2.4. Computational studies.
In order to get a better understanding of the behaviour of these chiral catalysts, we have carried out DFT23 calculations on the model reaction depicted in Scheme 1 and entry 7 of Table 1, using the DFT functional usually denoted as B3LYP.24 Silver atoms were treated using the Hay and Wadt effective core potential and basis set25 (denoted as LANL2DZ), whereas the remaining elements were described using the standard 6-31G* split-valence basis set.26 All the calculations were carried out using the GAUSSIAN 03 suite of programs.27 The graphics shown in Figures 6 and 7 were built up using the Maestro interface.28 In the model systems A-C we have included the cationic part of the presumed active complex (S)-16 as well as a model azomethine ylide and maleimide B.
The chief geometric features of complex (S
are gathered in Figure 6. The azomethine ylide part of (S
shows different distances for the two C-N bonds. These distances are compatible with an iminium-enolate structure as shown in Scheme 6, thus anticipating quite asynchronous transition structures in the reaction with the dipolarophile. It is also observed that the metallic centre is coordinated to the two phosphorus atoms of the catalysts and to the oxygen and nitrogen atoms of the azomethine ylide. This coordination pattern leads to the blockage of the Re
face of (S
by one of the phenyl groups of the phosphine unit (Figure 6). This steric hindrance is also generated by the (S
)-Binap moiety, in which the two -naphthyl subunits form a dihedral angle of ca. 75 deg.