Magnetic and structural properties of dinuclear singly bridged-phenoxido metal ( II ) complexes †

The reaction of a methanolic solution containing the bi-compartmental phenolic ligand 2,6-bis[bis(2-pyridylmethyl)aminomethyl]-4-chlorophenol (L-OH) with MCl2·nH2O in the presence of NH4PF6 or NaClO4 afforded the dinuclear bridged-phenoxido dichlorido-metal(II) complexes [Co2(μ-L O)(H2O)2Cl2][Co2(μ-L O)(MeOH)2Cl2](PF6)2 (1), [Ni2(μ-L O)(MeOH)2Cl2]PF6 (2), [Ni2(μ-L O)(MeOH)(H2O)Cl2]ClO4·1.25H2O (3), [Cu2(μ-L O)Cl2]PF6·1/2MeOH (4) and [Zn2(μ-L O)Cl2]PF6·MeOH (5). The complexes were characterized by elemental microanalyses, conductivity measurements, IR and UV-Vis spectroscopy, mass spectrometry and single crystal X-ray crystallography. Each M(II) center within the dinuclear complex cations is octahedrally coordinated in complexes 1–3, and five-coordinated distorted square pyramidal in 4 and 5. Magnetic susceptibility measurements at variable temperature of the complexes 1–4 revealed weak to moderate antiferromagnetic coupling with |J| values = 8.38, 39.0, 30.2 and 0.79 cm, respectively. The results of DFT calculations correlate well with the experimentally determined antiferromagnetic coupling and show that the magnetic exchange coupling occurs mainly through the phenoxido bridge M–O–M. Implications of geometry around the central metal ion, M⋯M distance, M–O–M bond angle and overlapping of magnetic orbitals on the magnetic exchange coupling are discussed.


Introduction
The design of compartmental ligands capable of providing symmetrical and asymmetrical bimetallic cores is a growing topic.][3][4][5][6][7][8][9][10][11][12][13] With focus on compartmental ligands that are derived from phenolic containing compounds, these ligands have been launched to study the phosphodiester hydrolysis and DNA cleavage [1][2][3][4] and to model purple acid phosphatases, 5,6 Zn phosphesterases, 7,8 Mn catalases, 9 catecholase oxidases, 10,11 metallo-β-lactamases (MβL) 12 and hemocyanin. 13The use of these model compounds was very helpful to gain insight into biological systems and to elucidate some structural features about these systems.][16][17][18] Among many different types of binucleating compounds are phenol-based compartmental ligands, which possess two pendant chelating arms attached to the 2-and 6-positions of the phenol ring.][5][6][7][8][9][10][11][12][13] The distance between the two metal ions bridged via the phenoxido group is a crucial parameter in mediating the magnetic interaction between the two paramagnetic metallic centers.Also, their close proximity allows the cooperation between metal ions, where the distance between the two bridged metal ions are within the range of 2.9-4.0Å, providing an excellent pathway for a strong antiferromagnetic coupling between 3d [7][8][9] centers.In addition, the benzene ring present in these systems allows great synthetic flexibility, especially in tuning the solubility of the compounds.
As it was indicated above, the phenol-based compounds with 2,6-pendant chelating coordinating arms have the tendency to bind two similar or dissimilar metal ions through the deprotonated phenolic group.3][54] Therefore, this study was undertaken to explore the coordination properties of this class of compounds using bis[bis(2-pyridylmethyl)aminomethyl]-4-chlorophenol (L Cl -OH) (Chart 1) with Co(II), Ni(II), Cu(II) and Zn(II) metal ions.

IR spectra of the complexes
The IR spectra of the complexes display some general characteristic features:

UV-Vis spectra of the complexes
The UV-Vis spectra of the complexes 1-4 were measured in 1) displays three distinct bands at around 564, 500 and 463 nm.These bands are typical for six-coordinate high spin Co(II) complexes and they can be assigned to the spin-allowed transitions 4 T 2g (F) ← 4 T 1g (F), 4 T 1g (P) ← 4 T 1g (F) and 4 A 2g (F) ← 4 T 1g (F), respectively.Similarly, three bands were also detected for the complex cations [Ni  3 T 2g (F) ← 3 A 2g (F), 3 T 1g (F) ← 3 A 2g (F) and 3 T 1g (P) ← 3 A 2g (F), respectively. 57e spectrum of copper complex [Cu 2 (µ-L Cl O)Cl 2 ]PF 6 • 1/2MeOH (4) reveals the presence of two maxima at 457 and 695 nm.The broad single band at 695 nm suggests a distorted square pyramidal (SP) environment around the central Cu 2+ ion. 58In general, the visible spectra of the five-coordinate SP Cu(II) complexes are most likely producing a broad band over the 550-700 nm range (d xz , d yz → d x 2 −y 2) which occasionally may or may not be associated with a low-energy shoulder at λ > 800 nm, whereas the presence of a single d-d band at λ > 800 nm (d xy , d x 2 −y 2 → d z 2) with a high-energy shoulder is typical for trigonal bipyramidal (TBP) stereochemistry. 58,59nterestingly, the geometries of the complexes 1-4, as determined by UV-Vis spectroscopy in CH 3 CN solution, were in complete agreement with those obtained by single crystal X-ray crystallography.

Mass spectral characterization of complexes
ESI-Mass spectra of the five complexes were recorded in acetonitrile and all are shown in the ESI section (Fig. S1 ] + , which observed above in complex 4 when CH 3 CN was used as a solvent in measuring mass spectra, have been recently reported in some dinuclear cobalt(II) complexes. 2Complexes 1-5 reveal also the presence of peaks at m/z 356.03, 356.04, 356.04, 356.01 and 278.17, respectively.Although we were unable to identify the origin of these species most likely they are attributed to ligand fragments.The acetonitrile mass spectrometry of [Cu 2 (L Me -O) 2 (OCH 3 )](ClO 4 ) 2 revealed a peak at m/z 357.1, which is located at about the same positions as for complexes 1-4 was incorrectly assigned for [Cu 2 (L Me -O) 2 (OH)] 2+ ion. 27In addition to these peaks, the hexafluorophosphate complexes 1, 2, 4 and 5 displayed an m/z peak at 144.97 (100%) for PF    4) Å] (Fig. S6 †).

Magnetic properties of complexes 1-4
The experimental magnetic data of 1, depicted in Fig. 6, shows dominant antiferromagnetic exchange within the Co(II) dimer as evidenced by a decrease of the effective magnetic moment from 6.46μ B (300 K) to 0.88μ B (1.9 K) and also by the maximum of M mol vs. T curve located at 21 K, which serves as a fingerprint for antiferromagnetically coupled homospin dimers. 61urthermore, the significant zero-field splitting (ZFS) is expected in hexa-coordinate Co(II) complexes 62 and therefore the following spin Hamiltonian was postulated where the isotropic exchange ( J), the zero-field splitting (D) and Zeeman term (g) are included.Then, the molar magnetization in a given direction of magnetic field B a = B•(sin θcos φ, sin θsin φ, cos θ) was calculated as Z is the partition function resulting from diagonalization of the spin Hamiltonian matrix.Finally, the integral (orientational) average of molar magnetization was calculated by eqn (3) in order to properly simulate experimental powder magnetization data.
Moreover, the small amount of monomeric paramagnetic impurity (PI) which accounts for an increase of molar magnetization (mean susceptibility) below 3 K was taken into consideration by eqn ( 4) where M PI was calculated using the Brillouin function.Both temperature and field dependent magnetic data were included into fitting procedure which resulted in these parameters for 1: J = −8.38 cm −1 , D = 25.7 cm −1 , g = 2.39, χ TIP = 4.9 × 10 −9 m 3 mol −1 , x PI = 0.79% (χ TIP stands for the temperature-independent paramagnetism).The results confirmed the moderate antiferromagnetic exchange between Co(II) atoms and the substantial role of magnetic anisotropy as deduced from the axial zero-field splitting parameter D. The small discrepancies between experimental and calculated data are ascribed to the existence of two dimeric units within the asymmetric unit and also to the fact that one of the dimers forms supramolecular tetramers through O-H⋯Cl hydrogen bonds (Fig. S11 †).This probably creates more complex magnetic exchange pathways in the solid state.
The magnetic data for the dinuclear nickel(II) complexes 2 and 3 are plotted in Fig. 7, and Fig. S12, † respectively.In complex 2, the effective magnetic moment drops on cooling from 4.17μ B (300 K) to 0.26μ B (1.9 K) and the maximum of M mol vs. T curve was found at 57 K, thus confirming strong antiferromagnetic exchange with S = 0 ground state.Such strong antiferromagnetic exchange means that excited molecular spin states S = 1 and S = 2, which bear information about magnetic anisotropy (D), are too high in energy, and therefore low temperature isothermal magnetization data (M mol /N A μ B < 0.1, Fig. 7) are non-informative concerning this issue.Therefore, only temperature data were used during the fitting procedure, which resulted in J = −39.0cm −1 , g = 2.19, χ TIP = 2.2 × 10 −9 m 3 mol −1 , x PI = 0.51%.Compound 3 shows very similar magnetic properties and the same magnetic analysis resulted in J = −30.2cm −1 , g = 2.24, χ TIP = 3.4 × 10 −9 m 3 mol −1 , x PI = 0.37% (Fig. S12 †).
In contrast to the above discussed results, the copper(II) complex 4 exhibited different magnetic behaviour (Fig. 8).The effective magnetic moment is almost constant in whole temperature range (μ eff ≈ 2.57μ B ) and only below 10 K the small drop of μ eff is observed (μ eff = 2.32μ B at T = 1.9 K).Moreover, there is no maximum on M mol vs. T curve.Consequently, we can presume only a very weak antiferromagnetic exchange in 4. The magnetic analysis of both temperature and field dependent data resulted in J = −0.79cm −1 , g = 2.10, χ TIP = 0.24 × 10 −9 m 3 mol −1 , thus confirming a very weak magnetic exchange between Cu(II) atoms of the antiferromagnetic nature.

Evaluation of magnetic properties using DFT calculations
In order to support the data of magnetic results which are showing large differences in the isotropic exchange mediated by the L Cl -OH ligand for various metal atoms and possibly to get insight into the exchange mechanism, ab initio calculations based on DFT theory were used to calculate the J parameters in the dinuclear moieties , and [Cu 2 (µ-L Cl O)Cl 2 ] + of 4. The ORCA 3.0.1 computational package was used for all the calculations. 63Well established B3LYP functional 64 and def2-TZVP(-f ) basis set 65 were used to calculate the energy difference Δ, between high spin (HS) and broken-symmetry (BS) spin states: For the above mentioned structurally characterized dinuclear molecular fragments of 1-4 complexes, the following spin Hamiltonian for a dinuclear was used All the calculations utilized the RI approximation with the decontracted auxiliary def2-TZV/J Coulomb fitting basis set and the chain-of-spheres (RIJCOSX) approximation to exact exchange. 66Increased integration grids (Grid5 and GridX5in ORCA convention) and tight SCF convergence criteria were also used.The isotropic exchange J values were calculated by Ruiz's approach (eqn ( 7)) 67 and also by a more general Yamaguchi's approach (eqn (8)) 68 : The results of DFT calculations are summarized in Table 1.The J values calculated by DFT correlate well with the experimentally determined antiferromagnetic exchange in 1-4, and in the case of Co(II) and Ni(II) complexes the J's calculated by the Ruiz's approach are also very close to those found from magnetic analysis, whereas in Cu(II) compound, the experimentally determined J value is between J Ruiz and J Yam values.
In case of of 4, the spin densities (Fig. 9) and also the non-orthogonal magnetic orbitals (Fig. 10) were visualized with the help of software Gabedit. 69vidently, the unpaired electron of each Cu(II) atom is localized in the d x 2 −y 2 orbital, which lies within the CuN 3 Cl plane.Therefore, there is a very weak overlap between magnetic orbitals of Cu(II) atoms, which results in such very weak antiferromagnetic exchange.On the contrary, the two unpaired electrons of Ni(II) atom are localized in the d x 2 −y 2 and d z 2 orbitals, leading to efficient overlap between magnetic orbitals through the bridged phenoxido Ni-O-Ni bond resulting in a strong antiferromagnetic exchange.Also, a consensus between DFT and magnetic analysis results in Ni(II) compounds 2 and 3 was acquired, showing that antiferromagnetic exchange is weaker in the case of 3.This trend can be related to the   1).However, the effect of different solvent molecules (methanol or water) on magnetic exchange cannot be excluded.

Structure parameters and magnetic coupling
1][72][73] Copper(II) complexes in which a singly bridged phenoxido group is the only bridge exhibit very weak antiferromagnetic coupling as this was the case in complex 4 ( J = −0.79cm −1 ).This has been observed in a number of related complexes such as [Cu 2 (µ-L Me -O)Cl 2 ]ClO 4 ( J = 0 cm −1 ), 53    Table 1 DFT-calculated net Mulliken spin densities (ρ), expected values <S 2 >, and isotropic exchange parameters ( J) from high-spin (HS) and broken symmetry spin (BS) states of the dinuclear molecular fragments based on X-ray structures of 1-4 weak ferromagnetic coupling was also reported in some related systems: [Co 2 (µ-L1 NO2 -O)(µ-OAc) 2 ]PF 6 ( J = +3.09cm −1 ) and [Co 2 (µ-L Br -O)(µ-OAc) 2 ]PF 6 ( J = +0.78cm −1 ) (see Chart 1). 75o obvious correlation was found between the magnitude or sign of J and the nature of the substituents at the phenolic ring. 75In dicobalt(II) compounds, several factors were addressed to account for this weak interactions.These include the Co(II)-Co(II) and Co-O( phenoxido) bond distances as well as Co-O-Co bond angle, 2,71,74,76,77 but unlike coupled dicopper and dinickel(II) complexes, [78][79][80] the magneto-structural relationship for dicobalt(II) is not well resolved.The results of dicobalt(II) and dicopper(II) compounds clearly indicate that the µ-phenoxido bridge is a poor mediator for exchange magnetic interaction in this series of complexes.

Materials and physical measurements
The compound bis(2-pyridylmethyl)amine (DPA) was purchased from TCI-America.All other chemicals were commercially available and used without further purification.Infrared spectra were recorded on a JASCO FTIR-480 plus spectrometer as KBr pellets.Electronic spectra were recorded using an Agilent 8453 HP diode array UV-Vis spectrophotometer. 1 H and 13 C NMR spectra were obtained at room temperature on a Varian 400 NMR spectrometer operating at 400 MHz ( 1 H) and 100 MHz ( 13 C). 1 H and 13 C NMR chemical shifts (δ) are reported in ppm and were referenced internally to residual solvent resonances (DMSO-d 6 : δ H = 2.49, δ C = 39.4 ppm).ESI-MS were measured on LC-MS Varian Saturn 2200 Spectrometer.The conductivity measurements were performed using Mettler Toledo Seven Easy conductivity meter and the cell constant was determined by the aid of 1413 μS cm −1 conductivity standard.The molar conductivity of the complexes were determined from Λ M = (1.0 × 10 3 κ)/M, where κ = cell constant and M is the molar concentration of the complex.Elemental analyses were carried out by the Atlantic Microlaboratory, Norcross, Georgia U.S.A. Magnetic measurements of cobalt(II) (1) and nickel(II) compounds (2 and 3) were performed with a PPMS Dynacool VSM magnetometer (Quantum Design, Inc.) (T = 1.9-300K at B = 1 T; B = 0-9 T at T = 2, 5 and 10 K), while the copper(II) complex 4 was measured on an MPMS XL7 SQUID magnetometer (Quantum Design, Inc.) (T = 1.9-300K at B = 1 T; B = 0-5 T at T = 2 and 5 K).The magnetic data were corrected for diamagnetic susceptibilities and the signal of the sample holder.Caution: Salts of perchlorate and their metal complexes are potentially explosive and should be handled with great care and in small quantities.
Synthesis of 2,6-bis[bis(2-pyridylmethyl)aminomethyl]-4-chlorophenol (L Cl -OH).To 2,6-bis(chloromethyl)-4-chlorophenol (1.12 g, 5 mmol) dissolved in anhydrous CH 3 CN (50 mL), bis(2-pyridylmethyl)amine (2.00 g, 10 mmol) was added.The mixture was treated with anhydrous K 2 CO 3 (2.10 g, 15 mmol) and magnetically stirred under gentle reflux for 3 days, during which the color turned light yellow and a white precipitate was formed.This mixture was cooled in the refrigerator and then filtered off to remove KCl and unreacted K 2 CO 3 .The solvent was removed with a rotary evaporator under reduced pressure and the resulting dark brown liquid was solidified when stored over P 2 O 5 in a desiccator under vacuum.Several crystallizations from Et 2 O with the aid of activated charcoal afforded yellow oil which was then solidified to produce the desired product as a pale yellow solid (yield: 2.2 g, 80%).Characterization: mp = 102-104 °C, Selected IR (KBr, cm

Syntheses of metal(II) complexes (1-5)
A general method was used to synthesize the dinuclear dichloro metal(II) complexes (1-5).To a mixture containing MCl 2 •nH 2 O (M = Co, n = 6; M = Ni, n = 6; M = Cu, n = 2; M = Zn, n = 0) (0.40 mmol) and 2,6-bis[bis(2-pyridylmethyl)aminomethyl]-4-chlorophenol (L Cl -OH) (0.110 g, 0.20 mmol) dissolved in MeOH (20 mL), NH 4 PF 6 (80 mg, 0.50 mmol) or NaClO 4 (61 mg, 0.50 mmol) in case of complex 3 was added and the resulting solution was heated on a steam-bath for 5-10 min.The resulting solution was filtered while hot through celite and then allowed to crystallize at room temperature.The precipitate which was obtained over a period of 1-6 h was collected by filtration, washed with propan-2-ol and Et 2 O, and then dried at room temperature.Long needles ( pink for Co, bluish green for Ni, light green for Cu and colorless for Zn) suitable for X-ray structure determination were obtained from dilute methanolic solutions of complexes 1, 3, 4 and 5.In case of 2, recrystallization of the complex from CH 3 CN afforded bluish green single crystals of X-ray quality. [

X-ray crystal structure analysis
The X-ray single-crystal data of compounds 1-5 were collected on a Bruker-AXS APEX CCD diffractometer at 100(2) K.The crystallographic data, conditions retained for the intensity data collection and some features of the structure refinements are listed in Table 2.The intensities were collected with Mo-Kα radiation (λ = 0.71073 Å).Data processing, Lorentz-polarization and absorption corrections were performed using APEX, and the SADABS computer programs. 83The structures were solved by direct methods and refined by full-matrix least-squares methods on F 2 , using the SHELXTL program package. 84All non-hydrogen atoms were refined anisotropically.The hydrogen atoms were located from difference Fourier maps, assigned with isotropic displacement factors and included in the final refinement cycles by use of HFIX (parent C atom) or DFIX (parent O atom) utility of the SHELXTL program.Molecular plots were performed with the Mercury program. 85In case of 1, data affected by twin components (8.6%) were excluded from refinement.In case of 3, U ij constraints were applied for disordered water oxygen molecules and their hydrogen atoms excluded from refinements.

( 1 )
Medium broad absorption band in the 4430-4450 cm −1 region due to the stretching frequency ν(O-H) of the coordinated and/or solvent of crystallization CH 3 OH and H 2 O. (2) A series of strong to medium absorption bands over the 1610-1430 cm −1 region attributable to the pyridyl groups {ν(CvC) and ν(CvN)}.(3) The very sharp strong absorption band observed in complexes 1, 2, 4 and 5 around 840 cm −1 is assigned to the ν(P-F) of PF 6 − anion, whereas Ni(II) complex 3 displays two strong absorption bands at 1120 and 1097 cm −1 due to the stretching frequency ν(Cl-O) of the perchlorate counter anion.The split of the perchlorate band in [Ni 2 (µ-L Cl O) (MeOH)(H 2 O)Cl 2 ]ClO 4 •1.25H 2 O (3) may be attributed to the strong involvement of the ClO 4 − ions in H-bonding with the aqua or MeOH molecules (see X-ray section).This interaction reduces the symmetry of the ClO 4 − ion from T d to C 3υ or C 2υ .

Fig. 8
Fig.8The magnetic data for 4: Left: the temperature dependence of the effective magnetic moment and molar magnetization measured at B = 1 T. Right: the isothermal magnetizations measured at T = 2 and 5 K. Open circles represent the experimental data and solid lines represent the best fit using eqn (1) with J = −0.79cm −1 , g = 2.10, χ TIP = 0.24 × 10 −9 m 3 mol −1 .