Aldehydes and ketones influence reactivity and selectivity in nickel-catalysed Suzuki–Miyaura reactions†‡§

The energetically-favorable coordination of aldehydes and ketones – but not esters or amides – to Ni0 during Suzuki–Miyaura reactions can lead either to exquisite selectivity and enhanced reactivity, or to inhibition of the reaction. Aryl halides where the C–X bond is connected to the same π-system as an aldehyde or ketone undergo unexpectedly rapid oxidative addition to [Ni(COD)(dppf)] (1), and are selectively cross-coupled during competition reactions. When aldehydes and ketones are present in the form of exogenous additives, the cross-coupling reaction is inhibited to an extent that depends on the strength of the coordination of the pendant carbonyl group to Ni0. This work advances our understanding of how common functional groups interact with Ni0 catalysts and how these interactions affect workhorse catalytic reactions in academia and industry.


4-methyl-3-(trifluoromethyl)-1,1'-biphenyl
Synthesised according to the General Procedure A using 3-bromobenzotrifluoride (180 μL, 290 mg, 1.3 mmol), 4-tolylboronic acid (206.1 mg, 1.5 mmol), [PdCl2(dppf)] (40.1 mg, 4.2 mol%), and K3PO4 (808.9 mg, 3.8 mmol) in 5 mL toluene. The desired product was purified via flash column chromatography (eluting with hexane) to yield a colourless oil that solidified upon drying under high vacuum (252.8 mg, 83%). In a round bottom flask equipped with a stirrer bar, aldehyde (1 eq.) and ketone (1 eq.) were dissolved in EtOH (14 mL) or a mixture of EtOH (25 mL) and THF (25 mL). Once dissolved, 10 % w/v solution of NaOH (6 mL) was added. The mixture was stirred for 5-15 minutes at room temperature and the resulting chalcone was filtered and washed with EtOH (3 x 5 mL) to give a white or off-white solid. GC-MS was carried out to check conversion and crude material was carried through to the next step. Allylic Alcohol. In a round bottom flask equipped with a stirrer bar, chalcone (1 eq.) and CeCl3.7H2O (1 eq.) were dissolved in MeOH (10 mL) and THF (50 mL). Once dissolved, the mixture was cooled to 0 o C with an ice bath. NaBH4 (1.5 eq.) was slowly added in portions. Once all of the NaBH4 was added, the mixture was warmed to room temperature and stirred for 15 minutes. The reaction was neutralised to pH 7 with 1 M HCl and distilled water (100 mL) was added. The mixture was extracted 3 times with Et2O (3 x 50 mL). The combined organic phases were dried over MgSO4, which was filtered and the Et2O removed under vacuum to furnish the allylic alcohol as a white oil. NMR was carried out to check conversion and crude material was carried through to the next step. Rearrangement to Saturated Chalcone. In a microwave vial equipped with a stirrer bar, allylic alcohol (1 eq.), [IrCl(IPr)(COD)] (0.1 mol%) and KOH (10 mol%) were dissolved (KOH suspended) in THF (2 -4 mL). The reaction was heated in a Biotage Initiator Microwave Synthesiser at 150 o C for 2 -4 hours. The resulting mixture was filtered through celite and the solvent removed under reduced pressure. 1 H NMR was carried out to check conversion. The desired compound was either: purified via flash column S9 chromatography, eluting with hexane to yield a white or off-white solid; or recrystallised from hot hexane and filtered to yield a white or off-white solid

GC-FID CALIBRATION
The GC-FID apparatus was calibrated for each analyte using a series of standards, accurately prepared, containing varying ratios of internal standard and analyte. In each case, a plot of the relative peak areas versus the molar ratio gave a straight line, and the slope of this line was used as the response factor.

Substrate
Internal Standard Response Factor

DESIGN OF EXPERIMENTS DATA FOR REACTION OPTIMISATION
For each reaction, solid components were loaded into a microwave tube equipped with a stir bar, sealed with a septum-fitted crimp-cap, and evacuated and backfilled with argon or nitrogen several times. The liquid reagents and the reaction solvent were added via syringe through the septum. The reactions were then heated, with stirring, for 18 h. After this time, the reaction was cooled to room temperature, and an accurately-known mass of dodecane or tetradecane was added. A sample of the solution was then diluted in chloroform for analysis by GC-FID. The DoE study was initially conducted using the dppe ligand, but comparable results are obtained using dppf under the same conditions. All other work was conducted with dppf as the model nickel(0) complex for kinetic studies and ligand binding studies used a dppf ligand.

Initial Screen
Run S13 Data were analysed using DesignExpert 8. From this, it was deduced that the catalyst loading had a positive effect on conversion, though it was not as significant as the temperature effect. Overall, this initial screen gave positive results, since the centre points appeared to proceed to full conversion. In order to further probe the reaction conditions, and potentially reduce factors such as catalyst loading, the experiment design was augmented to narrow in on optimised conditions.

Competition Reactions with Boronic Acids
Reaction of 1 equiv. p-bromotoluene with 1 equiv. of each of two boronic acids. Yields of each product determined by calibrated GC-FID analysis. Results are quoted as the average of two replicates.

DATA FROM ROBUSTNESS SCREENING REACTIONS
A microwave tube equipped with a stir bar was charged with 7 (5 mol%), K3PO4 (3 equiv.), p-tolB(OH)2 (1.1 equiv.) and the additive (if solid). The tube was sealed with a crimp cap and evacuated and backfilled with nitrogen or argon. 4-(Trifluoromethyl)bromobenzene was added via syringe (0.25 mmol, 1 equiv.), followed by the additive (if liquid), anhydrous toluene (1 mL), and degassed distilled water (10 equiv.). The reaction was heated to 85 °C with stirring for 2 h. Upon cooling, the tube was opened, a known mass of tetradecane or dodecane was added, and the mixture was stirred briefly. A sample was withdrawn, diluted with chloroform, and analysed by GC-FID.

EQUILIBRIUM CONSTANTS FOR THE BINDING OF ALDEHYDES AND KETONES TO NICKEL(0)
Equilibrium

KINETIC DATA FOR OXIDATIVE ADDITION TO NICKEL(0)
Kinetic data were obtained in the same manner as that used for our previous paper. 17 Liquid substrates were added neat to a septum-fitted NMR tube containing a solution of [Ni(COD)(dppf)] (1) in benzene-d6 that had been equilibrated at 20 °C. For solid substrates, a solution of [Ni(COD)(dppf)] in benzene-d6, equilibrated at 20 °C, was added to the solid substrate. 31 P NMR spectra were acquired at intervals, with a long D1 (25 seconds) and without 1 H decoupling. All kinetic data showed pseudo-first order behaviour in [Ni(COD)(dppf)] and so plots of ln(1) versus time yielded kobs. Each experiment was performed in duplicate and so rate constants are quoted as the average of these two replicates. Data for 5-Cl, 5-Br, 5-I, and 6-Cl can be found in our previous manuscript.

ENERGIES AND COORDINATES FROM DFT CALCULATIONS
Calculations were carried out using Gaussian09 Rev. D01 using desktop machines and the ARCHIE-WeSt High-Performance Computer. The B3LYP functional was used with Grimme's D3 corrections to account for dispersive interactions. [18][19][20] Solvation (toluene) was treated using the SMD implicit solvation model. 21 The LANL2TZ(f) basis set was used for Ni and Fe, LANL2DZ(d,p) was used for Br, and 6-31G(d) was used for all other atoms for optimisation. [22][23][24][25] Geometry optimisation was carried out in solvent, without symmetry constraints. The nature of each stationary point was verified using frequency calculations. Energies were then refined using single point energy calculations with 6-311+G(d,p) for all atoms except Ni, Fe, and Br, which were treated as noted above. Coordinates can be found in a separate Supporting Information file specifically for this content. The energies of intermediates A, B, and C for a series of substrates are tabulated on this page. Energies are tabulated on the following pages in Hartrees. Electronic energies (E), and corrections to enthalpy (Hcorr) and free energy (Gcorr) are reported here with the smaller basis set (6-31G(d) on H/C/N/O/F/P), along with electronic energies using the larger 6-311+G(d,p) basis set on these atoms (denoted E').