Dioxygen Reactivity of Copper(I)/Manganese(II)-Porphyrin Assemblies: Mechanistic Studies and Cooperative Activation of O2

The oxidation of transition metals such as manganese and copper by dioxygen (O2) is of great interest to chemists and biochemists for fundamental and practical reasons. In this report, the O2 reactivities of 1:1 and 1:2 mixtures of [(TPP)MnII] (1; TPP: Tetraphenylporphyrin) and [(tmpa)CuI(MeCN)]+ (2; TMPA: Tris(2-pyridylmethyl)amine) in 2-methyltetrahydrofuran (MeTHF) are described. Variable-temperature (−110 °C to room temperature) absorption spectroscopic measurements support that, at low temperature, oxygenation of the (TPP)Mn/Cu mixtures leads to rapid formation of a cupric superoxo intermediate, [(tmpa)CuII(O2•–)]+ (3), independent of the presence of the manganese porphyrin complex (1). Complex 3 subsequently reacts with 1 to form a heterobinuclear μ-peroxo species, [(tmpa)CuII–(O22–)–MnIII(TPP)]+ (4; λmax = 443 nm), which thermally converts to a μ-oxo complex, [(tmpa)CuII–O–MnIII(TPP)]+ (5; λmax = 434 and 466 nm), confirmed by electrospray ionization mass spectrometry and nuclear magnetic resonance spectroscopy. In the 1:2 (TPP)Mn/Cu mixture, 4 is subsequently attacked by a second equivalent of 3, giving a bis-μ-peroxo species, i.e., [(tmpa)CuII−(O22−)−MnIV(TPP)−(O22−)−CuII(tmpa)]2+ (7; λmax = 420 nm and δpyrrolic = −44.90 ppm). The final decomposition product of the (TPP)Mn/Cu/O2 chemistry in MeTHF is [(TPP)MnIII(MeTHF)2]+ (6), whose X-ray structure is also presented and compared to literature analogs.


Introduction
Dioxygen (O 2 ) binding, reduction, and activation at metalloenzyme active sites are vital for aerobic life. Manganese-containing enzymes, in particular, facilitate a wide variety of biological redox processes through interactions of manganese with O 2 and its reduced derivatives such as superoxide (O 2 •-) and peroxide (O 2 2-) [1][2][3]. Because of its relative abundance and multiple accessible oxidation states, manganese can facilitate biochemical multielectron redox conversions and is essential for a variety of biological redox processes such as photosynthetic O 2 evolution, antioxidant defense mechanisms regulating reactive oxygen species (ROS), and DNA synthesis [1][2][3][4]. Several Mn-containing enzymes involved in facilitating these redox processes include manganese superoxide dismutase (SOD), manganese catalase, the oxygen-evolving complex (OEC) in photosystem II, and Mn/Mn or Mn/Fe ribonucleotide reductase (RNR) [2,3,5]. The nuclearity of such manganese active sites varies, with more complex redox processes typically occurring at multinuclear sites with two or more metal centers [6,7].

Dioxygen Chemistry of [(TPP)Mn II ]
In coordinating solvents such as 2-methyltetrahydrofuran (MeTHF), [(TPP)Mn II ] is rather unreactive toward O 2 at room temperature. As shown in Figure 1a, despite long O 2 exposure, the UV-vis spectrum of the reaction mixture mostly retained the Soret (433 nm) and Q-band (568 and 606 nm) absorption characteristics of the manganous complex. The UV-vis spectroscopic changes of [(TPP)Mn II ] oxygenation in MeTHF at −90 • C are shown in Figure 1b. At low temperature, a small but noticeable decrease in the 434 nm Soret band, along with the formation of a minor (TPP)Mn III species (λ max = 470 nm), were observed. Allowing the solution to warm to room temperature yielded a product with features at 378, 400, 423, and 470 nm. We posit this final product as a [(TPP)Mn III (MeTHF) 2 ] + species, based on its characteristic "split Soret band" and other absorption features similar to those of an authentic MeTHF solution of [(TPP)Mn III (THF) 2 ]SbF 6 (λ max = 377, 399, 418, 467 nm) ( Figures S1 and S2). We note that in the high-spin manganese(III) porphyrins, the normally intense Soret band splits into two less-intense bands; one of which is a prominent peak at lower energy (~470 nm), and the other one is a very broad band that occurs at higher energy (~380 nm). The ratios of bands well as the λ max values for these bands are particularly sensitive to the nature of the axial ligand(s) [32,33]. The X-ray structure of [(TPP)Mn III (MeTHF) 2 ]SbF 6 , as well as 1 H-NMR and IR data further support the identity of this product species as [(TPP)Mn III (MeTHF) 2 ] + (vide infra).   Investigations of the independent O 2 chemistry for the copper complex, [(tmpa)Cu I (MeCN)][B(C 6 F 5 ) 4 ], have been described previously [28,30]. The O 2 adduct observed from  low-temperature reactions of this copper complex bearing the tripodal TMPA ligand with  O 2 was solvent-, temperature-, and (Figure 2b). We formulated this as a heterobinuclear peroxo complex, [(tmpa)Cu II -(O 2 2-)-Mn III (TPP)] + . The reaction proceeded through initial formation of a cupric superoxo species, [(tmpa)Cu II (O 2 •-)] + , followed by a fast electron transfer from the manganous center to the superoxide moiety, forming the bridged peroxo species. This µ-peroxo Mn III /Cu II complex was not stable at higher temperatures, primarily leading to the formation of a µ-oxo complex, [(tmpa)Cu II -O-Mn III (TPP)] + . By analogy to the process shown to occur in similar bridged peroxo systems, it was presumed that this transformation occurred through a disproportionation reaction of two µ-peroxo species that generated two µ-oxo complexes and released O 2 [27,34].
It is important to note that, under the given experimental conditions, i.e., low temperature (−110 • C) and low concentration (µM range), the oxygenation reaction of [(tmpa)Cu I (MeCN)][B(C 6 F 5 ) 4 ] exclusively resulted in the formation of the cupric superoxo species [29,30,35]. This strongly suggests that the electron transfer from a second cuprous complex to [(tmpa)Cu II (O 2 •-)] + is significantly less favored than from a [(TPP)Mn II ] complex. We can, therefore, rule out the formation of the dicopper peroxo species, i.e., The remaining question about the capability of the manganese(II) porphyrin precursor to react with more than an equimolar amount of the superoxide was addressed by reacting a 1:2 mixture of [(TPP)Mn II ] and [(tmpa)Cu I (MeCN)][B(C 6 F 5 ) 4 ] with dioxygen. Here, at room temperature, the oxygenation reaction and formation of the final solvated (TPP)Mn III product were faster compared with those of an equimolar mixture, with nearly all of [(TPP)Mn III (MeTHF) 2 ] + formed within 1 h after O 2 bubbling rather than over the 3 h observed for the 1:1 mixture (Figures 2a and 3a).  (Table 1), supporting the oxidation state of manganese(IV) in this intermediate [17,[36][37][38]. Further indication that the bis-µ-peroxo intermediate is a manganese(IV) complex was derived from the resonance observed for its pyrrolic protons using 1 H-NMR spectroscopy, vide infra. Upon warming, the bis-µ-peroxo intermediate may disproportionate to form a bis-µ-oxo adduct,  It is worth mentioning that quantitative analyses of UV-vis spectra of the oxygenation products of [(TPP)Mn II ] in the presence of 0 to 2 equivalents of [(tmpa)Cu I (MeCN)] + , at either low or room temperature, confirmed the generation of one equivalent of the final product, [(TPP)Mn III (MeTHF) 2 ] + , with features near 380, 400, and 470 nm in high yields (~100% yield).

X-ray Structure of [(TPP)Mn III (MeTHF) 2 ]SbF 6
For the present studies, the molecular structure of an authentic sample of the proposed final product, [(TPP)Mn III (MeTHF) 2 ] + , was also obtained. Dark red crystals of [(TPP)Mn III (MeTHF) 2 ]SbF 6 ·2MeTHF were grown by slow diffusion of heptane into a MeTHF solution of [(TPP)Mn III (THF) 2 ]SbF 6 (see Materials and Methods, Table S1). The complex crystallizes in a tetragonal crystal system with the P4 3 2 1 2 space group. A perspective view of the complex, along with the selected structural and geometrical parameters, are given in Figure 4; the molecular packing in the unit cell is shown in Figure S3. The hexacoordinate Mn center lies perfectly in the plane of the porphyrin and is axially ligated by two MeTHF molecules. The length of the bond between the manganese and the axially-ligated MeTHF molecule, Mn−O ax, , of 2.272(3) Å was within the range reported for other hexacoordinate Mn(III) porphyrins with two O-based ligands (Table S2) [42][43][44][45][46][47]. Moreover, the longer Mn−O ax distances, as compared to Mn−N por (i.e., average of 2.008 Å) were in accord with the presence of a high-spin tetragonally elongated Mn(III) center. The elongation of the bonds to the axial oxygen sites has been ascribed to a singly occupied axially antibonding d z 2 orbital that renders a ground electronic state of (d xz , d yz ) 2 (d xy ) 1 (d z2 ) 1 [48][49][50].

Nuclear Magnetic Resonance (NMR) Spectroscopy
The  Upon O 2 bubbling, several new signals appeared in the upfield and downfield regions in the spectrum, which resulted from oxidation of the two metal centers ( Figure 5 Left and Figure S11). After 1 min of O 2 bubbling, the 1 H-NMR spectrum showed a distinct pyrrolic peak at δ = 36.06 ppm, which corresponds to the presence of a peroxo Mn(III) center [24,51], in agreement with our UV-vis spectroscopic results (vide supra). This proposed µ-peroxo Mn III /Cu II intermediate was unstable and the corresponding pyrrolic resonance disappeared over time (Figure 5 Left).
The upfield region displayed the three most intense peaks at δ = −4.23, −19.60, and −34.65 ppm. The signal at δ = −4.23 ppm likely originated from the TMPA moiety of the attached copper center as the significant upfield shift of this peak indicated that the protons of this chelate were directly located above the porphyrin core, as observed in similar oxo-bridged heterobinuclear systems [52][53][54]. The second upfield broad signal at δ = −19.60 ppm could be ascribed to [(TPP)Mn III −O−Cu II (tmpa)] + and may correspond to a combination of the pyrrolic, as well as part of the attached, TMPA protons. The third upfield signal at δ = −34.65 ppm corresponded to the pyrrolic protons of [(TPP)Mn III (THF) 2 ] + in the reaction mixture, which was identical to that of the authentic [(TPP)Mn III (THF) 2 ] + sample, confirming the presence of a high-spin (S = 2) Mn(III) center ( Figure S8). All proton peaks were integrated with respect to the δ = −34.65 ppm peak for relative comparisons. As the reaction progressed, the signal at δ = −34.65 ppm continued growing and the intensities of the other signals in the upfield region decreased, which suggested decomposition to [(TPP)Mn III (THF) 2 ] + , consistent with the overall reactivity patterns observed in our UV-vis studies. Appearance of the small peaks in the δ = 10 to 26 ppm region over time ( Figure 5 left and Figure S11b-g Left) that correspond to the TMPA protons in a free [(tmpa)Cu II (X)] + complex [52] further confirmed the dissociation of the Mn/Cu assembly.
To further support our supposition that an individual homobinuclear copper species did not form during the oxygenation of the mixture of manganese and copper complexes, we also monitored the independent dioxygen reactivity of [(tmpa)Cu I (MeCN)][B(C 6 F 5 ) 4 ] through 1 H-NMR spectroscopy ( Figures S12 and S13). The cuprous complex showed three distinct signals between δ = 8.84 and 7.48 ppm, corresponding to the protons of the pyridyl arms, while the methyl-protons resonated at δ   Figure 6), which further confirms the formation of the Mn/Cu assembly. To try to provide further insight, the oxygenation of a 1:2 mixture of the reduced complexes did not lead to the mass spectrometric detection of a bis-µ-oxo adduct, most likely due to its even lower stability and, therefore, easier reduction/fragmentation.  (Figures S14-S17). The presence of bound MeTHF in [(TPP)Mn III (MeTHF) 2 ] + was confirmed by C-H stretching bands at or near 2966 and 2864 cm −1 . The high-frequency metal-sensitive IR bands of the TPP complexes have been previously found to, slightly but regularly, vary depending on the nature of the metal ion [56]. Here, the metal-sensitive TPP bands at 1598, 1487, 1342, and 1010 cm −1 were identical or within 2 cm −1 for the 1:1 and 1:2 oxygenation products, and similar to those of authentic [(TPP)Mn III (MeTHF) 2 ]SbF 6 (Table S3), further confirming the presence of a Mn(III) center.

Conclusions
The foregoing results demonstrate the cooperative activation of dioxygen by the copper(I)/manganese(II)-porphyrin systems. A combined array of variable-temperature UV-vis, NMR, IR as well as ESI-MS and X-ray crystallographic analyses collectively

General Methods
Chemicals were purchased commercially and used without further purification, unless noted otherwise. Acetonitrile (MeCN), dichloromethane (DCM), tetrahydrofuran (THF), and 2-methyltetrahydrofuran (MeTHF) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deuterated solvents (acetone-d 6 , CD 2 Cl 2 , and THF-d 8 ) were purchased from Cambridge Isotope Laboratories (Tewksbury, MA, USA). Commercial ACS grade solvents were used for chromatography and extractions. For the reactions, all solvents were purified by an Innovative Technologies (Newburyport, MA, USA) or Inert PureSolv Micro (Amesbury, MA, USA) solvent purification system. Solvents were then deoxygenated by bubbling with argon for 1 h, followed by storage over 3 or 5 Å molecular sieves for at least 72 h prior to use. Deionized water was purified by a PURELAB flex 1 Analytical Ultrapure Water System (ELGA) to obtain nanopure water with a specific resistance of 18.2 MΩ cm at room temperature. Air-sensitive compounds were prepared and handled under a dry, oxygen-free argon atmosphere using standard Schlenk techniques or under nitrogen atmosphere in a Vacuum Atmospheres (Hawthorne, CA, USA) OMNI-Lab inert atmosphere (<0.5 ppm of O 2 and H 2 O) glovebox. Ultra-high purity grade oxygen gas was purchased from Airgas (Greensboro, NC, USA) and dried by passing through a drying column containing Drierite desiccant and 3 Å activated molecular sieves. For the NMR experiments, dry O 2 gas was transferred and stored in a capped 50-mL Schlenk flask, then added into the metal complex solutions via a three-way long syringe needle.
UV-vis absorption spectra were recorded on an Agilent (Wilmington, DE, USA) Cary-60 spectrophotometer equipped with a Unisoku (Osaka, Japan) CoolSpeK USP-203-B cryostat using 4-mm modified Schlenk cuvettes. Infrared (IR) spectra of neat solid samples were obtained using a Thermo Scientific (West Palm Beach, FL, USA) Nicolet iS5 Fourier Transform IR (FT-IR) spectrometer equipped with an iD7 attenuated total reflection (ATR) accessory. NMR spectra were recorded on a JEOL (Peabody, MA, USA) 500 MHz spectrometer, and the chemical shifts were referenced against NMR solvent residual shifts (e.g., THF-d 8 at s hased commercially and used without further purification, unetonitrile (MeCN), dichloromethane (DCM), tetrahydrofuran hydrofuran (MeTHF) were purchased from Sigma-Aldrich (St. ated solvents (acetone-d6, CD2Cl2, and THF-d8) were purchased Laboratories (Tewksbury, MA, USA). Commercial ACS grade hromatography and extractions. For the reactions, all solvents ative Technologies (Newburyport, MA, USA) or Inert PureSolv SA) solvent purification system. Solvents were then deoxygenon for 1 h, followed by storage over 3 or 5 Å molecular sieves e. Deionized water was purified by a PURELAB flex 1 Analytical (ELGA) to obtain nanopure water with a specific resistance of perature. Air-sensitive compounds were prepared and handled argon atmosphere using standard Schlenk techniques or under Vacuum Atmospheres (Hawthorne, CA, USA) OMNI-Lab inert O2 and H2O) glovebox. Ultra-high purity grade oxygen gas was reensboro, NC, USA) and dried by passing through a drying te desiccant and 3 Å activated molecular sieves. For the NMR as transferred and stored in a capped 50-mL Schlenk flask, then plex solutions via a three-way long syringe needle.

General Methods
Chemicals were purchased commercially and used without further purification, unless noted otherwise. Acetonitrile (MeCN), dichloromethane (DCM), tetrahydrofuran (THF), and 2-methyltetrahydrofuran (MeTHF) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deuterated solvents (acetone-d6, CD2Cl2, and THF-d8) were purchased from Cambridge Isotope Laboratories (Tewksbury, MA, USA). Commercial ACS grade solvents were used for chromatography and extractions. For the reactions, all solvents were purified by an Innovative Technologies (Newburyport, MA, USA) or Inert PureSolv Micro (Amesbury, MA, USA) solvent purification system. Solvents were then deoxygenated by bubbling with argon for 1 h, followed by storage over 3 or 5 Å molecular sieves for at least 72 h prior to use. Deionized water was purified by a PURELAB flex 1 Analytical Ultrapure Water System (ELGA) to obtain nanopure water with a specific resistance of 18.2 MΩ cm at room temperature. Air-sensitive compounds were prepared and handled under a dry, oxygen-free argon atmosphere using standard Schlenk techniques or under nitrogen atmosphere in a Vacuum Atmospheres (Hawthorne, CA, USA) OMNI-Lab inert atmosphere (<0.5 ppm of O2 and H2O) glovebox. Ultra-high purity grade oxygen gas was purchased from Airgas (Greensboro, NC, USA) and dried by passing through a drying column containing Drierite desiccant and 3 Å activated molecular sieves. For the NMR experiments, dry O2 gas was transferred and stored in a capped 50-mL Schlenk flask, then added into the metal complex solutions via a three-way long syringe needle.
UV-vis absorption spectra were recorded on an Agilent (Wilmington, DE, USA) Cary-60 spectrophotometer equipped with a Unisoku (Osaka, Japan) CoolSpeK USP-203-B cryostat using 4-mm modified Schlenk cuvettes. Infrared (IR) spectra of neat solid samples were obtained using a Thermo Scientific (West Palm Beach, FL, USA) Nicolet iS5 Fourier Transform IR (FT-IR) spectrometer equipped with an iD7 attenuated total reflection (ATR) accessory. NMR spectra were recorded on a JEOL (Peabody, MA, USA) 500 MHz spectrometer, and the chemical shifts were referenced against NMR solvent residual shifts (e.g., THF-d8 at = 1.72 ppm) and/or tetramethylsilane (TMS at = 0.00 ppm). Electrospray ionization mass spectrometry (ESI-MS) data was collected in positive ion mode on a Thermo Fisher Scientific (Waltham, MA, USA) Q Exactive Plus system.

Synthesis and Characterization
The  [58] were synthesized and characterized following previously described methods.

Synthesis of [(TPP)Mn II ]
= 0.00 ppm). Electrospray ionization mass spectrometry (ESI-MS) data was collected in positive ion mode on a Thermo Fisher Scientific (Waltham, MA, USA) Q Exactive Plus system.

Synthesis of [(TPP)Mn II ]
The complex was prepared following slight modification of the literature procedure reported for the ferric choro complexes bearing similar porphyrin rings [59,60]. Using standard Schlenk techniques, a solution of [(TPP)Mn III Cl] (500 mg, 0.711 mmol) in DCM (200 mL) was mixed with a solution of sodium dithionite (22 g, 0.126 mol) in water (100 mL) for 1 h by bubbling argon. The reaction mixture was allowed to sit for approximately 20 min to allow the separation of the two layers. The DCM layer was filtered through sodium sulfate to remove residual water, then dried under vacuum, producing the deep-purple microcrystalline product. Yield: 416 mg, 91%. UV-vis (λ max , nm and transferred to an NMR tube inside the glovebox. The NMR sample sealed with a rubber septum was taken out, and the spectrum prior to dioxygen bubbling was recorded at room temperature. For oxygenation, 4 mL of dry O 2 gas was bubbled into the metal complex solution in the NMR tube using a Hamilton gastight syringe equipped with a three-way valve. After O 2 bubbling, the first spectrum was recorded within 2 min of mixing, and the reaction was monitored over 18 h ( Figure 5 and Figure S11). The same steps were repeated for the 1:2 mixture NMR sample which was prepared by dissolving 7. In a control experiment, in order to confirm the O 2 reactivity of the authentic mononuclear Cu(I) complex in our experimental conditions, NMR studies were carried out following a slightly modified version of the literature procedure [55]. Inside the glovebox, [(tmpa)Cu I (MeCN)][B(C 6 F 5 ) 4 ] (24.2 mg, 0.022 mmol) was dissolved in THF-d 8 (800 µL) and transferred to an NMR tube. Oxygenation of the Cu(I) complex solution was carried out using methods described above. After O 2 bubbling, the first spectrum was recorded within 2 min of mixing; the reaction was monitored over 6 h (Figures S12 and S13).

Crystallographic Studies
Suitable X-ray quality single crystals of [(TPP)Mn III (MeTHF) 2 ]SbF 6 were obtained by transferring a 7 µM solution of [(TPP)Mn III (THF) 2 ]SbF 6 in MeTHF into a 5-mm glass tube, layering with heptane, and storing it in the glovebox at room temperature. All reflection intensities were measured at 100(2) K using a Gemini R diffractometer (equipped with Atlas detector) with MoKα radiation (λ = 0.71073 Å) under the program CrysAlisPro (Version CrysAlisPro 1.171.38.43f, Rigaku OD, 2015). The same program (but a different version viz. CrysAlisPro 1.171.40.53, Rigaku OD, 2019) was used to refine the cell dimensions and for data reduction. The temperature of the data collection was controlled using the system Cryojet (manufactured by Oxford Instruments, Abingdon, UK). The structure was solved with the program SHELXT-2018/2 and was refined on F 2 by full-matrix leastsquares technique using the SHELXL-2018/3 program package [61]. Numerical absorption correction based on Gaussian integration was applied using a multifaceted crystal model by CrysAlisPro. Non-hydrogen atoms were refined anisotropically. In the refinement, hydrogen atoms were treated as riding atoms using SHELXL default parameters.
CSD 2128108 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via https://www.ccdc.cam.ac.uk/data_request/cif.

Acknowledgments:
The authors gratefully acknowledge financial support provided in the form of startup funds, the Spartans ADVANCE Research Award, and the URSCO Undergraduate Research and Creativity Award (URCA) from the University of North Carolina at Greensboro. The Joint School of Nanoscience and Nanoengineering is acknowledged for providing access to the X-ray diffraction facility.