|Home | About | Journals | Submit | Contact Us | Français|
Enzymatic prenyl transfers are heavily trafficked biosynthetic pathways,1 yet synthetic catalysts for direct C-H prenylation are unknown.2-4 Having recently developed protocols for catalytic carbonyl prenylation employing 1,1-dimethylallene as the prenyl donor,5 catalytic ortho-C-H insertion of 1,1-dimethylallene was viewed as a potential strategy for the direct prenylation of aryl C-H bonds. Due largely to the pioneering efforts of Murai,6 Lewis base directed catalytic C-H activation initiated olefin insertions and alkyne insertions are well developed. To our knowledge, analogous C-H activation initiated insertions of allenes remain unexplored.2-4,6,7 Here, we report that cationic iridium complexes catalyze the coupling of 1,1-dimethylallene to aromatic carboxamides, heteroaromatic carboxamides and α,β-unsaturated carboxamides to furnish the products of C-H prenylation in good isolated yield as single isomers.8
To avoid complications potentially arising from over-prenylation, initial studies focused on amides derived from ortho-methoxybenzoic acid. To our delight, it was found that the cationic iridium complex assembled from [Ir(cod)2]BArF and rac-BINAP catalyzes the coupling of N-benzyl amide 1a and 1,1-dimethylallene in THF solvent at 120 °C in 78% isolated yield. Cationic iridium complexes were required for coupling. One complication, however, is that prenylation of N-benzyl amide 1a delivers a 3:1 mixture of olefin regioisomers 2a and 3a, respectively (Table 1, entry 1). Based on the notion that isomerization occurs by way of amide-directed allylic C-H insertion of 2a, it was reasoned that tertiary amides may suppress isomerization by twisting out of plane upon introduction of the prenyl moiety, which should attenuate the amides directing influence. Consistent with this hypothesis, it was found that primary amides consistently provide good yields of prenylation product, but as mixtures of olefin regioisomers 2a and 3a, whereas tertiary amides, such the dibenzyl amide 1b and amide 1d derived from morpholine, do not participate in the coupling (Table 1, entries 2 and 3). Presumably, amides 1b and 1d are too large and nonbonded interactions evident in advance of prenylation are sufficiently strong to twist the amide out of plane and prohibit coupling. It was postulated that amide 1e derived from pyrrolidine should be smaller and more Lewis basic, potentially providing an ideal combination of steric and electronic features. Indeed, upon exposure of amide 1e to 1,1-dimethylallene under the aforementioned conditions, the desired product of C-H prenylation 2a was obtained in 70% isolated yield as a single isomer (Table 1, entry 5).
To evaluate the scope of this process, these conditions were applied to a diverse set of aromatic carboxamides 1e-1m. In each case, good to excellent yields of the desired aryl C-H prenylation product were obtained. Remarkably, ortho-substituted aryl carboxamides are not required to suppress isomerization or over-prenylation (Table 2, adducts 2g, 2i, 2j-m). Indeed, the parent benzamide 1m is prenylated under these conditions in 84% isolated yield using only 2 mol% of the iridium catalyst. Only in the case of the meta-methoxy derivative 1f was a mixture of isomeric adducts obtained, yet the corresponding meta-nitro derivative 1k undergoes prenylation to furnish a single isomeric product 2k.9 Additionally, the naphthoic amide 1l prenylates at the more hindered position. These data suggest electronic features of the substrate strongly influence regioselectivity.
To further evaluate the scope of this process, catalytic aryl C-H prenylation of heterocyclic aromatic carboxamides 1n-1p was attempted.10 In each case, the prenylated adducts 2n-2p are obtained in good isolated yields as single isomers. The conversion of indole 3-carboxamide 1p to the 2-prenyl derivative 2p is significant given the abundance of prenylated indoles in Nature (Table 3).11 α,β-Unsaturated carboxamides 1q-1s also were examined.12 The desired adducts 2q-2s were generated in good isolated yields as single isomers (Table 4).
To corroborate the catalytic mechanism, deuterio-1m was subjected to standard conditions for C-H prenylation. As anticipated, deuterium was transferred to the vinylic position of the adduct deuterio-2m (eqn. 1). This result is consistent with a catalytic mechanism involving ortho-C-H oxidative addition followed by allene hydrometallation to furnish an aryl-allyl iridium complex, which upon C-C reductive elimination from the primary σ-allyliridium haptomer delivers the product of prenylation with regeneration of cationic iridium(I). Incomplete levels of deuterium incorporation may arise via β-hydride elimination from the tertiary σ-allyliridium haptomer of the aryl-allyl iridium intermediate.
In summary, we report the first catalytic C-H activation initiated C-C coupling of allenes, as demonstrated by the direct prenylation of aromatic, heteroaromatic and α,β-unsaturated carboxamides. Future studies will focus on the development of related transformations where redistribution of hydrogen serves to promote byproduct-free C-C bond formation.
Acknowledgment is made to the Robert A. Welch Foundation and the NIH-NIGMS (RO1-GM069445) for partial support of this research. Eduardas Skucas thanks the Livingston Endowment for a graduate fellowship. Yongjian Zhang acknowledges generous financial support from Shanghai Jiao Tong University.
Supporting information available: Experimental procedures and spectral data for new compounds. This material is available free of charge via the internet at http://pubs.acs.org.