Pyrrolidinone-fused Cyclohexenones by Regioselective Dearomatising Anionic Cyclisation of 2-, 3- or 4-Methoxybenzamides

Abstract

On treatment with t-BuLi in the presence of HMPA, 2-, 3-or 4-methoxybenzamides (o-, m-or p-anisamides) bearing Nbenzyl substituents cyclise with dearomatisation to give methyl dienyl ethers which hydrolyse to give single stereo-and regioisomers of pyrrolidine-fused cyclohexenones. The bicyclic enones are versatile synthetic intermediates which undergo transformations such as stereoselective reduction and conjugate addition, regioselective Baeyer-Villiger oxidation, bromination and hydrogenation.The lithiation a to nitrogen of aromatic amides bearing benzylic substituents has been known for many years, with Fraser 1 and Durst 2 reporting the a-lithiation by LDA of N-benzylamides 1a, 2 and 3. In 1999, we reported 3 that the organolithiums derived from simple N-benzyl benzamides 1a and 1b are unstable at -78 °C in the presence of HMPA: they undergo a cyclisation reaction with loss of aromaticity in the benzamide ring, leading to pyrrolidinefused cyclohexadienes 5 and 6 (Scheme 1). The intermediate in the cyclisation reactions is an extended enolate 4, which unfortunately does not react regioselectively with electrophiles -a serious limitation to the synthetic application of the dearomatising reaction.In this Letter, we now report that benzamides 7-9 and 11 bearing methoxy groups also cyclise with dearomatisation, but that the regioselectivity of the electrophilic quench step is immaterial to the formation of single final product because hydrolysis of a mixture of first-formed dienyl ethers usually generates a single regioisomer of a cyclohexenone. The cyclohexenone products are versatile synthetic intermediates which undergo stereoselective functionalisation. 4The four anisamides (methoxybenzamides) 7, 8, 9a and 9b were made by acylation of N-benzyl-t-butyl amine or N-p-methoxybenzyl-t-butyl amine. Amide 11 was made by ortholithiation 5 and benzylation of the secondary benzamide 10 (Scheme 2). They were each treated under our standard conditions 3 for dearomatising cyclisation (Method A: t-BuLi, THF, HMPA, -78 °C, or, for 8 and 11, -40°C) and the reactions were quenched with aqueous ammonium chloride (Scheme 3).The o-anisamide 7 was the only amide to cyclise with a lack of stereoselectivity, giving 76% of the cyclised product 12 as a mixture of stereoisomers, 6 which were separable by flash chromatography on silica, though slightly unstable. Each of exo-and endo-12 was obtained as a single regioisomer only. The m-anisamide 8, which in principle could cyclise onto either the 2-or the 6-position of the ring, similarly gave only a single regioisomer 14, but the cyclisation required temperatures of -40 °C to return good yields of 14. The p-anisamides 9a, 9b and 11 cyclised stereoselectively, but sometimes gave mixtures of regioisomers arising from a-or g-attack on the enolate intermediate. 16a and 16b were typically obtained as single regioisomers; 7 19a and 19g were obtained in a 5:1 ratio; and 18a and 18g (produced by alkylating the intermediate enolate with MeI) were obtained in a 1:2 ratio. N O Ph R N O Ph R 1a, R ...

The Synthesis of (−)-Isodomoic Acid C

“…To employ this cyclization in the synthesis of isodomoic acid C, we made amide 5 from cumylamine 4 on a 10−20 g scale and cyclized it in 2.5 g batches. Treatment of 5 in THF at −78 °C with N -lithioamine 6 by our published method 13a promoted asymmetric deprotonation and cyclization to an enol ether which was hydrolyzed in situ to yield enone 7 (Scheme ) in 86% ee (by HPLC). Recrystallization of 7 from ethyl acetate improved the enantiomeric excess to >99%.…”

Approaches to the Formation of Condensed Isoindolones

“…These isoindolones can be converted in further steps, to (-)-kainic acid among others [12][13][14][15][16]. On treatment with t-BuLi in the presence of N',N'-dimethyl-N',N'-propyleneurea (DMPU) or hexamethylphosphoric triamide (HMPA), amides 9 form benzyllithium intermediates.…”

Cyclisations of Organolithiums onto Aromatic Rings

“…Enolate 73 can be alkylated or protonated both regio-and stereoselectively, giving a 6,5cis-fused product and avoiding dearomatising the second aromatic ring of the naphthalene system. On the other hand, 77 was generally alkylated or protonated non-regioselectively, and indeed alkylation was also non-stereoselective, 33,44 forming difficult-to-separate mixtures of dienes. We had some limited success in controlling the regioselectivity of the protonation, but we realised that if the dienes were functionalised with methoxy groups, they would both be enol ethers 81 and therefore hydrolysable to give enones, with the regiochemistry of the double bonds in the hydrolysed products being under thermodynamic control.…”

“…Indeed, cyclisation of 75 (R = OMe) gave the enone 82 in 71-73% overall yield, even without isolation of the intermediate enol ethers 81 (Scheme 20) 44. …”