Über die Wechselwirkung zwischen Lithium‐enolaten und sekundären Aminen in Lösung und im Kristall

Abstract

On the Interaction between Lithium Enolates and Secondary Amines in Solution and in the CrystalWhen lithium derivatives generated by lithium diisopropylamide (LDA) are quenched with a D+ source (large excess of ROD), there is frequently only partial deuterium incorporation (non‐stoichiometric effect, Scheme 1). Experiments with tert‐butyl 2‐methylpropionate (1) and its deuterated derivative (2‐D)‐1 and LDA indicate that the lack of deuteration is not caused by an isotope effect (Scheme 2). A H‐bonded complex between the amine (R2NH) and the Li‐derivative was envisioned as being responsible for the observed effects. Some crystalline Li‐enolate solvates with secondary amines were isolated, and gave rise to only partially deuterated products upon quenching with AcOD/CD3OD (Table 1). Enolates containing the Li‐chelating diamine N,N,N′‐trimethylethylenediamine (TriMEDA) turned out to be especially suitable for the isolation of single crystals. Thus, two X‐ray crystal‐structure analyses were undertaken to obtain detailed information about the geometry and bonding in such complexes. The structures of the dimeric N,N‐dimethyl‐propionamide lithium (Z)‐enolate ([2 (Z)‐4·2 TriMEDA]; R value 0,049, see Eqn. 2, Figs. 2, 4 and (5) and of the dimeric pinacolone lithium enolate ([2 6·2 TriMEDA]; R value 0,082, see Eqn. 3 and Figs. 3, and 6) were determined. In both centrosymmetric dimeric structures there is a Li‐O‐Li‐O four‐membered ring and a H‐bridge between TriMEDA and the enolate moiety. In the case of the amide enolate (Z)‐4·TriMEDA the bridge points from the amine N‐atom to the pyramidalized, enamine‐type N‐atom of the enolate (N …︁ N distance 3.173(1) Å, H …︁ N distance 2.21 (Å). In the complex 6·TriMEDA of the ketone enolate there is a H‐bridge pointing from the amine N‐atom to the enolate double bond (N …︁ Cterminal distance 3.45(1) Å, H …︁ Cterminal distance 2.62 Å; see Table 3). The geometries of the TriMEDA‐Li fragments in the two structures are compared with those of standard TMEDA‐Li moieties (Table 2). The two structures are discussed and interpreted using the structure correlation principle; the intermolecular interactions seen in these complexes are considered as points on the reaction coordinates of conversions between the species involved. Thus, they resemble the approach of a secondary amine in the protonation of an enamine N‐atom and of an enolate C‐atom. At the same time, they represent the final stages of deprotonation of an enammonium ion and of a ketone. The NH …︁ N bridge in 4 is in agreement with experimental observations and with calculations by Müller, indicating that enamines are kinetically protonated on the N‐atom. From the pinacolone enolate structures which are now known, (a) possible transition states of deprotonation of carbonyl compounds by Li‐amide bases (Figs. 7 and 8), (b) the hypothetical structure of a (i‐Pr)2NH/ethyl tert‐butyl ketone (E)‐enolate (Fig. 8b and Fig. in Footnote 18), c) the approach of the two trigonal centers in an aldol‐addition reaction, and the d) trajectory of electrophile attack on a double bond (Fig. 9) are deduced using simple molecular modeling without energy minimization. A comparison with and a discussion of the experimental and theoretical investigations about these same processes by Ireland, Corey, Dauben, Williard, and Houk are given. Finally, we believe that the H‐bonding discovered in the crystal structures (Z)‐4·TriMEDA and 6·TriMEDA is relevant to what Cram once called the ‘conducted‐tour’ mechanism of base‐catalyzed racemizations.