The enamide group is an important substructure that is often found in natural products [1] and synthetic drugs that have, amongst others, sedative, [2] cytotoxic, [3] or anti-inflammatory [4] properties. Moreover, they are versatile synthetic intermediates that can serve as substrates for polymerizations, [5] [4+2] cycloadditions, [6,7] cross-coupling reactions, [8] Heck olefinations, [9] halogenations, [8] enantioselective additions, [10] or asymmetric hydrogenations. [11] However, a regioand stereoselective construction of enamide substructures is not at all trivial. Traditional syntheses-for example, from carbonyl compounds and amides [12] or from hydroxylamines and acetic anhydride [13] -require harsh conditions and yield mixtures of E/Z products. Metal-catalyzed coupling reactions of vinyl halides, [14] vinyl triflates, [15] or vinyl ethers [16] proceed under milder conditions but suffer from the limited availability of these substrates.
In our opinion, a catalytic addition of amides to alkynes would be an ideal synthetic entry to enamides, since it would use readily available starting materials and be inherently atom-economic (Scheme 1). Related addition reactions of carboxylates, [17] water, [18] and amines [19] are well-established. [20] However, to the best of our knowledge, there is only one reported protocol in the literature for a catalytic hydroamidation reaction of terminal alkynes: [21] Watanabe et al. found that in the presence of Ru 3 CO 12 /trialkylphosphane catalysts, a very limited range of formanilides and acetanilides can be added to 1-hexyne, albeit at extremely high temperatures (180 8C) and under pressure. [22] Clearly, much more effective catalyst systems are required to allow an application of this reaction in organic synthesis.
To identify a catalyst system for the desired hydroamidation reaction, we chose the reaction of 1-hexyne (1 a) with 2pyrrolidinone (2 a) as a model system and investigated the catalytic activity of several ruthenium complexes under various conditions (Scheme 2, Table 1). As anticipated, no