Dual Gold Catalysis

Abstract

For more than a decade the innovative field of homogeneous catalysis by gold was dominated by the interaction of the substrate molecule with one gold center, in most cases in mononuclear gold complexes. The initial interaction was typically a π-coordination of a carbon-carbon double bond to the gold, an activation of the unsaturated substrate molecule by a π-acidic metal center. Only recently clear evidence for reactions that involve the activation of organic substrates by two gold centers was obtained. In that new class of gold-catalyzed reactions the two gold centers interact with the substrate in a very different way. One gold complex is σ-bonded to a terminal alkynyl group in the substrate, the other one is π-coordinated. Only in a few cases, a combination π-coordination and σ-coordination to the same alkyne, which is the energetically preferred mode of interaction with two gold centers, initiates the reaction. In most of the cases, the reaction proceeds through an intermediate with one alkyne σ-bonded to one gold complex and a different alkyne π-coordinated to the second gold complex. Experimental and computational results for many new reactions provide a clear picture of the overall sequence of elemental steps of these conversions; some of the steps are unprecedented in organometallic catalysis and chemistry. For example, the reaction of diynes can involve gold vinylidene-like species as very reactive substructures formed by a 5-exo-dig cyclization. NBO analysis indicates no gold-carbon double bond character in these "vinylidenes". In other reactions, a 6-endo-dig cyclization is energetically preferred; after that gold aryne complexes are not local minima but transition states of a 1,2-shift of gold. Computational studies showed a good correlation of the cyclization mode with the aromaticity of the intermediate. For both the 5-exo-dig and the 6-endo-dig cyclization modes, the intermediates are able to react even with unactivated alkyl-C,H bonds, in low yields even in intermolecular reactions. The final step of the catalytic cycles is also remarkable, because the protodeauration has to occur with the next alkyne substrate molecule. Only then the next gold acetylide is formed directly and a loss of selectivity can be avoided. A computational study suggests that two gold complexes are on the substrate throughout the catalytic cycle. The most recent results indicate that analogous intermediates can be accessible by the reaction of other electrophiles with gold acetylides. With regard to organic synthesis, the overall catalytic conversions open up a universe of new possibilities. Selective C,H-activations now allow to one use usually innocent alkyl side chains for additional anellation reactions by an sp(3)-C,H activation. The C,H activation can even be combined with halogen transfer reactions, directly providing vinyl iodides as versatile building blocks. Short and efficient routes to different carbo- and heterocycles including benzocyclobutenes, fulvenes, and pentalenes demonstrate the synthetic potential not only for total synthesis but also for material science.