DPPH Free Radical Scavenging Activity, and their Synthetic, Structural, Electronic and Magnetic Studies of Two [Cu^II₂(thb)] Compounds, (thb = N,N,N´,N´-tetrakis-{3-[(2-hydroxibenzyliden)-amine]propyl}-1,4- butanodiamine)]

Branched polymers are three-dimensional molecular structures, which have transformed the classic polymer concept, leading to develop a new field in chemistry [1-18]. In the biological world, dendritic molecules are present at different length scales, while in synthetic chemistry, there are many possibilities to synthesize branched macromolecules by changing the terminal groups of the different cores obtained. This strategy has allowed the modification of some physical and chemical properties and it has been applied to obtain a large number of new families of polymers with diverse applications [3-18]. Particularly, the poly(propylenimine) (PPI) is one of the first families of commercially available synthetic dendrimers with a well-defined structure that can be functionalized up to the 64^th generation. The PPI core is 1,4-diaminobutane (DAB) which has been transformed to 4, 8, 16, 32 and 64 generations, containing the same number of terminal amine (Am) groups [1,2]. So far, the spatial treats of these highly mobile molecules have been inferred; however, the spatial structures are still unknown. Importantly the covalent dendrimers structures have shown interesting applications, such as, antibacterial [6], nanoparticles containing metals [7-14], catalysts for different reactions [13,14], metal-encapsulators [14-18], nanoparticles templates for growing carbon nanotubes [18], among others. Hence our interest in this molecular polymer. We previously report the preparation and characterization of the branched ligand N,N,N´,N´-tetrakis-{3-[(2-hydroxibenzyliden)- amine]propyl}-1,4-butanodiamine) thb [19], used here to obtain a Cu^II complex in two different crystalline phases, [Cu^II₂ (thb)] 1a and 1b, both characterized electronic and magnetically, with considerably good antioxidant activity as shown below resolving, the 1b spatial structure. Herein, the scavenging activity of [Cu^II₂ (thb)] is studied, using the known free scavenging assay with the stable radical DPPH by UV-Vis [20]. The antioxidant activity relationship of coordination compounds containing transitional metals ions, like Copper and Chromium, has been proven in both, in vitro and in vivo models, for example, 1,1-diphenyl-2-picrylhydrazyl (DPPH) kinetic analysis, γ-radiolysis of rat liver microsomes, 2,2- azobis(2-amidinopropane)dihydrochloride (AAPH)-induced low density lipoprotein (LDL) oxidation, as well, AAPH-induced linoleic oxidation. Free radicals are, in most cases, produced physiologically by the cells. One example is the formation of the superoxide anion radical (O2- ·) which is formed during the mitochondrial respiratory chain reaction and the nitric oxide radical (NO) is made by the activation of nitric oxide synthase [21]. In some cases, the overproduction of radicals, as well as the formation of toxic radicals causes damage of biological molecules, resulting in many disease conditions. Oxidative stress is generated by the increase of free radical production and it has been linked to inflammatory processes. Hence, coordination compounds with radical scavenging activity are used to inhibit the free radical activity then to prevent several diseases [22].

Materials and methods

1a, 1b solutions were prepared by dissolving crystalline samples. Electronic spectra were measured with a Shimadzu UV-3100S spectrophotometer at 298 K (λ=200-1200 nm) in CH₂Cl₂ solution and CH₃OH solutions for 1a, 1b, at ca. 10^-5M. A Nicolet Magna-IR 750 spectrophotometer (?=400-4000 cm^-1) was employed to monitor the infrared spectra using KBr pellets. ¹H- and 13C-NMR spectra of 1 and only ¹H-NMR of 1a, 1b were recorded on a JEOL Eclipse-400 at rt using CDCl₃ solutions and TMS as reference. ¹H-NMR study of 1a, 1b, was carried out in a range from 100 to -100 ppm with acquisition time of 1 sec and 22000 repetitions. ESR spectra at X-band frequencies (9 GHz) for 1a, 1b were obtained with a JEOL JES-RES 3X spectrometer, at different temperatures from 300 K to 77 K, on polycrystalline powder samples and using 2,2-diphenyl-1-picrylhydrazyl (DPPH) as primary standard. ESR spectra at 77 K of 1a, 1b in CH₂Cl₂ solution using different concentrations (10^-3 to 10^-4M) were recorded with 2 min scan time, 1 × 0.1 mT modulation, 0.03 ms time constant, 1 mW power, 2.5 × 100 mT fieldwidth, 5 × 10 gain for 1a; with 2 min scan time, 1 × 0.1 mT modulation, 0.03 ms time constant, 1 mW power, 1.5 × 100 mT fieldwidth, 4 × 10 gain for 1b. X-band power variation experiment at 77 K for 1b powder sample was carried out from 0.01 to 160 mW. Measurements were performed in a magnetic field of 1 kOe, in the range 2-300 K, using a gelatine capsule and a Quantum Design Magnetometer. Calculated Pascal constants for 1a, 1b were about 264 × 10^-6 emu.

General procedure for synthesis of compounds 1a, 1b

Synthesis of 1a, 1b was carried out using the direct synthesis method [23,24]. A stoichiometric ratio of 2:1, Cu⁰ 0.6022 mmol:1 0.3011 mmol, respectively, and dimethylsulfoxide (2 ml) were placed in a flask. The reaction mixture is kept under stirring at rt for five days, and then filtered to give a green powder, which is dissolved in methanol. Green crystals are formed in the solution. The crystals of 1b were separated manually. The yield for 1a chloroform crystals is lower (yield: 30%) than that for 1b methanol crystals (yield: 60%) and the latter are suitable for X-ray crystallography.

1a: mp 165°C; Found: C, 61.49; H, 6.07; N, 9.74. Calc. for Cu₂C₄₄H₅₂N₆O₄: C 61.8; H 6.08; N 9.82; UV-Vis λ_max(CHCl₃)/nm (ε/M^-1cm^-1), 224 (30848.33), 241 (28545.11), 262.61 (9989.74), 277.31 (11015.94), 290.23 (6220.03), 365 (6345.69), 447.57 (18.33), 648.31 (112.57), 811.39 (486.92), 959.02 (348.18); IR (KBr) ?_max/cm^-1, 447.4 and 426.2 (Cu-N), 470.5 (Cu-O), 1633.4 (C=N), 1600.7 (C=C); ¹H-NMR δ (CDCl₃; (CH₃)₄Si): 89.09, 76.02, 54.74, 21.28, 8.89, 3.49, 2.62, 1.55, ?25.96, ?49.90, ?54.74, ?77.24, ?86.75; ESR (powdered sample) at 300 K/77 K: 2.117/2.118; χ_MT at 182 K: 0.7181 emu mol^-1 K (spin only 0.3760 emu mol^-1 K).

1b: mp 215°C; Found: C, 61.48; H, 6.05; N, 9.63. Calc. for Cu₂C₄₄H₅₂N₆O₄: C 61.8; H 6.08; N 9.82; UV-Vis λ_max(CHCl₃)/nm (ε/M^-1cm^-1), 224 (31372.4), 239 (26073.85), 255.17 (11371.97), 276.52 (10722.74), 297.47 (5110.05), 364.72 (5487.02), 446.79 (20.94), 642.57 (194.71), 808.55 (277.49), 938.28 (193.72); IR (KBr) ?_max/cm^-1, 455.1 y 420.4 (Cu-N); 472.5 (Cu-O); 1635.4 (C=N); 1598.7 (C=C); ¹H-NMR δ(CDCl₃; (CH₃)₄Si): 89.09, 76.02, 54.74, 21.28, 8.89, 3.49, 2.62, 1.55, ?25.96, ?49.90, ?54.74, ?77.24, ?86.75; ESR (powdered sample) at 300 K/77 K: g₁(300/77) = 2.190/2.192, g₂(300/77) = 2.101/2.103, g₃(300/77) = 2.053/2.052; χ_MT at 200 K: 0.4630 emu mol^-1 K (spin only 0.3760 emu mol^-1 K).

X-ray diffraction data of 1b

A green needle crystal mounted on glass fibre was studied with an Oxford Diffraction Gemini "A" diffractometer with a CCD area detector (λ_MoKα= 0.71073 Å, monochromator: graphite) source equipped with a sealed tube X-ray source at 130 K. Unit cell constants were determined with a set of 15/3 narrow frame/runs (1° in ) scans. A data sets consisted of 337 frames of intensity data collected with a frame width of 1° in , a counting time of 100 s/frame, and a crystal-to-detector distance of 55.00 mm. The double pass method of scanning used to exclude any noise. The collected frames integrated by using an orientation matrix determined from the narrow frame scans.

CrysAlisPro and CrysAlis RED software packages were used for data collection and data integration [25]. Analysis of the integrated data did not reveal any decay. Final cell parameters were determined by a global refinement of 6928 reflections (θ < 26°). Collected data were corrected for absorption effects by using analytical numeric absorption correction with a multifaceted crystal model based on expressions upon the Laue symmetry using equivalent reflections [26]. Structure solution and refinement were carried out with the programs: SHELXS97, SHELXL97 [27], and the software used to prepare material for publication was WinGX [28,29].

Full-matrix least-squares refinement was carried out by minimizing (Fo²-Fc²)². All non-hydrogen atoms were refined anisotropically. The H atom (H–O) of the hydroxo group was located in a difference map and refined isotropically with U_iso(H) = 1.5 U_eq for (O). H atoms attached to C atoms were placed in geometrically idealized positions and refined as riding on their parent atoms, with C–H = 0.95–0.99 Å and U_iso (H) = 1.2Ueq(C), or 1.5 U_eq (C) for aromatic, methylene and methyl groups. Crystal data and experimental details of the structure determination listed in Table S3.

DPPH radical scavenging assay

The hydrogen donating or radical scavenging ability of Cu^II₂(thb) was evaluated using a stable radical, 1,1-diphenyl-2-picryl hydrazyl (DPPH, Sigma, St. Louis, MO, USA). The reaction mixture was composed of 1.5 mL of 0.2 and 0.3 mM DPPH solution and 20 μL of the test compound in a total volume of 3 mL of methanolic solution. Stock solutions of Cu^II₂(thb) were prepared in 100% methanol and the final concentrations ranged from 5–100 μM. This assay was performed in triplicate in the two different phases of the Cu^II₂(thb) compound named 1a/1b and the free radical (thb) at 0.2 mM. The reaction mixtures were incubated at room temperature for 15 minutes then the solutions were measured by a Shimadzu UV-3100S spectrophotometer UV-Vis (λ =200–1100 nm).

Synthesis of [Cu^II₂(thb)] 1a, 1b

The organic ligand N,N,N’,N’-tetrakis-{3-[(2-hydroxibenzyliden)- amane]propyl}-1,4-butanodiamine (thb) was synthetized as previously reported [19]. 1a and 1b were obtained by direct synthesis method [18,23,24]. The absence of residual salts favours the formation of pure compounds, separated as solid samples from the reaction medium [24,30,31]. In the present work, this methodology has been essential in order to obtain two polymorphs of the complex [Cu₂ II(thb)], 1a and 1b (Scheme 1), one of which, 1b, was obtained as a single crystal. The 1a molecular structure must be the same as 1b because in solution both gave the same spectroscopic response and the same antioxidant activity, but in solid state, the spectroscopic results are different, so we conclude that 1a and 1b conatin the same molecules with different network arrangement.

Scheme 1: Crystallography of 1b.

Crystallography of 1b

The asymmetric unit consist of two Cu^II-(hidroxibenzyliden) molecules of the 1b compound. Four methanol crystallization solvent molecules stabilize to four dinuclear Cu^II complexes in the monoclinic crystal system with P2₁/c space group. The 1b compound has two main coordination sites; each site contains two salen groups and a tertiary amine group, acting as a five-dentate ligand around the Cu^II ions, (Figure 1). The dinuclear unit the geometry around each Cu^II ion can be described as an intermediary between a square-pyramidal (sp) and a trigonal bipyramidal (tbp) geometry, consistently with the geometric τ parameters of 0.58 found for Cu1 and of 0.63 for Cu₂ [32,33]. Axially, Cu1 is coordinated to N4 and N₆, while Cu₂ is coordinated to N1 and N3 of the imine groups. The equatorial plane around the Cu^II ions is formed by N5, O1 and O2 for Cu1, and by N2, O3 and O₄ for Cu₂. The longest bond distances are 2.345 (3) and 2.337 (3) Å for Cu-N5 and Cu₂-N2 respectively.

Figure 1: (a) Molecular structure of the dinuclear unit complex 1b, and b) thermal ellipsoids at 40% probability level for 1b. H atoms are omitted for clarity.

Intermolecular interactions of type O-H…H are observed on crystalline structure of 1b which stabilize the network crystalline (Figure 2a): i) interactions between the hydrogen (H1D) atom of methanol solvent and the oxygen atom (O1) of thb, H1D···O1 (2.812(5) Å); ii) interactions between the thb hydrogen (H7) atom and the oxygen (O₄) atom of a nearby thb molecule, H7···O₄ (3.481(6) Å); iii) interactions between the thb hydrogen (H15) atom and the oxygen (O3) atom of a nearby thb molecule, H15···O3 (2.675(3) Å); iv) interactions between the thb hydrogen (H17) atom and the oxygen (O3) atom of a nearby thb molecule, H17···O3 (3.585(6) Å). All the interactions construct an infinite layer along the a-b plane, forming a three-dimensional supramolecular arrangement, (Figure 2c). The crystalline structure of 1b consists of molecules parallel; to the b-axis, which are stacked along the a-axis (Figure 2b-c).

Figure 2: a) Molecules of 1b along the a-axis; b) One dimensional 1b (1D) zigzag chain Cu₂-O₄···N₆-Cu1-N4 ···O3 showing the Cu^II-phenolato top groups in the opposite direction of the Cu^II-phenolato bottom groups; c) View of 1b along the a-axis showing the Cu^II-phenolato heads with the same angle between consecutive Cu^II.

Thb phenolato top heads form zig-zag chains due the Cu₂-O₄···N₆- Cu1 inter-dinuclear interactions [O₄···N₆ distance: 4.164(9) Å; N4···O3 distance: 4.442(4) Å], meanwhile the phenolato at the bottom heads of each Cu^II-thb molecule form, zig-zag chains due the Cu1-N₆···O₄- Cu₂ inter-dinuclear interactions [O₄···N₆ distance: 4.164(9) Å; O3···N4 distance: 4.442(4) Å] (Figure 2c). Finally, the Cu-Cu intramolecular distance was 10.1391(9) Å maintaining the Cu ions isolated from each other [33,34].

UV-Vis and IR spectroscopy

The UV-Vis spectra of 1a and 1b (Supplementary information in Figure S1a, b), in CH₃OH solutions are very similar. The hydroxyl of the phenolate groups are deprotonated by coordination with the copper ion, therefore, the has two electron lone pairs, which showed one electronic transition in the UV-Vis spectra. On the other hand, the N lone electrons pair are used to form the coordination bond with the copper ion. The other transitions are assigned to two metalligand charge transfer transitions and one ligand-metal charge transfer transition (Supplementary Information in Table S2). Supplementary Information inset in Figure S1 showed the d_xz, d_yz² →d_z², d_x² _-y²→d_z² and dM_xy→ d_z² transitions for 1a at 811 nm with shoulders at 648 nm and 959 nm, and for 1b (inset in Figure 3a) at 808 nm, 642 nm and 938 nm. These transitions are characteristic of five-coordinated Cu^II compounds with distorted geometries [34-36]. The d-d transitions for 1a and 1b spectra helped to build an OM diagram, which is showed in Figure 3b. The oscillator strength, f, values for the d-d transitions were calculated and their energies values are among those found for d-d forbidden transitions and the reported values for energies corresponding to pure sp and pure tbp geometries [35,36]. The reported values for 1a and 1b showed in Supplementary information in Table S2 were very close. When 1a, 1b in methanol solutions were treated with NaBH₄ to Cu^I, the colour of both solutions changed from green to brown, suggesting the formation of Cu^I-clusters. For these solutions the d-d and CT transitions in the UV-Vis spectra disappeared, and an exponential curve emerged, where the Mie Plasmon pick at 590 nm was not observed. Similar spectra were obtained for compounds with particles size < 5 nm and non-isolated Cu^I atoms; therefore, the molecule size of 1a and 1b were < 5 nm (Supplementary Information, in Figure S2) [7,10].

Figure 3: χ^-1_M vs T curves for a) 1a (?), and b) 1b (?). The insets show μ_eff vs T curves. The fit lines using modified the Curie-Weiss model are in red for both phases.

IR spectra of 1a and 1b are characteristic of metal transition compounds showing characteristic peaks of the thb, which shifted due to the presence of Cu^II ions. The strong electronegative of Cu^II removes not only the electron density of the atoms in the first coordination sphere but also from neighbouring atoms through an inductive effect (Supplementary Information, in Figure S3a and b) [33,37]. Both spectra did not show the ν_O-H vibration peak indicating that the coordination of the phenol oxygen atoms to Cu^II as phenolato groups. This feature was in agreement with the 1a, 1b UV-Vis spectra, which did not show the two n(N)→π* (imino-phenolato) transition bands, observed in the thb UV-Vis spectrum, and only one band corresponding to the n(O)→π* (imino-phenolato) transition observed in the thb ligand UV-Vis spectrum [19].

¹H-NMR of 1a, 1b

Previously, our group reported ¹H-NMR studies of paramagnetic compounds with and without aromatic protons, which are very similar to the ¹H-NMR spectra of 1a, 1b [36-38]. Supplementary Information in Figure S4 showed the ¹H-NMR spectra with three important regions, showing different relaxation times. Two paramagnetic zones at downfield (100 to 10 ppm zone I) and one at upfield (-90 to -1 ppm, zone II), where the broad bands are contained; meanwhile the diamagnetic zone (0 to 10 ppm, zone III) contains narrower lines. In diamagnetic zone III, the signals correspond to the aromatic protons, which are outside of a sphere with radius >5 Å from the metal ions, while the broader lines in zones I, II in ¹H-NMR correspond to all protons within a sphere with average radius <5 Å from the metal ions. ¹H-NMR spectra of 1a, 1b were characteristics of paramagnetic complexes, with broad lines due the dipolar relaxation and the hyperfine protons coupling, which were sufficiently close to the metal and/or for which electron spin nuclear spin relaxation times are shorter than 5 ns [39-44].

Magnetic measurements

χ^-1_M vs T bulk measurements from 2 to 300 K are presented in Figure 3a, b. The linear behaviour of χ^-1_M vs T, in a temperature range of 2 to 83 K for 1a and 2 K to 133 K for 1b, shows that both phases are paramagnetic. Insets in Figure 3a, b, show the μ_eff vs T plots, indicating that the μeff values are temperature-independent between 13 K to 133 K for 1a and between 11 K to 206 for 1b.

In order to have better insight of the magnetic behaviour of both compounds, different models for fitting the susceptibility were used, including the Ising [45], Bonner-Fisher [46], Myers [47] and Bleany- Bowers models [48]. A good fitting (r² = 0.999 for both compounds) was obtained using the Curie-Weiss model (C-W) [49], with constant values of C_{1a/1b C-W} = 0.82/0.66 cm3 K mol^-1 and Curie-Weiss temperature values of θ_{1a/1b C-W} = 0 ± 1 K, (Support Information, in Figure S5 a and b, red solid lines). However, the best fitting (r² > 0.999) for both compounds was obtained with a modified Curie-Weiss model (modified C-W) (eq 1) [50] (Figure 4a, b red solid lines), with Curie constant values of C_{1a/1b modified C-W} = 0.80/0.70 cm³ K mol^-1 and modified Curie-Weiss temperature values of θ_{1a/1b modified C-W} = 0.6 ± 1 K.

   (1)

Figure 4: ESR spectra of powder samples show the broadening of the lines with the decreasing of the temperature for (a) 1a; (b) 1b, respectively.

where θ_{modified C-W} is a correction to the Curie-Weiss temperature, T is the temperature, αT is a phenomenological susceptibility and along with χ_{modified C-W} give the total Pascal constants. It is clear that the modified Curie-Weiss model fitted better, however, both models are adequate.

Regarding 1b, the Cu^II ions in the dinuclear molecules were well separated as it was shown in the crystallographic section, therefore, the molecules can be considered as monomers as they are. The same conclusion can be applied for 1a. The diamagnetic hydrocarbonated bridge supports the isolated ions, avoiding the intramolecular and intermolecular magnetic interactions, which is consistent with the magnetic studies.

ESR measurements

X-band ESR spectra for 1a, 1b in powder samples, were recorded at 77 K and 300 K (Supplementary information, in Figure S6). For any paramagnet the Zeeman interaction is expressed via the g-tensor which has three mutually orthogonal main values of g. The deviation of the g-values from free electron value (ge = 2.0023) carries information about the orbital angular momentum of the electron and information about the electronic structure [50]. At 77 K/300 K 1a ESR spectra (Figure S6a) showed a broad singlet with g value of g⁷⁷/g300 = 2.118/2.117. ESR of 1b in power samples at 77 K/300 K (Supplementary information, in Figure S6b) showed a rhombic spectra with g values of g₁ ⁷⁷/g₁ ³⁰⁰ = 2.192/2.190, g₂ ⁷⁷/g₂ ³⁰⁰= 2.103/2.101 and g₃ ⁷⁷/g₃ ³⁰⁰ = 2.052/2.053. All spectra did not show any additional resonances at zero field or at any other field. Neither hyperfine nor superhyperfine interactions were observed [49]. The obtained g-values of 1a, 1b are in agreement with those reported for Cu^II-dendrimers complexes [51-55]. In order to corroborate the rhombic spectrum for 1b, we carried out power variation experiments in which all the spectra features increased at the same rate as the microwave power. These results clearly indicate the presence of only one Cu^II specie. (Supplementary Information, in Figure S7). The ESR spectra differences indicate that structurally there is a higher asymmetry for 1b in the solid state. However, both spectra showed no well-define signals produced by magnetic interactions or by short relaxation times present in the samples [54,56,57]. It should be noted that g values for 1a, 1b did not change when the temperature was decreased to 77 K, this happens because the magnetic structure of the Cu^II-thb complex remains unchanged.

Furthermore, the sequences of powder ESR spectra of 1a, 1b shown in Figures 4a and 4b are snapshots of the ESR characteristics that were displayed when the temperature was changed. The general morphology of the ESR lineshape remains unchanged, despite of the mentioned marked spectral-detailed changes, however, at lower temperature there was an increase of the signals, due to the peaks implying a decreasing in the branch mobility. It is noteworthy that at 77 K, the ESR linewidth spectra increased 19% for 1a and 20% for 1b, when comparing with the spectra at 300 K. This behaviour is explained by the spin-lattice relaxation mechanisms for flexible macromolecules, dipolar spread and the hyperfine interactions for paramagnetic compounds [56,57].

According to Mabbs [53], the coincidence of the principal values of the g and  tensors depend on the point symmetry at the metal. Considering the relationship among the g and  tensors, the symmetry point of paramagnets and the type of ESR spectrum, it can be concluded that the local symmetry around the Cu^II ion for 1a is <T where all the tensor axes are coincident g_xx=g_yy=g_zz, A_xx=A_yy=A_zz, and for 1b, the Cu^II local symmetry is <D3h, with tensor axes g_xx≠g_yy≠g_zz, A_xx≠A_yy≠A_zz. The point symmetry for 1b is agree with the results obtained in the UV-Vis and X-ray measurements.

It is important to observe that the ESR spectra of 1a and 1b samples in solutions at different concentrations showed the same ESR singlet spectrum with g values, modified by Δg ≤ 0.013 when the concentration was changed (Figure 5a, b).

Figure 5: ESR spectra of a) 1a and b) 1b in CH₂Cl₂ at 77 K at different concentrations.

Both spectra did not show hyperfine interaction, indicating dominant dipolar interactions. The fact that in solid ESR studies were different for 1a and 1b, but the spectra in solution became similar confirmed that 1a and 1b are two different crystalline phases with the same molecular structure.

DPPH free radical scavenging assay to prove biological activity

DPPH radical is commonly used as a radical molecule to evaluate the free radical scavenging capacity of several chemicals. The DPPH is characterized by a signal at 514 nm by UV-Vis, and it is able easily to accept an electron or hydrogen atom (proton) to become a stable diamagnetic molecule. Antioxidants, in the other hand, easily donate hydrogens to DPPH, in our case this role is played by the donation of an electron by Cu^II. Thus, the loss of the DPPH transition signal intensity in the presence of antioxidants is clearly directly proportional with the concentration (or number) of protons accepted. As shown in Figure 6, 1a and 1b were able to rapidly suppress the signal intensity after addition of the [Cu^II₂(thb)] in concentration dependent manner. As negative control, we analyze the free ligand at 0.1 mM giving no inhibition at all of the DPPH activity.

Figure 6: Lost of DPPH free radical activity by 1a and 1b in a concentration dependant manner.

The DPPH assay proved the scavenging skill of both phases of [Cu^II₂(thb)], however, to test the real strength of them we calculated the IC₅₀, which define the concentration needed to diminish certain activity at its half. An antioxidant reference is the Vitamin E analog, Trolox, which is already in the market. By comparing the IC₅₀ [Cu^II₂ (thb)] = 16.5 μM versus the IC₅₀ Trolox = 36.9 μM in undebatable that [Cu^II₂(thb)] is strong as Troxol inhibition wise.

Mechanism of scavenging by [Cu(thb)] 1a, 1b

In order to justify the loss of the free electron in the DPPH, 1a/1b must be able to gain it. We followed this reduction process from Cu^II → Cu^I by UV-Vis, Figure 7a and b. As we detailed above, the d_xz, d_yz ²→d_z ², d_x ² _-y ²→d_z ² and dxy→ dz ² transitions for 1a/1b appear around 810 nm (Supplementary Information, in Figure S8). The increase of DPPH concentration from 0.2 to 0.3 mM fade away the d-d transitions. This evidence describe that the scavenging compound is positively the Cu^II ions of 1a and 1b.

Figure 7: a) UV/Vis spectra of the DPPH free radical signal suppression by addition of 1a and 1b. b) Proposal of scavenging process by dz 2 orbital of Cu^II ion.

The spatial, electronic and magnetic measurements in solid and solution samples allowed us to conclude that two different crystalline phases 1a and 1b were obtained. The crystalline structure of 1b is one of the first obtained for branched molecules, to our knowledge, and showed that 1b consist of a dinuclear complexes, where the Cu ions are separated by a hydrocarbonated chain. Therefore, the magnetic data could be fitted by the Curie-Weiss and modified Curie-Weiss models, both models. Both models are adequate for complexes without exchange interactions. The above was confirmed by the ESR spectra, which showed g values ~2.1 for both phases, which is consistent with a spin state s = ½. For solid samples of 1a and 1b, the ESR spectra are completely different; in solution, both spectra were similar. The antioxidant activity of 1a and 1b was observed in UV-Vis spectra by the DPPH free radical scavenging assay showing a decreasing in the forbidden bands d-d and quantifying its capacity as IC₅₀ [Cu^II₂ (thb)] = 16.5 μM respect to IC₅₀ Trolox = 36.9 μM. Both compounds gave the same antioxidant behavior observed through the UV-Vis spectra DPPH test, proving that 1a and 1b are the same molecules with crystalline arrangements. Structurally, the intermediate sp and tbp geometries give possibilities to the DPPH to approach enough, allowing the free electron to be transferred into the Cu^II centre. Electronically, it is confirmed that the Cu^II reduction occurs in the d_z² orbital, being the highest energetically and the only one available to accept one electron, moreover, d_z² has the ideal geometry to make it possible. Magnetically, the ions communication is almost null, characteristic that should be noted. The spectroscopic and magnetic studies of the novel Cu^IIpolymers, 1a and 1b, provide much information about the electronic, magnetic and spatial structures and its correlation with the antioxidant activity, being the more important contribution of this paper.

Secretaría de Educación Pública, Subsecretaría de Educación Superior, and Vicerrectoría de Investigación y Estudios de Posgrado from BUAP, Project REOY-NAT11, 12/2011, 2012-1, have supported this work. The authors are grateful to Dr. Sylvain Bernès (Universidad Autónoma de Nuevo León, México) for his participation in discussions on the correlation between spatial structure and magnetic behaviour and Dr. Roberto Escudero (Instituto de Investigaciones de Materiales, UNAM) for the magnetic measurements.