Photocatalytic Oxygenation of Substrates by Dioxygen with Protonated Manganese(III) Corrolazine

Abstract

UV-vis spectral titrations of a manganese(III) corrolazine complex [Mn^III(TBP₈Cz): TBP₈Cz = octakis(p-tert-butylphenyl)corrolazinato³⁻] with HOTf in benzonitrile (PhCN) indicate mono- and diprotonation of Mn^III(TBP₈Cz) to give Mn^III(OTf)(TBP₈Cz(H)) and [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf] with the protonation constants of 9.0 × 10⁶ M⁻¹ and 4.7 × 10³ M⁻¹, respectively. The protonated site of Mn^III(OTf)(TBP₈Cz(H)) and [Mn^III(OTf)(H₂O)-(TBP₈Cz(H)₂)][OTf] were identified by X-ray crystal structures of the mono- and diprotonated complexes. In the presence of HOTf, the monoprotonated manganese(III) corrolazine complex [Mn^III(OTf)(TBP₈Cz(H))] acts as an efficient photocatalytic catalyst for the oxidation of hexamethylbenzene and thioanisole by O₂ to the corresponding alcohol and sulfoxide with 563 and 902 TON, respectively. Femtosecond laser flash photolysis measurements of Mn^III(OTf)(TBP₈Cz(H)) and [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf] in the presence of O₂ revealed the formation of a tripquintet excited state, which was rapidly converted to a tripseptet excited state. The tripseptet excited state of Mn^III(OTf)(TBP₈Cz(H)) reacted with O₂ with a diffusion-limited rate constant to produce the putative Mn^IV(O₂^•−)(OTf)(TBP₈Cz(H)), whereas the tripseptet excited state of [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf] exhibited no reactivity toward O₂. In the presence of HOTf, Mn^V(O)(TBP₈Cz) can oxidize not only HMB but also mesitylene to the corresponding alcohols, accompanied by regeneration of Mn^III(OTf)(TBP₈Cz(H)). This thermal reaction was examined for a kinetic isotope effect, and essentially no KIE (1.1) was observed for the oxidation of mesitylene-d₁₂, suggesting a proton-coupled electron transfer (PCET) mechanism is operative in this case. Thus, the monoprotonated manganese(III) corrolazine complex, Mn^III(OTf)(TBP₈Cz(H)), acts as an efficient photocatalytic for the oxidation of HMB by O₂ to the alcohol.

Graphical Abstract

Photocatalytic oxygenation of substrates by a manganese complex in the presence of HOTf has been studied. The protonated site of the mono- and diprotonated Mn^III complexes were characterized by single crystal XRD. In the presence of HOTf, the monoprotonated manganese(III) corrolazine complex acts as an efficient photocatalytic catalyst for the oxidation of hexamethylbenzene and thioanisole by O₂ whereas the diprotonated manganese(III) corrolazine complex exhibited no reactivity toward O₂

INTRODUCTION

Oxidation reactions are essential in the metabolism of organic substrates and many enzymes are known to catalyze these reactions in biological systems. High-valent metal-oxo complexes have been implicated as important reactive intermediates in these reactions for heme and non-heme iron enzymes.^1–14 To help clarify related oxidation mechanisms and to control the reactivity, many synthetic high-valent metal-oxo complexes have been prepared using oxidants such as iodosylarenes, peroxy acids, and hydrogen peroxide.^3–14 Dioxygen (O₂) is an ideal reagent for the production of these metal-oxo complexes, and for the subsequent oxygenation of substrates because of its abundant availability and nontoxicity.^15,16 The mechanisms of oxidation of substrates by high-valent metal-oxo complexes have been studied extensively.^17–23 High-valent manganese-oxo complexes have attracted special attention because they are postulated as important intermediates for water oxidation in the oxygen-evolving center (OEC) of photosystem II.^24–32

We have shown that a well-characterized manganese(V)-oxo complex can be prepared by visible light irradiation of a manganese(III) corrolazine [Mn^III(TBP₈Cz): TBP₈Cz³⁻ = octakis(p-tert-butylphenyl)corrolazinato³⁻] in the presence of toluene derivatives with dioxygen.^15,16 Hexamethylbenzene (HMB) was shown to be oxidized to pentamethylbenzyl alcohol during this reaction, serving as a proton/electron source to assist with O₂ activation. However, the produced Mn^V(O)(TBP₈Cz) was unreactive toward HMB, which prevented the regeneration of the Mn^III starting complex and removed the possibility for catalytic turnover. Only in the case of the O-atom acceptor PPh₃, or 9,10-dihydro-1-methylacridine, which has a very weak C–H bond was photocatalysis observed.^16,33 The reactivity of high-valent metal-oxo complexes to oxidize substrates has been reported to be much enhanced by the presence of acids,^34–42 suggesting that the addition of an H⁺ source could activate the Mn^V(O) complex and possibly lead to catalytic turnover.

We showed that addition of a strong proton source [H(OEt₂)₂]⁺[B(C₆F₅)₄]⁻ (H⁺[B(C₆F₅)₄]⁻) allowed for the photocatalytic oxidation of HMB by Mn^III(TBP₈Cz) with O₂ as oxidant.⁴³ The addition of one and two equivalents of H⁺[B(C₆F₅)₄]⁻ to Mn^III(TBP₈Cz) formed two new species that were characterized by X-ray diffraction (XRD) as the monoprotonated [Mn^III(H₂O)-(TBP₈Cz(H))][B(C₆F₅)₄], and the diprotonated [Mn^III(H₂O)(TBP₈Cz(H)₂)][B(C₆F₅)₄]₂, respectively. The monoprotonated [Mn^III(H₂O)(TBP₈Cz(H))][B(C₆F₅)₄] is protonated at the meso-N atom adjacent to the direct pyrrole-pyrrole bond of the corrolazine ligand. For the di-protonated [Mn^III(H₂O)(TBP₈Cz(H)₂)][B(C₆F₅)₄]₂ complex, both H-atoms were found to be located on the opposite meso-N atoms adjacent to the direct pyrrole–pyrrole bond. The mono-protonated Mn^III complex was catalytically active with a turnover number (TON) of 18 for pentamethylbenzyl alcohol and 9 for pentamethylbenzaldehyde. Interestingly, the diprotonated [Mn^III(H₂O)(TBP₈Cz(H)₂)]²⁺ complex showed no catalytic activity, indicating that catalytic turnover was highly dependent on the level of protonation.

In this work we present several new findings on the proton-assisted, photoactivated oxygenation chemistry catalyzed by manganese corrolazine. We demonstrate that substitution of the H⁺[B(C₆F₅)₄]⁻ proton donor with triflic acid (HOTf) leads to dramatic changes in catalytic activity, and generates new protonated Mn(III) corrolazine complexes, Mn^III(OTf)(TBP₈Cz(H)) and [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf], which were characterized by X-ray diffraction. The crystal structures reveal that protonation occurs on the meso nitrogen atoms as seen for H⁺[B(C₆F₅)₄]⁻, but significant changes in axial ligation state are observed with HOTf, including coordination of the anionic triflate and formation of both 5- and 6-coordinate species. Coordination of OTf- provides a possible mechanism for enhancing catalytic reactivity through a proposed Mn^V(O)(TBP₈Cz) intermediate.^44,45,51 In our previous report on proton-assisted catalysis, we suggested a mechanism that involved a short-lived tripseptet excited state of the monoprotonated Mn^III complex as a key intermediate. This mechanism was based on prior photochemical characterization of the parent Mn^III complex. However, direct spectroscopic evidence for the proposed photochemical excited states of the protonated Mn^III species was absent. Herein we present femtosecond laser flash photolysis measurements on both the monoprotonated Mn^III(OTf)(TBP₈Cz(H)) and the diprotonated [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)]-[OTf], which reveal shorter-lived excited states (⁷T₁) in the presence of O₂ that provide direct evidence for the proposed mechanism of photocatalytic oxygenation. We also demonstrate new catalytic activity for the monoprotonated complex. This complex was shown to be an excellent catalyst for the selective S-oxygenation of thioanisole with only O₂ and light as reagents.

▪ EXPERIMENTAL SECTION

Materials

The starting material Mn^III(TBP₈Cz) was synthesized according to published procedures.⁴⁶ The commercially available reagents (hexamethylbenzene and trifluoromethane-sulfonic acid) were purchased with the best available purity and used without further purification. Benzonitrile (PhCN) was purchased with the best available purity from Wako Pure Chemical Industries, Ltd. and dried according to literature procedures⁴⁷ and distilled under Ar prior to use.

Spectral and Kinetic Measurements

The formation of protonated species of Mn^III(TBP₈Cz) and Mn^V(O)(TBP₈Cz) were examined from the change in the UV-vis spectra of Mn^III(TBP₈Cz) (λ_max = 695 nm, ε_max = 3.5 × 10⁴ M⁻¹ cm⁻¹) and Mn^V(O)(TBP₈Cz) (λ_max = 634 nm, ε_max = 2.0 × 10⁴ M⁻¹ cm⁻¹)¹⁶ by spectral titration with HOTf at 298 K using a Hewlett-Packard HP8453 diode array spectrophotometer. The first and second binding constants of HOTf with Mn^III(TBP₈Cz) were determined in PhCN from the UV-vis spectral change due to the formation of the protonated species Mn^III(OTf)(TBP₈Cz(H)) (λ_max = 725 nm) and [Mn^III(OTf)(H₂O)-(TBP₈Cz(H)₂)][OTf] (λ_max = 745 nm). The first binding constant of HOTf with Mn^V(O)(TBP₈Cz) was determined in PhCN from the UV-vis spectral change due to the formation of the protonated species Mn^IV(OH)(OTf)(TBP₈Cz^•+) (λ_max = 795 nm).

Crystallization of Mn^III(OTf)(TBP₈Cz(H))

To a solution of Mn^III(TBP₈Cz) (20 mg, 14 µmol) in CH₂Cl₂ (1.0 mL) was added Sc(OTf)₃ (70 mg, 1.4 × 10⁻⁴ mol, 10 equiv) dissolved in CH₃CN (0.4 mL). A color change from brown to reddish-brown was observed. The solution was then filtered through celite and eluted with CH₂Cl₂ (~1.0 mL). X-ray quality crystals were obtained by slow evaporation of this solution for about 4 weeks. Characterization by elemental analysis indicating the monoprotonated Mn^III(OTf)(TBP₈Cz(H))•CH₂Cl₂; calcd. for C₉₈H₁₀₇Cl₂F₃MnN₇O₃S: C, 71.52; H, 6.55; N, 5.96. Found: C, 71.65; H, 6.53; N, 6.01.

Crystallization of [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf]

To a solution of Mn^III(TBP₈Cz) (1.5 mg, 1.1 µmol) in CH₂Cl₂ (0.5 mL) was added 2 equiv HOTf. A color change from brown to red was observed. The solution was transferred to a NMR tube and layered with n-heptane. X-ray quality crystals were obtained after a month. Elemental analysis was performed for [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf⁻]•CH₂Cl₂•C₇H₁₆; calcd. for C₁₀₆H₁₂₆Cl₂F₆MnN₇O₇S₂: C, 66.51; H, 6.64; N, 5.12. Found: C, 66.28; H, 6.32; N, 5.56.

X-ray Crystallography

Mn^III(OTf)(TBP₈Cz(H))

All reflection intensities were measured at 110(2) K using a KM4/Xcalibur (detector: Sapphire3) with enhance graphite-monochromated Mo Kα radiation (λ = 0.71073 Å) under the program CrysAlisPro (Version 1.171.35.11 Oxford Diffraction Ltd., 2011). The program CrysAlisPro (Version 1.171.35.11, Oxford Diffraction Ltd., 2011) was used to refine the cell dimensions. Data reduction was done using the program CrysAlisPro (Version 1.171.35.11, Oxford Diffraction Ltd., 2011). The structure was solved with the program SHELXS-97 and was refined on F² with SHELXL-97.⁴⁸ Analytical numeric absorption corrections based on a multifaceted crystal model were applied using CrysAlisPro (Version 1.171.35.11, Oxford Diffraction Ltd., 2011). The temperature of the data collection was controlled using the system Cryojet (manufactured by Oxford Instruments). The H atoms (except when specified) were placed at calculated positions using the instructions AFIX 43 or AFIX 137 with isotropic displacement parameters having values 1.2 or 1.5 times U_eq of the attached C atoms. The H atom attached to N5 was located from difference Fourier maps, and its atomic coordinates were refined freely. Seven of the eight 4-tert-butylphenyl groups are either wholly disordered (over two orientations) or partially disordered around the tert-butyl groups (over two orientations). The coordinated counterion is also disordered over two orientations.

Mn^III(OTf)(TBP₈Cz(H))

Fw = 1560.88, black plate, 0.58 × 0.33 × 0.06 mm³, monoclinic, P2₁/n (no. 14), a = 19.9174(5), b = 21.8922(4), c = 20.3238(5) Å, β = 110.584(3)°, V = 8296.1(3) ų, Z = 4, D_x = 1.250 g cm⁻³, µ= 0.247 mm⁻¹, abs. corr. range: 0.923–0.987. 45329 Reflections were measured up to a resolution of (sin θ/λ)_max = 0.59 Å⁻¹. 14603 Reflections were unique (R_int = 0.0440), of which 10025 were observed [I > 2σ(I)]. 1482 Parameters were refined using 1634 restraints. R1/wR2 [I > 2σ(I)]: 0.0624/0.1552. R1/wR2 [all refl.]: 0.0977/0.1734. S = 1.041. Residual electron density found between −0.53 and 0.85 e Å⁻³.

[Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf]

All reflection intensities were measured at 110(2) K using a SuperNova diffractometer (equipped with Atlas detector) with Cu Kα radiation (λ = 1.54178 Å) under the program CrysAlisPro (Version 1.171.36.32 Agilent Technologies, 2013). The same program was used to refine the cell dimensions and for data reduction. The structure was solved and refined on F² with SHELXL-2013.⁴⁸ Analytical numeric absorption correction based on a multifaceted crystal model was applied using CrysAlisPro. The temperature of the data collection was controlled using the system Cryojet (manufactured by Oxford Instruments). The H atoms were placed at calculated positions (unless otherwise specified) using the instructions AFIX 23, AFIX 43 or AFIX 137 with isotropic displacement parameters having values 1.2 times U_eq of the attached C or N atoms. The H atoms attached to O1W and O2W/O2W were found from difference fourier maps, and their coordinates were refined freely using the DFIX restraints. The tert-butyl group C53→C56, the tert-butylphenyl group C77→C86, and the hydronium cation are found to be disordered over two orientations, and the occupancy factors of the major components of the disorder refine to 0.50(3), 0.588(6), and 0.634(4), respectively. The crystal lattice contains some disordered solvent molecules (DCM and heptane). All orientations of the three non-fully occupied DCM molecules were successfully modeled, and the asymmetric unit contains ca. 2.53 DCM molecules per Mn complex. The heptane molecule was found to be disordered, and its contribution has been taken out in the final refinement.⁴⁹

[Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf]

Fw = 2111.86, dark brown-black block, 0.35 × 0.30 × 0.21 mm³, triclinic, P−1 (no.2), a = 18.3393(4), b = 18.9149(3), c = 21.2275(3), α = 91.1721(13), β = 114.7271(17), γ = 118.8374(19)°, V = 5625.8(2) ų, Z = 2, D_x = 1.247 g cm⁻³, µ = 3.176 mm⁻¹, T_min – T_max: 0.445–0.619. 72524 Reflections were measured up to a resolution of (sin θ/λ)_max = 0.62 Å⁻¹. 22105 Reflections were unique (R_int = 0.0242), of which 20831 were observed [I > 2σ(I)]. 1549 Parameters were refined using 987 restraints. R1/wR2 [I > 2σ( I)]: 0.0500/0.1381. R1/wR2 [all refl.]: 0.0524/0.1403. S = 1.036. Residual electron density found between −0.57 and 1.25 e Å⁻³.

Photocatalytic Reactivity

The catalytic reactivity of Mn^III (TBP₈Cz) (1.0 × 10⁻⁵ M) was examined under a large excess of a substrate (hexamethylbenzene or thioanisole: 0 – 0.3 M) in the presence of HOTf (0 – 1.0 × 10⁻⁴ M) in PhCN (2.0 mL). This solution was transferred to a glass cuvette equipped with a stir bar. The generated products of the reaction solution were analyzed by GC-MS. An aliquot was taken from the product solution and injected directly into the GC-MS for analysis. All peaks of interest were identified by comparison of retention times and co-injection with authentic samples. The amounts of products were quantified by comparison against a known amount of detected products using a calibration curve consisting of a plot of product mole versus area. Mass spectra were recorded on a JEOL JMS-700T Tandem MS station, and the GC-MS analyses were carried out by using a Shimadzu GCMS-QP2000 gas chromatograph mass spectrometer. GC-MS conditions in these experiments were performed as follows: an initial oven temperature of 60 °C was held for 1 min and then raised 30 °C min⁻¹ for 7.3 min until a temperature of 250 °C was reached, which was then held for further 10 min. The products pentamethylbenzyl alcohol and pentamethylbenzaldehyde were identified by comparison with a standard sample. The product yields were quantified by comparison against a known amount of detected products using a calibration curve consisting of a plot of product mole versus area. Calibration curves were prepared by using concentrations in the same range as that observed in the actual reaction mixtures.

Quantum Yield Determination

A standard actinometer (potassium ferrioxalate)⁵⁰ was used for the quantum yield determination of photocatalytic oxidation of HMB by O₂ with Mn^III (TBP₈Cz) in O₂-saturated PhCN. Typically, a square quartz cuvette (10 mm i.d.), which contained an O₂-saturated PhCN solution (2.0 mL) of Mn^III (TBP₈Cz) (1.0 × 10⁻⁵ M) and HMB (0.4 M) in the presence of HOTf (4.0 × 10⁻⁵ M) was irradiated with monochromatic light of λ =450 nm from a Shimadzu RF-5300PC fluorescence spectrophotometer. Under the conditions of actinometry experiments, the actinometer and Mn^III(TBP₈Cz) absorbed essentially all of the incident light at λ = 450 nm. The light intensity of monochromatized light at λ = 450 nm was determined to be 3.7 × 10⁻⁸ einstein s⁻¹. The quantum yields were determined by the amount of generated a product.

Femtosecond Laser Flash Photolysis Measurements

Measurements of transient absorption spectra of Mn^III(OTf)(TBP₈Cz(H)) and [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf] were performed according to the following procedures. An N₂- or O₂-saturated PhCN solution containing Mn^III(TBP₈Cz) (8.4 × 10⁻⁵ M) and HOTf (1.7 × 10⁻⁴ or 1.2 × 10⁻² M) was excited using an ultrafast source, Integra-C (Quantronix Corp.), an optical parametric amplifier, TOPAS (Light Conversion Ltd.), and a commercially available optical detection system, Helios provided by Ultrafast Systems LLC. The source for the pump and probe pulses were derived from the fundamental output of Integra-C (λ = 786 nm, 2 mJ/pulse and fwhm = 130 fs) at a repetition rate of 1 kHz. 75% of the fundamental output of the laser was introduced into a second harmonic generation (SHG) unit: Apollo (Ultrafast Systems) for excitation light generation at λ = 393 nm, while the rest of the output was used for white light generation. The laser pulse was focused on a sapphire plate of 3 mm thickness and then white light continuum covering the visible region from λ = 410 nm to 800 nm was generated via self-phase modulation. A variable neutral density filter, an optical aperture, and a pair of polarizer were inserted in the path in order to generate stable white light continuum. Prior to generating the probe continuum, the laser pulse was fed to a delay line that provides an experimental time window of 3.2 ns with a maximum step resolution of 7 fs. In our experiments, a wavelength at λ = 393 nm of SHG output was irradiated at the sample cell with a spot size of 1 mm diameter where it was merged with the white probe pulse in a close angle (< 10 °). The probe beam after passing through the 2 mm sample cell was focused on a fiber optic cable that was connected to a CMOS spectrograph for recording the time-resolved spectra (λ = 410 – 800 nm). Typically, 1500 excitation pulses were averaged for 3 seconds to obtain the transient spectrum at a set delay time. Kinetic traces at appropriate wavelengths were assembled from the time-resolved spectral data. The decay rate of the tripquintet (⁵T₁) obeyed the first-order kinetics given by eq 1,

$$\mathrm{\Delta }\text{Abs}=A_1\text{exp}(-k_{1}t)+A_2$$ (1)

where A₁ and A₂ are pre-exponential factors for the absorbance changes and the final absorbance, and k₁ is the rate constant of the decay of the tripquintet (⁵T₁) after irradiation. The slower decay rate of the tripseptet (⁷T₁) also obeyed the first-order kinetics given by eq 2, where A₃ is the final absorbance of ⁷T₁ and k₂ is the rate constant of the decay of ⁷T₁. All measurements were conducted at room temperature, 298 K.

$$\mathrm{\Delta }\text{Abs}=A_1\text{exp}(-k_{1}t)+A_2\text{exp}(-k_{2}t)+A_3$$ (2)

RESULTS AND DISCUSSION

Protonation of Mn^III(TBP₈Cz)

Protonation of Mn^III(TBP₈Cz) was examined by addition of HOTf to a PhCN solution of Mn^III(TBP₈Cz) (1.0 × 10⁻⁵ M). The formation of a new species with a Q-band at λ_max = 725 nm, was monitored by UV-vis as shown in Figure 1a, where HOTf was added up to 1 equiv of Mn^III(TBP₈Cz). A color change from brown to reddish-brown was observed. This spectral change was similar to that seen for the addition of 1 equiv of H⁺[B(C₆F₅)₄]⁻, and indicated the formation of the monoprotonated complex, Mn^III(OTf)(TBP₈Cz(H)) (1). Addition of excess HOTf (1.0 × 10⁻⁵ – 1.0 × 10⁻² M) resulted in a change in the Q-band region to give λ_max = 745 nm with the isosbestic points as shown in Figure 2. These data were consistent with the formation of the diprotonated, [Mn^III(OTf)(TBP₈Cz(H)₂)]-[OTf] (2) and again similar to the spectral changes observed for H⁺[B(C₆F₅)₄]⁻.

Figure 1.

(a) UV-vis spectral changes and (b) plot for determination of the binding constant of the conversion from Mn^III(TBP₈Cz) (black line, 1.0 × 10⁻⁵ M) to Mn^III(OTf)(TBP₈Cz(H)) (blue line) by spectroscopic titration of HOTf (0 – 2.0 × 10⁻⁵ M) monitored at 695 nm (Mn^III(TBP₈Cz), black dots) and 725 nm (Mn^III(OTf)(TBP₈Cz(H)), blue dots) in a PhCN solution at room temperature.

Figure 2.

(a) UV-vis spectral changes and (b) plot for determination of the binding constant of the conversion from Mn^III(OTf)(TBP₈Cz(H)) (blue line, 1.0 × 10⁻⁵ M) to [Mn^III(OTf)(TBP₈-Cz(H)₂)][OTf] (red line) by spectroscopic titration of HOTf (1.0 × 10⁻⁵ – 1.0 × 10⁻² M) monitored at 725 nm (Mn^III(OTf)(TBP₈Cz(H)), blue dots) and 745 nm ([Mn^III(OTf)(TBP₈Cz-(H)₂)][OTf], red dots) in a PhCN solution at room temperature.

The monoprotonation equilibrium constant (K₁) for 1 was determined by fitting of the titration curves in Figure 1b, according to the equilibrium expression in eq 3. This analysis gave

$${\text{Mn}}^{\text{III}}({\text{TBP}}_8\text{Cz})+{\mathrm{H}}^{+}+{\text{OTf}}^{-}\stackrel{K_1}{\rightleftharpoons }{\text{Mn}}^{\text{III}}(\text{OTf})({\text{TBP}}_8\text{Cz}(\mathrm{H}))$$ (3)

K₁ = 9.0 × 10⁶ M⁻¹, indicating strong binding of H⁺ to Mn^III (TBP₈Cz). The second equilibrium constant (K₂) for the protonation equilibrium in eq 4, was determined to be 4.7 × 10³ M⁻¹ by

$${\text{Mn}}^{\text{III}}(\text{OTf})({\text{TBP}}_8\text{Cz}(\mathrm{H}))+{\mathrm{H}}^{+}+{\text{OTf}}^{-}\stackrel{K_2}{\rightleftharpoons }[{\text{Mn}}^{\text{III}}(\text{OTf})({\mathrm{H}}_2\mathrm{O})({\text{TBP}}_8\text{Cz}{(\mathrm{H})}_2)][\text{OTf}]$$ (4)

fitting a plot of reciprocal absorbance difference vs 1/[HOTf] from the titration curve in Figure 2b (see Figure S1 in the Supporting Information (SI)). The K₁ value for the first protonation to give 1 is much larger than the K₂ value for the second species 2.

The protonated complexes 1 and 2 were successfully crystallized and characterized by single crystal X-ray Diffraction (XRD) (Figure 3). The monoprotonated complex was originally crystallized from reaction of Mn^III (TBP₈Cz) with 10 equiv Sc(OTf)₃ in CH₂Cl₂/CH₃CN. The strongly Lewis acidic Sc³⁺ ion presumably reacts with exogenous H₂O to generate the proton source, and slow evaporation produced X-ray quality crystals of 1 after 4 weeks. The crystal structure revealed that monoprotonation occurs at one of the meso-N-atoms (N5(H)) adjacent to the direct pyrrole-pyrrole bond of the corrolazine ligand, and that the manganese center is five-coordinate with an axially-ligated triflate (Mn^III-O = 2.115(3) Å). The same crystals were subsequently obtained from reaction o Mn^III (TBP₈Cz) with 1 equiv HOTF and recrystallization from CH₂Cl₂/heptane. A unit cell measurement matched that for 1.

Figure 3.

Displacement ellipsoid plots (50% probability level) and chemical structures for (a) [Mn^III(OTf)(TBP₈Cz(H))] and (b) [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf] at 110(2) K.

Dissolution of crystalline 1 in PhCN yields the same UV-vis spectrum as seen for the in situ protonation o Mn^III (TBP₈Cz) by the addition of 1 equiv of HOTF (λ_max = 451, 725 nm) (Figure S2). Interestingly dissolution of crystalline 1 in CH₂Cl₂ (Figure S3) gives a UV-vis spectrum with λ_max = 443, 725 nm, similar to what was seen for [Mn^III (H₂O)(TBP₈Cz(H))][B(C₆F₅)₄] (λ_max = 446, 730 nm in CH₂Cl₂)⁴³ although the Q-band is shifted by 5 nm. A similar red-shift in the Q-band occurs upon axial ligation of anionic donors to Mn^III(H₂O)(TBP₈Cz) to give [Mn^III(X)(TBP₈Cz)]⁻ (e.g. X = CN⁻, F⁻)⁵¹.

X-ray quality crystals of the diprotonated complex 2 were obtained from the reaction of Mn^III(TBP₈Cz) with 2 equiv HOTf and recrystallization from CH₂Cl₂/heptanes (Figure 3). Both H-atoms were unambiguously located on the opposite meso-N atoms (N1 and N5). In contrast to the monoprotonated complex, the diprotonated complex has a six coordinate manganese center with an axially ligated water at 2.144(2) Å and a weakly bound triflate at 2.702(2) Å.

Dissolution of crystalline 2 in CH₂Cl₂ yielded a spectrum with λ_max at 454 and 758 nm (Figure S4). Comparing this spectrum to that seen for the diprotonated [Mn^III(H₂O)(TBP₈Cz(H)₂)]-[B(C₆F₅)₄]₂ (λ_max: 470, 763)⁴³ in the same solvent again reveals a red-shift of 5 nm in the Q-band, matching the differences seen for the two monoprotonated spectra. These results indicate that the triflate (⁻OTf) anion stays bound in solution for both the mono- and diprotonated complexes. When crystalline 2 was dissolved in PhCN, the UV-vis spectrum matched that for the monoprotonated complex, indicating that this solvent is basic enough to deprotonate the second meso-N atom.

Photocatalytic Oxygenations with O₂ by Mn^III(TBP₈Cz) in the Presence of HOTf

Previously it was shown that the addition of H⁺[B(C₆F₅)₄]⁻ to Mn^III(TBP₈Cz) allowed for photo-activated catalysis of HMB under ambient conditions, but only with modest turnover numbers and selectivity.⁴³ Replacement of H⁺[B(C₆F₅)₄]⁻ with the strong acid HOTf led to the photocatalytic oxygenation of HMB by O₂ and Mn^III(TBP₈Cz) in PhCN.

The addition of HOTf (2 equiv) to Mn^III(TBP₈Cz) in PhCN gave a spectrum (λ_max = 451, 725 nm) indicative of 1. Addition of HMB followed by photoirradiation initiated the catalytic reaction, which occured over 6 h with slow bleaching of the solution (Figure S5 in SI). Regarding the bleaching of the catalyst, we followed the reaction by UV-vis, which showed only the slow loss of the ground state complex. The catalytically active excited state was only present to a small extent at any time, and therefore the slow decomposition of the ground state (which is in large excess) did not change the effective concentration of the catalytically active excited state to a large extent. Analysis of the reaction mixture by removal of aliquots and analysis by GC showed a steady increase in the amount of oxidized products produced over time. Thus, there is an optimized concentration of HOTf for the phtoocatalytic oxygenation of HMB by O₂ with 1. The reason why the photocatlaytic oxygenation is prohibited in the presence of large concentrations of HOTf is discussed in relation with the phtoocatalytic reaction mechanism (vide infra). The production of oxidized products seemed only limited by the catalyst stability. The final turnover number (TON) was 563 for the alcohol PMB–OH and 9 for the aldehyde PMB– CHO (Figure S6 in SI), with a quantum yield of 0.125%. The catalytic turnover is dramatically increased by HOTf in comparison to H⁺[B(C₆F₅)₄]⁻, which gave only TON(PMB–OH) = 18 and TON(PMB–CHO) = 9.⁴³ In addition, the HOTf reaction is much more selective for production of the alcohol product. When the HMB was replaced by dueturated HMB, the reactivity was lower. The kinetic isotope effect (KIE) was determined to be 3.3 with linear increase of PMB-OH vs time, indicating H-atom abstraction is the rate determining step in the overall photocatalytic reaction.

The amount of PMB–OH produced after 1 h photoirradiation of a PhCN solution of Mn^III(TBP₈Cz) with HMB, O₂ and HOTf was proportional to concentrations of HMB (Figure 4a) and O₂ (Figure 4b). Thus, the rate-determining step in the photocatalytic oxygenation of HMB by O₂ may be the reaction of the photogenerated species from 1 with O₂ and HMB. If HOTf is added in excess (10 equiv), 2 is formed as seen by UV-vis. Photoirradiation of this complex in the presence of HMB/O₂ shows no oxidized products. This result is consistent with previous observations on the diprotonated Mn^III complex generated from H⁺[B(C₆F₅)₄]⁻, which also showed no catalytical activity.⁴³ The amount of PMB–OH produced by 30 min photoirradiation with a Xenon lamp increased with increasing concentration of HOTf to reach a maximum and then decreased with further increase in concentrationof HOTf as shown in Figure 5 because of the formation of the diprotonated complex 2 which shows no catalytic activity for oxygenation of HMB by O₂ in the presence of large excess HOTf (Figure S6b).

Figure 4.

(a) Plots of the amount of produced pentamethylbenzyl alcohol for photocatalytic oxygenation of HMB (0 – 4.0 × 10⁻² M) by Mn^III(TBP₈Cz) (1.0 × 10⁻⁵ M) in the presence of HOTf(4.0 × 10⁻⁵ M) in an 02-saturated PhCN solution and (b) Plots of the amount of produced pentamethylbenzyl alcohol for photocatalytic oxygenation of HMB (2.0 × 10⁻² M) by Mn^III(TBP₈Cz) (1.0 × 10⁻⁵ M) in the presence of HOTf (4.0 × 10⁻⁵ M) in a PhCN solution vs the oxygen concentration (0–8.5 × 10⁻³ M) under photoirradiation (white light) for 1 h at room temperature.

Figure 5.

Plots of the amount of produced pentamethylbenzyl alcohol for photocatalytic oxygenation of HMB (2.0 × 10⁻² M) by Mn^III(TBP₈Cz) (1.0 × 10⁻⁵ M) in the presence of HOTf (0 – 1.0 × 10⁻⁴ M) under photoirradiation (white light) for 30 min in an O₂-saturated PhCN solution at room temperature.

Photocatalytic Sulfoxidation by Mn^III(TBP₈Cz) in the Presence of HOTf

In our previous studies we focused exclusively on C–H functionalization with toluene derivatives. In the current work we wanted to determine if photocataytic oxygenation could be obtained with an O-atom acceptor substrate, and the thioether substrate PhSMe was examined. There have been some reports involving iron complexes reacting with O₂ to give catalytic sulfoxidation of thioether substrates, although they often involved poorly controlled radical-type pathways.⁵² The photoactivation of diiron(III)-µ-oxo-bridged porphyrins has also lead to catalytic sulfoxidation where dioxygen is used to regenerate the µ-oxo complex.⁵³ Dioxygen has been utilized by iron complexes for catalytic sulfoxidation by the formation of a peracid oxidant.⁵⁴ However, there remains few examples of sulfoxidation mediated by O₂ and well-defined transition metal catalysts that avoid co-additives or radical-type pathways. When HMB was replaced by thioanisole in the photocatalytic reaction with Mn^III(OTf)(TBP₈Cz(H)), a single S-oxygenated product, methylphenylsulfoxide (PhS(O)Me), was obtained (Scheme 2). Excellent TONs (902) and conversion (98%) of PhSMe to PhS(O)Me was found (Figure S7a) whereas negligible amount of PhS(O)Me (TON: 14) was obtained in the absence of HOTf (Figure S7b). Control experiments confirmed that no corresponding products were observed without photoirradiation or without the Mn^III complex (see Figure S8 in SI). These results show that Mn^III(TBP₈Cz) is an efficient and selective catalyst for the sulfoxidation of thioanisole, and requires only air, light, and an H⁺ source without the need for a co-reductant.

Scheme 2.

Catalytic Oxygenation of Thioanisole by O₂ with Mn^III(TBP₈Cz) and HOTf under Photoirradiation Conditions

Femtosecond Transient Absorption Measurements

In order to clarify the photodynamics of 1 and 2, femtosecond laser flash photolysis measurements were performed in the absence and presence of O₂ in PhCN. The Mn^III ion is a high spin (S = 2) species, and coupling between the metal d electrons and the π electrons of the corrolazine ring leads to a singquintet (⁵S₀) ground state derived from the lowest excited ring (π,π*) singlet for 1 and 2 due to the coupling between unpaired electrons of the metal with the π electrons of the corrolazine ring. A “tripmultiplet” manifold (³T₁, ⁵T₁, ⁷T₁) is derived from the lowest ring (π,π*) triplet.⁵⁵ Femtosecond laser excitation of 1 leads to a new maximum absorption at 610 nm, which can be assigned to the tripquintet (⁵T₁) excited state (Figure 6). The absorption due to the tripquintet (⁵T₁) decayed in a few ps with a rate constant of 3.1 × 10⁹ s⁻¹, and there was no effect of O₂ on the rate due to the fast intersystem crossing to the tripseptet excited state (⁷T₁) of 1 with a longer lifetime (Figure 6). An extremely rapid intersystem crossing process from the singquintet excited state (⁵S₁) to the tripquintet excited state (⁵T₁) resulting from the presence of unpaired electrons has been seen for a first row paramagnetic complex, Mn^III.⁵⁶ For example, in Mn^III porphyrins, the existence of two tripmultiplet levels was suggested where a tripquintet (⁵T₁) relaxes to a long-lived tripseptet (⁷T₁), which requires a spin conversion to go back to the quintet ground state. The absorption at 610 nm due to the tripseptet excited state (⁷T₁) decayed significantly faster in O₂-saturated PhCN (Figure 6c) as compared with the decay in the absence of O₂ (Figure 6b). The rate constant of the reaction of ⁷T₁ of 1 with O₂ was determined to be 5.0 × 10⁹ M⁻¹ s⁻¹, which is the same as that of the reaction of ⁷T₁ of neutral Mn^III(TBP₈Cz) with O₂.¹⁶ This reaction should produce the putative superoxo complex ([Mn^IV(O₂^•−)(TBP₈Cz(H))]⁺).

Figure 6.

(a) Transient absorbance spectral changes (red after 10 ps, green 100 ps, blue 500 ps, and black 3000 ps) after photoexcitation of Mn^III(OTf)(TBP₈Cz(H)) in PhCN. Time profile of the generation and decay of {Mn^III(OTf)(TBP₈Cz(H))}^* at λ = 610 nm under (b) N₂ and (c) O₂. The black lines are exponential fitting given in eqs 1 and 2.

Femtosecond laser flash photolysis measurements of 2 were also performed to characterize the excited state of the diprotonated species in the absence and presence of O₂ in PhCN, as shown in Figure 7. The decay rate constant of the intersystem crossing from the tripquintet excited state (⁵T₁) to the tripseptet excited state (⁷T₁) was determined to be 9.4 × 10⁹ s⁻¹ at 745 nm in both the absence and presence of O₂. In contrast to the case of Mn^III(TBP₈Cz) and 1, the decay rate constant for the tripseptet excited state (⁷T₁) of 2 was slower (2.2 × 10⁹ s⁻¹ at 745 nm) and showed no O₂ dependence. Thus, the tripseptet excited state (⁷T₁) of 2 does not react with O₂, in line with the observation that 2 is not catalytically active. We hypothesize that the one electron oxidation potential of ⁷T₁ of 2 may be too high for reaction with O₂.

Figure 7.

(a) Transient absorbance spectral changes (red after 10 ps, green 100 ps, blue 500 ps, and black 3000 ps) after photoexcitation of [Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf] in PhCN. Time profile of the generation and decay of {[Mn^III(OTf)(H₂O)(TBP₈Cz(H)₂)][OTf]}^* at λ = 745 nm under (b) N₂ and (c) O₂. The black lines are exponential fitting given in eqs 1 and 2.

Oxidation Potentials of Mn^III(TBP₈Cz), Mn^III(OTf)(TBP₈Cz(H)) and [Mn^III(OTf)-(H₂O)(TBP₈Cz(H)₂)][OTf]

The effects of protonation of Mn^III(TBP₈Cz) on the one-electron oxidation potentials were examined by cyclic voltammetry measurements of Mn^III(TBP₈Cz), 1, and 2, and the peak potentials (referenced to SCE) are shown in Figure 8. The first one-electron oxidation potential of Mn^III(TBP₈Cz) was determined to be E_ox = 0.78 V vs SCE in PhCN, which was shifted to the positive direction by the addition of HOTf: E_ox = 1.04 V for 1 and E_ox = 1.36 V for 2. However, the second one-electron oxidation potential was not positively shifted by the addition of HOTf: E_ox = 1.23 V for Mn^III(TBP₈Cz) and E_ox = 1.20 V for 1 probably because of the counter anion effect of OTf⁻. Although the energies of ⁷T₁ of Mn^III(TBP₈Cz), 1 and 2 have yet to be determined because of the absence of phosphorescence, the E_ox value of 2 in the ground state, which is 0.32 V higher than that of 1, is consistent with the observation that the ⁷T₁ of 2 is unreactive towards O₂ in contrast to 1.

Figure 8.

Cyclic voltammograms of (a) Mn^III(TBP₈Cz) (upper, 1.0 × 10⁻³ M) in the absence of HOTf and in the presence of HOTf (below, 1.0 × 10⁻³ M), and (b) Mn^III(TBP₈Cz) (1.0 × 10⁻³ M) in the presence of HOTf (0.1 M) in PhCN containing 0.1 M TBAP. Scan rate = 0.10 V s⁻¹.

Reaction Mechanism

Based on the study of the photocatalytic oxygenation of HMB by O₂ with 1 and photodynamics of 1 (vide supra), the mechanism of photocatalytic oxygenation of HMB by O₂ with 1 is proposed as shown in Scheme 3.

Scheme 3.

Mechanism of Photocatalytic Oxygenation of HMB by 1 with O₂

In the presence of 1 equiv of HOTf, Mn^III(TBP₈Cz) is converted to the monoprotonated complex 1, which is then excited to the tripquintet state {Mn^III(OTf)(TBP₈Cz(H))}^* (⁵T₁) upon photoirradiation. The ⁵T₁ of 1 is converted rapidly by intersystem crossing (ISC) to the tripseptet excited state (⁷T₁). Electron transfer from ⁷T₁ to O₂ occurs to produce the superoxo complex [Mn^IV(O₂^•−)(OTf)(TBP₈Cz(H))], which abstracts a hydrogen atom from HMB to produce the hydroperoxo complex [Mn^IV(OOH)(OTf)(TBP₈Cz(H))] and pentamethylbenzyl radical, in competition with the back electron-transfer (k_–et) to regenerate the ground state 1 and O₂. C–H activation by a metal superoxo species has been demonstrated with chromium and copper non-heme model complexes.⁵⁷ In our case, the subsequent homolytic O–O bond cleavage and combination with benzyl radical to yield PMB–OH is accompanied by generation of high-valent Mn^V(O)(OTf)(TBP₈Cz(H)), followed by conversion to Mn^IV(OH)(OTf)(TBP₈Cz^•+).⁴³ This species, in the presence of excess H⁺, is proposed to react with another equivalent of substrate to produce PMB–OH, regenerate 1, and close the catalytic cycle.

When 1 is further protonated to 2 in the presence of a large excess of HOTf no catalytic activity is observed. The tripseptet excited state (⁷T₁) of 2 is favored to go back to the ground state rather than react with O₂, possibly due to the higher oxidation potential for 2 as compared with that of 1 (vide supra) as shown in Scheme 4. As a result, the photocatalytic oxygenation of HMB by O₂ with Mn^III(TBP₈Cz) does not occur in the presence of large excess HOTf (10 equiv HOTf). In such a case, the involvement of singlet oxygen is unlikely, because the formation of singlet oxygen may not be affected by the acid. In addition, the efficient photocatalytic oxygenation of thioanisole by O₂ with Mn^III(TBP₈Cz) and HOTf in PhCN (Scheme 2) also suggests no involvment of singlet oxygen, because the reactions of alkyl and aryl sulfides with singlet oxygen in aprotic solvents such as PhCN is known to be sluggish.⁵⁸

Scheme 4.

Mechanism of Photocatalytic Oxygenation of HMB by 2 with O₂

Protonation of Mn^V(O)(TBP₈Cz)

To further study the proposed final step in the catalysis to regenerate the Mn^III resting state, the reaction of isolated Mn^V(O)(TBP₈Cz) with acid was studied. Protonation of Mn^V(O)(TBP₈Cz) was examined by reaction of Mn^V(O)(TBP₈Cz) with HOTf in PhCN as shown in Figure 9a. The broadening and decrease in the intensity of the Soret band at 420 nm and the Q band at 634 nm together with the appearance of a relatively weak band in the near-IR region at 795 nm is characteristic of the formation of a porphyrinoid π-radical cation.⁴³ The new spectrum matches that observed previously upon the addition of the Lewis or Brønsted acids (Zn^II(OTf)₂, B(C₆F₅)₃, H⁺) to the Mn^V(O) complex.^42,43 These acids stabilize a high-spin triplet (S = 1) (or quintet S = 2) state with an electronic configuration best described as a manganese(IV) corrolazine π-radical cation.⁴² Protonation of the terminal oxo ligand should weaken the Mn-O π -bonding and destabilize the Mn^V oxidation state, favoring an Mn^IV(O)(π-cation-radical) configuration. The formation of Mn^IV(OH)(OTf)(TBP₈Cz^•+) (3) (λ_max = 795 nm) was monitored by UV-vis spectroscopy as shown in Figure 9, where HOTf was added up to 1 equiv. The equilibrium constant for the addition of acid to Mn^V(O)(TBP₈Cz) was determined to be 1.0 × 10⁷ M⁻¹ by a plot of absorbance vs concentration of HOTf in Figure 9b. No further spectral changes were observed in the presence of large excess HOTf (Figure S9).

Figure 9.

(a) UV-vis spectral changes and (b) plot for the conversion of Mn^V(O)(TBP₈Cz) (blue line, 1.0 × 10⁻⁵ M) to Mn^IV(OH)(OTf)(TBP₈Cz^•+) (red line) by spectroscopic titration of HOTf (0 – 2.0 × 10⁻⁵ M) monitored at 634 nm (Mn^V(O)(TBP₈Cz), blue dots) and 795 nm (Mn^IV(OH)(OTf)(TBP₈Cz^•+), red dots) in PhCN at 298 K.

The addition of HMB to 3 and monitoring the reaction by UV-Vis showed slow decay of 3 and the growth of Mn^III over 14 hours. In the presence of excess of HOTf, however, 3 is reduced by HMB to produce 1 (Figure S10b in SI) and PMB-OH (yield: 85%) by GC-MS. Rates of the oxidation of HMB by Mn^V(O)(TBP₈Cz) in the presence of excess HOTf were determined from an increase in absorbance at 725 nm due to 1 in a PhCN at 298 K, obeying pseudo-first-order kinetics in the presence of large exess HMB and HOTf (Figure S11 and Figure S12 in SI). The observed pseudo-first-order rate constants are proportional to concentrations of HOTf and HMB to afford the second-order rate constants from the slopes of the linear plots as 1.4 × 10⁻³ M⁻¹ s⁻¹ (Figure 10a) and 5.3 × 10⁻⁴ M⁻¹ s⁻¹ (Figure 10b), respectively.

Figure 10.

(a) Plot of the observed pseudo-first-order rate constant (k_obs) vs concentrations of HOTf for the oxidation of HMB (2.0 × 10⁻² M) by Mn^V(O)(TBP₈Cz) (1.0 × 10⁻⁵ M) in the presence of HOTf (2.0 × 10⁻⁵ – 1.0 × 10⁻⁴ M) in PhCN at 298 K. (b) Plot of the observed pseudo-first-order rate constant (k_obs) vs concentrations of HMB for the oxidation of HMB (0 – 1.0 × 10⁻² M) by Mn^V(O)(TBP₈Cz) in the presence of HOTf (3.0 × 10⁻⁵ M) in PhCN at 298 K.

When HMB was replaced by mesitylene (C₆H₃(CH₃)₃), 3 was also reduced by mesitylene in the presence of excess HOTf. A very small deuterium kinetic isotope effect (KIE) close to unity (KIE = 1.1) was observed when C₆H₃(CH₃)₃ was replaced by the deuterated compound (C₆D₃(CD₃)₃) as shown in Figure 11. The KIE close to unity indicates that the oxidation of mesitylene as well as HMB by 3 proceeds via proton-coupled electron transfer from mesitylene and HMB to 3 rather than hydrogen atom transfer from mesitylene and HMB to 3.

Figure 11.

Plots of pseudo-first-order rate constants vs concentrations of mesitylene (C₆H₃(CH₃)₃) (black) and mesitylene-d₁₂ (C₆D₃(CD₃)₃) (blue) in the oxidation of C₆H₃(CH₃)₃ and (C₆D₃(CD₃)₃) by Mn^V(O)(TBP₈Cz) (3.0 × 10⁻⁵ M) in the presence of HOTf (1.5 × 10⁻⁴ M) in PhCN at 298 K.

CONCLUSIONS

In conclusion, Mn^III(OTf)(TBP₈Cz(H)) is capable of the photocatalytic oxygenation of HMB and PhSMe under ambient conditions, and was found to be a more robust, efficient and selective catalyst than the monoprotonated [Mn^III(H₂O)(TBP₈Cz(H))][B(C₆F₅)₄]. Thus changing the acid from H⁺[B(C₆F₅)₄]⁻ to an acid with a different conjugate base resulted in a dramatic change in catalytic reactivity. The major difference in conjugate bases is the ability of OTf⁻ to coordinate to the Mn center, and suggests that OTf⁻ coordination has a significant influence on the catalytic activity of the system. The dynamics of the photoexcited states were similar for both acids, and the same photoexcited state mechanism can be invoked for both H⁺[B(C₆F₅)₄]⁻ and HOTf. The proton-coupled electron transfer from HMB to Mn^IV(OH)(OTf)(TBP₈Cz^•+) occurs to yield pentamethylbenzyl alcohol, accompanied by regeneration of Mn^III(OTf)(TBP₈Cz(H)) to complete the catalytic cycle. The PCET oxidation of HMB by Mn^IV(OH)(OTf)(TBP₈Cz^•+) is enhanced with increasing concentration of HOTf. In the presence of large excess HOTf, however, Mn^III(OTf)(TBP₈Cz(H)) is converted the diprotonated complex [Mn^III(OTf)(TBP₈Cz(H)₂)][OTf], when ⁷T₁ of [Mn^III(OTf)(TBP₈Cz(H)₂)][OTf] cannot react with O₂ to produce the superoxo complex because of the high oxidation potential of [Mn^III(OTf)(TBP₈Cz(H)₂)][OTf] and thereby the catalytic cycle is stopped. It would be of interest in future efforts to examine other acids in order to elucidate the effect of the H⁺ source and conjugate base on the catalysis.

Scheme 1.

Catalytic Oxygenation of Hexamethylbenzene by O₂ with Mn^III(TBP₈Cz) and HOTf under Photoirradiation Conditions