The Reactivity of Isomeric Nitrenium Lewis Acids with Phosphines, Carbenes, and Phosphide

Abstract Alkylation of spiro[fluorene‐9,3’‐indazole] at N(1) and N(2) with tBuCl affords the nitrenium cations [C6H4N2(tBu)C(C12H8)][BF4], 1 and 2, respectively. Compound 1 converts to 2 over the temperature range 303–323 K with a free energy barrier of 28±5 kcal mol−1. Reaction of 1 with PMe3 afforded the N‐bound phosphine adduct [C6H4N(tBu)N(PMe3)C(C12H8)]BF4] 3. However, phosphines attack 2 at the para‐carbon atom of the aryl group with concurrent cleavage of N(2)−C(1) bond and proton migration to C(1) affording [(R3P)C6H3NN(tBu)CH(C12H8)][BF4] (R=Me 4, nBu 5). Analogous reactions of 1 and 2 with the carbene SIMes prompt attack at the para‐carbon with concurrent loss of H. affording the radical cation salts [(SIMes)C6H3N(tBu)NC(C12H8).][BF4] 6 and [(SIMes)C6H3NN(tBu)C(C12H8).][BF4] 7, whereas reaction of 2 with BAC gives the Lewis acid‐base adduct, [C6H4N(BAC)N(tBu)C(C12H8)][BF4] 8. Finally, reactions of 1 and 2 with KPPh2 result in electron transfer affording (PPh2)2 and the persistent radicals C6H4N(tBu)NC(C12H8). and C6H4NN(tBu)C(C12H8).. The detailed reaction mechanisms are also explored by extensive DFT calculations.


General Remarks
All reactions and work-up procedures were performed under an inert atmosphere of dry, oxygen-free N 2 by means of standard Schlenk techniques or glovebox techniques (MBraun glovebox equipped with a -35 °C freezer) unless otherwise specified. All glassware was ovendried and cooled under vacuum before use. Dichloromethane (CH 2 Cl 2 ), 1,2-difluorobenzene (ODFB), and toluene were distilled over CaH 2 and tetrahydrofuran (THF) was distilled over Na/benzophenone. Pentane and hexane were collected from a Grubbs-type column system manufactured by Innovative Technology and degassed. Solvents were stored over activated 4 Å molecular sieves. Molecular sieves, type 4 Å (pellets, 3.2 mm diameter) purchased from Sigma Aldrich were activated prior to usage by iteratively heating under vacuum for 24 hours. CDCl 3 purchased from Cambridge Isotope Laboratories was vacuum distilled over CaH 2 . Unless otherwise mentioned, chemicals were purchased from Sigma Aldrich or TCI. Spiro[fluorene-9,3'-indazole], S1 1,3-Bis(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2-ylidene (SIMes), S2 and bis(diisopropylamino)cyclopropenylidene (BAC) S3 were prepared according to previously reported synthetic procedures. Tertbutyl chloride was degassed and stored over activated 4 Å molecular sieves in a Schlenk flask prior to use. NMR spectra were recorded at room temperature (298K) unless otherwise mentioned on a Bruker Avance III 400 MHz, an Agilent DD2 500, and an Agilent DD2 600 Spectrometers. Spectra were referenced to the residual solvent signals (CDCl 3 : 1 H= 7.26 ppm and 13 C = 77.2 ppm). Chemical shifts (δ) are reported in ppm and coupling constants (J) are listed as absolute values in Hz. Multiplicities are reported as singlet (s), doublet (d), triplet (t), quartet (q), multiplet (m), overlapping (ov), and broad (br). Electron paramagnetic resonance (EPR) measurements were performed at 298 K using a Bruker ECS-EMX X-band EPR spectrometer equipped with an EP4119HS cavity. Simulations were performed using PEST WinSIM software. High resolution mass spectrometry was performed in house employing electrospray ionisation techniques in positive ion mode on an AB/Sciex QStarXL mass spectrometer (ESI).

Compound 1 and 2
Tert-butyl chloride (165 µL, 1.50 mmol) was added dropwise to a dichloromethane solution of spiro[fluorene-9,3'-indazole] (335.4 mg, 1.25 mmol) and AgBF 4 (243.3 mg, 1.25 mmol). Upon addition, the solution turned to dark red with white precipitates and was allowed to stir at room temperature for 10 min. The suspension was filtered and the volatiles were removed under vacuum. The orange solid was washed with dichloromethane and toluene (v/v=1/4), dried under vacuum. The crude mixture was again dissolved in dichloromethane (2 mL), layer with pentane (0.8 mL), and stored at -35 o C overnight. Compound 1 was decanted, dried under vacuum, and obtained as bright orange crystals. To the remaining aliquot, 1 mL of pentane was added and the solution was stored at -35 o C overnight again. Compound 2 was decanted, dried under vacuum, and obtained as a dusty orange powder. If 1 and 2 still have a trace amount of impurities, further recrystallization from saturated CH 2 Cl 2 solution layering with pentane would yield pristine products. Figure S1: 1 H (CDCl 3 ) NMR spectrum of the crude mixture.

Compound 3
To a solution of 1 (41.2 mg, 0.10 mmol) in dichloromethane (3 mL), a THF solution of PMe 3 (1.0 M, 0.12 mL, 0.12 mmol) was added dropwise. The solution was allowed to stir at ambient temperature for 10 mins. All volatiles were removed in vacuo and the residue was washed with hexane (2 X 1 mL) and dried under vacuum. Compound 3 was obtained as a white powder. Single crystals suitable for X-ray diffraction were grown from liquid diffusion of pentane into a saturated dichloromethane solution at -35 °C. Yield: 46.3 mg (95% isolated yield). 1

Compound 4
To a solution of 2 (41.2 mg 0.1 mmol) in dichloromethane (3 mL), a THF solution of PMe 3 (1.0 M, 0.12 mL, 0.12 mmol) was added dropwise. The solution was allowed to stir at room temperature for 10 min. All volatiles were removed in vacuo and the residue was washed with hexane (2 X 1 mL) and dried under vacuum. Compound 4 was obtained as an orange powder. Crystals were grown from liquid diffusion of pentane into a saturated dichloromethane solution at room temperature. Yield: 44.1 mg (92% isolated yield). 1

Compound 5
To a solution of 2 (41.2 mg 0.1 mmol) in dichloromethane (3 mL), nBu 3 P (22.7 mg, 0.11 mmol) was added dropwise. The solution was allowed to stir at room temperature for 10 min. All volatiles were removed in vacuo and the residue was washed with hexane (2 X 1 mL) and dried under vacuum. Compound 5 was obtained as an orange powder. Single crystals suitable for X-ray diffraction were grown from liquid diffusion of pentane into a saturated dichloromethane solution at -35 °C. Yield: 241.3 mg (95% isolated yield). 1

Compound 6
To a solution of 1 (41.2 mg, 0.1 mmol) in ODFB (2 mL), a solution of SIMes (30.8 mg, 0.1 mmol) in ODFB was added dropwise. The solution was allowed to stir at ambient temperature for 10 min. All volatiles were removed in vacuo and the residue was washed with hexane (2 X 1 mL) and dried under vacuum. Compound 6 was obtained as a purple powder. Yield: 48.3 mg (67% isolated yield). Compound 6 decomposes readily in halogenated solvents. Attempts to observe the product by high-resolution mass spectrometry failed due to the instability of the compound under mass spectrometry conditions, by either ESI or DART methods. In positive mode ESI, [M-1] + and [M-2] + were observed. Figure S25: Experimental and simulated EPR spectra of 6. S17 Table S1.
Experimental and simulated EPR data of 6.  Scheme S1. proposed redox scheme of 6.

Compound 7
To a solution of 2 (41.2 mg, 0.1 mmol) in ODFB (2 mL), a solution of SIMes (30.8 mg, 0.1 mmol) in ODFB was added dropwise. The solution was allowed to stir at ambient temperature for 10 min. All volatiles were removed in vacuo and the residue was washed with hexane (2 X 1 mL) and dried under vacuum. Compound 7 was obtained as a red powder. Single crystals suitable for X-ray diffraction were grown from liquid diffusion of pentane into a saturated THF solution at -35 °C. Yield: 57.6 mg (80% isolated yield). Compound 7 decomposes readily in halogenated solvents. Attempts to observe the product by highresolution mass spectrometry failed due to the instability of the compound under mass spectrometry conditions, by either ESI or DART methods. In positive mode ESI, [M-1] + and [M-2] + were observed. S18 Figure S26: Experimental and simulated EPR spectra of 7.

Compound 8
To a solution of 2 (41.2 mg, 0.1 mmol) in ODFB (2 mL), a solution of BAC (24.0 mg, 0.1 mmol) in ODFB was added dropwise. The solution was allowed to stir at ambient temperature for 10 min. All volatiles were removed in vacuo and the residue was washed with hexane (2 X 1 mL) and dried under vacuum.

X-ray Diffraction Studies
Single crystals were coated with paratone oil, mounted on a cryoloop and frozen under a stream of cold nitrogen. Data were collected on a Bruker Apex2 X-ray diffractometer at 150(2) K for compound 1-3 and 5-7 crystals using graphite monochromated Mo-Kα radiation (0.71073 Å). Data were collected using Bruker APEX-2 software and processed using SHELX and an absorption correction applied using multi-scan within the APEX-2 program. Compound 4 was collected with a Cu microsurce and Bruker CMOS PHOTON II detector gave very weak data. However the structure obtained is of sufficient precision to allow us to identify and confirm the compound synthesised. All structures were solved and refined by  Table S3.

Kinetic data
Compound 1 dissolves in chloroform, CH 2 Cl 2 , THF, and ODFB. In order to be able to monitor the reactions by 1 H NMR spectroscopy, we chose affordable CDCl 3 as solvent.
Compound 1 was dissolved in CDCl 3 in a sealed NMR tube. The concentrations of 1 and 2 were monitored by 1 H NMR spectroscopy at 50 o C (323 K) over 175 mins. 1,2-Dichloroethane was used as the internal standard.  The initial experiment run to full consumption showed the kinetic profile of the formation of 2 to be linear until ca. 33 minutes ( Figure S31). As such, only aliquots taken in the first 33 minutes (n = 5) were analyzed for all parallel experiments from 303 to 323 K. 1 (ca. 4.4 mg) was dissolved in 0.6 mL of CDCl 3 in a sealed NMR tube. The concentration of 2 was monitored by 1 H NMR spectroscopy. 1,2-Dichloroethane (0.0399 M) was used as the internal standard. Figure S33: Concentrations of compound 2 under various temperatures. Table S4.
Rate constants derived from kinetic experiments ( Figure S33).  Figure S34: Eyring plot of data from table S4.

Observation of H 2 in reaction of 2 and SIMes
2 (10.0 mg, 0. 024mmol) was carefully transferred to a J-Young tube. A solution of SIMes (12.8 mg, 0.024 mmol) in C 6 D 6 (0.5 mL) was added. Due to the poor solubility of compound 2 in C 6 D 6 , the reaction preceded very slowly and the reaction was monitored at room temperature by 1 H NMR spectroscopy.

Electrochemistry
Cyclic voltammetry experiments performed using BASi-Epsilon RDE-2 model. A standard three-electrode cell configuration was employed using a glassy graphite working electrode, a platinum wire counter electrode, and a silver wire serving as a reference electrode. Formal redox potentials were referenced to the ferrocenium/ferrocene redox couple.

Computational details
The quantum chemical DFT calculations have been performed with the TURBOMOLE 7.3 suite of programs 1 The structures are fully optimized at the TPSS-D3/def2-TZVP + COSMO(CHCl 3 ) level of theory, which combines the TPSS meta-GGA density functional 2 with the BJ-damped DFT-D3 dispersion correction 3, 4 and the def2-TZVP basis set, 5, 6 using the Conductor-like Screening Model (COSMO) continuum solvation model 7 for CHCl 3 solvent (dielectric constant ε = 4.8 and solvent diameter R solv = 3.17 Å). The density-fitting RI-J approach 5,8,9 is used to accelerate the geometry optimization and numerical harmonic frequency calculations 10 in solution. The optimized structures are characterized by frequency analysis to identify the nature of located stationary points (no imaginary frequency for true minima and only one imaginary frequency for transition state) and to provide thermal corrections (at 298.15 K and 1 atm) according to the modified ideal gas−rigid rotor− harmonic oscillator model. 11 This choice of dispersion-corrected meta-GGA functional makes the efficient exploration of all potential reaction paths possible.
The final solvation free energies in CHCl 3 are computed with the COSMO-RS solvation model 12 (parameter file: BP_TZVP_C30_1601.ctd) using the COSMOtherm program package [13 on the above TPSS-D3 optimized structures, and corrected by +1.89 kcal•mol −1 to account for higher reference solute concentration of 1 mol•L −1 usually used in solution. To check the effects of the chosen DFT functional on the reaction energies and barriers, singlepoint calculations at the meta-GGA TPSS-D3 2 and hybrid-meta-GGA PW6B95-D3 14 levels are performed using a larger def2-QZVP basis set. 6,15 The final reaction Gibbs free energies (∆G) are determined from the electronic single-point energies plus TPSS-D3 thermal corrections and COSMO-RS solvation free energies. The computed reaction free energies from both DFT functionals are in good mutual agreement of -0.9 ± 1.9 for reaction free energies (average and standard deviations, see Table S1 below) despite 3.5 ± 2.2 kcal/mol higher barriers at PW6B95-D3 level as expected. In our discussion, higher-level PW6B95-D3 Gibbs free energies (in kcal/mol, at 298.15 K and 1 mol/L concentration) will be used in our discussion unless specified otherwise. Table S7.