A First Asymmetric Diamination of Olefins

Abstract

An asymmetric diamination of olefins employing stable imidoosmium(VIII) complexes is described for the first time. The reaction is based on the use of chiral acrylic esters and related substrates with enantiopure menthol and 8-phenyl menthol as chiral auxiliaries and gives rise to osmaimidazolidines with good to excellent diastereomeric ratios (up to 95:5). Removal of the osmium centre and the chiral auxiliary leads to novel enantiopure 1,2,3-trisubstituted compounds.Metal-mediated diamination of unfunctionalised olefins dates back to 1974. In seminal work, Barluenga and Aznar described the use of thallium(III) salts to promote addition of anilines to C=C double bonds. 1 Reports on palladium by Bäckvall, 2 osmium by Sharpless 3 and mercury by Barluenga 4 followed quickly. However, further development of this particular reaction has been extremely scarce 5,6 and a direct asymmetric diamination of olefins has remained unknown to date. Since chiral vicinal diamines are of upmost importance in synthetic, biological and medicinal chemistry, 7 the development of a stereoselective diamine synthesis from olefins represents an important research goal. We describe here first results on asymmetric diamination of olefins (ADA) based on dioxobis(t-butylimido)osmium (1) and oxotris(t-butylimido)osmium (2). 3,8 Oxidative enantioselective C-C bond transformation is before all the success story of Os(VIII) chemistry, and Sharpless' work on dihydroxylation (AD) and aminohydroxylation (AA) has set highest standards. 9 In view of the original report on Os-mediated diamination that included the reaction of 1 and 2 with fumarates, 3 we have increased the range of substrates to unsymmetrical olefins such as acrylic and cinnamic esters and amides, which also yield stable osmaimidazolidine products. 10 Attempts to develop an asymmetric version of this diamination by use of chiral non-racemic ligands in either stoichiometric or substoichiometric amounts met with little success. For example, diamination of dimethyl fumarate or methyl cinnamate in the presence of 1 and equimolar amounts of the standard Sharpless ligands (DHQD) 2 PHAL and (DHQ) 2 PHAL led essentially to racemic products. 10 Apparently, the diminished electrophilicity of 1 and 2 as compared to OsO 4 and O 3 OsNt-Bu prevents complexation. Preliminary NMR control experiments (in CDCl 3 and C 6 D 6 at room temperature) revealed that the equilibrium for a mixture of 1 and (DHQD) 2 PHAL is located exclusively on the side of the uncomplexed compounds. This indicates that asymmetric synthesis via chiral ligand induction is not suitable for the present aim of olefin diamination. To this end, chiral auxiliary incorporation proved superior, and among several chiral alcohols that were envisaged, (-)-8-phenylmenthol performed best. Thus, reaction of complex 1 with the chiral non-racemic ester 3a resulted in the formation of two diastereoisomers 4a and 5a (Scheme 1 and Table 1). 11

Methods for direct alkene diamination, new & old

“…219,220 In common with Sharpless, Muñiz has found that electron-deficient alkenes are the most favored substrates for diamination, but has also demonstrated that in addition to acrylate and cinnamate esters 315 , α,β-unsaturated ketones, aldehydes, amides 318 , nitriles and even 3-pyridyl acrylates are suitable substrates (Scheme 81). 221 Liberation of the 1,2-diamines from their corresponding metallaimidazolidines can be accomplished through reduction with LiAlH 4 or NaBH 4 . 219 …”

Rhodium(II)‐Catalyzed Oxidative Amination

“…Both osmium-and selenium-based reagents have been demonstrated to promote this transformation, but application in synthesis is limited by the stoichiometric nature of these processes and by an abridged substrate scope [106][107][108]. Li has described a rhodium-catalyzed diamination reaction of alkenes using TsNCl 2 as the terminal oxidant (Scheme 17.36) [109].…”

Stereochemically Defined Osmium Centers from Asymmetric Diamination of Olefins: Mechanistic Implications for Osmium‐Mediated Acrylate Functionalization

“…We have previously employed (À)-8-phenylmenthol as auxiliary in asymmetric diaminations with the related bisimido complex. [10,11] In the present case, diamination of (À)-8-phenylmenthyl cinnamate and crotonate yielded four stereoisomers in different ratios of about 7:3:1:1 and 6:3:1:1, respectively. In contrast, (À)-8-phenylmenthyl acrylate 5 a and the corresponding 8-phenylmenthyl methacrylate reacted faster and produced an almost equal mixture of only two diastereomers (ratio = 1:1:0:0) that could be separated by semipreparative HPLC (Scheme 2).…”

“…Among these reactions, Sharpless dihydroxylation and aminohydroxylation have been highly acclaimed, [8] whereas related diamination reactions have received only scant attention. [9][10][11] We have recently reported on the stereoselective diamination of olefins by reaction of bis(N-tert-butylimido)dioxoosmium(viii) with fumaric and acrylic ester derivatives [10] to give stable monomeric osmaimidazolidines.[11] When the corresponding trisimidoosmium complex 1 is allowed to react with unsymmetrical esters such as methyl cinnamate 2 a, clean formation of two new compounds is observed (Scheme 1). These were identified as a pair of diastereomeric chiral-at-metal osmaimidazolidines.…”

“…Regarding the formation of the two diastereomers from 1 and the chiral acrylate 5 within a [3+2] process, the necessary stereoselection would be assumed to arise from an efficient p interaction between the phenyl group of the terpene and the Si face of the olefin, as observed in corresponding diamination reactions with the related bisimido reagent. [10,11] Such a selection step with an approach of the achiral 1 to the free Re face of the substrate would result in two diastereomeric products 6 and 8 with opposite absolute configuration at the central Os atom but with the same absolute configuration in the backbone of the osmaimidazolidine (Figure 2). However, the two products 6 and 7 differ both in absolute configuration at Os and in the osmaimidazolidine backbone, and these absolute configurations are inconsistent with a concerted, syn-specific [3+2] mechanism of olefin face differ- entiation.…”