Enantioselective α-Arylation of Ketones via a Novel Cu(I)–Bis(phosphine) Dioxide Catalytic System

A novel catalytic system based on copper(I) and chiral bis(phosphine) dioxides is described. This allows the arylation of silyl enol ethers to access enolizable α-arylated ketones in good yields and enantiomeric excess up to 95%. Noncyclic ketones are amenable substrates with this method, which complements other approaches based on palladium catalysis. Optimization of the ligand structure is accomplished via rational design driven by correlation analysis. Preliminary mechanistic hypotheses are also evaluated in order to identify the role of chiral bis(phosphine) dioxides.


General Information
All reactions were carried out in oven-or flame-dried glassware under an atmosphere of dry Nitrogen unless otherwise noted. Except as otherwise indicated, all reactions were magnetically stirred and monitored by analytical thin layer chromatography (TLC) using Merck pre-coated silica gel plates with F254 indicator. Visualization was accomplished by UV light (254 nm), with combination of Cerium Ammonium Molybdate solution as an indicator. Flash column chromatography was performed using silica gel pore size 60 Å, 230-400 mesh particle size, 40-63 μm particle size or Aluminium oxide 90 active neutral. Yields refer to chromatographically and spectrographically pure compounds, unless otherwise noted. Commercial grade reagents and solvents were used without further purification. 1 H NMR, 13 C NMR, 19 F and 31 P spectra were recorded on Bruker Avance300 spectrometer. The proton spectra are reported as follows δ (position of proton, multiplicity, coupling constant J, number of protons). Multiplicities are indicated by s (singlet), d (doublet), t (triplet), q (quartet), p (quintet), h (septet), m (multiplet) and br (broad). Enantiomeric excess was determined on a Shimadzu HPLC SPD-10A with a variable wavelength detector using chiral stationary phase columns (0.46 cm x 25 cm) from Phenomenex or Daicel. Unless otherwise noted, all reagents were obtained commercially and used without further purification.

S4
The reaction was tested in different solvents. No conversion was obtained but for DCM, CHCl3, or DCE.
Finally, several ligands were tested in order to optimize the enantioselectivity:

General procedure A: preparation of the silyl enol ethers
A solution of diisopropylamine (1.95 mL, 13.96 mmol) in THF (40 mL) was cooled to 0 °C and a 2.5 M solution of nbutyllithium in n-hexane (5.58 mL, 13.96 mmol) was added slowly by a syringe. This mixture was stirred for 20 min at 0 °C and subsequently cooled to -78 °C. A solution of the corresponding ketone (11.63 mmol) in THF (5 ml) was added by a syringe over a 2 min to the mixture. The solution was stirred for 1 h at -78 °C, then TMSCl (1.77 mL, 13.96 mmol) was added dropwise. The reaction mixture was subsequently allowed to warm up slowly to room temperature and it was stirred overnight. The reaction mixture was quenched with a saturated aqueous solution of NaHCO3 (10 mL) and diluted with n-hexane. The layers were separated and the aqueous phase was extracted with hexane (2 x 20 mL). The combined organic layers were washed with H2O (2 x 15 mL) and brine, dried over anhydrous Na2SO4, filtered, and concentrated to obtain the crude product. This was purified by column chromatography (6% H2O deactivated 90 neutral aluminium oxide and hexane as eluent). The isomer ratio were determined by 1 H NMR spectroscopy of the crude products.

S8
Characterization of the arylated ketones General procedure B: α-Arylation of ketones In a N2-filled glove box, an oven dried vial was charged with (CuOTf)2.Tol (4 mol%). The vial was sealed with a ptfe septum screw cap and was removed from the glove box. We observed that adding other solid reagents into the vial out of the glovebox exposing the Cu salt to the air, provided identical (sometimes slightly superior) results. Therefore, the phosphine oxide ligand 3x (12 mol%), the appropriate mesitylaryliodonium triflate 5 (1.0 equiv.), and a magnetic stirrer were quickly added to the vial out of the glovebox for simplicity. The vial was then sealed and filled with nitrogen and then dichloromethane (0.2 M). After 1 min the appropriate silyl enol ether 4 (2 equiv.) was added via syringe and the reaction mixture stirred for 16 h at room temperature. The reaction was quenched by addition of diethyl ether (2 ml), filtered through a short pad of silica gel, and washed with diethyl ether (ca. 8 ml). A stock solution of the internal standard in DCM was added, and the mixture was concentrated under vacuum. NMR yields were determined on the crude product (internal standard: dimethyl terephthalate, 6H, 3.96 ppm and 4H, 8.1 ppm). Purification by column chromatography on silica gel using the solvent mixture reported below afforded the desired α-arylated ketone (see yields below). The enantiomeric purity was determined by chiral HPLC analysis on the purified product.
Synthesis of the racemic compounds for HPLC peaks attribution was prepared according to the same procedure using rac-SEGPHOSO 3n instead of (R)-3x as ligand.

(3S,4R)-3-(naphthalen-2-yl)heptan-4-ol (13)
In accordance to a reported procedure, a solution of lithium aluminium hydride (1.0 M in hexanes; 99 µL, 0.099 mmol) was slowly added to a −78 °C solution of (S)-6ae (20 mg, 0.083 mmol) in THF (1 mL) in a 10-mL RBF under argon. The reaction mixture was allowed to warm to 0 °C and then it was stirred at 0 °C for 2 h. Next, water (0.5 mL) was added dropwise to the reaction mixture at 0 °C to quench the reaction. Aqueous HCl (1.0 N; 0.5 mL) was added, and the mixture was extracted with EtOAc (3x5 mL). The combined organic extracts were dried over NaSO4, filtered, and concentrated. The residue was purified by flash chromatography on silica gel, which afforded the alcohol as colourless oil (20 mg

Multidimensional correlation analysis
Geometry optimizations and frequency calculations were carried out using Gaussian 16. 21 Vibrational frequencies and intensities were calculated at the M06-2X/6-31G(d) level of theory. Sterimol values B1, B5, and L were calculated for the M06/2X optimized geometries using Molecular Modeling Pro®. 22 Buried volumes V% were evaluated using SambVca 2.0. 23 Multidimensional regression analyses were performed using Matlab®. 24 Parameters for the ligands were calculated from the whole structure as truncated versions provided biased structural features. The graphical description of the parameters acquired is reported in Figure S1. The values of the parameters collected for ligands 3a-3x are reported in Table S1 together with the selectivity they have shown in our benchmark reaction (see Reaction optimization section). Here, the selectivity is expressed as ΔΔG ‡ in kcal/mol. ΔΔG ‡ values were calculated using the formula ΔΔG ‡ = -RTln(er) where R is the gas constant, T is temperature, and er is the enantiomeric ratio.

S18
Preliminary single parameter correlations showed that POas was the only parameter providing a trend with the observed enantioselectivity. However, a good correlation (R 2 =0.85) could be obtained only by removing structurally biased ligands ( Figure S2). This suggests that the selectivity is due to both electronic and geometrical/steric features of the ligand.
Therefore, we reasoned that a comprehensive model could be obtained by multidimensional correlation analysis.

Model development
The

Kinetic studies
Evaluation of the initial rates for our benchmark reaction in the presence of different amounts of ligand 3a was accomplished by reaction NMR profiling as follow: In a N2-filled glove box, an oven dried vial was charged with (CuOTf)2.Tol (2.5 mol%) and a magnetic stir bar. The vial was sealed with a ptfe septum screw cap and was removed from the glove box. Ligand 3a (x mol%, see and placed into the NMR instrument. NMR spectra were recorded every 5 min with the first spectrum being recorded after exactly 10 min from the addition of 4a. Concentration of the product 6a against time are reported below for a set of 5 experiments with [3a] ranging from 6 to 30 mol% (Table S2). In Figure S3 the reaction profiles are plotted. Using the first 4 experimental points (from 10 to 25 min) for each profile, an initial rate kobs was determined by linear fitting with Excel. The plot of the reciprocal 1/kobs against shows the kinetic order of 3a to be -1.