Si–H addition followed by C–H bond activation induced by a terminal thorium imidometallocene: a combined experimental and computational study†

Department of Chemistry, Beijing Normal U gzi@bnu.edu.cn; dcfang@bnu.edu.cn; Fax: College of Chemistry and Chemical Engin 400715, China Institut für Anorganische und Analyti Braunschweig, Hagenring 30, 38106 Brau tu-bs.de; Fax: +49-531-3915387; Tel: +49-53 † Electronic supplementary information ( 1, cartesian coordinates of all stationary p level and kinetic study. CCDC 977309–977 in CIF or other electronic format see DOI Cite this: Chem. Sci., 2014, 5, 3165


Introduction
Terminal imido complexes of actinide-metals containing an An]N functionality have received widespread attention over the last 20 years because of their unique structural properties and their potential application in group transfer reactions and catalysis. 1-3 Whereas many uranium imido complexes have been prepared and structurally characterized, only a few of them show signicant reactivity. 2h-k,s,t,x,3f,g,j,l,o,y In contrast, thorium imido complexes have remained rare and not much is known about their reaction chemistry. 2d,3h,k,q This is surprising since thorium has a 7s 2 6d 2 ground state electron conguration, and one might expect a similar reactivity to that of group 3 and 4 metals, such as Sc, Ti, Zr and Hf, for which several complexes with M]N bond have been prepared. 4,5 However, the underlying question remains whether 5f-orbitals contribute to the bonding in thorium organometallics and whether Th 4+ should be considered as a transition metal or as an actinide element. 6 To answer this question we have recently prepared the base-free terminal thorium imido complex [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th] N(p-tolyl) (1). 7,8 Complex 1 shows a rich reaction chemistry such as the activation of elemental sulfur (S 8 ). 7 Furthermore it is an important intermediate in the catalytic hydroamination of internal acetylenes, 7 an efficient catalyst for the trimerization of PhCN, 7 and a useful precursor for the preparation of oxido and suldo thorium metallocenes [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th]E (E ¼ O, S) by cycloaddition-elimination reactions with Ph 2 C]E (E ¼ O, S) or CS 2 . 9 Encouraged by this broad reactivity, we are now focusing on small molecule activation. Herein, the rst Si-H bond activations by an actinide imido complex and the reactivity of the resulting thorium amido hydrido complexes are reported. Furthermore, the difference between thorium and early transition metal imido complexes will be addressed.

General methods
All reactions and product manipulations were carried out under an atmosphere of dry dinitrogen with rigid exclusion of air and moisture using standard Schlenk or cannula techniques, or in a glove box. All organic solvents were freshly distilled from sodium benzophenone ketyl immediately prior to use. PhSiH 3 and Ph 2 SiH 2 were freshly distilled from CaH 2 immediately prior to use. [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th]N(p-tolyl) (1) was prepared according to literature methods. 7 All other chemicals were purchased from Aldrich Chemical Co. and Beijing Chemical Co. and used as received unless otherwise noted. Infrared spectra were obtained from KBr pellets on an Avatar 360 Fourier transform spectrometer. 1 H and 13 C NMR spectra were recorded on a Bruker AV 400 spectrometer at 400 and 100 MHz, respectively. All chemical shis are reported in d units with reference to the residual protons of the deuterated solvents, which are internal standards, for proton and carbon chemical shis. Melting points were measured on an X-6 melting point apparatus and were uncorrected. Elemental analyses were performed on a Vario EL elemental analyzer.
Method B NMR scale. To a J. Young NMR tube charged with [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th]N(p-tolyl) (1; 16 mg, 0.02 mmol) and C 6 D 6 (0.5 mL), an excess of Ph 2 SiH 2 was added. Complete conversion to 3a was observed by 1 H NMR spectroscopy. Preparation SiHPh 2 ] (3a; 247 mg, 0.25 mmol) was stirred at 70 C for three days, the solvent was removed. The residue was extracted with n-hexane (10 mL Â 2) and ltered. The volume of the ltrate was reduced to ca.  . Aer this solution was stirred at room temperature for 2 hours, the solvent was removed. The residue was extracted with n-hexane (10 mL Â 2) and ltered. The volume of the ltrate was reduced to ca.

X-ray crystallography
Single-crystal X-ray diffraction measurements were carried out on a Bruker SMART CCD diffractometer using graphite monochromated Mo Ka radiation (l ¼ 0.71073Å). An empirical absorption correction was applied using the SADABS program. 10 All structures were solved by direct methods and rened by fullmatrix least squares on F 2 using the SHELXL-97 program package. 11 The hydride atom in 2a was located from a difference-Fourier map and rened isotropically. Hydrogen atoms were geometrically xed using the riding model. Disordered solvents in the voids of 2a and 6 were modeled or removed by using the SQUEEZE program. 12 Crystallographic details for 2a, 4 and 6 are summarized in Table 1.

Computational methods
All calculations were carried out with the Gaussian 09 program (G09), 13 employing the B3PW91 method, plus polarizable continuum model (PCM) and D3 (ref. 14) (denoted as B3PW91-PCM + D3), with standard 6-31G(d) basis set for C, H, N and Si elements and Stuttgart RLC ECP from EMSL basis set exchange (https://bse.pnl.gov/bse/portal) for Th element, 15 to fully optimize the geometries of reactants, complexes, transition state, intermediates, and product structures, and to mimic experimental toluene-solvent conditions (dielectricity constant 3 ¼ 2.379). All stationary points were subsequently characterized by vibrational analyses, from which their respective zero-point (vibrational) energy (ZPE) were extracted and used in the relative energy determinations; in addition to ensure that the reactant, complex, intermediate, product and transition state structures resided at minima and 1st order saddle points, respectively, on their potential energy hyper surfaces. (3a), respectively (Scheme 1). Consistent with a Si-H bond addition across the Th]N bond the syn-isomer is formed exclusively. However, in contrast to scandium 4f and titanium 5k,l imido complexes, the reaction of 1 with silanes is irreversible, which is presumably a consequence of the more polarized actinide imido bond (see ESI †). Previous studies have clearly established that the 5f orbitals play a key role in the bonding of actinide complexes, which results in a very polarized An]E bond, whereas this is not the case for group 4 metal complexes. 6c Therefore, actinide An]E bonds exhibit different reactivity patterns compared to those of early transition metals, as illustrated by reactivity of (4). The average Th-C (ring) distance is 2.882(6)Å, and the Cp (cent)-Th-Cp (cent) angle is 140.6(2) . The Th-N distance of 2.310(5)Å is shorter than that (2.387(2)Å found in 2a. Furthermore, the Th-C(40) distance of 2.420(3)Å is also signicantly shorter than the reported Th-C aryl bond distances (2.548(2)-2.654(14)Å). [22][23][24] Interestingly, only four structurally characterized thorium aryl complexes have been reported, and in all of them the thoriumaryl bond is supported by chelating ligands. [22][23][24] As demonstrated above, complex 1 irreversibly activates Si-H bonds to yield the amido hydrido complexes 2 and 3. Complex 3 undergoes a reversible intramolecular C-H bond activation to give 4 and H 2 . DFT calculations were performed at the B3PW91 level of theory to further understand the observed process. To properly describe the observed reactivity it was necessary to include solvent and dispersion effects (see ESI †). The reaction of 1 with Ph 2 SiH 2 to 3a is exergonic with DG (343 K) ¼ À17.6 kcal mol À1 and proceeds concerted via the 4-membered transition state TS formed by the Th]N and Si-H moieties. The Si-N and Th-H bonds in TS are 2.343 and 2.299Å, respectively, and ca. 0.61 and 0.23Å longer than those in the product 3a (see ESI †). The barrier for this reaction is DG ‡ ¼ 16.7 kcal mol À1 (343 K) (15.2 kcal mol À1 at 298 K) (Fig. 3); and these values are consistent with the rapid and irreversible formation of 3a at ambient temperature. The transformation of 3a to 4 + H 2 is more complicated and involves the anti-intermediate 3b and two transition states (TSa and TSb). The barrier for the conversion of 3a to 3b via TSa is 33.7 kcal mol À1 at 343 K and also represents the rate-limiting step in this reaction. The computed barrier is consistent with the experimental observation that this reaction only occurs at elevated temperatures, and it is also in reasonable agreement with the experimentally estimated barrier of 26.8 kcal mol À1 (see ESI †). The deviation between the computed and experimental values is mainly attributed to the low solubility of H 2 in organic solvent, which complicates the kinetic evaluation because the actual H 2 concentration cannot be determined with certainty (see ESI †). However, for TSa a closer inspection reveals an interesting feature, since the syn-/anti-intermediates are interconverted not by rotation of the -N(R)(p-tolyl) moiety along the Th-N axis, but by a Th-H movement (Fig. 3). This is a direct consequence of the severe steric bulk of the two 1,2,4-(Me 3 C) 3 C 5 H 2 ligands, which hinders the rotation along the Th-N axis. Nevertheless, once complex 3b is formed, H 2 is released via TSb and with a low barrier of 12.6 kcal mol À1 (Fig. 3). This makes the anti-isomer 3b difficult to isolate, since H 2 release from the reaction mixture shis the equilibrium to the nal product 4. On the contrary, H 2 addition to complex 4 forms a mixture of 3b and 3a.    Furthermore, similar to the amido hydrido scandium complex, 4f amido hydrido thorium complexes can also insert unsaturated substrates such as carbodiimides into the Th-H bond. For example, treatment of 2a with 2 equiv. of N,N 0 -dicyclohexylcarbodiimide (DCC) rapidly forms the metallacycle [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th[N(p-tolyl)C(]NC 6 H 11 )-N(C 6 H 11 )] (6) and azomethine (C 6 H 11 )N]CHN(C 6 H 11 )(SiH 2 Ph) (5) in quantitative conversion (Scheme 1). Furthermore, analogous to group 4 imido complexes, 25 thorium imido 1 can react with DCC to yield metallacycle 6 (Scheme 1), whereas 2a reacts with 1 equiv. of DCC to give 6 in only 50% conversion. These observations suggest that the insertion product [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th [N(p-tolyl)SiHPh 2 ][h 3 -N(C 6 H 11 )CHN(C 6 H 11 )] is unstable and eliminates azomethine 5 to form the imido complex 1, which then reacts with DCC to the metallacycle 6 by an [2 + 2] cycloaddition reaction (Scheme 1).

Conclusions
In conclusion, the rst example of a Si-H bond activation by a terminal actinide imido complex has been comprehensively studied. In contrast to scandium 4f and titanium 5k,l imido complexes, silanes such as PhSiH 3 and Ph 2 SiH 2 add irreversibly to the thorium imido [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th]N(p-tolyl) (1), supporting the notion that Th 4+ behaves more like an actinide than a transition metal. 6c DFT studies reveal that the 1,2-addition proceeds in a concerted, 4-membered transition state to give amido hydrido complexes 2 and 3. These compounds are reactive species, as illustrated by the formation of H 2 and [h 5 -1,2,4-(Me 3 C) 3 C 5 H 2 ] 2 Th[h 2 -N,C-{N(p-MeC 6 H 3 )(SiHPh 2 )}] (4) via an intramolecular aromatic C-H bond activation in 3, and by the insertion of unsaturated substrates such as DCC into the Th-H bond of 2. The development of new actinide imido complexes and the exploration of thorium amido hydrido complexes in catalysis are ongoing projects in these laboratories.