The separate, and combined, infhrences of ether, silyl ether and cyclic ketal groups at positions 3 and 5 on the mode and efficiency of Gheptenoyl radical cyclizations have been studied with a view to the eventual synthesis of 1~@dihydroxyvit D,. With a view to the eventual synthesis of the importa
Acyl radical cyclizations in synthesis. part 2 further substituent effects on the mode and efficiency of cyclizaion of 6-heptenoyl radicals.
โ Scribed by David Crich; K. Angeline Eustace; Simon M. Fortt; Timodthy J. Ritchie
- Publisher
- Elsevier Science
- Year
- 1990
- Tongue
- French
- Weight
- 769 KB
- Volume
- 46
- Category
- Article
- ISSN
- 0040-4020
No coin nor oath required. For personal study only.
โฆ Synopsis
Abstrach In bn attempt to determine the factors affecting exe-lendo-selectivity in the cycllzation of 6-heptenoyl radicals various heteroatom substituted selenol esters were prepared and reacted with tributyltin hydride. The incorporation of a 7-phenylthio moiety results in clean, high yielding, cycllxation in the exe_ mode. Evidence is adduced for the reversibility of 6-heptenoyl radical cyclbzations. In part 1' of this series we recorded the effects of ether and cyclic ketal functions at the 3-and/or 5positlons of the 6-heptenoyl radical on the mode and efficiency of it's cyclixation. We report here, in full: the results of some further experiments involving heteroatom substituted 6-heptenoyl and related radicals undertaken with a view to the gaining of greater insight into this problem. The 5,Sethylenedioxy and J-ethoxy substituted selenol esters (1) and (3) undergo preferentlal &Qmode cyclixation to the cycloheptanones (2) and (5) on treatment with tributyltin hydride and a catalytic quantity of azoisobutyronitrite (AIBN) in benzene at retlw (Table, entries 1 and 2).' PhSc 3 0 (1) PhSc OEt 0 c, 'OEt 2135 2136 D. CRICH etd. These unanticipated observations might be ascribed to any one of the three possibilities outlined in scheme 1: i) kinetic cyclization in the -mode fdlowed by rapid ring expansion to the overall gn& product (Scheme 1, path a) as is known3 to happen with the closely related vinyl radical cyclizations; ii) direct *mode cyclization (scheme 1, path b); or iii) reversible cycliition in the -mode followed by eventual endo_cyclization (Scheme 1, path c). Ring expansion of radicals of the type B (Scheme 1) to C has been reported' when the group 2 is electron withdrawing (CO,R, CN) and X = Y -II and also in the related example of the expansion of radical (6) to (7) (Scheme 2): Scheme 2 Direct *mode cycliition (Scheme 1, path b) might be favoored by the adoption of an appropriate conformation by the allylic ether moiety.' Both ring expansion and direct &mode cycliition might also be driven by the stabilisation which it has been suggested ' is afforded to carbon centred radicals by a p C-O bond. Acyl radical cyclixatlons in synthesis -II 2137 The removal of the ether and /or ketal groups from the 5-(allyllc) position to the 3-position as in (8) and (11) results in a drama& increase in yield coupled with a change In cycliaation mode (Table, entries 3 and 4).' PhSc 'BuPhgO (11) Clearly the increased yield is readily attributable to the 'Thorpe&gold' effect but the change in overall cyclixation mode may be a consequence of the removal of steric hindrance at the internal alkene position; of the removal of conformational constraints imposed by allylic ethers; or of the lack of a stabiiising @ C-O bond for the endo_pmduct. In an attempt to distinguish between conformatlonal effects due to the allylic oxygens in A(!Icheme 1) and extra stabilization afforded by the g C-O bonds in C (Scheme 1) as the overriding regiodirecting factor in the cyclixation of (1) and (3) we decided to prepare the seienoi eater (14) and to study it's AIBN initiated reaction with tributyltln hydride. The removal of the ailylic oxygen from an exe-cyclic position as in (l)+(2) and (3wS) to an endo_cyclic position as in (14)+(16) would not result in the less of any stabiilsation of the f&-mode ring closed radical by a /S C-O bond. Thus if this stablhsation were the predominant factor then (16) should predominate over (15). On the other hand it seemed reasonable to assume that any conformational effect due to the allylic oxygen in (1) and (3) when translated to (12) would favour -mode cyclixatkm. Furthermore it is known from the prototypicai' !i-hexenyl radical rearrangement that the inclusion of an allyllc oxygen vy&l& the chain increases that rate of ~cycllxation substantially more than that of ~cyclixatlon (Scheme 3).' ' k,fldo 4.1 x Id i' krxo exo:endo S7:I (25%) eXo :rmdo v85:i (25ยฐC) Scheme 3. In tbe event treatment of selenol ester (14), prepared by the action of sodium phenyk~en~t~th~y)b (~i~rn pheny~en~e) on the acyi chloride derived from 3.~~yioxyp~pionk acid, with tributyltin hydride and AIBN in benxene at reflux resulted in the isolation of the oxacycioheptanone (16) and the oxacyciohexanone (15) in 15 and 12% yields respectively (Table, entry 5). In a similar vein the seienoi esters (17~(19) were prepared from ~iy~~~ic acid, S-~ly~iey~k acid and N-afetyi-N-atlytan~~nnic acid in the fhst two cases by reaction of the derived acyi chloridtur with sodium phenykeienide and in the latter case by the reaction of a mixed anhydride" prepared from the acid and kobutyl e~o~f~rna~ with sodium p~n~en~~ In each case the onty product kohrted on reaction with tributyitin hydride and AXBN at WC resulted from -mode cyclization (Table, entries 6-S). Reaction of (17) with the activated allyistannane (23)" in a tandem intiintermdecuiar radical addition procear also provided the exo_mode product (Table, entry 9). Analogous results have been observed by other workers with closely related systems.D Clearly the inclusion of two further sp% centres in the chain overrides any other effect and directs eycliition to the -mode.
๐ SIMILAR VOLUMES