[NiIII(OMe)]-mediated reductive activation of CO2 affording a Ni(κ1-OCO) complex

We report a novel pathway for the reductive activation of CO2 by the [NiIII(OMe)(P(C6H3-3-SiMe3-2-S)3)]– complex, yielding the [NiIII(κ1-OCO˙–)(P(C6H3-3-SiMe3-2-S)3)]– complex.


Introduction
Carbon dioxide, the waste from human activity embodying the nature of high thermodynamic stability and chemical inertness, is expected to be employed as an inexpensive and potential feedstock of C 1 sources for the regeneration of valuable chemicals and fuel. 1,2 Nature developed carbon monoxide dehydrogenase (CODH) to harbor a Ni-Fe cluster for the reversible interconversion between CO 2 and CO. [3][4][5] To gain insight into the mechanism for the conversion of CO 2 to CO in CODH, several Ni-CO 2 adducts derived from the reaction of a low-valent Ni complex and CO 2 were reported. [6][7][8][9] The direct electrochemical reduction of CO 2 affords oxalate, carbon monoxide, formic acid, methanol, methane, and ethylene. 10 To gain insight into the transformation of CO 2 at a molecular level, the chemistry of the activation of CO 2 via nucleophilic attack/interaction on the polarized C center, in addition to the reduction of the coordinated CO 2 ligand by low-valence transition metal complexes, has grown explosively over past years. 11 The versatile chemical transformations of CO 2 , i.e. insertion of CO 2 producing bicarbonate/acetate/formate, [12][13][14][15][16][17][18] cleavage of CO 2 yielding m-CO/m-oxo transition-metal complexes, [19][20][21][22][23] reduction of CO 2 affording CO/HCOOH/CH 3 OH/CH 4 /C 2 H 4 /C 2 H 6 /methylene, and electrocatalysis of CO 2 converting it into oxalate, were well documented. [24][25][26] The direct electrochemical reduction or electrocatalytic transformation of CO 2 for the mass production of valuable chemicals and fuel, however, relies on the consumption of sustainable electric potential energy. Here we show a novel pathway for the reductive activation of CO 2 by a mononuclear Ni(III) complex [Ni III (OMe)(P(C 6 H 3 -3-SiMe 3 -2-S) 3 )] À . 27 This [Ni III -(OMe)]mediated reduction of CO 2 yields the complex Ni III (k 1 -OCOc À ), evidenced by single-crystal X-ray diffraction, EPR, SQUID, Ni/S Kedge X-ray absorption spectroscopy, IR and Ni valence-to-core X-ray emission spectroscopy. The ionic [Ni III (OMe)] core provides a kinetic pathway to induce the binding of CO 2 and trigger the subsequent reduction of CO 2 by the nucleophilic [OMe] À in the immediate vicinity. The covalent [Ni III (SPh)] core as well as Ni(II) center in complexes [Ni II (L)(P(C 6 H 3 -3-SiMe 3 -2-S) 3 )] À (L ¼ CO or N 2 H 4 ), in contrast, are inert toward CO 2 . 28

Results and discussion
Synthesis and characterization of nickel k 1 -OCO complex When CO 2(g) was bubbled into the thermally stable [Ni III (OMe)(PS 3 27 a pronounced color change from blue green to yellow green occurred to yield the O-bound k 1 -CO 2 complex [Ni(k 1 -OCO)(PS 3 3 )] À via the classical insertion or b-H migration mechanisms (Scheme 1a). 9,[12][13][14][15][16] The accompanied formation of [cOMe] in the reaction described above was corroborated using the spin-trapping reagent DMPO (ESI Fig. S1 †). 29 The IR n OCO stretching peak at 2177 cm À1 (KBr) (n OCO : 2226 cm À1 in THF) exhibited by complex 2 supports the formation of [Ni(k 1 -OCO)(PS 3 )] À , which is consistent with the isotopic shi of the IR n OCO stretching peak to 2117 cm À1 (KBr) observed in the 13 CO 2 labeling experiment (ESI Fig. S2 †). The conversion of complex 1 to complex 2 under a CO 2 atmosphere was also monitored by UV-vis spectrometry; the intense bands at 419 and 605 nm disappeared with the simultaneous formation of absorption bands at 425 and 610 nm (THF) (ESI Fig. S3 †). The green needle crystals of complex 2 were isolated when complex 2 was recrystallized from THF-diethyl ether at room temperature. As shown in Scheme 1b, treatment of complex 2 with CO (g) led to the formation of the reported complex [Ni II (CO)(PS 3 )] À accompanied by the release of CO 2(g) characterized by IR and GC (Fig. 1). 28 To   (4)  X-ray absorption/emission spectrum A Ni and S K-edge X-ray absorption spectroscopic (XAS) study of complex 2 was further attempted to investigate its electronic structure using complexes 1 and 4 as reference complexes. As shown in Fig. 3A, the Ni K-edge XAS of complex 2 (8333.1 eV) together with analogous complexes 1 (8332.9 eV) and 4 (8332.7 eV) shows a similar Ni 1s -to-Ni 3d transition energy. Accordingly, the  Table 1, the intensityweighted average energy of the S 1s -to-Ni 3d transitions in combination with the S 1s -to-S * C-S transition energy demonstrate that the Ni 3d manifold orbital energy of complex 2 is 0.3 eV higher than those of complexes 1 and 4. 33,34 With further regard to the Ni 1s -to-Ni 3d pre-edge transition energy observed in the Ni K-edge XAS, Ni Z eff of complexes 1, 2, and 4 are all comparable. That is, the Ni and S K-edge XAS study supports the [Ni III :CO 2 c À ] electronic structure in complex 2. As shown in Fig. S6, † the cyclic voltammogram of a 2 mM solution of complex 4 in CH 3 CN indicates a reversible interconversion between Ni III /Ni II at E 1/2 ¼ À0.58 V and an irreversible oxidation at E pa ¼ À0.21 V, whereas complex 2 exhibits a reversible interconversion between Ni III /Ni II at E 1/2 ¼ À0.70 V and an irreversible oxidation at E pa ¼ À0.29 V (vs. Fc/Fc + ).

)] À (2), instead of complexes [Ni(OC(O)OCH 3 )-(PS 3 )] À or [Ni(OC(O)H)(PS
With regard to complex 4 as an isolobal equivalent to [Ni III :CO 2 ], complex 2 is a Ni III complex bearing a 17-valenceelectron [CO 2 ]c À ligand. The signicantly lower intensity of the second S 1s -to-Ni 3d transition peak observed in the S K-edge XAS spectrum of complex 2, compared to that of complex 4, discloses that the one extra electron shared by the axial Ni 3d orbital and 2p * u orbital of CO 2 leads to a strengthening of the Ni III -CO 2 c À bond and stabilizes the coordination of k 1 -[CO 2 ]c À toward the Ni III center (Fig. 3B and Table 1). As observed in complex [Ni III (L)(PS 3 )] À (L ¼ OMe, SEt, SPh), complex 4 displays an EPR silence at 300 K, an axial signal at g ¼ 2.27 and 2.04 at 77 K, and an effective magnetic moment of 1.74 m B at 300 K ( Fig. 4C  and S7A †). 27,28,35 The stabilization of the [CO 2 ]c À radical through coordination to the Ni III center in complex 2 was further evidenced by EPR spectroscopy.
As shown in Fig. 4A and B, the EPR spectrum of complex 2 at 77 K apparently resembles a combination of the typical EPR signal of [Ni III (L) (PS 3 )] À (g ¼ 2.31, 2.03, and 2.00) and the [CO 2 ]c À radical with a contribution of Ni 3d leading to the observed g anisotropy ( Fig. 4A and B). 36 The spin quantitation of complex 2, using complex 4 as a reference, demonstrates that the electronic structure of complex 2 is best described as a resonance hybrid between [Ni III :CO 2 c À ] and [Ni II :CO 2 ], which is supported by the effective magnetic moment of 1.59 m B exhibited by complex 2 at 300 K (ESI Fig. S7B and S7C †).
The experimental valence-to-core X-ray emission (V2C XES) spectra of complexes 2 and 4 are presented in Fig. 3C. In comparison with complex 4, the broad V2C transition peak of complex 2 at 8330.0 eV shis from 8328.8 eV upon replacement of the [NCO] À by the [CO 2 ]c À ligand. DFT calculation was further pursued to verify the nature of the V2C transition(s). As shown in ESI Fig. S8A and S8B, † the DFT calculated V2C XES spectra resembles the experimental V2C features and, in particular, the trend of the energy shi comparing complexes 2 and 4. The contribution of the 4s g , 3s u , and 1p g orbitals of [NCO] À and Ni 3d -S 3p orbitals results in the V2C features of complex 4. 37 For complex 2, the absence of transitions from the 3s u and 1p g orbitals and an additional transition from the occupied 2p u   Complex [Ni(L)(P(C 6 H 3 -3-SiMe 3 -2-S) 3 )] À , embedded in a distorted trigonal bipyramidal geometry, features a wealth of chemical reactivity tailored by the oxidation state of Ni and coordinating ligand L (L ¼ OPh, SPh, SePh and Cl for Ni III ; L ¼ CO, N 2 H 4 for Ni II ). 27,28,35 33,34 Despite the labile nature of CO and N 2 H 4 , the inert reactivity of the Ni II center toward CO 2 demonstrates that the lowered Ni 3d manifold orbitals in Ni III complex 1 attracts the binding of weak s-donor CO 2 and triggers the subsequent reduction of CO 2 by the nucleophilic [OMe] À in the immediate vicinity. The reactivity of complex 1 toward CO 2 , affording an O-bound [Ni III :CO 2 c À ] species, uncovers a novel strategy for the immobilization and reductive activation of CO 2 , contrary to the typical interaction of unoccupied CO 2 2p * u orbitals with lled high-lying metal d orbitals in low-valence metal complexes. 38,39 Theoretically, lowering the energy of the 2p * u (6a 1 ) (LUMO) orbital on CO 2 for interaction with nickel orbitals binding by way of the O]C-unit may be responsible for the coordinated CO 2 reduction and the nonlinearity of the triatomic CO 2 molecule which contains 17 valence electrons, as reported by McGlynn and co-workers. 37 These results illustrate aspects of how a coordinated ligand and the electronic state of the nickel center work in concert to trigger coordination and activation of CO 2 .

Conclusions
Complex 1, with the inherent combination of an electrophilic [Ni III (PS 3 )] core and a properly positioned [OMe] À nucleophile, was employed to provide an optimum electronic condition to trap and activate CO 2 to afford complex 2, containing the O-coordinated [k 1 -CO 2 ]c À ligand. The Ni III -mediated reduction of CO 2 by an adjacent [OMe] À ligand immobilizes CO 2 in the form of [Ni III :CO 2 c À ] and may open a novel CO 2 activation pathway promoting the subsequent incorporation of CO 2 in the buildup of functionalized products. Table 1 Ni 1s / Ni 3d , S 1s / Ni 3d , S 1s / S * C-S transition energy and S 1s / Ni 3d transition intensity of complexes 1, 2, 4, and [Ni(SPh)(PS 3 )] À , derived from the Ni and S K-edge X-ray absorption spectroscopy Complexes Ni 1s / Ni 3d energy a (eV) S 1s / Ni 3d energy b (eV) S 1s / Ni 3d intensity b S 1s / S * a The peak energy is determined by the minimum of the second derivative. b The peak energy and intensity is determined based on the spectral deconvolution. c The intensity-weighted average energy is given here. d Calculated from the difference of the thiolate peak energy and the intensity-weighted pre-edge peak energy.