Magnetostructural Properties of Some Doubly-Bridged Phenoxido Copper(II) Complexes

Three new tripod tetradentate phenolate-amines (H2L1, H2L4 and H2L9), together with seven more already related published ligands, were synthesized, and characterized. With these ligands, two new dinuclear doubly-bridged-phenoxido copper(II) complexes (3, 4), and six more complexes (1, 2, 5–8), a new trinuclear complex (9) with an alternative doubly-bridged-phenoxido and –methoxido, as well as the 1D polymer (10) were synthesized, and their molecular structures were characterized by spectroscopic methods and X-ray single crystal crystallography. The Cu(II) centers in these complexes exhibit distorted square-pyramidal arrangement in 1–4, mixed square pyramidal and square planar in 5, 6, and 9, and distorted octahedral (5+1) arrangements in 7 and 8. The temperature dependence magnetic susceptibility study over the temperature range 2–300 K revealed moderate–relatively strong antiferromagnetic coupling (AF) (|J| = 289–145 cm−1) in complexes 1–6, weak-moderate AF (|J| = 59 cm−1) in the trinuclear complex 9, but weak AF interactions (|J| = 3.6 & 4.6 cm−1) were obtained in 7 and 8. No correlation was found between the exchange coupling J and the geometrical structural parameters of the four-membered Cu2O2 rings.

Herein, we describe a general effective procedure for the design of novel series of tripodal phenolate compounds containing pyridyl or aliphatic amine arms with N 3 O and N 2 O 2 chromophores, where the donor O atoms are provided by one or two phenolate groups (Scheme 1) and their corresponding doubly-bridged phenoxido-Cu(II) complexes. The magnetic properties of the complexes were investigated at variable temperatures and the evaluated exchange coupling constants, J, are discussed in relation to their molecular structures.

Synthesis of the Ligands and Complexes
The new tetradentate tripodal-phenolate amines (H 2 L 1 , H 2 L 4 and H 2 L 9 ) used in this work, together with related ligands (Scheme 1), were synthesized by a standard general procedure. In a typical experiment, a methanolic mixture containing 2,4-disubstituted phenol and the corresponding amine (2:1), as well as two molar amounts of an aqueous 37% HCHO and triethylamine (Et 3 N) are used, whereas in the case of 2-methoxy-phenols and pyridyl derivatives (HL 7 and HL 8 ligands, Scheme 1), equimolar ratios were employed for all reagents. Refluxing the solutions for 3 to 4 days, followed by evaporation of the solution resulted in the formation of the solid products in reasonable yields, but about 30% were observed in the bromo-phenolate H 2 L 3 . We should mention that too much evaporation of the resulting solutions may lead to the formation of an oily product and/or a semi-solid, which is hard to recrystallize and leads to impure products. The pure desired products were obtained upon recrystallization from ethyl acetate and activated charcoal (see Experimental section). The synthesis and characterization of H 2 L 2 , H 2 L 3 , H 2 L 5 , H 2 L 6 , HL 7 , HL 8 , and H 2 L 10 were recently reported by our group [30].
The doubly-bridged phenoxido Cu(II) complexes 1-6, 9, and the polymeric 1D 10 were obtatined by the reaction of a methanolic solution of Cu(NO 3 ) 2 ·3H 2 O with the corresponding tripodal-phenolate amine ligand and Et 3 N in the stochiometric 1:1:2 molar ratio. In the cationic complexes 7 and 8, copper(II) perchlorate was used with the bipyridyl-phenolates HL 7 and HL 8 , and Et 3 N (1:1:1), respectively. These reactions afforded the desired complexes in moderate to good yields (55-90%). An illustration for the synthesis of the doubly-bridged phenoxido moieties in the complexes derived from N,N-dialkylethylenediamines is shown in Scheme 2. Interestingly, although all double bridged phenoxido compounds display green or olive-green color, brownish-green and purple colored complexes were obtained in case of the trimeric complex 9 and the 1D-polymeric 10. Single crystals suitable for X-ray structural determination were obtained either from dilute methanolic solution and/or recrystallization from CH 3 CN or acetone. The complexes are slightly soluble in MeOH, but are more soluble in less polar solvents, such as CH 3 CN and DMSO. The isolated compounds were characterized by elemental microanalyses, spectroscopic techniques, and conductivity measurements as well as by single X-ray crystallography for the copper complexes. Scheme 2. Synthesis of the tripodal phenolate amine ligands and the doubly-bridged phenoxido-Cu(II) coordination core of complexes derived from N,N-dialkylethylenediamine backbone.

Characterization of the Ligands
Some very general features exist in the IR spectra of the tetradentate tripodal phenolate amine ligands, such as a very weak broad band or a shoulder over the frequency region 3140-3230 cm −1 , which was attributed to the stretching vibration, ν(O-H) of the phenolic groups, in addition to a series of weak to very weak bands observed over the range 2700-3050 cm −1 due to ν(C-H) stretching of the aliphatic and aromatic groups. The moderate intense band detected around the 1590-1600 cm −1 region is assigned to ν(C=N), whereas the moderate to strong series of bands shown in 1200-1590 cm −1 are most likely attributed to ν(C=C, C-O). The ESI-MS of the ligands in MeOH, showed the 100% m/z base peak that corresponds to the protonation of parent ligand (H 2 L n + H] + (n =2-5, 9, 10). The 1 H NMR (d 6 -DMSO) spectra displayed peak positions at 7.1-6.8 (protons-ph); 5.0-4.9 (phenolate-protons), 3.8-3.7 (CH 2 -py); 3.5-3.3 (CH 2 -ph); 2.5-2.1 ppm (CH 3 -ph). The pyridyl protons reveal their signals at δ = 8.7-7.1 in compounds H 2 L 1 , HL 4 and HL 8 . The hydroxyl phenolate protons were clearly observed in H 2 L 1 and H 2 L 5 compounds but was not seen in most compounds, most likely due to their low solubility in DMSO.

Characterization of the Complexes
The IR spectra of the complexes under investigation were almost similar to the spectral pattern observed in their parent ligands, except for the disappearance of the broad band or shoulder in the region 3140-3230 cm −1 of the ν(O-H) of the phenolic groups upon its deprotonation and/or coordination to the Cu 2+ ion. The two cationic perchlorate complexes The UV-vis spectra of the complexes under investigation, measured in DMSO and/or CH 3 CN at room temperature, displayed a single broad/shoulder band in the 610-750 nm region and a strong absorption band over wavelength region 410-500 nm. The latter intense band can be assigned to the bridged phenoxido charge transfer transition (L-O → Cu II LMCT) [12,20,63,64], whereas the former low intense broad band is attributed to d-d transition in five-coordinate Cu(II) complexes, which was occasionally accompanied with a weaker intense broad band around 850-890 nm. The d-d transition feature in solution is consistent with a distorted square pyramidal geometry (SP) around the central Cu(II) ion [32,48,49]. Thus, the distorted SP geometrical assignments observed in DMSO, CH 3 CN, or acetone solution, were retained in the solid state (see X-ray section). The solution spectra of the complexes in these media did not show any appreciable change over the one-week period, reflecting their high stability.
The position of λ max in the 610-750 nm region can be used as a criterion for the ligand field strength of the tripodal phenolate ligands; λ max decreases in the order: 9 (λ max = 748 nm) > 2 (λ max = 733 nm) > 5 (λ max = 708 nm) > 10 (λ max = 687 nm) > 6 (λ max = 655 nm) > 4 ≈ 8 (λ max = 630 nm) > 7 (λ max = 625 nm) > 1 (λ max = 620 nm). This means that the ligand field strength is decreasing in the reverse order and the strongest fields are for phenolate compounds containing pyridyl arms. However, we should mention that broadening and close location of this band in some complexes did not permit precise prediction of their actual ligand field strength, but results are consistent with previous results [65,66]. In general, ligand field strengths decrease with increasing chelate ring size, the presence of electronegative chlorine or bromine atoms in the phenolate groups, and/or steric effect on the coordinated N-aliphatic amine arms, which tends to reduce the electron density on the coordinated centers and hence its ligand field [65,66].

Magnetic Properties
According to the crystal structures, the complexes under investigations can be divided into six groups as follows: Group 1: Centrosymmetric di-µ-phenoxido-bridged dinuclear complexes with a distorted square-pyramidal geometries (1 and 2). Complex [Cu 2 (L 1 ) 2 ] (1) shows a distorted square-pyramidal arrangement (SP) (τ 5 = 0.23) around each copper(II) center with an axial Cu-N distance of 2.310(3) Å, Cu···Cu distance and Cu-O-Cu angle are 3.0915(7) Å and 103.44(9) • , respectively [30]. Thus, Complex 1 is expected to exhibit antiferromagnetic interaction (AF) between the two S = 1 2 spins through the two phenoxido bridges. Tem-perature dependence of magnetic susceptibilities, measured under an external magnetic field of 0.5 T, ranging from 2 K to 300 K are represented in Figure 3 in the form of χ M and µ M vs T plots, where χ M is the magnetic susceptibility per Cu 2 unit, µ M is the magnetic moment per Cu 2 unit, and T is the absolute temperature. The magnetic moment of 1.30 BM at 300 K is considerably lower than the spin-only value (2.45 BM) for two non-interacting copper(II) S = 1 2 ions. The magnetic moment exhibits a continuous decrease with lowering the temperature over the range 2-300 K and reaches 0.28 BM at 2 K. This magnetic behavior reveals significant AF interaction between the copper(II) ions. The magnetic data were analyzed by the Bleaney-Bowers equation based on the Heisenberg model: where g is the g value, J is the exchange coupling constant between the two copper(II) ions, p is the fraction of mononuclear copper(II) impurity, and Nα is the temperature-independent paramagnetism for each copper(II) ion [8]. The best-fitting parameters (g = 2.1 (fixed), 2J = −578 cm −1 , p = 0.016, and Nα = 60 × 10 −6 cm 3 mol −1 (fixed)) are in complete support for strong AF interaction. The crystal structure of [Cu 2 (L 2 ) 2 ] (2) also shows the same crystallographic similarity as 1 (τ = 0.24) [68] around each copper(II) center with the axial Cu-N distance of 2.3580(17) Å. The Cu···Cu distance and Cu-O-Cu angle are 3.0064(5) Å and 99.12(6) • , respectively. The µ M of 2 is 1.62 BM per dinuclear unit at 300 K, which is also considerably lower than the spin-only value and decreases with lowering of the temperature and reaching to 0.12 BM at 2 K. The best-fitting parameters (g = 2.10, 2J = −403 cm −1 , p = 0.0007, and Nα = 60 × 10 −6 cm 3 mol −1 ) disclose the expected strong AF interaction. The absolute 2J value is smaller than that of 1, but it reflects the strong AF interaction in 1 compared 2 and is in harmony with the associated wider Cu-O-Cu angle observed in 1.
Group 2: Pseudo-symmetric di-µ-phenoxido-bridged dinuclear complexes with distorted SP geometries (3 and 4). The crystal structure of [Cu 2 (L 3 ) 2 ] (3) showed that the complex exhibits distorted SP arrangements (τ = 0.06 and 0.37) around two copper(II) centers with axial Cu-N distances of 2.180(10) and 2.235(10) Å. The Cu···Cu distance is 3.047(2) Å and Cu-O-Cu angles are 98.8(4) and 100.4(4) • . Complex 3 crystallizes in noncentrosymmetric with different distortion around each copper(II) center. The magnetic properties of 3 is illustrated in Figure 4. The magnetic moment of 2.07 BM at 300 K is a little higher than those of 1 and 2, and it exhibits a gradual decrease with lowering the temperature, reaching 0.20 BM at 2 K. The magnetic data were analyzed by Equation (1), and the obtained best-fitting parameters (g = 2.14(1), 2J = −291 cm −1 , p = 0.009, and Nα = 60 × 10 −6 cm 3 mol −1 (fixed)) demonstrates the AF interaction between the two copper(II) centers. The absolute 2J value of 3 is smaller than those of 1 and 2, but with a weaker interaction. This may be attributed to the different distortion between the two copper(II) coordination environments, which causes poor overlap between the two magnetic orbitals compared to 1 and 2.  Figure 5). The best-fitting parameters to Equation (1) (g = 2.1 (fixed), 2J = −352 cm −1 , p = 0.036, and Nα = 60 × 10 −6 cm 3 mol −1 (fixed)) are consistent with considerable AF interaction between the two copper(II) ions. The absolute 2J value is comparable to those obtained in 3 and 4.  (7) and 96.93(7) • , respectively. The magnitude of the magnetic moment of 6 is 1.68 BM at 300 K. This value is lower than the spin-only value for copper(II), S = 1 2 ion, and the magnetic moment exhibits a continuous decrease with lowering temperature and reaches 0.11 BM at 2 K. The best-fitting parameters to Equation (1) (g = 2.10 (fixed), 2J = −500 cm −1 , p = 0.0009, and Na = 170 × 10 −6 cm 3 mol −1 ) showed a significant AF interaction between the two copper(II) ions. The magnetic behavior of 6 should be considered as similar to that of 5, irrespective of the relatively large -J value of 5, taking into consideration the similarity of the Cu-O-Cu angle, Cu-O-Cu-O dihedral angle, phenyl ring torsion angle (Cu-O-C-C angle), and phenyl out-of-plane shift angle t (Table 2), then one should expect strong AF interaction between the copper(II) centers in the two compounds 5 and 6.   Group 4: Centrosymmetric di(phenoxido)-bridged dinuclear complexes with a distorted (5 + 1) octahedral geometries (7 and 8). The crystal structural analysis of [Cu 2 (L 7 ) 2 ](ClO 4 ) 2 (7) shows a distorted octahedral (5 + 1) geometry with a semi-coordination Cu-OMe of 2.785(4) Å around each copper(II) center. If this semi-coordination is neglected as a substantial bond because of its long distance, then the coordination geometry can be regarded as a distorted square-pyramidal geometry with the τ 5 value of 0.41. Therefore, taking into consideration the long distance of the axial Cu-O(phenoxido) bond of 2.213(3) Å, as a result the magnetic orbital of each copper(II) ion may exist around the basal N 3 O plane mainly and not in the direction of the axial phenoxido-bridging bond, which leads to a poor magnetic interaction. The Cu···Cu distance and Cu-O-Cu angle are 3.130 Å and 97.72 (14) • , respectively. The magnetic data of 7 are illustrated in Figure 6. The µ B (2.65 BM at 300 K) is higher than the spin-only value (2.45 BM) for two non-interacting copper(II), S = 1 2 ions. The magnetic moment exhibits a continuous decrease with lowering temperature and reaches 0.51 BM at 2 K, suggesting an AF interaction. The best-fitting parameters (g = 2.14, 2J = −7.2 cm −1 , and Nα = 60 × 10 −6 cm 3 mol −1 (fixed)) are consistent with a weak AF interaction between the two copper(II) ions. Complex [Cu 2 (L 8 ) 2 ](ClO 4 ) 2 (8) exhibits very close crystallographic parameters like 7; distorted octahedral (5 + 1) arrangement with a semi-coordination of Cu-OMe of 2.791(3) Å around each copper(II) ion. The coordination geometry of this complex is similar to 7 and can be regarded as indicated above as a distorted SP geometry (τ 5 value of 0.40) with a relatively long axial phenoxido-bridging of Cu-O distance of 2.215(2) Å. The Cu···Cu distance and the Cu-O-Cu angle are 3.098(2) Å and 96.63(10) • , respectively. The similarity was also extended to the magnetic properties of 7. The best-fitting parameters (g = 2.1 (fixed), 2J = −9.2 cm −1 , and Nα = 60 × 10 −6 cm 3 mol −1 (fixed)) shows a weak AF interaction between the two copper(II) ions.
In 7 and 8, the Cu 2 O 2 bridging ring with the axial phenoxido-bridges is planar, as can be seen from the Cu-O-Cu-O dihedral angle of 0 • (Table 2), and thus the magnetic orbitals of each copper(II) ion should be parallel to each other. This suggests that the phenoxido oxygens are rather axially located relative to the basal dx 2 -y 2 orbital plane. Such axial bridging oxygens cannot effectively participate in mediating the magnetic interaction through the phenoxido-bridges.
Group 5: linear trinuclear complex 9. The crystal structure of [Cu 3 (L 9 ) 2 (µ-OCH 3 ) 2 ]·CH 3 COCH 3 (9) shows that the molecule has a µ-phenoxido-µ-methoxido-bridged trinuclear copper(II) complex with a linear arrangement of the three copper(II) ions. The magnetic data (Figure 7) revealed a magnetic moment of 2.50 BM at 300 K, which is lower than the spin-only value (3.00 BM) for three non-interacting copper(II), S = 1 2 ions. The magnetic moment exhibits a gradual decrease with lowering temperature and reaches 1.46 BM at 2 K. This behavior suggests a remarkable AF interaction between the copper(II) ions. The magnetic analysis was carried out with the susceptibility equation for the linear trinuclear copper(II) ions based on H = −2J Cu3 (S 1 ·S 2 + S 2 ·S 3 ). However, the fitting was not good, and rather better fitting was obtained when the magnetic data were analyzed by the dinuclear model using Equation (1). The results indicated that the magnetic data of 9 contain some impurities arising from the presence of dinuclear copper(II) species. Therefore, the magnetic data were analyzed with the aid of Equation (2) of linear trinuclear and dinuclear copper(II) model.
where p Cu2 is the fraction of dinuclear copper(II) impurity. The best-fitting parameters (g = 2.1 (fixed), J Cu3 = −59 cm −1 , p Cu2 = 0.315, 2J = −592 cm −1 , and Nα = 60 × 10 −6 cm 3 mol −1 (fixed)) revealed the existence of a considerable amount of dinuclear copper(II) species in the sample.  [9] with Cu···Cu distance of 7.269 Å. The temperature dependence of magnetic susceptibilities χ A and magnetic moments µ A are represented in Figure 8, where χ A and µ A are per Cu unit. The µ A value of 1.83 BM at 300 K is a little bit higher than the spin-only value (1.73 BM) for copper(II), S = 1 2 ion. The µ A vs. T plot shows a slight gradual increase with a lowering of the temperature over the range 50-300 K and a gradual decrease on lowering the temperature from 50 to 2 K, reaching, 1.70 BM at 2 K. While the increase of the µ A value suggests F interaction between two adjacent Cu(II) centers via the L 10 anion ligand, the corresponding decrease in µ A values indicates an AF intermolecular interaction. The magnetic data were analyzed by the molecular field approximation (Equation (3) [8]), where for this series Equation (4) for the Heisenberg model for ferromagnetically coupled S = 1 2 ions (derived by Baker et al. [10]), considering the magnetic interaction between the neighboring chain molecules as zJ' (z = number of interacting neighbors), was used: where x = J/2kT. The solid line in Figure 8 shows the calculated curve with best-fitting parameters of g = 2.1 (fixed), J = 3.5 cm −1 , Nα = 60 × 10 −6 cm 3 mol −1 (fixed), and zJ' = −2.9 cm −1 .

Conclusions
In this work eight dinuclear doubly-bridged-phenoxido copper(II) complexes revealed antiferromagnetic coupling (AF) varied from strong AF interaction (−J = 289-145 cm −1 ) for complexes 1-6 to very weak AF interaction (−J = 3.6 & 4.6 cm −1 ) in 7 and 8. The geometry around the central Cu(II) centers exhibits distorted square-pyramidal arrangement in the first four complexes, mixed square pyramidal and distorted square planar in 5 and 6, and distorted pseudo octahedral (5 + 1) arrangements in 7 and 8. Attempts were made to correlate the exchange coupling (J) to the structural parameters of the non-planar Cu 2 O 2 four-membered rings. These parameters include the Cu···Cu distances . Unfortunately, no satisfied correlations were observed between J and the mentioned geometrical parameters [69,70]. This is most likely due to the influence of the steric environment incorporated by the substituents into phenolate groups, and the terminal coordinated dialkylamine, and hence the geometrical arrangement around the central Cu(II) ions as well as the coplanarity of the phenolate groups with the bridging Cu 2 O 2 moiety [53,70,71]. These factors may play a key role in reducing the efficient overlap between the 3d Cu 2+ and 2p O-phenoxido magnetic orbitals bearing the unpaired electrons. In general and focusing the discussion only on Cu(II) complexes containing mono-and doubly-bridged phenoxido compounds with no other bridging ligands, our results agree with the magnetic properties determined in many phenoxido-bridged copper complexes, where relatively strong AF couplings were reported [1,54,69,70,72], regardless of the fact that some of these compounds were ferromagnetically coupled [1,54,72,73].
In contrast and unlike the bridged bis(phenoxido) dinuclear copper(II) complexes, where no correlation was observed between J and Cu-O-Cu bond angle, linear relationships were obtained in the corresponding bis(hydroxido) [74] and bis(alkoxido) [75] bridged copper(II) complexes. In addition, a similar linear correlation was reported in bis(phenoxido) macrocyclic complexes in which the unique conjugated π-electron inherited into the macrocycle skeleton constrained the ligand to adopt a planar configuration [71]. On the other hand, a relatively moderate AF coupling (-J = 59 cm −1 ) was evaluated in the trinuclear 9, where alternative bridged phenoxido and methoxido groups were determined. However, in the 1D polymer 10, a weak ferromagnetic interaction (J = + 3.5 cm −1 ) was obtained. Interestingly, in the latter complex, the Cu(II) centers were linked via the N,N-dimethylpropyl arms with long Cu···Cu distance (7.269 Å) and a tetrahedral distorted square planar four-coordinate geometry around each Cu(II) center [30].
Electronic spectra were recorded using an Agilent 8453 HP diode array UV-Vis spectrophotometer. Infrared spectra were recorded on a Cary 630 (ATR-IR) spectrometer. 1  The temperature dependent magnetic susceptibilities were measured over 2-300 K at the constant field of 0.5 T with a Quantum Design MPMS 3 (installed at Shimane University) for 1-3, 4, 6-10, and with a Quantum Design MPMS-7 (installed at Institute for Molecular Science (IMS)) for 5. The measured data were corrected for diamagnetic contributions [76].

Synthesis of the Organic Ligands
The new ligands H 2 L 3 , H 2 L 4 and H 2 L 9 illustrated in Scheme 1, and their characterization are given below, whereas the rest of the ligands shown in this scheme (H 2 L 1 , H 2 L 2 , H 2 L 5 , H 2 L 6 , and H 2 L 10 as well as HL 7 and H 2 L 8 ) were recently reported [30]

Synthesis of Copper(II) Complexes
A general method was used to synthesize the copper(II) complexes 3, 4, and 9: to a mixture containing the appropriate tripod phenolate-amine ligand (0.5 mmol) and Et 3 N (0.102 g, 1.00 mmol) dissolved in MeOH (20-30 mL), Cu(NO 3 ) 2 ·3H 2 O (0.122 g, 0.5 mmol) was added, and the resulting solution was heated for 5-10 min, filtered while hot through celite and allowed to crystallize at room temperature. The precipitate, which was collected by filtration, was washed with Et 2 O and dried in air. Crystals suitable for X-ray analysis were obtained from dilute methanolic solutions or by recrystallization from CH 3 CN, but acetone was used in complex 4.
Anal: calcd for C 47

X-ray Crystal Structure Analysis
The X-ray single-crystal data of the title compounds 3, 4, and 9 were collected on a Bruker-AXS APEX II 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 3. Data collections were performed with Mo-Kα radiation (λ= 0.71073 Å); data processing, Lorentz-polarization, and absorption corrections were performed using APEX and the SADABS computer programs [77][78][79]. The structures were solved by direct methods and refined by full-matrix least-squares methods on F 2 , using the SHELX [80][81][82] program library. Additional programs used: Mercury and PLATON [83]. Packing plots are given in the Supplementary Materials (Figures S1-S3).  (Figures S1-S3) and Figure S4 for temperature dependence of magnetic susceptibility and magnetic moment µ eff of 2. Table S1: Crystallographic data and processing parameters of 3, 4 and 9.