A Bioinspired Nonheme FeIII–(O22–)–CuII Complex with an St = 1 Ground State

Cytochrome c oxidase (CcO) is a heme copper oxidase (HCO) that catalyzes the natural reduction of oxygen to water. A profound understanding of some of the elementary steps leading to the intricate 4e–/4H+ reduction of O2 is presently lacking. A total spin St = 1 FeIII–(O22–)–CuII (IP) intermediate is proposed to reduce the overpotentials associated with the reductive O–O bond rupture by allowing electron transfer from a tyrosine moiety without the necessity of any spin-surface crossing. Direct evidence of the involvement of IP in the CcO catalytic cycle is, however, missing. A number of heme copper peroxido complexes have been prepared as synthetic models of IP, but all of them possess the catalytically nonrelevant St = 0 ground state resulting from antiferromagnetic coupling between the S = 1/2 FeIII and CuII centers. In a complete nonheme approach, we now report the spectroscopic characterization and reactivity of the FeIII–(O22–)–CuII intermediates 1 and 2, which differ only by a single −CH3 versus −H substituent on the central amine of the tridentate ligands binding to copper. Complex 1 with an end-on peroxido core and ferromagnetically (St = 1) coupled FeIII and CuII centers performs H-bonding-mediated O–O bond cleavage in the presence of phenol to generate oxoiron(IV) and exchange-coupled copper(II) and PhO• moieties. In contrast, the μ-η2:η1 peroxido complex 2, with a St = 0 ground state, is unreactive toward phenol. Thus, the implications for spin topology contributions to O–O bond cleavage, as proposed for the heme FeIII–(O22–)–CuII intermediate in CcO, can be extended to nonheme chemistry.


■ INTRODUCTION
−3 CcO contains a unique heterobimetallic heme copper active site, the metal centers of which are separated by >4.0 Å. 4,5 The four electrons required to fully reduce dioxygen to water are supplied by iron (Fe II → Fe IV ), copper (Cu I → Cu II ), and a tyrosine residue (Tyr−OH → Tyr−O • ), which is covalently tethered to one of the histidine moieties ligated to Cu.The consensus mechanism of CcO, shown in Scheme 1, involves the initial binding of dioxygen to the reduced (R) form of the enzyme to form transient intermediate A, which is then rapidly converted to P M . 3,5The electronic structure of P M is unambiguously assigned as a ferryl-heme-cupric-hydroxide-tyrosyl-radical species based on various spectral and chemical evidence. 6owever, the identity of the species formed before the O−O cleaved intermediate P M has proved controversial.−15 Accordingly, the lack of empirical vibrational data does not allow for a definite assessment of the degree of O 2 reduction for intermediate A. Furthermore, various theoretical and kinetic studies have hypothesized the involvement of a total spin S t = 1 heme Fe III −O 2 −Cu II peroxido intermediate (I P ) during the A to P M conversion; 16−19 however, direct spectroscopic evidence of such intermediate in the CcO catalytic cycle has stayed elusive.Notably, the ferromagnetic coupling between the low-spin S = 1/2 Fe III and S = 1/2 Cu II centers in the S t = 1 heme Fe III − O 2 −Cu II peroxido intermediate to form the exchange-coupled Cu II −OH and Tyr • products in CcO is discussed as the prerequisite necessary for the efficient Tyr-mediated reductive O−O bond rupture without any spin-surface crossing. 18,20,28−28 Nevertheless, the acute Fe−O−O−Cu dihedral angle in all of these complexes have led to a catalytically nonrelevant antiferromagnetically coupled S t = 0 ground state of the low-spin heme Fe III −O 2 −Cu II adducts. 14,15−31 But the exact nature of the O 2reduced moiety and its formation in these "as isolated" structures have been disputed.In particular, the reported O−O and Fe/Cu−O bond lengths in these structures are inconsistent with the peroxido assignment. 5One of the "as isolated" structures was reported to exhibit a UV−vis absorption feature at 650 nm, which when excited at this wavelength showed a resonance Raman (rRaman) active band at 755 cm −1 . 32This band, which was initially thought to be a ν(O−O) stretching mode, was later assigned to a mode of the His-419 ligand based on rRaman spectroscopy. 33This provided further doubt to the proposed binding of the peroxido ligand in the "as isolated" form of CcO.
Herein, we report the synthesis, spectroscopic characterization, and reactivity of a nonheme [(CH 3 CN)(transcyclam)Fe III (μ-η 1 :η 1 -O 2 )Cu II (AN)] 3+ (1; AN = 3,3′-iminobis-(N,N-dimethylpropyl-amine)) complex, which may improve our understanding of the properties of the elusive catalytically relevant ferromagnetically coupled (S t = 1) low-spin heme  34 Figure 1B shows the UV−vis spectrum after reaction of one equivalent of the well-characterized [Cu I (AN)]BF 4 complex 28,35 with 3, which led to the formation of a deep blue species, 1, with an absorption maximum at 615 nm (ε = 1.5 L mmol −1 cm −1 ) and a half-life time t 1/2 = 10 min at −90 °C.The 14 K Moßbauer spectrum of 1 (Figure 1A) shows a main species in ∼70% yield with parameters (isomer shift δ = 0.26 mm s −1 , quadrupole splitting |ΔE Q | = 2.11 mm s −1 ) that are typical for a low-spin (S = 1/2) Fe III center.However, the low-temperature, low-field (5 K, 0.06 T) Moßbauer spectrum of 1 (Figure S1, top, blue lines) does not display a magnetic splitting but instead a broadened, asymmetric doublet, characteristic of an integer spin with a positive zero-field splitting (ZFS) parameter.This is confirmed by applied magnetic field Moßbauer studies of 1 (Figure 1A), where the spectra are best fit in the slowrelaxation limit with a total spin S t = 1 with a small axial zerofield splitting D t = 1.5 cm −1 and an anisotropic hyperfine coupling tensor.The asymmetry parameter η = 0.43 determined for 1 differs from the reported parameters for the axially symmetric [Fe IV (O)(trans-cyclam)(CH 3 CN)]OTf 2 (η = 0) complex. 34The Mossbauer spectra of the minor species (Figure S1, green lines) are completely compatible with an axial Fe IV �O complex.Altogether, the Moßbauer measurements propose a species with a low-spin Fe III that is ferromagnetically coupled to a Cu II center, giving rise to a S t = 1 state in 1.
X-ray Absorption Near-Edge Structure (XANES) and EPR Spectroscopy.In addition to Fe III , the presence of a Cu II center 1 is confirmed by the X-ray absorption near-edge structure (XANES) data (Figure S2).The XANES spectra at the Cu K-edge revealed an edge energy of 8986.2 eV, which is shifted by ca.2.5 eV to higher energies relative to the [Cu I (AN)]BF 4 starting compound (8983.6eV), confirming a divalent oxidation state for copper in 1. Complex 1 is EPR silent in perpendicular mode at the X-band (Figure S3), consistent with its S t = 1 ground state with considerable zerofield splitting; the signals shown in Figure S3A correspond to minor contributions (∼15% based on spin quantification; Figure S3B) arising from monomeric high-and low-spin Fe III and Cu II impurities.
rRaman Spectroscopy.Confirmation for the presence of a peroxido moiety in 1 comes from rRaman measurements (in acetone/CH 3 CN 10:1 v/v solution at −90 °C upon excitation at 407 nm) as shown in Figure 1C.They exhibit a characteristic but very low-energy ν(O−O) vibrational mode at 739 cm −1 ( 16 O 2 ; 700 cm −1 with 18 O 2 ) and a diagnostic ν(Fe−O) mode at 635 cm −1 ( 16 O 2 ; 608 cm −1 with 18 O 2 ). 36heoretical Studies.For a better understanding of the actual peroxido binding mode in 1, density functional theory (DFT) calculations were performed to identify optimized structures (Figure 2) and verify if they can theoretically reproduce the experimental results.Due to the flexibility of the mononucleating AN and cyclam ligands, a variety of different conformations and rotamers with and without an additional solvent molecule (CH 3 CN) binding to the Fe III center were needed to be considered.These contain the cyclam ligand to Fe in either a folded (cis-V) conformation with alternating directions of the amino hydrogens (above vs below the ring) or a planar (trans-III) conformation with propylene-linked amino hydrogens pointing in the same and ethylene-linked ones in different directions. 37 The same approach has been pursued for complex 2 (vide infra), in which the amino hydrogen of the AN ligand is replaced with a methyl group (MeAN).For each conformational combination, the lowest-energy structure, along with their relative energies, geometric parameters, Raman data of 18 O-sensitive modes, and UV−vis absorption spectra are presented in Table S1.
All geometry optimizations yielded Cu II in a distorted tetrahedral coordination environment, ligated by only one of the peroxido oxygens.Independent of the conformation of the Cu ligand, starting structures with trans-III-cyclam converged to a μ-η Most interestingly, depending on the O−O binding mode, the conformers can be divided into two groups according to both the calculated UV−vis spectra and the vibrational frequencies of the peroxido bridge (Table S1 and Figure 3).The Fe-cyclam conformation was found to be the controlling factor in determining the spectroscopic properties of the Fe III − O 2 −Cu II core in 1.While all models show strong absorption below 450 nm, only the trans-III-cyclam, i.e., the end-on μ-  S7 and S8, respectively, for their Cartesian coordinates.In particular, considering the respective lowest-energy structures (Figure 2; see Table S7), the calculated spectroscopic properties of 1-trans-b (λ LMCT cal = 617 nm, ν(O−O) cal = 737 cm −1 , and Δ 16/18 O cal = −37 cm −1 ) are in excellent agreement with the experimental UV−vis and rRaman features of 1.Thus, although the corresponding cis and trans models cannot be compared energetically due to the presence of an additional axial CH 3 CN ligand to the Fe III ion in the trans structure, complex 1 can be safely assigned to an end-on μ- 2. This is also supported by the identification of another 18 O-sensitive mode, involving the Fe− O stretching vibration in the range of 608−627 cm −1 (Δ 16/18 O cal = −28 to −30 cm −1 ) for the end-on peroxido structures, in close proximity to the experimentally observed 18 O-sensitive band at 635 cm −1 (Δ 16/18 O = −27 cm −1 ) for 1.The corresponding Fe−O vibration is calculated at 565−592 cm −1 (Δ 16/18 O cal = −16 to −22 cm −1 ) for the μ-η 2 :η 1 peroxido structures, which are significantly downshifted in energy relative to that of the experimentally observed value for 1.
The absorption spectrum of 1 was also analyzed in detail.For the absorption band peaking at 617 nm, we find two main contributions at 610 and 649 nm in the calculated spectrum (Figure 3C).From the TD-DFT difference density (Figure 3E) for the electronic transition at 610 nm, we readily can identify it as a ligand-to-metal charge-transfer (LMCT) band from the peroxido bridge to the Fe III ion.Closer inspection by means of the corresponding natural transition orbitals (NTOs, Figure S6) shows that it is characterized by two orbital pairs: mainly an O 2 2− π* v → Fe 3d xz excitation (from the peroxido-Fe π-bonding MO into its π-antibonding counterpart, 65%), and a minor excitation (20%) from a mixed Fe 3d xy /Cu 3d z 2 MO into a strongly delocalized σ-antibonding MO involving Fe 3d z 2 , O 2 2− π σ , Cu 3d x 2 −y 2 and orbitals on the N-donor atoms of both the AN and cyclam ligands.For the 649 nm band, the difference density and the NTOs indicate donor and acceptor orbitals delocalized mainly over the Cu ion, the peroxido ligand, and the nitrogens of the AN ligand.The donor orbital is composed of the Cu 3d z 2 orbital with substantial admixture from the peroxido oxygens (π* v ), while the acceptor orbital comprises the Cu 3d x 2 y 2 and the σ-antibonding ligand orbitals from the AN nitrogens and peroxido oxygens (π* σ ).Thus, in the presented nonheme Fe III −Cu II -peroxido species 1, the specific O−O bridging motif is essential in defining both these transitions, such that a 600−650 nm band appears as a spectroscopic marker for this geometry.Me group, a μ-η 2 :η 1 peroxido structure results in the corresponding [(cis-cyclam)Fe III (μ-η 2 :η 1 -O 2 )Cu II (MeAN)] 3+ complex 2. Experimentally, the iron center in 2 is also lowspin Fe III as evident from zero-field Moßbauer studies (δ = 0.24 mm s −1 and ΔE Q = −2.51mm s −1 ), but unlike the S t = 1 ground state in 1, a S t = 0 assignment is deduced for 2 based on the applied field Moßbauer, which is also consistent with EPR studies (Figures S9 and S10).The different peroxido binding motifs and coupling situations in 1 and 2 are reflected in their different spectroscopic properties (Figure 3A/B).Complex 2 lacks any UV−vis absorption feature above 400 nm and exhibits a higher peroxidic stretching frequency, ν(O−O) = 856 cm −1 ( 16 O 2 ; 803 cm −1 for 18 O 2 , Figure S11), relative to that of 1 (ν(O−O) = 739 cm −1 ).Our theoretical studies predict a 2-cis-b structure with a μ-η 2 :η 1 peroxido binding motif (Figure 2, Table S8), which can reproduce the experimentally observed UV−vis absorption and the O−O stretching modes in 2 (Figure 3 and Table S1).1.99 Å).The apparent shortest Fe−O bonds in 1 and 2 from the EXAFS simulations may in part reflect contributions of Fe IV �O species in the samples (see below).Also, for the copper spectra, the simulations revealed two O atoms at 1.87/ 1.84 and 2.57/2.55Å in 1/2, respectively (DFT: Cu−O2: 1.91/2.00Å, Cu−O1: 2.62/2.93Å) (see Tables S2−S4 for details). 38We note that in the presence of multiple species in the solution samples of the complexes, perfect agreement between interatomic distances from XAS and DFT is not expected, as a limited number of variable parameters (e.g., bond lengths) have to be used in EXAFS fit analysis to avoid data overinterpretation.Furthermore, low concentrations of 1 and 2 and the presence of scatterers from the ligand scaffold in similar distances did not allow for a reliable determination of Fe•••Cu distances above 3.5 Å via EXAFS.
Axial Ligand Effect on the Peroxido Binding Motif.As suggested by DFT, the binding of an axial CH 3 CN ligand triggers the isomerization of the cis-cyclam moiety in 2 to trans-cyclam in 1, with the concomitant change of the peroxido binding motif at Fe from side-on to end-on (Scheme 2 and Figure 2).Consistent with this suggestion, the addition of 1,5dicyclohexylimidazole (DCHIm) to a CH 3 CN/acetone solution of 2 at −90 °C led to the generation of a blue− green species 2-DCHIm (t 1/2 = 40 s at −90 °C) that shows the characteristic low-energy absorption feature at 640 nm (Figure 4A) associated with the μ-η 1 :η 1 peroxido binding motif.The axial CH 3 CN ligand in 1 can also be replaced by DCHIm as evidenced by the red shift of the 615 nm band in 1 to 640 nm in 1-DCHIm (t 1/2 = 60 s at −90 °C).The Fe III centers in 1-DCHIm and 2-DCHIm are electronically very similar as evidenced from their near identical Moßbauer parameters: for 1-DCHIm, δ = 0.28 mm s −1 and |ΔE Q | = 1.87 mm s −1 , and for 2-DCHIm, δ = 0.29 mm s −1 and |ΔE Q | = 1.81 mm s −1 (Figure 4B, Figure S12 and Table S6).Both are silent in conventional X-band EPR measurements in perpendicular mode indicating an integer total spin state (Figure S13).Although the metastable nature of 1-DCHIm and 2-DCHIm prevented us from measuring their ν(O−O) vibration modes, based on the similarity in their absorption and Moßbauer features, the presence of a [(DCHIm)(trans-cyclam)Fe III (μ-η 1 :η 1 -O 2 )-Cu II (AN/MeAN)] 3+ motif can safely be concluded in both the complexes.
Reactivity Studies of 1 versus 2 with PhOH.The different electronic structures of 1 and 2 are also reflected in their reactivity properties.For example, complex 1, and not complex 2, exhibits reactions with phenol, possibly stressing the importance of the S t = 1 ground state and the weak −NH secondary interaction in the hydrogen atom transfer (HAT) reactivity of the peroxido complexes.Addition of phenol to 1 showed a fast reaction that could be monitored by UV−vis (Figure 5A).The typical absorption feature of 1 at 615 nm vanished, while a new purple species 1-PhO • is observed with an absorption maximum at 540 nm (ε = 1.05 L mmol −1 cm −1 ) and a broad feature between 600 and 1000 nm.When substituted phenols (p-kresol, p-methoxyphenol, and pchlorophenol) were added to 1, a shift of the absorption maximum (1-p-Me PhO • : λ max = 525 nm, 1-p-OMe PhO • : λ max = 580 nm, and 1-p-Cl PhO • : λ max = 570 nm) can be observed, corroborating the binding of phenol-derived species (Figure S14).Gas chromatography-mass spectrometry (GC-MS) analyses of the reaction mixtures at room temperature revealed the formation of coupled phenol products (4,4′-bis(2,6-di-tertbutylphenol) (19%), 3,3′,5,5′-tetra-tert-butyldiphenoquinone (10%), and 2,6-di-tert-butyl-1,4-benzoquinone (10%) for reaction with 2,6-di-tert-butylphenol and 5,5′-dimethoxy[1,1′biphenyl]-2,2′-diol (∼20%) for reaction with 4-methoxyphenol) in moderate to high yields (Figures S15, S16 and Table S5).Further spectroscopic investigations show that the oxidative coupling of the phenol is accompanied by the reductive cleavage of the peroxido bridge in 1 and the formation of copper(II) and iron(III)-phenoxyl radical species.The initially formed oxoiron(IV) species presumably reacts with excess PhOH to generate the iron(III)-phenoxyl radical species (Scheme 3).Indeed, the Moßbauer spectrum of 1-PhO • (Figure S17) shows the formation of a major Fe III S6); 34 the differences in δ and ΔE Q values are plausibly attributed to additional interactions from the Cu II , PhO • , or OH moieties (see Scheme 3).The EPR spectrum reveals the formation of multiple overlapping S = 1/2 species (Figure S18) arising from Cu II centers in near-quantitative yields.rRaman studies were performed (Figures 5B and S19) to confirm the 1-X PhO • assignment.Two features were observed for 1-PhO • at 815 and 586 cm −1 , which showed no shift when 18 O labeled 1 was used in the reaction, but shifted to 804 and 574 cm −1 , respectively, when 18 O-labeled PhOH was used.The positions of these bands were also found to be sensitive to the nature of the para-substituents (Figure S19).Based on a recent study where we spectroscopically characterized a phenoxyl radical bound to an Fe II center 39 and previous reports of metal bound phenoxyl radicals, 40,41 we assign the 586 and 815 cm −1 bands to Fe−O and in-plane phenyl ring bending vibrations, respectively, for a metal bound PhO  S6 and Figures S22−S24).Consistent with our presumption, an authentic sample of [(DCHIm)(trans-cyclam)Fe IV �O] 2+ obtained by replacement of the CH 3 CN ligand in [(CH 3 CN)(trans-cyclam)Fe IV �O] 2+ by DCHIm at −50 °C did not show any reactivity with PhOH (Figure S24).Notably, both 1-DCHIm and 2-DCHIm show increased yields of the coupled phenol products compared to 1 (Table S5 and Figure S25), reflecting the release of the phenoxyl radical instead of coupling to the iron center.
Reactions of 1 and 2 with PPh 3 .Intermediate 1 also reacts at −90 °C with PPh 3 forming OPPh 3 (identified by GC-MS, Figure S26) and a new species 1-Cu with a UV−vis absorption feature at 675 nm (ε = 0.19 L mmol −1 cm −1 ) that further decays above −50 °C (Figure S27).Notably, both 2 and 3 also reacted with PPh 3 , albeit with rates,respectively, 1 and 3 orders of magnitude slower than 1 (Figure S28−S30).This confirms that the O−O bond in 1 is more activated relative to that in 2 and 3.In particular, the higher reaction rates of  S6). 34The XANES spectra at the Fe K-edge of 1-Cu and 2-Cu (Figure S32) show an intense preedge peak at about 7115 eV that is typical for oxoiron complexes.EXAFS analysis additionally revealed a short iron− oxygen bond of 1.62 Å in ca.50% of the complexes (Figure S33, Table S3), which is indicative of an Fe IV �O species.The ESI-MS spectrum of 1-Cu shows a signal at m/z = 706.1 with mass and isotope distribution pattern consistent with a [Fe(O)(cyclam)(OTf)Cu(AN)(Cl)] + assignment, which is shifted by two and one mass units upon 18 O-and 57 Felabeling, respectively (Figure S33).1-Cu is EPR silent in perpendicular mode X-band EPR (Figure S34).All of these point to the formation of Fe IV �O, Cu I , and PPh 3 O products in the reaction of the S t = 1 [(CH 3 CN)(trans-cyclam)Fe III (μη 1 :η 1 -O 2 )Cu II (AN)] 3+ (1) with PPh 3 .

■ CONCLUSIONS
In summary, a nonheme iron(III)-superoxo complex 3 reacts with [Cu I (AN)] generating a low-spin iron(III)-peroxidocopper(II) complex 1 with a S t = 1 ground state. 1 undergoes efficient O−O bond cleavage in the presence of H-bonded PhOH and a strong axial ligand DCHIm leading to the stoichiometric formation of an oxoiron(IV) species.This process has some relevance to the CcO catalytic cycle.A key matter that remains unsolved in CcO is the reaction coordinate connecting intermediates A and P M .−45 In the present study, complex 1 not only reproduces the proposed S t = 1 [Fe III (μ-η 1 :η 1 -O 2 )Cu II ] motif, suggested for I P , but the −NH group of the AN ligand also provides the platform for HAT from a H-bonded phenol that is required for the efficient O−O bond cleavage step.The fast and efficient O−O cleavage Scheme 3. Proposed Mechanism for the Reaction of 1 and 1-DCHIm with Phenol via a H-Bond-Assisted Mechanism occurring in 1 can be attributed to the observed ferromagnetic coupling between the low-spin Fe III and Cu II centers in 1, which, as previously proposed based on theoretical studies 16−19 in I p ensures efficient reductive O−O rupture to form S = 1 Fe IV �O and antiferromagnetically coupled Cu II and PhO • via electron transfer from PhO − without the necessity of any spin-surface crossing.However, the exchange-coupled Cu II −PhO • species is presumably transient in our case and decays to mononuclear Cu II (detected in EPR, data not shown) and coupled phenol products.Consistent with this proposition, a related low-spin Fe III -peroxido-Cu II species 2 in otherwise similar Fe-and Cu-coordination environments, but with an S t = 0 ground state due to antiferromagnetic coupling between the Fe III and Cu II centers, did not show any reactivity with PhOH.
Complexes 1 and 2 differ in their peroxido binding mode to Fe III (end-on vs μ-η 2 :η 1 in 1 and 2, respectively), which is also reflected in their ν(O−O) modes and the transitions of the O 2 2− to Fe III charge transfer in the UV−vis absorption spectrum.Considering the substantial differences in structural parameters for the two peroxido bridging motifs and of the Fe coordination geometries, it is apparent that metal−ligand orbital interactions and MO energies will be largely different in 1 and 2. The fact that in 2, the Fe III ion is directly ligated by both instead of only one O donor leads to a considerably different interaction of the peroxo π* with the Fe 3d orbitals.For example, the π* v orbital exhibits stronger interactions with the Fe 3d xy and 3d yz orbitals at the expense of the interaction with its 3d xz orbital compared to the situation in complex 1.It can thus be understood that the dominant excitation within the strongest LMCT transition (610 nm) in 1 does not exist as such in 2 due to a mitigated π-bonding interaction between the O 2 2− π* v and the Fe 3d xz orbital.
Comparison of 1 to the very well characterized heme-based iron-peroxo-copper adduct [(DCHIm)(F 8 )Fe−O 2 −Cu(AN)] + by Karlin and Solomon, 12,28,42 which they have identified to possess a μ-η 1 :η 1 /end-on binding mode, shows that they resemble each other in that they both feature similar LMCT excitations from the O 2 2− π* v orbital into the Fe 3d xz and Cu 3d x 2 −y 2 orbitals, albeit at significantly lower energies (789 and 951 nm, respectively) and of lower intensity.As orbitals from the equatorial ligands to the Fe III ions contribute only insignificantly in both cases, comparison between the heme and nonheme complexes is legitimate in this respect.The higher energy of the transition from the π-bonding O In heme-copper model chemistry, a lower value for ν(O−O) is in general correlated with a side-on binding mode to one or both metal centers.Table 1 shows selected examples of hemecopper model complexes (note that the [(F 8 )Fe−O 2 − Cu(AN)] system also incorporates the AN ligand at the copper site) with a linear correlation between the observed O− O stretching frequency and the Fe•••Cu-distance, corresponding to the assignment from side-on 11,13,24 to end-on 11,27,28 via a μ-η 2 :η 1 -(O 2 ) 22,27 binding mode. 5,12Similar observations and assignments were also made for other heme-based models.−33 While the investigated structures and trapped intermediate states are not directly involved in the catalytic cycle of CcO, they are thought to provide valuable insights into the character of actual intermediate I P .All studies agree on a long Fe•••Cu distance of about 4.6−4.8Å, which would only allow for an end-on bound peroxido unit, but the O−O-bond length could not be determined precisely. 5The O−O frequency that is of course correlated with the bond length is also a matter of discussion.In initial studies, a rRaman feature at 750 cm −1 was observed in one resting oxidized CcO example and assigned to the O−O stretching mode, 7,32 but a recent study discussed this feature as the wagging of the axial histidine that is sensitive to the peroxido unit. 33One reason why the assignment of the 750 cm −1 band to the O−O mode was questioned was the discrepancy between its low energy, indicating a side-on binding mode in heme chemistry while the environment in the active center would presuppose an end-on bound O 2  Journal of the American Chemical Society
1 :η 1 (end-on) bridging motif with FeCu distances of 4.1−4.3Å and a trend of short (1.8 and 1.9 Å, respectively) and long (2.6−2.8Å) Fe•••O and Cu•••O distances, respectively.In contrast, those with cis-V-cyclam yielded a μη 2 :η 1 motif with the peroxido unit binding side-on to Fe and end-on to Cu and shorter Fe•••Cu distances of 3.7−3.9Å, two shorter Fe•••O distances (1.8 and 2.0 Å), and one short and one long Cu•••O distances (2.0 and 2.9 Å, respectively).For both binding modes, the O•••O distances of 1.43−1.45Å are very similar.Broken-symmetry (BS) DFT suggests triplet ground states and thus ferromagnetic Fe III −Cu II interactions for all of the structures.
Synthesis and Characterization of 2. An interesting structural aspect of 1-trans-b is the orientation of the amino hydrogen of the AN ligand in bent conformation, which is placed at distances of 3.92 and 4.11 Å from the peroxido O Cu and O Fe atoms, respectively, and possibly involved in a weak secondary interaction (Figure 2, Scheme 3).Notably, the deuteration of the NH-group of AN resulted in an increased half-life of 1, further supporting an interaction of the −NH group with the peroxido moiety in 1 (Figures S7−S8).However, this interaction is too weak to effect any significant change in the Fe−O or O−O vibration modes, as corroborated by theoretical (the Δ 1/2 H cal isotope shift upon deuteration of the amino hydrogen is only 0.2 cm −1 for the O−O stretch and 0.01 cm −1 for the Fe−O stretch) and experimental rRaman (data not shown) studies.When the N−H group is replaced by a N−

2 .
Scheme 2. Overview of the Discussed Complexes and Intermediates and Their Reactivity
• radical.The S t = 1 [(CH 3 CN)(trans-cyclam)Fe III (μ-η 1 :η 1 -O 2 )-Cu II (AN)] 3+ core in 1 thus undergoes a reductive O−O bond cleavage step via a net H-atom transfer from phenol to generate [(AN)Cu II −OH] + , [Fe IV (O)(trans-cyclam)-(CH 3 CN)] 2+ and PhO • (Scheme 3).A previous study on a heme Fe III −O 2 −Cu II species 39 has shown that the phenol can facilitate O−O cleavage via two possible mechanisms, which differ by the amount of proton transfer prior to the transition state.One mechanism involves nearly complete proton transfer from the phenol to the peroxo before the barrier.The second mechanism involves O−O homolysis by phenol H-bonding to peroxo, with the proton transfer occurring after the barrier.The two mechanisms can be distinguished by kinetic isotope effect (KIE) data for an H-atom-donating phenol inducing O− O cleavage.A KIE of 1.43 has been determined for the O−O bond cleavage step (Figure S20) in 1, consistent with the hydrogen-bonded mechanism (Scheme 3); 42 a much higher KIE (>5) is expected for complete proton transfer before the barrier.This H-bond assistance is further supported by an observed secondary KIE of 1.20 (Figure S20) in the phenol reaction upon deuteration of the −NH group of the AN ligand in 1.An increase of the reaction rate for the formation of 1-X PhO • with the pK a of the corresponding substituted phenols (k 2 = 0.38 for 1-p-Me PhO • , k 2 = 0.83 for 1-PhO • , k 2 = 1.53 for 1-p-Cl PhO • , Figure S21) is also consistent with the mechanism shown in Scheme 3. Effect of the Axial Ligand on PhOH Reactivity of 1.The complex [Fe IV (O)(trans-cyclam)(CH 3 CN)] 2+ undergoes a side reaction with excess PhOH to generate [(transcyclam)Fe III (PhO • )(OH)] 2+ .This reaction is presumably triggered by the replacement of the axial CH 3 CN ligand in [Fe IV (O)(trans-cyclam) (CH 3 CN)] 2+ by PhOH, followed by HAT by the Fe IV �O center.Indeed, the side reaction is completely stopped in the reaction of 1-DCHIm with PhOH, where replacement of the stronger axial ligand DCHIm by PhOH is not favored, and the Moßbauer analysis of the reaction mixture shows the formation of [Fe IV (O)(transcyclam)(DCHIm)] 2+ in near-quantitative yields (δ = 0.05 mm s −1 , |ΔE Q | = 2.00 mm s −1 , Table 1 and 2, relative to 3, demonstrate the positive influence of Cu in the reductive O−O bond cleavage reaction.The absorption spectrum and Moßbauer parameters (Figure S31, δ = 0.07 mm s −1 , |ΔE Q | = 2.08 mm s −1 ) of 1-Cu are similar compared to that reported for [Fe IV (O)(trans-cyclam)-(CH 3 CN)]OTf 2 (Table

2 −.
The present report of the S t = 1 [(CH 3 CN)-(trans-cyclam)Fe III (μ-η 1 :η 1 -O 2 )Cu II (AN)] 3+ complex 1, which can reproduce both the UV−vis absorption and rRaman features associated with the as-isolated CcO, corroborates that a bridging peroxido moiety in an end-on bound O 2 2− structure and with a significantly weak O−O bond can be accommodated within the Fe•••Cu distance of >4.0 Å that is established in CcO.Whether the 750 cm −1 signal in the resting state of CcO can be possibly assigned as the O−O-vibration mode associated with an end-on Fe III −O 2 −Cu II peroxido binding motif is now an intriguing question, which needs further investigation.■ASSOCIATED CONTENT* sı Supporting InformationThe Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.4c04492.Instrumental and physical methods, synthesis of compounds and intermediates, reactivity and kinetic data, theoretical studies, Figures S1−S34 (further characterization and reactivity data), and Tables S1− S8 (reactivity and DFT data) (PDF)

Table 1 .
33mparison of the Frequencies of the O−O Stretching Modes and Selected Bond Lengths of Fe−O 2 −Cu Intermediates Alternatively assigned as an O-sensitive histidine breathing mode.33 a