Microsolvation and 13C-Li NMR coupling.

The empirical expression (1)J(CLi) = L[n(a + d)](-1) is proposed; it claims a reciprocal dependence of the NMR coupling constant (1)J((13)C, Li) in a C-Li compound on two factors: (i) the number n of lithium nuclei in bonding contact with the observed carbanion center and (ii) the sum (a + d) of the numbers a of anions and d of donor ligands coordinated at the Li nucleus that generates the observed (1)J(CLi) value. The expression was derived from integrations of separate NMR resonances of coordinated and free monodentate donor ligands (t-BuOMe, Et2O, or THF) in toluene solutions of dimeric and monomeric 2-(alpha-aryl-alpha-lithiomethylidene)-1,1,3,3-tetramethylindan at moderately low temperatures. This unusually slow ligand interchange is ascribed to steric congestion in these compounds, which is further characterized by measurements of nuclear Overhauser correlations and by solid-state structures of the dimers bearing only one donor per lithium atom (d = 1). Increasing microsolvation numbers d are also accompanied by typical changes of the NMR chemical shifts delta (positive for the carbanionic (13)C(alpha), negative for C(para) and p-H). The aforementioned empirical expression for (1)J(CLi) appears to be applicable to other cases of solvated monomeric, dimeric, or tetrameric C-Li compounds (alkyl, alkenyl, alkynyl, and aryl) and even to unsolvated (d approximately 0) trimeric, tetrameric, or hexameric organolithium aggregates, indicating that (1)J(CLi) might serve as a tool for assessing unknown microsolvation numbers. The importance of obtaining evidence about the (13)C NMR C-Li multiplet splitting of both the nonfluxional and fluxional aggregates is emphasized.