Organolithium (RLi) and -magnesium (R 2 Mg) compounds are kinetically poor bases for proton removal from weak carbon acids. If an amine like N,N,N',N'-tetramethylethylendiamine (TMEDA) is added, the barrier is sometimes lowered, but the nucleophility of the organometallic compound remains a problem. This can be ameliorated by using metal amides like lithium diisopropylamide (LDA) and bis(diisopropylamido)magnesium (DA 2 Mg). These poor nucleophiles are still kinetically effective bases for deprotonation of weakly acidic CH groups (pK a % 30 ยฑ 35). [1] However, as the pK a of the liberated amine and that of the CH acid (e.g. 1) are similar, such deprotonations are nowhere near stoichiometric (e.g. Scheme 1 ). This is unsatisfactory.
Scheme 1. The equilibrium reaction of LDA with amide 1.
We now introduce alkylmagnesium amides, here specifically BuMgNiPr 2 , denoted hereafter as BuMgDA, as an effective solution to this problem. We prepare BuMgDA simply by adding 1.0 equivalent of anhydrous diisopropylamine (DAH) to commercial [4] dibutylmagnesium in heptane (ca. 1.0 m) at room temperature and then stirring the solution for five minutes at 50 8C. BuMgDA [3b] in heptane is quite reactive. Replacing heptane, all or in part, with THF (after the base has been formed) increases this usefully. Unlike Bu 2 Mg and many organolithium bases, BuMgDA is stable even in refluxing THF for many hours. BuMgDA, like DA 2 Mg, deprotonates/metalates amide-activated [5] cyclopropyl-CH (a, b, or beyond) and cubyl-CH (ortho), but the BuMgDA deprotonations are driven to completion by irreversible formation of butane.
It is instructive to compare the metalation of the cyclopropylcarboxamide 2 using Bu 2 Mg solutions in heptane treated first with 0, 0.5, or 1.0 equivalent of DAH. Bu 2 Mg itself reacted only slowly; [7] mostly starting material was recovered. In the other two cases, when at least some BuMgDA [3b] was present, the overall deprotonation/metalation/carboxylation proceeded in high yield, but the final