|Home | About | Journals | Submit | Contact Us | Français|
The dearomatization of arenes is recognized as a chemical transformation of fundamental importance for organic chemists since it allows efficient access to the alicyclic frameworks present in many biologically active compounds.1,2 In this context, complexation of the aromatic ring with stoichiometric amounts of transition metals,3 the oxidation of phenols4 and the reduction using metals in solution5,6 have been widely investigated. Despite the importance of these methods, their requirement to use stoichiometric amounts of metals or reactive reagents and/or the additional complexation/decomplexation steps needed to release the desired product remains a limitation. Therefore, the development of new dearomatization methods that operate under catalytic conditions and with high stereocontrol would be extremely valuable for the synthetic organic community.7 Herein, we present the first asymmetric transition-metal-catalyzed8 dearomatization to form all-carbon quaternary stereocenter.9
Our initial idea is depicted in Scheme 1. As shown, the deprotonation of aniline I would be expected to increase the electron density in the adjacent aromatic ring, allowing the intramolecular electrophilic aromatic substitution-type reaction with the palladium (II) center to generate 3aH-indole derivative III.
To ascertain the feasibility of this hypothesis, we started our investigations with the transformation of 1a into 6a-phenyl-6aH-benzo[a]carbazole 2a using various Pd sources and phosphine ligands.10 After some initial experimentation, we found that compound 2a was produced in 98% yield when 1a was treated with 3 mol% of Pd(OAc)2, and 4.5 mol% of SPhos in dioxane at 80°C using LiOt-Bu as a base.11
With these results in hand, we next focused our efforts on the use of different chiral ligands to effect an asymmetric version of this transformation. The use of bidentate ligands resulted only in the recovery of starting material 1a (Table 1, entries 1-2).12 Better results were obtained, however, when monodentate ligands were used (Table 1, entries 4-6). Indeed, when MOP (L4) was employed, compound 2a was obtained in 90% yield, albeit in only 21% ee. The enantioselectivity could be increased to 90% ee by using KenPhos13 (L5) as a ligand (Table 1, entry 6). Further optimization of these conditions led to the formation of benzocarbazole 2a in 96% yield and 93% ee when 1a was treated with Pd(dba)2 and KenPhos (L5) in THF (0.1 M) in presence of LiOt-Bu as a base (Table 1, entry13).
Encouraged by these initial findings, we next examined the scope of this transformation. First, we evaluated the effect of different substituents on the benzene ring. As shown in Table 2, both electron-donating (entry 2) and electron-withdrawing substituents (entries 3-4) formed the corresponding benzocarbazole derivatives 2a-f in good yields and enantioselectivities. Under these reaction conditions it was possible to obtain chlorine-substituted compound 2e in 65% isolated yield and 89% ee. The efficacy of this method decreased, however, with the more sterically hindered ortho-substituted benzene derivatives. Thus, the reaction of o-Me substituted 1f provided incomplete conversion to the corresponding benzocarbazole 2f in 62% yield and 66% ee.
Following these experiments, we focused our attention on the substitution on the naphthalene ring. Our initial protocol using L5 as a ligand provided the desired products in good yield even though in moderate enantioselectivity. After careful investigation we found that the use of bulkier ligand L6, in which one methyl group on the nitrogen had been replaced by an i-Pr residue, gave better results.14 Indeed, when compounds 3a-d were exposed to the standard conditions using L6 in lieu of L5, benzocarbazoles 4a-d were obtained in high yield and enantioselectivity (Table 3, entries 1-4). Again, the use of a more sterically hindered substrate resulted in a diminished ee (Table 3, entry 5). Compound 4c proved to be crystalline, allowing the determination of the absolute configuration by means of X-Ray crystallographic analysis.15
The synthetic potential of these benzocarbazole derivatives is shown in Scheme 2. As depicted, the 1,2-addition of MeLi gave rise to the corresponding compound in 80% isolated yield as a 9:1 mixture of diastereomers (based on GC and GC-MS analysis of the crude reaction mixture) which were separated by column chromatography.16 Protection of the secondary amine provided the crystalline compound 5 enantiomerically pure after crystallization.
In conclusion, we report the first asymmetric palladium-catalyzed intramolecular dearomatization reaction. The application of this new method to naphthalene derivatives led us to obtain benzocarbazole derivatives in high yields and enantioselectivities, making this method suitable for synthetic purposes. Further investigations into the mechanism of this reaction as well as extensions of the substrate scope are ongoing in our laboratories.
We thank the National Institutes of Health (GM 58160) for financial support of this work. J.G.-F. and F.K thank the Spanish M.E.C. and German Academic Exchange Service (DAAD) for the respective fellowships. We also thank Merck, Boehringer Ingelheim and Amgen for unrestricted support, as well as BASF for Pd(OAc)2. The Varian 300 MHz used in this work was purchased with funding from the National Institutes of Health (GM 1S10RR13886-01).
Supporting Information Available: Experimental procedures and spectral data for all compounds. This material is available free of charge via the Internet at http://pubs.acs.org.