|Home | About | Journals | Submit | Contact Us | Français|
Mononuclear non-heme iron oxygenases catalyze a diverse array of important metabolic transformations that require the binding and activation of dioxygen. In the catalytic cycles of the non-heme iron enzymes, several intermediates, such as iron(III)-hydroperoxo, iron(IV)-oxo, and iron(V)-oxo species, have been proposed as active oxidants that effect the oxygenation of organic substrates. Among the proposed intermediates, synthetic non-heme iron(III)-hydroperoxo species were shown to be a sluggish oxidant that cannot oxygenate organic substrates, such as alkanes and olefins.
The first non-heme iron(IV)-oxo complex was characterized spectroscopically by Wieghardt and co-workers. Subsequently, a crystal structure of a non-heme iron(IV)-oxo complex was reported by Münck, Nam, Que, and their co-workers. Since then, a handful of iron(IV)-oxo complexes have been synthesized and investigated in various oxidation reactions.[6,7] Thus, it has been well demonstrated that non-heme iron(IV)-oxo species are capable of oxygenating organic substrates, such as alkanes, alkenes, alcohols, and sulfides.
Iron(V)-oxo species have also been inferred as active oxidants in the oxygenation of organic substrates by non-heme iron catalysts. Although such an iron(V)-oxo intermediate was synthesized and characterized with various spectroscopic techniques, evidence for the involvement of iron(V)-oxo species as active oxidants in catalytic oxygenation reactions was indirect and mainly from product analysis, such as a selectivity of alcohol over ketone in alkane hydroxylation, a KIE value of >4, and the incorporation of 18O from H218O into products in labeled water experiments. However, iron(IV)-oxo complexes have also shown very large KIE values and the oxygen exchange with H218O. Therefore, it is difficult to distinguish the nature of intermediate(s) (i.e., iron(IV)-oxo vs iron(V)-oxo species) in catalytic oxygenation reactions.
In this work, we have generated a new non-heme iron(IV)-oxo complex, [Fe(IV)(O)(BQEN)]2+ (1) (BQEN = N,N’-dimethyl-N,N’-bis(8-quinolyl)ethane-1,2-diamine) (see Figure 1), that shows a high reactivity in the oxidation of alkanes and alcohols. Reactivity studies revealed that the alkane and alcohol oxidations occur via a hydrogen-atom (H-atom) abstraction mechanism. We have also obtained strong evidence that an intermediate different from the iron(IV)-oxo species is involved in the catalytic oxidation of alkanes and alcohols by an iron catalyst and a terminal oxidant.
The reaction of Fe(II)(BQEN)(CF3SO3)2 with CH3CO3H in CH3CN at 0 °C produced a green intermediate 1 with λmax at 740 nm (Figure 2a). This chromophore resembles those associated with S = 1 non-heme iron(IV)-oxo complexes.[4–7,12] The EPR spectrum of 1 appeared silent, as observed in other iron(IV)-oxo species.[5–7] The electrospray ionization mass spectrum (ESI-MS) of 1 exhibits a prominent ion peak at m/z = 563.0 (Figure 2b), which upshifts to m/z = 565.0 upon introduction of H218O into the solution of 1 (Supporting Information (SI), Figure S1 for mass spectral changes of 1 at different incubation times), indicating that 1 contains an iron-oxo group which exchanges its oxygen with H218O at a slow rate. These ESI-MS data are consistent with the formulation of 1 as [Fe(IV)(O)(BQEN)(CF3SO3)]+ (calculated m/z of 563.1).
The formation of an Fe(IV)-oxo species in the reaction of [Fe(II)(BQEN)]2+ and CH3CO3H was further confirmed by X-ray absorption spectroscopy (XAS) (Figure 3a). The Fe K XAS pre-edge and edge of intermediate 1 show an increase in energy of ~2 eV relative to the starting material, as well as a notable increase in pre-edge intensity, from ~7 to ~21 units. The edge shift and increase in pre-edge intensity are a result of greater Zeff and distortion from centrosymmetry, indicating the presence of a short Fe(IV)=O bond.[12c,13] EXAFS data for [Fe(II)(BQEN)]2+ and 1 show a notable intensity decrease upon oxidation as well as a sharpening of the first shell feature in the Fourier transform (Figure 3b; also see SI for detailed discussion). The best EXAFS fits to [Fe(II)(BQEN)]2+ (fit 3) and 1 (fit 4) are given in SI, Table S2. The EXAFS data of [Fe(II)(BQEN)]2+ fits to a 6-coordinate first shell of light atoms (nitrogen/carbon) at 1.97 Å, whereas a short 1.67 Å Fe(IV)=O bond is required to accurately fit the EXAFS data for 1. The decrease in EXAFS amplitude and Fourier transform intensity could be a result of the combination of the formation of a short Fe(IV)=O and by the presence of a mixture due to incomplete conversion to 1. The short Fe=O bond causes a lengthening of the trans Fe-N bond (see SI on XAS Results and Analysis: DFT), resulting in a less ordered first shell. Sharpening of the Fourier transform over the first shell might be a result of the EXAFS wave contribution from the short Fe(IV)=O bond, as the short EXAFS wave overlaps with the main first shell feature, creating destructive interference at high and low k or from interference from the signals of the starting material. A more detailed analysis of both the pre-edge and EXAFS fits is given in the Supporting Information.
The intermediate 1 has a significant lifetime (t1/2 ~0.5 h) at 0 °C, but decayed at a fast rate upon addition of substrates, such as triphenylmethane, indan, tetralin, cumene, ethylbenzene, and benzyl alcohol. Pseudo-first-order fitting of the kinetic data allowed us to determine kobs values, and the pseudo-first-order rate constants increased linearly with the increase of substrate concentration (SI, Figure S4); the second-order rate constants, k2, at 0 °C were 2.4 × 10−1 M−1 s−1 for triphenylmethane, 4.5 × 10−1 M−1 s−1 for indan, 4.3 × 10−1 M−1 s−1 for tetralin, 2.7 × 10−2 M−1 s−1 for cumene, 2.6 × 10−2 M−1 s−1 for ethylbenzene, and 6.7 × 10−1 M−1 s−1 for benzyl alcohol. When the log k2' values were plotted against the C-H bond dissociation energy (BDE) of the substrates (triphenylmethane, 81 kcal mol−1; indan, 82 kcal mol−1; tetralin, 82 kcal mol−1; cumene, 85 kcal mol−1; ethylbenzene, 87 kcal mol−1), we have obtained a linear correlation between the reaction rates and the C-H BDE of substrates (Figure 4). The observation that the rate constants decrease with the increase of the C-H BDE of substrates implicates an H-atom abstraction as the rate-determining step for the oxidation. Further evidence for the H-atom abstraction mechanism was obtained from a measurement of KIE values in the oxidation of cumene, ethylbenzene, and benzyl alcohol; KIE values of 25(2) for benzyl alcohol, >11 for cumene, and >10 for ethylbenzene are consistent with C-H bond cleavage being the rate-determining step (SI, Table S3). It is worth noting that large KIE values of 30 – 50 were reported in the oxidation of ethylbenzene and benzyl alcohol by [FeIV(O)(N4Py)]2+ (N4Py = N,N-bis(2-pyridylmethyl)-N-bis(2-pyridyl)methylamine)[7a,b] and in the cleavage of the target C-H bond of taurine by the FeIV=O intermediate of TauD. The good correlation between reaction rates and BDE of substrates and large KIE values in the C-H bond oxidation of alkanes and alcohol support the notion that the oxidation of alkanes and alcohols by 1 occurs via an H-atom abstraction mechanism.[7,17]
Interestingly, when CH3CO3H was added to a reaction solution containing [Fe(II)(BQEN)]2+ and indan at 0 °C, the formation of 1 was not observed (SI, Figure S5a). Considering the reaction rate of 1 with indan at 0 °C (SI, Figure S5b), this result implies that a highly reactive intermediate 2 was generated in the reaction of [Fe(II)(BQEN)]2+and CH3CO3H and then reacted fast with indan. If 1 is the sole intermediate generated in the reaction of [Fe(BQEN)]2+and CH3CO3H, we should observe the formation of 1 even in the presence of indan and then the decay of 1 due to the reaction with the substrate as shown in SI, Figure S5b.
The following observations further support the involvement of 2 as an active species in the reaction of [Fe(II)(BQEN)]2+and CH3CO3H under catalytic conditions. When the oxidation of benzyl alcohol by [Fe(II)(BQEN)]2+and CH3CO3H was carried out at 0 °C, a deep green color developed within 60 s (λmax = 660 nm) (SI, Figure S6a). The ESI-MS of the green species 3 exhibits a prominent peak at m/z = 520.1 (calculated m/z of 520.2) (SI, Figure S6b), whose mass and isotope distribution pattern corresponds to [Fe(III)(BQEN)(OCH2(C6H4)O)]+ (see the proposed structure 3 in Scheme 1). The EPR spectrum of 3 shows a strong signal at g = 4.3, typical of a high-spin (S = 5/2) FeIII species (SI, Figure S6c). The spectroscopic observations indicate that the aromatic ring hydroxylation of benzyl alcohol took place at the ortho position, affording an iron(III) complex bearing a 2-hydroxybenzyl alcohol ligand (Scheme 1, pathway E). This proposal is based on the recent reports that reactions of non-heme iron(II) complexes bearing two cis-labile sites, such as [Fe(TPA)]2+ (TPA = tris(2-pyridylmethyl)amine) and [Fe(BPMEN)]2+ (BPMEN = N,N’-dimethyl-N,N’-bis(2-pyridylmethyl)ethane-1,2-diamine), react with perbenzoic acids or benzoic acid plus H2O2, affording iron(III)-salicylate complexes as products. We therefore propose that the aromatic ring hydroxylation is a preferred pathway in the oxidation of benzyl alcohol by the intermediate 2. In contrast, the formation of 3 was not observed in the reaction of 1 and benzyl alcohol, demonstrating that 1 does not hydroxylate the aromatic ring of benzyl alcohol but oxidizes benzyl alcohol to benzaldehyde (Scheme 1, pathway D).
By carrying out labeled water experiments, the intermediate 2 was found to contain an iron-oxo group exchangeable with H218O at a fast rate. We have shown above that the iron-oxo group of 1 exchanges with H218O at a slow rate (e.g., ~50% 18O-exchange after 20 min incubation; see SI, Figure S1) (Scheme 1, pathway F). Interestingly, when 1 was generated in the presence of H218O in the reaction of [Fe(II)(BQEN)]2+ and CH3CO3H and analyzed immediately (e.g., less than 30 s) by ESI-MS, we found that 1 contained ~63% 18O derived from H218O (Figure 2b, inset). These results imply that 2 exchanges its oxygen with H218O at a fast rate (Scheme 1, pathway G), followed by the conversion of 2 to 1 (Scheme 1, pathway H). The present results further indicate that the proposed iron(V)-oxo species exchanges its oxygen with H218O much faster than the corresponding iron(IV)-oxo species. This speculation is of interest since the dependence of the oxygen exchange of metal-oxo species on the oxidation states of metal ions has never been explored previously, and we are currently investigating the effect of the metal oxidation states on the oxygen exchange between high-valent metal-oxo complexes and H218O under the identical conditions (e.g., the same ligand structure and solvent systems).
Finally, we propose the mechanism for the formation of 1 and 2 in the reaction of [Fe(II)(BQEN)]2+ and CH3CO3H as follows: [Fe(II)(BQEN)]2+ is oxidized to [Fe(III)(BQEN)]3+ by CH3CO3H. Then, [Fe(III)(BQEN)(OO(O)CCH3)]2+ is generated by the reaction of [Fe(III)(BQEN)]3+ and CH3CO3H. The O-O bond heterolysis of [Fe(III)(BQEN)(OO(O)CCH3)]2+ produces 2, followed by one e− reduction of the unstable 2 that leads to the formation of 1. Nonetheless, other high-valent iron-oxo species, such as [Fe(IV)(OH)2(BQEN)]2+ and a high-spin (S = 2) [Fe(IV)(O)(BQEN)]2+, can be considered as plausible structures for 2.
In conclusion, we have reported the synthesis, characterization, and reactivities of a mononuclear non-heme iron(IV)-oxo complex, [Fe(IV)(O)(BQEN)]2+ (1). We have also provided strong evidence that an intermediate different from 1, possibly an iron(V)-oxo species (2), is involved as an active oxidant in the catalytic oxidation of alkanes and alcohols by [Fe(II)(BQEN)]2+ and peracetic acid. The reactivity of 2 is greater than that of 1, the product distributions observed in the reactions of 1 and 2 are different, and 2 exchanges its oxygen with H218O much faster than 1. Future studies will focus on attempts to characterize the intermediate 2 and understand the reactivity of 2 in various oxygenation reactions.
**The research at EWU was supported by KOSEF/MOST through the CRI Program, Korea, and that at Stanford University by NIH grants RR-01209 (K.O.H.) and GM 40392 (E.I.S.). SSRL operations is funded by the U.S. DOE BES, and the SSRL SMB program by NIH NCRR BTP and DOE BER. The project described was supported by Grant Number 5 P41 RR001209 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH) and its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH.
Jihae Yoon, Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Korea.
Samuel A. Wilson, Department of Chemistry, Stanford University, Stanford, CA 94305, USA.
Yu Kyeong Jang, Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Korea.
Dr. Mi Sook Seo, Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Korea.
Kasi Nehru, Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Korea.
Britt Hedman, Department of Chemistry, Stanford University, Stanford, CA 94305, USA.
Dr. Keith O. Hodgson, Department of Chemistry, Stanford University, Stanford, CA 94305, USA.
Dr. Eckhard Bill, Max-Planck-Institute für Bioanorganische Chemie, Stiftstrasse 34-36, D-45470 Mülheim an der Ruhr, Germany.
Dr. Edward I. Solomon, Department of Chemistry, Stanford University, Stanford, CA 94305, USA.
Dr. Wonwoo Nam, Department of Chemistry and Nano Science, Ewha Womans University, Seoul 120-750, Korea.