Structural and Spectroscopic Study of New Copper(II) and Zinc(II) Complexes of Coumarin Oxyacetate Ligands and Determination of Their Antimicrobial Activity

Tackling antimicrobial resistance is of increasing concern in a post-pandemic world where overuse of antibiotics has increased the threat of another pandemic caused by antimicrobial-resistant pathogens. Derivatives of coumarins, a naturally occurring bioactive compound, and its metal complexes have proven therapeutic potential as antimicrobial agents and in this study a series of copper(II) and zinc(II) complexes of coumarin oxyacetate ligands were synthesised and characterised by spectroscopic techniques (IR, 1H, 13C NMR, UV-Vis) and by X-ray crystallography for two of the zinc complexes. The experimental spectroscopic data were then interpreted on the basis of molecular structure modelling and subsequent spectra simulation using the density functional theory method to identify the coordination mode in solution for the metal ions in the complexes. Interestingly, the solid-state coordination environment of the zinc complexes is in good agreement with the simulated solution state, which has not been the case in our previous studies of these ligands when coordinated to silver(I). Previous studies had indicated excellent antimicrobial activity for Ag(I) analogues of these ligands and related copper and zinc complexes of coumarin-derived ligands, but in this study none of the complexes displayed antimicrobial activity against the clinically relevant methicillin-resistant Staphylococcus aureus (MRSA), Pseudomonas aeruginosa and Candida albicans.


C NMR Signals in ppm, Peak Multiplicity, Coupling Constant in Hz
Compound C 2 C 3 C 4 C 5 C 6 C 7 C 8 C 9 C 10 C 11 C 12 C 15
The complexes were characterized by decomposition temperatures, microanalysis, atomic absorption spectroscopy, thermogravimetric analysis, conductivity and magnetic moment measurements, IR, UV-Vis and 1 H, 13 C NMR spectroscopies and X-ray crystallography. The complexes obtained were prepared in good yields ranging from 60% to 72%. They showed good solubility in DMSO at ambient temperature and sparing solubility in ethanol:water 50:50 mixture at 60 • C. Analytical data for the complexes given in Table 1 were within an acceptable range of the theoretical values. The metal content for these complexes (Table 1) were estimated via atomic absorption spectroscopy and were later confirmed by thermogravimetric analysis (TGA). The obtained results showed a 1:2 ratio between metal-to-ligand along with four water molecules in each of the Zn(II) and Cu(II) complexes. The thermograms recorded for the Zn(II) and Cu(II) complexes (in this study) revealed identical thermal decomposition behaviour of the 2-[2-oxo-2H-chromen-3-yl]oxy acetic acid-based Zn(II) and Cu(II) complexes, Figure S1. Thermogravimetric analysis of the metal(II) complexes studied showed the loss of four water molecules in a single step transition at the temperature range ca. 80 to 250 • C. The significant mass loss in a single step and at relative higher temperature range indicated that possibly all of the water molecules were of the same type, i.e., water of coordination. However, Karaliota et al. [32] have suggested the loss of both type water molecules of coordination and crystallisation in a single step transition. In their complex, lattice water molecules created an extensive hydrogen-bonded network which could account well for their higher removal temperature, probably overlapping with the removal temperature of coordinated water molecules. All of the Zn(II) and Cu(II) complexes showed a significant loss in mass after 270-285 • C, representing decomposition of the coumarin-based ligand. Finally, the complexes decomposed gradually with the formation of metal-oxide above 450-500 • C. The TGA data obtained Molecules 2023, 28, 4560 9 of 23 along with microanalytical and atomic absorption data were used to deduce the empirical formulae for these complexes ( Table 1).
The molar conductivity values of Zn(II) and Cu(II) complexes in DMSO at 25 • C were in the range 1.11 to 2.97 Scm 2 mol −1 and at 37 • C (physiological temperature) these values were slightly higher: 1.33 ÷ 3.08 Scm 2 mol −1 (Table S2 in Supplementary Materials). The conductivity values at both temperatures were too low to account for any dissociation of the metal complexes in DMSO, hence these complexes were considered nonelectrolyte in nature, similar to other reported coumarin-and metal-based complexes [7][8][9][10][11].  (19), diffraction quality single crystals were obtained by recrystallisation from ethanol-water (50:50) solution mixture. Despite repeat attempts to isolate crystals of the copper complexes 21-30, crystals of suitable quality for XRD could not be obtained. It is notable that no other coumarin metal complexes were found in the March 2022 version of the Cambridge Structural Database [33], which suggests that they are challenging to crystallize.
Crystal data and structure refinement details of complexes 12 and 19 are given in the Supplementary Materials, Tables S3 and S4, respectively, and selected bond lengths and bond angles are presented in Table 2 and hydrogen bond parameters in Table S5. The X-ray crystal structures of 12 and 19 together with the atom numbering scheme and the crystal packing structure diagrams are shown in Figures 1 and 2, respectively. The crystal structures of the complexes revealed that they have a pseudo-octahedral coordination geometry, with Zn(II) ion having six contacts within its coordination sphere. In both 12 and 19, Zn lies on inversion centres. Two of the contacts around Zn(II) ion were constituted by two deprotonated coumarin-based ligands C-4oxy-acet (2) for complex 12 and 4CF 3 -C-7oxyacet (9) for complex 19 bonded in a monodentate manner and the remaining four contacts were with four water molecules. In the complexes 12 and 19, the ligands 2 [Zn1-O3 distance 2.0732(9) Å, C1-O3-Zn1 angles 125.61 (8) • ] and 9 [Zn1-O2 distance 2.0486(15) Å, C4-O2-Zn angles 127.95(13) • ], respectively, were coordinated to the Zn(II) ion via the deprotonated carboxylate group arranged in a trans orientation. The Zn(II)-O carboxylate bond distances for these complexes were found to be slightly longer than the Zn(II)-O carboxylate (2.0058 Å) for the reported crystal structure of acetobis(L-arginine)Zn(II) acetate trihydrate by Nolan et al. [34]. However, Chen et al. reported comparable Zn(II)-O carboxylate bond distances (2.082-2.095 Å) for one of their octahedral Zn(II) complexes derived from zwitterionic carboxylates [35].  The 1:2, metal-to-ligand ratio and the presence of four coordinated water molecules in complexes 12 and 19 were initially observed by atomic absorption spectroscopy, elemental and thermogravimetric analyses prior to recrystallisation. However, complex 19 showed two additional water molecules of crystallisation which were not observed in the powdered samples ( Figure 2). The four water molecules were coordinated to the metal ion via the oxygen atoms. The Zn(II)-Owater bond distances (2.082-2.115 Å) were comparable to those reported for other octahedral Zn(II) complexes [35]. The axial water molecules in complex 12 [O1-H11W···O4, distance 1.85 (2)   The 1:2, metal-to-ligand ratio and the presence of four coordinated water molecules in complexes 12 and 19 were initially observed by atomic absorption spectroscopy, elemental and thermogravimetric analyses prior to recrystallisation. However, complex 19 showed two additional water molecules of crystallisation which were not observed in the powdered samples ( Figure 2). The four water molecules were coordinated to the metal ion via the oxygen atoms. The Zn(II)-O water bond distances (2.082-2.115 Å) were comparable to those reported for other octahedral Zn(II) complexes [35]. The axial water molecules in complex 12 [O1-H11W···O4, distance 1.85 (2) 12 and 19, respectively. These angles originated between two different planes constituted by Zn(II)-O carboxylate and C carboxylate -C methylene (Figures 1b  and 2b). Due to this torsional angle, the derivatised coumarin molecules are aligned along a diagonal [−1,0,1]. Coumarin molecules were observed to be planar in both of the complexes and the heterocycle C5O plane centroid to C6 phenyl plane centroid distance between intermolecular aromatic rings of complexes 12 and 19 was 3.7960(7) Å and 3.6597(10) Å, respectively. Two of the four water ligands on 19 are disordered (O2a 49%, O2b 51% relative occupancy). Selected experimental IR frequencies of the Zn(II) (11, 12 and 19) and the Cu(II) (22) complexes studied, and the corresponding ligands 1, 2 and 9, are given in Table 3. In order to analyse the observed IR spectral data, calculations of the IR spectra and the relative weights of internal coordinates to the normal modes were performed for the [Zn(C-3oxy-acet) 2 (H 2 O) 4 ] (11) and [Cu(C-4oxy-acet) 2 (H 2 O) 4 ] (22) complexes and the relative 1, 2 and 9 ligands at the molecular level. Interpretation of the observed IR spectral data of the complexes was not a trivial task since in the solid state (Figures 1 and 2), strong intra-and intermolecular hydrogen bonds are formed (due to the presence of the electrondonor carboxylic and carbonylic groups and electron acceptor water hydrogens) and they specifically perturb the bond lengths of the compounds studied. The available Xray structures of [Zn(C-4oxy-acet) 2 (Table 3). An inspection of calculated structural data for 12 in gas phase and in solid state showed that the solid state-calculated bond lengths are in better agreement with the experimental ones (Table S6 in Supplementary Materials). The experimental IR spectra (2000-400 cm −1 ) of C-4oxy-acetH (2) and [Zn(C-4oxy-acet) 2 (H 2 O) 4 ] (12) compared to the calculated one of the Zn(II) complex are shown in Figure S2 in Supplementary Materials). Similar IR spectra were expected for the crystalline and powder Zn(II) complexes and this supposition was confirmed by the IR spectra of both samples of [Zn(4CF 3 -C-7oxy-acet) 2 (H 2 O) 4 ] (19) complex ( Figure S3). The observed similar behaviour in the IR spectra of the Cu(II) (21)(22)(23)(24)(25)(26)(27)(28)(29)(30) and the Zn(II) (11)(12)(13)(14)(15)(16)(17)(18)(19)(20) complexes indicated that the coordination mode in the Cu(II) complexes is identical to that already found for the Zn(II) complexes of the same series of ligands.
The IR spectra of the free ligands exhibited a medium intensity band at ca. 3434 to 3420 cm −1 , which is attributed to the ν(OH) carboxylate vibration. In the spectra of the metal complexes, this band should disappear since the active form of the coordinating ligand is the deprotonated one. However, broad and medium intensity bands at ca. 3500 to 3400 cm −1 were observed (Table 3) and they were assigned to the ν(O-H) stretching vibrations of water molecules in the inner and outer coordination sphere of the complexes as it was indicated by the X-ray structures of 12 and 19 and by the thermogravimetric analysis. The coordination of water molecules was manifested also by the presence of a new band at~816 cm −1 in the IR spectrum of the Zn(II) complex 12, corresponding to the rocking (OH 2 ) vibration, Figure S2. The two type stretching C=O vibrations (ν(C=O) coum and ν(C=O) COOH )) of the ligands (1-10) appear as a doublet or a broad band with a very high intensity in the range 1713-1757 cm −1 and their order could be changeable depending on the involved hydrogen-bonded interactions.
According to the calculations, the lactone carbonyl stretching vibrations produce very intensive bands in the IR region 1750-1670 cm −1 for the Zn(II) and Cu(II) complexes studied. There is no clear trend of ν(C=O) frequency shift going from ligands to their Zn(II) and Cu(II) complexes because the ν(C=O) vibrational mode is perturbed in addition to various weak interactions (Table 3). For example, the vibrational band corresponding to the lactone carbonyl ν(C=O) mode of 12 appears at 1675 cm −1 and it is shifted to lower wavenumbers by~50 cm −1 as compared to that of its neutral ligand (at 1720 cm −1 ). The wavenumber shift could be explained with the elongation of C=O bond (ligand 2 1.214 Å (calc)→ complex 12 1.230 (exp), 1.245 (calc)), which is produced by the intermolecular bonding (C=O coum ···H w~1 .93 Å (exp)) as is seen by X-ray structure of 12 ( Figure 1). The IR spectrum of complex 19 showed the ν(C=O) band at 1727 cm −1 . The higher ν(C=O) frequency for 19 (1727 cm −1 ) as compared to that of 12 (1675 cm −1 ) is in agreement with the C=O bond length 1.212 Å and 1.230 Å, respectively, and this result could be explained with weaker intermolecular (C=O···H~2.40 Å) interactions in 19 than (C=O···H~1.94 Å) in 12.
The prediction and interpretation of the stretching vibrations of the carboxylic group, ν(COO) as and ν(COO) s , also appeared complicated in the metal complexes studied due to the solid-state effect and intramolecular bonding of the uncoordinated carboxylic oxygen with the aqueous hydrogens. Moreover, these bands are medium, partially overlapped and not well indicative. The solid-state calculations of Zn(II) complexes 12 (19), predicted that the ν(COO) as and ν(COO) s vibrations appear as strong band at 1603 (1567) cm −1 and medium band at 1340 (1347) cm −1 ( Table 3). The predicted ν(COO) for the Cu(II) and Zn(II) complexes of~220-300 cm −1 are consistent with the experimental data for other unidentate carboxylate metal complexes [36] and confirm the monodentate ligand binding to the Zn(II) and Cu(II) through the oxygen atom of the deprotonated carboxylate group.
Further evidence of the mononuclear structure of the Cu(II) complexes is the room temperature magnetic moments of the powdered samples (Table S7 in
DFT/TDDFT molecular modelling calculations were performed for [Zn(C-4oxy-acet) aqueous hydrogens. In the crystal structure, where solid-state packing forces dominate, the carboxylic oxygen interacts mainly with the axial aqueous hydrogen. As expected for many Cu(d 9 ) complexes, the [Cu(C-4oxy-acet) 2 (H 2 O) 4 ] complex has axial elongated octahedral geometry (pseudo-octahedral) both in gas phase and in solution. In solution, the intramolecular hydrogen bond between the carboxylic oxygen and equatorial aqueous hydrogen atom stabilizes the coumarin and equatorial water ligands in the plane (Figure 3). in solution the Zn/Cu-Ow bonds slightly decrease and the Zn/Cu-Ocoum bonds increase by 0.02 Å. The pseudo-octahedral geometry of [Zn(C-4oxy-acet)2(H2O)4] complex is kept in solution and the uncoordinated oxygen forms hydrogen bonds with axial and equatorial aqueous hydrogens. In the crystal structure, where solid-state packing forces dominate, the carboxylic oxygen interacts mainly with the axial aqueous hydrogen. As expected for many Cu(d 9 ) complexes, the [Cu(C-4oxy-acet)2(H2O)4] complex has axial elongated octahedral geometry (pseudo-octahedral) both in gas phase and in solution. In solution, the intramolecular hydrogen bond between the carboxylic oxygen and equatorial aqueous hydrogen atom stabilizes the coumarin and equatorial water ligands in the plane ( Figure  3). The electronic spectral data revealed that the observed absorption bands of the coumarin ligands (1)(2)(3)(4)(5)(6)(7)(8)(9)(10) are in the range from 340 to 240 nm and they are assigned to one electron ππ* transition (Table 4). All these bands are kept in the corresponding Zn(II) and Cu(II) complexes without visible shifts (up to 10 nm). Only an increase of the molar absorptivity values (M −1 cm −1 ) was observed for the Zn(II) and Cu(II) complexes ( Figure S4), possibly due to the polarisation in the ligands caused by the metal-ligand binding [38]. It is evident that the vertical excitation energies due to intraligand electron transitions are not affected by the metal ion in the complexes. A survey of the molecular orbitals involved in the transitions mentioned above showed an absence of the carboxylic carbon and oxygen orbitals, whose energies should be most influenced by ligand-metal interaction. In the absorption spectra of the Cu(II) complexes (21)(22)(23)(24)(25)(26)(27)(28)(29)(30), one additional broad band was observed at 785 -750 nm and it was assigned to d-d transition ( Figure S5). In virtue of the molecular orbital and vertical excitation energy calculations of Cu(C-4oxy-acet)2(H2O)4 (22), the unpaired electron occupies the dx2-y2 orbital with Ag symmetry and accordingly the open-shell orbital is dx2-y2 orbital (x and y axes are directed along the M-L bonds). This finding is in agreement with the previous suggestions for monodentate Cu(II) complexes The electronic spectral data revealed that the observed absorption bands of the coumarin ligands (1)(2)(3)(4)(5)(6)(7)(8)(9)(10) are in the range from 340 to 240 nm and they are assigned to one electron ππ* transition (Table 4). All these bands are kept in the corresponding Zn(II) and Cu(II) complexes without visible shifts (up to 10 nm). Only an increase of the molar absorptivity values (M −1 cm −1 ) was observed for the Zn(II) and Cu(II) complexes ( Figure  S4), possibly due to the polarisation in the ligands caused by the metal-ligand binding [38]. It is evident that the vertical excitation energies due to intraligand electron transitions are not affected by the metal ion in the complexes. A survey of the molecular orbitals involved in the transitions mentioned above showed an absence of the carboxylic carbon and oxygen orbitals, whose energies should be most influenced by ligand-metal interaction. In the absorption spectra of the Cu(II) complexes (21)(22)(23)(24)(25)(26)(27)(28)(29)(30), one additional broad band was observed at 785-750 nm and it was assigned to d-d transition ( Figure S5). In virtue of the molecular orbital and vertical excitation energy calculations of Cu(C-4oxy-acet) 2 (H 2 O) 4 (22), the unpaired electron occupies the d x2-y2 orbital with A g symmetry and accordingly the open-shell orbital is d x2-y2 orbital (x and y axes are directed along the M-L bonds). This finding is in agreement with the previous suggestions for monodentate Cu(II) complexes with axial elongated octahedron [39]. Four spin-allowed d-d transitions with increasing energy are possible in the visible region: d xy d x2-y2 , d yz d x2-y2 , d xz d x2-y2 or d z2 d x2-y2 for the Cu(II) complex. These transitions are forbidden by symmetry rules but obviously one of them is activated by vibro-electronic interaction and symmetry lowering.

Experimental and Theoretical 1 H, 13 C NMR Spectral Analysis of Zn(II) Complexes of 2-(2-oxo-2H-chromen-Substituted-yl)oxy Acetic Acid Ligands
Formation of Zn(II) complexes in solution was also confirmed by means of 1 H and 13 C NMR spectroscopy. The 1 H NMR and 13 C NMR spectral data of the selected Zn(II) complexes (11,12,19) compared to those of the corresponding ligands 1, 2 and 9 are presented in Tables 5 and 6, respectively. The stacked 1 H NMR and 13 C NMR spectra of [Zn(C-4oxy-acet) 2 (H 2 O) 4 ] (12) and C-4oxy-acetH (2) are shown in Figure S6 and Figure S7, respectively. To give reliable peak assignments and to confirm the suggested metal complex structure, 1 H and 13 C NMR chemical shift calculations were performed for selected [Zn(C-3oxy-acet) 2  The 1 H NMR peak at~13 ppm in the NMR spectra of the oxy acetic coumarin ligands assigned to carboxylic hydrogen suggested an elongated O-H bond due to a strong solvation of this group. The model calculations showed that only simulation of the microsolvation around the COOH group (which is free by intramolecular hydrogen bond interaction) could reproduce such a downfield 1 H NMR peak at~13 ppm. The signal corresponding to the carboxylic acid exchangeable hydrogen was absent in the 1 H NMR spectra of 2-(2oxo-2H-chromen-substituted-yl)oxy acetic acid-based Zn(II) complexes, relative to the free ligands, suggesting metal binding through the carboxylate group ( Figure S6). As expected, the 1 H NMR spectra of the complexes showed upfield chemical shifts for most of the protons relative to the free acids with the largest upfield chemical shift observed for H 11 , which is next to the coordinating carboxylate group. A similar large shielding effect on H 11 was observed for the Ag(I) complexes of the same series of ligands [7]. Hence it could be concluded that the upfield shift of H 11 supposes coordination of the ligand through the carboxylate group, but this is not indicative of the coordination mode, monodentate or bidentate bridging. The aromatic protons are only slightly influenced upon complexation, which is consistent with their relative remoteness from the coordination centre.
The carbons C 12 (of the carboxylate group) and C 11 (near to the carboxylate group) exhibited the most significant downfield shifts (~2 ppm) in the Zn(II) complexes in agreement with the monodentate carboxylic oxygen coordination to the Zn(II) predicted by the molecular modelling in solution. It should be mentioned that a similar downfield shift was observed for C 11 , but not for C 12 in the Ag(I) complexes with the same ligands, where bidentate bridging through carboxylic group was realized. Hence, we can conclude that a downfield chemical shift of C 12 is characteristic for the monodentate carboxylate binding ( Figure S7). The observed deshielding of C 11 and C 12 i.e., decrease of their electron density upon metal complexation, is due to the polarisation effect of Zn(II) cation or to ligand-Zn(II) electron transition (donor-acceptor L-M interaction). The 13 C chemical shifts for the aromatic carbons were slightly shielded or deshielded.

Antimicrobial Studies
The antibacterial activity of the ligands used in the present study was studied in our previous investigation and no significant antibacterial or antifungal activity against the strains tested was found [7]. The Ag(I) complexes of these ligands, however, showed significant activity, particularly against the clinically relevant C. albicans, MRSA and P. aeruginosa. In addition, a number of coumarin-and quinolinone-derived Cu(II) complexes studied by our groups had previously shown excellent activity against C. albicans in particular. A recent series of studies showed that Zn(II) complexes have excellent activity against a number of bacterial strains and the Cu(II) complexes of coumarin derivatives have excellent antibacterial and antifungal activity [30,31]. That work, coupled with other studies on the antimicrobial activity of Cu(II) carboxylate complexes and analogous Zn(II) complexes in some cases, was the rationale for the preparation of the series of Cu(II) and Zn(II) complexes isolated here. Unexpectedly, all of the Zn(II) and Cu(II) complexes tested in this study showed no antimicrobial or antifungal activity against the strains tested (Table S8). In our previous studies, the most active Cu(II) complexes were the ones with the best aqueous solubility. The Cu(II) complexes of coumarin derivatives isolated in the present study, although largely soluble in DMSO, also have sparing solubility in 50:50 water:ethanol mixture. The lack of activity of the Cu(II) and Zn(II) complexes studied here against all three species stands in stark contrast to the activity of their Ag(I) analogues. The role of Ag(I) complexes as antimicrobial agents has been postulated to be multifactorial, involving a number of possible mechanisms [40]. One mechanism is related to the controlled release of silver ions at the target site. The nature of the Ag(I) carrier, whether nanoparticle or ligand, can then be tuned to target the active site or pathogen. If release of the metal ion is the basis of this series of complexes activity, then the complexes isolated here may not release their metal ion due to the difference in coordination mode of the Ag(I) complexes to the Cu(II) and Zn(II) complexes studied. In the Ag(I) complexes, the oxyacetate ligands bind via a chelating carboxylate mode but in the complexes isolated here, only unidentate bonding is noted [40]. The difference in bonding mode may well have an effect on the release of the metal ion. A more recent review on the antimicrobial role of Cu(II) complexes postulates three possible mechanisms of action for antimicrobial activity including: (1) Cu(II) complexes undergoing Fenton-like reactions to generative reactive oxygen species; (2) eliciting an inflammatory response which targets pathogens; or (3) the release of free Cu(II) or Cu(I) from the complex at the target site again resulting in Fenton-like chemistry in the presence of intracellular co-factors or else in the form of copper nanoparticles targeting cell membranes of pathogenic species [41]. Many previous studies in particular have shown that copper ions are toxic to microbial species and a recent review indicates that release of Cu(II) and reduction in situ to Cu(I) can damage the membrane and infiltrate the cell, inducing an oxidative stress response involving endogenous ROS [41,42].
The antimicrobial role of zinc complexes has focused much more on Zn-nanoparticles, as reviewed by Almoudi et al., but studies on zinc complexes indicate that the role of Zn(II) lies in interference in metabolic pathways within the microbial cell [43]. We have also previously speculated that the antimicrobial activity of Cu(II) complexes may be related to their ability to act as biomimetics and all previous copper coumarin complexes investigated by us, which showed biomimetic activity, did not have octahedral structures [30,44,45]. The active Zn(II) carboxylate complexes isolated by other groups have a variety of geometries and coordination environments although many, including those previously isolated by our group, had octahedral geometry [46]. The lack of activity of the complexes isolated here may be a result of the coordination modes for the central metal ions, which mean that the complexes lack the ability to act as biomimetics or that they are simply not getting into the microbial cells. Further mechanistic studies would be needed to elucidate this and a comparison between these and active Cu(II) and Zn(II) complexes in terms of biomimetic ability, localisation in cells etc. may help identify a new therapeutic target for metal-based complexes.

Experimental Methods
Chemicals and solvents were purchased from Sigma-Aldrich Co. (Dorset, UK) and used without further purification. Melting points were recorded on a Stuart Scientific SMP-1 apparatus (up to 300 • C). Thin layer chromatographic (TLC) was carried out on Silica gel 60F 254 aluminium-backed plates using a pre-saturated elution tank. The compounds were detected under ultraviolet light and stained using iodine crystals in a closed chamber. Infrared spectra of solids (in KBr disc) were recorded on a Nicolet Avatar 320 FT-IR spectrophotometer operating at a resolution of 2 cm −1 in the region 4000-400 cm −1 . Bruker Biospin 300 and 500 MHz FT-NMR spectrometers were used to record 1 H NMR spectra (-5 to 15 ppm with a resolution of 0.18 Hz) and 13 C NMR spectra (-33 to 233 ppm with a resolution of 2.40 Hz) as solutions in d 6 -DMSO, using tetramethylsilane (TMS) as the internal reference standard. The signal assignments were made using standard techniques including DEPT, COSY and CH-Shift and where necessary HSQC, HMBC and NOESY experiments were also run to facilitate peak assignments. Microanalytical data were provided by the Microanalytical Laboratory, Department of Chemistry, University College Dublin, Belfield, Dublin 4, Ireland, using a Perkin-Elmer CHN analyser. The metal (zinc and copper) content was determined using atomic absorption spectroscopy on a Perkin Elmer 460AAS instrument. The zinc and copper standards and samples were analysed at 213.90 and 324.70 nm, respectively. All samples were analysed in triplicate. UV-Vis spectra of the complexes and their corresponding ligands were recorded on a MSC UV-T80 spectrophotometer in dimethyl sulfoxide (DMSO) solvent (260-800 nm). The molar conductivity of the complexes and their corresponding ligands was measured on an EUTOH INSTRUMENTS Cyberscan-500 digital conductometer, using 4 mM solution of compounds in DMSO at 25 • C. Solid-state magnetic susceptibility measurements for Cu(II) complexes were carried out at room temperature using a Johnson Matthey Magnetic Susceptibility Balance with [HgCo(SCN) 4 ] as a reference standard, in the Chemistry Department, Maynooth University, Co. Kildare, Ireland.
A single crystal was coated in Paratone-N heavy oil then mounted on a Hampton Research Cryoloop and placed in a cold stream of nitrogen gas (100 K) on a Bruker X8 Apex2 CCD diffractometer running the Apex2 software. Data collection consisted of, on average, 3000 frames from a combination of phi and omega scans. After unit cell determination, intensity data were integrated using SAINT then scaled with SADABS (TWINABS if the crystal was not single) [47]. OLEX2 [34] with the SHELX suites of programs were used for structure solution and refinement. Pictures were prepared using the graphics program XP [48,49]. The quality of the dataset was checked using the IUCr checkcif facility. Crystallographic data have been deposited at the Cambridge Crystallographic Data Centre; 2253752 (12) and 2253753 (19).

Computational Details
Molecular structures, vibrational (IR), NMR and absorption spectra calculations of the model Zn(II) complexes Zn(C-3oxy-acet) 2 (H 2 O) 4 (11), Zn(C-4oxy-acet) 2 (H 2 O) 4 (12) and Cu(C-4oxy-acet) 2 (H 2 O) 4 (22) were carried out with DFT/TDDFT method using Gaussian16 program code [50] and the geometrical and spectroscopic data were visualized by the ChemCraft program [51]. The X-ray structure for Zn(C-4oxy-acet) 2 (H 2 O) 4 (12) was used as an input geometry for calculations of this complex at molecular level. By analogy, a monodentate carboxylate binding of the coumarin ligand was simulated in the model complexes 11 and 12, with general formula M(L) 2 (H 2 O) 4 M = Zn(II), Cu(II). For comparison, the calculations were performed for the corresponding isolated neutral ligands C-3oxy-acetH (1) and C-4oxy-acet (2). Split valence plus polarisation basis set 6-31G(d) was applied for the main group elements. To give a better description of the wave functions in the intermolecular region and hence to improve the estimation of the Zn(II)/Cu(II)-coumarin ligand interactions, diffuse functions were added to the standard basis set (one s and one p set) for all O atoms as well as for the two acetoxy carbon, C11 and C12, atoms (see a scheme under Table S1). For Zn and Cu, 6-31++G(d) basis set was employed [52]. The combined basis set, denoted as CB1, was used in combination with the Lee, Yang, and Parr correlation functional (LYP) [53] and Becke's three parameters exchange functional (B3) [54]. The adequacy of the B3LYP method for prediction of geometry parameters and vibrational spectra of coumarin derivatives and their metal complexes was proven in our previous investigations [55,56].
The 1 H and 13 C NMR chemical shift calculations of the model compounds 1, 2, 11 and 12 were carried out in DMSO at B3LYP/6-31+G(d,p) and B3LYP/6-31++G(2d,p) levels of theory. The UV-Vis absorption spectra of the ligands and their Cu(II) and Zn(II) complexes were simulated in the region 200-400 nm by means of vertical excitation energy calculations in DMSO using TDDFT/B3LYP/CB1 method. NMR and UV-Vis spectra calculations in DMSO were carried out with the optimized geometries in the same solvent which was simulated by the Polarisable Continuum Model (PCM) [57][58][59].
For reliable interpretation of the experimental solid-state infrared spectra, harmonic vibrational wavenumbers of the simulated crystallographic structures 12 and 19 were computed. The periodic DFT calculations were performed using the Vienna Ab initio simulation package (VASP) [60]. Electron exchange correlation interactions were treated using the generalized gradient approximation as parametrized by Perdew, Burke and Ernzerhof [61]. The electron-ion interactions were described using the projector-augmentedwave (PAW) method [62]. The vibrational frequencies were calculated using density functional perturbation theory (DFPT) (linear response theory) [63][64][65].

Antimicrobial Studies
The ligands (1-10), Zn(II) and Cu(II) complexes (11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) and the commercially available drug vancomycin were screened for their antimicrobial activity. The activity of the ligands has been reported previously [7]. S. aureus (NCIMB 12702) was obtained from the National Collection of Industrial, Marine and Food. MRSA strain ATCC4300 and P. aeruginosa strain ATCC 27853 were obtained from the American Type Culture Collection. P. aeruginosa were cultured in Luria Bertani (LB) broth (Sigma) while MRSA was cultured in nutrient broth. Bacteria were stored in broth with 20% glycerol at -80 o C and passaged twice on agar plate (LB agar, TSA or nutrient agar respectively) before being used in an experiment. MIC 50 were determined by broth microdilution. Stock solutions of each compound to be tested were prepared in DMSO and serially diluted in triplicate using broth (100 µL/well) at a concentration range of 0.195-200 µM on microtitre plates, after which the final concentration of DMSO in the cell suspension was not greater than 1%. Overnight cultures (10 mL) of bacterial strains were inoculated in broth (100 mL) and incubated at 37 o C with agitation, optical density at 600 nm (OD 600 ) was measured and strains were then diluted to 1 x 10 6 CFU/mL and seeded at 100 µL/well in microtiter well plates. Bacteria were also treated with a solvent control (0.5% DMSO) in control wells. Following 24 h incubation, the absorbance was read in a microtiter plate reader. Each compound was screened at each concentration at least in triplicate and three independent experiments were carried out.
The isolated ligands of these complexes were tested previously under the same conditions and with the same strains as their metal complexes [7]. All complexes were tested for their anti-Candida activity using a broth microdilution susceptibility protocol established by the National Committee for Clinical Laboratory Standards (NCCLS), document M27-A2, with slight modifications. The M27-A2 method was altered by substituting antibiotic medium 3 (Oxoid Ltd.) for RPMI 1640 medium [66]. All of the test compounds were prepared in 2% DMSO solution as the compounds were soluble only in DMSO. The maximum concentration of the DMSO, after serial dilution, in the test wells was 0.5%. The inhibitory effect of DMSO on the growth of C. albicans was also examined and it was shown that 0.5% DMSO did not affect the growth of this particular strain (ATCC 10231) of C. albicans. The ligands, complexes and a commercially available antifungal drug Amphotericin B were tested against C. albicans (ATCC 10231). The screening was carried out according to the broth microdilution reference method [66].

Conclusions
The preparation of novel Zn(II) (11)(12)(13)(14)(15)(16)(17)(18)(19)(20) and Cu(II) (21-30) complexes of coumarinderived oxyacetate ligands (1-10) was undertaken to improve the solubility profile of coumarin-derived Cu(II) and Zn(II) complexes towards increasing the antimicrobial activity of this type of complex. The complexes were synthesized and characterized in detail by both experimental and computational methods. The new Zn(II) and Cu(II) complexes correspond to the general formula [M(L) 2 (19) were determined by X-ray crystallographic measurements, which revealed a monodentate carboxylate binding of the two coumarin ligands to the Zn(II) in the monomer complexes. The experimental and calculated solid-state IR, solution UV-Vis and NMR spectra predicted similar ligand coordination behaviour in all the studied Zn(II) and Cu(II) complexes. Unfortunately, none of the newly synthesized Zn(II) and Cu(II) complexes of coumarin oxyacetate ligands displayed the expected antimicrobial activity against the clinically relevant MRSA, Pseudomonas Aeruginosa and Candida albicans. A comparative analysis of the coordination behaviour of the ligands and the resulted coordination polyhedron in the solid state and solution was carried out to explain the unexpected lack of biological activity of the studied Zn(II) and Cu(II) complexes. DFT structure modelling in solution predicted molecular structures for the Zn(II) and Cu(II) complexes which are in very good agreement with their solid-state coordination geometries, indicating that the solid-state structure of these complexes is preserved in solution. This structural behaviour is different from that of other Zn(II), Cu(II) and Ag(I) complexes of coumarin ligands, studied by us previously, which have shown quite different ligand coordination to the metal centre in solution and in the solid state (in particular extensive contacts to Ag(I) ion), but have revealed excellent antimicrobial activity. In the complexes isolated here, octahedral coordination geometry is maintained in solution and that coordination sphere may hinder the biomimetic activity of the complexes or hinder their incorporation into the target cells. A SARS study could not be completed but a comparison of biomimetic activity of these complexes relative to previous coumarin-based Cu(II) and Zn(II) antimicrobial complexes isolated by our group may help elucidate the mechanism of action of this type of complex and its therapeutic target.