Absence of stereoelectronic control in the hydrolysis of fully and partially N-alkylated cyclic amidinium ions

Article abstract of DOI:10.1021/ja00061a013

Deslongchamps' hypothesis of stereoelectronic control states that preferential cleavage of a tetrahedral intermediate occurs when a leaving group is antiperiplanar to two lone pairs. Yet substantial amounts of lactams were produced from hydrolysis of cyclic amidines, through cleavage of the exocyclic C-N bond that is antiperiplanar to only one lone pair and syn to the other. It may be that proton transfer catalyzes nitrogen inversion, leading to an intermediate with two antiperiplanar lone pairs, such that the product distribution is indeed consistent with stereoelectronic control. To exclude this possibility, three cyclic amidinium ions (1a,b,c) with methyl groups at both nitrogens were synthesized, and their hydrolysis products in NaOD/D₂O were analyzed by NMR. Although stereoelectronic control favors ring opening to amino amide 3, substantial amounts (40-90%) of lactam 4 are again produced. To exclude the possibility that lactam formation is due to a steric effect, hydrolysis of a cyclic N,N′-dimethyl amidine (2) was also studied, and 26-35% lactam is still produced. These results cannot be rationalized by assuming reaction via a boat conformer, since the lone pair of the ring nitrogen does not become antiperiplanar even in the boat form. Thus stereoelectronic control is not operative in amidine hydrolysis.