Abstract
Flexible model results for the NH2 inversion in vinylamine are compared with microwave data published by Lovas et al. [3]. The information available on the spacing of inversion levels is found to be sufficient to rule out a planar equilibrium structure. Shifts of rotational constants upon excitation of the inversion are calculated for two predicted paths of motion derived from PRDDO geometry optimization results obtained by MiiIler & Brown [2]. Partial geometry optimization yields a path inconsistent with experiment but the path obtained from complete optimization produces shifts in excellent agreement with the observed shifts. A potential energy function compatible with experimental data on spacing of inversion levels is calculated from the observed inertia defects. The inversion angle at equilibrium is slightly smaller than the angle predicted from PRDDO calculations, and the barrier to inversion, ΔV = 1.08 ± 0.07 kcal/mol, is approximately half the PRDDO value.