A copper complex supported by an N2S-tridentate ligand inducing efficient heterolytic O-O bond cleavage of alkylhydroperoxide.

We have recently reported a copper(II)-superoxide complex supported by an N3-tridentate ligand (L(N3)), which exhibits a similar structure and reactivity to those of a putative reactive intermediate involved in the catalytic reactions of copper monooxygenases such as peptidylglycine α-hydroxylating monooxygenase (PHM) and dopamine β-monooxygenase (DβM). In this study, we have synthesised and characterised copper complexes supported by a related sulphur-containing ligand (L(N2S)) to get insight into the notable electronic effect of the sulphur donor atom in the reaction with cumene hydroperoxide, inducing efficient heterolytic O-O bond cleavage.


Introduction
Copper monooxygenases such as peptidylglycine α-hydroxylating monooxygenase (PHM), dopamine β-monooxygenase (DβM) and tyramine β-monooxygenase (TβM) employ two uncoupled, mononuclear copper sites referred to as Cu M and Cu H . 1-6 Cu M serves as the reaction site for dioxygen activation and substrate oxygenation, whereas Cu H functions as an electron acceptor from the physiological reductant (ascorbate). The Cu M site is ligated by two histidine imidazoles and one methionine sulphur (N 2 S donor set), whereas Cu H is coordinated by three histidine imidazoles (N 3 donor set). 3,4 These coordination environments are somewhat unusual, since most of the reaction centres of O 2 -activating copper proteins (type-2 and type-3 Cu proteins) consist of nitrogen rich donor sets, whereas the large number of electron transfer proteins (type-1 Cu proteins) involve a sulphur-containing ligand environment. 7,8 Thus, such a unique feature of these copper monooxygenases has long been a target in synthetic bioinorganic modelling studies. [9][10][11][12][13][14][15][16][17][18][19][20] So far, a series of mononuclear copper-dioxygen adducts have been characterised as model compounds of possible reactive intermediates of copper monooxygenases. [21][22][23][24][25] Among them, a mononuclear copper(II)-superoxide complex supported by a simple N 3 -ligand L N3 (1-[2-(2-pyridyl)ethyl]-1,5-diazacyclooctane derivatives, Fig. 1) is a unique example, which exhibits both a similar structural feature (distorted tetrahedral copper(II) with an end-on superoxide ligand) and a reactivity (aliphatic hydroxylation) to those of the putative reactive intermediate involved in the enzymatic reactions. 26,27 In this study, we have synthesised and characterised the copper(I) and copper(II) complexes supported by L N2S (Fig. 1) in order to get insights into the effects of the sulphur atom on the structure and reactivity. Ligand L N2S has the same molecular framework as that of L N3 , but one of the alkylamine nitrogen atoms is replaced by a sulphur atom, allowing us to simply access the electronic effects of the donor atoms.

Synthesis and characterization
The ligand L N2S was synthesized in three steps as shown in Scheme 1. Treatment of 2-(2-aminoethyl)pyridine and 3-chloro-1-propanol in CH 3 CN gave a bis(3-hydroxypropyl)-amine derivative, which was then converted to the dichloro derivative by the reaction with SOCl 2 . Ring closure using Na 2 S gave the ligand L N2S in a moderate yield.
The copper(I) complex [Cu I (L N2S )](PF 6 ) (1 N2S ) was prepared by treating L N2S with an equimolar amount of [Cu I (CH 3 CN) 4 ]-(PF 6 ) in THF. The crystal structure of 1 N2S is shown in Fig. 2. The crystallographic data and the selected bond lengths and angles of 1 N2S are presented in Table S1 † and the figure caption, respectively. 1 N2S exhibits a slightly distorted T-shaped structure ligated by the two nitrogen atoms N(1) and N(2), and one sulphur atom S(1) of the supporting ligand. The sum of the three angles around the copper center [∠S(1)-Cu(1)-N(1), ∠S(1)-Cu(1)-N(2), ∠N(1)-Cu(I)-N(2)] is equal to 360°, indicating that the copper ion and the three donor atoms exist in the same plane. The copper(I) complex 1 N3 supported by L N3 exhibited a similar T-shaped structure, demonstrating that the molecular framework of the ligands can highly stabilise such a three coordinate T-shaped structure of copper(I). 26,27 Although the precise structure of the reduced enzyme is not known, the EXAFS data have suggested that the reduced Cu M is 3-coordinate, or 4-coordinate with a weakly bound water molecule, where the Cu M -S Met distance is 2.24 Å. 4 In this context, 1 N2S represents a good structural model of reduced Cu M , even though the Cu-S distance in 1 N2S (2.203 Å) is slightly shorter than that in the enzyme (2.24 Å).
1 N2S showed a relatively intense absorption band at 375 nm (ε = 1058 M −1 cm −1 ) ascribable to a Cu I to S charge transfer (MLCT) (Fig. 3), and exhibited a reversible Cu I /Cu II redox couple at 0.23 V vs. Ag/AgNO 3 in acetone (Fig. 3, inset) that is higher than that of 1 N3 (0.11 V for R = -CH 2 CH 2 Ph) under the same conditions. 27 Copper(II) complexes 2 N2S ·X (X = coordinating counter anion; Cl, N 3 , NO 2 , OTf ) were also prepared as the model compounds of resting enzymes in the copper(II) oxidation state. The crystal structures of 2 N2S ·Cl, 2 N2S ·N 3 , 2 N2S ·NO 2 and 2 N2S ·OTf have been determined as shown in Fig. S1, S2, † and     All the copper(II) complexes exhibit five-coordinate structures with τ = 0.01-0.55, 28 among which 2 N2S ·Cl and 2 N2S ·N 3 form coordination polymer chain structures, where the coordinating counter anion (Cl − and N 3 − ) acts as the bridging ligand ( Fig. S1 †). On the other hand, 2 N2S ·OTf and 2 N2S ·NO 2 exhibit a monomeric structure as shown in Fig. 4 and Fig. S2. † Thus, the structure and chemical properties were examined using 2 N2S ·OTf in more detail.
The copper(II) complex 2 N2S ·OTf exhibits a distorted square pyramidal structure coordinated by N(1), N(2), S(1) and O(4) in the equatorial plane and O(1) occupying the axial position (τ = 0.30), which is in accord with the observed ligand field absorption band at 660 nm (ε = 222 M −1 cm −1 , Fig. 5) and a typical EPR signals of Cu II with a tetragonal geometry (Fig. 6, the EPR parameters are provided in the figure caption). The structural feature is quite different from that of the copper(II) complexes supported by L N3 , which enforces the copper(II) center having a four-coordinate distorted tetrahedral geometry. 26,27 Thus, the replacement of the donor atom from nitrogen to sulphur resulted in a significant change in the coordination geometry of the metal center from the four-coordinate distorted tetrahedral to the five-coordinate square pyramid or trigonal bipyramidal. In this case, Cu-S distance is 2.286 Å, which is longer than that in the copper(I) complex (2.203 Å). Such an elongation of the Cu II -S bond demonstrates the strong preference of the Cu I -S bond as compared to the Cu II -S bond. However, the Cu II -S distance in the model complex (2.286 Å) is much shorter than that in the resting state (copper(II) oxidation state) of the enzyme, that is, 2.68 Å, although the coordination geometries are similar to each other. 3

Reactivity
First, the reaction of 1 N2S and O 2 was examined in acetone at low temperature (−85°C). However, no spectral change was observed. Even at room temperature or in other solvents such as THF and CH 2 Cl 2 , 1 N2S showed no reactivity toward O 2 . This result is in sharp contrast to the high O 2 -reactivity of 1 N3 , which gives a mononuclear copper(II)-superoxide complex. 26,27 Such a big difference in the O 2 -reactivity between the two ligand systems could be attributed to the higher oxidation potential of 1 N2S (0.23 V) as compared to that of 1 N3 (0.11 V). As mentioned above, the Cu-S distance in the enzyme is much longer, especially in the copper(II) oxidation state, which may cause the less copper(I)-stabilizing effect of the sulphur donor atom in the enzymatic system. Therefore, Cu M may be able to activate O 2 for the substrate oxygenation reaction even in the presence of the Met ligand.
On the other hand, 1 N2S smoothly reacted with cumene hydroperoxide (CmOOH) in a 1 : 1 ratio to give CmOH (cumyl alcohol) as the major product (98%) together with a trace amount of acetophenone (PhC(O)Me), where CmOOH was completely consumed (Fig. S3 †). In the case of 1 N3 , on the other hand, only 44% of CmOH was obtained together with a small amount of PhC(O)Me (6%), and 50% of CmOOH was recovered (Fig. S4 †). Thus, the stoichiometry in the 1 N3 case was 2 : 1.
We have recently examined the reactivity of a series of copper complexes using CmOOH as a proof for searching the O-O breaking pattern in the copper-peroxide adducts. ‡ 29,30 It has been demonstrated that copper(I) complexes supported by nitrogen ligands react with CmOOH in a 2 : 1 ratio to induce the heterolytic O-O bond cleavage of the peroxide to give CmOH as the major product. In this case, each copper(I)  ). In contrast, heterolytic O-O bond cleavage gives cumyl alcohol (CmOH) after oxyanion protonation. 36,37 complex formally acts as a one-electron donor to induce total two-electron reduction of CmOOH, thus providing the heterolytic cleavage product, CmOH, in a ∼50% yield based on the copper(I) starting material (stoichiometry of Cu : CmOOH is 2 : 1). 30 Thus, 1 N3 (N 3 -donor ligand) giving CmOH in a ∼50% yield is involved in this category.
In this respect, the reactivity of 1 N2S (N 2 S-donor ligand) is different, showing the 1 : 1 stoichiometry to give CmOH in a ∼100% yield based on the copper(I) starting material. Such a unique reactivity of 1 N2S can be attributed to the presence of the sulphur atom. In order to get insight into the effect of the sulphur atom, the reaction was examined by UV-vis, resonance Raman and EPR spectra as follows. Fig. 7 shows the UV-vis spectral change for the reaction of 1 N2S and an equimolar amount of CmOOH in acetone. After about 1000 s, an intermediate A exhibiting characteristic absorption bands at 336, 465 and 625 nm was developed (Fig. 7, solid line). This intermediate was gradually converted to another intermediate B having absorption bands at 336, 415, and 635 nm (Fig. 7, dashed line). When the solution was warmed to room temperature, the absorption bands in the visible region disappeared, suggesting formation of a copper(I) product C (Fig. 7, dotted line). The EPR spectrum of the final reaction solution was essentially EPR silent, being consistent with the formation of a copper(I) complex (Fig. S8 †).
Then, the reaction was examined by resonance Raman spectrum at the low temperature. As shown in Fig. S5, † isotopesensitive Raman bands were observed at 887, 827 and 610 cm −1 , which shifted to 794 and 603 cm −1 , when 18  The sulphoxide ligand L N2SO was obtained nearly quantitatively after the work-up treatment of the reaction mixture with NH 4 OH aq (demetalation). The formation of L N2SO was confirmed by 1 H-NMR, IR and ESI-mass spectra ( Fig. S5 and S6, † also see the Experimental section), and the isotope labeling experiment using Cm 18 O 2 H confirmed that the oxygen atom incorporated into the ligand originated from the CmOOH used (Fig. S6 †). Oxygenated product L N2SO was not obtained at all in the reaction of ligand itself and CmOOH under the same experimental conditions. Although mechanistic details of the enzymatic reaction remain obscure, such a role of sulphur

Conclusions
In summary, copper(I) and copper(II) complexes of the N 2 S ligand (L N2S ) have been synthesised and characterised in order to get insights into the effects of the sulphur atom of the Met ligand in the enzyme active site. Structural examinations have suggested that the sulphur coordination in the present model system is stronger than that in the enzymatic system, causing higher stability of the copper(I) oxidation state toward O 2 . Nonetheless, a reactivity study using cumen hydroperoxide has clearly indicated that the sulphur atom helps in the O-O bond heterolysis of the peroxide intermediate. Such an effect of sulphur might be important to facilitate the enzymatic reactions. The present result also suggests that the sulphur atom of the Met ligand in the enzymatic system can be oxygenated to sulphoxide during the catalytic cycles. In other words, posttranslational modification of the Met sulphur could occur in the enzyme active site.

Materials and physical measurements
The reagents and the solvents used in this study, except the ligands, cumene hydroperoxide and the copper complexes, were commercial products of the highest available purity and were further purified by standard methods, if necessary. 32 Cumene hydroperoxide (CmOOH) used in this study was prepared by a reported procedure, 33 and purified by silica gel column chromatography (eluent: AcOEt-Hexane). 18  UV-visible spectra were taken in acetone or in acetonitrile on a Hewlett Packard 8453 photodiode array spectrophotometer equipped with a Unisoku thermostated cell holder designed for low temperature measurements (USP-203). 1 H-NMR spectra were recorded on a JEOL ECS 400 or a Varian UNITY INOVA 600 MHz. ESR spectra were recorded on a BRUKER E-500 spectrometer at −196°C equipped with a variable temperature cell holder or a JEOL JES-FA100. A Mn IImaker was used as the reference, and experimental error in the EPR parameter is ±0.001. Mass spectra were recorded with a JEOL JMS-700T Tandem MS station or a JEOL JMS-700. ESI-MS (electrospray ionization mass spectra) measurements were performed on a Mariner ESI-TOF instrument. Cyclic voltammetric measurements were performed with a Hokuto Denko HZ-3000 or HZ-7000 in deaerated solvent containing 0.10 M n-Bu 4 PF 6 as a supporting electrolyte. A GC working (BAS) electrode was polished with BAS polishing alumina suspension and rinsed with acetone before use. The counter electrode was a platinum wire. The measured potentials were recorded with respect to an Ag/AgNO 3 (0.01 M) reference electrode.
Resonance Raman scattering was excited 441.6 nm from a He-Cd laser (KIMMON KOHA, K5651R). Resonance Raman scattering was dispersed by a single polychromator (Ritsu Oyo Kogaku, MC-100) and was detected by a liquid nitrogen cooled CCD detector (HORIBA JOBIN YVON, Symphony 1024 × 128 Cryogenic Front Illuminated CCD Detector). The resonance Raman measurements were carried out using a rotated cylindrical cell thermostated at −80°C or a rotating NMR tube (outer diameter = 5 mm) thermostated at −40°C by flashing cold nitrogen gas. A 135°back-scattering geometry was used.

Synthesis
N-(Pyridine-2-yl)ethyl-bis(3-hydroxypropyl)amine. A 200 mL round-bottom flask was charged with 2-(2-aminoethyl)pyridine (4.0 g, 33 mmol), 3-chloro-1-propanol (13 g, 0.13 mol), sodium carbonate (35 g, 0.33 mol), tetrabutylammonium bromide (∼5 mg), water (70 mL) and CH 3 CN (70 mL). The mixture was refluxed for 2 days. The reaction mixture was cooled to room temperature and extracted with CHCl 3 (3 × 50 mL), and the combined organic fractions were dried over Na 2 SO 4 . After removal of Na 2 SO 4 by filtration, evaporation of the solvent gave a brown oil, from which the pure product was isolated by alumina column chromatography (eluent: AcOEt-MeOH = 95 : 5) in 70% (6.5 g). N-(Pyridine-2-yl)ethyl-bis(3-chloropropyl)amine. N-(Pyridine-2-yl)ethyl-bis(3-hydroxypropyl)amine (0.36 g, 1.3 mmol) in benzene (30 mL) was cooled to 0°C. At this temperature, 0.5 mL of thionyl chloride was added dropwise under aerobic conditions. After an initial exothermal reaction, the mixture was refluxed for 4 h, then brought down to room temperature and left to stand overnight. After cooling down again to 0°C, a solution of sodium carbonate (10%) was poured until the pH became basic. The organic layer was then separated, and dried over Na 2 SO 4 . After removal of Na 2 SO 4 by filtration, evaporation of the solvent gave a brown oil, from which the pure product was isolated by alumina column chromatography (eluent: AcOEt) in 78% (0.28 g). After stirring for 5 min, the insoluble material was then removed by filtration. The filtrate was concentrated to give a green oily material. Single crystals were obtained for the X-ray analysis by diffusion of n-hexane into a CH 2  Other copper(II) complexes 2 N2S ·Cl, 2 N2S ·N 3 , and 2 N2S ·NO 2 were prepared in a similar manner, as described in the ESI. †

Product analysis
Decomposition products of cumene hydroperoxide. An acetone solution (3.0 mL) of 1 N2S (0.5 mM) was cooled to −85°C, and then an acetone solution of CmOOH (1 equiv.) was added to the solution under a nitrogen atmosphere. The resulting mixture was stirred for 1 h at −85°C. After warming the solution to room temperature, anisole (0.5 mM) was added as an internal standard, and then decomposition products derived from cumene hydroperoxide were analyzed by using a HPLC system consisting of a Shimadzu 10A series. Reverse phase chromatography was performed on an ODS column (Cosmosil 5C 18 -AR-II, 250 mm × 4.6 mm, Nacalai tesque) at room temperature with an acetonitrile-water (40 : 60) mixed solvent as the mobile phase at a constant flow rate of 0.5 mL min −1 . The yields of products were determined by comparing the integrated peak areas of the products with that of the internal standard (anisole) using calibration lines. † Hydroxylated ligand (L N2SO ). An acetone solution (3.0 mL) of 1 N2S (12 mg, 27 μmol) was cooled to −85°C using a Unisoku thermostated cell holder designed for low temperature measurement. Then, an acetone solution of CmOOH (2 equiv.) was added to the solution under a nitrogen atmosphere. The resulting mixture was stirred for 1 h at −85°C, and then gradually warmed to room temperature. A mixture of organic materials was obtained by treating the reaction mixture with an NH 3 aq and subsequent extraction by CHCl 3 . The combined organic layer was dried over Na 2 SO 4 . After removal of Na 2 SO 4 by filtration, evaporation of the solvent gave a yellow oil. Each single crystal obtained was mounted on a CryoLoop (Hampton Research Co.) with a mineral oil, and all data of X-ray diffraction were collected at −170°C on a Rigaku R-AXIS--RAPID diffractometer using filtered Mo-Kα radiation. The structures were solved by direct methods SIR 92 or SIR 2008 and expanded using Fourier techniques. 34,35 The non-hydrogen atoms were refined anisotropically by full-matrix leastsquares on F 2 . The hydrogen atoms were attached at idealized positions on carbon atoms and were not refined. All structures in the final stages of refinement showed no movement in the atom positions. The calculations were performed using Single-Crystal Structure Analysis Software, version 4.0.