Studies on intramolecular (4+3) cycloadditions of geminally-tethered epoxy enolsilanes and the synthesis of himandrine

Abstract

Epoxides geminally-tethered by three carbon atoms to diene substrates were prepared to investigate their intramolecular (4+3) cycloadditions (Scheme 1). Under silyl triflate catalysis, seven such substrates (2.13, 2.15-2.20) underwent exo-selective (4+3) cycloadditions to provide perhydroazulene bicyclo[5.3.0]decane frameworks with up to 82% yield. Even hindered furans underwent intramolecular (4+3) cycloadditions successfully. Cycloaddition failed only for those substrates which are hindered on both the furan and the enolsilane. Substrates that have terminal substituents in the enolsilane moiety undergo cycloaddition to generate epimeric Type i and Type ii cycloadducts which were shown to undergo equilibration at the stereocenter α-to the carbonyl group. However, the cycloadditions of epoxide substrates such as 2.14 geminally-tethered by four carbon atoms to diene substrates that could afford bicyclo[5.4.0]undecane systems did not proceed. Substrates tethered to thiophenes or benzenes such as 2.25 and 2.27 underwent Friedel–Crafts alkylation only, instead of cycloaddition. The reaction of epoxide substrates such as 2.72 geminally-tethered to double bonds did not undergo intramolecular (3+2) cycloaddition to generate carbocycles but instead resulted in a tetrahydrofuran cyclization product 2.73 in a moderate yield and with good diastereoselectivity. Based on the results, a stepwise reaction mechanism was proposed to rationalize the formation and diastereoselectivity of these reactions. These results underscore that the length of the tether leading to the formation of the first ring, is critical to a successful cycloaddition. We have also further investigated the intramolecular (4+3) cycloaddition of pyrroles aiming to apply it as a key reaction in the synthesis of the natural product, himandrine. We have achieved the synthesis of enantiomerically enriched epoxy enolsilane intermediate 3.215 in 13 steps from pyridine oxide 3.174. Some key reactions included a photochemical rearrangement of 3.174 to yield a pyrrole carboxyaldehyde 3.168, an organocatalytic Michael addition-Wittig reaction to afford cyclohexenone 3.200 in 91% ee, and a decarboxylative allylation to generate trans-3.208. However, it was found that intramolecular (4+3) cycloaddition of 3.215 did not proceed (Scheme 2). We also prepared in 11 steps from the commercially available 2-formyl pyrrole, a related substrate 3.251 which was less hindered by having a less-substituted pyrrole. In addition, decarboxylative allylation generated the cis-3.230, instead of the trans-isomer, ultimately leading to 3.251 also being a cis-isomer. The intramolecular (4+3) cycloaddition of 3.252 under silyl triflate catalysis afforded cycloadduct 3.253, albeit in low yield. From these studies, it was learned that the A ring is a factor obstructing the intramolecular (4+3) cycloaddition due to the lack of flexibility of the resulting decalin system. A model study of a Type V transannular (4+3) cycloaddition was initiated but could not be completed due to the failure to induce a successful epoxidation in the presence of the electron-rich pyrrole. Further experiments and investigations are needed to complete the total synthesis of himandrine.