1,3-Disubstituted 7-azaindoles were synthesized from 2,6-dichloropyridine using DoM and intramolecular aromatic substitution after epoxide opening by an amine. Even the sterically demanding adamantylamine may be incorporated leading to derivatives which are not accessible by alkylation of the parent compound.In nature 7-azaindoles (1H-pyrrolo[2,3-b]pyridines) appear only as a substructure in a small number of alkaloids. However, much interest in this heterocyclic moiety has arisen in recent pharmacological programs, wherein it serves as bioisoster of indole or purine. 1 Numerous publications on its synthesis and derivatization reflect the increasing attention paid to this heterocyclic system. 2,3 In connection with a medicinal chemistry program, we required a practical synthesis of 1,3-substituted azaindoles of the general formula 1 (Scheme 1). Inspired by two reports on the synthesis of indoles from 2-(2-chlorophenyl)oxiranes 4 and 2-(2-bromophenyl)oxiranes, 5 we decided to develop a new azaindole synthesis based on the retrosynthetic analysis depicted in Scheme 1. An amine should open the epoxide 2 to an amino alcohol, which expels the adjacent halogen by intramolecular nucleophilic aromatic substitution. Subsequent dehydration should yield the desired heterocycle 1. We planned to start with 2,6-dichloropyridine (4) for the synthesis of oxiranes 2. The directed ortho metalation (DoM) of 4 should allow its functionalization in the 3-position. 6The lithiation 7 of fluoro-and chloropyridines 8 has been thoroughly studied by Quéguiner 9 and Schlosser. 10 In our hands, after deprotonation of 2,6-dichloropyridine (4) with LDA, the addition to ketones like acetone, acetophenone, or trifluoroacetone occurred smoothly (Scheme 2), and the desired tertiary alcohols 5 could be easily separated from unconsumed starting material by column filtration due to their higher polarity. In agreement with repeated reports in the literature, we detected small amounts of the regioisomeric adducts resulting from lithiation at the 4-position. In the course of our investigations it turned out to be unnecessary to separate the regioisomers, since the minor component leads to a polar, unproductive intermediate in the last step of the sequence. The dehydration of 5 yielding a-substituted styrene derivatives 6 was performed in a mixture of acetic and sulfuric acid (3:1) at 130°C for 30 minutes. The elimination products were isolated after basic work-up and did not need further purification. Unfortunately, 5c failed to dehydrate due to the strong electron-withdrawing effect of the trifluoromethyl group. Even after heating in concentrated sulfuric acid to 120°C for 14 hours we recovered the alcohol 5c almost quantitatively. With other reagents like thionyl chloride, thionyl chloride/pyridine 11 or Burgess' reagent we were also unsuccessful, demonstrating that this approach is not applicable for the synthesis of 3-trifluoromethyl 7-azaindole.The epoxides 2 were generated from the alkenes by oxidation with MCPBA in dichloromethane, and were essen...