2-Trimethylsilylamidopyridine complexes of uranium(IV)

Abstract Reaction of 2-trimethylsilylaminopyridine with n-butyl lithium or potassium hydride affords the alkali metal ligand transfer reagents [{Li(C5H4N-2-NSiMe3)(THF)}2] (1) and [K(C5H4N-2-NSiMe3)(THF)0.25] (2), respectively. The solid state structure of complex 1 was determined revealing a dimeric structure in the solid state constructed around a centrosymmetric trans-μ-amide-lithium Li2N2 core. The synthetic utility of 1 and 2 was demonstrated by their reactions with UCl4 and UI4(OEt2)2 to give the corresponding uranium(IV) complexes [U(Cl)(C5H4N-2-NSiMe3)3] (3) and [U(I)(C5H4N-2-NSiMe3)3] (4), respectively. Crystallographic analyses of 3 and 4 revealed heteroleptic monomeric complexes where all three trimethylsilyl groups are ‘up’ with respect to the halide co-ligand. The three 2-trimethylsilylamidopyridine ligands in 3 and 4 are arranged in a twist-propeller orientation around each uranium centre giving approximate C3 symmetry down the uranium-halide bond vector but crystals of 3 and 4 are racemic. Attempts to reduce 4 to give the hypothetical uranium(III) complex [U(C5H4N-2-NSiMe3)3] were unsuccessful and instead the only product that could be isolated from a variety of reactions was the homoleptic uranium(IV) complex [U(C5H4N-2-NSiMe3)4] (5). Complex 5 exhibits fluxionality in solution which was studied by variable-temperature 1H NMR spectroscopy revealing decoalescence at low temperature which is consistent with the presence of a structure in solution that is analogous to the solid state structure. Complexes 1–5 have been characterised by NMR and FTIR spectroscopy, Evans method magnetic moment determinations, CHN microanalyses, and X-ray crystallography for 1 and 3–5.

Here, we report the synthesis of two new alkali metal 2-amidopyridine ligand transfer reagents and demonstrate their utility in the preparation of two heteroleptic uranium(IV) L 3 UX complexes. Attempts to reduce these L 3 UX complexes to homoleptic uranium(III) L 3 U complexes were, however, unsuccessful and only the homoleptic uranium(IV) L 4 U complex could be isolated from reaction mixtures. We selected the 2-trimethylsilylamidopyridine ligand because of its close relationship to the exemplar bis(trimethylsilyl)amide ligand that has found extensive use in f-element chemistry [8].

Alkali metal ligand transfer reagents
Prior work by some of us [3][4][5][6][7] resulted in the synthesis of alkali metal complexes of 2-trimethylsilylamidopyridine that incorporated crown ethers. The object of that study was to elucidate structural changes brought about in the solid state by variation of the crown or alkali metal. However, we anticipated that crown ethers would complicate the preparation of uranium derivatives. Therefore, we investigated the preparation of alkali metal complexes of 2-trimethylsilylamidopyridine in the presence of the co-ligand THF since this is a typical solvent for uranium complexes. Accordingly, we treated 2-trimethylsilylaminopyridine, prepared as 0020-1693/$ -see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2011.09.005 described previously [3], with one equivalent of n-butyl lithium in THF. Volatiles were removed and the resulting solid was washed with hexanes to afford an analytically pure off-white solid in 80% yield formulated as [{Li(C 5 H 4 N-2-NSiMe 3 )(THF)} 2 ] (1), Scheme 1. The 1 H, 13 C, 7 Li, and 29 Si NMR spectra of 1 confirm the 1:1 2-trimethylsilylamidopyridine:THF ratio and are as expected but are not particularly informative as to the nature of 1. Therefore, we conducted an X-ray diffraction study on crystals obtained from a concentrated solution in hexane.
The molecular structure of 1 is illustrated in Fig. 1 and selected bond lengths are listed in Table 1. Complex 1 adopts a dimeric structure in the solid state, constructed around a classic trans-Li 2 N 2 four-membered ring involving the amide nitrogens. Each lithium atom adopts an approximately tetrahedral geometry, which is distorted principally by the tight bite angle of the 2amidopyridine ligand [66.95(17)°]. A molecule of THF completes the coordination sphere of lithium. This dimeric fragment can be regarded as a fragment of an infinite lithium amide ladder, following well established structural building principles [48], which has been intercepted by the THF and intramolecular coordination of the pyridyl group. Two molecules of 1 crystallise in the crystallographic asymmetric unit but their metrical parameters are very similar so we focus our discussion on one of the molecules only. The Li(1)-N(1) and Li(1)-N(2), and Li(1)-N(2A) bond lengths are 2.053(6), 2.114(6) and 2.089(6) Å, respectively, and these compare well to previous Li-N bond distances in related systems [49]. The Li(1)-O(1) bond length of 1.940(6) Å is unremarkable.
Although lithium amides are excellent ligand transfer reagents, reactions with electropositive metals such as f-block elements often result in the formation of 'ate' salt occlusion complexes because of the small size of lithium [50]. However, the corresponding potassium salts rarely give occlusion complexes due to the large radius of potassium and have the added benefit of being more reactive effecting straightforward ligand transfer [51]. We therefore investigated the synthesis of the potassium congener of 1. Following a similar procedure to the synthesis of 1, but substituting n-butyl lithium with potassium hydride, afforded an off-white powder in 97% yield. NMR spectroscopy and CHN microanalyses suggest that this complex is best formulated as [K(C 5 H 4 N-2-NSiMe 3 )(THF) 0.25 ] (2). Given the large size of potassium it is likely that the structure of 2 involves a complex aggregate [52]. However, crystalline material suitable for X-ray diffraction has not been obtained so the structure of 2 remains unclear. However, 2 is analytically pure, rendering it suitable for use as a ligand transfer reagent.

Uranium derivatives
Since one objective of this work was to isolate homoleptic uranium(III) L 3 U complexes we investigated the synthesis of the corresponding heteroleptic L 3 UX precursor complexes. Treatment of UCl 4 with three equivalents of 1 in cold (À78°C) THF afforded, after filtration and work-up, green crystals of the heteroleptic uranium(IV) complex formulated as [U(Cl)(C 5 H 4 N-2-NSiMe 3 ) 3 ] (3) in 63% isolated yield, Scheme 2. The 1 H NMR spectrum of 3 spans the relatively narrow range of +12 to 0 ppm, which suggests a relatively symmetrical environment at uranium in 3. The magnetic moment of 3 in benzene solution at 298 K was found to be 2.54 l B which is consistent with the uranium(IV) formulation of 3 since molecular uranium(IV) complexes tend to exhibit magnetic
1 moments in the range 2.5-3.0 l B per uranium centre [53]. To confirm the structure of 3 we determined the crystal structure of crystals grown from a saturated solution in hexane. Complex 3 is mononuclear in the solid state and the structure is shown in Fig. 2 with selected bond lengths given in Table 1. The uranium centre adopts an irregular 7-coordinate geometry which is constrained by the acute bite angles of the three chelating 2-amidopyridine ligands [55.74(11)°av.]. The three amide centres are coplanar with respect to the uranium centre, and each other, and the sum of the three N amide -U-N amide angles is 359.89 (12) (1) bond length of 2.6494(9) Å is in the range of previously reported uranium-chloride distances [49]. The three 2-amidopyridine ligands arrange around a given uranium centre in a twist-propeller orientation. This results in an approximate C 3 rotation axis aligned along the uranium-chloride vector and is reminiscent of tris(amidinate)uranium halide complexes [9,14]. This imparts chirality to individual uranium centres, but crystals of 3 are racemic overall since there is no chiral induction in the formation of 3. Furthermore, this twist-propeller arrangement results in all the trimethylsilyl groups pointing 'up' with respect to the chloride co-ligand.
Having established that the heteroleptic complex 3 could be isolated, we targeted the iodide congener since we anticipated this to be a better reagent for reduction chemistry. Analogously to the preparation of 3, we repeated the synthesis but substituted UCl 4 with UI 4 (OEt 2 ) 2 and 2 for 1. Following filtration and work-up a brown solid was isolated from which brown [U(I)(C 5 H 4 N-2-NSiMe 3 ) 3 ] (4) was obtained in 82% crystalline yield, Scheme 3. The 1 H NMR spectrum of 4 spans the range +19 to À17 ppm, which is over twice the range of 3. The magnetic moment of 4 in benzene solution at 298 K was found to be 2.80 l B , and although this magnetic moment is higher than 3 it falls well within the range of reported magnetic moments for uranium(IV) complexes [53]. It is germane to note that we have previously observed magnetic moments for triamidoamine uranium(IV) complexes with soft pseudo-halide ligands that are lower than the analogous chloride complexes [26], but a thorough understanding of the factors which determine uranium orbital magnetism is still an ongoing challenge.
We determined the structure of complex 4 and this is depicted in Fig. 3 and selected bond lengths are tabulated in Table 1. Complex 4 is essentially isostructural to 3 except for the obvious replacement of chloride by iodide. As in 3, the three amide atoms in 4 are co-planar with respect to uranium and each other [av. bite angle = 55.76 (16) With the synthesis of 4 accomplished, we investigated its reduction in an attempt to prepare a homoleptic uranium(III) [U(C 5 H 4 N-2-NSiMe 3 ) 3 ] derivative, Scheme 4. We investigated the reduction of 4 with potassium graphite, potassium naphthalenide, or potassium mirror in THF and toluene but in all cases yellow crystals, indicative of uranium(IV), were isolated from hexane in an average yield of 32%. Since potassium iodide was eliminated, and purple solutions were observed, we reasoned that reduction to uranium(III) was proceeding, but subsequent disproportionation and ligand redistribution were occurring during work-up. In an attempt to prevent disproportionation and ligand scrambling we conducted reduction reactions in 1,2-dimethoxyethane reasoning that this may block coordination sites that would enable dimerisation and the aforementioned disproportionation/ligand redistribution to occur. However, this method also generated yellow crystals characteristic of uranium(IV), which is supported by the observed solution magnetic moment of 2.87 l B . Given that the closely related uranium(III) com- [54] may be routinely prepared and isolated we posit that the pyridyl group destabilises the trivalent state of uranium. This destabilisation promotes disproportionation and ligand redistribution. Thus, the tetravalent state of uranium is stabilised even though the pyridyl groups may in principle saturate the coordination sphere of uranium and suppress  disproportionation and ligand scrambling. Although pyridine can be considered a 'softer' donor than 'hard' alkoxides it is a strong donor ligand and it would appear that its strong donor nature is sufficient to push the U(III)/(IV) redox couple of uranium, which is already favourable, in these 2-amidopyridine systems past the cusp of stability. This presumably outweighs any stabilisation effects gained by saturating the coordination sphere of uranium such that the trivalent system is destabilised with respect to the tetravalent state.
The yellow crystals isolated from the reduction reactions were all determined to be the homoleptic uranium(IV) complex [U(C 5 H 4 N-2-NSiMe 3 ) 4 ] (5), Scheme 5, and the structure is illustrated in Fig. 4 and selected bond lengths are given in Table 1. The uranium centre in 5 adopts an irregular 8-coordinate geometry and the four 2-amidopyridine ligands are geared such that the trimethylsilyl groups mesh efficiently. The U-N amide and U-N amine distances average 2.429(12) and 2.514(12) Å, respectively, which are substantially longer than observed in 3 and 4, reflecting the higher steric congestion and higher coordination number at uranium in 5. This is reflected in the more narrow bite angle of the 2-amidopyridine ligands in 5 [av. 54.68(4)°] compared to 3 and 4.
At room temperature the 1 H NMR spectrum of 5, Fig. 5, exhibits one major broad resonance (fwhm $ 2700 Hz) along with two minor resonances but apart from a small quantity of sharp diamagnetic impurity resonances it is otherwise apparently featureless which suggests one or more fluxional processes are operating in solution for 5. Upon cooling a sample of 5 in toluene-d 8 , the major and minor resonances collapse into two broad features centred at À5 and À20 ppm. At 263-253 K four broad features in the range +20 to À15 ppm are evident and as the temperature is lowered further these resonances become sharper and other resonances become apparent until at the low temperature limit of 213 K a spectrum exhibiting four trimethylsilyl resonances and 16 pyridyl C-H resonances is observed. This suggests that at low temperature a structure analogous to the solid state structure is present in solution, where the interlocked nature of the trimethylsilyl groups renders the ligand environments magnetically inequivalent, whereas at higher temperatures this species undergoes fluxional exchange behaviour which renders all four of the 2-amidopyridyl ligands equivalent on the NMR timescale. The overlapping nature of the resonances results in this process being somewhat underdetermined, which prevents precise thermodynamic parameters from being extracted, but line-shape analysis yields an estimated DG à value of $50 kJ mol À1 in the coalescence regime.

Summary and conclusions
To conclude, we have prepared two new 2-trimethylsilylamidopyridine alkali metal ligand transfer reagents and demonstrated their utility in the synthesis of two heteroleptic uranium(IV) L 3 UX (X = Cl, I) complexes which are stable. Attempts to isolate the hypothetical homoleptic uranium(III) L 3 U complex under a variety of conditions resulted in disproportionation and ligand redistribution affording the corresponding homoleptic uranium(IV) L 4 U complex. This instability may be rationalised on the basis that the strong donor pyridyl group destabilises the trivalent state of uranium and    promotes disproportionation and ligand redistribution to access the tetravalent state of uranium. On the one hand it was anticipated that the pyridyl donors would saturate the coordination sphere of uranium suppressing disproportionation and ligand scrambling, but this is apparently outweighed by the destabilising effect that this strongly donating group has on the U(III)/(IV) redox couple which already favours the tetravalent state.

General
All manipulations were carried out using standard Schlenk techniques, or an MBraun UniLab glovebox, under an atmosphere of dry nitrogen. Solvents were dried by passage through activated alumina towers and degassed before use. All solvents were stored over potassium mirrors, with the exception of ethers which were stored over activated 4 Å molecular sieves. Deuterated solvents were distilled from potassium, degassed by three freeze-pumpthaw cycles and stored under nitrogen. The compounds C 5 H 4 N-2-N(H)SiMe 3 [3], UCl 4 [55], and UI 4 (OEt 2 ) 2 [56] were prepared according to published procedures. 1 H NMR spectra were recorded on a Bruker 400 spectrometer operating at 400.2 MHz; chemical shifts are quoted in ppm and are relative to TMS. FTIR spectra were recorded on a Bruker Tensor 27 spectrometer. Elemental microanalyses were carried out by Tong Liu at the University of Nottingham.

Synthesis of [U(Cl)(C 5 H 4 N-2-NSiMe 3 ) 3 ] (3)
THF (40 ml) was added slowly to a stirring mixture of 1 (1.47 g, 3.00 mmol) and UCl 4 (0.38 g, 1.00 mmol) at À78°C. The mixture was allowed to warm to room temperature with stirring over 16 h. Volatiles were removed in vacuo and the product was extracted into hot toluene (30 ml). The mixture was filtered to remove the LiCl

Synthesis of [U(I)(C 5 H 4 N-2-NSiMe 3 ) 3 ] (4)
THF (40 ml) was added slowly to a stirring mixture of 2 (4.00 g, 18.00 mmol) and UI 4 (OEt 2 ) 2 (5.36 g, 6.00 mmol) at -78°C. The mixture was allowed to warm to room temperature with stirring over 16 h. Volatiles were removed in vacuo and the product was extracted into hot toluene (30 ml). The mixture was filtered to remove the LiCl precipitate and volatiles were removed in vacuo to yield 4 as a brown solid. Brown crystals were grown from a saturated solution of hexanes. Yield Method A: THF (20 ml) was added slowly to a stirring mixture of 4 (0.86 g, 1.00 mmol) and KC 8 (0.18 g, 1.30 mmol) at À78°C. The mixture was allowed to warm to room temperature with stirring over 16 h. The solution was filtered to remove the KI and C 8 precipitate. Volatiles were removed in vacuo and the product was extracted into hexanes. Yellow crystals were grown from a saturated solution of hexanes. Method B: Toluene (20 ml) was added slowly to a stirring mixture of 4 (0.86 g, 1.00 mmol) and KC 8 (0.18 g, 1.30 mmol) at À78°C. The mixture was allowed to warm to room temperature with stirring over 16 h. The solution was filtered to remove the KI and C 8 precipitate. Volatiles were removed in vacuo and the product extracted into hexanes. Yellow crystals were grown from a saturated solution of hexanes. Method C: DME (20 ml) was added slowly to a stirring mixture of 4 (0.86 g, 1.00 mmol) and KC 8 (0.18 g, 1.30 mmol) at À78°C. The mixture was allowed to warm to room temperature with stirring over 16 h. The solution was filtered to remove the KI and C 8 precipitate. Volatiles were removed in vacuo and the product extracted into hexanes. Yellow crystals were grown from a saturated solution of hexanes. Method D: A solution of potassium naphthalenide (0.17 g, 1.00 mmol) in THF (10 ml) was added to a stirring solution of 4 (0.86 g, 1.00 mmol) in THF (10 ml) at À78°C. The mixture was allowed to warm to room temperature with stirring over 16 h. The volatiles were removed in vacuo and the product was extracted into hexanes (20 ml), and filtered to remove the KI precipitate. Yellow crystals were grown from a saturated solution of hexanes. Method E: A solution of 4 (0.86 g, 1.00 mmol) in toluene (20 ml) was vigorously stirred over a potassium mirror (20 fold excess) for 2 days. The mixture was filtered and volatiles removed in vacuo. Yellow crystals were grown from a saturated solution of hexanes to give 5. Average yield of 0.29 g, 32%. Anal. Calc. for C 32

X-ray crystallography
Crystal data for compounds 1 and 3-5 are given in Table 2 and bond lengths and angles are listed in Table 1. Crystals were examined variously on a Bruker APEX CCD area detector diffractometer using graphite-monochromated Mo Ka radiation (k = 0.71073 Å), or on an Oxford Diffraction SuperNova Atlas CCD diffractometer using mirror-monochromated CuKa radiation (k = 1.5418 Å). Intensities were integrated from data recorded on 0.3°(APEX) or 1°( SuperNova) frames by x rotation. Cell parameters were refined from the observed positions of all strong reflections in each data set. Semi-empirical absorption correction based on symmetryequivalent and repeat reflections (APEX) or Gaussian grid face-indexed absorption correction with a beam profile correction (Supernova) were applied. The structures were solved variously by direct and heavy atom methods and were refined by full-matrix least-squares on all unique F 2 values, with anisotropic displacement parameters for all non-hydrogen atoms, and with constrained riding hydrogen geometries; U iso (H) was set at 1.2 (1.5 for methyl groups) times U eq of the parent atom. The largest features in final difference syntheses were close to heavy atoms and were of no chemical significance. Programs were Bruker AXS SMART [57] and CrysAlisPro [58] (control), Bruker AXS SAINT [57] and CRYSALISPRO [58] (integration), and SHELXTL [59] and OLEX2 [60] were employed for structure solution and refinement and for molecular graphics.