complex 1a. Complex 1b was prepared in an analogous manner from propynal. Scheme 1With the desired aldehyde in hand, we reacted this with a series of nucleophiles. All the reactions outlined in Scheme 2 proceeded quickly at low temperature (-78°C) in THF to give the corresponding alcohols. The diastereoisomeric excess was measured directly in most cases since the complexes were easily separable by chromatography. With inseparable complexes, the diasteromeric excess was estimated from the 1 H NMR spectra. The results show that excellent diastereoisomeric excesses are achieved with all the nucleophiles we studied. Scheme 2In order to account for the stereocontrol, we obtained Xray crystal structures of the starting metal alkyne complex and one of the products. 11 The aldehyde, 1a, is shown in Figure 1. The complex a...

Preparation of NewN-Heterocyclic Carbene Metal-Alkyne Complexes and Application to a Stereocontrolled Pauson-Khand Reaction

“…With the structural evidence in hand, initial synthetic research focused on its potential use of the new complex 4e in asymmetric PKRs. This has previously been achieved thermally with heterobimetallic complexes, 7,8 and via the use of N-oxides such as N-methylmorpholine-N-oxide (NMO) with desymmeterised bis-cobalt complexes. 16,17 The L-(+)-menthol-alkyne complex (Scheme 2) formed separable diastereoisomers when reacted with the carbene ligand with, surprisingly, an approximately 3:1 de in favour of the early eluting isomer.…”

“…This diastereoselectivity is intriguing since even the displacement of the cobalt tricarbonyl group by CpMo(CO) 2 only achieved a 1:1 ratio. 7,8 The diastereoisomers were separated by column chromatography then used in Pauson-Khand reactions under both N-oxide and thermal conditions. As illustrated in Table 2, the N-oxide conditions did not give useful yields of the cyclopentenone product when the reaction was run at room temperature.…”

“…When we used the separated diastereoisomers of the complex, these gave rise to distinct cyclopentenones, as we have noted previously for the heterobimetallic-alkyne complexes. 7,8 Of particular note is that the unsubstituted biscobalt-hexacarbonyl complex gives a de of less than 10%, arising from the menthol auxiliary. 7 The stereocontrol achieved here, therefore, must arise from the configuration of the IPr-cobalt-alkyne core of the molecule.…”

“…6 Alternatively, chirality may be achieved via the replacement of one of the cobalt vertices by a different metal core, forming a heterobimetallic complex, such as 3 (Figure 1). These heterobimetallic complexes have been successfully employed in organic synthesis including asymmetric versions of the Pauson-Khand reaction, 7,8 stereocontrolled Nicholas reactions 9 and recently as an alkyne based chiral auxiliary. 10 We also have an interest in imidazolium derived carbene ligands, and were intrigued to see whether we could use these to desymmeterise the metal-alkyne complexes.…”

“…These heterobimetallic complexes have been successfully employed in organic synthesis including asymmetric versions of the Pauson-Khand reaction, 7,8 stereocontrolled Nicholas reactions 9 and recently as an alkyne based chiral auxiliary. 10 We also have an interest in imidazolium derived carbene ligands, and were intrigued to see whether we could use these to desymmeterise the metal-alkyne complexes. Of particular interest were the reported enhanced thermal and kinetic stabilities, and reactivities of many catalytic orga-nometallic complexes when N-heterocyclic imidazolium ligands were used as phosphine mimics.…”