A family of complexes with N-scorpionate-type and other N-donor ligands obtained in situ from pyrazole derivative and zerovalent cobalt. Physicochemical and cytotoxicity studies

Abstract

In situ syntheses and X-ray structures of three novel cationic–anionic complexes: [CoL¹Br][ZnL²L³ZnBr₅] (1), and [CoL¹Cl][ZnL³Br₃] (2) and [CoL¹Cl][ZnL³Cl₃] (3) L¹ = N,N,N-tris(3,5-dimethylpyrazol-1-ylmethyl)amine, L² = hexamethylenetetramine (urotropine), L³ = 3,5-dimethylpyrazole, have been reported. The presence of three different organic ligands (L¹, L² and L³) in isolated complexes results from various reactions taking place in the system which contains zerovalent cobalt and 1-hydroxymethyl-3,5-dimethylpyrazole as starting materials, in the presence of Zn(II) ions. The scorpionate-type ligand (L¹) formed in situ, possesses four potential donor sites, specifically three nitrogen donor atoms from the pyrazole rings, and one from tertiary amine, all of which coordinate to Co(II). They form a distorted trigonal bipyramidal [CoL¹X]⁺ cation whereas anionic parts include tetrahedrally coordinated zinc(II). The crystal structures and electronic (UV-Vis), infrared (FT-IR) spectra and thermal investigation of the isolated complexes have been analysed and discussed. Finally, the biological activity of 1–3 complexes was assessed. All the tested complexes expressed higher selectivity towards human cancer cells than towards human normal cells and showed a substantial antitumor activity against: colorectal adenocarcinomas Caco-2 and SW 620, hepatocellular carcinoma Hep G2 and lung carcinoma A549.

---

Introduction

A family of complexes built with poly(pyrazolyl)borates (Tp or scorpionates), introduced by Trofimenko,^1,2 was the inspiration to explore this original synthetic approach, which led to further research of their analogs and development of this major concept in coordination chemistry.^3–7

Commonly investigated multipodal ligands, such as tris(pyrazol-1-ylalkyl)amines containing nitrogen atom instead of original boron, exhibit a configuration similar to those found for anionic poly(1-pyrazolyl)borates^1,2 and neutral poly(1-pyrazolyl)methanes^3,4. For this reason, they are often called analogs of scorpionates (or N-scorpionate-type and/or N-scorpionate like ligands).^5,6 Tris- and bis-N-pyrazole derivatives with nitrogen atoms engaged in coordination are multipodal, flexible ligands capable of creating, with transition metal ions, rigid structures with five and six-membered rings.^7–17 They are able to coordinate with almost all transition metal ions, forming new compounds with unusual physicochemical and biological properties.^5,6,10,11,17 Their complexes have wide applications, e.g. as models for bioinorganic systems (metalloenzymes): CoCa, “blue-copper proteins”^12,18–20, molecular wires,²¹ selective catalysts or their precursors.^11,22,23

Among various preparative methods of metal complexes with N-scorpionate-type ligands,^6,9–17 we have explored an original, one-pot synthetic approach starting from the system containing zerovalent metals.^24–30 In addition, we have applied 1-hydroxymethyl-3,5-dimethylpyrazole (L) as a proligand resulting in formation of scorpionate-type complexes contained N,N-bis(3,5-dimethylpyrazol-1-ylmethyl)amine^27,28 and N,N,N-tris(3,5-dimethylpyrazol-1-ylmethyl)amine as ligands.^24–26,29 It should be emphasized that in contrast to the widely used multi-step reactions developed by Driessen et al.,¹² this one-step and one-pot synthetic protocol is a significant simplification of the known procedures.

So far, we have obtained a family of cationic-anionic and neutral complexes containing these scorpionate-type ligands in various Co(II) coordinate species: (CN = 6,^22,23,25 CN = 5 ^24–27 and CN = 4 ²⁶): which were formed in situ. It was found that the composition and structure of these complexes were dependent not only on the above mentioned substrates but also on the Mⁿ⁺ and ammonium salts, NH₄X.^24–27

Recently, we have found that in the [Co⁰–1-hydroxymethyl-3,5-dimethylpirazole-CdCl₂/ZnCl₂–NH₄I] system, exchanging cadmium ion for zinc ion resulted in the isolation of the product of a completely different composition and structure.^24,25 It was an unexpected result in view of a general chemical and structural similarity of Zn(II) and Cd(II).³¹

Thus, the main goals of the current study are the following: (i) to extend the knowledge of coordination chemistry of ligands created in situ from 1-hydroxymethyl-3,5-dimethylpirazole, (ii) to investigate the role of halogen ion used in our synthetic approach, (iii) to recognize and examine the biological activity and cytotoxicity profile of the isolated complexes.

Therefore, in this work, we have described a synthesis and full characteristics of three new complexes built with a N-scorpionate-type tetradentate ligand formed in situ through one pot synthesis, zerovalent cobalt, 1-hydroxymethy-3,5-dimethylpirazole, ZnCl₂ in the presence of NH₄Cl and NH₄Br. Similarities and differences of structural, spectroscopic and thermal behaviour of three obtained cationic–anionic complexes: [CoL¹Br][Zn(μ-uro)ZnL³Br₅] (1), [CoL¹Cl][ZnL³Br₃] (2) and [CoL¹Cl][ZnL³Cl₃] (3), where L¹ = N,N,N-tris(3,5-dimethylpyrazol-1-ylmethyl)amine, L² = urotropine, heksamethylenetetraamine and L³ = 3,5-dimethylpyrazole, have been studied (Scheme 1). None of these complexes has been reported so far.

Scheme 1 Ligands formed in situ.

Special attention was paid to the relationship between structural characteristics and their spectroscopic and thermal data. Given that cationic–anionic complexes tend to attract attention because of their substantial biological activity,³² cytotoxicity profiles of all the obtained complexes were assessed, and cell selectivity studies and anti-tumour activity of compounds were performed.

Results and discussion

Synthesis and structural studies of 1–3

Three complexes were obtained in one pot synthesis from the system: [Co⁰–1-hydroxymethyl-3,5-dimethylpyrazole (L)–ZnCl₂–NH₄X] (Scheme 2). The isolated (1–3) species include cobalt and zinc ions in cationic and anionic parts, respectively. The cationic part contains pentacoordinate (CN = 5) cobalt ion exclusively on +2 oxidation state, formed from zerovalent powdered metal whereas the anionic part includes tetrahedrally coordinated zinc. Both ions were coordinated by organic ligands formed in situ from proligand 1-hydroxymethyl-3,5-dimethylpyrazole (L) (Scheme 2). In our synthetic procedure, both Co(II) ions and as many as three types of organic ligands: N,N,N-tris-(3,5-dimethylpyrazol-1-ylmethyl)amine (L¹), hexamethylenetetraamine (urotropine, L²) and 3,5-dimethylpyrazole (L³) were formed (Schemes 1–5). The parent ligand i.e. 1-hydroxymethyl-3,5-dimethylpyrazole (L) decomposes into the 3,5-dimethylpyrazole (L³) and formaldehyde (Scheme 4). The cleavage of the C(sp³)–N(pyrazole) bond in the 1-hydroxymethyl-3,5-dimethylpyrazole (L) leads to not only 3,5-dimethylpyrazole (L³) (similarly it happens with Ru(II) and Pt(II) complexes^34,35 and Cu(II)³⁶) but also formaldehyde, which is indispensable to create urotropine (L²).³⁷ The presence of NH₄⁺ cations derived from ammonium salts (NH₄Cl and NH₄Br) (indirectly³⁸) leads to formation of both cobalt(II) ions but also to releasing ammonia molecules which were further indispensable for the synthesis of both ligands N,N,N-tris(3,5-dimethylpyrazol-1-ylmethyl)amine (L¹), but also hexamethylenetetraamine urotropine (L²) (see eqn (1) below). This tetradentate, tripodal amine ligand (L¹), formed in situ according to Mannich condensation,³³ being tertiary belongs to the family of tris(polypyrazolyl) tripodal ligands and has coordination features similar to poly(pyrazol-1-yl)borates (scorpionates), but instead of boron it contains nitrogen as a central atom. The bonding properties of scorpionate-like ligands such as N,N,N-tris-(3,5-dimethylpyrazol-1-ylmethyl)amine are very attractive because of their multipodal coordination via four nitrogen atoms (N₃N′), which lead to a rigid, asymmetric architecture of cation complexes. The synthesis of the ligands obtained in situ follows the route depicted in Schemes 3–5.

Scheme 2 Isolation procedure for 1–3 complexes.

Scheme 3 Precursor L and formation of L¹.³³

Scheme 4 Precursor L and formation of L³.^34–36

Scheme 5 The route of formation of L² (urotropine).³⁷

Below, we present proposed reactions taking place during the isolation of a solid product from the system.

Co⁰ + 2NH₄⁺ + O₂ → Co²⁺ + 2NH₃ + 2H₂O ³⁸ (1)

Cation formation:

Co²⁺ + L¹ + X⁻ → [CoL¹X]⁺ (this work) and^24–27 (2)

Anion formation:

Zn²⁺ + 5Br⁻ + C₅N₂H₈(L²) + C₆N₄H₁₂(L³) → [Zn₂Br₅L²L³]⁻ (3)

[CoL¹Br]⁺ + [Zn₂Br₅L²L³]⁻ → [CoL¹Br][Zn₂Br₅L²L³]↓ (4)

Crystal structure

All of the studied compounds, [CoL¹Br][ZnL³(μ-uro)ZnBr₅] (1), [CoL¹Cl][ZnL³Br₃] (2) and [CoL¹Cl][ZnL³Cl₃] (3), crystallize in monoclinic symmetry and P2₁/c space group (Table 1, Fig. 1).

Table 1 Crystal data and structure refinement for [CoL¹Br][ZnL²L³ZnBr₅] (1), [CoL¹Cl][ZnL³Br₃] (2) and [CoL¹Cl][ZnL³Cl₃] (3)

	1	2	3
Chemical formula	C18H27BrCoN7·C11H20Br5N6Zn2	C23H35Br3ClCoN9Zn	C18H27ClCoN7·C5H8Cl3N2Zn
Mr	1246.93	837.08	703.70
a, b, c (Å)	8.7475 (3), 26.5384 (11), 18.0526 (8)	13.9778 (4), 12.0741 (3), 20.2696 (6)	14.4641 (7), 11.5385 (5), 20.015 (8)
α, β, γ (°)	90, 93.994 (4), 90	90, 110.222 (3), 90	90, 109.778 (4), 90
V (Å^3)	4180.6 (3)	3210.02 (15)	3143.4 (2)
μ (mm^−1)	7.31	5.11	1.66
Crystal size (mm)	0.52 × 0.13 × 0.12	0.32 × 0.29 × 0.26	0.42 × 0.30 × 0.28
Tmin, Tmax	0.225, 0.522	0.257, 0.426	0.521, 0.682
No. of measured, independent and observed [I > 2σ(I)] reflections	45 902, 7371, 4086	37 176, 6299, 3622	27 093, 7130, 4881
Rint	0.060	0.045	0.031
(sin θ/λ)max (Å^−1)	0.595	0.617	0.649
R[F^2 > 2σ(F^2)], wR(F^2), S	0.066, 0.210, 0.94	0.076, 0.278, 1.03	0.041, 0.119, 1.09
No. of reflections	7371	6299	7130
No. of parameters	468	351	351
Δmax, Δmin (e Å^−3)	1.53, −2.09	1.54, −2.35	0.57, −0.50

---

Fig. 1 Asymmetric unit of the unit cell for (a) [CoL¹Br][ZnL²L³ZnBr₅] (1), (b) [CoL¹Cl][ZnL³Br₃] (2) and (c) [CoL¹Cl][ZnL³Cl₃] (3). Hydrogen atom were omitted for clarity. Displacement ellipsoids are drawn at 30% probability.

The crystal structure analysis of 1 revealed that cobalt(II) ion is five-coordinate, where four coordination sites are occupied by nitrogen atoms of the L¹ ligand and the bromide anion lies at the last site. Similarly to our previous studies, the created cobalt(II) ion is displaced from the base of the [CoNN′₃X] trigonal bipyramid of 0.51 Å due to the stiffness of the organic ligand and its multipodal coordination. A charge of [CoL¹Br]⁺ ion is balanced by relatively large dinuclear complex ion [ZnL³(μ-uro)ZnBr₅]⁻. Here, the urotropine (L²) is a bridging ligand linking two zinc coordination centers. This fact stays in contrast to our previous studies on [CoL¹Cl][Zn(uro)I₃],²⁴ where urotropine was a monodentate ligand.³⁷ In the same system [Co/ZnCl₂–1-hydroxymethyl-3,5-dimethylpyrazole–NH₄Br] as compound 1 was isolated, the crystals of [CoL¹Cl][ZnL³Br₃] (2) were formed. On the other hand, when ammonium bromide was exchanged with chloride salt, compound 3 was isolated but formation of isostructural crystals to 1 was not observed. Compound 2 is isostructural to 3 and its unit cell volume is greater by 67 ų than 3, which results from a larger ionic radius of a bromide anion than that of chloride.

In all the presented crystal structures, cobalt(II) ion is five-coordinated and the coordination sites are arranged in a trigonal bipyramid around the metal center. Comparison of the geometry data for the coordination sphere of the cobalt(II) ion is given in Tables 2 and 3.

Table 2 Selected bond lengths (Å) for [CoL¹Br][ZnL²L³ZnBr₅] (1), [CoL¹Cl][ZnL³Br₃] (2) and [CoL¹Cl][ZnL³Cl₃] (3)

1		2		3
Co1–N12	2.079 (7)	Co1–N12	2.057 (7)	Co1–N12	2.076 (2)
Co1–N22	2.080 (7)	Co1–N22	2.078 (9)	Co1–N22	2.088 (3)
Co1–N32	2.057 (7)	Co1–N32	2.058 (8)	Co1–N32	2.058 (2)
Co1–N1	2.342 (6)	Co1–N1	2.266 (6)	Co1–N1	2.278 (2)
Co1–Br1	2.327 (2)	Co1–Cl1	2.272 (3)	Co1–Cl1	2.2479 (10)
Zn1–N41	2.020 (8)	Zn1–N41	2.032 (7)	Zn1–N41	2.034 (2)
Zn1–N51	2.124 (7)	Zn1–Br2	2.3813 (13)	Zn1–Cl2	2.2662 (8)
Zn1–Br3	2.3238 (16)	Zn1–Br3	2.3331 (17)	Zn1–Cl3	2.2509 (8)
Zn1–Br2	2.3580 (16)	Zn1–Br4	2.3384 (14)	Zn1–Cl4	2.2212 (9)
Zn2–N52	2.112 (7)
Zn2–Br5	2.332 (2)
Zn2–Br6	2.3328 (19)
Zn2–Br4	2.3398 (15)

---

Table 3 Selected geometric parameters for the coordination sphere of the cobalt(II) ion in [CoL¹Br][ZnL²L³ZnBr₅] (1), [CoL¹Cl][ZnL³Br₃] (2), [CoL¹Cl][ZnL³Cl₃] (3) and in other similar complexes

	Namine–Co–X	Npz–Co–Npz	τ	d(Co-plane3Npz)
[CoL^1Br][Zn2L^2L^3Br5] (1)	175.90(17)	119.2(3)	0.945	0.521 (1)
[CoL^1Cl][ZnL^3Br3] (2)	178.37(18)	116.8(3)	1.026	0.499 (1)
[CoL^1Cl][ZnL^2Cl3] (3)	178.44(6)	115.77(9)	1.044	0.500 (1)
[CoL^1Cl][ZnL^2I3]^24	178.3(2)	116.6(3)	1.028	0.534 (1)
[CoL^1Cl][CdI4]^25	180.0	114.10(9)	1.098	0.510 (1)
[CoL^1Cl][CdI4]^25	180.0	114.24(13)	1.096	0.507 (1)
[CoL^1NCS]2[V6O11(MeO)8]·4toluene^26	178.1(2)	116.8(2)	1.022	0.473 (1)
[CoL^1NCS]2[Co(NCS)4]·MeOH^26	178.23(17)	117.13(15)	1.018	0.481 (1)
[CoL^1NCS]2[Co(NCS)4]·MeOH^26	176.51(19)	116.62(19)	0.990	0.460 (1)

---

It was found that Co–N_amine and Co–N_pz distances are longer in 1 than in 2 and 3, probably due to less electronegativity of bromide than that of chloride. Furthermore, it was found that all the Co–N_pz bond distances are shorter than that of Co–N_amine showing an elongation of the trigonal bipyramid. The N_amine–Co–N_pz and N_pz–Co–N_pz angles are comparable in all the presented complexes, but the τ-parameter^10,39 is slightly lower for the [(CoL¹)Br]⁺ ion (1) than in [CoL¹Cl]⁺ (2 and 3). Since the τ-parameter is equal to 1 for the ideal trigonal pyramid, it suggests that the deformation of the coordination sphere of the cobalt(II) ion is rather weak. However, it appears that the τ-parameter is not a sufficient factor to determine such a deformation because it shows only an angular distortion. In the [CoL¹Br]⁺ and [CoL¹Cl]⁺ ions, the τ-parameters are comparable, but the main distortion is connected to the displacement of the Co²⁺ ion from the plane of three N_pz atoms which define the base of the trigonal bipyramid. Similar distortions and geometry parameters were observed earlier, also for the [CoL³NCS]⁺ ion (Table 3).

In both 2 and 3, the Co²⁺ ion is displaced from the base of the trigonal bipyramid of 0.50 Å. So, comparing the geometry of the [CoL¹Br]⁺ (1) and [CoL¹Cl]⁺ (2 and 3) ions, a substitution of the halogen atoms does not significantly affect the geometry of cobalt(II) coordination sphere (Table 2). The same concerns [ZnL³X]⁻ anion. The only important difference is noticed for Co–X and Zn–X bond lengths and this fact is connected to a larger ionic radius of the bromide and chloride anions.

In all the crystals presented here, geometry parameters indicate a weak nature of the N–H⋯Br/Cl intermolecular interactions (Table S1†). The strongest ones are visualized on the normalized Hirshfeld map (Fig. 2).^40,41 Since no strong hydrogen bonds occur in the crystals, these weak hydrogen bonds appear to be the most essential ones from the molecular self-assembly point of view. Comparison of the geometric parameters indicates that weaker interactions occur in 1 and 2 than in 3 due to less electronegativity of bromide than that of chloride.

Fig. 2 Normalized Hirshfeld surface around (a) [CoL¹Br]⁺ ion in 1, (b) [ZnL³Br₂(μ-uro)ZnBr₃]⁻ anion in 1, (c) [CoL¹Cl]⁺ ion in 2 and (d) [CoL¹Cl]⁺ in 3.

Infrared spectra

The analysis of IR spectra confirms the existence of new ligands formed in situ. The comparison of selected vibrational bands of proligand (L)^42,43 and 1–3 complexes confirm that during the condensation process of 1-hydroxymethyl-3,5-dimethylpyrazole (L) and NH₃ molecule, N-scorpinate-type ligand (L¹ = N,N,N tris(3,5-dimethylpyrazol-1-ylmethyl)amine) was formed. According to this, the OH stretching band at 3252 cm⁻¹ appearing in proligand disappears in the spectra of 1–3. Additionally, the spectra of 1–3 complexes exhibit corresponding IR bands, characteristic of the new N-scorpionate ligand L¹ (see Experimental part).^24–26,42 Moreover, the IR spectra of (2) and (3) are similar, which well correlates with the crystal structures data for these complexes and confirms the presence of the same ligands. In contrast, the IR spectra of (1), contain new stretching bands at 1227 cm⁻¹, 990 cm⁻¹ and 661 cm⁻¹ characteristic of a new ligand L² formed in situ (where L² = urotropine). For uncoordinated heksamethylenetetraamine (L²), these bands are observed at higher frequencies (1236 cm⁻¹, 1010 cm⁻¹ and 671 cm⁻¹).³⁷

Electronic spectra and magnetic properties

All the isolated complexes were characterized by electronic reflectance spectra. The results are collected in Table 4 and Fig. 3. The analysis of electronic spectra in the solid state, in the d–d range supports the geometry of the environment of Co(II) determined by X-ray diffraction. The spectra of 1–3 complexes are very similar to each other, but not identical (Fig. 3).

Table 4 The electronic spectra of 1–3 complexes

Compound	Band position (νmax/cm^−1)
	ν1	ν2	ν3	ν4	CT
[CoL^1Br][Zn2L^2L^3Br5] (1)	6600	11 540	16 850	19 030	33 775
					39 016
					44 579
[CoL^1Cl][ZnL^3Br3] (2)	6470	11 670	16 650	19 590	33 775
					38 590
					43 250
[CoL^1Cl][ZnL^2Cl3] (3)	6300	11 620	16 590	19 590	33 300
					39 020
					43 900

---

Fig. 3 Diffuse reflectance spectra of 1–3 complexes.

The reflectance spectra of 1–3 exhibit in visible part four well-resolved bands which can be assigned to the following transitions: ⁴E₁′′(F) ← ⁴A₂′(F) (ν₁), ⁴E₁′(F) ← ⁴A₂′(F) (ν₂), ⁴A₂′(P) ← ⁴A₂′(F) (ν₃); ⁴E′′(P) ← ⁴A₂′(F) (ν₄), characteristic of a high-spin five coordinated cobalt(II) complex of trigonal bipiramidal geometry.⁴⁴ These data correspond well with reflectance spectra exhibited by complexes with chromophore {CoN₄Br},⁴⁵ {CoN₄Cl}.⁴⁶ An atypical arrangement of the coordination sphere of cobalt(II) due to the peculiar chemical structure of L¹, the structures of 1–3 complexes are slightly modified due to C_3v point group symmetry.⁴⁴ Distortion from ideal trigonal bipiramidal geometry is consistent with the selected geometric τ parameter^10,39 presented in Table 3.

The values of effective magnetic moments at room temperature (μ_eff = 4.56 MB for 1), (μ_eff = 4.17 MB for 2) and (μ_eff = 4.12 MB for 3), are very close to that of spin-only value expected for trigonal bipyramidal complexes having one Co(II) ion with three unpaired electrons.^45,47 These values correspond well to those generally reported for mononuclear trigonal bipyramidal high spin cobalt(II) complexes with the {N₄Br}⁴⁵ and {N₄Cl}⁴⁶ donor set. These complexes exhibit magnetic moments in the range of 4.1–4.8 μB.^45,47

Thermal analysis

To examine the thermal stability and structure similarities and differences of 1–3 complexes, thermogravimetric analysis has been carried out in ca. 30–1000 °C range in air, using TG, DTG, SDTA curves. The thermogravimetric data and proposed decomposition were collected in Table S2.†

The results show that the decomposition of (1–3) in air proceeds in multi-stages (Fig. 4). Thermogravimetric analyses show that all the isolated complexes are stable but compound 1 is less (up to ca. 170 °C) thermally stable in comparison to 2 and 3, which are stable up to about 190 °C (Fig. 4). The thermal decomposition of complex 1 proceeds in four stages. The first step in TG curve in temperature range (170–300 °C), accompanied with weight loss of 9.95% (calc. 8.99%) corresponds to a release of urotropine (L²) ligand. This bridging N-coordinated ligand (L²) is weaker bound to Zn(II) ions than tetradentate amine (L¹) to Co(II) ions which decompose at a higher temperature (Table S2†). Further heating leads to the formation of mixture of ZnO and ZnCo₂O₄ as final products, confirmed from their XRD patterns.⁴⁸

Fig. 4 Thermogravimetric curves of 1–3 compounds.

Thermal decomposition behavior of complex 2 is quite different than that of 1 but, as expected, very similar to that of compound 3. The thermal decomposition of 2 and 3 occurs in two, well-defined steps which are attributed to decomposition of the part of tetradentate amine (L¹) and pyrazole ligand (L³). Further heating leads to oxides of central ions. The total mass loss occurring up to 900 °C is in agreement with formation of the mixture of ZnO and Co₃O₄ as a final residue identified on the basis of ICDD using XRAYAN package.⁴⁸

Study of anti-proliferative activity and selectivity against tumour and normal cells

In order to find out if the newly synthesised complexes possess biological activity, we exposed human cultured cells to 1–3 compounds. Anti-proliferative MTT assay was used for evaluation of inhibitory potency of complexes after 72 h of exposure to cells. The inhibitory effect was investigated towards a panel of human cancer cells (colorectal adenocarcinomas Caco-2 and SW 620, hepatocellular carcinoma Hep G2 and lung carcinoma A549). The influence on tumour cells was compared to the effect measured in normal human fibroblasts (BJ) culture (mortal cells) (Table 5). The effect of treatment with increasing concentrations of the tested compounds was compared to reference salts (ZnCl₂, NH₄Br, NH₄Cl), which were used as substrates for synthesis and CoCl₂, a source of Co(II) ions.

Table 5 IC₅₀ values [μM] of tested compounds at inhibiting the proliferation of normal human fibroblasts (BJ) and human tumor cells (CaCo-2, SW 620, Hep G2, A549) as determined by the MTT assay. Results are means ± SD (n = 3)

Compound	BJ fibroblasts	CaCo-2	SW 620	HEP G2	A549
1	22.7 ± 3.5	2.4 ± 0.7	5.5 ± 0.9	55.4 ± 5.1	22.4 ± 3.1
2	24.3 ± 2.9	1.3 ± 0.3	11.7 ± 1.6	20.8 ± 2.0	21.3 ± 2.2
3	24.2 ± 3.2	0.7 ± 0.3	6.5 ± 1.1	22.4 ± 3.4	22.3 ± 3.8
[CoL^1(NCS)2]^26	19.2 ± 3.1	1.4 ± 0.3	192.9 ± 11.1	23.2 ± 3.4	23.4 ± 3.4
ZnCl2	98.7 ± 18.7	131.0 ± 11.1	236.9 ± 24.2	232.3 ± 26.0	142.6 ± 17.2
CoCl2	282.5 ± 28.4	n.d	356.4 ± 23.9	160.2 ± 18.1	281.5 ± 16.7
NH4Br	n.d	n.d	n.d	n.d	n.d
NH4Cl	n.d	n.d	n.d	n.d	n.d
Cisplatin	13.0 ± 1.9	5.6 ± 0.6	12.0 ± 1.1	21.3 ± 2.4	16.9 ± 1.2

---

Additionally, the inhibitory activities of 1–3 compounds which contained both Co(II) and Zn(II) ions have been compared to the homonuclear complex [CoL¹(NCS)₂],²⁶ isolated by us earlier, which contained only Co(II) and scorpionate-type ligand (L¹). The anti-proliferative activities of all tested compounds were compared to the reference cisplatin.

IC₅₀ values, calculated from dose–survival curves, are shown in Table 5. In general, the newly synthesized 1–3 complexes express a strong anti-proliferative potency towards all the tested cancer cell lines. Notably, cationic–anionic 1–3 complexes are less cytotoxic against normal human fibroblasts (BJ), compared to cancer cells, with only one exception: 1 exerts a higher inhibitory effect towards mortal cells than towards Hep G2 carcinoma (with IC₅₀ 22.7 ± 3.5 vs. 55.8 ± 5.1, respectively).

Our results revealed an anti-proliferative potential of the new synthesized 1–3 complexes on cancer cells. What is particularly worth noting is the inhibition of cancer cells growth by 1–3 complexes in SW 620. IC₅₀ SW 620 values for 1, 2 and 3 are in general higher by one order of magnitude than the value measured for [CoL¹(NCS)₂]²⁶ (Table 5). The obtained data point out that the presence of Zn(II) in the structure may enhance the inhibitory potency against neoplastic cells (SW 620).

What is more, the anti-proliferative action of 1, 2 and 3 towards cancer cells are one or two orders of magnitude higher comparing with inorganic salts counterparts (Table 5).

Taking into account the influence of hexamethylenetetramine (urotropine) moiety on cytotoxic properties, the obtained results show that the presence of urotropine in the structure of 1 may influence the anti-proliferative action as well as the selectivity of the compound towards distinct-originated tumor cells, especially Hep G2 and SW 620. The cytotoxicity of 1 towards Hep G2 is lower when compared to the effect of the other tested 2 and 3 complexes in this cell line. On the contrary, the inhibition of SW 620 cells caused by 1 is twice as high higher than that caused by 2 (Table 5). The obtained data show that the presence of urotropine (L²) in complex 1 may influence its action in the cell. It should be stressed that hexamethylenetetramine (urotropine) has been clinically used and tested in anticancer therapies with promising effects, especially when combined treatment with other drugs, such as cisplatin, has been implemented.^49,50

Additionally, our results show that 1, 2 and 3 compounds reveal selectivity, especially towards colorectal adenocarcinomas, Caco-2 and SW 620. In the performed experiments, CaCo-2 cells (at Duke's B stage) are the most susceptible to the tested complexes, with the lowest IC₅₀ values. At the same time, the pronounced inhibition of SW 620 cells (Duke's C stage) by 1, 2 and 3 was measured (Table 5). The anti-proliferative action of all the tested complexes (1–3) in the invasive and resistant to treatment SW 620 cell line is of interest, since successful inhibition of metastatic colorectal cancer is extremely difficult. The obtained result is a relevant finding and anti-neoplastic mechanism of influence should be elucidated.⁵¹

Comparing the effect of the isolated 1, 2 and 3 complexes with well-studied cytotoxicity of cisplatin,⁵¹ our in situ synthesized compounds are as effective (or even more efficient) in inhibiting the growth of cancer cells as cisplatin (Fig. 5). It is worth noting that cytotoxicities of 1 and 3 in both colon adenocarcinoma cells exceed at least twice that of the cisplatin. At the same time, the inhibitory effects in A549 are comparable to that of cisplatin.

Fig. 5 Comparative plot of IC₅₀ values [μM] of 1, 2,3 complexes and [Co L¹(NCS)₂]²² and cisplatin against normal human fibroblasts (BJ) and human tumor cell lines (Caco-2, SW 620, Hep G2, A549).

It was well published that cisplatin exerts a toxic effect in normal cells, causing detrimental effects during therapy.⁵² In contrast to cisplatin, our complexes 1 and 3 are less cytotoxic towards normal cells than cisplatin (in fibroblasts, IC₅₀ 22.7 ± 3.5 for 1 and IC₅₀ 24.2 ± 3.2 for 3 vs. 13.0 ± 1.9 for cisplatin) (Table 5).

Summing up, the isolated 1, 2 and 3 complexes show higher selectivity towards cancer cells than towards normal cells. The tested complexes express a potent anti-proliferative activity towards tumour cells with the relevant cytotoxicity (even greater than that of cisplatin), which additionally makes them interesting candidates for further investigation of biological properties.

Experimental

Materials and measurements

All reagents were purchased from commercial sources and used without further purification. The experiments were carried out in air atmosphere. Cobalt powder, ZnCl₂, NH₄Cl, NH₄Br, and 1-hydroxymethy-3,5-dimethylpyrazole were purchased from Aldrich Chemical Company.

Elemental analyses were performed with a Perkin Elmer Elemental Analyzer 2400 CHN and AES-ICP 3410 emission spectrometer (Co) using appropriate Aldrich standards. IR spectra were recorded on a Perkin-Elmer FTIR 1600 (4000–400 cm⁻¹) and Perkin-Elmer FTIR 2000 (600–100 cm⁻¹) spectrophotometers in KBr pellets and nujol mull, respectively. The electronic reflectance spectra in the range 50 000–5000 cm⁻¹ were measured on a Cary 500 Scan (Varian) UV-VIS-NIR Spectrophotometer. Thermal decomposition was carried out in air, over 25–1000 °C range with a heating rate of 5 deg min⁻¹ using Metter Toledo TGA/SDTA 851 type derivatograph. Calcined Al₂O₃ was used as the reference material. The measurements were carried out in air using Al calorimeter which was calibrated with pure indium and zinc. Magnetic moments were measured using a MSB-MKI instrument (Sherwood Scientific Ltd) at ambient temperature with Co[Hg(SCN)₄] as standard.

General synthetic procedure

All the complexes were prepared according to the general procedure described previously.^20–25 Cobalt powder (325 mesh), 1-hydroxymethyl-3,5-dimethylpyrazole (L), NH₄X (X = Cl⁻ and Br⁻) and ZnCl₂ were used in 2 : 4: 6 : 1 molar ratio for 1–3, respectively, similarly as in ref. 24–27. Cobalt powder (0.0295 g, 0.5 mmol) was added to a methanol solution (70 cm³) of 1-hydroxymethyl-3,5-dimethylpyrazole (0.1274 g, 1.0 mmol) followed by ZnCl₂ (0.0340 g, 0.25 mmol) (compounds 1–2) and NH₄X, (1.5 mmol) (NH₄Cl: 0.0804 g) or (NH₄Br: 0.1470 g) (compound 1–2). All reagents were heated to 50–60 °C and magnetically stirred until almost complete dissolution of the cobalt powder was observed (ca. 5–6 h). The resulting dark violet solution was kept at room temperature and after ca. 6–9 days deep violet crystals of compound 1 appeared. Subsequently, after filtration of compound 1, under a slow evaporation of filtrate 1 at room temperature, violet crystals precipitated after one week, and were filtered off. To obtain crystals suitable for X-ray crystallographic measurements, compounds 1 and 2 were recrystallized from the mixture of methanol and isopropanol (1 : 1).

To check if the molar ratio affects the type of isolated products, compounds 1–2 were also prepared using different molar ratios of the substrates (1 : 3: 3 : 1). By using different amounts of ZnCl₂ and 1-hydroxymethyl-3,5-dimethylpyrazole (L), the same products were precipitated. The procedure for obtaining compound 3 was the same as that for complex 1 but NH₄Cl was used as the NH₄⁺ source. The reagent molar ratio (Co⁰ : L : ZnCl₂ : NH₄Cl) was also (2 : 4 : 6 : 1) and (1 : 3 : 3 : 1) and the same products were isolated. The resulting dark violet solution of reagents was kept at room temperature and after ca. 8–10 days deep violet crystals of compound 3 appeared, which were suitable for X-ray crystallographic measurements.

The compound [CoL¹Br][Zn₂L² L³Br₅] (1) is soluble in CH₃OH, CH₃CN, DMF, and isopropanol, (yield:40%). Elemental analysis: found (%): C 28.37; H 3.98; N, 14.83; and calculated (%): C 27.93; H 3.77; N 14.60; for C₂₉H₄₇N₁₃Br₆CoZn₂, mp 249 °C; IR: ν(CN_pyrazole + NN) 1248 cm⁻¹; ν(C–N) 1280 cm⁻¹; ν(CC) + (C–N)_{pyrazole ring} 1551 cm⁻¹.

The compound [CoL¹Cl][ZnL³Br₃] (2) is soluble in CH₃OH, CH₃CN, DMF, and isopropanol, (yield:45%). Elemental analysis: found (%): C 34.24; H 4.35; N, 15.42; and calculated (%): C 33.95; H 4.32; N 15.04; for C₂₃H₃₉N₉Br₃ClCoZn, mp 223 °C; IR: ν(CN_pyrazole + NN) 1248 cm⁻¹; ν(C–N) 1270 cm⁻¹; ν(CC) + (C–N)_{pyrazole ring} 1553 cm⁻¹.

The compound [CoL¹Cl][ZnL³Cl₃] (3) is soluble in CH₃OH, CH₃CN, DMF, and isopropanol, (yield:40%). Elemental analysis: found (%): C 38.81; H 5.07; N, 17.91; and calculated (%): C 39.18; H 5.14; N 17.89; for C₂₄H₃₆N₉Cl₄CoZn, mp 250 °C; IR: ν(CN_pyrazole + NN) 1248 cm⁻¹; ν(C–N) 1279 cm⁻¹; ν(CC) + (C–N)_{pyrazole ring} 1553 cm⁻¹.

Crystal structure determination

X-ray diffraction data were collected on a KUMA Diffraction KM-4 four-circle single crystal diffractometer equipped with a CCD detector using graphite-monochromatized MoK_α radiation (λ = 0.71073 Å). The experiments were carried out at 295 K. The raw data were treated with the CrysAlis Data Reduction Program taking into account an absorption correction. The intensities of the reflection were corrected for Lorentz and polarization effects. The crystal structures were solved by direct methods⁵³ and refined by full-matrix least-squares method using SHELXL-2014 and ShelXle programs^53,54 (Table 1). Non-hydrogen atoms were refined using anisotropic displacement parameters. H-atoms were visible on the Fourier difference maps, but placed by geometry and allowed to refine “riding on” the parent atom.

Biological studies

Cell cultures. Human cell lines were derived from the American Type Cell Culture collection, ATCC (LGC Standards-ATCC Teddington, Great Britain), ATCC designations were as follows: BJ, normal adherent human skin fibroblasts, CRL-2522; HEP G2, hepatocellular carcinoma, HB-8065; A549, lung carcinoma, HTB-37; Caco-2, Duke's type B colorectal adenocarcinoma, CCL-228; SW 620, Duke's type C colorectal adenocarcinoma, CCL-227. The cells were grown as monolayer cultures in Eagle's Minimum Essential Medium, EMEM (the skin fibroblasts BJ and HEP G2 cells) or in Dulbecco's Modified Eagle's Medium, DMEM (A549, SW 480 and SW 620 were cultured) (Sigma-Aldrich, Seelze, Germany). All the media were supplemented with 10% v/v FBS (PAA Laboratories GmbH, Austria) and with antibiotic solution (100 IU mL⁻¹ penicillin, 0.1 mg mL⁻¹ streptomycin, Gibco Laboratories, NY, USA, Minerva Biolabs, Berlin, Germany). The cells were kept under standard culture conditions (37 °C in a humidified atmosphere of 5% CO₂).

Anti-proliferative assay. The inhibitory activity of compounds was determined by using the colorimetric MTT assay MTT, 3-[4,5-dimethylthiazol-2yl]-2,5-diphenyl tetrazolium bromide purchased from Sigma-Aldrich, Seelze, Germany. The assays were performed as described previously.⁵⁵ Briefly, 200 μL of a suspension of exponentially dividing cells was placed in each well of a 96-well microtiter plate (BD Biosciences, CA, USA) at density 1 × 10⁵ cells per mL of culture medium and incubated overnight. Then the medium in each well was replaced with a new one, also containing the adequate volume of a stock solution of tested compounds or the adequate amounts of dimethylsulfoxide (DMSO) solvent (DMSO concentrations in the cell media were always the same, the dilutions were from 10⁻³ to 10⁻⁷ M). The cells were then exposed to the compounds for 72 h. The cells cultured in the medium and the solvent were positive control (100% of growth). After incubation, the medium was removed and MTT reagent was added to each well and incubated for 1–2 h. MTT formazan generated during incubation was dissolved and the absorbance was recorded at 570 nm (the reference wavelength was 630 nm) using a microplate reader Infinite M200 Pro, Tecan, Austria. IC₅₀ values, which indicated anti-proliferative potency of tested compounds, were calculated from concentration–response curves (IC₅₀ was the concentration [μM L⁻¹] of a tested compound required to decrease the cell proliferation to 50% of the control).³⁰ All the data were expressed as means ± SD.

Conclusions

In summary, we have described here one pot synthetic pathways which generated in situ cationic–anionic type complexes containing three different ligands: N,N,N-tris(3,5-dimethylpyrazolylmethyl)amine (L¹), hexamethylenetetramine (urotropine) (L²) and 3,5-dimethylpyrazole (L³) also formed in situ. The most important aspect of our research is one-pot, one-step reaction which leads to complexes obtained in situ with N,N,N-tris-(pyrazolylmethyl)amine formed in situ, in contrast to the widely used multi-step reactions developed by Driessen et al.¹²

This work is a new contribution to understanding coordinating performance of ligands obtained in situ, in relation to ions such as Co(II) obtained from zerovalent metal and Zn(II). Additionally, this study illustrates the ability of N-scorpionate-like N,N,N-tris(3,5-dimethylpyrazol-1-ylmethyl)-amine to adopt special geometries which lead to a rigid, asymmetric architecture of cationic Co(II) complexes. Their positive charge is compensated by tetrahedral, anionic Zn(II) complexes. Cobalt(II) is five-coordinate and exists in trigonal bipyramidal cation [CoN₃N′X]⁺, where four coordination sites are occupied by the nitrogen atoms of the L¹ ligand and the bromide (or chloride) anion lies at the last site. The halogen ions (Cl⁻ and Br⁻) participate in coordination, both in Co(II) as well as Zn(II) ions. It is noteworthy that in the presence of NH₄I²⁴ and NH₄Br, urotropine (L²) and inorganic and organic complexes [Zn(uro)I₃]⁻ (ref. 24) and [Zn(μ-uro)(L³)ZnBr₅] were formed, whereas in the presence of NH₄Cl only [ZnL³Cl₃] complex was formed. Formation of a urotropine ligand was not observed in this case.

It is interesting to note that N,N,N-tris(3,5-dimethyl-1-pyrazolymethyl)amine has been shown to act as a tripodal tetradentate ligand, coordinating through the amine N-atom and three pyrazolyl N-atoms only with Co(II) ions and not Zn(II). The urotropine obtained in situ coordinates exclusively to Zn(II) ions and it is a bidentate, bridging ligand. In this paper, we propose pathways describing the whole process.

The isolated 1, 2 and 3 complexes express the substantial anti-tumour activity towards a panel of human cancer cells (colorectal adenocarcinomas Caco-2 and SW 620, hepatocellular carcinoma Hep G2, lung carcinoma) with the relevant cytotoxicity (even greater than that of cisplatin). The tested complexes show a higher selectivity towards cancer cells than towards normal cells, which additionally makes them interesting candidates for further investigation of biological investigations.

Acknowledgements

Anna Adach is thankful to students, Ms E. Pietura and M. Dziewisz for their help in the synthesis of the complexes and Dr W. Surga for the thermal measurements.

References

1. (a) S. Trofimenko, J. Am. Chem. Soc., 1967, 89, 6288–6297; (b) S. Trofimenko, Chem. Rev., 1993, 93, 943–980.
2. S. Trofimenko, Polyhedron, 2004, 32, 197–203.
3. (a) C. Pettinari and R. Pettinari, Coord. Chem. Rev., 2005, 249, 525–543; (b) C. Pettinari, A. Tabacarub and S. Galli, Coord. Chem. Rev., 2016, 307, 1–31; (c) C. Pettrenari and N. Masciocchi, J. Organomet. Chem., 2005, 690, 1871–1877; (d) F. Marchetti, C. Pettinari, A. Cerquetella, A. Cingolani, R. Pettinari, M. Monari, R. Wanke, M. L. Kuznetsov and A. J. L. Pombeiro, Inorg. Chem., 2009, 48, 6097.
4. (a) C. Santini and M. Pellei, Curr. Bioact. Compd., 2009, 5, 243; (b) M. Pellei, G. Papini, G. G. Lobbia and C. Santini, Curr. Bioact. Compd., 2009, 5, 321; (c) R. F. Semeniuc and D. L. Reger, Eur. J. Inorg. Chem., 2016 DOI:10.1002/ejic.201600116; (d) L. M. D. R. S. Martins and A. J. L. Pombeiro, Eur. J. Inorg. Chem., 2016 DOI:10.1002/ejic.201600053; (e) B. A. McKeown, J. P. Lee, J. Mei, T. R. Cundari and T. B. Gunnoe, Eur. J. Inorg. Chem., 2016 DOI:10.1002/ejic.201501470; (f) D. M. Lyubov, A. V. Cherkasov, G. K. Fukin and A. A. Trifonov, Organometallics, 2016, 35, 126–137.
5. J. Reglinski and M. D. Spicer, Coord. Chem. Rev., 2015, 297–298, 181–207.
6. J. Krzystek, D. C. Swenson, S. A. Zvyagin, D. Smirnov, A. Ożarowski and J. Telser, J. Am. Chem. Soc., 2010, 132, 5241–5253.
7. G. Zamora, J. Pons and J. Ros, Inorg. Chim. Acta, 2004, 357, 2899–2904.
8. E. M. Hahn, A. Casini and F. E. Kühn, Coord. Chem. Rev., 2014, 276, 97–111.
9. (a) S. S. Massoud, F. R. Louka, Y. K. Obaid, R. Vicente, J. Ribas, R. C. Fischer and F. A. Mautner, J. Chem. Soc., Dalton Trans., 2013, 42, 3868–3878; (b) S. S. Massoud, L. L. Quan, K. Gatter, J. Albering and R. C. Fischer, Polyhedron, 2012, 31, 601–606.
10. K. Fujisawa, S. Chiba, Y. Miyashita and K. Okamoto, Eur. J. Inorg. Chem., 2009, 3921–3934.
11. E. Haldón, M. Delgado-Rebollo, A. Priet, E. Álvarez, C. Maya, M. C. Nicasio and P. J. Pérez, Inorg. Chem., 2014, 53, 4192–4201.
12. (a) G. J. Driel, W. L. Driessen and J. Reedijk, Inorg. Chem., 1985, 24, 2919–2925; (b) G. J. Kleywegt, W. G. R. Wiesmeijer, G. J. Van Driel, W. L. Driessen and J. Reedijk, J. Chem. Soc., Dalton Trans., 1985, 10, 2177–2184.
13. C. Benelli, I. Bertini, M. Vaira and F. Mani, Inorg. Chem., 1984, 23, 1422–1425.
14. G. Yang, J. Chem. Crystallogr., 2004, 34, 269–273.
15. H. Yang, Y. Tang, Z.-F. Shang, X.-L. Han and Z.-H. Zhang, Polyhedron, 2009, 28, 3491–3498.
16. (a) B. Chakraborty and T. K. Paine, Inorg. Chim., Acta, 2011, 378, 231–238; (b) D.-Y. Wu, W. Huang, L. Wang and G. Wu, Z. Anorg. Allg. Chem., 2012, 638, 401–404.
17. (a) B. Barszcz, J. Masternak and W. Sawka-Dobrowolska, J. Chem. Soc., Dalton Trans., 2013, 42, 5960–5963; (b) J. Masternak, B. Barszcz, W. Sawka-Dobrowolska, J. Wietrzyk, J. Jezierska and M. Mielczarek, RSC Adv., 2014, 4, 43962–43972.
18. M. Scarpellini, A. J. Wu, J. W. Kampf and V. L. Pecoraro, Inorg. Chem., 2005, 44, 5001–5010.
19. M. Scarpellini, J. Gätjens, O. J. Martin, J. W. Kampf, S. E. Sherman and V. L. Pecoraro, Inorg. Chem., 2008, 47, 3584–3593.
20. N. V. Fischer, G. Türkoglu and N. Burzlaff, Curr. Bioact. Compd., 2009, 5, 277–295.
21. M. Daoudi, N. Ben Larbi, A. Kerbal, B. Bennani, J. P. Launay, J. Bonvoisin, T. Ben-Hadda and P. H. Dixneuf, Tetrahedron, 2006, 62, 3123–3127.
22. M. El Kodadi, F. Malek, R. Touzani and A. Ramdani, Catal. Commun., 2008, 9, 966–969.
23. (a) N. Boussalah, R. Touzani, I. Bouabdallah, S. El Kadiri and S. Ghalem, J. Mol. Catal., 2009, 306, 113–117; (b) H. Yang, H. Gao, K. Wang and Z. Zhan, Transition Met. Chem., 2006, 31, 958–963.
24. A. Adach, M. Daszkiewicz, B. Barszcz, M. Cieślak-Golonka and G. Maciejewska, Inorg. Chem. Commun., 2010, 13, 361–364.
25. A. Adach, M. Daszkiewicz and B. Barszcz, Struct. Chem., 2010, 21, 331–336.
26. A. Adach, M. Daszkiewicz and M. Cieślak-Golonka, Polyhedron, 2012, 47, 104–111.
27. A. Adach, M. Daszkiewicz, M. Duczmal and Z. Staszak, Inorg. Chem. Commun., 2013, 35, 22–26.
28. A. Adach, M. Daszkiewicz, M. Cieślak-Golonka, T. Misiaszek and D. Grabka, Polyhedron, 2014, 78, 31–39.
29. A. Adach, M. Daszkiewicz and B. Barszcz, Polyhedron, 2015, 95, 60–68.
30. A. Adach, M. Daszkiewicz, M. Tyszka-Czochara and B. Barszcz, RSC Adv., 2015, 5, 85470–85479.
31. B. Barszcz, Coord. Chem. Rev., 2005, 249, 2259–2276.
32. V. Bobbarala, A search for antibacterial agents, InTech, Rijeka, Croatia, 2012.
33. S. G. Subramaniapillai, J. Chem. Sci., 2013, 125, 467–482.
34. D. Ghorai and G. Mani, Inorg. Chem., 2014, 53, 4117–4129.
35. (a) A. Romero, A. Vegas, A. Santos and A. M. Cuadro, J. Chem. Soc., Dalton Trans., 1987, 183; (b) A. Boixassa, J. Pons, X. Solans, M. Font-Bardia and J. Ros, Inorg. Chim. Acta, 2003, 355, 254–263.
36. B. Barszcz, T. Głowiak, J. Jezierska and A. Tomkiewicz, Polyhedron, 2004, 23, 1309–1316.
37. (a) A. M. Kirillov, Coord. Chem. Rev., 2011, 255, 1603; (b) F. B. Stocker, Inorg. Chem., 1991, 30, 1472; (c) A. N. Semakin, A. Y. Sukhorukov, Y. V. Nelyubina, Y. A. Khomutova, S. L. Ioffe and V. A. Tartakovsky, J. Org. Chem., 2014, 79, 6079–6086.
38. (a) D. S. Nesterov, O. V. Nesterova, V. N. Kokozay and A. J. L. Pombeiro, Eur. J. Inorg. Chem., 2014, 4496–4517; (b) V. V. Semenaka, O. V. Nesterova, V. N. Kokozay, R. I. Zybatyuk, O. V. Shishkin, R. B. Carlos, J. Gómez-García, J. M. Clemente-Juan and J. Jezierska, Polyhedron, 2010, 29, 1326–1336.
39. A. W. Addison, T. N. Rao, J. Reedijk, J. Rijn and G. C. Verschoor, J. Chem. Soc., Dalton Trans., 1984, 1349–1356.
40. F. L. Hirshfeld, Theor. Chim. Acta, 1977, 4, 129–138.
41. J. J. McKinnon, D. Jayatilaka and M. A. Spackman, Chem. Commun., 2007, 3814–3816.
42. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, John Wiley, New York, 4th edn, 2009.
43. V. M. Leovac, Z. D. Tomić, A. K. Kovàcs, M. D. Jaksović, L. S. Jovanović and K. M. Szécsényi, J. Organomet. Chem., 2008, 693, 77–86.
44. A. B. P., Lever Inorganic Electronic Spectroscopy, 2nd edn, Elsevier, Amsterdam, New York, Tokyo, 1984.
45. F. Meyer, K. Heinze, B. Nuber and L. Zsolna, J. Chem. Soc., Dalton Trans., 1998, 207–213.
46. Z. He, D. C. Craig and S. B Colbran, J. Chem. Soc., Dalton Trans., 2002, 4224.
47. J.-W. Lim, M. Mikuriya and H. Sakiyama, Bull. Chem. Soc. Jpn., 2001, 74, 2131–2132.
48. Powder Diffraction File, JCPDS; ICDD, 1601 Park Lane Swarthmore, PA 19081, 531, File No.1–1149, No.5-664, No. 2–770.
49. S. I. Masunaga, K. Kono, J. Nakamura, K. Tano, H. Yoshida, M. Watanabe, G. Kashino, M. Suzuki, Y. Kinashi, Y. Liu and K. Ono, Oncol. Rep., 2009, 21, 1307–1312.
50. S. I. Masunaga, K. Tano, J. Nakamura, M. Watanabe, G. Kashino, M. Suzuki and K. Y. K. Ono, Exp. Ther. Med., 2010, 1, 169–174.
51. M. J. Hannon, Pure Appl. Chem., 2007, 79, 2243–2261.
52. J. Reedijk, Eur. J. Inorg. Chem., 2009, 1303–1312.
53. G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008, 64, 112–122.
54. C. B. Hübschle, G. M. Sheldrick and B. Dittrich, J. Appl. Crystallogr., 2011, 44, 1281–1284.
55. M. Tyszka-Czochara, P. Paśko, W. Reczyński, M. Szlósarczyk, B. Bystrowska and W. Opoka, Biol. Trace Elem. Res., 2014, 160, 123–131.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 836769 (1), 1434391 (2) and 836764 (3). For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra06439f

---