The intermediates and transition state for the deprotonation of acetonitrile, 7, with dimeric lithiumamide [LiNH 2 ] 2, 8, which leads to the dianion complex [CH 2 CN/NH 2 ] 2 - 2 Li +, 14, have been calculated by ab initio (MP2(full)6-31+G*//MP2(full)6-31+G*), DFT (Becke3LYP/6-311+G*, DGAUSS DZP/A1), and semiempirical methods. This reaction is initiated by the formation of a 4-ring-dimer/nitrile complex, 9, which rearranges to give an “open dimer” complex 10 . This step is (energetically) the most expensive, requiring 11.4 kcal/mol. The complex 10 is a suitable precursor for easy C to N proton transfer via transition state 11, which lies just 1.2 kcal/mol above 10 . The reaction proceeds via 12, a second “open dimer” and finally, after extrusion of NH 3, yields 14 . The PM3 and ab initio results for the relative energies of these structures are in acceptable agreement; the only exception is the overestimation of the stability of the complex 9 by PM3 (Figure 1). Therefore, PM3 allows comparison of the inter- and intramolecular versions of the deprotonation step with inclusion of solvation, Schemes 3 and 4 and Figure 2, indicating that the intramolecular pathway is preferred by 9 kcal/mol. To assess the reliability of the PM3 method in predicting the geometries of larger complexes in this series, PM3 and MNDO were applied to simulate the X-ray structures of model intermediates such as LiC 6 H 5 CHCN·LDA·2 TMEDA (Figure 3), [ t -BuCN·LiN(SiMe 3 ) 2 ] 2 (Figure 4), and [(LiC 6 H 5 CHCN·TMEDA) 2 ·C 6 H 6 ] (Figure 5). In all cases the LiX subunits are described significantly better by PM3 than by MNDO.
No takes yet. Share an insight, caveat, or question.
Koch et al. (1996) studied this question.