Redox chemistry and H-atom abstraction reactivity of a terminal zirconium(iv) oxo compound mediated by an appended cobalt(i) center

The reactivity of the terminal zirconium(iv) oxo complex, O 
<svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="23.636364pt" height="16.000000pt" viewBox="0 0 23.636364 16.000000" preserveAspectRatio="xMidYMid meet"><metadata>
Created by potrace 1.16, written by Peter Selinger 2001-2019
</metadata><g transform="translate(1.000000,15.000000) scale(0.015909,-0.015909)" fill="currentColor" stroke="none"><path d="M80 600 l0 -40 600 0 600 0 0 40 0 40 -600 0 -600 0 0 -40z M80 440 l0 -40 600 0 600 0 0 40 0 40 -600 0 -600 0 0 -40z M80 280 l0 -40 600 0 600 0 0 40 0 40 -600 0 -600 0 0 -40z"/></g></svg>
 Zr(MesNPiPr2)3CoCNtBu (2), is explored, revealing unique redox activity imparted by the pendent redox active cobalt(i) center. Oxo complex 2 can be chemically reduced using Na/Hg or Ph3C• to afford the ZrIV/Co0 complexes [(μ-Na)OZr(MesNPiPr2)3CoCNtBu]2 (3) and Ph3COZr(MesNPiPr2)3CoCNtBu (4), respectively. Based on the cyclic voltammogram of 2, Ph3˙ should not be sufficiently reducing to achieve the chemical reduction of 2, but sufficient driving force for the reaction is provided by the nucleophilicity of the terminal oxo fragment and its affinity to bind Ph3C+. Accordingly, 2 reacts readily with [Ph3C][BPh4] and Ph3CCl to afford [Ph3COZr(MesNPiPr2)3CoCNtBu][BPh4] ([5][BPh4]) and Ph3COZr(MesNPiPr2)3CoCl (6), respectively. The chemical oxidation of 2 is also investigated, revealing that oxidation of 2 is accompanied by immediate hydrogen atom abstraction to afford the hydroxide complex [HOZr(MesNPiPr2)3CoCNtBu]+ ([9]+). Thus it is posited that the transient [OZr(MesNPiPr2)3CoCNtBu]+ [2]+ cation generated upon oxidation combines the basicity of a nucleophilic early metal oxo fragment with the oxidizing power of the appended cobalt center to facilitate H-atom abstraction.


General consideration
All manipulations were carried out under an inert atmosphere using a nitrogen-filled glovebox or standard Schlenk techniques unless otherwise noted. All glassware was oven-dried prior to use. Diethyl ether was obtained as HPLC grade without inhibitors. All protio solvents were degassed by sparging with ultra-high purity argon and dried via passage through columns of drying agents using a Seca solvent purification system from Pure Process Technologies.
Benzene-d 6 and dichloromethane-d 2 were dried with CaH 2 overnight, vacuum transferred to an oven-dried vessel, and freeze-pump-thawed, then stored over 4 Å molecular sieves.
(THF)Zr(MesNP i Pr 2 ) 3 CoN 2 , 1 (THF)Zr(MesNP i Pr 2 ) 3 CoCN t Bu, 2 (4) Oxo compound 2 (74.5 mg, 0.0745 mmol) was added to fluorobenzene (1 mL) and stirred for 2 minutes to form a cloudy bright green suspension. Gomberg's dimer (18.1 mg, 0.0372 mmol) was dissolved in fluorobenzene (1 mL) to produce a light-yellow solution that was added into the stirring suspension of 2. After approximately 15 min of stirring, the suspension turned from green to yellow, and the reaction was allowed to stir for 2 h to ensure reaction completion. The cloudy fluorobenzene suspension was filtered to collect the yellow solid product, which was washed with fluorobenzene (0.5 mL). The product was then extracted into diethyl ether (6 mL) and filtered through a plug of glass microfiber filter paper, and the solvent was evaporated from the filtrate under vacuum to obtain pure solid product. Yield: 59.9 mg, 64.7%. A single crystal suitable for X-ray diffraction was grown from a saturated solution of 4  Figure S6. Experimental (black) and simulated (red) X-band EPR spectra of compound 4 showing spin localization on 59 Co (I = 7/2). Spectrum was obtained in frozen fluorobenzene at 30 K with two scans (power attenuation = 30 dB, modulation amplitude = 10 G, modulation frequency = 100 kHz).   Figure S15. 1  atm). 10 The color of the solution quickly changed from dark brown to light yellow. The reaction was stirred for 5 min. The solvent was evaporated in vacuo, and the brown-yellow residue was extracted into pentane. The resulting cloudy light yellow suspension was filtered through a plug of glass microfiber filter paper and the filtrate was then concentrated in vacuo and stored at -30 °C to afford yellow crystalline product. Crystals suitable for X-ray diffraction were grown in the same fashion using slightly more pentane solvent to slow the crystal growth rate. Owing to the reactivity of 8 with air, moisture, and protic solvents, satisfactory elemental analysis and/or high-resolution ESI-MS data could not be obtained. Figure S21. 1 H NMR spectrum (400 MHz, C 6 D 6 ) of HOZr(MesNP i Pr 2 ) 3 CoCN t Bu (8). Peaks corresponding to residual solvents, in this case C 6 D 5 H, THF, and pentane, are labelled.

Additional X-ray data collection, solution, and refinement details for 3
The sodium atom was disordered across three positions. The relative occupancy was refined freely, with an EADP constraint on the thermal parameters and a SUMP instruction for their combined occupancy that converged at 81%, 12% and 7%. During the refinement, electron density difference maps revealed disordered solvent that could not be successfully modeled with or without restraints. Thus, the structure factors were modified using the PLATON SQUEEZE 11 technique, in order to produce a "solvate-free" structure factor set. PLATON reported a total electron density of 98 eand total solvent accessible volume of 510 Å 3 , representing a mixture of two THF and pentane molecules. Figure S32. Fully labelled displacement ellipsoid (50%) diagram of 4•3C 6 H 5 F. The whole molecule and the fluorobenzene solvent molecule were found to be fully disordered; disordered fluorobenzene solvate molecules, and hydrogens are omitted for clarity.

Additional X-ray data collection, solution, and refinement details for 6
There is one B-level alert in the Checkcif that corresponds to the number of reflections missing at low resolution (289/313 92.0 %). One of the fluorobenzene molecules present was located on a special position, so had to be modeled with the PART -1 line. This accounts for the one-half of hydrogen, and fluorine in the formula. Electron density difference maps revealed that there was disordered solvent that could not be successfully modeled, so the structure factors were modified using the PLATON SQUEEZE 11 technique. PLATON reported a total electron density of 92 eand total solvent accessible volume of 430 Å 3 , likely representing 2 fluorobenzene molecules per unit cell. Figure S35.  6 ] was found to be disordered. Hydrogens have been omitted for clarity.

Additional X-ray data collection, solution, and refinement details for [7][PF 6 ]
There is one B-level alert in the Checkcif that corresponds to the number of reflections missing at low resolution (256/279 91.8 %). During the refinement, it became clear that the reflections corresponding to the Miller indices (2,3,2), (-1,0,3), (0,1,3), (-2,1,1), (1,2,1), and (-1,2,2) were affected by the shadow of the beamstop (error = >10), and have been omitted. The hexafluorophosphate counteranion was found to be fully disordered and was modeled with twocomponent disorder, where the sum of the major and minor components was constrained to be one. Figure S36. Fully labelled displacement ellipsoid diagram of 8. Hydrogens except for the hydroxide H have been omitted for clarity.

Additional X-ray data collection, solution, and refinement details for 8
Crystals of this compound exhibited very weak diffraction patterns, so a large number of reflections were unobserved. Nonetheless, the structure itself is unequivocally determined. Disordered solvent could not be adequately modeled with or without restraints. Thus, the structure factors were modified using the PLATON SQUEEZE technique 12 in order to produce a "solvate-free" structure factor set. PLATON reported a total electron density of 1708 eand total solvent accessible volume of 5637 Å 3 , likely representing several molecules of pentane. Figure S37. Fully labelled displacement ellipsoid diagram of [9][BAr F 4 ], with occupational disorder modeled of hydroxide trace iodide (~1%). Hydrogens except for the hydroxide H have been omitted for clarity.

]
There is one B-level alert in the Checkcif that corresponds to the number of reflections missing at low resolution (239/366 92.6%). During the refinement, it became clear that the reflections corresponding to the Miller indices (0,2,0), (2,0,1), (0,2,2), (1,-1,2), (1,2,1), (-1,0,3) and (2,0,0) were affected by the shadow of the beamstop (error = >10), and have been omitted. There is occupational disorder of the hydroxide with iodide (~1%), and this was modeled with twocomponent disorder, where the sum of the major and minor components was constrained to be one. The hydrogen atom of the hydroxide was restrained with DFIX commands. One of the -CF 3 groups of the [BAr F 4 ]counteranion was found to be disordered and was modeled with twocomponent disorder, where the sum of the major and minor components was constrained to be one. Electron density difference maps revealed that there was disordered solvent that could not be successfully modeled, so the structure factors were modified using the PLATON SQUEEZE 11 technique. PLATON reported a total electron density of 92 eand total solvent accessible volume of 430 Å 3 , likely representing 2 fluorobenzene molecules per unit cell.

Computational Details
All calculations were performed using Gaussian16 for Linux operating system. 13 Density functional theory calculations were carried out using a combination of Becke's 1988 gradientcorrected exchange functional 14 and Perdew's 1986 electron correlation functional (BP86). 15 A mixed-basis set was employed, using the LANL2TZ(f) triple-ζ basis set with effective core potentials for cobalt and zirconium, [16][17][18] Gaussian16's internal 6-311+G(d) for heteroatoms (nitrogen, oxygen, and phosphorus), and Gaussian16's internal LANL2DZ basis set (equivalent to D95V) 19 for carbon and hydrogen. Using crystallographically determined geometries as a starting point and modifying as needed, the geometries were optimized to a minimum, followed by analytical frequency calculations to confirm that no imaginary frequencies were present. XYZ coordinates of optimized geometries are provided in Tables S4-S5.  Figure S38. 19 F and 31 P{ 1 H} NMR spectra of the reaction mixture between 2 and FcPF 6 , revealing multiple PF 6derived byproducts that contain both 31 P and 19 F nuclei. The doublet centered at -56 ppm in the 19 F NMR spectrum (J PF = 900 Hz) and the quintet centered at -68 ppm in the 31 P NMR spectrum (J PF =900 Hz) are tentatively assigned to a species of the general form PF 4 OR (where R could potentially = H) based on several literature reports (Makromol. Chem. 1979, 180, 1509-1519; J. Am. Chem. Soc. 1997, 119, 3918-3928).