A Redox Transmetalation Step in Nickel-Catalyzed C–C Coupling Reactions

Ni-catalyzed C–H functionalization reactions are becoming efficient routes to access a variety of functionalized arenes, yet the mechanisms of these catalytic C–C coupling reactions are not well understood. Here, we report the catalytic and stoichiometric arylation reactions of a nickel(II) metallacycle. Treatment of this species with silver(I)–aryl complexes results in facile arylation, consistent with a redox transmetalation step. Additionally, treatment with electrophilic coupling partners generates C–C and C–S bonds. We anticipate that this redox transmetalation step may be relevant to other coupling reactions that employ silver salts as additives.

), and all coupling constants (J) are reported in Hz. 1 IR spectra were recorded on a PerkinElmer (Spectrum 100) FT-IR spectrometer. High resolution mass spectra were obtained on a Thermofisher ScientificQ Exactive mass spectrometer. Cyclic voltammograms were recorded in a nitrogen-filled glovebox using a PINE WaveNow portable potentiostat. All potentials are reported versus ferrocene/ferrocenium. UV-Visible spectra were recorded on a Shimadzu UV-1800 spectrophotometer using quartz cuvettes (path length = 1 cm) or a Pine Research Honeycomb Spectroelectrochemical Cell (path length = 0.17 cm) for the spectroelectrochemical studies consisting of a honeycomb patterned electrode card with a platinum working electrode and two counter electrode bands. Column chromatography was performed using Silicycle SiliaFlash P60 silica gel. Elemental Analyses were performed by Atlantic Microlab, Inc., Norcross, GA.

II. Synthesis and Characterization of Nickel Complexes 1 and 2a
2-(Quinolin-8-yl)isoindoline-1,3-dione. A round bottom flask was charged with phthalic anhydride (0.74 g, 5.0 mmol) and 8-aminoquinoline (0.72 g, 5.0 mmol). Acetic acid (30 mL) was added and the mixture was heated at reflux for 2 hours. The reaction mixture was allowed to cool to room temperature and an equal volume of water (30 mL) was added. The mixture was allowed to sit overnight and off-white needles crystallized. The mixture was filtered via a Büchner funnel and washed with water (750 mL). The resulting solids were dried under vacuum and subsequently recrystallized from CHCl3 to yield 987 mg of 2-(quinolin-8-yl)isoindoline-1,3-dione as an offwhite crystalline solid (3.60 mmol, 72 % Complex 1. In a N2 filled glovebox, an oven dried 20 mL scintillation vial equipped with a stir bar was charged with 2-(quinolin-8-yl)isoindoline-1,3-dione (274 mg, 1.00 mmol) and THF (20 mL). An oven dried 100 mL round bottom flask also equipped with a stir bar was charged with THF (15 mL) and Ni(COD)2 (275 mg, 1.00 mmol). The solution of 2-(quinolin-8-yl)isoindoline-1,3-dione was then added dropwise to the Ni(COD)2 solution with stirring. After complete addition of the 2-(quinolin-8-yl)isoindoline-1,3-dione solution, the reaction mixture was allowed to stir for 3 h at room temperature, during which a red-orange solid precipitated. After completion, Et2O (~20 mL) was added to effect further precipitation of the product. The reaction mixture was filtered by vacuum filtration through a Büchner funnel and the red-orange residue washed with Et2O (~50 mL

Complex 2a.
In a N2 filled glovebox, complex 1 (483 mg, 1.45 mmol) was added to an oven dried 40 mL pressure tube with a stir bar. Toluene (20 mL) was added and the pressure tube was sealed with a Teflon screw cap. The reaction mixture was then removed from the glovebox and placed in an oil bath pre-heated to 160 °C. After 2 h, the reaction mixture was removed from the oil bath and allowed to cool to room temperature. The reaction mixture was allowed to sit undisturbed for 1 week, during which a green precipitate settled out and yellow crystals formed in the clear yellow toluene layer. The pressure tube was opened on the benchtop and a pipet was used to remove the first batch of yellow crystals. The remaining mixture was poured into a 50 mL beaker and ethyl acetate was added in ~5 x 10 mL portions. The yellow crystals were allowed to settle to the bottom of the beaker and the layer of ethyl acetate was decanted along with any unwanted green precipitate. This process was repeated until only the pure yellow crystals remained. After obtaining an electrolyte cyclic voltammogram, the analytes were dissolved in the electrolyte. After the completion of the electrochemical experiments, an internal standard (decamethylferrocene (Cp*2Fe)) was dissolved in the analyte solution and an additional cyclic voltammogram was collected. The decamethylferrocene exhibited poor solubility in the analyte (<3.3mg/10mL) and was only used for referencing peak potentials.

III. Exchange Reactions of Complex of 2a (Scheme 4 in manuscript)
Synthesis of Complex 2b. To a sample of complex 2a (10.0 mg, 0.0300 mmol) in a 50 mL round bottom flask was added acetonitrile (2 mL). CO(g) release was indicated by the generation of tiny bubbles. The solution was sonicated for 2 min, after which the solvent was removed by rotary evaporation to reveal a yellow-orange solid.

S7
Complex 2b-d3. To a sample of complex 2a (3.2 mg, 9.6 μmol) in a 50 mL round bottom flask was added CD3CN (1.5 mL). The solution was sonicated for 1 min, resulting in a yellow solution. The solvent was removed by rotary evaporation to reveal an orange solid residue. The flask with residue was transferred to a high vacuum line for ~ 2 h to remove residual solvent. The sample was taken in CD2Cl2 and the yellow suspension filtered through a pipet containing a plug of cotton, and packed with Celite (0.5 cm) directly into an NMR tube.

2b-d3
Complex 2c. To a stirred yellow suspension of complex 2a (34.2 mg, 0.103 mmol) in dry acetone (20 mL) in a 50 mL round bottom flask was added 4-picoline (35 μL, 0.36 mmol) in one portion. Immediately, the mixture became an orange solution. The solution was stirred at room temperature for 1 h after which the solvent was removed under vacuum. The orange residue was dissolved in acetone (1 mL) and an ochre colored powdery solid was precipitated with a mixture of pentane : Et2O (3:2, 15 mL). The solid was filtered via a Hirsch funnel and quickly rinsed with 5 mL pentane and dried under vacuum to yield complex 2c in 94% yield (38.4 mg, 0.096 mmol). The 1 H NMR spectroscopic parameters are consistent with reported values.

Stoichiometric activity of Complex 2a.
An oven-dried 50 mL Schlenk tube equipped with a stir bar was charged with 4-methyl-2-phenyl-1,3-thiazole-5-carboxylic acid (39.5 mg, 0.180 mmol), Na2CO3 (19.1 mg, 0.180 mmol), and dry DMA (1 mL). The reaction tube was placed in a pre-heated oil bath at 110 °C and stirred for 30 min. The solvent was removed under reduced pressure to dryness and the resulting carboxylate salt was used without further purification. Complex 2a (20.1 mg, 0.060 mmol), Ag2CO3 (33.0 mg, 0.120 mmol), and PivOH (9.1 mg, 0.089 mmol) were added and the tube was evacuated and backfilled with nitrogen three times after which DMA (1.5 mL) was added via syringe. The reaction mixture was stirred at 110 °C for 24 h. Upon completion, the reaction tube was cooled to room temperature. The solution was diluted with ethyl acetate (25 mL) and poured into a 250 mL separatory funnel, HCl (1 N, 10 mL) was added and the layers separated. The aqueous layer was further extracted with ethyl acetate (2 x 25 mL). The combined organic layers were washed with water (20 mL) and brine (20 mL), dried over Na2SO4, filtered and concentrated under vacuum. The dark brown crude material was dissolved in CDCl3 with 1,3,5-trimethoxybenzene (1.4 mg, 8.3 μmol) as an internal standard and analyzed by 1 H NMR spectroscopy. The average yield of 2,6-bis(4-methyl-2-phenyl-1,3-thiazol-5-yl)-N-(quinolin-8-yl)benzamide was calculated to be 49% (0.0295 mol). The spectroscopic data are consistent with literature values. 4

V. Reactions of 2a with Silver(I)-Aryl Complexes to Generate Products 3a-b (Scheme 6 in manuscript)
(2,3,4,5,6-pentafluorophenyl)silver(acetonitrile). The title compound was prepared by a modification of a literature synthesis. 5 AgF (317 mg, 2.50 mmol) was added to an oven-dried 20 mL vial with anhydrous MeCN (5 mL) in a N2 filled glovebox. The solution was allowed to stir for 15 minutes. Trimethyl(pentafluorophenyl)silane (0.486 mL, 2.55 mmol) was then added in one portion and the reaction mixture was allowed to stir for 1 h. The 20 mL reaction vial was then placed in a freezer at -30 °C for 1 day. The acetonitrile was decanted from the 20 mL reaction vial leaving a pale grey slurry. The grey slurry was then dried under vacuum to yield 584 mg (1.85 mmol, 74%) of the title compound as a pale grey crystalline solid. General procedure for the reactions of complex 2a with silver(I)-aryl complexes. In a N2 filled glovebox a solution of the silver(I)-aryl (0.025 mmol, 1.0 equiv or 0.050 mmol, 2.0 equiv) in anhydrous DMA (1 mL) was added dropwise to a mixture of complex 2a (8.3 mg, 0.025 mmol) and anhydrous DMA (1 mL). The resulting mixture was stirred at room temperature for 1 h. The reaction mixture was then removed from the glovebox, diluted with ethyl acetate (15 mL), poured into a 100 mL separatory funnel and HCl (2 N, 10 mL) added and the layers separated. The aqueous layer was further extracted with ethyl acetate (2 x 15 mL). The combined organic layers were washed with water (20 mL) and brine (15 mL), dried over Na2SO4, filtered, and concentrated under vacuum. To the crude residue was added 1,3,5-trimethoxybenzene (1.8-2.3 mg, 0.011-0.014 mmol) and the sample dissolved in CDCl3 for 1 H NMR analysis.
Compound 3c. When 1 equiv of (2,3,4,5,6-pentafluorophenyl)silver(acetonitrile) was used, a 63% yield was calculated by 1 H NMR spectroscopy (see Figure S6 below). Under these conditions 12% yield of the diarylated product was also observed (confirmed by mass spectrometry). These yields are the average of two independent runs giving 65% and 60% yields of 3c and 13% and 11% yields of the diarylated product.
When 2 equiv of (2,3,4,5,6-pentafluorophenyl)silver(acetonitrile) were used, a 46% yield of 3c and a 38% yield of diarylated product were calculated by 1 H NMR spectroscopy. These yields are the average of two independent runs giving 46% and 46% yields of 3c and 38% and 37% yields of the diarylated product.

VI. Reactions of 2a with Coupling Partners to Generate Products 3e-g (Scheme 8 in manuscript)
Compound 3e. To a mixture of complex 2a (8.3 mg, 0.025 mmol) and anhydrous DMA (1 mL) in a 1 dram vial with a stir bar in a N2 filled glovebox, was added a solution of phenyl disulfide (5.4 mg, 0.025 mmol) in anhydrous DMA (1 mL) dropwise. The vial was tightly capped and sealed with electrical tape. The vial was then removed from the glovebox and placed in a preheated oil bath at 110 °C for 1 hour. The reaction mixture was then allowed to cool to room temperature and the crude mixture was poured into a 100 mL separatory funnel. To the solution, was added HCl (2 N, 10 mL) and the layers separated. The aqueous layer was further extracted with ethyl acetate (2 x 15 mL). The combined organic layers were washed with water (20 mL) and brine (15 mL), dried over Na2SO4, filtered, and concentrated under vacuum. Then 1,3,5-trimethoxybenzene (2.0 mg, 0.012 mmol) was added to the residue and the crude mixture was dissolved in CDCl3 for 1 H NMR analysis. The spectroscopic data are consistent with literature values 8 and a 60% yield was calculated by 1 H NMR spectroscopy (see Figure S8 below).  (1 mL) dropwise. The mixture was stirred at room temperature for 1 h. The reaction vial was then removed from the glovebox and poured into a 100 mL separatory funnel. To the separatory funnel, water (25 mL) was added followed by aqueous HCl (2 N, 3 mL) and the resulting mixture was separated and the aqueous layer was subsequently extracted with ethyl acetate (2 x 30 mL). The combined organic layers were washed with water (3 x 30 mL) and brine (~15 mL), dried over Na2SO4, filtered, and concentrated under vacuum. Then 1,3,5-trimethoxybenzene (2.0 mg, 0.012 mmol) was added to the residue and the crude mixture was dissolved in CDCl3 for 1 H NMR analysis. The spectroscopic data are consistent with literature values 9 and a 62% yield was calculated by 1 H NMR spectroscopy (see Figure S11 below).  In a N2 filled glovebox, a 1 dram vial equipped with a stir bar was charged with complex 2a (19.7 mg, 0.059 mmol) and anhydrous DMA (1.5 mL) and the mixture stirred. To this mixture was added a solution of PhI(OAc)2 (19.4 mg, 0.059 mmol) in anhydrous DMA (1.5 mL) via pipet. The resulting solution was stirred at room temperature for 1 h. At the end of this period, the reaction mixture was removed from the glovebox and diluted with ethyl acetate (15 mL) then poured into a 100 mL separatory funnel. HCl (2 N, 10 mL) was then added and the layers separated. The aqueous layer was further extracted with ethyl acetate (2 x 15 mL). The combined organic layers were washed with water (20 mL) and brine (15 mL), dried over Na2SO4, filtered, and concentrated under vacuum. To the residue was added 1,3,5-trimethoxybenzene (1.4 mg, 0.0083 mmol) and the crude mixture dissolved in CDCl3 for 1 H NMR analysis. A 49% yield was calculated by 1 H NMR spectroscopy (see Figure S10 below) and is the average of two independent runs giving 48% and 50% yields The crude material was purified by silica gel column chromatography (gradient elution, 100% hexanes to hexanes : ethyl acetate (4:1, v/v)) and 3g was isolated for further characterization.  PhI(OAc) 2 S17 Figure S12. Example crude 1 H NMR spectrum of the reaction of complex 2a with (diacetoxyiodo)benzene taken in CDCl3 at 600 MHz. The product peaks used are indicated with blue circles ( ) and the internal standard is indicated with an asterisk (*).

Stoichiometric reaction of complex 2a and Zn(C6F5)2 at room temperature.
In a N2 filled glovebox, to a solution of complex 2a (10.9 mg, 0.033 mmol) in anhydrous DMA (1.0 mL) in a 1 dram vial equipped with a stir bar was added a solution of Zn(C6F5)2 (13.2 mg, 0.033 mmol) in anhydrous DMA (1.0 mL). The resulting solution was stirred at room temperature for 1.5 h. At the end of this period, the reaction mixture was removed from the glovebox and diluted with ethyl acetate (15 mL) then poured into a 100 mL separatory funnel. HCl (2 N, 10 mL) was then added and the layers separated. The aqueous layer was further extracted with ethyl acetate (2 x 15 mL). The combined organic layers were washed with water (20 mL) and brine (15 mL), dried over Na2SO4, filtered, and concentrated under vacuum. The crude material was dissolved in CDCl3 with 1,3,5-trimethoxybenzene (1.6 mg, 9.5 μmol) as an internal standard and analyzed by

Stoichiometric reactions of complex 2a and Zn(C6F5)2 at 110 °C.
Reaction in a Vial. In a N2 filled glovebox, a solution of complex 2a ( Reaction in a J. Young NMR tube. In a N2 filled glovebox, a solution of complex 2a (5.1 mg, 0.015 mmol) in anhydrous proteo-DMA (0.5 mL) was prepared in a J. Young NMR tube and CD2Cl2 (0.2 mL) added. The tube was sealed, then removed from the glovebox and an initial spectrum was acquired. The tube was returned to the glovebox and Zn(C6F5)2 (6.0 mg, 0.015 mmol) added and the tube resealed. The tube was removed from the glovebox and a spectrum acquired at room temperature. The tube was then heated at 110 °C in a pre-heated oil bath and spectra acquired after 1 h and 2 h. CD 2 Cl 2 S20 filtrate was placed in a centrifuge tube and the residual silver removed following centrifuging of the sample and decanting the solution. This was repeated once more. The solvent was then removed in vacuo to reveal an orange-brown residue. 1 H NMR analysis in acetone-d6 with 1,3,5trimethoxybenzene (2.2 mg, 13.1 μmol) as an internal standard revealed compound 3c in 7% yield, along with other unidentified components.

Stoichiometric reaction of complex 2c and (MeCN)Ag(C6F5) at -42 °C.
The recently reported related 4-picoline-bound nickelacycle 3 was also explored in stoichiometric reactions with the silver-aryl species in an attempt to obtain an isolable Ni III intermediate.
In The residue was taken in THF (5 mL) and the mixture filtered through Celite in a fritted Büchner funnel, and the pad of Celite washed with THF (4 mL). The brown filtrate was placed in a centrifuge tube and the residual silver removed following centrifuging of the sample and decanting the solution. This was repeated once more. The solvent was then removed under vacuum to reveal an orange-brown residue. The crude material was dissolved in acetone-d6 with 1,3,5trimethoxybenzene (1.4 mg, 8.3 μmol) as an internal standard. A 23% yield of compound 3c was obtained by 1 H NMR spectroscopy along with other unidentified components.

UV-Vis Spectroelectrochemical study of Complex 2a in DMA
To obtain UV-visible spectra of species generated upon oxidation of 2a.

Description of the X-ray Structural Analysis of 2a [(C16H10N2O)Ni(CO)].
A long yellow parallelepiped crystal of 2a (C16H10N2O)Ni(CO) was coated in polybutene oil (Sigma-Aldrich) and placed on the end of a MiTeGen loop. The sample was cooled to 100 K with an Oxford Cryostream 700 system and optically aligned on a Bruker AXS D8 Venture fixed-chi X-ray diffractometer equipped with a Triumph monochromator, a Mo Kα radiation source (l = 0.71073 Å), and a PHOTON 100 CMOS detector. Three sets of 12 frames each were collected using the omega scan method with a 10 second exposure time. Integration of these frames followed by reflection indexing and least-squares refinement produced a crystal orientation matrix for the monoclinic crystal lattice that was used for the structural analysis.
Data collection consisted of the measurement of a total of 740 frames in four runs using omega scans with the detector held at 5.00 cm from the crystal. Frame scan parameters are summarized in Table S1 below: The APEX3 software program (version 2016.9-0) 11 was used for diffractometer control, preliminary frame scans, indexing, orientation matrix calculations, least-squares refinement of cell parameters, and the data collection. The frames were integrated with the Bruker SAINT software package using a narrow-frame algorithm. The integration of the data using a monoclinic unit cell yielded a total of 29130 reflections to a maximum θ angle of 30.09° (0.71 Å resolution), of which 3891 were independent (average redundancy 7.487, completeness = 99.8%, Rint = 3.05%, Rsig = 1.96%) and 3417 (87.82%) were greater than 2σ(F 2 ). The final cell constants of a = 10.9298(5) Å, b = 7.1659(3) Å, c = 17.2345(7) Å, β = 101.0570(10)°, volume = 1324.78(10) Å 3 , are based upon the refinement of the XYZ-centroids of 9902 reflections above 20 σ(I) with 6.174° < 2θ < 60.11°. Data were corrected for absorption effects using the multi-scan method (SADABS). The ratio of minimum to maximum apparent transmission was 0.859. The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.541 and 0.921.
The structure was solved by using the intrinsic phasing routine available in the APEX3 software 11 and refined using the programs provided by SHELXL-2014/7. 12 The crystallographic asymmetric unit consists of only a molecule of (C16H10N2O)Ni(CO). Idealized positions for the aromatic hydrogen atoms were included as fixed contributions using a riding model with isotropic temperature factors set at 1.2 times that of the adjacent carbon atom. Full-matrix least-squares refinement, based upon the minimization of Swi |Fo 2 -Fc 2 | 2 , with weighting wi -1 = [s 2 (Fo 2 ) + (0.0254 P) 2 + 1.0711 P], where P = (Max (Fo 2 , 0) + 2 Fc 2 )/3. 12 The final anisotropic full-matrix least-squares refinement on F 2 with 199 variables converged at R1 = 2.48 % for the 3417 data with I>2σ(I) and wR2 = 6.11 % for all data. The goodness-of-fit was 1.033. 13

S24
A correction for secondary extinction was not applied. The largest peak in the final difference electron density synthesis was 0.537 e -/Å 3 and the largest hole was -0.290 e -/Å 3 with an RMS deviation of 0.063 e -/Å 3 . The linear absorption coefficient, atomic scattering factors, and anomalous dispersion corrections were calculated from values found in the International Tables of X-ray Crystallography. 14

Calculations on Pentafluorophenyl System
Redox Transmetalation. In addition to the calculations discussed in the main text of the manuscript, DFT calculations were performed on (DMA)AgC6F5 and 2d. Similar to the calculations performed on the (DMA)AgPh NO2 system, a Ni-Ag adduct was calculated, and the association was found to be exergonic by -10.9 kcal/mol, nearly identical to the energy calculated for the corresponding adduct with (DMA)AgPh NO2 . We were unable to find transition states for the adduct formation or for the transfer of the C6F5 group to the Ni center, but we found that the transfer of the C6F5 to the Ni center is mildly endergonic (+6.8 kcal). The findings are illustrated in Figure S15, below.

C-C Coupling.
We found that C-C coupling from LNiC6F5 had a barrier of 21.7 kcal/mol, much higher than that seen for LNiPh NO2 . The difference lies in the coordination of the NO2 group in compound 4. Indeed, we found that binding a DMA molecule to LNi(C6F5) was exergonic by -9.5 kcal/mol, and led to a much lower C-C coupling barrier (+8.2 kcal/mol), that is in better agreement with the low barrier to coupling for 4. Clearly a 5-coordinate Ni III species is needed for rapid C-C coupling in catalysis. The findings are illustrated in Figure S16, below.