Generation of boryl-nitroxide radicals from a boraalkene via the nitroso ene reaction

Examples of isolated boron substituted nitroxide radicals are rare. The reaction of the reactive cyclic boraalkene 3 with nitrosobenzene yields a mixture of the [2 + 2] cycloaddition product 4a, the B-nitroxide radicals 5a and 6a and the azoxybenzene co-product 7avia a bora nitroso ene reaction pathway, the boron analogue of the nitroso ene reaction. The products were separated by flash chromatography, and the B-nitroxide radicals were characterized by X-ray diffraction and EPR spectroscopy. Radical 5a was shown to be a hydrogen atom abstractor. Both the B-nitroxide radicals are more easily oxidized compared to e.g. TEMPO, as shown by cyclic voltammetry.


S3
Materials: Unless otherwise noted, all chemicals were purchased from commercially available sources and used as received. Compounds 1 and 3 were prepared according to procedures descried in the literature. 5

Synthesis and characterization of compound 2
In a Young NMR tube, compound 1 (9.6 mg, 0.02 mmol, 1.0 equiv.) and PhNO (2.2 mg, 0.02 mmol, 1.0 equiv.) were mixed and C6D6 (0.6 mL) was added. After the resulting mixture was shaken at r.t. for 5 min., it was dried in vacuo. The remaining residue was washed with pentane (3 × 1 mL) and dried in vacuo, to give compound 2 as a pale-white solid (10.5 mg, 89%). Crystals suitable for the X-ray single crystal structure analysis of compound 2 were obtained from a solution of the obtained pale-white solid in C6D6/pentane (v/v < 1:5) at r.t.  Figure S4. 19  was used for the X-ray crystallographic analysis. The X-ray intensity data were measured on a Kappa CCD APEXII Bruker APEXII Diffractometer system equipped with a fine-focus sealed tube Cu sealed tube (CuKα, λ = 1.54178 Å) and a graphite monochromator. A total of 1649 frames were collected. The total exposure time was 18.69 hours. The frames were integrated with the Bruker SAINT software package using a wide-frame algorithm. The integration of the data using a monoclinic unit cell yielded a total of 35334 reflections to a maximum θ angle of 66.77° (0.84 Å resolution), of which 5118 were independent (average redundancy 6.904, completeness = 98.5%, Rint = 10.34%, Rsig = 5.99%) and 3320 (64.87%) were greater than 2σ(F 2 ). The final cell constants of a = 8.5951(4) Å, b = 13.5070(7) Å, c = 25.5281(10) Å, β = 99.612(3)°, volume = 2922.1(2) Å 3 , are based upon the refinement of the XYZ-centroids of 3651 reflections above 20 σ(I) with 7.024° < 2θ < 131.3°. Data were corrected for absorption effects using the Multi-Scan method (SADABS). The ratio of minimum to maximum apparent transmission was 0.864. The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.8880 and 0.9660. The structure was solved and refined using the Bruker SHELXTL Software Package, using the space group P21/c, with Z = 4 for the formula unit, C33H27BF5N3O. The final anisotropic fullmatrix least-squares refinement on F 2 with 394 variables converged at R1 = 5.21%, for the observed data and wR2 = 14.53% for all data. The goodness-of-fit was 1.026. The largest peak in the final difference electron density synthesis was 0.307 e -/Å 3 and the largest hole was -0.270 e -/Å 3 with an RMS deviation of 0.068 e -/Å 3 . On the basis of the final model, the calculated density was 1.335 g/cm 3 and F(000), 1216 e -. CCDC Nr.: 2157861. S9 Figure S6. Crystal structure of compound 2 (thermal ellipsoids: 30% probability). S10 2. Synthesis and characterization of compounds 4a, 5a and 6a 1 st Experiment: Compound 3 (28.3 mg, 0.05 mmol, 1.0 equiv.) and PhNO (10.7 mg, 0.1 mmol, 2.0 equiv.) were mixed and C6D6 (0.5 mL) was added. The resulting mixture was stirred at r.t. for 1h and directly used for 1 H, 19 F, and 11 B NMR experiments ( Figure S7-S9).
(2) Deeply red crystals suitable for the X-ray single crystal structure analysis of compound 5a were obtained from a solution of the obtained brown solid in CH2Cl2/pentane at r.t.
(3) Orange to red crystals suitable for the X-ray single crystal structure analysis of compound 6a were obtained from a solution of the obtained brown solid in CH2Cl2/pentane at r.t.  Figure S8. 19

UV-vis spectra of compounds 5a and 6a
Absorption spectra were recorded on a JASCO V-730 two-beam UV/Vis spectrometer.

Cyclic voltammograms of compounds 5a and 6a
The Cyclic voltammograms were recorded on a CHI600E Electrochemical Workstation.
Samples (compounds 5a, 6a, and TEMPO) were prepared in CH3CN solution (c: 1 × 10 -3 mol/L) with the Bu4NPF6 as additive (c: 1 ×10 -1 mol/L).  Table S1. g-factor and hyperfine coupling constants (in G and MHz) for 5a, yielding the optimum agreement with the experimental CW EPR spectrum and corresponding second derivative EPR spectrum (see Figure S29). EPR results: The liquid-state CW EPR spectrum in Figure S30 (left) shows a hyperfine splitting into several lines with a g-factor of 2.00604. The hyperfine splitting is caused by one 14 N nucleus (I = 1), one 11 B nucleus (I = 3/2) and the four 1 H nuclei (I = ½, 2:1:1) of the phenylene ring. Table S2. g-factor and hyperfine coupling constants (in G and MHz) for 6a, yielding the optimum agreement with the experimental CW EPR spectrum and the corresponding second derivative EPR spectrum (see Figure S30). The EPR spectra depicted in Figure S29-S30 (left) correspond to the first derivatives of the absorption EPR signal, because of the field modulation technique as applied here. It is known from literature that the resolution can be artificially enhanced by plotting the second derivative ( Figure S29-S30 (right)). 11,12,13,14,15 Thus, using a simultaneous simulation of the first and second derivates, the hyperfine coupling constants for the radicals 5a and 6a could be determined (Table S1 and S2). The differences in the EPR spectra of the radicals 5a and 6a are caused by the different coordination and additional bonding at the phenyl/phenylene ring. The calculated spin density via DFT in Figure S31a is significant at the oxygen where the radical is located and at the oxygen bonded nitrogen nucleus. Moreover, the spin density is higher at the nitrogen nucleus than at the boron nucleus, which correlates well with the differences in hyperfine coupling constants, cf. Table S1 and S2, that is, higher hyperfine coupling constants values are caused by a higher probability density of the electron at the nucleus. Furthermore, spin density is located at the conjugated phenyl ring, explaining the presence of hyperfine couplings to the abundant 1 H of the phenyl ring. In Figure S31b, the spin density located at the conjugated phenylene ring is a slightly higher, which correlates with the experimental determined hyperfine coupling constant for the two equivalent protons of 3.44 G (cf . Table S2).   was stirred at r.t. for 10 min and directly used for the EPR experiment (see Figure S35). Figure S35. Liquid-state CW EPR spectrum of the reaction mixture recorded in toluene at X-band and at room temperature.     impurities]

S42
Step 2: After that the solvent was removed in vacuo and the residue was extracted with pentane (5 mL 19 F, and 11 B NMR spectra were recorded (see Figure S44-S46).  Figure   S47). 20 Figure S47. Liquid-state CW EPR spectrum of the reaction mixture recorded in toluene at X-band and at room temperature. Table S4. g-factor and hyperfine coupling constants (in G) for the reaction mixture, yielding the optimum agreement with the experimental CW EPR spectrum.
[Comment: crystalline compound 9 has extremely poor solubility in C6D6, which would cause very slow reaction rate at r.t.] the obtained crude product (entries 1-6)