The mechanism of the ketene-imine (Staudinger) reaction in its centennial:: Still an unsolved problem?

Although Staudinger reported the reaction between ketenes and imines 100 years ago (1907), this process is still the most general and useful method for the synthesis of beta-lactams and their derivatives. This reaction is a [2 + 2] thermal cycloaddition in which two chiral centers may be generated in one preparative step. Staudinger reactions involving alpha,beta-unsaturated imines or ketenes have issues concerning the [2 + 2] or [4 + 2] periselectivity of the reaction, This Account discusses how the main factors that determine the regiochemical and stereochemical outcomes of this reaction were elucidated with computational and experimental data. This fruitful interplay between theory and experiment has revealed that the [2 + 2] cycloaddition is actually a two-step process. The first step is a nucleophilic addition of the nitrogen atom of the imine on the sp-hybridized carbon atom of the ketene. This attack forms a zwitterionic intermediate that evolves toward the final beta-lactam cycloadduct. The second step can be viewed as a four-electron conrotatory electrocyclization that is subject to torquoelectronic effects. When alpha,beta-unsaturated imines are used, the zwitterionic intermediates yield either the corresponding 4-vinyl-beta-lactams or the alternative 3,4-dihydropyridin-2(1H)-ones. In this latter case, the cyclization step consists of a thermal disrotatory electrocyclization. In the context of stereoselectivity, it is usually assumed that the first step takes place through the less hindered side of the ketene. The cis-trans selectivity of the reaction depends on the geometry of the imine. As the general rule, (E)-imines form cis-beta-lactams whereas (Z)-imines yield trans-beta-lactams. Most of the experimental results point to the two-step model. The asymmetric torquoselectivity of the conrotatory ring closure of the second step accounts for the stereochemical discrimination in the reaction of chiral ketenes or chiral imines. Nevertheless, recent studies have revealed that isomerization paths in the imine or in the zwitterion may determine the stereochemistry of the reaction. Thus, if the rotation about the N1-C4 bond of the zwitterion intermediate is faster than the cyclization, the formation of trans-beta-lactams from (E)-imines is biased. Alternatively, in some cases, the (E)-(Z) isomerization of the starting imines prior to the cycloaddition steps also results in the formation of trans-cycloadducts. Although the main variables that govern the outcome of the reaction have been elucidated, there are still several aspects of the reaction yet to be disclosed. Finally, the discovery of the catalytic version of the reaction is a new and formidable mechanistic challenge and will be a nice playground for forthcoming theoretical-experimental discussions.