Synthesis and crystal structures of extremely bulky phosphinoamido and phosphinoamino germanium(II) chloride complexes

The chemistry of low-oxidation-state and/or low-coordination-number main group metal compounds has rapidly expanded since the turn of the millennium (Fischer and Power, 2010). The growing interest in such compounds stems from their intrinsically high reactivity, which has afforded them applications as ‘transition metal-like’ reagents in synthesis, catalysis, etc. (Power, 2010; Asay et al., 2011). The stabilization of most low-oxidation-state main group compounds has required the employment of very bulky ligand systems, which provide kinetic stabilization to those compounds. Although most attention has focused on bidentate and higher dentate ligands for this task (Jones, 2010; Tsai, 2012), we have investigated the use of extremely bulky monodentate amide ligands, e.g. -N(Ar)(SiR₃), (Ar=C₆H₂{C(H)Ph₂}₂R′-2,6,4; R′=Me (Ar*) or Prⁱ (Ar^†); R=Me, Prⁱ, Ph), in this area over the past several years (Li et al., 2011a; Wong et al., 2013). The rationale here is that such ligands have the steric bulk to stabilize very-low-oxidation-state main group systems, while leaving the main group center coordinatively unsaturated and therefore open to reaction with a variety of unsaturated substrates. This has proved to be the case, and a variety of unprecedented main group complex types have been accessed. From group 14, these include the first examples of monomeric amido-germanium halides, e.g. [(SiMe₃)(Ar*)NGeCl] (Li et al., 2011a), amido-digermynes, e.g. [{(SiPrⁱ₃)(Ar^†)NGe}₂] (Li et al., 2011b; Hadlington et al., 2013), and two-coordinate germanium hydrides, e.g. [{Si(OBu^t)₃}(Ar^†)NGeH] (Hadlington et al., 2015). Moreover, examples of the latter complex type have proven to act as efficient catalysts for the hydroboration of aldehydes and ketones, exhibiting activities that rival those of most transition metal-based catalysts used for such transformations (Hadlington et al., 2014).

We wished to extend the library of extremely bulky amide ligands at our disposal to include examples that have P-functionalities α to the amido center. The idea here was that these amides could act as hemi-labile ligands toward coordinated main group metal centers, thus increasing the stability of the metal complex without reducing its reactivity significantly. To this end, we have recently developed a range of the target ligands, e.g. L^† (=[Ar^†(PPh₂)N]^-), which we have shown can act as monodentate κ¹-N-donor ligands toward low-oxidation-state silicon centers, while also exhibiting weak intramolecular P···Si interactions (Böttcher and Jones, 2015). We wished to extend the range of such complexes to those derived from germanium, e.g. [L^†GeCl], which could be used as precursors to amido-digermyne and/or amido-germanium hydride complexes. Although attempts to prepare [L^†GeCl] were not successful, they did lead to the low-yield formation of the related complexes, [GeCl₂(κ¹-P-L^†GeCl)] and [GeCl₂(κ¹-P-L^†H)], the X-ray crystal structures of which are reported herein.

Reaction of one equivalent of the potassium amide complex, KL^†, with GeCl₂·dioxane in THF did not give [L^†GeCl], but instead led to a mixture of products that included the secondary amine L^†H as the major component. Workup of the reaction mixture led to isolation of a few yellow crystals of the toluene soluble amido-germanium chloride complex, [GeCl₂(κ¹-P-L^†GeCl)] 1, and the hexane soluble amino-germanium chloride, [GeCl₂(κ¹-P-L^†H)] 2 (Scheme 1). It is not known what causes the generation of L^†H in this reaction, but it is reproducible and is not due to the presence of adventitious water. It is possible that compound 1 is formed via reaction of initially generated [L^†GeCl] with GeCl₂·dioxane, whereas compound 2 could result from reaction of in situ formed L^†H with GeCl₂·dioxane. With that said, attempts to obtain a higher isolated yield of 1 by reacting KL^† with two equivalents of GeCl₂·dioxane were unsuccessful. Similarly, reacting L^†H directly with GeCl₂·dioxane afforded a mixture of products containing only a small amount of 2, as determined by a ³¹P NMR spectroscopic analysis of the reaction solution.

Scheme 1:

Preparation of compounds 1 and 2 (Ar^†=C₆H₂{C(H)Ph₂}₂Prⁱ-2,6,4).

Given the extremely low isolated yields of 1 and 2, very few spectroscopic data could be obtained for the complexes. However, by monitoring both the 1:1 and 1:2 reactions by ³¹P NMR spectroscopy (C₆D₆ solutions), broad signals at δ 45.2 and 22.3 ppm can be tentatively assigned to 1 and 2, respectively, as the lower field signal predominates in the 1:2 reaction. The only other signal observed in these spectra is due to the secondary amine, L^†H (δ=38.2 ppm), and this confirms the amine is the major component in both reaction mixtures.

Both 1 and 2 were crystallographically characterized, and their molecular structures are depicted in Figures 1 and 2 (see also Table 1). Compound 1 is monomeric and can be considered as an adduct between the tertiary phosphine center of the germanium amide, L^†GeCl, and the GeCl₂ fragment. The geometric parameters within the L^†GeCl fragment of the molecule are similar to those in [(SiPrⁱ₃)(Ar^†)NGeCl] (Hadlington et al., 2013) and are suggestive of a high s-character lone pair of electrons at its germanium center [Σ angles at Ge(2)=280.7°]. The dative P→Ge bond in the compound, 2.5725(8) Å, is slightly longer than those in the handful of other crystallographically characterized phosphine adducts of GeCl₂, e.g. 2.5087(7) Å in [GeCl₂(PCy₃)] (Swarnakar et al., 2014), presumably due to the imposing steric bulk of the L^†GeCl fragment. The intramolecular Cl(1)···Ge(2) distance (3.084(1) Å) is well within the sum of the van der Waals radii for the two elements [3.86 Å (Mantina et al., 2009), cf. sum of covalent radii=2.20 Å (Pyykkö and Atsumi, 2009)], and therefore, there is a possibility of an interaction between the atoms (cf. 4-coordinate adducts of GeCl₂, e.g. [(κ²-N,N′-tmeda)GeCl₂] (Cheng et al., 2010). However, it is also possible that Cl(1) is directed toward Ge(2) for steric reasons.

Figure 1:

Thermal ellipsoid plot (30% probability surface) of the molecular structure of 1 (hydrogen atoms omitted).

Selected bond lengths (Å) and angles (°): Ge(1)-N(1) 1.903(2), Ge(1)-Cl(1) 2.2405(8), P(1)-N(1) 1.706(2), P(1)-Ge(2) 2.5725(8), Ge(2)-Cl(3) 2.2785(9), Ge(2)-Cl(2) 2.3038(9), Cl(1)···Ge(2) 3.084(1), N(1)-Ge(1)-Cl(1) 96.55(6), Cl(3)-Ge(2)-Cl(2) 97.12(3), Cl(3)-Ge(2)-P(1) 91.06(3), Cl(2)-Ge(2)-P(1) 92.49(2), Ge(1)-Cl(1)···Ge(2) 108.6(1).

Figure 2:

Thermal ellipsoid plot (30% probability surface) of the molecular structure of 2 [hydrogen atoms, except H(1), omitted].

Selected bond lengths (Å) and angles (°): Ge(1)-Cl(1) 2.2451(7), Ge(1)-Cl(2) 2.2734(8), Ge(1)-P(1) 2.5435(8), P(1)-N(1) 1.670(2), Cl(1)-Ge(1)-Cl(2) 96.32(3), Cl(1)-Ge(1)-P(1) 96.87(2), Cl(2)-Ge(1)-P(1) 90.00(3), N(1)-P(1)-Ge(1) 110.74(7).

In the solid state, compound 2 is monomeric and can be considered as an adduct between the P-center of the protonated ligand, L^†H, and the GeCl₂ fragment. Interestingly, the NH moiety lies on the opposite side of the molecule to the GeCl₂ fragment, whereas both the NGeCl and GeCl₂ units in 1 lie on the same side of that molecule. This might add weight to the proposal of a weak intramolecular Cl···Ge interaction in the latter. The Ge center in 2 (Σ angles=283.2°) exhibits a similar degree of pyramidalization to Ge(2) in 1, although its P-Ge distance (2.5435(8) Å) is slightly shorter, presumably because of the lesser steric strain in 2.

In summary, complexes of extremely bulky phosphinoamido and phosphinoamino ligands with germanium(II) chloride fragments have been prepared and structurally characterized. The complexes represent rare examples of phosphine adducts of GeCl₂, and one possibly exhibits an intramolecular Cl···Ge interaction in the solid state. We continue to develop the application of bulky amides to the stabilization of low oxidation state/low coordination number main group compounds and will report on our efforts in this direction in due course.