Stereoselectivity of ring closure of substituted hex-5-enyl radicals

Athelstan L. J. Beckwith and the Flowering of Hex-5-enyl Radical Cyclization Chemistry. The Adelaide Years

“…[29] Cyclizations of hex-5-enyl radical bearing simple alkyl or alkenyl substituents at the non-reacting C-2 15, C-3 16, and C-4 17 positions (Scheme 7) also proceed at enhanced rates compared with the parent unsubstituted radical 1. [22,[31][32][33] Ph.D. student Tony Lawrence conducted a systematic study of the effect of methyl substituents at all three positions and found modest rate enhancements (1.9-3.5) for single methyl groups and more substantial ones (10.8-16.3) for geminal dimethyl groups. Other singly substituted (vinyl, allyl, propyl) hexenyl radicals had roughly comparable relative cyclization rates to those with single methyl groups at the same position.…”

“…[26] Fairly quickly, examples of hexenyl systems with substituents at C-2 and C-3 undergoing stereoselective cyclizations were produced by other members of the Beckwith group. Hexenyl radicals with substituents at the 2-position 15 (R ¼ methyl, [22,33] vinyl [33] ) cyclized preferentially to the trans-1,3disubstituted cyclopentanes 19t, while those with 3-substituents 16 (R ¼ methyl, [22,33] vinyl, [33] allyl, [31] propyl [33] ) gave mainly cis-1,3-disubstituted cyclopentanes 19c (Scheme 7). The major: minor isomer ratios varied with position and substituent and ranged from 1.7 to 4.8.…”

“…The equatorial conformers 15e-18e will lead to the major products, trans (19t and 20t) from the 2-and 4-substituted radicals 15 and 17, and cis (19c and 20c) from the 1-and 3substituted radicals 18 and 16, and in each case the axial conformers 15a-18a will produce the minor opposite stereoisomer (Scheme 8). [33] At the time this model was first proposed, there were no known examples of 4-substituted hex-5-enyl radical 17 Hex-5-enyl Radical Cyclization Chemistry cyclizations, and the predominant formation of a trans-1,2disubstituted product from such a cyclization was only a prediction. Tony Lawrence was quickly set the task of testing this prediction, and soon showed that 4-methylhex-5-enyl radical 17 (R ¼ CH 3 ) performed as predicted, producing a high selectivity (4.8:1) for cyclization to the trans-product 20t (R ¼ CH 3 ).…”

“…Tony Lawrence was quickly set the task of testing this prediction, and soon showed that 4-methylhex-5-enyl radical 17 (R ¼ CH 3 ) performed as predicted, producing a high selectivity (4.8:1) for cyclization to the trans-product 20t (R ¼ CH 3 ). [22,33] There was initially some doubt whether 1-substituted hexenyl radicals should be included in this model because there was a feeling in some quarters that the disadvantage of a strong eclipsing interaction between C-6 and R in the equatorial-like transition state conformer 18e would overwhelm the disadvantage of an axial conformer 18a, and thus the observed cispreference must arise from other causes, such as the secondary orbital interaction proposed earlier. However, later theoretical work at ANU by Athel and Ph.D. student Carl Schiesser confirmed that the R-C-6 eclipsing interaction was in fact less important than the non-bonded interactions implicit in the above-described model, and that these latter did indeed favour conformer 18e.…”

“…[33] Radical 37 cyclized to an initial mixture of cis-and transcyclopentane isomers 38c and 38t, with the former predominating and undergoing further cyclization to form only the exo-and endo-isomers (1.7:1) of the 2-norbornylmethyl radical, 39 and 40. The second ring closure now constitutes a hexenyl cyclization, as opposed to heptenyl for 34c, but the apparent differences in regioselectivity (all exo) and stereoselectivity (two epimers) are nevertheless remarkable.…”