A tailor-made ligand to mimic the active site of diiron enzymes: an air-oxidized high-valent Fe^III h.s.(μ-O)₂Fe^IV h.s. species

Abstract

The design and first application of a new dinucleating ligand system to mimic high-valent oxidation states of O₂-dependent diiron enzymes is presented.

---

Methane monooxygenase (MMO) and ribonucleotide reductase (RNR) belong to a class of metalloenzymes which activate dioxygen at nonheme diiron active sites.¹ The active species in MMO, Q, is proposed to consist of a Fe^IV(μ-O)₂Fe^IV core with Fe^IV h.s. (S_i = 2) ions.² A similar, oxo-bridged Fe^IIIFe^IV species, X, is found in the catalytic cycle of RNR.³ There have been numerous efforts to mimic the active sites of Q and X.^4,5 Besides principal questions regarding the structures and catalytic cycles of the metalloenzymes, this interest is motivated to obtain conceptional insights for the development of new biomimetic homogenous catalysts.

We have started a program to stabilize highly oxidized diiron units by the use of strongly electron-donating ligand sets.⁶ This was inspired by fully characterized high-valent Fe^III(μ-O)₂Fe^IV complexes using substituted tpa ligands.⁵ Later, linear tetradentate ligands like BPMCN have been employed,⁷ which correspond to a formal substitution of a pyridine by a t-amine, which is more electron-donating than pyridine (Scheme 1a).

Our initial approach has been based on the further substitution of the remaining pyridine donors by strong σ- and π-donating phenolates to further stabilize highly oxidized species leading to the simple ligand H₂L (Scheme 1a) and resulted in the formation of the dinuclear [LFe^III(μ-O)Fe^IIIL] (1) and mononuclear ferric complexes.⁶1 can be reversibly oxidized to 1⁺ and 1²⁺ at relatively low potentials (E¹ = 0.27 V, E² = 0.44 V vs. Fc⁺/Fc). 1²⁺ accumulates the same number of oxidation equivalents as Q. Detailed investigations revealed that (i) the oxidations are ligand- and not metal-centered and (ii) the dinuclear complexes dissociate upon oxidation. Herein, we present an optimization of the ligand design (Scheme 1b) based on the results of the first generation complexes. Moreover, we apply a ligand of the second generation successfully to stabilize a high-valent diiron core.

Scheme 1 Design strategies for the 1^st generation (a) and 2^nd generation (b) ligands.

1. The ligand should offer four coordination sites per metal center. The use of a tripodal ligand topology forces two open sites in a cis-position suitable for a (μ-O)₂ unit.

2. To avoid dissociation into mononuclear species the two ligand units should be covalently linked.

3. Based on a related complex⁸ and simple modeling attempts, we identified ethylene spacers as reasonable bridging units.

4. Donor groups bearing more than one organic rest should be t-amines. The remaining terminal coordination sites should consist of strong but innocent σ- and π-donating groups, like biomimetic carboxylate donors. This leads to the bis(tetradentate) dinucleating ligand H₄julia.⁹

Reaction of H₄julia with Fe^III salts in water‡ allowed the isolation of red crystals of [(julia){Fe(OH₂)(μ-O)Fe(OH₂)}]·6H₂O (2·6H₂O; Fig. 1a).§ The ligand julia⁴⁻ coordinates an {Fe(OH₂)(μ-O)Fe(OH₂)}⁴⁺ core in a dinucleating fashion, exhibiting an Fe1–O3–Fe2 angle of 167° and a dihedral angle O81–Fe1–Fe2–O82 of 59°. The FT-IR spectrum of 2 exhibits a band at 889 cm⁻¹ assigned to the ν_as(Fe–O–Fe) stretching mode. The Mössbauer spectrum recorded at 80 K exhibits one quadrupole doublet with δ = 0.47 mm s⁻¹ and |ΔE_Q| = 2.04 mm s⁻¹ (Fig. S1, ESI†) and the coupling constant (H = −2JS₁S₂) is J = −101 cm⁻¹ (Fig. S2, ESI†). These data are consistent with an oxo-bridged diiron(III) h.s. complex.

Fig. 1 (a) Molecular structure of 2. Except for the coordinated water molecules, all hydrogen atoms have been omitted for clarity. Selected interatomic distances [Å]: Fe1–O1 2.0366(7), Fe1–O2 1.9994(7), Fe1–O3 1.7692(7), Fe1–O81 2.1421(7), Fe1–N1 2.2117(8), Fe1–N2 2.2324(8), Fe2–O41 2.0307(7), Fe2–O42 2.0273(7), Fe2–O3 1.7686(7), Fe2–O82 2.0991(7), Fe2–N41 2.1989(8), Fe2–N42 2.2357(8), Fe1–Fe2 3.5156(2). pH-dependent spectroscopic characterization of 2 in aqueous solution: Mössbauer spectra recorded at 80 K in (b–d) at the given pH-values. UV/Vis/NIR absorption spectrum (black dots) at pH = 5.95 (e) with a Gaussian fit (sum of the Gaussians: red line) and during addition of [Me₄N]OH (f).

A 1 mM solution of 2 in water has a pH of 5.93 indicating a slight acidity of the coordinated water molecules. In order to correlate electrochemical behaviour with molecular structures we performed a pH-dependent spectroscopic characterization of 2 in aqueous solution. The Mössbauer spectrum of a frozen aqueous solution of ⁵⁷Fe-enriched 2 at 80 K exhibits one quadrupole doublet (δ = 0.47 mm s⁻¹, |ΔE_Q| = 1.97 mm s⁻¹, Fig. 1b). The values for 2 in solution are closer to those for 2 in the solid state spectrum than to those of the dmso complex [(julia){Fe(dmso)(μ-O)Fe(dmso)}] (3) (Fig. S3, ESI†) obtained from dissolution of 2 in dmso (δ = 0.47 mm s⁻¹, |ΔE_Q| = 1.91 mm s⁻¹; Fig. S4, ESI†). While the substitution of weakly bound water molecules by dmso introduces such a minor change, the even smaller change by going from the solid state to solution indicates that the structure of 2 remains almost unchanged by dissolution in water.

The UV/vis/NIR spectrum of 2 dissolved in water at pH = 5.95 exhibits strong absorption features above 25 000 cm⁻¹ (Fig. 1e). The Gaussian resolved absorptions at 25 500, 28 900, and 32 800 cm⁻¹ are assigned to oxo LMCTs,¹⁰ while the band at 36 600 cm⁻¹ has a strong carboxylate LMCT character.¹¹

Upon increasing the pH by adding [Me₄N]OH, the absorption bands change mainly in the pH-ranges 7.0–8.4 and 10.5–11.7 with no further variations observed above pH = 12.0 (Fig. 1f). This behavior indicates the formation of two distinct deprotonated species. The evaluation of the ε-vs.-pH plot (Fig. S5, ESI†) allows the estimation of the pK_a values as 7.7 ± 0.3 and 11.4 ± 0.4 for the formation of a mono-deprotonated (4⁻−) and of a double-deprotonated species (5²⁻2−), respectively. The mono-deprotonated species may contain a (μ-O₂H₃)⁻ bridge¹² (4a⁻−) or an Fe^III(μ-O)(μ-OH)Fe^III core (4b⁻−),¹³ while the double-deprotonated species may consist of two terminal hydroxo ligands (5a²2⁻−) or an Fe^III(μ-O)₂Fe^III core (5b²2⁻−).

The Mössbauer spectra of the aqueous solutions at pH = 10.1 and pH = 13.0 exhibit sharp quadrupole doublets with δ = 0.47 mm s⁻¹, |ΔE_Q| = 1.68 mm s⁻¹ (Fig. 1c) and δ = 0.46 mm s⁻¹, |ΔE_Q| = 0.70 mm s⁻¹ (Fig. 1d), respectively. The small change in the quadrupole splitting going from pH = 5.9 to 10.1 is more consistent with the formation of 4a⁻− than of 4b⁻−, which should exhibit a stronger deviation. This is corroborated by the UV/vis spectrum of 4⁻−, where a characteristic band for a Fe^III(μ-O)(μ-OH)Fe^III core at ∼18 000 cm⁻¹ is absent.^12,13 The assignment of the species at pH = 13.0 is less straightforward because (1) the electric field gradient of both potential species is mainly caused by two strong ⁻OX (X = Fe or H) donors in a cis arrangement and (2) 5a²⁻2− and 5b²2⁻− have been detected in ESI-MS experiments (Fig. S6, ESI†). However, we strongly prefer the formulation as 5b²2⁻− due to the lack of characteristic purple LMCT bands for RO⁻ ligands and the inherent instability of the highly nucleophilic Fe^III–OH unit while the Fe^III(μ-O)₂Fe^III has already been described.^13,14

Electrochemical studies were performed in aqueous solution at pH = 5.9 (2), pH = 9.8 (4⁻−), and pH = 12.8 (5²⁻2−) (Fig. 2a). The CV of 2 exhibits two irreversible reductions with E^p,red₁ = −0.25 and E^p,red₂ = −0.56 V vs.SHE. Deprotonation of 2 to 4a⁻− results in a cathodic shift of the first reduction process to E^p,red₁ = −0.92 V vs.SHE due to the formation of a negatively charged species. Contrarily, the double-deprotonated species 5b²⁻ exhibits one almost reversible electron-transfer wave at E^1/2 = −0.75 V vs.SHE, which is shifted anodically relative to 4a⁻. As the reduction to Fe^IIIFe^II should be shifted again cathodically compared to 4a⁻− due to an increase of the negative charge, this redox step must therefore be assigned to an oxidation to Fe^IIIFe^IV. This potential corresponds to a cathodic shift by −1.56 V compared to that for the analogous oxidation of [(tpa^6-Me)Fe^III(μ-O)₂Fe^III(tpa^6-Me)]²⁺.¹⁴ Qualitatively, this shift is consistent with the argument that a redoxactive unit (here the Fe^III(μ-O)₂Fe^III core) is easier to oxidize the more negative the charge of the complex is (2+ for the tpa-complex and 2− for 5b²2⁻−). A simple electrostatic model for a charged spherical particle of radius r in a medium with a dielectricity constant ε_E considering only the difference of the solvation energy of an n+ and an (n−1)+ charge provides a quantitative estimate¹⁵ of −1.68 V for this shift (see, ESI†). This rough estimate strongly supports the assignment to an oxidation of 5b²2⁻− at E^1/2 = −0.75 V.

Fig. 2 (a) CVs of 5 mM aqueous solutions of 2 (0.05 M KClO₄, scan rates 200 mV s⁻¹) at the given pH values. (b) 24 K EPR spectrum of a frozen 5 mM aqueous solution of 2 and pH = 13.0 prepared without exclusion of atmospheric dioxygen.

The potential should be low enough to facilitate oxidation by O₂ (E = 0.46 V vs.SHE at pH = 13). This is corroborated by an EPR spectrum of an aqueous solution of 2 at pH = 13.0, i.e.5b²⁻2−, that exhibits a reproducible axial S_t = 1/2 spectrum (Fig. 2b) with g_⊥ = 2.0066 and g_∥ = 2.0982 (g_av = 2.038). Preparation of a sample under the same protocol but under anaerobic conditions is EPR silent. Opening the unfreezed sample to the atmosphere results in the same S_t = 1/2 spectrum again. This indicates the oxidation of the complex by atmospheric dioxygen in agreement with the low redox potential. The strong anisotropy and g-splitting in the EPR spectrum discard an organic radical. The only monomeric iron species which could exhibit g_i > 2.0 is a Fe^III l.s. species that can be ruled out due to the donor set present. Based on spin-projection analysis¹⁶ the observed gS_t^=1/2 > 2.0 only agrees with an antiferromagnetically coupled Fe^III h.s. (S_i = 5/2, g_i ≈ 2.0) Fe^IV h.s. (S_i = 2, g_i < 2.0) spin system.

The spin-concentration of the EPR spectrum is only 0.24%. On the one hand, this is related to the O₂ concentration in O₂ saturated water, which is only 4.4% of the complex concentration; i.e. the oxidation of 5b²2⁻− by O₂ has a yield of 5.5%. On the other hand, the formation of rust from solutions of 2 above pH ≈ 8 and even faster above pH ≈ 12 directly after base addition is observed. This is not surprising as the high negative charge and high electron density at the Fe^III ions destabilize 5b²2⁻− so that oxidation to Fe^IV or decomplexation with formation of rust is favored. Water as a solvent also precludes the use of lower temperatures to stabilize 5b²2⁻−. This intrinsically low concentration of the oxidized species in water prevents further characterization by Mössbauer or other spectroscopies.

In summary, based on the results of the complexes with ligands of the first generation, we identified a new dinucleating lead motive for ligands of the second generation to stabilize high-valent diiron sites. Herein, the characterization of the first complex [(julia){Fe(OH₂)(μ-O)Fe(OH₂)}] in the solid state and in aqueous solution is described. The spectroscopic data are in accordance with a pH-dependent equilibrium between [(julia){Fe^III(OH₂)(μ-O)Fe^III(OH₂)}], [(julia){Fe^III(μ-O)(μ-O₂H₃)Fe^III}]⁻, and [(julia){Fe^III(μ-O)₂Fe^III}]²⁻, while the electrochemical data demonstrate that the (μ-O)₂-bridged species can be oxidized by atmospheric oxygen to a high-valent {Fe^III h.s.(μ-O)₂Fe^IV h.s.}³⁺ species. Thus, the metal-centered oxidation by dioxygen, the formation of the high-valent {Fe^III h.s.(μ-O)₂Fe^IV h.s.}³⁺ core containing the rare Fe^IV h.s. ion, and its relative high stability (prepared at room temperature!) prove the success of our ligand design to mimic the active sites of non-heme diiron enzymes. In order to stabilize and further characterize this high-valent species, experiments in other solvents are currently performed in our laboratories.

Notes and references

1. A. C. Rosenzweig, C. A. Frederick, S. J. Lippard and P. Nordlund, Nature, 1993, 366, 537–543; P. Nordlund, B.-M. Sjöberg and H. Eklund, Nature, 1990, 345, 393–398; E. I. Solomon, T. C. Brunold, M. I. Davis, J. N. Kemsley, S.-K. Lee, N. Lehnert, F. Neese, A. J. Skulan, Y.-S. Yang and J. Zhou, Chem. Rev., 2000, 100, 235–349.
2. M. Merkx, D. A. Kopp, M. H. Sazinsky, J. L. Blazyk, J. Müller and S. J. Lippard, Angew. Chem., Int. Ed., 2001, 40, 2782–2807.
3. B. E. Sturgeon, D. Burdi, S. Chen, B. H. Huynh, D. E. Edmondson, J. Stubbe and B. M. Hoffman, J. Am. Chem. Soc., 1996, 118, 7551–7557.
4. D. Lee, J. Du Bois, D. Petasis, M. P. Hendrich, C. Krebs, B. H. Huynh and S. J. Lippard, J. Am. Chem. Soc., 1999, 121, 9893–9894; A. Ghosh, F. T. de Oliveira, T. Yano, T. Nishioka, E. S. Beach, I. Kinoshita, E. Münck, A. D. Ryabov, C. P. Horwitz and T. J. Collins, J. Am. Chem. Soc., 2005, 127, 2505–2513; G. Xue, D. Wang, R. De Hont, A. T. Fiedler, X. Shan, E. Münck and L. Que, Jr., Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 20713–20718; D. Wang, E. R. Farquhar, A. Stubna, E. Münck and L. Que, Jr., Nat. Chem., 2009, 1, 145–150.
5. Y. Dong, H. Fujii, M. P. Hendrich, R. A. Leising, G. Pan, C. R. Randall, E. C. Wilkinson, Y. Zang, L. Que, Jr., B. G. Fox, K. Kauffmann and E. Münck, J. Am. Chem. Soc., 1995, 117, 2778–2792; H.-F. Hsu, Y. Dong, L. Shu, V. G. J. Young and L. Que, Jr., J. Am. Chem. Soc., 1999, 121, 5230–5237.
6. T. Glaser, R. H. Pawelke and M. Heidemeier, Z. Anorg. Allg. Chem., 2003, 629, 2274–2281; J. B. H. Strautmann, S. DeBeer George, E. Bothe, E. Bill, T. Weyhermüller, A. Stammler, H. Bögge and T. Glaser, Inorg. Chem., 2008, 47, 6804–6824; J. B. H. Strautmann, C.-G. Frhr. v. Richthofen, S. DeBeer George, E. Bothe, E. Bill and T. Glaser, Chem. Commun., 2009, 2637–2639.
7. M. Costas, J.-U. Rohde, A. Stubna, R. Y. N. Ho, H. Quaroni, E. Münck and L. Que, Jr., J. Am. Chem. Soc., 2001, 123, 12931–12932; M. P. Jensen, M. Costas, R. Y. N. Ho, J. Kaizer, A. Mairata i Payeras, E. Münck, L. Que, Jr., J.-U. Rohde and A. Stubna, J. Am. Chem. Soc., 2005, 127, 10512–10525.
8. K.-O. Schäfer, R. Bittl, W. Zweygart, F. Lendzian, G. Haselhorst, T. Weyhermüller, K. Wieghardt and W. Lubitz, J. Am. Chem. Soc., 1998, 120, 13104–13120.
9. G. Schwarzenbach, H. Senn and G. Anderegg, Helv. Chim. Acta, 1957, 40, 1886–1900.
10. C. A. Brown, G. J. Remar, R. L. Musselman and E. I. Solomon, Inorg. Chem., 1995, 34, 688–717.
11. K. Wieghardt, U. Bossek, P. Chaudhuri, W. Herrmann, B. C. Menke and J. Weiss, Inorg. Chem., 1982, 21, 4308–4314.
12. A. Hazell, K. B. Jensen, C. J. McKenzie and H. Toftlund, Inorg. Chem., 1994, 33, 3127–3134.
13. H. Zheng, Y. Zang, Y. Dong, V. G. J. Young and L. Que, Jr., J. Am. Chem. Soc., 1999, 121, 2226–2235; S. Taktak, V. Kryatov and E. V. Rybak-Akomova, Inorg. Chem., 2004, 43, 7196–7209.
14. H. Zheng, S. J. Yoo, E. Münck and L. Que, Jr., J. Am. Chem. Soc., 2000, 122, 3789–3790.
15. M. Born, Z. Phys., 1920, 1, 45–48; T. Beissel, F. Birkelbach, E. Bill, T. Glaser, F. Kesting, C. Krebs, T. Weyhermüller, K. Wieghardt, C. Butzlaff and A. X. Trautwein, J. Am. Chem. Soc., 1996, 118, 12376–12390.
16. A. Bencini and D. Gatteschi, Electron Paramagnetic Resonance of Exchanged Coupled Systems, Springer-Verlag, Berlin, 1990.

---

Footnotes

† Electronic supplementary information (ESI) available: Description of physical measurements, molecular structure of 3 (CCDC 743430), Mössbauer spectra of 2 and 3 as solid. CCDC 743429 and 743430. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c0cc03098h

‡ Preparation of [(julia){Fe(H₂O)(μ-O)Fe(H₂O)}]·6H₂O. A solution of FeCl₃·6H₂O (147 mg, 0.543 mmol) in water (4 mL) is added to a solution of H₄julia·4HCl (150 mg, 0.272 mmol) in water (4 mL). Then, triethylamine (275 mg, 272 mmol) dissolved in water (4 mL) was added. The solution was stirred for 45 minutes, filtrated and kept at −2 °C. Within two days, red crystals (85 mg, 46%) suitable for single-crystal X-ray diffraction deposit (Found: C 28.6%, H 6.2%, N 8.3%. Calc. for C₁₆H₄₂N₄O₁₇Fe₂: C 28.5%, H 6.3%, N 8.3%).

§ Crystal data for 2·6H₂O: C₁₆H₄₂N₄O₁₇Fe₂, M_r = 674.21 g mol⁻¹, 0.40 × 0.20 × 0.18 mm³, monoclinic, P2₁/c, a = 10.4386(3), b = 24.5460(7), c = 12.1232(3) Å, β = 115.012(1)°, V = 2814.97(13) ų, Z = 4, ρ_calcd = 1.591 g cm⁻³, μ = 1.111 mm⁻¹, Mo_Kα radiation (0.71073 Å), T = 173(2) K, 2θ_max = 60.00°, 39 993 reflections measured, 8198 independent reflections, R_int = 0.0169, R = 0.0196, wR2 = 0.0529, residual max./min. electron densities 0.401/−0.268 e Å⁻³, CCDC 743429.

---