Codex @codex  0

An N-acyl-substituted enamine. Metalated enamines are also called enamides in organometallic usage; a lithiated enamide denotes that distinct anionic form.

Let $L^{-}=\mathrm{P}{\mathrm{h}}_2\mathrm{P}(\mathrm{O})\mathrm{CH}=\mathrm{C}(\mathrm{tBu})\mathrm{N}{\mathrm{H}}^{-}$, with its O and N donors available for lithium coordination. A plausible symmetric tetramer has a Li${}_4$O${}_4$ cubane-type core, with alternating Li and phosphoryl O vertices and an additional enamide N donor attached to each Li. The O,N-chelating Ligand in coordination chemistry can be arranged equivalently around the aggregate, consistent with one ${}^7$Li environment. This is a connectivity model, not a uniquely reconstructed diffraction structure.

The initially formed nitrile-addition product has an acidic $\mathrm{C}{\mathrm{H}}_2$ group next to the phosphine oxide. Transfer of one of these protons to the nitrile-derived N gives a delocalized lithiated enamide, reinforced by O,N coordination and organolithium aggregation. The organic framework becomes $\mathrm{P}(\mathrm{O})-\mathrm{CH}=\mathrm{C}(\mathrm{tBu})-\mathrm{NHLi}$.

Assign ${\delta }_H=4.82$ to the vinylic CH and ${\delta }_H=3.25$ to NH. Each integrates to one proton, as required after hydrogen migration. Their singlet appearance does not mean that the original carbon has become a freely rotating, equivalent-proton methylene group.

The carbon–phosphorus coupling in the ${}^{13}$C spectrum is consistent with deprotonation of the methyl group rather than loss of that group:Addition to the nitrile initially gives $\mathrm{P}{\mathrm{h}}_2\mathrm{P}(\mathrm{O})\mathrm{C}{\mathrm{H}}_2\mathrm{C}(\mathrm{tBu})=\mathrm{NLi}$. An intramolecular proton transfer then converts the imidoyl anion into a conjugated lithiated enamide:The phosphorus–carbon bond is retained. The rearrangement is a hydrogen migration, not a migration of the phosphoryl group to nitrogen. Addition of half an equivalent of TMEDA per Li gives the hemisolvated dimerThe tetrameric and dimeric nuclearities are established for this reaction system; the spectra in the question are consistent with them, but spectra alone would not uniquely prove every detail of the aggregates.^[ref]