Green-synthesis and characterization for new Schiff-base complexes; spectroscopy, conductometry, Hirshfeld properties and biological assay enhanced by in-silico study

2.1 Instrumentation

The utilized spectral and analytical techniques were schematically depicted for ease (Scheme 1S). Furthermore, the measurements conditions were either presented in the scheme or in discussion part. While, the metal content was analysed by known analytical methods (Vogel, 1989)

2.2 Materials and reagents

All materials used in this scientific work have PDH nature and extremely pure (Merck or Sigma& Aldrich). For instance, 2-amino thiazole, dietyl malonate, hydrazine hydrate, 2-acetyl pyridine, PdCl₂, UO₂(OAc)₂·4H₂O, Cu(OAc)₂·H₂O, Co(OAc)₂·4H₂O and Zn(OAc)₂·H₂O, were the chemicals used. Also, the solvents (diethyl ether, absolute ethyl alcohol, DMSO and DMF) were of spectroscopic nature and used as it is without pre-treatments.

2.3 Synthesis stage

The aimed Schiff base ligand was prepared as shown in Scheme 1. Then this ligand was utilized in facile-green strategy through ball milling technique to prepare UO₂(II), Co(II), Cu(II), Pd(II) and Zn(II) complexes. Such environmentally preferable approach excreted high yielded products (∼89%) with high purity. This feature was judged initially from homogeneity of solid colours as well as elemental analysis (Table 1).

Close

Scheme 1

The outline of the synthesis of ligand (H₂L) and its complexes.

Export to PPT

2.4 Molecular optimization

DMOL3 program in Materials Studio software was used to obtain cluster calculations onto isolated solid compounds [11]. DFT (density functional theory) was the suitable one applied to configure the prepared compounds via DNP basis sets (Frisch et al., 2009; Hehre et al., 1986; Kuca et al., 2018). The excellent exchange-correlation functional was determined based on GGA as well as PBEPBE functional (Willian et al., 2016; Gonçalves et al., 2014; de Lima et al., 2017). Also, MEP map was projected by applying B3LYP/6-31G (d) level in DFT theory. While, UO₂(II) complex was excluded from this study, due to failure to find the suitable basis set for this purpose.

2.5 Conductometric measurements

The conductometric titrations protocol among the ligand and Cu(II) ion, were conducted as exhibited in Scheme 2S over 1 mmol/L Cu(II) solution.

2.6 Biological studies

2.7 Molecular docking

This computerized usage aims to predict the biological efficiency for selected compounds in diseases-treatment and assess the compatibility with practical results. The ligand and its Zn(II) complex were chosen for this computation against four PDB crystal-structures (1cca, 2hq6, 6iv7 and 1p3j) (Goodin and McRee, 1993). 1cca was the triad of cytochrome peroxidase controls the reduction potential and coupling of the tryptophan free radical to the heme. 2hq6 was the structure of cyclophilin-like domain of the serologically defined colon cancer. 6iv7 was the crystal structure of dependent enzyme from aspergillus flavus, while 1p3j was the adenylate kinase from bacillus subtilis. An advanced software (MOE, Vs. 2015) was applied for this simulation to raise the accuracy of results (Abdellattif et al., 2018). Such accomplished through intensive orientation stages till reach the optimum conditions for docking. Firstly, each tested compound must face energy minimization process, then the atomic charges and potential energy must be adjusted. Secondly, the compound ready for final regulation for other parameters through MMFF94x force field, then the compound saved as MDB format (El-Metwaly et al., 2019). After that each protein crystal also faced orientation as, adding H-atoms, connecting receptors then fixing the potential energy over receptors. Therefore, search for receptor-sites, build the dummies over the helix and start the docking process (Althagafi et al., 2019). Each process was the mean of 30 poses that accomplished along various times. Some of these poses may be rejected immediately that clashed with protein-sites. The true poses have been adjusted by London dG-scoring function, which already improved twice times through triangle-matcher. The effective docking path was judged based on bond length, which must be ≤3.5 Å. Finally, all docking data as well as interaction patterns were exported to rank inhibition property of compounds.

3.1 Common properties

The ligand as well as its solid complexes were analysed to propose their molecular formulae (Table 1). The minimum ratio was suggested (1 M:1L) within the complexes, based on elemental analyses. In DMSO solvent, all complexes exhibited non-conducting property during molar conductivity measurements (1 mmol, Λ_m). This is strongly expected especially with using anions that having a strong tendency for covalence (OAc^- & Cl^-).

3.2 IR & ¹H NMR spectra

The new Schiff base prepared (H₂L) was spectrally investigated to assert on its chemical formula, which truly achieved. By the way, its IR spectrum exhibited bands at 1688, 1668, 1635, 1581, 649, 3270 and 3187 cm⁻¹. Such bands have been assigned sequentially to υ(C⚌O)₁, υ(C⚌O)₂, υ(C⚌N)₁, υ(C⚌N)₂, υ(C⚌N)_py, υ(NH)¹ and υ(NH)² (Narang and Singh, 1996) (Table 2, Fig. 1S). In furtherance of this, the ¹H NMR spectrum (Fig. 1) exhibited two characteristic signals at δ(ppm) = 12.33 and 10.96, which refer to NH¹ and NH², sequentially. The multiple-signals appeared at 7.19–8.55 (ppm) region, refer to aromatic-hydrogen atoms. The two singlet-signals exhibited at 3.68 and 2.29 ppm, refer to CH₂ and CH₃, respectively. Moreover, the molecular formula of the ligand was confirmed by mass-spectral analysis (Fig. 2S), which displayed molecular ion peak (m/z = 303) agrees comfortably with that proposed (303.23).

Close

Fig. 1

¹H NMR spectra of H₂L(A) and its Pd(II) complex (B).

Export to PPT

Consequently, this ligand which possess an extensive presence for donor-sites, has a good opportunity for multiple the mode of bonding with various metal-ions as follow;

i) The IR spectrum of Pd(II) complex reflects lower shift for υ(C⚌N)_py, υ(C⚌N)₁, υ(C⚌O)₁ and υ(C⚌O)₂ vibrations. This signifies their coordination through neutral poly-dentate binding type (Fekri and Zaky, 2014). Furthermore, its ¹ HNMR spectrum (Fig. 1) displayed signals for NH¹ and NH² that almost un-shifted (11.35 & 9.05 ppm), which confirm the lack of their participation in coordination. ii) The IR spectrum of UO₂(II) complex reflects lower shift for υ(C⚌O)₂ and υ(C⚌N)₁ vibrations only, which signifies neutral bi-dentate binding type (Zaky and Fekri 2015). As well as, bands attribute to υ_as(O—C—O) & υ_s(O—C—O) vibrations in mono-dentate covalence-acetate, were appeared at 1559 & 1351 cm⁻¹. iii) The IR spectrum of Co(II) complex reflects lower shift for υ(C⚌N)₁ and υ(C⚌N)_py vibrations beside the complete obscure for υ(C⚌O)_2. This denotes their coordination through mono-negative tri-dentate binding type. The disappeared vibration of υ(C⚌O)₂ was immediately attached by presence of υ(C⚌N*) & υ(C—O)₂ bands. Otherwise, the presence of υ_as(O—C—O) & υ_s(O—C—O) vibrations at 1525 & 1420 cm⁻¹ (Δυ = 105 cm⁻¹), signifies the contribution of bi-dentate acetate group. iv) The IR spectra for Cu(II) and Zn(II) complexes appeared fairly analogous in lower-shifted presence of υ(C⚌N)₁ and υ(C⚌N)₂ vibrations. As well as, the similarity extended to the obscure of υ(C⚌O)_a and υ(C⚌O)₂ vibrations with simultaneous appearance of υ(C⚌N*) & υ(C—O) bands. This denotes bi-negative poly-dentate binding type within the two complexes. Finally, the water molecules were proposed based on ρ_r(H₂O) and ρ_w(H₂O) bands (Rakha et al., 1996), which supported thermally (part 3.5). In addition to, the new bands assign for υ(M—N) and υ(M—O) vibrations were appeared at 400–600 cm^-1region.

3.3 Magnetic moments & electronic spectra

UV–Vis spectral analysis was executed for most complexes (Fig. 3S), to establish the structural forms based on electronic transitions displayed (Table 3). The octahedral geometry was the only form proposed for all complexes. Also, the ligand field parameters within Co(II) complex were calculated as follow; 10Dq (8790 cm⁻¹), B (925 cm⁻¹) and β(0.95) using known equations (Zaky, 2011). The values fall in the normal range for octahedral configuration which possess highly ionic nature of Co-L bonds.

3.4 SEM, EDX and XRD analysis

SEM analysis was carried out to evaluate the surface morphology for Zn(II), Pd(II) and UO₂(II) complexes (Lever, 1968), as representative examples. Such investigated complexes had asymmetric broken ice rock shape morphology. Where the particles of about micrometer size are too randomly distributed over these ice rock shape structures. Also, the chemical composition of the complexes was defined by energy-dispersive X-ray spectrometer (EDX) analysis as shown in Fig. 2. Also, Cu(II) and Co(II) complexes were examined at RT through X-ray diffraction by applying (Cu, Kα) radiation. The obtained diffraction patterns as shown in Fig. 3 were recorded in the range of 10° < 2θ < 80° (Khan et al., 2013). The patterns had sharp peaks that refer to the formation of well-defined distorted crystalline structure. Some crystal parameters were estimated by FWHM method on characteristic peak and applying known equations. Such parameters were the particulate sizes (0.591 &1.261 nm), the diffraction angles (21.96° & 23.32°), the d-spacing (2.059 &1.945 Å) and FWHM (0.2641 &0.1247) for Cu (II) and Co (II) complexes, respectively.

Close

Fig. 2

SEM and EDX views of (A) [Zn(L)(H₂O)₂]·2H₂O, (B) [Pd(H₂L)Cl₂] and (C) [UO₂(H₂L)(OAc)₂]·2H₂O complexes.

Export to PPT

Close

Fig. 3

X ray diffract grams patterns of (A) [Co(HL)(OAc)(H₂O)]·H₂O and (B) [Cu(L)(H₂O)₂]·2H₂O.

Export to PPT

3.5 Thermal analysis and kinetics

TGA curves for two chosen complexes, Co(II) and Cu(II) (Fig. 4S), were obtained over 20–800 °C range at constant heating rate (10 °C min⁻¹). The thermal instability character was dominated for the two complexes due to presence of crystal water molecules, which expelled at the first stage (≈40–120 °C). The successive three degradation stages were recorded till ≈550 °C, which sufficient to distort the complexes completely leaving the residual part of MO form (Table 1S). The kinetics parameters were calculated by two standard methods for confirmation as Coats-Redfern (Coats and Redfern, 1964) and Horowitz-Metzger (Horowitz and Metzger, 1963).

1. Coats-Redfern method

Coats-Redfern relation is as follow;

• B.

  Horowitz-Metzger Method

Horowitz-Metzger relation is as follow; $$ln[-ln(1-\alpha )]=\frac{{E_a}^{\theta }}{R{T_s}^2}(n=1)$$ $$ln\left[\frac{1-{(1-\alpha )}^{1-n}}{1-n}\right]=ln\left(\frac{\mathit{AR}{T_s}^2}{\beta }\right)-\frac{E_a}{R{T}_s}+\frac{{E_a}^{\theta }}{R{T}^2}(n\ne 1)$$ where θ = T-Ts, Ts is the DrTG peak temperature, T is the temperature consistent to the weight loss (Wt). In this method a straight line detected at the left hand sides of the last two equations versus θ with a slope of E_a/RT_s² (Yusuff and Sreekala, 1990) as shown in Fig. 6S & Table 3S.

• C.

  Thermodynamic parameters

ΔS, ΔH, A, E_a and ΔG were the parameters calculated based on activation energy values as well as pre-exponential factor (A) by using known equations (Frost and Pearson, 1961) as follow; ΔH = E-RT, ΔS = R ln h A /K_b T and ΔG = ΔH − TΔS. These parameters were obtained over all decomposition stages via Coats-Redfern and Horowitz- Metzger methods and revealed following information; i) successive increase in ΔH and E_a values from step to another, indicates the rigidity of the remaining moiety after decomposing water molecules in endothermic process. ii) The suitable order applied was n = 1. iii) Non-spontaneous characteristic was prevailed on most decomposition steps (−ve ΔS). iii) ΔG values increased from step to another in agreement with increasing the rigidity for the remaining part in sequenced steps.

3.6 Hirshfeld surface properties

Molecular packing within crystals can be imagined and studied by using crystal explorer 3.1 software (Turner et al., 2011) over exported files after orientation by VESTA package (Momma and Izumi, 2011). The 3D-crystal structure of all compounds were established over program-screen after energy minimization, through d _norm (A) and curvedness (B) styles (Fig. 4 &7S). Moreover, the cell-unit was built for the ligand through VESTA package, which appeared as a face-cantered cubic-cell unit. The degree of contact between molecules can be estimated from crystal model of normalized contact distance (d _norm). This is indicated by three colours, red, blue and white, which refer to strong, weak and moderate contact, respectively. This assumption is based on van der Waals radii for short (strong < 0) or long (weak) contact distance (Seth et al., 2011). Thus, the best molecular packing may be expected in Co(II) and Pd(II) crystals by contact lengths of −0.557 and −0.539 Å, respectively(Table 4S). Furthermore, the 3D-curvedness style for such complexes, which considered the function of square root curvature that appeared high, signify the ease of splitting surface into contact patches (Madhankumar et al., 2019). This assumption may be asserted from asphericity and surface area of crystals (Table 4). Moreover, 2D-fingerprint plots (Fig. 5 & 8S) were extracted according to elemental contribution in contact surface, which show the active role of O, S, N and Cl atoms (Skovsen et al., 2010). For instance, the crystal-fingerprint of the ligand (Fig. 5) displayed the following satisfactory contributions; N—H(4.1%), O—H(7.1%) and S—H(4.8%).

Close

Fig. 4

Hirshfeld surface maps as d _norm (A) , curvedness (B) and cubic-crystal (C) of the Schiff base ligand.

Export to PPT

Close

Fig. 5

2D fingerprint plots for contributing atoms in molecular packing of the ligand.

Export to PPT

3.7 Optimization process by applying DFT theory

The host-gust interaction between Schiff base and metal ions can be studied by applying DFT. The optimized structures for the ligand (H₂L) and its complexes were shown in (Fig. 9S). The HOMO and LUMO patterns were estimated (Fig. 10S) from chk files and appeared concentrated in two opposite zones in the ligand itself. Such appearance was fairly remaining in the metal ion complexes. Moreover, other important characteristics as; total energy, binding energy, dipole moment, E_LUMO & E_HOMO, were obtained (Table 5S). From them, the following information can be extracted;

1) The energy gap (Δ = E_LUMO − E_HOMO) values appeared significantly reduced in complexes. This indicates the ease of electronic transition within the complexes under metal ion impact. 2) The lower binding energy of all complexes indicates their stability in comparing to the free ligand. 3) Increasing dipole moment values over all complexes due to to the presence of metal atoms that coordinately bind with ligand sites. 4) Increasing in lengths of some bonds within the complexes is legally due to coordination with metal atoms (Fig. 11S).

3.8 Conductometric titrations

From specific conductance quantities (K_s), the molar conductance (/\_m) values were estimated (El-Shishtawi et al., 2010) by using the following equation; /\_m = [(K_s − K_solv) × K_cell × 1000]/C. Where K_s is the specific conductance of solution, K_solv is the specific conductance of solvent, K_cell is the cell-constant (=1), C is the molar concentration of Cu(II) solution and /\_m is the various Cu(II) concentrations in the lack of ligand. Therefore, (/\_m) was plotted vs. ( $\sqrt{\mathrm{C}}$ ) (Fig. 6) and displayed a reverse relation in between. By the way, the plots were established at various Cu(II) concentrations in absence of the ligand (A) or in its existence (B). Moreover, /\_o (limiting molar conductance) was determined for Cu(II) solution in the absence of the ligand at infinite dilution from the relation of (/\_m) and ( $\sqrt{\mathrm{C}}$ ) (Fig. 6A). Sequentially, the relation between /\_m and M/L (Fig. 7) was established, then the molar ratio between Cu(II) and the ligand (L) was obtained. As well as, the association constants of Cu(II) in the presence of ligand could be calculated by this equation (Hamada et al. 2009; Christensen et al., 1971) K_A = Λ₀² (Λ₀ − Λ_m)/4C_m^{2 2}γ_± Λ_m³S_(Z). Where, C_m is the metal concentration, S(Z) is the Fuoss-Shedlovsky factor of strong electrolytes (=1) (Covington and Dickinson, 1973) and γ_± is the activity co-efficient. The Gibbs free energy values of association (ΔG_A at 294.15 K) (El-Dossoki, 2008; Tekeda, 1983) have been calculated through association constant values by this equation; Δ G_A- =RT ln K_A (R the gas constant & T the absolute temperature) (Table 5). The association free energies obtained are small and spontaneously improved the presence of electrostatic attraction force in the solution. Moreover, the formation constants (K_f) for Cu(II)-H₂L complex can be calculated from this equation (Rahmi-Nasrabadi et al., 2009; Filippovic et al., 2009); K_f = Λ_m − Λ_obs/(Λ_obs − Λ_ML)[L]. Where, /\_m is the molar conductance of metal sulphate without ligand, /\_obs is the molar conductance of solution during titration and /\_ML is the molar conductance of the complex ion. Also, the Gibbs free energy values of formation for Cu(II) complex were calculated by this equation; ΔG_f = −RT ln K_f (Table 6)_. Fig. 7 displayed the two straight lines intersected at a point for 2 M:1L molar ratio of the complex formed. While, the Gibbs free energy values for complex formation were also estimated and displayed (Table 6).

Close

Fig. 6

The relation between molar conductance and ( $\sqrt{\mathrm{C}}$ ) of (A) CuSO₄ alone (B) in the presence of H₂L in ethanol at 294.15 K.

Export to PPT

Close

Fig. 7

The relation between molar conductance and the molar ratio (M/L) of (A) CuSO₄ in the presence of H₂L in ethanol at 294.15 K.

Export to PPT

3.9 Biological activity

The Schiff bases biological activity stimulated us to assume systematic studies on their complexation affinity and test their abilities compared to economically vital antifungal as well as antibacterial drugs (Thimmaiah et al., 1984; Nagar, 1990).

3.9.3

3.9.3 Cytotoxic study towards HCT-116 cell line

All compounds were tested against colorectal carcinoma cells line, the outcomes reveal the favorable influence of H₂L and its Zn(II) complex (Table 10, Fig. 8). This fact depends on the IC₅₀ values, which significantly reduced for the ligand and its complex (Yousef et al., 2011).

Table 10 Cytotoxicity (IC50) of tested compounds on HCT-116 cell line.

Compounds	In vitro Cytotoxicity IC50 (µg/ml)
	HCT-116
5-FU	5.2 ± 0.31
(1)	9.5 ± 0.78
(2)	40.5 ± 3.11
(3)	50.1 ± 4.65
(4)	58.7 ± 4.12
(5)	94.2 ± 5.76
(6)	15.8 ± 1.23

Close

Fig. 8

Dose-dependent effects of H₂L and it's Cu(II), Co(II), Zn(II), Pd(II), UO₂(II) complexes on cell viability of HCT-116 cell line.

Export to PPT

3.10 Drug-likeness study

Depending on molecular structure, we explored in-silico study through drug-likeness ability of some compounds to predict structural-activity relationship by Swiss ADME as a website. Such tool is easy to use for evaluating many of physical and pharmacological parameters (Shaaban et al., 2018). These parameters were displayed in Table 11 and computationally graphed (Fig. 13S). These parameters as, TPSA (topological polar surface area), which indicates the summation for the contribution of polar fragments in the compound surface. Also, Log P_o/w is the partition coefficient that significant in drug discovery and has a revers relation with the lipophilicity. In addition to, gastrointestinal absorption (GI), blood-brain barrier (BBB) and skin permeation (Log K_p) were estimated. Moreover, the synthetic accessibility score (SA), which appears in the range of 1–10 from ease to difficult synthesis, was also obtained. Regarding almost all compounds the data revealed the reduced polarity, significant lipophilic feature as well as the ease of their synthesis.

3.11 Docking simulation

Such in-silico study was implemented to confirm all above data that has been resulted from in to vitro in study. MOE-docking software is considered the most advanced tool provides strict confirmation for biological results. As well as this in-silico way adds a considerable view on the expected biological feature, which may be appeared also in-vitro study. Two compounds (H₂L and Zn(II) complex) were chosen to study their behaviour within pathogen-cells, to assess the extent of inhibition for cell-growth. The targeted cell-proteins have been chosen that belong free radicals (1cca), colorectal cancer (2hq6), aspergillus flavus (6iv7) and bacillus subtilis kinase (1p3j). This selection is based on inhibition effectiveness that recorded in-vitro screening (Goodin and McRee, 1993). All backup interaction information (Table 12) and supporting patterns (Fig. 9 & 14S-16S) were extracted from real paths for examination and comparison (Abdellattif et al., 2018). These significant notices were collected;

1. The binding sites were N10, 18, 24 and O1, 4, 13, 16

2. The responsible-coordinating receptors were Glutamate, Aspartate, Arginine, Lysine and Histidine

3. The interaction types were H-donor, H-acceptor and ionic in un-tricky paths (bond length ≤ 3.5 Å)

4. The values of scoring energy as well as the internal energies offer the following inhibitory order H₂L-1p3j ˃ H₂L-1cca ˃ Zn(II) complex-1cca ˃ Zn(II) complex-6iv7 ˃ H₂L-6iv7 ˃ Zn(II) complex-1p3j. This behaviour is distinctively agrees with the presented practical results.

Close

Fig. 9

Interaction validity (A) and surface maps (B)of H₂L and Zn(II) complex against 1cca protein.

Export to PPT

For sureness, the docking patterns (Fig. 9 & 14S-16S) were obtained for investigation, then the following notices were aggregated regarding 1cca (for example);

1. Side-chain acceptor type towards polar amino acids was appeared with H₂L, while side-chain donor type towards acidic receptors was appeared with Zn(II) complex.

2. Reduced ligand exposed surface was observed in comparing to that broadly appeared with receptors.

3. A deliberate proximity contour was observed around the two docking-complexes (El-Metwaly et al., 2019).