Acetic anhydride 99.8%, acetic acid glacial 99.5%, aniline 99%, N, N- dimethyl formamide 99%, phosphorus oxychloride 98%, methanol 99.5%, n-Hexane 99%, dichloromethane 98%, Ethyl acetate 99.5%, chloroform 99%, Triethylamine 99%, copper nitrate trihydrate 98%, nickel nitrate hexahydrate 98%, zinc chloride 98%, silver nitrate 99.9%, L-Ascorbic acid 99%, Dimethyl sulphoxide 99% and 2, 2-diphenyl-1-picrylhydrazyl (DPPH) were used. All the chemicals and reagents, analytical grade, were purchased from Loba chemie PVT. LtD (Mumbai, India).

The NMR spectra of the ligand were obtained using NMR Bruker Avance 400 spectrometer operating at 400 MHz using DMSO-d6 and CDCl₃. Chemical shifts (δ) are reported in ppm and the coupling constants (J) are reported in Hz. Fluorescence and UV-visible spectral data were measured using Agilent MY-18490002/PC spectrofluorophotometer and SM-1600 Spectrophotometer, respectively. The absorption spectra of the synthesized compounds were measured using 1.0 × 10^–5 M methanolic diluted solution. Elemental composition and morphology were carried out using scanning electron microscopy with energy dispersive X-ray (SEM-EDX, CARL ZE 155, OXFORD instrument’s EDX, USA Hitachi SU 70 Oxford Instruments 50 mm² X-Max silicon drift EDS detector, with resolution of 127 eV FWHM and detection limit of about 1 atomic % from depth of 0.3–3 μm). Mass spectra were recorded with SHIMADZU, LC-MS 8030 (model LCMS-8030, mass range m/z 10 to 2000, sensitivity resolution R < 0.7 FWHM). X-ray diffractometer (SHIMADZU model: XRD-7000 X-RAY DIFFRACTOMETER) was used to measure the X-ray diffraction. The patterns of the synthesized metal complexes were performed using powder sample with measurement conditions of X-ray tube target: Cu (λ = 1.5406 Å), voltage: 40.0 kV, current: 30.0 mA, divergence slit: 1.0^o, scatter slit: 1.0^o, receiving slit: 0.3 mm, scanning drive axis: 2θ, scan range: 5.0–80.0^o, scan mode: continuous scan, Scan speed: 3.0 ^o/min, sampling pitch: 0.02^o.

The Chekcell Graphical Powder Indexing (CCP14) program was used to calculate the miller indices, whereas the lattice parameters were determined using the CRYSFIR computer program. FTIR measurements were performed using Perkin-Elmer BX spectrometer (from 4000–400 cm⁻¹ and KBr pellets). Thermogravimetric analyses (TGA) data were recorded using DTG-60H SHIMADZU thermal analyzer under nitrogen-atmosphere (20 ml/min) and heating rate of 10^°C/min between 25–800°C. Molar conductance of the complexes were recorded at room temperature in 1 × 10^–3 M methanolic solution of the samples using conductometre (AD8000: resolution 0.1 mV (±699.9 mV)/1 mV (±2000 mV), 0.01, 0.1, 1 μS/cm; ppm, 0.01, 0.1 mS/cm; ppt 0.1°C, accuracy at 25°C ± 0.2 mV up to ±699.9 mV, ± 1 mV up to ±2000 mV), ±0.5°C). Melting point analysis was performed using digital auto melting point apparatus (Hanchen, model 934).

The ligand was prepared based on our previously reported procedure (Digafie et al., 2021) with minor modifications, in which 2-chloroquinoline-3-carbaldehyde (2.5 g, 0.013 mol) was added to 15 ml ethanol amine and refluxed at temperature ranges of 90–95°C for 2 h. The progress and completion of the reaction was monitored using TLC. After completion, the resulting mixture was cooled and then put into crushed ice. The precipitate was collected through suction filtration and washed with ice cold water to remove the excess amount of ethanolamine which served both as solvent and reagent as well as unreacted material, and finally dried at room temperature (Digafie et al., 2021). The ligand has a molecular formula of C₁₄H₁₇N₃O₂, with a yield 86% and yellow powder and melting point of 80–85°C. It has UV-Visible λ_max (methanol) of 383 nm, IR [ʋ cm⁻¹, KBr (pellet)]: 3368 ν(O-H), 3275 ν(N-H), 1639 ν(imine C=N). Composition: Calc. for C₁₄H₁₇N₃O₂; C 64.85; H 6.61; N 16.20; O 12.34%. Found C 64.71; H 6.65; N 16.08 and O 12.56%. ¹H NMR (400 MHz, DMSO-d6): δ _H 3.65 (8H, d, H-11, H-12, H-14 and H-15), 4.72 (1H, s, OH), 4.92 (1H, s, OH), 7.19 (1H, t, J = 7.25Hz, H-6), 7.55 (2H, m, H-5, H-8), 7.72 (1H, d, J = 8.36 Hz, H-7), 8.21 (1H, s, H-4), 8.5(1H, s, H-9), and 9.55 (1H, s, NH); ¹³C NMR (400 MHz, DMSO-d6): δ _C 43.4 (C-14), 60.5 (C-12), 61.2 (C-15), 63.7 (C-11), 117.2 (C-3), 121.9 (C-8), 122.4 (C-4a), 125.7 (C-5), 128.9 (C-6), 131.5 (C-7), 143.0 (C-4), 148.3 (C-8a), 155.4 (C-2), and 163.8 (C-9); DEPT-135 δC 43.4 (C-14 negative), 60.5 (C-12 negative), 61.2 (C-15 negative), 63.7 negative (C-11), 121.9 (C-8), 125.7 (C-5), 128.9 (C-6), 131.5 (C-7), 143.0 (C-4) and 163.8 (C-9). Composition: Calc. for C₁₄H₁₇N₃O₂; C 64.85; H 6.61; N 16.20; O 12.34%. Found C 64.71; H 6.65; N 16.08 and O 12.56% (Supplementary Figures S1–S3, S4B).

A drop of triethylamine was added to stirred solution of the ligand (0.25 g, 0.96 mmol,) in methanol (10 ml). After 30 min of stirring, a solution (0.96 mmol) of ZnCl₂ (0.13 g), Cu(NO₃)₂·3H₂O (0.232 g), and Ni(NO₃)₂.6H₂O (0.279 g) in methanol (10 ml,) was added dropwise to this solution separately in different flasks (Damena et al., 2022a; Damena et al., 2022b). The mixture was refluxed for 3 h and 3.5 h at 80°C, respectively for the Zn(II), Cu(II) and Ni(II) complexes. Progress and completion of the reaction was monitored with thin layer chromatography. After completion, the reaction mixture was cool down at room temperature and the precipitated product was filtered off, washed with ice cold methanol and dried at room temperature based on reported procedures (Zou et al., 2017; Ramachandran et al., 2018; Naz et al., 2020; Halevas et al., 2021; Sumalatha et al., 2021; Damena et al., 2022a). Finally, light yellow, deep green and reddish brown powder product was obtained respectively (Supplementary Figure S5). The proposed reaction mechanisms are presented in Scheme 1.

Zn(II) chloride, Cu(II) nitrate trihydrate, and Ni(II) nitrate hexahydrate standard solutions were pipetted into ten 50 ml volumetric flasks (0, 1, 2,... 10 ml), and aliquots of a similar standard solution of the precursor ligand were added (10, 9, 8... 0 ml). All absorbance were recorded at λ_max 380, 406, and 401 nm, respectively, for complexes 1, 2 and 3, at temperatures of 25, 30, 37 and 40°C. A drop of triethylamine was used to keep the pH of the mixture constant. The metal ion and free ligand mole fractions were changed between 0 and 1 for stoichiometric measurement. From this, the absorbance of the solutions was plotted with these mole fractions, and “n”, the average number of bound ligand, was obtained from the plot, where X _max was calculated using Eq. 1 (Ahmad et al., 2012; Shalaby and Mohamed, 2020). n = X max 1 − X max (1)

Additionally, spectroscopic analysis was used to estimate the complexes’ formation constants (Supplementary Eq. S1). The changes in enthalpy and entropy (ΔH and ΔS) were derived from the slope and intercept of the lnK vs. 1/T (van’t Hoff) plot, respectively, in order to evaluate the thermodynamic parameters (ΔG, ΔH, and ΔS) appropriately (Ahmad et al., 2012; Damena et al., 2014; Shalaby and Mohamed, 2020). Eqs 2, 3 were combined to get the reactions’ Gibbs free energy (ΔG). Δ G = Δ H − T Δ S (2) Δ G = − R T   ln ⁡ K (3)

Antibacterial activities of the newly synthesized Zn(II), Cu(II) and Ni(II) complexes (1–3) were evaluated using disc diffusion method against two Gram-positive (Staphylococcus aureus, ATCC25923 and Streptococcus pyogenes, ATCC19615) and two Gram-negative (Escherichia coli, ATCC 25922, and Pseudomonas aeruginosa, ATCC 27853) bacteria by following previously reported media preparation methods (Damena et al., 2022a; Damena et al., 2022b). The antibacterial activities were recorded for two sample concentrations (150 and 300 μg/ml) in DMSO. Ciprofloxacin and DMSO were used as a positive and negative control, respectively. The plates were incubated at 37°C for 48 h, and then the bacterial growth data were evaluated by measuring the inhibition zones according to literature (Kargar et al., 2021; Damena et al., 2022a; Damena et al., 2022b). All experiments were performed in triplicate, and the mean of the triplicates was reported. The bacterial activities of the synthesized complexes were confirmed by calculating activity index (AI) (El-Gammal et al., 2021), Eq. 4. %   A c t i v i t y   i n d e x   ( A I ) = M e a n   i n h i b i t i o n   z o n e   o f   c o m p o u n d s M e a n   i n h i b i t i o n   z o n e   o f   s t a n d a r d × 100 (4)

The radical scavenging activity study of the ligand and its Zn(II), Cu(II) and Ni(II) complexes were determined using a DPPH assay based on the reported studies (Fetoh et al., 2018; Kumar et al., 2020; Sumalatha et al., 2021). Accordingly, various samples (5, 10, 25, 40, 55, 70, 85, 100, and 115 μg/ml) and the assay concentration (40 ppm) was prepared, in which 2 ml of the assay solution was mixed with 2 ml of each of the titled sample compounds. The control was prepared from 2 ml of the assay (DPPH) solution and 2 ml of solvent (methanol). The vigorously shaken resulting mixtures were put into dark incubator (Labfreez: TSI-200) at 37°C for 30 min and absorbance was recorded at 517 nm in triplicates. The percentage of radical scavenging was determined from average absorbance using Eq. 5 based on reported studies (Kumar et al., 2020; Halevas et al., 2021): D P P H r a d i c a l   s c a v e n g i n g   a c t i v i t y   ( % ) = [ ( A i − A S ) A i ] × 100 % (5) where A_i and A_S are the absorbance of the control and sample with control solution, respectively. Finally, the half-maximal inhibitory concentration (IC₅₀) was determined from the slope and intercept of the plot of percent radical scavenging activity vs. concentration.

The bacterial activities evaluation data with triplicate measurements were determined as mean ± standard deviation, in which GraphPad Prism version 5.00 was used for the analysis (GraphPad Software, California, United States) (Mogana et al., 2020). Groups were analyzed for significant differences using analysis of variance (ANOVA) test for correlation with significance (p < 0.05) (Supplementary Table S1).

The molar conductance of complexes 1 – 3 were found to be 5.21, 18.57, and 15.07 Ω⁻¹mol⁻¹cm², respectively (Supplementary Table S2), indicating non-electrolytic nature of the complexes (Chandra and Kumar, 2005; Ismael et al., 2020). This is mainly due to the fact that the metal cations receive electrons from the ligand to make a net charge balance of zero. The molar conductance results can also provide information about the proposed structures of the complexes. Hence, the chloride test of Zn(II) complex was performed using a drop of silver nitrate solution. The absence of white precipitate confirmed that chloride ion was coordinated to central metal ion in the inner sphere of the Zn(II) complex, hence the proposed formula of this complex is represented with [Zn(H₂L)(Cl)] and this in line with reported studies (Ismael et al., 2020; Damena et al., 2022b). Although it was discovered that most metal complexes exhibited a non-electrolytic nature, copper (II) and nickel (II) complexes have substantially higher molar conductance than zinc (II) complexes. This could be as a result of the Zn(II) complex’s low solubility, which impacts its ion mobility and molar conductance. A previous study suggested that the low molar conductivity of the Zn(II) complex may potentially be due to the absence of anions beyond the coordination sphere (Condé et al., 2022). Furthermore, this analysis was found to be in line with the DFT optimized geometries (vide infra).

The Job’s plot for mole fraction of the ligand and the metal ions are presented in Supplementary Figure S6. It was found that the maximum point was recorded at a mole fraction (X) of 0.5, evidencing the synthesized complexes have 1:1 [M:H₃L] ratio (Supplementary Table S3). Similar results were obtained at temperature elevations up to 40°C (Ahmad et al., 2012; Shalaby and Mohamed, 2020). The thermodynamic parameters were determined from the plot of lnK vs. 1/T (Supplementary Figure S7). The negative values obtained for the change in Gibbs free energy and enthalpy showed that the complexes are thermally stable up to 40°C (Table 1). The metal-ligand interactions showed spontaneity and exothermic nature. This is due to larger negative values of Gibbs free energy (ΔG) and lower negative values of enthalpy change (ΔH) of the chemical reactions. In other cases, the complex formations are entropically favored due to positive values of change in entropy (ΔS) (Ahmad et al., 2012; Shalaby and Mohamed, 2020; Damena et al., 2022b). Overall, the formation constants of the complexes remain constant with increase in temperature, inferring that the complexes are stable up to 40°C (Supplementary Table S3). This is in line with the thermal analysis study of all the three complexes in which no mass loss was observed up to 100°C (vide infra).

After the successful synthesis of the complexes, the presence and disappearance of characteristic functional groups in the targeted compounds were identified from the FTIR spectral data (Supplementary Figures S8, S9). Accordingly, the spectra showed strong stretching band at 1639 cm⁻¹ which was assigned for imine ν(C=N) group (Im) of the ligand (Supplementary Figure S8A), but in the case of all complexes 1–3, this spectral band shifted towards higher frequency range of 1647–1688 cm⁻¹ (Supplementary Table S4, Supplementary Figures S8B, S9C,D). This confirms the involvement of the donor nitrogen atom of the imine group ν(C=N) coordinated with the metal ions (Senthil et al., 2012; El-Sonbati et al., 2019). Similar to this, the loss of the free ligand’s ν(O-H) stretching frequency at 3368 cm⁻¹ provides proof that the hydroxyl group’s oxygen atom participated in the creation of metal-oxygen bonds (Ali et al., 2019). In addition, the characteristic ν(N-H) stretching frequency of the ligand at 3275 cm⁻¹ has shifted to 3294, 3168, and 3197 cm⁻¹ for Zn(II), Cu(II) and Ni(II) complexes, respectively (Supplementary Table S4). This also indicates the participation of the amine group in the formation of a dative bond during the complex formation process (Ali et al., 2019; Indira et al., 2019).

The IR spectral data of Ni(II) and Cu(II) complexes have weak and broad spectral bands of stretching frequencies in the range 3673–3373 and 3664–3334 cm⁻¹, respectively (Supplementary Table S4). This could be attributed to the vibration of the water molecules that might be present as lattice water and coordination water, respectively. These are newly emerged medium stretching vibration bands at 524 cm⁻¹ ν(Zn-O) and 460 cm⁻¹ ν(Zn-N) for the Zn(II) complex; 626 cm⁻¹ ν(Cu-O) and 474 cm⁻¹ ν(Cu-N) for the Cu(II) complex; and 534 cm⁻¹ ν(Ni-O), 462 cm⁻¹ ν(Ni-N) for the Ni(II) complex. The DFT calculated frequencies are in close agreement with the corresponding experimental results (Supplementary Table S4). The visualization of the DFT calculated frequencies confirmed that all the three metal ions were involved in binding to the ONN donor atoms of the ligand during complex formation (Abd El-Halim et al., 2017; Ali et al., 2019; Indira et al., 2019). The presence of nitrate in both Cu(II) and Ni(II) complexes were confirmed by the strong and broad band IR peaks at 1380 and 1351 cm⁻¹, respectively, in agreement with previously reported studies for related complexes (Ali et al., 2019; Condé et al., 2022). Weak O-H bending vibration was observed in the range of 1437–1460 cm⁻¹ in all the three complexes, confirming the appearance of free bending hydroxyl (O-H) groups. Overall, the close agreement between the DFT calculated IR frequencies and the corresponding experimental results (Supplementary Table S4) further confirmed the analysis.

The electronic spectra of the free ligand and the associated Zn(II), Cu(II), and Ni(II) complexes revealed that the ligand absorption bands are primarily caused by π→π* and n→π* transitions in UV-Visible spectral data between 200 and 800 nm. The peak for Cu(II) and Ni(II) complexes are red shifted, whereas very small blue shift was observed for the Zn(II) complex. The free ligand showed absorption peaks at 231, 258 nm (π→π*) and 300, 383 nm (n→π*) (Table 2). The red shifts observed for the Cu(II) and Ni(II) complexes at 406 and 401 nm, respectively, are mainly due to ligand-based transitions and ligand to metal charge transfer (LMCT) (Sathisha et al., 2008; Güveli et al., 2014; Halevas et al., 2021). This is because the ligand has lone pair electrons and both Cu(II) and Ni(II) complexes have vacant d-orbitals (Dalal M Charge Transfer Spectra; Sallam et al., 2011). The small band gap energy (3.272 eV) of the Cu(II) complex resulted LMCT from the quinoline ring π-orbitals to the metal d-orbitals. Moreover, as it can be seen from the insets of the absorption plots presented in Figure 1, there is a very weak d→d transition around 480 nm in the case of the Ni(II) complex. Due to the predominance of inter-ligand electron transfers, the Zn(II) complex underwent only minor modification. This might be as a result of Zn(II) having a d¹⁰ electron configuration, which prevents it from taking part in the d→d transition (Ekennia et al., 2015). This is clearly observed from the inset for the experimental absorption plot of the Zn(II) complex. The analysis was supported using the TD-DFT calculated absorption spectra of the complexes. The calculated spectra also showed the same trends and are in very good agreement with the corresponding experimental results (Figure 1).

The characteristics strong bands that appeared between 259–383 nm in all the three metal complexes were attributed to intra-ligand (C-N or C-O group) electronic transitions, in line with literature (Ilhan et al., 2014; Mahmoud et al., 2015; Abd El-Halim et al., 2017). In addition, the strong absorption bands of all the three metal complexes in the UV region (229–258 nm) could be assigned to the N-quinoline ring (ligand-based π→π* transitions). The absorption bands above 400 nm could be assigned to the LMCT of Cu(II) and Ni(II) complexes (Table 2) in agreement with reported studies (Ilhan et al., 2014; Mahmoud et al., 2015; Abd El-Halim et al., 2017). Overall, the analysis of the electronic spectra revealed that d→d transition was dominated by LMCT phenomenon and hence the resulted colors of the complexes (Abd El-Halim et al., 2017). Moreover, the electronic transitions were analyzed based on the frontier molecular orbital (FMO) plots presented in Figure 2. The HOMO and LUMO of the H ₃ L reside on the quinoline ring, confirming the presence of π→π* electron transition. It is also observed that the electron densities of the HOMO reside on the amine part of the molecule and the LUMO reside on its imine part. This is due to the fact that the amine and imine part of the ligand are in the same plane making it suitable for metal coordination (Hamdani et al., 2020). The HOMO and LUMO of compounds 1 and 3 are delocalized over the metal centers (Zn and Ni) and the quinoline ring, respectively, inferring the presence of electron transition from metal to the ligand orbital systems for Zn(II) and weak d→d transition for Ni(II) complex. Since the d-orbitals of Zn(II) are fully occupied, there is no d→d transition in 1 (see inset of Figure 1).

Fluorescence study was performed at room temperature. The spectra of the ligand (H ₃ L) and its complexes 1–3 showed emission bands at 526, 608, 471, and 511 nm, respectively (Table 3 and Supplementary Figure S10). Previous reports indicated that the photoluminescence properties of Zn(II) complexes are mainly due to intraligand emissions because of the presence of d¹⁰ electron configuration (Aslkhademi et al., 2019; Szemik-Hojniak et al., 2020; Halevas et al., 2021). In the case of Cu(II) and Ni(II) complexes, the emission intensity enhancement could be mainly due to LMCT (Halevas et al., 2021; Zheng et al., 2022). The complexes showed hyperchromic (intense), hypsochromic (blue) shifts, and intense fluorescent bands than the ligand (Supplementary Figure S10). This could be due to an increase in the conformational rigidity of the ligand upon metal coordination. Hence, the metal complexes could potentially be used for photochemical applications (Mandewale et al., 2019; Szemik-Hojniak et al., 2020).

The powder X-ray diffraction (PXRD) patterns of the three synthesized complexes (1–3) have polycrystalline characteristic peaks (Figures 3A–C), in line with the previously reported studies (Nagesh et al., 2015; Morgan et al., 2017; Ramachandran et al., 2018; Sumalatha et al., 2021). The average crystallite size (D) evaluated from the XRD pattern according to Debye–Scherrer equation as reported in previous studies (Morgan et al., 2017; El-Sonbati et al., 2019; Sumalatha et al., 2021). D = K λ β ⁡ cos ⁡ θ (6)

The equation uses the reference peak width at angle θ, where λ is the wavelength of X-ray radiation (1.5406 Å), K is Scherrer constant (0.9) and β is the width at half maximum of the reference diffraction peak measured in degrees (El-Sonbati et al., 2019). The average crystallite sizes of complexes 1–3 were found to be 27.86, 33.54 and 37.40 nm, respectively, in good agreement with previous reports for related complexes (Morgan et al., 2017; Ramachandran et al., 2018). This calculation was done by taking average size of the three major peaks. In addition to this the main reflections were defined based on maxima at 2θ that correspond to d values in which the inter-planar spacing (d) was calculated by using Bragg’s equation, λ n = 2dsinθ. The calculated inter-planar d-spacing together with relative intensities with respect to most intense peak have been recorded and then h²+k²+l² values were determined and based on this value, absence of forbidden numbers (7, 15, 23, 71) tell us that cubic or orthorhombic system (translational symmetry) based on the unit cell calculated values. According, the Zn(II) complex showed absence of the forbidden numbers (7, 15, 23, 71) (Supplementary Table S5), indicating that it could be belong to the face centered cubic system with a, b and c values of 7.28, 7.24, and 6.83 Å, respectively (Morgan et al., 2017). Similarly, the calculated lattice parameters (a, b and c) for the Cu(II) complex were found to be 21.80, 3.86, 8.61, respectively. It was observed that the absence of forbidden numbers (7, 15, 23, 71) (Supplementary Table S5) shows that the Cu(II) complex could belong to orthorhombic systems (Nagesh et al., 2015; Ramachandran et al., 2018). Similar calculations were done for the Ni(II) complex and the results showed that the Ni(II) complex could also belong to orthorhombic systems. Hence, the calculated unit cell parameters for Ni(II) complex (a, b and c) were found to be 13.61, 7.47 and 7.02 Å, respectively (Szemik-Hojniak et al., 2020), which is in agreement with previously reported studies for related complexes (Refat et al., 2013; Nagesh et al., 2015; Morgan et al., 2017; El-Metwaly et al., 2019; Ramesh et al., 2020; Abumelha et al., 2021; Sumalatha et al., 2021).

The compositions of Zn(II), Cu(II) and Ni(II) complexes were analyses from Energy Dispersive X-ray (EDX) analysis, in which the experimental percentage of atoms were found to be very close to the theoretical results (Refat et al., 2013; Sumalatha et al., 2021). In the EDX spectrum of [Zn(H₂L)Cl] complex, five characteristic signals corresponding to atoms C, O, N, Cl and Zn were observed which confirmed the formation of pure CHZnNOCl (Figure 4A). Similarly, the spectrum of [Cu(H₂L)(H₂O)(NO₃)] showed four signals, which correspond to C, O, N, and Cu atoms, indicating pure CHCuNO (Figure 4B). The [Ni(H₂L)(NO₃)].2H₂O complex showed four signals which correspond to C, O, N, and Ni atoms, and indicate the formation of CHNiNO (Supplementary Figure S4A). The free ligand shows three characteristic signals, which clearly confirms the formation of CHNO compound (Supplementary Figure S4B).

The scanning electron microscopy (SEM) micrographs indicated that the Zn(II) complex shows agglomerates which appear to be stick-like (Figure 4C), whereas the Cu(II) complex shows tiny needles-like agglomerates (Figure 4D), whereas the Ni(II) complex shows flower-like agglomerate structures (Supplementary Figure S4C). From both the SEM and PXRD data, all the three complexes appear to be clear polycrystalline structures in agreement with reported studies (Refat et al., 2013; El-Sonbati et al., 2019; Sumalatha et al., 2021).

The mass spectrum of complex 1 showed a parent molecular ion peak at m/z 357.05 (found 357.98) which corresponds to the formula [C₁₄H₁₆ClN₃O₂Zn] (Supplementary Figure S11A) with a molecular weight of 359.13 g/mol. This complex has additional peaks at m/z 322.07 (8.75%) (found 322.05) and 259.63 (54.75%) (found 260.14), attributed to [C₁₄H₁₆N₃O₂Zn]⁺ and [C₁₄H₁₈N₃O₂]⁺ fragments, respectively. Similarly, complex 2 has a parent molecular ion peak at m/z 400.08 (found 401.05) which corresponds to the formula [C₁₄H₁₈CuN₄O₆] with a molecular weight of 401.86 g/mol (Supplementary Figure S11B). This complex also showed other peaks at m/z 322 95 (6.70%) (found 322.06), 260.01 (54.75%) (found 260.14) and 226.13 (14.75%) (found 226.13) corresponding to [C₁₄H₁₇CuN₃O₂]⁺, [C₁₄H₁₈N₃O₂]⁺ and [C₁₄H₁₆N₃]⁺ fragments respectively. Finally, complex 3 exhibited a parent molecular ion peak at m/z 413.09 (found 414.07) attributed to [C₁₄H₂₀N₄NiO₇] formula with a molecular weight of 415.02 g/mol (Supplementary Figure S11C). As usual this complex showed additional peaks at m/z 351.09 (10.9%) (found 352.08), 315.09 (30.70%) (found 316.06) and 260.33 (29.9%) (found 260.14) corresponding to [C₁₄H₂₀N₃NiO₄]⁺, [C₁₄H₁₆N₃NiO₂]⁺ and [C₁₄H₁₈N₃O₂]⁺ fragments, respectively. It is important to note that similar observations have been made for other related Ni(II) containing complexes (Bursey, 1974; Tyagi et al., 2015; Morgan et al., 2017; Indira et al., 2019; El-Sonbati et al., 2020). Overall, the analysis of the mass spectra and elemental compositions are in good agreement.

The weight loss of the complexes was measured at temperature ranges 25–800°C. The results are presented in Figures 5A–C, Supplementary Figures S12A–C and Table 4. The TGA diagram of complex 1 showed three decomposition steps (Figure 5A and Supplementary Figure S12A). The first step of degradation was observed in a temperature range of 255–350°C (DTA_max of 325), which indicates a mass loss of 17.80% (calcd. 17.96%) corresponding to the loss of chloroethane (C₂H₅Cl) like moiety. The complex was stable up to 200°C indicating the absence of both lattice and a water ligands in the specified complex (Nagesh et al., 2015; Halevas et al., 2021; Sumalatha et al., 2021). The second decomposition observed with a weight loss of 21% (calcd. 21.19%) was attributed to the loss of C₆H₄ fragment of the quinoline ring moiety at 360–520°C (DTA_max of 515). The third step of degradation was from mass loss of 29.99% (calcd. 30.38%) which corresponds to the loss of C₅H₅N₂O moiety at a temperature range of 535–765°C (DTA_max of 655). The actual weight loss occurred from all these steps are 68.79%, which is in good agreement with the calculated result of 69.53%. Gradual degradation was observed up to 765°C and the residue corresponds to zinc oxide (ZnO), 22.52% (calcd. 22.66%), and CHN imine moiety with 7.40% (calcd. 7.53%) of the complex.

The TGA diagram of complex 2 indicates three decomposition steps (Figure 5B and Supplementary Figure S12B). The first step of degradation was due to mass loss of 41.69% (calcd. 41.69%) corresponding to the elimination of one water molecule, C₃H₆N₂O and nitrate ion moiety at temperature ranges of 100–225°C (DTA_max of 213). The second step is due to weight loss of 7.26% (calcd. 7.00%) at temperature ranges 240–498°C (DTA_max of 245), which correspond to the elimination of C₂H₄ ethane like moiety. The final step occurs at temperature ranges of 520–640°C (DTA_max of 530) due to the loss of 32.14% (calcd. 31.89%) related to quinoline ring C₉H₆N moiety. The leaving residue of the degradation is CuO, which is 18.91% (calcd. 19.80%) of the complex (Refat et al., 2013; Nagesh et al., 2015; Ali et al., 2019). The overall actual weight loss is 81.09%, which is close to the calculated 80.33%.

The thermal decomposition of complex 3 showed four degradation steps (Figure 5C and Supplementary Figures S12C). The first degradation step was due to mass loss of 8.65% (calcd. 8.67%) at temperature ranges of 100–155°C (DTA_max of 125) which correspond to the elimination of two lattice water molecules. The second weight loss of 10.23% (calcd. 10.13%) was observed at a temperature ranges of 220–255°C (DTA_max of 229), which is attributed to the loss of C₂H₄N imine. The third step showed a mass loss of 22.44% (calcd. 22.41%) at temperature ranges of 260–361°C (DTA_max of 278) due to loss of (CH₃O + NO₃) methanol and nitrate ion moiety. The final step of this complex is due to mass loss of 28.74% (calcd. 28.96%) at a temperature ranges of 382–610°C (DTA_max of 508), which corresponds to the loss of quinoline ring moiety. The leaving residues of nickel oxide (NiO), representing 29.24% (calcd. 27.22%) of the complex. The actual mass loss from all these steps is 70.06%, which is in very good agreement with the calculated result (69.91%).

Overall, the percentage content of elements obtained from both the elemental and TGA analyses are in very good agreement for all the three complexes. The general degradation pattern of the Cu(II) and Zn(II) complexes arise in three stages, while that of Ni(II) complex occurred in four stages. The thermogram of complexes 1–3 beyond 765, 640 and 610°C, respectively, showed a straight line, indicating the formation of metal oxides (Ambala and Lincoln, 2020). All the complexes were stable up to 100°C without any weight loss, in line with the spectroscopically evaluated stability constants which do not change up to 40°C (Supplementary Table S5). This is a good indication that the complexes could be potentially important for biological applications. Overall, the results are in very good agreement with the formulae proposed from the analytical data (Scheme 1).