Vittorio
Pace‡
,
James P.
Rae‡
,
Hassan Y.
Harb
and
David J.
Procter
*
School of Chemistry, University of Manchester, Oxford Road, Manchester, M13 9PL, UK. E-mail: david.j.procter@manchester.ac.uk
First published on 15th April 2013
The scope of the asymmetric silyl transfer to unsaturated lactones utilising a C2-symmetric NHC–Cu(I) catalyst has been established and kinetic resolutions mediated by silyl transfer have been used to prepare enantiomerically enriched anti-4,5-disubstituted 5-membered lactones. The method has been exploited in an expedient synthesis of (+)-blastmycinone.
Although the Oestreich and Hoveyda protocols allow the asymmetric silylation of a wide range of cyclic and acyclic substrates (e.g. esters, ketones, nitriles), limited studies on silyl transfer to heterocyclic systems, and in particular α,β-unsaturated lactone substrates, have been described.6,8 Furthermore, in general, asymmetric conjugate additions to 5-membered substrates are known to be challenging.10 In this Communication we describe our studies to optimise and establish the scope of the Cu-catalysed asymmetric silyl transfer to unsaturated lactones. We also report the kinetic resolution of 5-substituted butenolides11 mediated by the silyl transfer.
Building on Oestreich's studies involving furanone 1,6a,c we began our investigation by studying silyl transfer to 1 using Hoveyda's protocol.12 Employing L1, a ligand that has previously been used in conjugate silyl transfer to carbocyclic systems (Table 1),8a poor enantiocontrol was observed in the silylation of 1 (entry 1). Subtle modifications of the C1-symmetric imidazolinium salt core did little to improve the outcome (entries 2–5). For furanone 1, enantioselectivities were improved by a switch to C2-symmetric ligands (entries 6–10). In particular, the best results were obtained using C2-symmetric ligands bearing extended aromatic systems (naphthyl and anthryl, entries 8 and 10) on the N-phenyl substituent. However, the presence of additional steric bulk had a detrimental effect on enantioinduction (entry 9). Thus, the use of a C2-symmetric ligand L8 gave the best results in silyl transfer to 1.
Entry | Ligand | Type | R | R1 | R2 | er |
---|---|---|---|---|---|---|
1 | L1 | C 1 | Me | Et | H | 62:38 |
2 | L2 | C 1 | H | Et | H | 56:44 |
3 | L3 | C 1 | i-Pr | Et | H | 60:40 |
4 | L4 | C 1 | H | Me | Me | 75:25 |
5 | L5 | C 1 | Me | Me | Me | 66.5:33.5 |
6 | L6 | C 2 | H | Ph | — | 80.5:19.5 |
7 | L7 | C 2 | Me | Ph | — | 72:28 |
8 | L8 | C 2 | 2-Naphthyl | H | — | 93:7 |
9 | L9 | C 2 | 2-Naphthyl | i-Pr | — | 54:46 |
10 | L10 | C 2 | 2-Anthryl | H | — | 92:8 |
With ligand choice completed, we turned our attention to further optimising the reaction conditions (Table 2). Although different Cu(I) sources13 did not have a significant effect on enantioinduction, copper salt selection was important in maximizing conversions due to their moisture sensitivity.
Entry | Cu salt | Additive | Conv. (%) | er |
---|---|---|---|---|
1 | CuCl | — | 71 | 93:7 |
2 | CuBr | — | 80 | 90.5:9.5 |
3 | CuBr·SMe2 | — | 68 | 89.5:10.5 |
4 | CuOTf | — | 91 | 89.5:10.5 |
5 | CuI | — | >98 | 93:7 |
6 | CuI | 4 Å MS | >98 | 93:7 |
In this sense, the use of CuI gave the best results (entry 5) and the addition of molecular sieves improved reproducibility. Importantly, we found that ‘glove-box free’ conditions could be used thus greatly simplifying operational procedures. With optimised conditions in hand, we applied the protocol to the asymmetric conjugate silylation of 5, 6 and 7-membered α,β-unsaturated lactones (Table 3). The desired β-silyl adducts were obtained in up to 96.5:3.5 er in good yield (entries 1–3). Interestingly, the 8-membered lactone did not react, presumably due to conformational effects (entry 4). When a 7-membered lactone was fused with an aromatic ring, there was a deleterious effect upon enantiocontrol although the yield remained high (entry 5). Moderate selectivity was also observed in the silylation of a 2-substituted lactone (entry 6).
To further explore the scope of the protocol, the kinetic resolution of a series of 5-substituted butenolides was carried out using Cu(I)–NHC catalysed asymmetric silyl transfer.14 Pleasingly, treatment of 5-substituted butenolides with 60–70 mol% of PhMe2SiBpin and the C2-symmetric catalyst derived from L8 and CuI afforded silylated products after kinetic resolution in good yields (up to a maximum of 50%), good enantiomeric ratios and as single anti-diastereoisomers (Table 4).15
Entry | Substrate | Conversiona (%) | Yieldb (%) | er | s |
---|---|---|---|---|---|
a Determined by 1H NMR. b Yield of isolated product. c Selectivity factor determined according to ref. 16. d 10 mol% catalyst loading, 0.7 equiv. (PhMe2SiBpin). | |||||
1 | 50 | 46 | 86:14 | 13 | |
2 | 52 | 50 | 90:10 | 25 | |
3d | 43 | 43 | 91:9 | 19 | |
4d | 48 | 43 | 86:14 | 12 | |
5 | 47 | 46 | 84:16 | 10 | |
6d | 49 | 48 | 89:11 | 18 | |
7d | 43 | 42 | 86:14 | 11 | |
8 | 45 | 41 | 88.5:11.5 | 15 |
The rate of addition to 5-substituted butenolides was slower than silyl transfer to unsubstituted lactones, presumably due to increased steric hindrance, therefore higher catalyst and silylborane loading was required.17 Primary alkyl, allyl, benzyl and phenyl substituents at the 5-position of butenolides were found to be compatible with the process. To our knowledge, these examples represent the first kinetic resolutions achieved by Cu-catalysed silyl transfer from a Si–B reagent.
To demonstrate the value of Cu(I)–NHC catalysed asymmetric silyl transfer to unsaturated lactones, we report a concise approach to (+)-blastmycinone, a natural product arising from the hydrolysis of the antibiotic (+)-antimycin A3 (Scheme 1).18 Alkylation of silylated lactone 3 (see entry 1, Table 4) provided 4 with three contiguous stereocenters as a single diastereoisomer. Fleming–Tamao oxidation2 then gave lactone 5, which after esterification afforded (+)-blastmycinone.19
Scheme 1 Catalytic asymmetric approach to (+)-blastmycinone. |
In summary, we have explored the scope of a convenient procedure for asymmetric silyl transfer to unsaturated lactones. The Cu(I)–NHC catalysed process delivers β-silylated lactones in good yields and enantioselectivities. In contrast to observations with other substrate classes, the use of C2-symmetric imidazolinium salts as NHC precursors was crucial for efficient asymmetric silyl transfer to unsaturated 5-membered lactones. Kinetic resolution using Cu-catalysed silyl transfer from a Si–B reagent has been applied to racemic 5-butenolides and affords products with good enantiocontrol and excellent diastereocontrol. The method has been used in an expedient asymmetric synthesis of (+)-blastmycinone.
We thank The Leverhulme Trust (V.P.) and the EPSRC (J.P.R.) for funding and Robyn Bullough for assistance optimising the conversion of 3–5.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3cc42160k |
‡ Contributed equally. |
This journal is © The Royal Society of Chemistry 2013 |