Synthesis, Structure, and Magnetic and Biological Properties of Copper(II) Complexes with 1,3,4-Thiadiazole Derivatives

New coordination compounds of copper(II) with 2,5-bis(ethylthio)-1,3,4-thiadiazole (L1) and 2,5-bis(pyridylmethylthio)-1,3,4-thiadiazole (L2) with compositions Cu(L1)2Br2, Cu(L1)(C2N3)2, Cu(L2)Cl2, and Cu(L2)Br2 were prepared. The complexes were identified and studied by CHN analysis, infrared (IR) spectroscopy, powder X-Ray diffraction (XRD), and static magnetic susceptibility. The crystal structures of Cu(II) complexes with L1 were determined. The structures of the coordination core of complexes Cu(L2)Cl2 and Cu(L2)Br2 were determined by Extended X-ray absorption fine structure (EXAFS) spectroscopy. Magnetization measurements have revealed various magnetic states in the studied complexes, ranging from an almost ideal paramagnet in Cu(L1)2Br2 to alternating-exchange antiferromagnetic chains in Cu(L1)(C2N3)2, where double dicyanamide bridges provide an unusually strong exchange interaction (J1/kB ≈ −23.5 K; J2/kB ≈ −20.2 K) between Cu(II) ions. The cytotoxic activity of copper(II) complexes with L2 was estimated on the human cell lines of breast adenocarcinoma (MCF-7) and hepatocellular carcinoma (HepG2).


Introduction
Polynitrogen-containing heterocycles are a promising class of ligands for the synthesis of 3d-block metal complexes that are strongly responsive to the changes in external conditions.Cu(II) complexes exhibit antiferro-or ferromagnetic exchange interactions between paramagnetic centres [1], while metal complexes with 3d n (n = 4-7) electron shells demonstrate spin crossover [2,3].

Synthesis and Characterization
Complexes Cu(L 1 )2Br2 (1), Cu(L 1 )(C2N3)2 (2), Cu(L 2 )Cl2 (3), and Cu(L 2 )Br2 (4) were i lated from ethanol solutions at various Cu(II) and L 1 , L 2 ratios.The L:Cu(II) ratio was lected empirically to obtain a phase with a required composition (Scheme 2).Accordi to the powder XRD data, the resulting Cu(II) compounds with both L 1 and L 2 are cryst line, but we succeeded in growing single crystals only for copper(II) complexes with The XRD-quality single crystals of complexes 1 and 2 with L 1 precipitated from the moth liquor after the separation of the polycrystalline phase.The XRD patterns obtained for t polycrystalline phases of 1 and 2 agree well with the theoretical XRD patterns calculat for the structural models of 1 and 2 (Figures S1 and S2).The IR spectra of L 1 , L 2 and those of the complexes with these ligands contain ν(C vibrations at 3070-3000 cm −1 (Table 1) and bands of ν(CH3) and ν(CH2) vibrations of t L 1 , L 2 ligands and the complexes at 3000-2850 cm −1 .The region of 1700-1450 cm −1 conta coordination-sensitive bands of stretching and bending vibrations of 1,3,4-thiadiaz rings.The positions and number of these bands in the spectra of complexes with L 1 , differ significantly from those of the ligands.This indicates that the nitrogen atoms of t heterocycle are coordinated to the copper(II) ion, which is also confirmed by the XRD d for complexes with L 1 .Both of the ligands and the complexes contain ν(S-C-S) bands the region of 1090-1030 cm −1 .

Synthesis and Characterization
Complexes Cu(L 1 )2Br2 (1), Cu(L 1 )(C2N3)2 (2), C lated from ethanol solutions at various Cu(II) and lected empirically to obtain a phase with a require to the powder XRD data, the resulting Cu(II) comp line, but we succeeded in growing single crystals The XRD-quality single crystals of complexes 1 and liquor after the separation of the polycrystalline ph polycrystalline phases of 1 and 2 agree well with t for the structural models of 1 and 2 (Figures S1 and  The IR spectra of L 1 , L 2 and those of the comp vibrations at 3070-3000 cm −1 (Table 1) and bands o L 1 , L 2 ligands and the complexes at 3000-2850 cm −1 .coordination-sensitive bands of stretching and b rings.The positions and number of these bands in differ significantly from those of the ligands.This i heterocycle are coordinated to the copper(II) ion, w for complexes with L 1 .Both of the ligands and the the region of 1090-1030 cm −1 .
The IR spectra of L 1 , L 2 and those of the complexes with these ligands contain ν(CH) vibrations at 3070-3000 cm −1 (Table 1) and bands of ν(CH 3 ) and ν(CH 2 ) vibrations of the L 1 , L 2 ligands and the complexes at 3000-2850 cm −1 .The region of 1700-1450 cm −1 contains coordination-sensitive bands of stretching and bending vibrations of 1,3,4-thiadiazole rings.The positions and number of these bands in the spectra of complexes with L 1 , L 2 differ significantly from those of the ligands.This indicates that the nitrogen atoms of the heterocycle are coordinated to the copper(II) ion, which is also confirmed by the XRD data for complexes with L 1 .Both of the ligands and the complexes contain ν(S-C-S) bands in the region of 1090-1030 cm −1 .

X-ray Structure Determination
The XRD patterns obtained for the polycrystalline phases of 1 and 2 agree well with the theoretical XRD patterns calculated for the structural models of 1 and 2 (Figures S1 and S2).According to the XRD data, the Cu(L 1 ) 2 Br 2 complex (1) is isostructural with its chloride analogue Cu(L 1 ) 2 Cl 2 that was prepared earlier [22].It crystallizes in the triclinic crystal system, space group P-1, Z = 2. Figure 1a shows the structure of complex 1.The central Cu 2+ ion coordinates nitrogen atoms of two monodentate L 1 ligands and two bromide ions of the {CuN 2 Br 2 } coordination polyhedron.The bond lengths in CuN 2 Br 2 agree well with the data from the CCDC structural database (version 5.43, November 2021) [23].The bond angles at the Cu 2+ ion vary from 91.98 • to 158.30 • , thereby indicating that the central atom occurs in a distorted square-planar environment with bromide ligands in trans positions (Figure 1b).The L 1 ligands in 1 occur in trans positions and are rotated almost perpendicular to the plane of the coordination polyhedron.In the structure of the complex, the atoms of one of the two thiolate groups are disordered over two positions S3A and S3A .

X-ray Structure Determination
The XRD patterns obtained for the polycrystalline phases of 1 and 2 agree well with the theoretical XRD patterns calculated for the structural models of 1 and 2 (Figures S1  and S2).According to the XRD data, the Cu(L 1 )2Br2 complex (1) is isostructural with its chloride analogue Cu(L 1 )2Cl2 that was prepared earlier [22].It crystallizes in the triclinic crystal system, space group Р-1, Z = 2.
Figure 1а shows the structure of complex 1.The central Cu 2+ ion coordinates nitrogen atoms of two monodentate L 1 ligands and two bromide ions of the {CuN2Br2} coordination polyhedron.The bond lengths in CuN2Br2 agree well with the data from the CCDC structural database (version 5.43, November 2021) [23].The bond angles at the Cu 2+ ion vary from 91.98° to 158.30°, thereby indicating that the central atom occurs in a distorted square-planar environment with bromide ligands in trans positions (Figure 1b).The L 1 ligands in 1 occur in trans positions and are rotated almost perpendicular to the plane of the coordination polyhedron.In the structure of the complex, the atoms of one of the two thiolate groups are disordered over two positions S3A and S3A'.Single crystals of Cu(L 2 )(C2N3)2 (2) also belong to the triclinic crystal system, space group Р-1, Z = 2, (Table S1).The structure is composed of polymeric 1D chains with repeating {Cu(L 2 )(μ-(C2N3)2} fragments (Figure 2).The Cu 2+ coordination sphere contains four N atoms of two equatorial dicyanamide ions with the Cu-N distances varying from 1.974(2) to 1.981(1) Å, and an apical N atom of the thiadiazole ring (the Cu-N distance is 2.243(2) Å).The average deviation of nitrogen atoms from the basal plane is 0.02 Å.The N atoms form a tetragonal pyramid, with the angle between the Cu1-N2 vector and the basal plane equal to 87.8°.Two neighbouring copper ions are connected by two bridging C2N3 -ions forming zigzag chains (Figure 3a,b).Each {Cu2(μ-C2N3)2} metallocycle in these polymer chains adopts a chair conformation with the Cu1 and Cu1′ atoms deviating from the plane by −0.567 Å and 0.567 Å, respectively.Single crystals of Cu(L 2 )(C 2 N 3 ) 2 (2) also belong to the triclinic crystal system, space group P-1, Z = 2, (Table S1).The structure is composed of polymeric 1D chains with repeating {Cu(L 2 )(µ-(C 2 N 3 ) 2 } fragments (Figure 2).The Cu 2+ coordination sphere contains four N atoms of two equatorial dicyanamide ions with the Cu-N distances varying from 1.974(2) to 1.981(1) Å, and an apical N atom of the thiadiazole ring (the Cu-N distance is 2.243(2) Å).The average deviation of nitrogen atoms from the basal plane is 0.02 Å.The N atoms form a tetragonal pyramid, with the angle between the Cu1-N2 vector and the basal plane equal to 87.It is worth emphasizing that the Cu…Cu distances inside each chain alternate and are equal to 6.950 Å and 7.050 Å (Figure 2).In turn, the distances between the chains are significantly larger, the shortest of them being equal to 7.840 Å (Figure 3b).There are also S…N contacts between the chains: the shortest of them is the contact between the sulphur atom of the thiadiazole ring and the dicyanamide nitrogen atom (S2…N4 = 3.178(2) Å) (shown by dashed lines in Figure 3a), which is shorter than the sum of their van der Waals radii (3.45 Å).It is worth emphasizing that the Cu…Cu distances inside each chain alternate and are equal to 6.950 Å and 7.050 Å (Figure 2).In turn, the distances between the chains are significantly larger, the shortest of them being equal to 7.840 Å (Figure 3b).There are also S…N contacts between the chains: the shortest of them is the contact between the sulphur atom of the thiadiazole ring and the dicyanamide nitrogen atom (S2…N4 = 3.178(2) Å) (shown by dashed lines in Figure 3a), which is shorter than the sum of their van der Waals radii (3.45 Å).

EXAFS Spectroscopy Determination
A structure of the coordination cores in complexes 3 and 4 was determined by EXAFS spectroscopy.The XANES and EXAFS spectra obtained for these compounds are shown in Figure 4a,b.It is worth emphasizing that the Cu. ..Cu distances inside each chain alternate and are equal to 6.950 Å and 7.050 Å (Figure 2).In turn, the distances between the chains are significantly larger, the shortest of them being equal to 7.840 Å (Figure 3b).There are also S. ..N contacts between the chains: the shortest of them is the contact between the sulphur atom of the thiadiazole ring and the dicyanamide nitrogen atom (S2. ..N4 = 3.178(2) Å) (shown by dashed lines in Figure 3a), which is shorter than the sum of their van der Waals radii (3.45 Å).

EXAFS Spectroscopy Determination
A structure of the coordination cores in complexes 3 and 4 was determined by EXAFS spectroscopy.The XANES and EXAFS spectra obtained for these compounds are shown in Figure 4a,b.Figures 5 and 6 show the k 2 -weight EXAFS spectra fitted with the EXCURVE program [24], with the Δk ranging from 3 to 11 Å −1 .The initial models of the EXAFS spectra were based on the XRD data for the mononuclear Cu(L 1 )2Br2 complex and cupric chloride complex representing a 1D coordination polymer, which was formed at the expense of the triple bridge between two neighbouring copper(II) cations (double chlorido bridge and the tetrazole ring N3,N4-bridge) [25].EXAFS fitting results are presented in Table 2 for complex 3 and in Table 3 for 4. Note that the EXAFS allow us to determine average coordination numbers, interatomic distances, and Debye-Waller factors for the scattering atom (Cu).The parameters were determined from the EXAFS data with an accuracy of ±1% for interatomic distances, ±10% for coordination numbers, and ±20% for Debye-Waller factors.Figures 5 and 6 show the k 2 -weight EXAFS spectra fitted with the EXCURVE program [24], with the ∆k ranging from 3 to 11 Å −1 .The initial models of the EXAFS spectra were based on the XRD data for the mononuclear Cu(L 1 ) 2 Br 2 complex and cupric chloride complex representing a 1D coordination polymer, which was formed at the expense of the triple bridge between two neighbouring copper(II) cations (double chlorido bridge and the tetrazole ring N3,N4-bridge) [25].EXAFS fitting results are presented in Table 2 for complex 3 and in Table 3 for 4. Note that the EXAFS allow us to determine average coordination numbers, interatomic distances, and Debye-Waller factors for the scattering atom (Cu).The parameters were determined from the EXAFS data with an accuracy of ±1% for interatomic distances, ±10% for coordination numbers, and ±20% for Debye-Waller factors.Figures 5 and 6 show the k 2 -weight EXAFS spectra fitted with the EXCURVE progr [24], with the Δk ranging from 3 to 11 Å −1 .The initial models of the EXAFS spectra w based on the XRD data for the mononuclear Cu(L 1 )2Br2 complex and cupric chloride co plex representing a 1D coordination polymer, which was formed at the expense of triple bridge between two neighbouring copper(II) cations (double chlorido bridge and tetrazole ring N3,N4-bridge) [25].EXAFS fitting results are presented in Table 2 for comp 3 and in Table 3 for 4. Note that the EXAFS allow us to determine average coordinat numbers, interatomic distances, and Debye-Waller factors for the scattering atom (Cu).T parameters were determined from the EXAFS data with an accuracy of ±1% for interatom distances, ±10% for coordination numbers, and ±20% for Debye-Waller factors.
is the index characterizing the fitting qualit is the index characterizing the fitting qualit

Magnetic Properties
For all the complexes under investigation, the dependences of the molar magne susceptibility χ on temperature and magnetic field were measured in the ranges T = 1.7 330 K and H = 0-10 kOe.The data obtained under zero-field cooled and field-cooled c ditions revealed perfect magneto-thermal reversibility in all samples, which pointed the absence of any ferromagnetic or spin-freezing phenomena and allowed a straightf ward analysis of the χ(T,H) dependences.After subtracting the diamagnetic core con bution, the remaining paramagnetic component of the magnetic susceptibility, χp(T), w analysed using the Curie-Weiss dependence χ (T) = N µ eff 3k (T − θ) ⁄ to obtain is the index characterizing the fitting quality.
is the index characterizing the fitting quality.

Magnetic Properties
For all the complexes under investigation, the dependences of the molar magnetic susceptibility χ on temperature and magnetic field were measured in the ranges T = 1.77-330K and H = 0-10 kOe.The data obtained under zero-field cooled and field-cooled conditions revealed perfect magneto-thermal reversibility in all samples, which pointed to the absence of any ferromagnetic or spin-freezing phenomena and allowed a straightforward analysis of the χ(T,H) dependences.After subtracting the diamagnetic core contribution, the remaining paramagnetic component of the magnetic susceptibility, χ p (T), was analysed using the Curie-Weiss dependence χ p (T) = N A µ 2 eff /3k B (T − θ) to obtain insights into the spin states of Cu ions and the exchange interactions between them.
The simplest magnetic behaviour, barely distinguishable from that of an ideal paramagnet, has been found for complex 1 (Figure 7).Indeed, the measured (1/χ p )(T) curve of 1 turns out to be almost perfectly linear and goes straight to the origin, tending to intersect the temperature axis at |θ| less than 0.05 K. Given the small value of θ, we can calculate the effective magnetic moment µ eff neglecting the exchange interactions between copper ions.The resulting µ eff amounts to ≈ 1.83 µ B at T = 300 K, which is close to the spin-only value of 1.73 µ B expected for Cu 2+ (S = 1/2) ions, with a slight excess being caused by the contribution from orbital moments.As can be seen in Figure 7, µ eff measured at H = 1 kOe decreases smoothly to µ eff ≈ 1.80 µ B at T = 20 K and somewhat faster at lower temperatures down to µ eff ≈ 1.78 µ B at T = 1.77K; the latter implies the presence of a tiny, barely detectable antiferromagnetic (AF) interaction between Cu 2+ ions.
Int. J. Mol.Sci.2023, 24, x FOR PEER REVIEW 7 of 20 the effective magnetic moment µeff neglecting the exchange interactions between copper ions.The resulting µeff amounts to ≈ 1.83 µB at T = 300 K, which is close to the spin-only value of 1.73 µB expected for Cu 2+ (S = 1/2) ions, with a slight excess being caused by the contribution from orbital moments.As can be seen in Figure 7, µeff measured at H = 1 kOe decreases smoothly to µeff ≈ 1.80 µB at T = 20 K and somewhat faster at lower temperatures down to µeff ≈ 1.78 µB at T = 1.77K; the la er implies the presence of a tiny, barely detectable antiferromagnetic (AF) interaction between Cu 2+ ions.A noticeably stronger AF exchange interaction is observed in complexes 3 and 4 (Figure 8a,b).Curie-Weiss fi ing of the high-temperature χp(T) data results in the θ values of ≈ -6.4 K and ≈ -5.9 K for 3 and 4, respectively.When we select for fi ing the low-temperature region, where the χp data are much less sensitive to such uncertain parameters as the Van Vleck paramagnetism of Cu 2+ ions and the temperature-dependent orbital contribution, the Weiss constant θ diminishes to -0.70 K and -0.64 K, yet still remains sizable.The evaluated effective magnetic moment of 3 reaches µeff ≈ 1.87 µB at T = 300 K (µeff ≈ 1.86 µB for θ ≈ -0.70 K), which is typical for Cu 2+ (S = 1/2) ions, and slightly exceeds the spinonly value of ≈1.73 µB (Figure 8a).Although the shape of the µeff(T) curve in 4 is remarkably similar to that of 3, the absolute µeff values turn out to be noticeably lower (Figure 8b).The la er implies an admixture of a non-magnetic impurity phase in the studied sample of 4.
The most interesting magnetic behaviour was revealed for complex 2 (Figure 9a,b).In contrast to the aforementioned samples, 2 demonstrates a non-monotonic χp(T) dependence that passes through a broad peak at Tmax ≈ 10.45 K and then recovers a steep growth upon cooling below T ≈ 4.5 K (Figure 9a).Such a broad magnetic susceptibility peak is a fingerprint of low-dimensional antiferromagnetic systems, such as dimers or 1D chains [26][27][28][29][30], where the AF exchange interactions are capable of providing local spin correlations, but the absence of 3D links prevents a phase transition to the long-range ordered AF state.In its turn, the low-temperature χp growth is often observed in such systems owing to the presence of monomers or chain fragments with an odd number of ions, which behave as isolated paramagnetic centres.The Curie-Weiss fi ing of χp(T) data in the high-temperature region of 40-300 K, i.e., far above Tmax, results in the θ value of ≈ −16 K and µeff ≈ 1.91-1.92µB (Figure 9b).A detailed analysis of the observed magnetic behaviour side-by-side with structural data will be given in Section 3. A noticeably stronger AF exchange interaction is observed in complexes 3 and 4 (Figure 8a,b).Curie-Weiss fitting of the high-temperature χ p (T) data results in the θ values of ≈−6.4 K and ≈−5.9K for 3 and 4, respectively.When we select for fitting the lowtemperature region, where the χ p data are much less sensitive to such uncertain parameters as the Van Vleck paramagnetism of Cu 2+ ions and the temperature-dependent orbital contribution, the Weiss constant θ diminishes to −0.70 K and −0.64 K, yet still remains sizable.The evaluated effective magnetic moment of 3 reaches µ eff ≈ 1.87 µ B at T = 300 K (µ eff ≈ 1.86 µ B for θ ≈ −0.70 K), which is typical for Cu 2+ (S = 1/2) ions, and slightly exceeds the spin-only value of ≈1.73 µ B (Figure 8a).Although the shape of the µ eff (T) curve in 4 is remarkably similar to that of 3, the absolute µ eff values turn out to be noticeably lower (Figure 8b).The latter implies an admixture of a non-magnetic impurity phase in the studied sample of 4.
The most interesting magnetic behaviour was revealed for complex 2 (Figure 9a,b).In contrast to the aforementioned samples, 2 demonstrates a non-monotonic χ p (T) dependence that passes through a broad peak at T max ≈ 10.45 K and then recovers a steep growth upon cooling below T ≈ 4.5 K (Figure 9a).Such a broad magnetic susceptibility peak is a fingerprint of low-dimensional antiferromagnetic systems, such as dimers or 1D chains [26][27][28][29][30], where the AF exchange interactions are capable of providing local spin correlations, but the absence of 3D links prevents a phase transition to the long-range ordered AF state.In its turn, the low-temperature χ p growth is often observed in such systems owing to the presence of monomers or chain fragments with an odd number of ions, which behave as isolated paramagnetic centres.The Curie-Weiss fitting of χ p (T) data in the high-temperature region of 40-300 K, i.e., far above T max , results in the θ value of ≈−16 K and µ eff ≈ 1.91-1.92µ B (Figure 9b).A detailed analysis of the observed magnetic behaviour side-by-side with structural data will be given in Section 3.

In Vitro Cytotoxic Activity
The effect of the novel compounds on the viability of MCF-7 and HepG2 cells was evaluated using the Hoechst 33342/PI dual staining method and subsequent differentiation of the cells into live, dead, and apoptotic ones.The half maximal inhibitory concentration IC50 was defined as the concentration of the substance at which the percentage of live cells was equal to 50%.To determine the IC50 value, experimental dependences of cell survival (%) on the substance concentration (µМ) were fi ed by nonlinear functions.

In Vitro Cytotoxic Activity
The effect of the novel compounds on the viability of MCF-7 and HepG2 cells was evaluated using the Hoechst 33342/PI dual staining method and subsequent differentiation of the cells into live, dead, and apoptotic ones.The half maximal inhibitory concentration IC50 was defined as the concentration of the substance at which the percentage of live cells was equal to 50%.To determine the IC50 value, experimental dependences of cell survival (%) on the substance concentration (µМ) were fi ed by nonlinear functions.

In Vitro Cytotoxic Activity
The effect of the novel compounds on the viability of MCF-7 and HepG2 cells was evaluated using the Hoechst 33342/PI dual staining method and subsequent differentiation of the cells into live, dead, and apoptotic ones.The half maximal inhibitory concentration IC 50 was defined as the concentration of the substance at which the percentage of live cells was equal to 50%.To determine the IC 50 value, experimental dependences of cell survival (%) on the substance concentration (µM) were fitted by nonlinear functions.
The copper(II) halide complexes Cu(L 2 )Cl 2 and Cu(L 2 )Br 2 did not affect the viability of HepG2 cells (Figure 10a,b, respectively), but had a dose-dependent cytotoxic effect on MCF-7 cells after a 48 h period of incubation (Figure 10c,d, respectively).The IC 50 values characterizing the cytotoxic activity of the complexes on the MCF-7 line are 41.7 ± 0.4 and 39.0 ± 0.4 µM, respectively.Both compounds exhibit similar cytotoxic effects.Thus, the cytotoxic activity of the copper(II) complex with L 2 does not depend on the halide ion in its composition.At the same time, it is worth emphasizing that the studied complexes selectively affect the cell lines of different origins.
Int. J. Mol.Sci.2023, 24, x FOR PEER REVIEW 9 of 20 The copper(II) halide complexes Cu(L 2 )Cl2 and Cu(L 2 )Br2 did not affect the viability of HepG2 cells (Figure 10а,b, respectively), but had a dose-dependent cytotoxic effect on MCF-7 cells after a 48 h period of incubation (Figure 10 c,d, respectively).The IC50 values characterizing the cytotoxic activity of the complexes on the MCF-7 line are 41.7 ± 0.4 and 39.0 ± 0.4 µМ, respectively.Both compounds exhibit similar cytotoxic effects.Thus, the cytotoxic activity of the copper(II) complex with L 2 does not depend on the halide ion in its composition.At the same time, it is worth emphasizing that the studied complexes selectively affect the cell lines of different origins.

Structure of Complexes 1 and 2
Literature review reveals that, similar to 1 complexes, Cu(tda)2Hal2 (Hal = Cl, Br) (tda = 2,5-bis(methylthio)-1,3,4-thiadiazole), whose ligand contains a thiomethylate group instead of a thioethylate group, are centrosymmetric, i.e., their Hal-Cu-Hal and N-Cu-N angles are strictly equal to 180° [30].Moreover, the thiadiazole ligands in the Cu(tda)2Hal2 complexes are rotated by 180° relative to the N-Cu-N line (Figure 11).In the present case of complex 1, the ligands are located opposite to each other (the corresponding angle is ~ 4°).Different geometries of molecules cause the crystal packing in the complexes to be also different; namely, Cu(tda)2Hal2 contains intermolecular Hal•••S contacts (Hal = Cl, Br), while the distances in 1 are significantly larger (the minimal distance in 1 is 3.74 Å, Figure 1b).11).In the present case of complex 1, the ligands are located opposite to each other (the corresponding angle is ~4• ).Different geometries of molecules cause the crystal packing in the complexes to be also different; namely, Cu(tda) 2 Hal 2 contains intermolecular Hal• • • S contacts (Hal = Cl, Br), while the distances in 1 are significantly larger (the minimal distance in 1 is 3.74 Å, Figure 1b).The CCDC database contains no structures of 1D polymer chain {Cu2(µ-C2N3)2} fragments where the copper(II) coordination number wo At the same time, 12 discrete binuclear copper complexes with such found.For instance, in the structure of one of these complexes, [Cu2( [31], the Cu-Cu distance within dimers is 7.134(1) Å, while the shortest dimers is 6.752(1) Å.In the [Cu2(medpt)2(C2N3)2](ClO4)2 binuclear com Cu distance is 6.975 Å within dimers and as large as 8.562 Å between the polymer Cu(L2)2(C2N3)2 studied in our earlier work [33], the distances bet connected by cyanamide bridges are 7.30 Å and 7.707 Å.
The coordination environment of Cu(II) depends on the nature of l ber of donor atoms, charge, skeletal rigidity, etc.), reaction conditions, an secondary bonding interactions forming a specific crystal packing.Thes tively affect the organization of a coordination polyhedron, sometimes ways.In a recent paper [22], we have shown that CuCl2 and L 1 can for penta-coordinated complexes upon slight variation of the reaction cond hexa-coordinated complexes [Cu(tda)(C2N3)2] and [Cu(tda)2(C2N3)2] ar similar conditions [30].To obtain an idea for which coordination numbe performed a CSD survey of crystal structures comprising a fragment {C = κ-N heterocyclic ligand with the N having three neighbours (i.e., the ized), x, y = 2-4 (Table S1).Tetra-coordinated fragment {CuL2Br2} has 268 coordinated ones have a total of 124 hits, suggesting the former fragm more common.Note that hexa-coordinated fragments have only a tota likely a consequence of high steric hindrance of heavy Br atoms.Compl amide are less common than bromide ones.The CSD survey of structu least one dicyanamide coordinated to Cu revealed that penta-coordinat hit) dominate over tetra-(10 hits) and hexa-coordinated ones (24 hits; Ta Cu polyhedra in structures 1 and 2 belong to the most common coordinat tetracoordinated for the bromide, and pentacoordinated for the dicyana 3.2.Structure of Complexes 3 and 4 According to the EXAFS Data The CCDC database contains no structures of 1D polymer chains with repeating {Cu 2 (µ-C 2 N 3 ) 2 } fragments where the copper(II) coordination number would be equal to 5. At the same time, 12 discrete binuclear copper complexes with such fragments were found.For instance, in the structure of one of these complexes, [Cu 2 (dmphen) 2 (C 2 N 3 ) 4 ] [31], the Cu-Cu distance within dimers is 7.134(1) Å, while the shortest distance between dimers is 6.752(1) Å.In the [Cu 2 (medpt) 2 (C 2 N 3 ) 2 ](ClO 4 ) 2 binuclear complex [32], the Cu-Cu distance is 6.975 Å within dimers and as large as 8.562 Å between them.In the layered polymer Cu(L 2 ) 2 (C 2 N 3 ) 2 studied in our earlier work [33], the distances between copper ions connected by cyanamide bridges are 7.30 Å and 7.707 Å.
The coordination environment of Cu(II) depends on the nature of ligands (the number of donor atoms, charge, skeletal rigidity, etc.), reaction conditions, and the presence of secondary bonding interactions forming a specific crystal packing.These factors cumulatively affect the organization of a coordination polyhedron, sometimes in unpredictable ways.In a recent paper [22], we have shown that CuCl 2 and L 1 can form either tetra-or penta-coordinated complexes upon slight variation of the reaction conditions.Penta-and hexa-coordinated complexes [Cu(tda)(C 2 N 3 ) 2 ] and [Cu(tda) 2 (C 2 N 3 ) 2 ] are also formed in similar conditions [30].To obtain an idea for which coordination number is preferred, we performed a CSD survey of crystal structures comprising a fragment {CuL x Br y }, where L = κ-N heterocyclic ligand with the N having three neighbours (i.e., the N is sp 2 -hybridized), x, y = 2-4 (Table S1).Tetra-coordinated fragment {CuL 2 Br 2 } has 268 hits, while pentacoordinated ones have a total of 124 hits, suggesting the former fragment is noticeably more common.Note that hexa-coordinated fragments have only a total of 39 hits; this is likely a consequence of high steric hindrance of heavy Br atoms.Complexes with dicyanamide are less common than bromide ones.The CSD survey of structures comprising at least one dicyanamide coordinated to Cu revealed that penta-coordinated complexes (61 hit) dominate over tetra-(10 hits) and hexa-coordinated ones (24 hits; Table S2).Thus, the Cu polyhedra in structures 1 and 2 belong to the most common coordination environment: tetracoordinated for the bromide, and pentacoordinated for the dicyanamide.

Structure of Complexes 3 and 4 According to the EXAFS Data
The experimental XANES and EXAFS spectra (Figure 4a,b) of Cu(L 2 )Cl 2 (3) and Cu(L 2 )Br 2 (4) differ significantly, thus indicating that Cu(II) ions occur in different coordination in these compounds.The simulations of experimental EXAFS spectra (Figures 5 and 6) have shown that complex 3 is well described by the trimer structure (Figure 12a), while 4 conforms to the dimer one (Figure 12b).The fi ing results of EXAFS spectra for 3 signify the formation of a trimer, as follows from the obtained coordination numbers (C.N.) and the interatomic distances (Table 2).Copper(II) ions form three different coordination sites: CuNCl3, CuN2Cl3, and CuN2Cl4 (Figure 12a).The EXAFS spectrum of the complex reveals two different average values for Cu-Cl' (2.26 Å) and Cu-Cl" (2.67 Å) distances with the corresponding average C.N. values equal to 2.6 and 0.8, respectively.The total coordination number of Cu(II) is 3.4, which is very close to the average C.N. value of Cu(II) in the case of trimers (3.3(3)).Each Cu(II) ion coordinates nearest nitrogen atoms (N') at an average distance of 1.97 Å (Cu-N') and remote nitrogen atoms (N") at an average distance of 3.00 Å (Cu-N") (Table 2).In the trimer model, the average C.N. value of Cu(II) should be ~1.67,both for the nearest and the remote nitrogen atoms.Thus, the C.N. values of 1.6 obtained from the fi ing of EXAFS spectra both for Cu-N' and Cu-N" (Table 2) agree very well with the trinuclear structure of the Cu(L 2 )Cl2 complex.
According to EXAFS spectra fi ing for 4, the dimer is the most likely model (Figure 12b).Each Cu(II) ion in the dimer coordinates two nitrogen atoms N' + N" with distances 2.00 Å and 2.94 Å, respectively (Table 3).The coordination is supplemented by two bromide ions (Figure 12b).The average Cu-Br distance is 2.43 Å, which is in good agreement with the XRD data for the Cu(L 1 )2Br2 (1) mononuclear complex.

Magnetic Properties of the Complexes
With the structural data in hand, we can discuss the mechanisms of interactions between Cu 2+ ions in the studied complexes.In the case of Cu(L 1 )2Br2 (1), the situation is quite simple: the large distance between Cu 2+ ions in its crystal structure and the absence of pathways capable of providing exchange interaction leave the metal ions almost perfectly isolated from each other.The observed AF interaction is so weak (|θ| < 0.05 K) that one can hardly conclude with certainty whether it is of exchange or dipole-dipole origin.A bit stronger yet still very weak (θ ≈ -0.2 K) AF interaction was recently reported for the isostructural Cu(L 1 )2Cl2 complex [22], confirming the magnetically dilute state of Cu 2+ ions in this family of compounds.
In contrast to 1, the explanation of magnetic properties of 3 and 4 (Figure 8) is challenging.According to the EXAFS data, these complexes possess remarkably different structures, with Cu 2+ ions in 3 and 4 being arranged in trimer and dimer blocks, respectively.Given the short distance between Cu 2+ ions in the crystal structure (≈ 3.3 Å in 3, and ≈ 3.89 Å in 4), one would naturally expect a pronounced behaviour of magnetic trimers in 3 and dimers in 4. In the former case, the effective magnetic moment should decrease by The fitting results of EXAFS spectra for 3 signify the formation of a trimer, as follows from the obtained coordination numbers (C.N.) and the interatomic distances (Table 2).Copper(II) ions form three different coordination sites: CuNCl 3 , CuN 2 Cl 3 , and CuN 2 Cl 4 (Figure 12a).The EXAFS spectrum of the complex reveals two different average values for Cu-Cl (2.26 Å) and Cu-Cl (2.67 Å) distances with the corresponding average C.N. values equal to 2.6 and 0.8, respectively.The total coordination number of Cu(II) is 3.4, which is very close to the average C.N. value of Cu(II) in the case of trimers (3.3(3)).Each Cu(II) ion coordinates nearest nitrogen atoms (N ) at an average distance of 1.97 Å (Cu-N ) and remote nitrogen atoms (N ) at an average distance of 3.00 Å (Cu-N ) (Table 2).In the trimer model, the average C.N. value of Cu(II) should be ~1.67,both for the nearest and the remote nitrogen atoms.Thus, the C.N. values of 1.6 obtained from the fitting of EXAFS spectra both for Cu-N and Cu-N (Table 2) agree very well with the trinuclear structure of the Cu(L 2 )Cl 2 complex.
According to EXAFS spectra fitting for 4, the dimer is the most likely model (Figure 12b).Each Cu(II) ion in the dimer coordinates two nitrogen atoms N + N with distances 2.00 Å and 2.94 Å, respectively (Table 3).The coordination is supplemented by two bromide ions (Figure 12b).The average Cu-Br distance is 2.43 Å, which is in good agreement with the XRD data for the Cu(L 1 ) 2 Br 2 (1) mononuclear complex.

Magnetic Properties of the Complexes
With the structural data in hand, we can discuss the mechanisms of interactions between Cu 2+ ions in the studied complexes.In the case of Cu(L 1 ) 2 Br 2 (1), the situation is quite simple: the large distance between Cu 2+ ions in its crystal structure and the absence of pathways capable of providing exchange interaction leave the metal ions almost perfectly isolated from each other.The observed AF interaction is so weak (|θ| < 0.05 K) that one can hardly conclude with certainty whether it is of exchange or dipole-dipole origin.A bit stronger yet still very weak (θ ≈ −0.2 K) AF interaction was recently reported for the isostructural Cu(L 1 ) 2 Cl 2 complex [22], confirming the magnetically dilute state of Cu 2+ ions in this family of compounds.
In contrast to 1, the explanation of magnetic properties of 3 and 4 (Figure 8) is challenging.According to the EXAFS data, these complexes possess remarkably different structures, with Cu 2+ ions in 3 and 4 being arranged in trimer and dimer blocks, respectively.Given the short distance between Cu 2+ ions in the crystal structure (≈3.3 Å in 3, and ≈3.89 Å in 4), one would naturally expect a pronounced behaviour of magnetic trimers in 3 and dimers in 4. In the former case, the effective magnetic moment should decrease by √ 3 times with cooling, since AF trimers behave at low temperatures (k B T << J) essentially as paramag-netic S = 1/2 centres.In the latter case, the magnetic susceptibility should pass through a maximum at a certain temperature and then drop to zero in the zero-T limit, where AF dimers turn into the ground-state singlet state [26,34].Apparently, these expectations fail; the χ p (T) data (Figure 8) for neither 3 nor for 4 follow the expressions developed for AF dimers and trimers [26,34].To make a side-by-side comparison of complexes 3 and 4 whose magnetic behaviours should allegedly be very different, we plotted their µ eff (T) dependences in Figure 13.As can be seen, the µ eff (T) curves of 3 and 4 turn out to be surprisingly similar and resemble the behaviour of a uniform isotropic antiferromagnet, with just the exchange interaction in 4 being slightly (by ~10%) weaker than in 3. Additional information can be obtained from the low-T M(H) data that should follow the theoretical dependence M(H) = N A gµ B SB S gµ B k B T SH based on the Brillouin function B S (x).A fit taking into account the isotropic AF exchange interaction (θ ≈ −0.70 K for 3) and the g-factor of Cu 2+ ions g ≈ 2.20 gives a pretty good description of the M(H) data for 3 (inset in Figure 13) without any additional adjustable parameters.This gives a solid evidence that the magnetic system of 3 is composed of paramagnetic S = 1/2 centres whose number is equal to the number of Cu 2+ ions, but not of trimers.The same is apparently true for the complex 4. The only apparent way to reconcile the observations that the Cu 2+ ions in 3 and 4 act magnetically as individual centres, while structurally they are arranged in trimers and dimers, is to assume that the bonds inside trimers and dimers are ineffective in delivering the exchange interaction.The latter turns the dinuclear and trinuclear structural blocks into sets of individual Cu 2+ ions that interact with ions from neighbouring molecules.
where AF dimers turn into the ground-state singlet state [26,34].Apparently, these tations fail; the χp(T) data (Figure 8) for neither 3 nor for 4 follow the expressions oped for AF dimers and trimers [26,34].To make a side-by-side comparison of com 3 and 4 whose magnetic behaviours should allegedly be very different, we plo e µeff(T) dependences in Figure 13.As can be seen, the µeff(T) curves of 3 and 4 turn be surprisingly similar and resemble the behaviour of a uniform isotropic antifer net, with just the exchange interaction in 4 being slightly (by ~10%) weaker than in ditional information can be obtained from the low-T M(H) data that should foll theoretical dependence M(H) = N gµ SB ( µ SH) based on the Brillouin functio A fit taking into account the isotropic AF exchange interaction (θ ≈ −0.70 K for 3) a g-factor of Cu 2+ ions g ≈ 2.20 gives a pre y good description of the M(H) data for in Figure 13) without any additional adjustable parameters.This gives a solid ev that the magnetic system of 3 is composed of paramagnetic S = 1/2 centres whose n is equal to the number of Cu 2+ ions, but not of trimers.The same is apparently true complex 4. The only apparent way to reconcile the observations that the Cu 2+ ions i 4 act magnetically as individual centres, while structurally they are arranged in and dimers, is to assume that the bonds inside trimers and dimers are ineffective in ering the exchange interaction.The la er turns the dinuclear and trinuclear str blocks into sets of individual Cu 2+ ions that interact with ions from neighbouring cules.At a first glance, the complex Cu(e da)(C2N3)2 (2) exhibits a rather ordinary m behaviour of AF chains of Cu 2+ ions supplemented by a Curie-Wiess contribution from monomers and chain fragments consisting of an odd number of Cu 2+ ions [27- At a first glance, the complex Cu(ettda)(C 2 N 3 ) 2 (2) exhibits a rather ordinary magnetic behaviour of AF chains of Cu 2+ ions supplemented by a Curie-Wiess contribution χ CW (T) from monomers and chain fragments consisting of an odd number of Cu 2+ ions [27][28][29][30]35].A common approach to fit the magnetic susceptibility χ ch (T) from uniform AF Heisenberg chains described by the Heisenberg Hamiltonian H = −J ∑ i → S i • → S i+1 (J < 0 is for the AF exchange) is to use numerical calculations [27,29,35] or a well-known polynomial approximation [28].A fine tuning of the model fit is usually achieved by introducing an additional inter-chain interaction [28][29][30] or by modifying J or the g-factor [29,30].However, these approaches failed to produce a fair-quality fit to the χ p (T) data of 2, and so did the dimer model [26].
To make a more detailed analysis of the χ p (T) data of 2, one needs, firstly, to separate the contributions from Cu 2+ chains, χ ch (T), and from individual paramagnetic (PM) centres, χ CW (T).This can be performed by examining the low-T magnetization curve M(H) (Figure 14a), since the chain susceptibility at available fields should be virtually independent of H, while the magnetization of PM centres follows the dependence M(H) = N A gµ B SB S gµ B k B T SH .As can be seen in Figure 14a, the measured χ(H)/χ(0) is very close to the behaviour of isolated S = 1/2 centres, implying that it is the PM centres that dominate the low-T magnetization.Calculations show that 5.9% of Cu 2+ ions in the sample 2 behave as isolated S = 1/2 centres; the evaluated χ CW (T) curve is shown by the solid line in Figure 14b.
Int. J. Mol.Sci.2023, 24, x FOR PEER REVIEW 13 of 20 chains described by the Heisenberg Hamiltonian H = −J ∑ S ⃗ ⋅ S ⃗ (J < 0 is for the AF exchange) is to use numerical calculations [27,29,35] or a well-known polynomial approximation [28].A fine tuning of the model fit is usually achieved by introducing an additional inter-chain interaction [28][29][30] or by modifying J or the g-factor [29,30].However, these approaches failed to produce a fair-quality fit to the χp(T) data of 2, and so did the dimer model [26].
To make a more detailed analysis of the χp(T) data of 2, one needs, firstly, to separate the contributions from Cu 2+ chains, χch(T), and from individual paramagnetic (PM) centres, χCW(T).This can be performed by examining the low-T magnetization curve M(H) (Figure 14a), since the chain susceptibility at available fields should be virtually independent of H, while the magnetization of PM centres follows the dependence M(H) = N gµ SB ( µ SH).As can be seen in Figure 14a, the measured χ(H)/χ(0) is very close to the behaviour of isolated S = 1/2 centres, implying that it is the PM centres that dominate the low-T magnetization.Calculations show that 5.9% of Cu 2+ ions in the sample 2 behave as isolated S = 1/2 centres; the evaluated χCW(T) curve is shown by the solid line in Figure 14b.The χch(T) contribution (red symbols in Figure 14b) obtained by subtracting χCW(T) from the raw χp(T) data passes through a maximum at T′max ≈ 14.1 K with a subsequent decrease to zero with further cooling.This behaviour is much different from that of uniform AF S = 1/2 chains, where the magnetic susceptibility drops by only one quarter in the zero-T limit, and points to the formation of a spin gap in the Cu 2+ chains in the complex 2. The χ ch (T) contribution (red symbols in Figure 14b) obtained by subtracting χ CW (T) from the raw χ p (T) data passes through a maximum at T max ≈ 14.1 K with a subsequent decrease to zero with further cooling.This behaviour is much different from that of uniform AF S = 1/2 chains, where the magnetic susceptibility drops by only one quarter in the zero-T limit, and points to the formation of a spin gap in the Cu 2+ chains in the complex 2. Indeed, the lowest temperature part of the χ ch (T) curve can be well described by a simple expression for chains with a spin gap: χ∆(T) = aT −0.5 exp(−∆/k B T) [29,36], with a spin gap ∆/k B ≈ 7 K (inset in Figure 14b).Furthermore, the χ ch (T) data in a wide temperature range can be well fitted by a polynomial expression [37], χ alt (T), suggested for alternating-exchange AF chains described by the Hamiltonian H [28,29,36,37] (dashed blue lines in Figure 14b).The best fit was obtained for the alternation parameter δ = 0.86 and the largest exchange integral J/k B = −23.5 K, which means that the exchange interaction J/k B along the Cu 2+ chains in 2 alternates from −23.5 K to −20.2 K.It is worth noting that, according to numerical calculations of Ref. [29], a spin gap emerging in a partially dimerized S = 1/2 AF chain with J/k B = −23.5K and δ = 0.86 should amount to ∆/k B ≈ 6.3 K, which agrees well with our fitting result of ∆/k B ≈ 7 K.
As a matter of fact, the formation of alternating AF chains in the complex 2 could be expected straight from its structural data, which show that the distance between Cu 2+ ions along the chain alternates from 6.95 Å to 7.05 Å (Figure 2).A surprising feature that could hardly be expected is the remarkably strong exchange interaction found for dicyanamidebridged Cu 2+ ions in 2. Usually, the ability of dicyanamide bridges to transfer exchange interactions between copper ions is considered to be very poor, resulting in the J/k B values of about 1 K or below [31,32].To the best of our knowledge, the strongest coupling of Cu 2+ ions double-bridged by dicyanamide was found for [Cu 2 (dmphen) 2 (C 2 N 3 ) 4 ] (dmphen = 2,9dimethyl-1,10-phenanthroline) with J/k B ≈ 4.75 K, where the susceptibility exhibited maximum at T max ≈ 3 K [31]; this was suggested to represent the upper limit of the AF interaction through dicyanamide bridges.The present work demonstrates the ability of dicyanamide to deliver an almost five times stronger exchange interaction (J/k B ≈ −23.5 K; T max ≈ 14.1 K).

Cytotoxic Properties of the Complexes
The study of cytotoxic effects of new copper(II) halide complexes with 2,5-bis(pyridylmethylthio)-1,3,4-thiadiazole showed that complexes Cu(L 2 )Cl 2 (3) and Cu(L 2 )Br 2 (4) have no effect on the viability of HepG2 hepatocellular carcinoma cells in the concentration range from 1 to 50 µM after a 48 h exposure.The HepG2 line is quite commonly used for the early in vitro assessment of potential hepatotoxicity of novel molecules [38].Note, however, that the activity and expression of some xenobiotic-metabolizing enzymes in these cells are significantly lower than those of non-tumour human liver samples [39][40][41].For example, CYP2C9, CYP2C19, and CYP3A4, which are mainly located in the liver, are diagnostic markers of hepatocellular carcinoma [42][43][44].Therefore, not all toxic effects of reactive metabolites can be correctly estimated on HepG2-based models.Under experimental conditions similar to those of the present work, classical medications carboplatin and cisplatin had a more significant effect on HepG2 cells than the new complexes (Table 4).The present study also showed that new copper(II) halide complexes have different effects on the cell lines of different origins: in the studied concentration range, the compounds have a cytotoxic effect on the MCF-7 breast adenocarcinoma cells but did not cause the death of the HepG2 cells.Since no difference between the effects of 3 and 4 on the cells was revealed, we conclude that their cytotoxic activity does not depend on the halide ion.Under similar experimental conditions, the IC 50 values (characterizing the cytotoxic activity) of carboplatin and cisplatin with respect to MCF-7 cells are similar to those of 3 and 4 (Table 4).
The in vitro cytotoxicity study demonstrated that the further search for complexes exhibiting cytotoxic and potential antitumor activity in this series of compounds is a promising direction of research.

Materials and Methods
All of the reagents and solvents were obtained from commercial sources and used without purification.The CHN analysis was performed at the analytical laboratory of the Nikolaev Institute of Inorganic Chemistry SB RAS using a EuroVector EURO EA 3000 analyser (Pavia, Italy).The IR absorption spectra were registered using a Scimitar FTS 2000 spectrometer in the range from 4000 to 400 cm −1 .The studied samples were prepared in the form of suspensions in nujol and in fluorinated oils.Found (%): C, 30.5;H, 2.2; N, 10.1.For C 14 H 12 Br 2 CuN 4 S 3 calculated (%): C, 30.3;H, 2.8; N, 10.1.

XRD
The powder XRD patterns of polycrystalline samples 1 and 2 were collected with a BrukerD8 Advance diffractometer (sealed 40 kV/40 mA ceramic tube, LYNXEYE XE T detector, energy-discriminated CuK α ) in the Bragg-Brentano geometry at room temperature.The samples were ground in heptane and deposited in the form of a ~0.1 mm thick layer on a plastic substrate.The experiment was conducted in the range of 5-70 • with a step of 0.03 • and the total counting time of 96 s/point.
The single-crystal XRD data for the crystals of 1 and 2 were collected at 150 K with a Bruker D8 Venture diffractometer (ω-and ϕ-scans with a step of 0.5 • , three-circle fixed-χ goniometer, CMOS PHOTON III detector, Mo-I µS 3.0 microfocus source, focusing Montel mirrors, λ = 0.71073 Å, MoKα radiation, N 2 -flow thermostat).The crystal structures were solved using the ShelXT [48] and refined using the ShelXL [49] programs assisted by Olex2 GUI [50].Hydrogen atoms were located geometrically and refined in the riding model.Atomic displacements for non-hydrogen atoms were refined in harmonic anisotropic approximation with the exception of atoms of disordered thiolate groups in 1 belonging to a minor position.The corresponding S-C and C-C bond distances were restrained to be similar for the major and minor parts.The structures of 1 and 2 were deposited to the Cambridge Crystallographic Data Centre (CCDC) as a supplementary publication (No. 2259018 and 2259019).More detailed data were deposited to the CCDC and can be obtained at https://www.ccdc.cam.ac.uk/structures/ (accessed on 16 August 2023).The main crystal data and refinement details for 1 and 2 are summarized in Table S3.The main interatomic distances and bond angles are listed in Table S4.

X-ray Absorption Spectroscopy
The X-ray absorption spectra (XAS) were recorded in the 8 beamline channel at the Siberian Synchrotron and Terahertz Radiation Center located in the Novosibirsk VEPP-3 storage ring at Budker Institute of Nuclear Physics SB RAS [51].The XAS spectra of the complexes were registered in the range from 150 eV, before the CuK edge, to 800 eV, after the CuK edge, in a standard transmission mode using ionization chambers filled with Ar/He and Xe serving as the monitoring and final detectors, respectively.A Si(111) slotted single crystal was used as a two-crystal monochromator.The storage ring operated with an energy of 2 GeV and a current of 70-140 mA.To perform the measurements, the complexes were mixed with a cellulose powder as a filler and pressed into pellets.The EXAFS data extraction (pre-edge subtraction, spline background removal) was performed using a VIPER 10.17 software package [52].

Magnetic Susceptibility
Magnetization measurements were carried out using a Quantum Design MPMS-XL SQUID magnetometer in the temperature range 1.77-300 K at magnetic fields H up to 10 kOe.To test the thermomagnetic reversibility, temperature dependences of the magnetization, M(T), were measured on heating the sample after it had been cooled either in zero magnetic field or in a given magnetic field, as well as upon cooling the sample.In order to determine the paramagnetic component of the molar magnetic susceptibility, χ p (T), the temperature-independent diamagnetic contribution, χ d , and a possible magnetization of ferromagnetic micro-impurities, χ FM (T) were evaluated and subtracted from the measured values of the total molar susceptibility χ = M/H.While χ d was calculated using the Pascal's additive scheme, χ FM (T), if any, was determined from the measured isothermal M(H) dependencies and the M(T) data taken at different magnetic fields.To determine the effective magnetic moment, µ eff , and the Weiss constant, θ, the paramagnetic susceptibility χ p (T) was analysed using the Curie-Weiss dependence χ p (T) = N A µ 2 eff /3k B (T − θ), where N A and k B are the Avogadro and Boltzmann numbers, respectively.

Figure 1 .
Figure 1.Structure of the molecule of complex 1 (а).Fragment of the packing of 1 with the minimal S•••Br distance (b).

Figure 1 .
Figure 1.Structure of the molecule of complex 1 (a).Fragment of the packing of 1 with the minimal S• • • Br distance (b).

Figure 3 .
Figure 3. (a) Packing of 1D polymer chains in 2. The S…N contacts between the sulphur atoms of the thiadiazole rings and the nitrogen atoms of dicyanamide are shown by dashed lines.(b) View along the chains in 2. The S-N, Cu-Cu, and S-S distances are shown by grey, green, and yellow dashed lines, respectively (the green dashed lines of 7.84 Å indicate the translation between identical chains along the а axis).

Figure 3 .
Figure 3. (a) Packing of 1D polymer chains in 2. The S. ..N contacts between the sulphur atoms of the thiadiazole rings and the nitrogen atoms of dicyanamide are shown by dashed lines.(b) View along the chains in 2. The S-N, Cu-Cu, and S-S distances are shown by grey, green, and yellow dashed lines, respectively (the green dashed lines of 7.84 Å indicate the translation between identical chains along the a axis).

Figure 5 .
Figure 5.Comparison of experimental (black line) and simulated (red line) EXAFS spectra (a) and "radial distribution functions" (b) for complex 3.

Figure 6 .
Figure 6.Comparison of experimental (black line) and simulated (red line) EXAFS spectra (a) and "radial distribution functions" (b) for complex 4.

Figure 7 .
Figure 7. Temperature dependencies of the effective magnetic moment µ eff and inversed magnetic susceptibility 1/χ p , measured for complex 1 at magnetic fields H = 1; 10 kOe.

Figure 9 .
Figure 9. Temperature dependencies of the magnetic susceptibility χ (a), effective magnetic moment µ eff (calculated for the θ = 0 case) and inversed magnetic susceptibility 1/χ p (b) measured for complex 2 at magnetic fields H = 1; 10 kOe.The inset in (a) presents the low-temperature χ(T) data on an enlarged scale.The dashed blue line in (b) shows the Curie-Weiss fit to the high-temperature χ p (T) data.

Figure 10 .
Figure 10.Influence of complexes 3 and 4 on the viability of HepG2 (a,b) and MCF-7 (c,d) human cells.

Figure 11 .
Figure 11.Schematic comparison of the geometry of complexes 1 and Cu(tda)2H

Figure 12 .
Figure 12.Structure of complexes 3 (a) and 4 (b) according to the EXAFS data.

Figure 12 .
Figure 12.Structure of complexes 3 (a) and 4 (b) according to the EXAFS data.

Figure 13 .
Figure 13.Temperature dependencies of the effective magnetic moment µeff of Cu(pyme d (red open circles) and Cu(pyme da)Br2 (4) (blue solid circles) measured at the magnetic fiel kOe.To compare the curves, the µeff data for 4 are multiplied by 1.132.The inset shows the m field dependences of the normalized magnetic susceptibility χ(H)/χ(0) measured for 3 (re circles) and 4 (blue solid circles) at T = 1.77K.The green dashed line demonstrates a the dependence based on the Brillouin function that takes into account an isotropic AF exchang action (θ ≈ −0.70 K) and the g-factor of Cu 2+ ions g ≈ 2.20.

Figure 13 .
Figure 13.Temperature dependencies of the effective magnetic moment µ eff of Cu(pymettda)Cl 2 (3) (red open circles) and Cu(pymettda)Br 2 (4) (blue solid circles) measured at the magnetic field H = 10 kOe.To compare the curves, the µ eff data for 4 are multiplied by 1.132.The inset shows the magnetic-field dependences of the normalized magnetic susceptibility χ(H)/χ(0) measured for 3 (red open circles) and 4 (blue solid circles) at T = 1.77K.The green dashed line demonstrates a theoretical dependence based on the Brillouin function that takes into account an isotropic AF exchange interaction (θ ≈ −0.70 K) and the g-factor of Cu 2+ ions g ≈ 2.20.

Figure 14 .
Figure 14.(a) Magnetic field dependences of the magnetization M and normalized magnetic susceptibility χ(H)/χ(0) measured for Cu(e da)(C2N3)2 (2) at T = 1.77K.The blue dashed line illustrates the susceptibility behaviour of non-interacting magnetic moments (S = 1/2; g = 2).(b) Temperature dependencies of the magnetic susceptibility χp measured for 2 at magnetic fields H = 1; 10 kOe are shown by the open and solid green circles, respectively.The magnetic susceptibility of AF chains (open and solid red circles) is determined by subtracting from the measured χp data of the Curie-Weiss contribution χCW associated with isolated paramagnetic centres (solid brown line).The dashed blue line presents a fit to the data of the alternating-exchange AF Heisenberg chain model with the alternation parameter δ = 0.86 and the largest exchange integral J/kB = -23.5 K.The inset in (b) shows the low-temperature data on an enlarged scale.In addition to the alternating-exchange AF chain fit, χalt, a simple asymptotic low-T fit, χΔ, related to a chain model with a spin gap Δ/kB = 7 K is shown by the dash-do ed line.

Figure 14 .
Figure 14.(a) Magnetic field dependences of the magnetization M and normalized magnetic susceptibility χ(H)/χ(0) measured for Cu(ettda)(C 2 N 3 ) 2 (2) at T = 1.77K.The blue dashed line illustrates the susceptibility behaviour of non-interacting magnetic moments (S = 1/2; g = 2).(b) Temperature dependencies of the magnetic susceptibility χ p measured for 2 at magnetic fields H = 1; 10 kOe are shown by the open and solid green circles, respectively.The magnetic susceptibility of AF chains (open and solid red circles) is determined by subtracting from the measured χ p data of the Curie-Weiss contribution χ CW associated with isolated paramagnetic centres (solid brown line).The dashed blue line presents a fit to the data of the alternating-exchange AF Heisenberg chain model with the alternation parameter δ = 0.86 and the largest exchange integral J/k B = −23.5 K.The inset in (b) shows the low-temperature data on an enlarged scale.In addition to the alternating-exchange AF chain fit, χ alt , a simple asymptotic low-T fit, χ ∆ , related to a chain model with a spin gap ∆/k B = 7 K is shown by the dash-dotted line.

Table 3 .
Local structural parameters of 4 according to the EXAFS data.

Table 3 .
Local structural parameters of 4 according to the EXAFS data.