Correction to Synthesis of Heterocycles via Pd-Ligand Controlled Cyclization of 2-chloro-N-(2-vinyl)aniline: Preparation of Carbazoles, Indoles, Dibenzazepines and Acridines

Dmitry Tsvelikhovsky and Stephen L. Buchwald^* [J. Am. Chem. Soc., 2010, 132, 14048–14051]

We recently reported a synthesis of carbazoles (8), acridines (9), and dibenzazepines (1) from common precursor 7 via Pd-catalysis.¹ The catalyst based on L2 gave exclusively 1, while catalysts based on L3 and L4 selectively formed 8 and 9 respectively (Figure 1).

Figure 1.

Previously reported, ligand controlled cyclizations of 2-chloro-N-(2-vinyl)aniline (7).

Upon reexamination of both the Pd/L3 and Pd/L4 catalyst systems, we have found that these systems are not 100% selective as reported.

The reaction of 7 using L3 gave 8 in 55% isolated yield (94% reported). The ¹H NMR spectrum of the crude reaction mixture showed 8, 1, and 9 in an approximate ratio of 1.0 : 0.25 : 0.20. This product distribution was largely insensitive to temperature (90–110 °C range) and other tested variables.

The reaction of 7 using L4 gave 9 in 45% isolated yield (87% reported). The ¹H NMR of the crude reaction mixture showed 9, 1, 8, and a previously unidentified compound in an approximate ratio of 1.0 : 0.18 : 0.05 : 0.60. The latter was identified as dimer 2 (X-ray)). Again, the product ratio was insensitive to slight variations in the reaction conditions.

The reaction of 7 using L2 furnished dibenzazepine 1 in 80% isolated yield (99% reported) and with good selectivity (~90%) by ¹H NMR analysis of the crude reaction mixture. The one-pot conversion of 2-chloroaniline and 2-bromostyrene to 1 (Table 1 in manuscript) occurred in 70% isolated yield (99% reported).

In summary, the formation of 1 from 7 is largely selective, but the yields may be somewhat lower than what is reported. The reactions that form 8 and 9, in contrast to what we originally reported, form mixtures of products with the major isomer being the one indicated in our published manuscript. The supposition that catalysts based on L2, L3 and L4 each produce a single product is not correct. We apologize for these errors and for any inconvenience to others that this may have caused. We thank Drs. Thomas Maimone and Meredeth McGowan for their invaluable assistance.

Figure 2.

Dimeric dibenzazepine side product 2.