Synthesis, Characterization, and Reactivities of Manganese(V)-Oxo Porphyrin Complexes

Abstract

The reactions of manganese(III) porphyrin complexes with terminal oxidants, such as m-chloroperbenzoic acid, iodosylarenes, and H₂O₂, produced high-valent manganese(V)-oxo porphyrins in the presence of base in organic solvents at room temperature. The manganese(V)-oxo porphyrins have been characterized with various spectroscopic techniques, including UV-vis, EPR, ¹H and ¹⁹F NMR, resonance Raman, and X-ray absorption spectroscopy. The combined spectroscopic results indicate that the manganese(V)-oxo porphyrins are diamagnetic low-spin (S = 0) species with a longer, weaker Mn-O bond than in previously reported Mn(V)-oxo complexes of non-porphyrin ligands. This is indicative of double bond character between the manganese(V) ion and the oxygen atom, and may be attributed to the presence of a trans-axial ligand. The [(Porp)Mn^V=O]⁺ species are stable in the presence of base at room temperature. The stability of the intermediates is dependent on base concentration. In the absence of base, (Porp)Mn^IV=O is generated instead of the [(Porp)Mn^V=O]⁺ species. The stability of the [(Porp)Mn^V=O]⁺ species also depends on the electronic nature of porphyrin ligands; [(Porp)Mn^V=O]⁺ complexes bearing electron-deficient porphyrin ligands are more stable than those bearing electron-rich porphyrins. Reactivity studies of manganese(V)-oxo porphyrins revealed that the intermediates are capable of oxygenating PPh₃ and thioanisoles, but not olefins and alkanes at room temperature. These results indicate that the oxidizing power of [(Porp)Mn^V=O]⁺ is low in the presence of base. However, when the [(Porp)Mn^V=O]⁺ complexes were associated with iodosylbenzene in the presence of olefins and alkanes, high yields of oxygenated products were obtained in the catalytic olefin epoxidation and alkane hydroxylation reactions. Mechanistic aspects, such as oxygen exchange between [(Porp)Mn^V=¹⁶O]⁺ and H₂¹⁸O, are also discussed.

Introduction

An important objective in understanding biological oxidation reactions by heme-containing monooxygenases is to elucidate the nature of reactive intermediates and the mechanism of oxygen atom transfer from the intermediates to organic substrates.¹ Since the catalytic cycle of cytochrome P450 (CYP 450) is believed to involve a high-valent iron(IV)-oxo porphyrin π-cation radical, [(Porp)^+·Fe^IV=O]⁺, as an active oxidant in oxidation reactions,² a number of iron(IV)-oxo porphyrin π-cation radicals have been synthesized by the reaction of iron(III) porphyrins with terminal oxidants such as m-chloroperbenzoic acid (m-CPBA) and iodosylarenes (ArIO), characterized with various spectroscopic techniques, and studied in a variety of oxidation reactions, including alkane hydroxylation and olefin epoxidation.^3,4

Synthetic manganese(III) porphyrins have also been extensively studied as CYP 450 models in oxygen atom transfer reactions.⁵ Although manganese porphyrins have shown promise as versatile catalysts in oxidation reactions over the past two decades, only recently have the key manganese(V)-oxo porphyrin intermediates been isolated, spectroscopically characterized, and studied in oxidation reactions.^6–9 Groves and co-workers reported the generation and characterization of the first manganese(V)-oxo porphyrin complexes in aqueous solution and the reactivities of these complexes in olefin epoxidation and in the oxidation of bromide and nitrite ions.⁶ Subsequently, Nam and co-workers demonstrated that a manganese(V)-oxo porphyrin can be generated with a biologically relevant oxidant, H₂O₂, in aqueous solution and that the formation of the intermediate depends markedly on the pH of reaction solutions.⁷ Very recently, Newcomb and co-workers reported the generation of manganese(V)-oxo complexes via laser flash photolysis methods in organic solvents and the kinetic studies of the intermediates in olefin epoxidation and alkane hydroxylation.⁸ Naruta and co-workers synthesized a dinuclear Mn^V=O porphyrin complex in the presence of base in organic solvents that showed an O₂-evolution via O-O bond formation between the manganese(V)-oxo moieties.⁹ In addition to the manganese(V)-oxo porphyrins, manganese(V)-oxo complexes bearing non-porphyrinic macrocycles, such as corrole and corrolazine, have been isolated and characterized.^10–15 Interestingly, the manganese(V)-oxo complexes of non-porphyrin ligands are very stable at room temperature and are poor oxidants in oxygen atom transfer reactions. Thus, insight into the chemical properties of the long-sought manganese(V)-oxo intermediates has been obtained by the recent developments in isolating and characterizing manganese(V)-oxo complexes of porphyrin and non-porphyrin ligands. In this paper, we report the generation and characterization of manganese(V)-oxo porphyrin complexes that are stable at room temperature in the presence of base in organic solvents. The stability and reactivities of the manganese(V)-oxo porphyrins have been investigated in detail in oxygenation reactions under stoichiometric and catalytic conditions.

Results and Discussion

Preparation and Characterization of Manganese(V)-Oxo Porphyrin Complexes

Addition of m-chloroperbenzoic acid (m-CPBA) to a reaction solution containing a manganese(III) porphyrin chloride, Mn(TDCPP)Cl (Figure 1 for the structures of manganese porphyrin complexes used in this study),¹⁶ and tetrabutylammonium hydroxide (TBAH) in a solvent mixture of CH₂Cl₂ and CH₃CN (1:1) at 25 °C resulted in the immediate generation of a manganese(V)-oxo porphyrin complex, [(TDCPP)Mn^V=O]⁺ 1a, with a strong and sharp Soret band at 444 nm and a Q band at 560 nm (Figure 2a; Supporting Information, Figure S1 for UV-vis spectra of other manganese(V)-oxo porphyrins; Table 1 for the data of UV-vis absorption bands).^6–9 The intermediate persisted for several hours (t_1/2 ~40 min) at 35 °C. The formation of 1a was also observed when Mn(TDCPP)Cl was treated with other oxidants such as iodosylarenes (i.e., PhIO and F₅PhIO) and H₂O₂ under the identical conditions.

Figure 1.

Structure of manganese(III) porphyrin complexes used in this study.

Figure 2.

(a) UV-vis spectra of Mn(TDCPP)Cl (0.15 mM) (red-colored dashed line), Mn(TDCPP)Cl (0.15 mM) in the presence of TBAH (3 mM) (blue-colored dotted line), and 1a (0.15 mM) (black-colored solid line). Inset shows a magnification of the Q-band region of the manganese porphyrins (1.5 mM). (b) ¹H NMR spectrum of 1a. 1a was prepared by reacting Mn(TDCPP)Cl (1.5 mM) with H₂O₂ (3 mM) in the presence of TBAH (30 mM) in CD₃CN at ambient temperature. Chemical shifts (in ppm) were referenced to TMS (δ = 0.0 ppm).

Table 1.

UV-vis Absorption Bands and Rate Constants for the Natural Decay of [(Porp)Mn^V=O]⁺ Complexes

	absorption bands		rate of natural decay, kobs (s−1)a
(Porp)MnV=O	Soret band (nm), log ε	Q-bands (nm), log ε
[(TPFPP)MnV=O]+	434 nm, 5.1(3)	553, 4.2(3)	1.3(2) × 10−4
[(TDFPP)MnV=O]+	436, 5.1(6)	552, 4.2(6)	1.9(3) × 10−4
[(TDCPP)MnV=O]+ 1a	444, 5.1(6)	560, 4.1(4)	4.6(4) × 10−4
[(TDMPP)MnV=O]+	438, 5.1(5)	562, 4.0(6)	8.5(6) × 10−4
		599, 3.8(3)
[(TMP)MnV=O]+	441, 5.1(3)	561, 4.0(4)	1.6(2) × 10−3
		598, 3.7(4)

^aFirst-order rate constants, k_obs, were determined by monitoring absorbance changes of Q-bands of [(Porp)Mn^V=O]⁺ (1.5 mM) in a 0.1-cm UV cell at 35 °C.

The manganese(V)-oxo porphyrin complex, 1a, was further characterized by EPR, ¹H and ¹⁹F NMR, resonance Raman, and X-ray absorption spectroscopy. The X-band EPR of 1a shows no signal (data not shown).^6b,7,9 The ¹H NMR spectrum of 1a in CD₃CN displays sharp ¹H resonances in the normal aromatic region (δ7 – 9 ppm) (Figure 2b). The β-pyrrole proton appears at 8.7 ppm as a sharp singlet, and the meta- and para-protons of the phenyl group appear at 7.9 and 8.1 ppm with the integral ratio of 2:1, respectively. Similarly, the ¹H NMR spectrum of [(TDFPP)Mn^V=O]⁺ exhibits sharp signals for β-pyrrole proton at 8.9 ppm and the meta- and para-protons of the phenyl group at 7.7 and 8.1 ppm with the integral ratio of 2:1, respectively (Supporting Information, Figure S2a). Further, the [(TDFPP)Mn^V=O]⁺ complex shows a sharp ¹⁹F NMR peak at −109 ppm for the phenyl fluorine (Supporting Information, Figure S2b).^11b,12 Such well-resolved signals in the ¹H and ¹⁹F NMR spectra demonstrate that 1a and other manganese(V)-oxo complexes are diamagnetic low-spin, d² species.^{6b,10b,11b,12}

Resonance Raman Analysis

The resonance Raman spectrum of 1a, measured in CH₃CN at ambient temperature with a 442-nm laser excitation, exhibited an isotope-sensitive band at 759 cm⁻¹, which shifted to 724 cm⁻¹ when 1a-¹⁸O was generated with H₂¹⁸O₂ or upon addition of H₂¹⁸O to the solution of 1a-¹⁶O (vide infra) (Figure 3a).⁹ The observed isotopic shift of 35 cm⁻¹ with ¹⁸O-substitution is in good agreement with the calculated value (Δν_calc = −34 cm⁻¹) from the Mn-O diatomic harmonic oscillator. Further, the observed Mn-O frequency at ~759 cm⁻¹ is consistent with double bond character between the manganese(V) ion and the oxygen atom (vide infra). Resonance Raman spectra were also obtained for [(TDFPP)Mn^V=O]⁺ and [(TPFPP)Mn^V=O]⁺. These data exhibit ν_Mn=O frequencies of 755 and 753 cm⁻¹, respectively, indicating that the Mn^V=O stretching frequency is not sensitive to the porphyrin ligand environment.¹⁷ It is worth noting that the observed Mn^V=O stretching frequency of 1a is lower than that of dinuclear (OH)Mn(V)=O species (ν_Mn=O = 791 cm⁻¹)⁹ but similar to that of Mn^IV(TMP)(O) (ν_Mn=O = 754 cm⁻¹).¹⁸ In addition, when we prepared the manganese(V)-oxo porphyrin complex, [(TM-2-PyP)Mn^V=O]⁵⁺,¹⁶ in aqueous solution as reported by Groves and co-workers^6b,6c and measured the resonance Raman band of the Mn-O moiety, we observed a strong, isotope-sensitive band at 727 cm⁻¹, which shifted to 696 cm⁻¹ when the manganese(V)-oxo porphyrin was generated in H₂¹⁸O (Figure 3b; Supporting Information, Figure S3 for UV-vis spectrum). A weak, isotope-sensitive band at 502 cm⁻¹, which shifted to 475 cm⁻¹ upon introduction of ¹⁸O, was also detected in the resonance Raman spectrum (Figure 3b). The bands at 729 and 502 cm⁻¹ are assigned to ν(Mn^V=O) and ν(Mn^V-OH), respectively.⁹ Different from the manganese(V)-oxo porphyrins, manganese(V)-oxo complexes bearing non-porphyrinic macrocycles exhibit a Mn-O stretching band at a higher frequency (e.g., ~980 cm⁻¹) in organic solvents, which indicates triple bond character between Mn(V) ion and oxo group.^10a,14b,14c The triple bond character of Mn-O moieties in 5-coordinate non-porphyrinic manganese(V)-oxo complexes has been confirmed by short Mn-O bond lengths (~1.56 Å).^{10c,14a,14c,14d,15} However, we have observed in the present study that the Mn-O stretching Raman bands of manganese(V)-oxo porphyrins are in the range of 760 – 790 cm⁻¹ in organic solvents. This indicates a longer, weaker Mn-O bond which suggests double bond character.⁹ The longer, weaker Mn-O bond (relative to known 5-coordinate Mn(V)-oxo complexes) has been further confirmed by X-ray absorption spectroscopic studies, which suggest that the weakening of the Mn-O bond is due to the presence of a sixth ligand trans to the Mn-oxo bond (vide infra).

Figure 3.

(a) Resonance Raman spectra of 1a-¹⁶O (blue line), 1a-¹⁸O (red line), and the difference between 1a-¹⁶O and 1a-¹⁸O (black line). (b) Resonance Raman spectra of [(TM-2-PyP)Mn^V=¹⁶O]⁵⁺ (blue line), [(TM-2-PyP)Mn^V=¹⁸O]⁵⁺ (red line), and the difference between [(TM-2-PyP)Mn^V=¹⁶O]⁵⁺ and [(TM-2-PyP)Mn^V=¹⁸O]⁵⁺ (black line). See Experimental Section for detailed reaction procedures.

X-ray Absorption Spectroscopy (XAS)/Extended X-ray Absorption Fine Structure (EXAFS) Results

A comparison of the normalized Mn K-edge XAS data for the Mn(III)-, Mn(IV)-, and Mn(V)(TDCPP) (1a) complexes is shown in Figure 4. The rising edge energy clearly increases across this series (increasing by ~2 eV upon oxidation of Mn(III) to Mn(IV), and by an additional ~1 eV ongoing to Mn(V)), consistent with the increasing effective nuclear charge on the manganese. Figure 5 shows a comparison of the normalized Mn K-edge XAS data for 1a to that of the previously reported Na[Mn(V)(HMPAB)(O)] (HMPAB = 1,2-bis(2-hydroxy-2-methylpropan-amido)benzene) complex.¹⁹ The rising edge positions are essentially identical, and thus confirms a Mn(V) oxidation state in the manganese porphyrin complex. However, the pre-edges (at ~6542 eV) of these two complexes are very different. The dramatic decrease in intensity requires an increased coordination number in 1a, compared to the 5-coordinate Na[Mn(V)(HMPAB)(O)] complex.¹⁹ The presence of a strong trans axial ligand would weaken the Mn(V)-oxo bond (thus lowering 4p_z mixing) and would also shift the Mn more into the equatorial plane (lowering 4p_x,p_y mixing), resulting in a more centrosymmetric Mn center. All of these factors contribute to a reduced electric dipole contribution and a weaker pre-edge intensity in 1a,²⁰ compared to the previously reported 5-coordinate Mn(V)-oxo complex.

Figure 4.

Comparison of the normalized Mn K-edge XAS data for the Mn(III)(TDCPP)(OH) (blue), Mn(IV)(TDCPP)(O) (purple), and [Mn(V)(TDCPP)(O)]⁺ (red) complexes.

Figure 5.

Comparison of the normalized Mn K-edge XAS data of the [Mn(V)(TDCPP)(O)]⁺ complex (red) to a 5-coordinate [Mn(V)(HMPAB)(O)]⁻ complex (blue).

The k³-weighted EXAFS data and fits for the Mn(III)-, Mn(IV)-, and Mn(V)(TDCPP) complexes are shown in Figure 6a. A comparison of the corresponding Fourier transforms (k = 2–11 Å⁻¹) is shown in Figure 6b. There are clear changes in the overall beat pattern of the EXAFS data and in the Fourier transforms upon oxidation. The best fits to the data are summarized in Table 2. For the Mn(III) complex, the EXAFS are best fit by inclusion of 5 Mn-N/O interactions at 2.01 Å, with additional outershell contributions from the porphyrin. Attempts to add a sixth ligand resulted in a slightly poorer fit (error increased from 0.45 to 0.54). Fits were also attempted in which the first shell was split into shorter and longer components; however, this resulted in the two distances coalescing to the same value. The EXAFS results are consistent with the Mn(III) complex having an axial hydroxide ligand.

Figure 6.

(a) Comparison of the EXAFS data (solid lines) and the fits to the data (dashed lines) for Mn(III)(TDCPP)(OH) (blue), Mn(IV)(TDCPP)(O) (purple), and [Mn(V)(TDCPP)(O)]⁺ (red). (b) The corresponding non-phase shift corrected Fourier transforms of the Mn(III)(TDCPP)(OH) (blue), Mn(IV)(TDCPP)(O) (purple), and [Mn(V)(TDCPP)(O)]⁺ (red) complexes.

Table 2.

EXAFS Fit Results

Mn(III)			Mn(IV)			Mn(V)
	R (Å)	σ2 (Å2)		R (Å)	σ2 (Å2)		R (Å)	σ2 (Å2)
5 Mn-N/O	2.01	0.0036	6 Mn-N/O	1.99	0.0044	2 Mn-O	1.68	0.0051
						4 Mn-N	2.04	0.0023
8 Mn-C	3.04	0.0036	8 Mn-C	3.01	0.0036	8 Mn-C	3.06	0.0056
16 Mn-C-N	3.14	0.0100	16 Mn-C-N	3.15	0.0073	16 Mn-C-N	3.18	0.0080
4 Mn-C	3.44	0.0013	4 Mn-C	3.41	0.0034	4 Mn-C	3.45	0.0043
16 Mn-C-N	4.32	0.0079	16 Mn-C-N	4.31	0.0071	16 Mn-C-N	4.38	0.0068
ΔE0	−4.61		ΔE0	−0.92		ΔE0	−0.56
errora	0.45		errora	0.44		errora	0.60

^aError is given by Σ[(χ_obsd - χ_calcd)²k⁶]/Σ[(χ_obsd²k⁶].

The Mn(IV) data are best fit by inclusion of 6 Mn-N/O interactions at 1.99 Å, with additional outershell contributions from the porphyrin ring. A fit which included only 5 Mn-N/O interactions gave an essentially identical error value; however, the increased coordination is supported by the decreased pre-edge intensity relative to the Mn(III) complex. The first shell could not be split into two components. A plausible trans-axial ligand bound to a Mn(IV) ion is m-chlorobenzoate (m-CBA) derived from m-CPBA which has been used to generate the Mn(IV)-oxo porphyrin complex (vide infra).

The [Mn(V)(TDCPP)(O)]⁺ complex (1a) is best fit by 2 Mn-O interactions at 1.68 Å and 4 Mn-N interactions at 2.04 Å, with additional outershell contributions from the porphyrin. Inclusion of only a single Mn-O at 1.68 Å results in a significant increase in the error (from 0.60 to 0.99). The larger Debye-Waller value on this component (as compared to the 4 Mn-N interactions at 2.04 Å) may suggest that the two trans axial Mn-O interactions are at slightly different distances; however, their separation is beyond the resolution limits of the data. The absence of the 1.68 Å Mn-O component gives an error of 3.60 and a significant low frequency component in the residual. No similar short component can be fit in the Mn(III) or Mn(IV) complexes.

The short Mn-O distances for the [Mn(V)(TDCPP)(O)]⁺ complex (1a) are 0.13 Å longer than the Mn-O in the five-coordinate Na[Mn(V)(O)(HMPAB)]¹⁹ and other non-porphyrinic Mn(V)-oxo complexes,^{10c,14a,14c,14d,15} which may be attributed to a trans effect due to the presence of the sixth ligand. Qualitatively the increase in distance is consistent with the change in Mn-O stretching frequency from the resonance Raman data. However, quantitatively Badger’s rule would predict an even larger change in distance (1.80 Å based on the change in frequency) than is observed.²¹ Density functional theory (DFT) calculations on 1a, both with and without a trans axial ligand, are currently in progress to understand the correlation between the EXAFS derived distances and the experimental frequencies.

In summary, we have prepared manganese(V)-oxo porphyrins in organic solvents and characterized their physical properties with various spectroscopic techniques. The manganese(V)-oxo porphyrins are diamagnetic low-spin (S = 0) species, as characterized by EPR and ¹H and ¹⁹F NMR spectroscopies. The Mn-O stretching frequency and the Mn-O bond length of 1a, determined by resonance Raman and X-ray absorption spectroscopy, respectively, are both indicative of a longer, weaker bond than has been observed in previously reported Mn(V)-oxo complexes bearing non-porphyrinic ligands. This suggests double bond character between the manganese(V) ion and the oxygen atom in porphyrin systems, as indicated by the weak Mn K-pre-edge and the EXAFS data. The weaker Mn-O bond may be attributed to a trans effect due to the presence of a sixth ligand, whereas the previously reported non-porphyrin Mn(V)-oxo complexes are all 5-coordinate. More detailed investigations including DFT calculations are underway to understand how the ligands of manganese(V)-oxo complexes influence the manganese(V)-oxo bond orders.

Effect of Base on the Stability of Manganese(V)-Oxo Porphyrins

It has been reported previously that the formation of manganese(V)-oxo porphyrins is markedly influenced by the pH of reaction solutions; the intermediates are generated at high pHs in aqueous solution^6,7 or in the presence of base in organic solvents.⁹ We therefore investigated the base effect on the generation of manganese(V)-oxo porphyrins, by carrying out reactions with Mn(TDCPP)Cl and m-CPBA in the presence of different amounts of TBAH. When Mn(TDCPP)Cl was reacted with m-CPBA in the absence of TBAH in a solvent mixture of CH₂Cl₂ and CH₃CN (1/1) at 25 °C, (TDCPP)Mn^IV=O 2a was generated (Supporting Information, Figure S4 for the UV-vis and EPR spectra of 2a).^6,7,22 In the presence of 10 equiv TBAH, 1a was formed but quickly disappeared, and the decay of 1a became slower with the increase of TBAH concentration (Supporting Information, Figure S5). Moreover, when 10 equiv HClO₄ was added to a solution of 1a which was generated in the presence of 20 equiv TBAH, the intermediate immediately reverted back to the starting [Mn^III(TDCPP)]⁺ complex. These results demonstrate that the role of base is to stabilize manganese(V)-oxo porphyrins. Although we do not know the exact role of base in increasing the stability of manganese(V)-oxo species, the reactivity of manganese(V)-oxo complexes may be controlled by the binding of hydroxide ion as an axial ligand (i.e., axial ligand effect).²³ The presence of a sixth trans axial ligand in the Mn(V) complex is supported by the XAS results (vide supra).

Reactivities of Manganese(V)-Oxo Porphyrins

We have shown above that manganese(V)-oxo porphyrins are highly stable in the presence of base; 1a decays slowly with a rate constant of 4.6(4) × 10⁻⁴ s⁻¹ at 35 °C (Table 1). This rate was not dependent on the concentration of [(Porp)Mn^V=O]⁺ species. However, the decay rate of manganese(V)-oxo porphyrins was dependent on the nature of porphyrin ligands; electron-rich manganese porphyrins decayed faster than electron-deficient manganese porphyrins (Table 1 for k_obs values for the decay of manganese(V)-oxo porphyrins).^6b It is of interest to note that the stability order of manganese(V)-oxo porphyrins is opposite to that of iron(IV)-oxo porphyrin π-cation radicals. In iron porphyrins, an iron(IV)-oxo porphyrin π-cation radical bearing an electron-rich porphyrin ligand is more stable than one bearing an electron-deficient porphyrin.^4d,24 Further, the stability order of manganese(V)-oxo porphyrins in the presence of base is opposite to the reactivity and stability order of [(Porp)Mn^V=O]⁺ complexes observed in the absence of base but similar to that of (Porp)Mn^IV=O species.^8b In the latter case, the inverted stability order of manganese(IV)-oxo porphyrins was correlated with ease of the disproportionation of (Porp)Mn^IV=O to [Mn^III(Porp)]⁺ and [(Porp)Mn^V=O]⁺ species.^{8b, 13}

We then investigated the reactivity of 1a in oxygen atom transfer reactions with various substrates such as triphenylphosphine (PPh₃), thioanisole (C₆H₅SCH₃), cyclooctene, and cyclooctane at 35 °C. Upon addition of 20 equiv PPh₃ to the solution of 1a at 35 °C, the intermediate reverted back to the starting [Mn^III(TDCPP)]⁺ complex immediately, and product analysis of the resulting solution revealed that Ph₃PO was produced quantitatively. With thioanisole, 1a reverted back to the starting complex, showing isosbestic points at 496 and 568 nm (Figure 7a). Pseudo-first-order fitting of the kinetic data allowed us to determine the k_obs value to be 7.5 × 10⁻³ s⁻¹ at 35 °C. The pseudo-first-order rate constants increased proportionally with thioanisole concentration, giving a second-order rate constant of 2.6(8) × 10⁻² M⁻¹s⁻¹ (Figure 7b). When pseudo-first-order rate constants were determined with various para-substituted thioanisoles and plotted against σ_p, a good correlation was observed with Hammett ρ value of – 0.65 (Figure 7c). The negative ρ value indicates the electrophilic character of the oxo group of 1a in oxygen atom transfer reactions.²⁵ Further, by determining rate constants from 283 to 308 K, we were able to calculate activation parameters of ΔH^‡ = 7(2) kcal mol⁻¹ and ΔS^‡ = −44(2) cal mol⁻¹ K⁻¹ for the oxidation of para-NH₂-thioanisole by 1a (Figure 7d).

Figure 7.

(a) UV-vis spectral changes of 1a (1.5 mM) upon addition of 100 equiv thioanisole at 35 °C. Inset shows absorbance traces monitored at 560 nm. (b) Plot of k_obs against thioanisole concentration to determine a second-order rate constant. (c) Hammett plot of log k_rel against σ_p of thioanisoles in the reactions of 1a (1.5 mM) and para-X-substituted thioanisoles (10 equiv to 1a) at 35 °C. (d) Plot of first-order-rate constants against 1/T to determine activation parameters.

When the reactivity of 1a was examined in olefin epoxidation and alkane hydroxylation, the rate of the disappearance of 1a was not affected by the addition of cyclooctene and cyclooctane to reaction solutions. Further, product analyses revealed that no oxygenated products were generated in these reactions, indicating that 1a is not capable of oxygenating olefins and alkanes under the conditions. These results are of interest since it has been generally believed that manganese(V)-oxo porphyrin complexes are highly reactive and the sole reactive species in the catalytic oxygenation of olefins and alkanes by manganese(III) porphyrins and terminal oxidants.⁵ The low reactivity of 1a observed in the present study may be ascribed to the binding of an anionic axial ligand (i.e., OH⁻) that would serve to decrease the electrophilicity of the Mn-oxo complex toward organic substrates (vide infra).

We have also investigated the oxygen exchange between the oxo group of 1a and H₂¹⁸O, by incubating 1a in the presence of H₂¹⁸O and then adding PPh₃ to the resulting solution.^26,27 The degree of ¹⁸O exchange was then determined by analyzing ¹⁶O and ¹⁸O percentages in Ph₃PO product (Scheme 1). Figure 8a shows that the amounts of ¹⁸O found in the Ph₃PO product increased with the incubation time of 1a (Supporting Information, Figure S6 for the analysis of ¹⁸O % in Ph₃PO). Figure 8b exhibits that the amounts of ¹⁸O incorporated into the product increased proportionally with the amounts of H₂¹⁸O in reaction solutions. The present results provide direct evidence that manganese(V)-oxo porphyrins exchange their oxygen atom with H₂¹⁸O.^6a,7 Interestingly, we found that the rate of oxygen exchange between manganese(V)-oxo porphyrins and H₂¹⁸O was extremely slow under the conditions; the calculated k_obs value of 4.9(2) × 10⁻² s⁻¹ determined in the presence of base in organic solvents at 25 °C is much slower than the estimated rate (k_exchange ≈ 10³ s⁻¹) of oxo-aqua interchange in an manganese(V)-oxo porphyrin in aqueous solution.^6a

Scheme 1.

Oxygen exchange between Mn(V)-oxo species and H₂¹⁸O.

Figure 8.

Plots of ¹⁸O (%) in Ph₃PO against (a) incubation time and (b) concentration of H₂¹⁸O for the oxygen exchange between 1a and H₂¹⁸O. Reaction conditions: (a) 1a (1.5 mM) was incubated with H₂¹⁸O (1.4 M, 6 µL of 95% ¹⁸O-enriched) in the presence of TBAH (30 mM) in a solvent mixture (0.4 mL) of CH₂Cl₂ and CH₃CN (1:1) at 25 °C. (b) Incubation time was 15 min with different amounts of H₂¹⁸O.

Catalytic Oxygenation of Olefins and Alkanes by Manganese(V)-Oxo Porphyrins and PhIO

We have used manganese(V)-oxo porphyrins as catalysts in olefin epoxidation and alkane hydroxylation by PhIO. We first generated 1a by reacting Mn(TDCPP)Cl with m-CPBA in the presence of 20 equiv TBAH. Then, organic substrates were added to the reaction solution, followed by adding solid PhIO (20 equiv). While the UV-vis spectrum of 1a was retained during the reaction, product analysis of the reaction mixture revealed that epoxides and alcohols were produced in the olefin epoxidation and alkane hydroxylation, respectively (Table 3). In the epoxidation of cyclohexene, cyclohexene oxide was produced predominantly with small amounts of allylic oxidation products such as cyclohexenol and cyclohexenone (entry 1). In cis-stilbene epoxidation, cis-stilbene oxide was the major product with the formation of small amounts of trans-stilbene oxide and benzaldehyde (entry 3). This result indicates that the olefin epoxidation is stereospecific, as reported in the epoxidation of cis-stilbene by Mn(TDCPP)Cl and PhIO.²⁸ Also, as observed in the iron porphyrin-catalyzed epoxidation of trans-stilbene by PhIO,²⁹ only a small amount of trans-stilbene oxide was produced in the epoxidation of trans-stilbene (entry 4). In the hydroxylation of alkanes by 1a and PhIO, equimolar amounts of alcohol and ketone products were produced (entries 5 and 6).

Table 3.

Catalytic Oxygenation of Hydrocarbons by 1a and PhIO^a

entry	substrate	products	yields (%)b,c
1	cyclohexene	cyclohexene oxide	24 ± 4
		cyclohexenol	2 ± 1
		cyclohexenone	3 ± 1
2	cyclooctene	cyclooctene oxide	36 ± 4
3	cis-stilbene	cis-stilbene oxide	27 ± 4
		trans-stilbene oxide	3 ± 1
		benzaldehyde	2 ± 1
4	trans-stilbene	trans-stilbene oxide	4 ± 1
		benzaldehyde	2 ± 1
5	cyclohexane	cyclohexanol	10 ± 3
		cyclohexanone	9 ± 3
6	cyclooctane	cyclooctanol	16 ± 4
		cyclooctanone	13 ± 3

^aReactions were run at least in triplicate, and the data reported represent the average of these reactions. See Experimental section for detailed reaction conditions.

^bYields were determined based on the amounts of PhIO added.

^cNo formation of oxygenated products was detected in the absence of the manganese catalyst.

Since we have demonstrated above that 1a does not oxygenate olefins and alkanes, the observation of the formation of oxygenated products in the catalytic olefin epoxidation and alkane hydroxylation by 1a and PhIO implies the generation of an active oxidant that is different from 1a. Similar to the present results, it has been reported previously that manganese(V)-oxo complexes of corrolazine and corrole ligands are inactive, but the Mn(V)-oxo complexes produce oxygenated products in the oxidation of olefins and sulfides by PhIO.^10a,11b Figure 9 depicts plausible oxidants that may be involved in the oxygenation reactions by 1a and PhIO. Those are (Porp)Mn^V(O)(OIPh) (3), [(Porp)Mn^VI=O]²⁺ (4), [(Porp)Mn^V=O]⁺ with a high-spin Mn(V) state (5), and [(Porp)Mn^V=O(X)]⁺ bearing a different axial ligand from 1a (6) (e.g., H₂O). Structure 3, a PhIO-manganese(V)-oxo adduct, has been frequently suggested as an active oxidant in metal complex-catalyzed oxidation reactions by PhIO, such as in the oxidation of sulfides by (corrolazine)Mn(V)=O and PhIO.^10b,30 Structure 4, a manganese(VI)-oxo porphyrin complex, has not yet been identified, but Golubkov and Gross recently reported the characterization of a (nitrido)Mn(VI) corrole complex.³¹ A high-spin Mn(V)-oxo complex, 5, has been proposed as an active oxidant based on DFT calculations on a (salen)Mn(V)-oxo intermediate in Jacobsen-Katsuki epoxidation reactions.³² According to the DFT calculations, manganese(V)-oxo complexes having different spin states (e.g., singlet, triplet, and quintet) show markedly different reactivities in oxidation reactions. Further, Shaik and co-workers conducted DFT calculations with oxoiron(IV) porphyrin π-cation radicals and proposed that the reactivity of the oxoiron(IV) porphyrins is significantly affected by the spin states of the intermediates (i.e., a low-spin doublet state and a high-spin quartet state).^1c,33 Finally, since the low reactivity of 1a may be due to the binding of hydroxide as an axial ligand, replacement of the axial hydroxide ligand upon addition of PhIO in the catalytic oxygenation reactions could generate an intermediate with high reactivity. Indeed, it has been well documented that the presence of neutral nitrogen bases (e.g., imidazoles) in manganese porphyrin-catalyzed reactions increases product yields dramatically.³⁴ Therefore, structure 6 may be a Mn(V)-oxo porphyrin complex bearing a different axial ligand (e.g., H₂O). At the present time, none of the proposed species in Figure 9 has been identified, and intensive mechanistic studies are needed to elucidate the exact nature of active oxidant(s) in manganese complex-catalyzed oxidation reactions.

Figure 9.

Plausible intermediates involved in the oxygenation of hydrocarbons by [(Porp)Mn^V=O]⁺ and PhIO.

Conclusion

Manganese(V)-oxo species have been frequently invoked as reactive species in the catalytic oxygenation of hydrocarbons by manganese(III) porphyrins and terminal oxidants.⁵ In the present study, we have prepared manganese(V)-oxo porphyrins that are highly stable at room temperature in the presence of base in organic solvents. The manganese(V)-oxo porphyrins were characterized with various spectroscopic techniques and found to be diamagnetic low-spin (S = 0) species with a longer, weaker Mn-O bond than that found in previously characterized Mn(V)-oxo complexes. This is suggestive of double bond character between the manganese(V) ion and the oxygen atom, and originates from the presence of a sixth trans axial ligand. The stability of the manganese(V)-oxo species was found to depend on the concentration of base and the electronic nature of porphyrin ligands, but not on the concentration of the manganese(V)-oxo species. The low-spin manganese(V)-oxo porphyrins showed a low reactivity in oxygen atom transfer reactions;³⁵ the intermediates are capable of oxygenating PPh₃ and thioanisoles but not olefins and alkanes. Moreover, the rate of oxygen exchange between the manganese(V)-oxo species and H₂¹⁸O in the presence of base in organic solvents was found to be very slow. These results are in contrast to previous suggestions that manganese(V)-oxo porphyrins are invariably highly reactive in oxygenation reactions and that the intermediates exchange their oxygen with H₂¹⁸O at a fast rate. We have also reported that manganese(V)-oxo porphyrins associated with terminal oxidants, such as PhIO, afforded high product yields in the oxygenation of olefins and alkanes. Future studies will focus on elucidating the effect(s) of base on the chemical properties of manganese(V)-oxo species and the nature of oxygenating intermediate(s) generated in the reactions of manganese(V)-oxo porphyrins and terminal oxidants.

Experimental Section

Materials

Dichloromethane (anhydrous) and acetonitrile (anhydrous) were obtained from Aldrich Chemical Co. and purified by distillation over CaH₂ prior to use. All reagents purchased from Aldrich were the best available purity and used without further purification unless otherwise indicated. m-CPBA was purified by washing with phosphate buffer (pH 7.4) followed by water and then dried under reduced pressure. Iodosylarenes were prepared according to published procedures.³⁶ The purities of the oxidants were determined by iodometric titration.³⁷ H₂¹⁸O (95% ¹⁸O-enriched) and H₂¹⁸O₂ (90% ¹⁸O-enriched, 2% H₂¹⁸O₂ in water) were purchased from ICON Services Inc. (Summit, NJ, USA). Mn(TDCPP)Cl, Mn(TDFPP)Cl, Mn(TPFPP)Cl, Mn(TMP)Cl, and Mn(TM-2-PyP)Cl₅ were obtained from Mid-Century Chemicals (Posen, Il, USA). Mn(TDMPP)Cl was synthesized according to published procedures.³⁸

Instrumentation

UV-vis spectra were recorded on a Hewlett Packard 8453 spectrophotometer equipped with a circulating water bath. EPR spectra were obtained on a JEOL JES-FA200 spectrometer at 4 K. ¹H NMR spectra were measured with Bruker DPX-250 spectrometer, and chemical shifts were reported as δ values from standard solvent peaks. ¹⁹F NMR spectrum was measured with Varian Unity-Inova 500 MHz spectrometer. Product analyses for the oxidation of PPh₃ and the epoxidation of cis- and trans-stilbenes were performed on DIONEX Summit Pump Series P580 equipped with a variable wavelength UV-200 detector (HPLC). Products were separated on Waters Symmetry C18 reverse phase column (4.6 × 250 mm), and detection was made at 215 and 254 nm. Product analyses for the oxidation of sulfides, the epoxidation of cyclohexene and cyclooctene, and the hydroxylation of alkanes were performed on Agilent Technologies 6890N gas chromatograph equipped with a FID detector (GC) and Hewlett-Packard 5890 II Plus gas chromatograph interfaced with Hewlett-Packard model 5989B mass spectrometer (GC-MS). LC-ESI MS spectra for the determination of ¹⁸O percentage in Ph₃PO in isotopically labeled H₂¹⁸O experiments were collected on a Finnigan Surveyor Integrated HPLC systems (PDA detector and LC pump) connected with Thermo Finnigan (San Jose, CA, USA) LCQ™ Advantage MAX quadrupole ion trap instrument. The separation of product was achieved by on-column injection to a Hypersil GOLD column (5 µm, 4.6 × 250 mm) using a MeOH:H₂O (3:1) eluent at a flow rate of 1 mL/min, at the spray voltage 4.7 kV and the capillary temperature at 220 °C.

Preparation of Manganese(V)- and Manganese(IV)-Oxo Complexes

In general, 1 (0.15 mM) was prepared by adding 2 equiv of m-CPBA (0.3 mM, diluted in 50 µL of CH₃CN) into a 0.1-cm UV cuvette containing a manganese(III) porphyrin chloride (0.15 mM) and TBAH (3.0 mM) in a solvent mixture (0.5 mL) of CH₃CN and CH₂Cl₂ (1:1) at 25 °C. 2 was generated by reacting Mn(TDCPP)Cl with 2 equiv of m-CPBA in a solvent mixture (0.5 mL) of CH₃CN and CH₂Cl₂ (1:1) at 25 °C. The formation of 1 and 2 was monitored by a UV-vis spectrophotometer, and the resulting solution was used immediately for further studies.

Resonance Raman Measurements

Samples for resonance Raman were prepared as follows: 1a-¹⁶O was prepared by reacting Mn(TDCPP)Cl (0.4 mM) with m-CPBA (0.8 mM) in the presence of TBAH (8 mM) in CH₃CN (0.5 mL) at 10 °C, whereas 1a-¹⁸O was prepared by reacting Mn(TDCPP)Cl (0.4 mM) with H₂¹⁸O₂ (0.8 mM) or exchanging the oxygen atom of 1a-¹⁶O with H₂¹⁸O (15 µL of 95% ¹⁸O-enriched). The [(TM-2-PyP)Mn^V=¹⁶O]⁵⁺ and [(TM-2-PyP)Mn^V=¹⁸O]⁵⁺ intermediates were prepared by reacting [Mn^III(TM-2-PyP)]⁵⁺ (0.4 mM) with PhI¹⁶O and PhI¹⁸O (0.8 mM, diluted in 10 µL of CH₃OH) in the presence of TBAH (8 mM) in H₂¹⁶O (0.5 mL) and H₂¹⁸O (0.5 mL), respectively, at 10 °C. The absence of Mn(IV)-oxo species was confirmed by taking the UV-vis spectra of resonance Raman samples. The samples were transferred in a quartz spinning cell pre-cooled at 10 °C. Resonance Raman spectra were obtained using a liquid nitrogen cooled CCD detector (model LN/CCD-1100-PB, Roper Scientific) attached to a 1-m single polychromator (model MC-100DG, Ritsu Oyo Kogaku). An excitation wavelength of 441.6-nm was provided by a He-Cd laser (model CD4805R, Kinmon Electric), with 4 mW power at the samples. All measurements were carried out with a quartz spinning cell (1000 rpm) at ~ 10 °C. Raman shifts were calibrated with indene, and the accuracy of the peak positions of the Raman bands was ±1 cm⁻¹.

XAS Data Collection and Analysis

XAS data for Mn(III)(TDCPP)(OH), Mn(IV)(TDCPP)(O), and [Mn(V)(TDCPP)(O)]⁺ (1a) were recorded at the Stanford Synchrotron Radiation Laboratory (SSRL) on focused beam line 9–3, under ring conditions of 3 GeV and 80–100 mA. A Si(220) monochromator (fully tuned) was used for energy selection. A Rh-coated mirror (set to a cutoff of 10 keV) was used for harmonic rejection. All XAS samples (~2 mM) were measured as solutions in CH₃CN. Samples were loaded into 2-mm Delrin XAS cells with Kapton windows and then frozen immediately in liquid nitrogen prior to XAS measurements. During XAS measurements, samples were maintained at a constant temperature of 10 K by an Oxford Instruments CF1208 continuous-flow liquid-helium cryostat.

Data were measured in fluorescence mode using a Canberra Ge 30-element array detector. XAS data were measured to k = 11 Å⁻¹ due to contributions from both Fe contamination and diffraction (resulting from a poor glass). Internal energy calibration was performed by simultaneous measurement of the absorption of a Mn foil placed between a second and third ionization chamber. The first inflection point of the Mn foil was assigned to 6539.0 eV. Samples were monitored for photoreduction throughout the course of data collection. Only those scans, which showed no evidence of photoreduction, were used in the final average. The data represent 8, 6, and 6 scan averages for the Mn(III)(TDCPP)(OH), Mn(IV)(TDCPP)(O), and [Mn(V)(TDCPP)(O)]⁺ complexes, respectively.

The averaged data were processed as described previously³⁹ by fitting a second-order polynomial to the post-edge region and subtracting this background from the entire spectrum. A three-region cubic spline was used to model the smooth background above the edge. Normalization of the data was achieved by subtracting the spline and normalizing the post-edge region to 1. The resultant EXAFS was k³-weighted to enhance the impact of high-k data.

Theoretical EXAFS signals χ(k) were calculated using FEFF (version 7.0)⁴⁰ and fit to the data using EXAFSPAK.⁴¹ The non-structural parameter E₀ was also allowed to vary but was restricted to a common value for every component in a given fit. The structural parameters varied during the refinements were the bond distance (R) and the bond variance (σ²). The σ² is related to the Debye-Waller factor, which is a measure of thermal vibration and to static disorder of the absorbers/scatters. Coordination numbers were systematically varied in the course of the analysis, but they were not allowed to vary within a given fit. Single scattering paths and the corresponding multiple scattering paths were linked during the course of refinements.

Reactions of [(TDCPP)Mn^V=O]⁺ (1a) with Organic Substrates

All reactions were run in a 0.1-cm UV cuvette by monitoring UV-vis spectral changes of reaction solutions. 1a was prepared by reacting Mn^III(TDCPP)Cl (1.5 mM) with 2 equiv m-CPBA (3 mM) in the presence of TBAH (30 mM) in a solvent mixture (0.4 mL) of CH₂Cl₂/CH₃CN (1/1) at 35 °C. Substrate (0.15 M) was then added to the solution of 1a. After the completion of the reaction, the reaction mixture was directly analyzed by HPLC, GC, and/or GC-MS. Product yields were determined by comparison against standard curves prepared with known authentic samples.

The labeled water, H₂¹⁸O, experiments for oxygen exchange between 1a and H₂¹⁸O were carried out as follows: 1a was prepared as described above. Then, appropriate amounts of H₂¹⁸O (95% ¹⁸O-enriched) were added to the solution of 1a, followed by incubating the resulting solution for the given time. After addition of PPh₃ (0.15 M) to the solution of 1a, the ¹⁸O-percentage in Ph₃PO product was determined by analyzing reaction solutions with LC-ESI MS, and the ¹⁶O and ¹⁸O compositions in Ph₃PO product were analyzed by the relative abundances of m/z = 279.2 for [Ph₃P¹⁶O + H⁺]⁺ and m/z = 281.2 for [Ph₃P¹⁸O + H⁺]⁺.

Catalytic Oxygenation of Organic Substrates by 1a and PhIO

1a was prepared by reacting Mn^III(TDCPP)Cl (1.5 mM) with 2 equiv m-CPBA (3 mM) in the presence of TBAH (30 mM) in a solvent mixture (0.5 mL) of CH₂Cl₂/CH₃CN (1/1) at 25 °C. Substrate (0.15 M, diluted in CH₂Cl₂/CH₃CN (1/1)) was added to the reaction solution, followed by the addition of solid PhIO (20 equiv to the catalyst). For comparison, control experiments were carried out in the absence of the manganese catalyst in the oxidation of cyclohexene and cyclohexane. After the solid PhIO disappeared completely (ca. 2 h), the resulting solution was directly analyzed by GC, HPLC, and/or GC-MS. Product yields were determined by comparison against standard curves prepared with known authentic samples.