A New Mechanism of Metal-Ligand Cooperative Catalysis in Transfer Hydrogenation of Ketones

Abstract

We report iridium catalysts IrCl(η⁵-Cp*)(κ²-(2-pyridyl)CH₂NSO₂C₆H₄X) (1-Me, X = CH₃ and 1-F, X = F) for transfer hydrogenation of ketones with 2-propanol that operate by a previously unseen metal-ligand cooperative mechanism. Under the reaction conditions, complexes 1 (1-Me and 1-F) derivatize to a series of catalytic intermediates: Ir(η⁵-Cp*)(κ²-(C₅H₄N)CHNSO₂Ar) (2), IrH(η⁵Cp*)(κ²-(2-pyridyl)CH₂NSO₂Ar) (3), and Ir(η⁵-Cp*)(κ³-(2-pyridyl)CH₂NSO₂Ar) (4). The structures of 1-Me and 4-Me were established by single-crystal X-ray diffraction. A rate-determining, concerted hydrogen transfer step (2 + R₂CHOH ⇄ 3 + R₂CO) is suggested by kinetic isotope effects, Eyring parameters (ΔH^≠ = 29.1(8) kcal mol^–1 and ΔS^≠ = −17(19) eu), proton-hydride fidelity, and DFT calculations. According to DFT, a nine-membered cyclic transition state is stabilized by an alcohol molecule that serves as a proton shuttle.

1. Introduction

Substrate activation via dual-site metal-ligand cooperation is a thought-provoking area in catalysis,^1–5 particularly because such mechanisms enable reactivity, selectivity, and tunability unavailable to metals alone: for example, asymmetric hydrogenation of unsaturated polar bonds, where cooperation between a ruthenium center and an amido ligand results in dihydrogen splitting and delivery of the proton and hydride to a substrate.^6–8 A different mechanism of cooperative hydrogen transfer was discovered by Milstein et. al.,⁹ who showed reversible dearomatization of pyridine-based pincer complexes A (Figure 1).

Figure 1.

Hydrogen transfer catalysts.

The community generally accepts that the Milstein PNP and PNN pincer complexes frequently operate via a metal-ligand cooperative mechanism, and the role of the ligand backbone in catalysis is being studied actively in these and related hydrogenation systems.^10–13 However, controversy surrounds many bifunctional mechanism proposals,^13–16 including documented cases of non-cooperative, metal-centered catalysis by complexes in which cooperation was designed or expected.^17–23 Our group previously reported a family of iridium- and ruthenium-based hydrogen transfer catalysts B–D (Figure 1), that employ PN and NC CH-acidic bidentate ligands.^24–27 Contrary to their original design and unlike their pincer congeners, B–D do not operate by the expected cooperative catalysis mechanism, and their deprotonated/dearomatized forms do not participate in any catalytic reaction that we have found.

A series of complexes based on isolobal Cp*Ir^III and (p-cymene)Ru^II fragments and bidentate pyridyl-sulfonamide ligands have demonstrated a high catalytic activity in transfer hydrogenation of quinolinium salts,²⁸ imines,²⁹ aldehydes, and ketones.³⁰ Here we describe two complexes of this class, 1-Me and 1-F (Figure 1), that enable transfer hydrogenation of ketones with 2-propanol, via metal-ligand cooperation, through reversible dearomatization of the pyridine moiety. The iridium atom and dearomatized N-picolylbenzenesulfonamide ligand cooperate in a concerted abstraction of C–H and O–H hydrogen atoms: this contrasts the PNN Milstein complexes in which chemistry happens more readily on the phosphorus side of the pincer. Since the concerted nature of proton-hydride transfer in Milstein-type catalytic systems is still an open question,¹³ and since B–D do not utilize the cooperative mechanism, we wanted to characterize our catalytic system with experimental and computational data.

2. Results and Discussion

2.1. Synthesis and Reactivity of Complexes 1

In this section we describe the synthesis of complexes 1 (1-Me and 1-F) and a reactivity study that revealed a series of catalytically relevant species 2, 3, and 4 (Schemes 1 and 2). We began this project by exploring the reactivity of 1-Me, but later we made a fluorinated analog 1-F for the purpose of ¹⁹F NMR studies (Scheme 1). These complexes were synthesized with high yields (> 90%) in respective reactions among pentamethylcyclopentadienyliridium(III) chloride [Cp*IrCl₂]₂, the corresponding N-(picolyl)-benzenesulfonamide ligand H[NN^X] (X = Me, F), and triethylamine. Complexes 1 are air and water tolerant, therefore an aqueous workup is efficient and convenient. The structure of 1-Me was established by single-crystal X-ray diffraction (Scheme 1). Both compounds are analogous in reactivity and spectral properties.

Scheme 1.

Synthesis of 1-Me and 1-F. Molecular structure of 1-Me shown with 50% probability ellipsoids.

Scheme 2.

Formation of catalytically competent complexes 2, 3, and 4. Molecular structure of 4-mMe shown with 50% probability ellipsoids.

We demonstrated temperature- and solvent-dependant dynamic behaviour of the CH₂ protons in complex 1-Me, which is a consequence of stereoinversion of the iridium atom enabled by reversible chloride dissociation. The rate of the stereoinversion correlates with the solvent dielectric constant; for example, according to ¹H NMR, in CD₂Cl₂ (ε = 9.1) the diastereotopic protons appear as two broad peaks at 4.76 and 4.52 ppm, whereas in CD₃OD (ε = 33) they give a sharp singlet at 4.56 ppm. We conducted a variable temperature NMR experiment in CD₂Cl₂ and (CD₃)₂CO/CD₃OD (5 : 1 by volume) solutions and found that while the coalescence temperature of the methylene peaks in CD₂Cl₂ is way above its boiling point, in (CD₃)₂CO/CD₃OD the peaks coalesce at −12 °C, which corresponds to the rate constant of iridium inversion k_c = 512 s^–1 and the free energy of activation ΔG^≠ = 12 kcal mol^–1 (Figure S28).

Complexes 1 undergo deprotonation at the methylene site by t-BuOK to give dark-blue solutions of the corresponding complexes 2 (Scheme 2). These compounds are moisture-, air-, and light-sensitive, and our isolation attempts led to their decomposition. Nevertheless, complexes 2 were characterized by ¹H NMR in CD₂Cl₂ solution, and their identity was established based on the characteristic singlets of the backbone CH fragment at 8.11 (1-Me) and 8.12 (1-F) ppm. This assignment is consistent with the COSY data that suggest spin isolation of the CH fragment (Figures S5 and S6). The deprotonation also effects the pyridine protons H³ and H⁴, whose peaks are shifted downfield to the olefin region and appear at 5.80 and 6.41 ppm, respectively, thus indicating the loss of pyridine aromaticity. Analogous dearomatization of the pyridine system was observed previously for the pincer complexes⁹ and our dehydrogenation catalysts C and D.^24,25 Complexes 2 are convenient precursors for the synthesis of catalytically relevant compounds 3 and 4 (Scheme 2). The hydride complexes 3 can be prepared by reacting the corresponding complex 2 with formic acid in CH₂Cl₂. According to ¹H NMR, the chemical shifts of the hydride ligands in 3-Me and 3-F (–10.39 and −10.36 ppm, respectively) illustrate a negligible effect of the methyl or fluorine group on the metallic catalytic center. IR spectra of 3-Me and 3-F are consistent with the Ir–H bond, as they demonstrate metal-hydride vibration bands at 2100 and 2118 cm^–1, respectively. We find that species 2 undergoes a slow transformation to 4 in a CH₂Cl₂ solution, and the process is accelerated by direct sunlight (Scheme 2). Single-crystal X-ray diffraction study suggests that 4-Me features an aryl–iridium bond that originates from phenylene metalation. Although the mechanistic details of this transformation are unknown, the intramolecular C–H activation is selective over intermolecular processes even in benzene or toluene solutions. Alternatively, 4-F can be accessed from the hydride complex 3-F by heating it with acetophenone and 2-propanol without a base (Scheme 2). Interestingly, 3-F does not transfer its hydride to acetophenone in the absence of added 2-propanol, and without the ketone that serves as a hydrogen acceptor, the conversion of 3-F reaches only 53%.

We performed a series of catalytic tests to compare the activity of complexes 1, 3, and 4, and to evaluate the effect of added base (Table 1). Entries 2, 6, and 7 show that complexes 1, 3, and 4 are equally effective in transfer hydrogenation of acetophenone with 2-propanol, since they all give the same yield of 1-phenylethanol within the error of experiment. Entry 3 shows that air does not deactivate the catalyst but slightly reduces the reaction yield to 85%. Entries 1, 2, 4, and 5 demonstrate that transfer hydrogenation by 1-Me is enabled by general base catalysis, since the base strength correlates with the acetophenone conversion. Importantly, bases that cannot deprotonate complexes 1 inhibit the reaction. In contrast to our catalytic system, a homolog of 1-Me with an ethylene group in the ligand backbone, reported by O’Connor et al.,^30d shows a different reactivity in transfer hydrogenation of ketones with 2-propanol: the complex facilitates the base-free reaction and is deactivated by a base.

Table 1.

Catalyst and Base Screening.^a

Entry	[Ir]	Base, conditions	Yield, %b
1	1-Me	NaH	95
2	1-Me	t-BuOK	92
3	1-Me	t-BuOK, air	85
4	1-Me	K2CO3	38
5	1-Me	CH3COOCs	10
6	3-Me	t-BuOK	88
7	4-F	t-BuOK	88

^aReaction conditions: acetophenone (60 mg, 0.500 mmol), 2-propanol (0.5 mL), a catalyst (5.00 μmol, 1 mol%), and a base (25.00 μmol, 5 mol%) were stirred in a closed reactor under nitrogen at 85 °C for 14 h.

^bThe yields of 1-phenylethanol were calculated from ¹H NMR spectra of the reaction samples.

We tracked speciation of precatalyst 1-F in 2-propanol–acetophenone transfer hydrogenation by variable temperature in situ ¹⁹F NMR study (Figure 2). The peaks of 2-F, 3-F, and 4-F were identified by spiking the reaction mixture with the pure compounds. When 1-F is treated with 2-propanol, acetophenone, and t-BuOK at room temperature the reaction mixture temporarily turns dark blue due to formation of 2-F, and then slowly turns yellow. At this point, the corresponding ¹⁹F NMR spectrum contains four peaks at −109.90, −110.40, −110.60, and −111.26 (3-F) ppm (Figure 2A). The hydride 3-F is the only form that persists during the reaction. When the mixture is brought to 85 °C, the catalytic transfer hydrogenation begins, the mixture turns dark blue again, and the peak of 2-F appears at −107.40 ppm (Figure 2B). Generation of 2-F in the catalytic mixture is temperature dependent, since it exists as a minor intermediate during the catalysis at 85 °C and disappears at room temperature (Figure 2C and 2D). Complex 4-F was detected as a minor catalytic species at −112.44 ppm. While the spectra contain signals that we were unable to isolate, we see that they equilibrate with the known species, and MALDI data show that these are ligated forms of the [Cp*Ir(κ²-(2-pyridyl)CH₂NSO₂C₆H₄F)]⁺ cation.

Figure 2.

¹⁹F NMR spectra of iridium species derived from 1-F in the reaction between 2-propanol and acetophenone. Species of unknown structure are marked with asterisks.

2.2. Kinetic Studies

Complexes 2 are the most reactive among the observed catalytic intermediates, and therefore we believe that the active catalyst systems involve these species. Our further kinetic and theoretical studies support this hypothesis.

Standard homogeneity tests,^31,32 i.e., mercury drop test, quantitative poisoning, and well-behaved kinetics suggest that transfer hydrogenation of acetophenone with 1 and 2-propanol proceeds via homogeneous catalysis, and therefore, the system is suitable for the NMR kinetic studies (Figure S21). We endeavored to learn the kinetic isotope effect (KIE) values for CH and OH groups of 2-propanol and the activation parameters (ΔH^≠ and ΔS^≠) to get an insight to the rate-determining step of the catalytic mechanism.

We measured the rates of the reaction under identical conditions (acetophenone, 1-Me, and t-BuOK) to enable KIE determinations among five 2-propanol isotopologues ((CH₃)₂CHOH, (CD₃)₂CHOH, (CD₃)₂CDOH, (CD₃)₂CHOD, and (CD₃)₂CDOD). Kinetic isotope effects were derived from the obtained rate constant values (Table 1). The combined isotope effect (k_CHOH/k_CDOD = 2.15(3)) matches the product of the average separate OH and CH isotope effects 2.16(6). Secondary KIE was determined by comparing the rate constants of (CH₃)₂CHOH and (CD₃)₂CHOH, k_CH3/k_CD3 = 1.69(18)/1.38(2) = 1.23(13). These combined data are consistent with a mechanism where the hydride and proton transfer occur in a single kinetically relevant step.

Proton-hydride fidelity in our catalytic system was established based on the ¹H NMR data obtained during the KIE study (Figure S18). Analysis of the deuteration sites in 1-phenylethanol product showed full deuteration of the CH site only when a C-deuterated isotopologue, (CD₃)₂CDOD or (CD₃)₂CDOH, is used. We also note that C-deuteration is independent of O-deuteration in the alcohol. Thus the absence of proton-hydride scrambling suggests a concerted proton-hydride abstraction.

An Eyring study was conducted to determine the activation parameters of 2-prpanol–acetopheone transfer hydrogenation. Variable temperature ¹H NMR kinetics were recorded to determine four rate constants within 50–90 °C (Figures S19 and S20). The enthalpy and entropy of activation were derived using Eyring equation and were found to be ΔH^≠ = 29.1(8) kcal mol^–1 and ΔS^≠ = −17(19) eu. The strongly negative ΔS^≠ value indicates an associative event in the rate-determining step.

2.3. Computational Study

We employed hybrid, dispersion-corrected DFT, and reaction path search methods to determine the structure of the transition state in the rate-determining step of 2-propanol–acetophenone transfer hydrogenation. Our DFT model replaces 2-isopropanol, Cp*, and Ts groups with H₂O, Cp, and Tf, respectively, to simplify the calculations. We considered two potential transition states TS and TS’: the former contains a nine-membered ring with an extra alcohol molecule acting as a proton shuttle (Scheme 3), whereas in the latter the hydrogen transfer happens directly between 2-propanol and the metal complex (Figure S23). The computed enthalpy of activation for TS is ΔH^≠ = 26.2 kcal mol⁻¹, which is consistent with the enthalpy of activation obtained in the Eyring study, ΔH^≠ = 29.1(8) kcal mol⁻¹. The other transition state TS’ has a much higher enthalpy of activation, ΔH^≠ = 42.6 kcal mol⁻¹, which is energetically unrealistic and in discord with the experiment. Therefore, we conclude that TS could be relevant to the rate-determining step, and that the catalyst most likely operates via a proton shuttle mechanism. Similar proton shuttles have been invoked in hydrogenation reactions of other bifunctional complexes.^33,34

Scheme 3.

Rate-determining step and its transition state.

3. Conclusions

In conclusion, we present the first example of bidentate transition metal complexes, 1-Me and 1-F, that operate a via metal-ligand cooperative mechanism previously known only in pincer scaffolds only. The structurally-novel iridium complexes supported by bidentate N-picolylbenzenesulfonamide ligands undergo reversible deprotonation and formation of 2, which appears to be the active catalyst. Kinetic isotope effect data and proton-hydride fidelity are consistent with the concerted nature of the hydrogen atom transfer from 2-propanol to 2. Eyring study and DFT calculations provided consistent activation parameter values, and the optimized structure of the nine-membered cyclic transition state, TS, which is stabilized by a second alcohol molecule that shuttles the proton from the alcohol to the catalyst.

4. Experimental

4.1. General Procedures

All synthetic procedures were performed under nitrogen either in a Vacuum Atmospheres glovebox or in a closed reactor. Acetophenone, 4-toluenesulfonyl chloride, 4-fluorobenzenesulfonyl chloride, 2-picolylamine, [Cp*IrCl₂]₂, potassium tert-butoxide, formic acid (97%), and NaHCO₃ were purchased from common vendors and used without further purification. NMR solvents (CD₂Cl₂, CDCl₃, and C₆D₆), 2-propanol, and triethylamine were dried and distilled over CaH₂. Hexane and CH₂Cl₂ were dried using a solvent purification system.

¹H, ¹³C, and ¹⁹F NMR spectra were acquired on Varian Mercury 400, VNMRS-500, and VNMRS-600 spectrometers and processed using MestReNova 12.0.1. All chemical shifts are reported in ppm and referenced to the residual ¹H or ¹³C solvent peaks. Following abbreviations are used: (s) singlet, (bs s) broad singlet, (d) doublet, (t) triplet, (dd) doublet of doublets, etc. NMR spectra of all metal complexes were taken in 8” J. Young tubes (Wilmad or Norell) with Teflon valve plugs. MALDI-MS spectra were acquired on Bruker Autoflex Speed MALDI Mass Spectrometer. X-ray crystallography data were obtained on a Bruker APEX DUO single-crystal diffractometer equipped with an APEX2 CCD detector, Mo fine-focus and Cu micro-focus X-ray sources. IR spectra were obtained using Jasco FT/IR-4600 FT-IR Spectrometer.

4.2. Catalytic Procedure

A solution of acetophenone (60 mg, 0.500 mmol), complex 1 (5.00 μmol, 1 mol%) and a base (25.00 μmol, 5 mol%, t-BuOK, KOH, NaH, K₂CO₃, or K₃PO₄) in isopropanol (0.5 mL) was stirred in a closed reactor under nitrogen at 85 °C.

4.3. Synthesis of Ligands

N-Picolylbenzenesulfonamide ligands H[NN^Me] and H[NN^F] were synthesized in good yields 85–90% according to a reported procedure by Prim et al.³⁵

H[NN^F]:

Colorless crystals. ¹H NMR (400 MHz, CDCl₃): δ 8.41 (d, J = 4.7 Hz, 1H, Py), 7.82 (dd, J = 9.0, 5.1 Hz, 2H, C₆H₄), 7.59 (td, J = 7.7, 1.7 Hz, 1H, Py), 7.20 −7.11 (m, 2H, Py), 7.07 (t, J = 8.6 Hz, 2H, C₆H₄), 6.52 (br s, 1H, NH), 4.25 (d, J = 5.0 Hz, 2H, CH₂). ¹³C{¹H} NMR (101 MHz, CDCl₃): δ 164.99 (d, J = 254.5 Hz), 154.87, 149.02, 137.07, 136.02 (d, J = 3.4 Hz), 129.97, 129.88, 122.82, 122.20, 116.31, 116.09, 47.50. ¹⁹F NMR (564 MHz, CDCl₃): δ −105.60. IR (KBr, cm⁻¹): 3075 (br), 2867 (br), 1591, 1492, 1335, 1165, 839, 552.

4.4. Synthesis 1-Me and 1-F

[Cp*IrCl₂]₂ (100 mg, 0.125 mmol), ligand (H[NN^Me] or H[NN^F], 0.250 mmol), and triethylamine (38 mg, 0.375 mmol) were stirred with CH₂Cl₂ (5 mL) for 3 hours at room temperature. Then, all volatiles were removed in vacuum, the residue was suspended in hexanes (5 mL) and filtered. The solid was washed with hexane (5 mL) and then with water (3 × 5 mL). The yellow solid was dried on the filter and then crystallized from CH₂Cl₂/hexanes to give the product.

1-Me:

Orange crystals (156 mg, 93%). Crystals suitable for X-ray analysis were obtained by slow evaporation of CH₂Cl₂/hexane solution. ¹H NMR (600 MHz, CD₂Cl₂): δ 8.51 (d, J = 5.2 Hz, 1H, Py), 7.79 (d, J = 8.2 Hz, 2H, C₆H₄), 7.70 (td, J = 7.7, 1.5 Hz, 1H, Py), 7.26 (t, J = 6.9 Hz, 1H, Py), 7.21 (d, J = 7.9 Hz, 1H, Py), 7.10 (d, J = 8.0 Hz, 2H, C₆H₄), 4.76 (br s, 1H, CH₂), 4.52 (br s, 1H, CH₂), 2.30 (s, 3H, Me), 1.69 (s, 15H, 5Me). ¹³C{¹H} NMR (101 MHz, CD₂Cl₂): δ 164.66, 151.57, 141.08, 140.91, 138.34, 128.98, 128.40, 124.97, 120.68, 86.82, 58.10, 21.40, 9.65. IR (KBr, cm¹): 1569, 1474, 1447, 1402, 1380, 1355, 1275, 1225, 1134, 1104, 1084, 1056, 1031, 1001, 961, 932, 816, 768, 723, 709, 661, 620, 591, 556, 536, 506, 485, 482, 477, 468, 452. MALDI-MS: m/z calcd for [C₂₃H₂₈IrN₂O₂S]⁺ 589.15, found 589.32.

1-F:

Yellow crystals (141 mg, 90%). ¹H NMR (600 MHz, CD₂Cl₂): δ 8.51 (d, J = 6.4 Hz, 1H, Py), 7.97 (dd, J = 9.0, 5.4 Hz, 2H, C₆H₄), 7.71 (td, J = 7.7, 1.5 Hz, 1H, Py), 7.27 (t, J = 7.3 Hz, 1H, Py), 7.22 (d, J = 8.5 Hz, 1H, Py), 6.96 (t, J = 8.9 Hz, 2H, C₆H₄), 4.78 (d, J = 16.9 Hz, 1H, CH₂), 4.47 (d, J = 18.0 Hz, 1H, CH₂), 1.69 (s, 15H, 5Me). ¹³C{¹H} NMR (151 MHz, CD₂Cl₂): δ 164.17 (d, J = 249.0 Hz), 164.39, 151.56, 140.08 (d, J = 2.9 Hz), 138.45, 131.02 (d, J = 8.7 Hz), 125.08, 120.74, 115.15 (d, J = 22.0 Hz), 86.93, 57.98, 9.65. ¹⁹F NMR (564 MHz, CD₂Cl₂): δ −111.48. IR (KBr, cm⁻¹): 1591, 1497, 1277, 1136, 940, 667, 558. MALDI-MS: m/z calcd for [C₂₂H₂₅FIrN₂O₂S]⁺ 593.13, found 593.35.

4.5. Synthesis of 2-Me and 2-F

Complex 1 (0.080 mmol) and potassium tert-butoxide (45 mg, 0.400 mmol) were stirred with CD₂Cl₂ (1.0 mL) for 20 minutes to give a dark blue solution of 2. It was filtered through a PTFE syringe filter and then analyzed by ¹H NMR. Complex 2 is moisture-, air-, and light-sensitive, therefore any isolation attempt led to decomposition.

2-Me:

¹H NMR (600 MHz, CD₂Cl₂): δ 8.25 (d, J = 6.8 Hz, 1H, Py), 8.11 (s, 1H, CH), 7.43 (d, J = 7.7 Hz, 2H, C₆H₄), 7.37 (d, J = 8.9 Hz, 1H, Py), 7.20 (d, J = 7.8 Hz, 2H, C₆H₄), 6.40 (d, J = 6.7 Hz, 1H, Py), 5.80 (t, J = 6.6 Hz, 1H, Py), 2.38 (s, 3H, Me), 1.78 (s, 15H, 5Me).

2-F:

¹H NMR (600 MHz, CD₂Cl₂): δ 8.25 (d, J = 7.1 Hz, 1H, Py), 8.12 (s, 1H, CH), 7.59 – 7.52 (m, 2H, C₆H₄), 7.39 (d, J = 9.0 Hz, 1H, Py), 7.12 – 7.05 (m, 2H, C₆H₄), 6.42 (ddd, J = 9.0, 6.2, 1.1 Hz, 1H, Py), 5.81 (t, J = 7.4 Hz, 1H, Py), 1.78 (s, 15H, 5Me). ¹⁹F NMR (564 MHz, CD₂Cl₂): δ – 107.79.

4.6. Synthesis of 3-Me and 3-F

Complex 1 (0.160 mmol) and potassium tert-butoxide (90 mg, 0.801 mmol) were stirred with CH₂Cl₂ (5 mL) for 20 minutes to give a dark blue solution of 2. Then, 2 drops of formic acid were added to cause a sharp color change to orange. The mixture was stirred for another 10 minutes with NaHCO₃ (50 mg) and then filtered through a PTFE syringe filter. The resulting solution was dried under vacuum and crystallized twice from CH₂Cl₂/hexanes to afford the product.

3-Me:

Pale-yellow crystals (89 mg, 95%). ¹H NMR (600 MHz, CD₂Cl₂): δ 8.35 (d, J = 5.7 Hz, 1H, Py), 7.55 (d, J = 8.2 Hz, 2H, C₆H₄), 7.53 (td, J = 7.7, 1.5 Hz, 1H, Py), 7.20 (d, J = 7.8 Hz, 1H, Py), 6.99 (d, J = 7.9 Hz, 2H, C₆H₄), 6.91 (t, J = 6.7 Hz, 1H, Py), 4.76 (d, J = 17.2 Hz, 1H, CH₂), 4.43 (d, J = 17.2 Hz, 1H, CH₂), 2.26 (s, 3H, Me), 1.81 (s, 15H, 5Me), −10.39 (s, 1H, IrH). ¹³C{¹H} NMR (151 MHz, CD₂Cl₂): δ 166.07, 152.71, 141.33, 140.20, 136.51, 128.48, 127.86, 123.46, 119.79, 87.73, 61.60, 21.34, 10.19. IR (KBr, cm⁻¹): 2907 (br), 2100 (ν_IrH), 1449, 1289, 1137, 1092, 667, 558. MALDI-MS: m/z calcd for [C₂₃H₂₉IrN₂O₂S]⁺ 590.16, found 589.3.

3-F:

Pale-yellow crystals (71 mg, 75%). ¹H NMR (600 MHz, CD₂Cl₂): δ 8.34 (d, J = 5.7 Hz, 1H, Py), 7.73 – 7.65 (m, 4H, C₆H₄), 7.53 (td, J = 7.7, 1.5 Hz, 1H, Py), 7.20 (d, J = 7.8 Hz, 1H, Py), 6.91 (t, J = 6.6 Hz, 1H, Py), 6.84 (t, J = 8.9 Hz, 2H, C₆H₄), 4.79 (d, J = 17.2 Hz, 1H, CH₂), 4.42 (d, J = 17.2 Hz, 1H, CH₂), 1.81 (s, 15H, 5Me), −10.36 (s, 1H, IrH). ¹³C{¹H} NMR (151 MHz, CD₂Cl₂): δ 165.69, 163.67 (d, J = 248.5 Hz), 152.87, 140.43, 136.65, 130.30 (d, J = 8.6 Hz), 123.53, 119.79, 114.61 (d, J = 22.0 Hz), 87.78, 61.54, 10.19. ¹⁹F NMR (564 MHz, CD₂Cl₂): δ – 112.03. IR (KBr, cm⁻¹): 2118 (ν_IrH), 1592, 1495, 1284, 1137, 670, 543.

4.7. Synthesis of 4-Me and 4-F

Complex 1 (0.160 mmol) and potassium tert-butoxide (90 mg, 0.801 mmol) were stirred with CH₂Cl₂ (6 mL) for 30 minutes to give a dark blue solution of 2. Then, the reaction was continued outside glovebox under direct sunlight for one hour. After the mixture turned red it was filtered through a PTFE syringe filter. The resulting solution was dried under vacuum and crystallized twice from CH₂Cl₂/hexanes to give an off-white crystalline material.

4-Me:

Pale-yellow powder (75 mg, 80%). Crystals suitable for X-ray analysis were obtained by slow evaporation of CH₂Cl₂/pentane solution. ¹H NMR (600 MHz, CD₂Cl₂): δ 8.27 – 8.23 (m, 1H), 8.11 (s, 1H), 7.47 – 7.40 (m, 2H), 7.38 (dt, J = 8.9, 1.4 Hz, 1H), 7.20 (d, J = 8.0 Hz, 2H), 6.41 (dd, J = 9.0, 6.2 Hz, 1H, CH₂), 5.80 (dd, J = 7.4, 6.2 Hz, 1H, CH₂), 2.38 (s, 3H, Me), 1.78 (s, 15H, 5Me). ¹³C{¹H} NMR (101 MHz, CD₂Cl₂): δ 166.92, 149.87, 146.58, 145.59, 139.72, 136.87, 136.00, 124.07, 124.02, 122.02, 119.64, 88.43, 60.57, 21.03, 8.63. IR (KBr, cm⁻¹): 2914, 1341, 1259, 1215, 1148, 1124, 1086, 849, 833, 802, 764, 728, 683, 664, 617, 601, 583, 559, 541, 510, 483, 464, 456. MALDI-MS: m/z calcd for [C₂₃H₂₇IrN₂O₂S]⁺ 588.14, found 587.3.

4-F:

Pale-yellow crystals (68 mg, 72%). ¹H NMR (600 MHz, CD₂Cl₂): δ 8.29 (d, J = 5.2 Hz, 1H, ArH), 7.58 (td, J = 7.7, 1.4 Hz, 1H, ArH), 7.26 – 7.12 (m, 3H, ArH), 7.03 (t, J = 6.6 Hz, 1H, ArH), 6.57 (td, J = 8.8, 2.5 Hz, 1H, ArH), 5.11 (d, J = 18.3 Hz, 1H, CH₂), 4.41 (d, J = 18.3 Hz, 1H, CH₂), 1.73 (s, 15H, 5Me). ¹³C{¹H} NMR (151 MHz, CD₂Cl₂): δ 167.25, 163.67 (d, J = 250.3 Hz), 150.38, 149.37, 145.85, 137.67, 124.75, 124.53 (d, J = 8.7 Hz), 121.50 (d, J = 17.1 Hz), 120.27, 110.40 (d, J = 23.1 Hz), 89.19, 61.01, 9.06. ¹⁹F NMR (564 MHz, CD₂Cl₂): δ – 113.54. IR (KBr, cm⁻¹): 2921, 1557, 1267, 1144, 862, 549.