Synthesis, Ligating Properties, Thermal Behavior, Computational and Biological Studies of Some Azo-transition Metal Complexes

Synthesis of new Fe(III), Co(II), Ni(II), and Cu(II) complexes of two azo ligands; 1-(phenyldiazenyl) naphthalene-2-ol (sudan orange R, HL1), and sodium 2-hydroxy-5-[(E)-(4-nitrophenyl) diazenyl]benzoate (alizarin yellow GG, HL2) have been reported. Stoichiometries of 1:2 and 1:3 (M:L) of the synthesized complexes were approved by total-reflection X-ray fluorescence technique (TXRF) and by elemental analyses. The geometry of complexes (octahedral and square planar) was typified by various spectroscopic, thermal, and magnetic techniques. The ESR spectroscopy showed that Cu(II) complexes are of different isotropic and rhombic symmetries with the existence of Cu–Cu ions interaction. TGA, DTA, and DSC analyses supported the multi-stage thermal decomposition mechanisms, where the thermal breakdown is ended by the formation of metal oxide in most cases. Moreover, chemical reactivity modeling using the density functional theory (DFT) method with the B3LYP/6–31 basis set, showed that metal complexes are more biologically active than their precursor ligands. The calculated lipophilicity character for metal complexes is in the range of 33.8–37.5 eV. Docking results revealed high scoring energy for [Fe(HL2)3].H2O complex and moderate inhibition strength of [Cu(L1)2].H2O complex versus 1bqb, 3t88, and 4esw proteins. Ultimately, the extent of biological effectiveness was endorsed experimentally against four microbial strains. The results are guidelines for toxicological investigations.


Introduction
Azo ligands are a class of compounds extensively used as indicators in analytical chemistry to show different colors in different media. They have an important role in qualitative and quantitative analyses. Most of these compounds possess an azo group conjugated to one or two arenes. The azo group has great interest for various reasons, such as models of biological systems, antifungals, and analytical reagents [1,2]. Due to the planarity of the azo bridge versus the nonplanarity of the rest of the system, the azobenzene moiety produces a large π-electron transmission effect and leads to high optical activity. Such activity favors the use of azo-benzene compounds as nonlinear optical (NLO) materials [3,4]. Besides, the (-N = N-) chromophore together with the conjugated system have induced light sensitivity to these compounds and hence they are incorporated in chemosensors and laser technologies [5]. Azo compounds are also engaged in some therapeutic applications due to their considerable DNA binding affinity [6]. Moreover, the coordination of transition metals with azo-containing ligands produces compounds with significant photocatalytic performances [7]. Azo-containing compounds and their metal complexes are reported to be involved in some biological reactions.
For example, alizarine yellow GG contains an azo-linked salicylate moiety which has a structural resemblance to the sulfasalazine drug and can interact with the antiporter protein in the cell membrane [8]. Also, sudan orange ligand is a lipophilic compound that has the capability of interacting with serum albumin which induces conformational changes in the structure of albumin [9]. This article is part of continuing investigations of the structural chemistry of biologically active complexes containing pyrimidine, purine, azo, nitro, nitroso, and azo-nitroso chromophoric groups [10][11][12][13][14]. The major interests of this paper are to study the thermal stability and antimicrobial effects of newly synthesized transition metal-azo complexes derived from sudan orange R (HL 1 ) and alizarin yellow GG (HL 2 ). The geometry and binding modes were determined by elemental and spectroscopic techniques. The thermal stability was evaluated by studying the decomposition behavior at different temperatures up to 600 ºC. Recently, the in-silico study has become an efficient and essential step to prognose the extent of reactivity of designed chemical structures in many research works [15,16]. As such, molecular docking was implemented against protein receptors to predict the biological feature of the synthesized compounds. Finally, based on the theoretical calculations, complexes were assessed by in vitro study for their antipathogenic

Synthesis of Metal Complexes
All the metal complexes were prepared in a similar manner, where a required mass equivalent to 10 mmol of transition metal chloride salt: CoCl 2 .6H 2 O (2.37 g), NiCl 2 .6H 2 O (2.37 g), CuCl 2 .2H 2 O (1.70 g) and FeCl 3 .6H 2 O (2.70 g) was dissolved in (20 ml) distilled water. Solution of the ligand in ethanol was added quantitatively to the metal chloride solution to obtain different (M:L) stoichiometries; (1:2) for Co 2+ , Ni 2+ and Cu 2+ while (1:3) for Fe 3+ . The reaction mixture was refluxed for 2 h, then cooled. The resulting complexes were filtered and dried at ~ 90 °C. The metal contents of the synthesized complexes were analyzed by three different techniques: the total-reflection X-ray fluorescence (TXRF), [19] atomic absorption spectrophotometric technique (AAS) and titrimetrically with standard EDTA solution using the appropriate indicator. The complexes are synthesized according to Schemes (1 and 2).

Theoretical Approach
Geometry optimization of selected synthesized compounds was achieved using Gaussian 09 software with Gaussian view for visualization and schematic numbering. The preferred conformers with the minimum energy were obtained by DFT with B3LYP/6-31G base-set without constraining the symmetry [20]. The yielded molecular indices gave excellent predictions about the chemical and biological features of the investigated compounds. Moreover, the ligand-protein receptor binding model for complexes was developed by the MOE-2015 software package. Three protein receptors have been chosen from the Protein Data Bank; 1bqb, 3t88, and 4esw which are corresponding to the threedimensional crystal structures of S. Aureus, E. Coli, and C. Albicans proteins, respectively [21][22][23]. Protein repair steps were staged before docking including the elimination of solvent molecules, the addition of hydrogens, chain fixation and selection of active sites of the protein backbone. Also, the tested compounds were oriented for docking by energy minimization, potential energy adaption, atomic charge and binding energy calculations using Merck Molecular Force Field, MMFF94x. This field has been approved to be satisfactory in dealing with a broad range of diverse structural arrangements to evaluate the stability of H-bonds and van der Waals adducts [24] Among thirty ligand-receptor poses, five conformers were selected which represent the best-implanted ligand molecule inside the protein active pocket with bond lengths less than 3.5 Å and of highest scoring energy.

Results and Discussion
Seven complexes derived from HL 1 and HL 2 azo-ligands are synthesized. The complexes are characterized by spectroscopic (IR, UV-Vis, ESR), thermal (TGA, DTA, DSC) and magnetic susceptibility measurements. Also, Molecular geometry and electronic parameters are determined by the DFT method of calculation.

FT-IR Study
The FT-IR spectra of the investigated azo-ligands and their complexes were studied to characterize the mode of bonding between the metal ion and the ligand (Table S1 and Figs. 1, S2). The IR spectrum of the HL 1 free ligand exhibited bands at 3462, 1622, 1499, 1449, 1321 and 1208 cm −1 , which are assigned to ν OH , ν C=C , ν C-C , ν N=N , ν C-N , and ν C-O vibrational modes, respectively [25]. The characteristic infrared spectral band of HL 1 complexes (Fig. S2) showed a remarkable change in the intensity and slight shift in the position of ν N=N bands upon complexation. Besides, the disappearance of the ν OH bands in the IR spectrum of the complexes (Fig. S2) suggests the deprotonation of the phenolic OH group. Noteworthy, the broadband that appeared in the region > 3300 cm −1 in the IR spectrum of the HL 1 complexes could be attributed to the water of crystallization which is approved later by thermal analysis. Accordingly, HL 1 behaves as a bidentate ligand, coordinating via the oxygen atom of the OH group in the naphthalene moiety after a loss of proton and the lone pair of electrons on the nitrogen of the azo group forming a stable six-membered ring. This mode of coordination is supported by the disappearance or lowering in the intensity of the bands due to bonded OH group [26] and the stretching frequency of the azo-group with the simultaneous appearance of new bands in the frequency ranges 608-557 cm −1 and 497-371 cm −1 of ν M-O and ν M-N , respectively [27].
Likewise, the characteristic ν OH band at 3468 cm −1 for the free ligand HL 2 was slightly shifted in all complexes (Table S1 and Fig. 1) indicating the involvement of the OH group in the complex formation without deprotonation. The ν C-O band at 1390 cm −1 in the free ligand exhibited a considerable change in the position and intensity upon complexation indicating the contribution of the carboxylic oxygen in the complex formation [28]. Additionally, the bands due to the scissoring and wagging of the carboxylate group in the range 900-600 cm −1 displayed noticeable changes in their intensities [29]. The band at 1495 cm −1 of the free ligand spectrum is due to the stretching vibration of the azo group, ν N=N , in an azobenzene-like structure in a trans configuration [30]. The ν N=N band is nearly unaffected through complexation which rules out the participation of the azo group in the chelation. The asymmetric and symmetric stretching vibrations of the carboxylate group at 1540 and 1410 cm −1 , respectively [31], appeared clearly in all HL 2 -complexes, implying the involvement of this group in the chelation moiety. Hence, all HL 2 -complexes are neutral, where the charge on the central metal ion is balanced by binding to the COO − groups of the coordinated ligands ( Fig. 1).

UV-Vis, NMR, MS, and Magnetic Susceptibility Studies
The UV-Vis. spectrum of the HL 1 -metal complexes showed three electronic absorption bands at 268, 308-320, 370-478 nm in DMF (Table 1 & Figs. S3-S6). The band at 268 nm is attributed to π → π * resulted from the interaction of π-electrons in the conjugated system. Also, the bands in the range 308-320 nm are assigned to n → π * developed from the non-bonding electrons on the azo group. However, the intense broad bands in the range 370-478 nm are due to the 2 MCl 2 .nH 2 O Reflux for 2hrs.

ESR of Copper (II) Complexes
The room temperature X-band powdered electronic spin resonance of the square planar [Cu(L 1 ) 2 ].H 2 O derived from HL 1 (Fig. 5) showed an anisotropic spectrum of rhombic type [34]. The g-values are g 1 = 2.639, g 2 = 1.788 and g 3 = 1.741. The < g > value is equal to 2.056. Here, the deviation of the < g > value from the free electron 2.0032 could be due to the misalignment of the local copper (II) environment and the probability of Cu-Cu ion exchange [32]. The R-parameter which is calculated by the relation R = (g 2 -g 1 )/ (g 3 -g 2 ) equals to 18.11 (R > 1) implying that the ground state is d z 2 [32]. However, the ESR spectra of the square planar [Cu(HL 2 ) 2 ] 0.2H 2 O complex derived from HL 2 is of isotropic spectral pattern of dynamic or pseudo rotational type of Jahn-Teller distortion (Fig. 5) at g iso = 1.85 with A = 333.33 × 10 -4 cm −1 . Isotropic spectrum is common for complexes containing misaligned "tetragonal" axes of magnetically dilute interaction [34]. The observed broadening of the resonance line suggests a heightened magnetic dipole interaction between the paramagnetic centers due to the polymeric nature of this complex [35].

Thermal Analysis (TGA, DTA, and DSC)
The TGA curves for the investigated azo-ligands and their complexes are collected in Figs. 6 & S14. The thermal decomposition pattern and assignment of the removed species are presented in Tables 2 & S2. A slight difference was  (Fig. S14). This may be attributed to a strong attachment of the outer sphere H 2 O to the coordination sphere through hydrogen bonds which delays its removal during the thermal decomposition [36].
Generally, the thermal decomposition routes of the complexes proceed in three main stages where the bond between the central metal ion and the ligands starts to dissociate after losing small molecules such as H 2 O, NO 2 , and CO 2 (Tables 2 & S2 Table 2) supports the bonding of the Fe(III) to the carboxylate group as indicated in the spectral data and similarly reported complexes in literature [37].
The DTA curves for the entitled azo-ligands and some of their complexes are presented in Figs. 7 & S15-S18. The considerably high thermal stability of some complexes was directly inferred from the first decomposition DTA peak maximum (T m ), Table 2  of the phenyl groups. The last stage is accompanied by an exothermic peak at 430.3 ºC due to the liberation of the naphthalene moieties of the ligand which is ended with the formation of NiO as a final product [38]. The exothermic nature of the thermal decomposition steps of [Ni(L 1 ) 2 ]. H 2 O complex is verified by the positive direction of DTA curves (Fig. 7). Also, the DTA curve for Fe-HL 2 complex, [Fe(HL 2 ) 3 ].H 2 O (Fig. S15) exhibits an exothermic peak at 99.7 ºC corresponding to two TGA steps assigned to dehydration of one water molecule in the atmosphere of the complex, and the loss of 2 NO 2 . The last step appeared as an exothermic peak at 318.5 ºC assigning the liberation of the rest of the ligand which ended with the formation of FeCO 3 . The DSC measurements for HL 2 and its Ni-complex, [Ni (HL 2 ) 2 (H 2 O) 2 ]0.4H 2 O, were carried out under a flow of N 2 at a heating rate of 10 °C min −1 in the temperature range of 25-200 °C. Apparently, both HL 2 and its Ni-complex exhibited distinct transitions from glassy disordered to the more ordered crystalline solid phase. HL 2 displayed a glass transition temperature, T g , at 57 °C (Fig. 8). This temperature represents the change in the heat capacity of the disordered solid without exhibiting any phase transition. However, [Ni (HL 2 ) 2 (H 2 O) 2 ]0.4H 2 O complex showed a higher glass transition temperature, T g = 85 º C, than the free ligand owing to the decrease in the free space of the complex as a result of coordination to the metal ion [39]. The crystallization temperature, T c , which is indicative of the transition from disordered to crystalline solid, appeared at 135 and 149 °C for HL 2 and its Ni-complex respectively. The melting process of the investigated compounds was not discerned within the studied temperature range (25-200 °C). The melting temperatures are 274 °C for HL 2 and > 300 °C for the Ni-complex as measured experimentally by Fisher-Johns melting-point apparatus. The elevated melting point values further support the thermal stability of the synthesized complexes up to at least 300 °C (Tables 1 and 2).

Computational Analysis
Optimization of the geometrical structures of all studied compounds to their minimum energy values was achieved using the density functional theory (DFT) method of calculation with B3LYP/6-31G basis sets. Electrophilic and nucleophilic regions on the ligand's surfaces were evaluated aiming to investigate the optimal binding sites for the interactions with metal ions that were proposed experimentally in this contribution. Also, the stability of the square planar and octahedral geometries of complexes was examined by DFT calculations without enforcing symmetry around the metal centers. The yielded data are displayed in Tables 3, S3-S5, Figs. 9, 10, 11 & S19-S23.

Molecular Geometry and Electronic Parameters
The optimization of geometries of HL 1 , HL 2 and their metal complexes by DFT calculation provided different possible conformers with diverse bond angles and bond lengths. The most stable conformers, i.e. the ones with the minimum structural energy, are given in Figs. 9 & S19-S21. The bond lengths of the functional groups of the optimized structure of HL 1 are stated in Table S3 as a (Table S4) exhibited slight elongation in the bond length relative to the free HL 1 (Table S3) (Table S4). This finding points to the square planar geometry around Cu(II) center, with a small discrepancy of 2-5º from the ideal square planar angles (90º) probably due to the bulkiness of the ligand structure [40,41]. Moreover, the molecular properties of the studied compounds such as the energies of the highest occupied (E HOMO ) and the lowest unoccupied (E LUMO ) molecular orbitals (Figs. 10 and S22) as well as the chemical reactivity parameters; electronegativity (χ), electrochemical potential   Table 3. Generally, the energy gap between HOMO and LUMO is an index for the reactivity of the compounds; such that, molecules with the least energy gap are soft and more reactive towards the metal ions and biological receptors compared to those having high values of energy gap [42]. According to the energy gap values of the studied ligands (Table 3), HL 2 is chemically more reactive than HL 1 . Also, the other calculated molecular parameters; χ, μ, η, S and ω point to the same order of reactivity. Furthermore, the calculated energy gap for the [Cu(L 1 ) 2 ]. H 2 O complex (Fig. 9) is quite small (1.594 eV) and hence, this complex is expected to possess some biological activity. Worth to mention that the low dipole moment of all complexes indicates their higher lipophilicity than the corresponding free ligand [43].

Molecular Electrostatic Potential (MEP)
The preferable sites for electrophilic and nucleophilic chemical interactions of a compound can be predicted by constructing the electrostatic potential map on its surface [44]. The color gradually alters from red which represents the regions with high electron density to blue which indicates the regions with electron deficiency. Figures 11  and S23 illustrate the molecular electrostatic potentials of HL 1 , HL 2 , and [Cu(L 1 ) 2 ].H 2 O complex. For instance, the most negative potential sites in HL 1 are N (19) and N(20) of the azo group, whereas, the positive potentials are on the hydrogen atoms of the rings with the most electrophilic site on H(18) (Fig. 11a). In the same manner, the active nucleophilic sites are located on the electronegative atoms O (25), O (28) and O(29) in HL 2 (Fig. 11c). Also, the low negative potential on the azo group of HL 2 ruled out its possibility of being a site of coordination with the metal ion which is in good harmony with the spectral data of the current research. Finally, constructing MEP for biologically active metal complexes is an important step in the drug designing procedure. Allocating the nucleophilic and electrophilic centers may shed light on their possible behavior with the cell membrane of microorganisms [44]. [Cu(L 1 ) 2 ].H 2 O complex showed a positive electrostatic potential on the metal core whereas the negative potentials are found in the outer region of the complex (Fig. S23).

Molecular Docking of Complexes
The activity of the synthesized metal complex as possible antipathogenic agents was estimated by molecular docking against three receptors: 1bqb, 3t88, and 4esw corresponding to S. Aureus, E. Coli, and C. Albicans proteins respectively [21][22][23]. The designated pathogenic micro-organisms are Gram-positive, Gram-negative bacteria and fungi that appear on the surfaces of various objects and cause superficial infections of human skin as well as mucous membranes infections. The docking results are displayed in Tables (Fig. S24). Also, the best docking poses (Fig. 12b) revealed high ligand occlusion in the electronic cloud of the protein surface that points to high interaction efficiency and great inhibition activity toward the selected pathogens. The simulated antipathogenic activity of the Cu(II) complex is in agreement with similar reported Cu(II) complexes [45,46] and hence was satisfactory to perform the in vitro biological activity which

Antimicrobial Assay of the Synthesized Metal Complexes
The synthesized metal complexes were screened for their activity of inhibition against one Gram-positive (S. Aureus), one Gram-negative (E. Coli) and two fungi (C. Albicans and A. Flavus) utilizing the diffusion agar technique [11,18]. The comparison between the inhibition zone diameter values of the synthesized complex with the standard drug values of Amphotericin B & Ampicillin ( Notably, Cu(II) complex showed more inhibitory effects than its free ligand HL 1 . The superior biological activity of the complexes relative to their precursor ligands is very common and can be explained by Tweedy's chelation theory [47], where the chelation between the metal ion and the ligand causes a decrease in the polarity of the chelate by delocalization of the π electrons [48]. The allocation of electrons over the entire complex surface enhances its lipophilicity and hence its penetration capability through the phospholipid bilayer membrane of the biological cells [49]. Moreover, metal ions disturb the transfer of electrons in the electron transport chain by acting as terminal electron acceptors within the microbial cell [50,51]. No activity was observed against A. Flavus. The yielded data of the isolated [Cu(L 1 ) 2 ].H 2 O complex proposed its antipathogenic potency against the tested strains. However, the less potency of HL 1 could be due to its higher polarity that was predicted theoretically which led to unsuccessful disperse through the bacterial cell wall or probably as a result of inactivation of HL 1 by certain bacterial enzymes [52]. Moreover, among HL 1 metal complexes, Cu(L 1 ) exhibited the most potent inhibition activity. Clearly, Co(II) and Ni(II) complexes of the same geometries and stoichiometries as Cu(II) complexes showed less potency against the same species, implying the importance of the type of metal and its radius on the biological efficiencies of complexes. Additionally, metal complexes of HL 2 displayed less biological activity than the corresponding HL 1 complexes which could be attributed to the impact of geometrical and molecular size on the antimicrobial performance. Furthermore, Fe(III) complex was

Conclusion
Seven new azo -Fe(III), Co(II), Ni(II), and Cu(II) transition metal complexes were synthesized and characterized by different spectroscopic techniques (IR, UV-Vis., mass spectra, magnetism, and ESR) and thermal analysis (TGA, DTA, and DSC). A molar ratio of 1:2 (M:L) was isolated for all complexes except the Fe (III) complex which exhibited a ratio of (1:3). Based on the data obtained from different measurements, square planar or octahedral geometries were the proposed structures for the studied complexes. ESR of Cu(II) complexes showed an anisotropic spectrum and pointed to the existence of exchange interaction between the Cu(II) centers. Additionally, all HL 1 metal complexes exhibited an extent of thermal stability up to ~ 300 ºC, implying their potential applications. DFT calculations of the bond angles confirmed a slight distortion of the square planar geometry of [Cu(L 1 ) 2 ].H 2 O that was proposed experimentally by ESR. Also, significant lipophilicity was predicted for all complexes compared to the free ligands. Molecular docking gave a good insight into the binding mode of the tested compounds with the adjacent amino acids inside the protein pocket. Among the metal complexes, Fe(III) complex showed the highest calculated scoring energy against the target proteins 1bqb, 3t88, and 4esw. The in vitro screening revealed moderate antimicrobial activity for some of the complexes, which assorted them as possible candidates for this purpose.

Conflict of interest
The authors declare no competing financial interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.