Insertion of an alkene into an ester: intramolecular oxyacylation reaction of alkenes through acyl C-O bond activation.

lation” could be an alternative to the aldol reaction. Atom economy and ester manipulation, however, are rarely compatible: esters usually fragment after reactions with nucleophiles, or decarbonylate when activated with transition metals. In the rare cases when the acyl C O bond is activated and decarbonylation is suppressed, the acyl metal alkoxide complexes can undergo additional transformations, but only with the expulsion of an alcohol (Scheme 2a) or ketone (Scheme 2b). We are aware of one example where acyl C O activation provided products containing the original atoms: Ohe s recent Pd-catalyzed nitrile insertion into an acyl C O bond, followed by rearrangement (Scheme 2c). The challenge of productive acyl C O bond activation is accentuated by the frequent reports of the reverse reaction: when acyl metal alkoxides are accessed by other means, they readily undergo reductive elimination to form esters. We postulated that a chelating group would prevent decarbonylation by stabilizing the acyl metal complex, I. Our approach was akin to stoichiometric acyl C O bond activation strategies in which metal chelating groups were used. We employed quinoline as a chelating group based on our previous successes in acyl C C bond activation. We designed an intramolecular reaction to avoid problems with regioselectivity and to increase local concentrations of alkene. Activation of 1a to I, followed by migratory insertion and reductive elimination would provide 2a, containing a cyclic ether with a ketone in a 1,3-relationship and a new fully substituted carbon center. Here, we report the first example of alkene insertion into the acyl C O bond of an ester. Beginning with 1a, we screened Rh complexes containing various counterions (Cl, BF4, OTf), with [Rh(cod)2]OTf (cod= cyclooctadiene) providing encouraging results (Table 1, entries 1–3). A byproduct observed in our initial study was the phenol 3a, resulting from a formal hydrolysis of 1a, though attempts to rigorously exclude water did not decrease the formation of 3a. Switching to 1,2-dicholorethane as solvent, [Rh(cod)2]BF4 showed good conversion, but gave a 1:1 mixture of 2a :3a (Table 1, entries 4, 5). The addition of bidentate phosphine ligands, particularly dppp, was effective at maintaining high conversion. Using the [Rh(cod)2BF4]/dppp catalyst system at higher temperature suppressed the formation of 3a (Table 1, entries 8–11). Using the conditions from Table 1, entry 11, we examined the scope of oxyacylation (Table 2). Both electron-donating and electron-withdrawing substituents on the aromatic linker gave products 2b–g in good yields (Table 2, entries 1–6), although longer reaction times were required for electronScheme 1. Alkene oxyacylation.