Identification and Spectroscopic Characterization of Nonheme Iron(III) Hypochlorite Intermediates

FeIII–hypohalite complexes have been implicated in a wide range of important enzyme-catalyzed halogenation reactions including the biosynthesis of natural products and antibiotics and post-translational modification of proteins. The absence of spectroscopic data on such species precludes their identification. Herein, we report the generation and spectroscopic characterization of nonheme FeIII–hypohalite intermediates of possible relevance to iron halogenases. We show that FeIII-OCl polypyridylamine complexes can be sufficiently stable at room temperature to be characterized by UV/Vis absorption, resonance Raman and EPR spectroscopies, and cryo-ESIMS. DFT methods rationalize the pathways to the formation of the FeIII-OCl, and ultimately FeIV=O, species and provide indirect evidence for a short-lived FeII-OCl intermediate. The species observed and the pathways involved offer insight into and, importantly, a spectroscopic database for the investigation of iron halogenases.


Physical methods
1 H NMR spectra (400 MHz) were recorded on a Varian Mercury Plus. Chemical shifts are denoted relative to the residual solvent peak ( 1 H NMR spectra D 2 O, 4.79 ppm). Elemental analyses were performed with a Foss-Heraeus CHN Rapid or a EuroVector Euro EA elemental analyzer. UV/Vis absorption spectra were recorded with a HP8453 spectrophotometer or a Specord600 (AnalytikJena) in 10 mm path length quartz cuvettes. EPR spectra (X-band, 9.46 GHz) were recorded on a Bruker ECS106 spectrometer in liquid nitrogen (77 K). Samples for measurement (300 μL) were transferred to EPR tubes, which were frozen in liquid nitrogen immediately. High resolution mass spectra (HRMS) were recorded on a Bruker MicrOTOF-Q II TM Instrumental at Serveis Tècnics of the University of Girona. A cryospray attachment was used for CSI-MS (cryospray mass spectrometry). The temperature of the nebulizing and drying gasses was set at 5 and 0 ºC, respectively. Samples were introduced into the mass spectrometer ion source by direct infusion using a syringe pump and were externally calibrated using sodium formate. The instrument was operated in the positive ion mode. Raman spectra were recorded at  exc 785 nm using a Perkin Elmer Raman Station at room temperature. Raman spectra at 473 nm (50 mW at source, Cobolt Lasers) were obtained in a 155 o backscattering arrangement with Raman scattering collected by a 2.5 cm diameter plano-convex lens (f = 7.5 cm). The collimated Raman scattering passed through an appropriate long pass edge filter (Semrock) and was focused by a second 2.5 cm diameter plano convex lens (f = 10 cm) into a Shamrock300i spectrograph (Andor Technology) with a 1200 L/mm grating blazed at 500 nm and acquired with an DU970N-BV CCD camera (Andor Technology). The spectral slit width was set to 60 μm. Data were recorded and processed using Solis (Andor Technology) with spectral calibration performed using the Raman spectrum of acetonitrile/toluene 50:50 (v:v). [3] Samples were held in quartz 10 mm path length cuvettes. Baseline correction was performed for all spectra.

UV/Vis absorption spectroscopy
Direct addition of 1 equiv of NaOCl to [(MeN4Py)Fe II (Cl)](Cl) at pH 2.2 led to a near complete loss in absorbance at 490 nm within 15 s, indicating a complete loss of the Fe(II) complex, after which a new band grew at 480 nm and reached a maximum absorbance within ca. 2 min. Subsequently a decrease but not complete loss in absorbance at 480 nm, concomitant with an increase in absorbance at 670 nm, was observed ( Figure S 1). Upon addition of a second equivalent of NaOCl, the band at 480 nm increased S3 initially with no change in absorbance at 670 nm. After the absorbance at 480 nm reached a maximum, the absorbance at 670 nm began to increase with a concomitant decrease in absorbance at 480 nm. The absorption band in the NIR region (i.e. at 670 nm) is typical of an Fe(IV)=O species. [4] Under acidic conditions the absorption at 670 nm is persistent but disappears rapidly upon an increase in pH ( Figure  S 1e).   7 The bands 653 and 580 cm -1 were shifted by 25 and 18 cm -1 , respectively, these shifts are close to those expected for an Fe-O bond (29 cm -1 for the band at 653 cm -1 and 26 cm -1 for the band at 580 cm -1 ) and O-Cl (26 cm -1 for the band at 653 cm -1 and 23 cm -1 for the band at 580 cm -1 ) modes. Hence definitive assignment of the mode cannot be made on the basis of isotope shift.

Figure S 12
Reaction between [(MeN4Py)Fe II (Cl)]Cl (4 mM in 18 OH 2 at pH 2.2) and Na 18 OCl (two additions of two equiv in 18 OH 2 ) followed by Raman spectroscopy at λ exc 473 nm. Spectra were normalized to the ClO 4 band at 934 cm -1 except for the initial spectrum. The legend is time in minutes after addition of Na 18 OCl.

Figure S 13
Intermediates generated upon the reaction of [(MeN4Py)Fe II (Cl)]Cl (4 mM at pH 2.2) with (a) Na 16 OCl in 16 OH 2 and (b) Na 18 OCl in 18 OH 2 followed by Raman spectroscopy at λ exc 473 nm Figure S 14 Raman Spectra of aqueous NaOCl before and after addition of NaBr at λ exc 785 nm. Spectra were normalized to the water band at ca. 1650 cm -1 .
NaOBr was employed to facilitate band assignments. Addition of 2 equiv of NaOBr [9] to an aqueous solution of [(MeN4Py)Fe II (Cl)](Cl) (1 mM in H 2 O at pH 2.2), results in substantial interference from fluorescence and, hence, it is not possible to obtain Raman spectra under the conditions employed. The solvent system was changed to water/acetonitrile (1:1) to circumvent this problem. As with water, addition of 2 equiv of NaOCl to a solution of [(MeN4Py)Fe II (Cl)](Cl) (1:1 water/acetonitrile at pH 2.2) shows bands at 580, 656, 676 and 843 cm -1 (Figure S 13). Addition of 2 equiv of NaOBr instead of NaOCl shows the bands 843, 673 and 629 cm -1 . The band at 843 cm -1 was not affected. The band at 676 was moderately shifted (3 cm -1 ) by replacement of Cl with Br, this band is also oxygen insensitive indicating that it is a Fe-N mode. The 656 cm -1 band was shifted to 629 cm -1 , the observed shift of 27 cm -1 is too small to be assign to be due to an O-Br vibrational mode. Interestingly, the band at 580 cm -1 was not observed. Even though Cong et al. [10] assigned the 786 cm -1 band in the heme (Fe III -OCl) system as an O-Cl vibrational mode, 18 O labelling and bromine labelling data support the assignment of the 580 cm -1 band as the O-Cl vibrational mode. The other oxygen sensitive mode at 656 cm -1 was tentatively assigned to an Fe III -O stretch. The shift of 27 cm -1 might be due to the effect of bromine on the Fe-O stretch (i.e. a change in force constant). Resonance Raman spectra were calculated for the Fe(III)-OCl (doublet) and the Fe(IV)=O (triplet) species using ORCA (version 3.0.2) and the IMDHO method as implemented in the orca_asa program. [11] The procedure used was as follows: The geometry optimization and the Hessian matrix were calculated with the Becke89 [12] and Parr86 13 (BP86) DFT functional using Def2-TZVP basis set. [14] The Hessian matrix was used in the TD-DFT [15] calculation, which was performed using the BH and HLYP functional with the TZVPP [14a] basis set for Iron and SV(P) [14a,16] basis set for C, H, N, O, Cl. 18 O isotope shifts were calculated with Gaussian. Figure

Computational Details
Computational studies were performed to gain insight into the electronic features of the complexes and the reaction mechanism. Each structure described here was found through a full geometry optimization and characterized as a local minimum with a frequency calculation. To assist with the interpretation of experimental resonance Raman spectra, we also calculated Raman vibrations from the frequencies. All calculations mentioned here were performed using Density Functional Theory (DFT) with the unrestricted B3LYP method [17] as implemented in the Gaussian 09c program package. [18] Single point calculations with B3LYP-D3 [19] were calculated using Jaguar V3.0 Rev 2. [20] All calculations utilized a triple-ζ quality basis set that includes LANL2TZ+(f) on iron and 6-311+G(d,p) on the rest of the atoms, basis set BS1. [21] Free energies (G) reported here use UB3LYP-D3 energies corrected with ZPE, thermal and entropic corrections from the frequency calculation at 298 K. The effect of solvent was tested through single point calculations using water as a solvent with the SMD solvation model as implemented in Gaussian.

Figure S 19
Scheme of the species studied. The number in superscript refers to the possible spin state of the molecule.
Part 1: Spin densities and Charges for N4Py-H ligand.