LiF-ThF

Other indications for the generation of phosphasilenes involved compounds 520 and 6.21 Both derivatives were generated by thermally induced elimination of LiF/THF from corresponding lithium(fluorosilyl)phospha-nylides. Whereas 5 was directly observed by means of 3,P-NMR spectroscopy, surprisingly, the sterically more hindered phosphasilene 6 was not detected due to its lower thermostability. Therefore, evidence for the existence of 6 was achieved solely by trapping experiments. Due to the more crowded substitution in 5 compared with 2, the dimerization process to give 7 is relatively slow (Eq. 1) but is complete within a few hours. 

The lithium derivative of the Ph-substituted aminofluorosilane crystallizes from THF as a monomeric LiF adduct of an iminosilane, (THF)3LiF (CMe3)2Si=NSiPh(CMe3)3.2I... 

Compounds 17 were formed by the reaction of Is(R)SiF2 (R = Is, /Bu) with two equivalents of [LiPH2(DME)] (DME = 1,2-dimethoxyethane). They do not show any tendency to eliminate LiF in THF solutions, surely because of the strong Si—F bond. Similar stability was observed for the nitrogen analogs.13... 

Based on the early racemic synthesis of 4 (cis series), it had already been demonstrated that 2-azetidinone ring closure could be achieved via nucleophilic attack of a lithium amine anion on a (3-ester. Cyclization could be accomplished with other strong bases, but sodium bistrimethylsilylamide was found to effect efficient cyclization without significant racemization at C3. During the search for experimentally convenient bases, it was noted that Noyori (Nakamura et al., 1983) reported that tetrabutylammonium fluoride (TBAF) as well as LiF, KF, and CsF could serve as the base in Aldol reactions. Treatment of 17a or 17b with TBAF trihydrate in THF did not affect cyclization. After much experimentation it was found that addition of A,0-bistrimethylacetamide (BSA) to 19 followed by TBAF addition, effected 2-azetidinone ring closure. Further optimization found that use of catalytic TBAF (< 1%) in methylene chloride afforded near quantitative cyclization. 

The silicon atom in Wiberg s silenes is a Lewis acidic center and can be coordinated with donors. These donors may be halide ions [like, e.g., in 104 LiX(T2-C-4)831, ethers or nitrogen bases84. Coordination with the donor stabilizes the silenes and, when the basicity suffices (LbOcBr, THF, NMe3, F ), the adducts serve as stock compounds from which the silenes may be liberated (equation 32). 92 decomposes at —100 °C, but its trimethylamine adduct is stable at room temperature85. Noticeable amounts of 97 are available from 97-LiBr already at — 78 °C, whereas from 97 LiF the temperature has to be raised to +30 °C. Structures of such silene adducts will be discussed in Section I.C.l. The silene adduct 124-TIIF has been discussed in Section I.A.2. 

Summary Lithiated aminofluorosilanes can often be regarded as LiF adducts of iminosilanes on account of their structures [1], They are suitable precursors to iminosilanes after fluorine-chlorine exchange followed by elimination of lithium chloride [1,2]. This synthesis shows the following features (a) the method is a general one in the sense that it can be applied to differently substituted aminosilanes (b) the iminosilanes are obtained, either as THF adducts or as free imines in good yield (c) the tedious and sometimes complicated synthesis of aminochlorosilanes as precursors is avoided, since the fluorine-chlorine exchange takes place in one step in the lithium derivative. 

After the addition of THF, LiF is eliminated. The intermediate silaamidide [13, 14] dimerizes in a [2+2]-cycloaddition to give the lithium(cyclodisilazane)diamide anion. The lithium which is part of the anion is only twofold coordinated. The Li cation is complexed by THF. 

The compounds 8 and 10 are formed by the reaction of IS2Sip2 with two equiv [LiEH2(DME)] (E = P, As DME = 1,2-dimethoxyethane) they show no tendency to eliminate LiF in THF solution, surely because of the inherent strength of the Si-F bond. Silylation/germylation of 8 furnished the compounds 9 and 11, respectively, which were subsequently lithiated on phosphorus and arsenic to yield 1 and 3, respectively. In the first instance, the nature of the substituent at phosphorus determines the structures of the precursors 1 and 3, respectively. This has been demonstrated by a study of a structure-reactivity relationship for several compounds 1 [6]. The latter investigation... 

---