Nickel(II) complexes of N-CH₂CF₃ cyclam derivatives as contrast agents for ¹⁹F magnetic resonance imaging

Abstract

Kinetically inert Ni(II) complexes of N¹,N⁸-bis(2,2,2-trifluoroethyl)cyclams with hydrogen atoms or phosphonic acid groups in the N⁴,N¹¹-positions show significant ¹⁹F NMR relaxation rate enhancement useful for 19-fluorine MRI imaging.

---

Magnetic resonance imaging (MRI) is one of the most common techniques in molecular imaging. It is based on the detection of the NMR signal originating from water protons in a tissue. To increase its sensitivity, paramagnetic contrast agents (CAs) are often applied.^1,2 They affect mainly the longitudinal (T₁) relaxation time of the ¹H signal, which leads to an increase in the intensity of the water proton MRI signal. However, essentially all tissues contain water and, thus, the background signal compromises the detection accuracy. This problem can be solved by using non-proton MRI and the ¹⁹F nucleus seems to be the most promising candidate.^3–6 Natural monoisotopic ¹⁹F has an NMR resonance frequency close to that of ¹H (40.08 MHz T⁻¹ for ¹⁹F compared to 42.58 MHz T⁻¹ for ¹H) and exhibits sensitivity comparable to ¹H (83%). Fluorine concentration in organisms is virtually zero and, therefore, the lack of background in fluorine-based images enables “hot-spot” imaging. The wider spectral range of the ¹⁹F nucleus (∼350 ppm) compared to ¹H (∼10 ppm) is also beneficial for some applications. Moreover, only small hardware and software adjustments of standard ¹H scanners are needed for ¹⁹F detection.⁶ This makes the nucleus very potent for e.g. cellular tracking of labelled cell cultures.^7–11

However, the ¹⁹F nucleus present in organic molecules has usually a very long T₁ relaxation time requiring a long delay between excitation pulses; this prolongs the total duration of imaging experiments to unrealistic lengths. Shortening of the T₁ relaxation time can result in significant shortening of the experimental time. However, it is necessary to take into account the concomitant shortening of the transversal (T₂ or ) relaxation time, which leads to signal broadening and can result in very fast loss of signal intensity. It has been shown that the introduction of highly paramagnetic lanthanide(III) ions to the close vicinity of the fluorine atom(s) leads to significant shortening of the relaxation times,¹² and the ratio is in the range of 0.3–0.9, which is suitable for MRI measurements.^12c,13

It is known that, despite the low overall electronic spin (S = 1) and magnetic momentum (μ_eff ∼ 3 B.M.) of Ni(II), this ion can induce a large paramagnetic chemical shift and relaxation enhancement comparable to that of lanthanide(III) ions with higher S and μ.¹⁴ Therefore, some Ni(II) complexes have been studied as MRI agents employable in the Chemical Exchange Saturation Transfer method.¹⁵ Here, we decided to study the ¹⁹F NMR relaxation properties of Ni(II) complexes. The Ni(II) ion fits perfectly in the cavity of 1,4,8,11-tetraazacyclotetradecane (cyclam) and cyclam derivatives are well-known to form Ni(II) complexes with high thermodynamic stability, especially with ligands having coordinating pendant arms enabling octahedral binding to the metal. The 2,2,2-trifluoroethyl side arm was chosen as a group containing a high number of equivalent fluorine atoms. Therefore, ligands 1 and H₄te2p-tfe₂ (Fig. 1) were suggested for testing the ¹⁹F NMR parameters of their complexes.

Fig. 1 Ligands studied in this work.

Acylation of 1,8-dibenzylcyclam with ethyl trifluoroacetate or trifluoroacetic anhydride yielded the corresponding bis(amide). It was followed by BH₃ reduction¹⁶ in diglyme at elevated temperature and the benzyl protecting groups were removed by Pd/C hydrogenolysis to obtain ligand 1. The reaction¹⁷ of amine 1 in neat P(OEt)₃ with CH₂O led to the tetraethyl bis(methylenephosphonate) cyclam derivative. The ethylester groups were removed by transesterification with trimethylsilylbromide¹⁸ followed by the silylester hydrolysis to yield H₄te2p-tfe₂, which was isolated in a zwitterionic form after ion exchange chromatography. Synthetic details and results of a single-crystal X-ray diffraction study of 1·2HCl·2H₂O and H₄te2p-tfe₂·4HBr·0.5H₂O are given in the ESI (Fig. S5 and S6†).

Ligand 1 (in the form of a hydrochloride) reacts with Ni(II) salts in aqueous solutions to give a light greenish-blue precipitate. The structure of this compound was determined by a single-crystal X-ray study as cis-[Ni(1)(Cl)₂] (see ESI Fig. S7†); the central ion is surrounded by four cyclam nitrogen atoms in the cis-V configuration¹⁹ with two-fold symmetry (d_Ni–N = 2.10 and 2.26 Å for secondary and tertiary amines, respectively) and cis-chloride anions coordinated with d_Ni–Cl = 2.42 Å.

However, the cis-[Ni(1)(Cl)₂] complex shows extremely low solubility in all solvents. Thus, the presence of chloride ions had to be avoided during the preparation of a water-soluble complex. Therefore, ligand 1 in the form of a free base and Ni(ClO₄)₂ was used for further complex preparation. The course of the reaction in the H₂O : DMSO 1 : 6.5 mixture at 50 °C was followed by ¹⁹F NMR spectroscopy (Fig. S2†). Such a solvent mixture was used to keep the reaction mixture fully homogeneous right from the beginning as compound 1 is poorly soluble in water. The reaction proceeds (at 50 °C) through an intermediate (δ_F = −22.9 ppm) and is completed during 90 min to give the final complex with δ_F = −29.3 ppm (at 50 °C). On cooling to 25 °C, the signal shifts to −26.5 ppm. No further ¹⁹F NMR spectral changes were observed upon heating the solution at 100 °C for several days.

To obtain an aqueous stock solution, prolonged heating (80 °C) of the suspension of ligand 1 with Ni(ClO₄)₂ in H₂O : MeOH 1 : 1 (with subsequent evaporation of MeOH) was used. It led to the formation of a light blue aqueous solution of a single product with δ_F = −26.2 ppm. When the aqueous solution of the complex prepared in H₂O : MeOH was mixed with the sample prepared in H₂O : DMSO, only one symmetric signal in ¹⁹F NMR was observed revealing that the species formed in both experiments are identical complexes.

Despite a number of attempts, we were not able to crystallize this light blue product. However, red single-crystals of trans-[Ni(1)](ClO₄)₂ (Fig. 2) were obtained when the blue aqueous solution of the complex was saturated with NaClO₄ and was left standing for a few weeks. In this complex, the cyclam ring is coordinated in the centrosymmetric trans-III configuration¹⁹ (d_Ni–N = 1.95 and 1.99 Å for secondary and tertiary amino groups, respectively). Consistent with the red colour, only very weak axial interaction with perchlorate anions located in distant positions (d_Ni–O = 2.83 Å) was observed.

Fig. 2 Structures of (A): trans-[Ni(1)]²⁺ and (B): cis-[Ni(1)(H₂O)₂]²⁺ complex cations found in the solid state structures of trans-[Ni(1)](ClO₄)₂ and cis-[Ni(1)(H₂O)₂](TsO)₂, respectively. Carbon-bound hydrogen atoms are omitted for clarity. Thermal ellipsoids are shown at the 60% probability level.

Dissolution of the red trans-[Ni(1)](ClO₄)₂ complex in water produced a light blue solution with δ_F = −19.3 ppm and the species slowly isomerizes with first-order kinetics (τ_1/2 ∼3.5 h, 25 °C) to a species with δ_F −26.3 ppm (Fig. S3†). The final species is identical to the original Ni(II)–1 complex, as it was confirmed by ¹⁹F NMR after the addition of a standard. Taking into account the isomerism of metal ion–cyclam complexes,¹⁹ red trans-[Ni(1)](ClO₄)₂ with the trans-III configuration probably forms a blue hexacoordinated trans-[Ni(1)(H₂O)₂]²⁺ species upon dissolution, and this complex is rearranged in solution to the cis-[Ni(1)(H₂O)₂]²⁺ species with the cis-V cyclam conformation.

Typically, the trans-III isomer of cyclam complexes is considered to be the thermodynamically most stable one,²⁰ and the preference of the cis-V cyclam conformation for the Ni(II)–1 complex is rather surprising. However, the higher stability of the cis-V isomer over the trans-III one was supported also by the isolation of cis-[Ni(1)(H₂O)₂](TsO)₂ (Fig. 2). The structure shows Ni–N distances of 2.07 and 2.09 Å for secondary amino groups, and 2.22 and 2.25 Å for tertiary ones, respectively, with two water molecules coordinated with d_Ni–O = 2.07 and 2.10 Å.

Based on the data presented above, one can conclude that the reaction of ligand 1 with Ni(ClO₄)₂ leads to the formation of the cis-[Ni(1)(H₂O)₂](ClO₄)₂ complex. The geometries of Ni(II) coordination polyhedra found in the solid state structures are compiled in Table S2† and are comparable to other [Ni(L)(H₂O)₂] complexes of cyclam derivatives.^21,22 The Ni–F distances found in all solid-state structures were in the range of 4.8–5.4 Å (Table S3†).

The complex of the second studied ligand, H₄te2p-tfe₂, was prepared by heating the ligand together with a slight excess of NiCl₂ in aq. ammonia (pH 10) at 75 °C for 24 h; the excess of Ni(II) ions was removed by column chromatography. As mentioned above, the course of the reaction was followed by ¹⁹F NMR (Fig. S4†), which showed a fast drop in the concentration of the free ligand (δ_F = −68.3 ppm), and the formation of an intermediate (δ_F = −41.1 ppm) and its slower rearrangement to the final product (δ_F = −26.4 ppm). The time-dependence of intensities of all three signals could be satisfactorily fitted using a monoexponential function (Fig. S4†) and showed comparable rate constants for all three processes (see the ESI†). Such behaviour points to the presence of an equilibrium between the free ligand and the intermediate with an irreversible (rate-determining) reaction step leading to the formation of the final complex.

The final product was isolated in the form of light blue crystals, which were identified as (NH₄){trans-[Ni(Hte2p-tfe₂)]}·3.25H₂O by single-crystal X-ray diffraction. Therefore, the complex species present in solution are expected to be trans-[Ni(H_nte2p-tfe₂)]ⁿ⁻² (n = 0, 1) depending on pH. The molecular structure of the complex anion is shown in Fig. 3, and geometric parameters of the coordination sphere of Ni(II) and Ni–F distances are listed in Tables S2 and S3.† The cyclam ring exhibits the trans-III configuration¹⁹ (d_Ni–N are ∼2.10 and 2.11 Å for amino groups bearing the methylenephosphonate pendant arms, and 2.22 and 2.23 Å for those substituted by trifluoroethyl groups) with the oxygen atoms of phosphonate groups occupying apical positions (Ni–O distances are 2.06 and 2.10 Å, respectively). The molecules of the complex are connected via short hydrogen bonds between the oxygen atoms of protonated and unprotonated phosphonate pendants (d_O⋯O = 2.47 Å), forming infinite chains, similar to what was found for analogous complexes of cyclam-methylenephosphonate derivatives.^23,24 Bonding distances and the overall molecular structure are very similar to those of Ni(II) complexes of analogous derivatives.^22,24

Fig. 3 Molecular structure of the trans-[Ni(Hte2p-tfe₂)]⁻ anion found in the crystal structure of trans-(NH₄)[Ni(Hte2p-tfe₂)]·3.25H₂O. Carbon-bound hydrogen atoms are omitted for clarity. Thermal ellipsoids are shown at the 60% probability level.

The thermodynamics of complexing properties of the phosphonate H₄te2p-tfe₂ ligand was studied by potentiometry (see the ESI and Table S4†). The comparison of ligand stepwise protonation constants (log K_1–4 10.86, 10.09, 5.60 and 4.73) with those of the N¹,N⁸-dimethyl-N⁴,N¹¹-bis(methylenephosphonate) analogue²⁵ (log K_1–4 11.47, 12.17, 7.20 and 6.33, Table S5†) points to significantly decreased ligand basicity caused by the presence of electron-withdrawing –CH₂CF₃ groups. Surprisingly, it affects not only the first two protonation constants corresponding to the ring amino groups but also those of the phosphonate moieties, probably as a result of a strong electron-withdrawing effect transferred through intramolecular hydrogen bonds, which are expected to have a geometry analogous to that found for related cyclam derivatives.²⁵ Equilibration of the Ni(II)–H₄te2p-tfe₂ system is relatively slow and, therefore, the out-of-cell titration method had to be used. As the complexation mechanism is not fully straightforward (see above), the samples used for the out-of-cell titration were heated at 50 °C for 2 weeks to ensure quantitative rearrangement of the intermediate to the final trans isomer. The time required for equilibration was checked by ¹⁹F NMR of separate samples. The stability constant, log K_NiL = 13.28, is about 2–7 orders of magnitude lower than the constants of complexes with related ligands (Table S5†),^26,27 mainly as a consequence of the lower ligand basicity. The distribution diagram of the system (Fig. S8†) shows that full Ni(II) complexation by H₄te2p-tfe₂ is completed at pH 7 and the complex is present at this pH almost entirely in a fully deprotonated form.

For possible in vitro/in vivo utilization, kinetic inertness is a more important parameter than thermodynamic stability. Kinetic inertness is often tested in acidic solutions as acid-assisted complex dissociation. Thus, the decomposition of both studied complexes was examined in 1 M aq. HCl at 37 and 80 °C. Both complexes are decomposed relatively slowly by HCl at 37 °C (τ_1/2 ∼8 and ∼10 h for cis-[Ni(1)(H₂O)₂]²⁺ and trans-[Ni(H_nte2p-tfe₂)]ⁿ⁻², respectively) but the decomplexation of the cis-[Ni(1)(H₂O)₂]²⁺ complex is substantially accelerated at the higher temperature (80 °C, τ_1/2 ∼ 3 min compared to ∼5 h for trans-[Ni(H_nte2p-tfe₂)]ⁿ⁻², Table S6†), as the presence of apically coordinated pendant arms in the H₄te2p-tfe₂ complex enhances kinetic inertness. High inertness has been observed for several Cu(II) complexes with analogous cyclam-based ligands,²⁸ and highly protonated species of several Ni(II)^22,24 complexes of phosphonated cyclam derivatives have been isolated even in the solid state. These results suggest sufficient complex stability under physiological conditions and warrant the possible use of the Ni(II)–H₄te2p-tfe₂ complex in in vitro/in vivo applications.

As the trans-[Ni(te2p-tfe₂)]²⁻ complex is kinetically inert and promises reasonable stability in vivo, its ¹⁹F MRI-related parameters were investigated. Although the cis-[Ni(1)(H₂O)₂]²⁺ complex is not suitable for any in vivo application due to its low solubility in chloride-containing media, it was studied as well for comparative purposes as there are no related data reported in the literature at all. The ¹⁹F NMR relaxation measurement of both Ni(II) complexes at B₀ = 7.05 T showed extreme shortening of ¹⁹F NMR T₁ relaxation times by 2–3 orders of magnitude compared with the values observed for the free ligands, 1 and H₄te2p-tfe₂ (Table 1).

Table 1 ¹⁹F NMR relaxation times^a and ¹H relaxivity of the studied compounds (pH 7, 25 °C)

Parameter	1	cis-[Ni(1)(H2O)2]^2+	H4te2p-tfe2	trans-[Ni(te2p-tfe2)]^2–
a T 1 was determined using inversion recovery pulse sequence; was determined from line-width using Lorentzian-shape fitting of the signal.
B 0 = 7.05 T (300 MHz for ^1H, 282 MHz for ^19F)
T 1(^19F)	0.8(3) s	1.72(1) ms	0.5(1) s	2.8(7) ms
	≈76 ms	≈0.82 ms	≈50 ms	≈0.90 ms
	0.1	0.48	0.1	0.32
r 1(^1H)	—	0.83(3)	—	0.18(1)
B 0 = 4.70 T (200 MHz for ^1H, 188 MHz for ^19F)
T 1(^19F)	0.82(1) s	1.2(1) ms	1.1(2) s	4.2(1.1) ms
	3.1(1) ms	0.62(1) ms	3.1(2) ms	1.1(1) ms
	0.0038	0.52	0.0028	0.26
r 1(^1H)	—	0.66(4) s^–1 mM^–1	—	0.12(2) s^–1 mM^–1

---

The suitability of the trans-[Ni(te2p-tfe₂)]²⁻complex for ¹⁹F MRI was tested by phantom visualization at B = 4.7 T. At this magnetic field, the T₁ of the complex is very short in the millisecond range with a still convenient relaxation times ratio (Table 1). The observed relaxation times are even slightly shorter than those reported for studied Ln(III) complexes – in the cases of highly paramagnetic Tb(III), Dy(III) and Ho(III) complexes with the estimated Ln(III)–F distance lying in the range of 5–7 Å, the reported T₁ is typically in the range of 7–11 ms at 4.7 T and room temperature.^12c,13 The very short relaxation time of the Ni(II) complex required optimization of the fast pulse sequence – for the visualization of the complex, a fast gradient echo sequence with TE = 1.3 ms and TR = 3 ms was used. The slowly relaxing samples (containing the free ligand and trifluoroethanol used as a standard) were best measured using a long turbospin echo sequence employing TE = 40 ms and TR = 2000 ms. For localization of the samples, the ¹H MRI scan (Fig. 4A) was also acquired. Fig. 4 shows the results of the MRI visualization. The brightness of aq. solution of the Ni(II) complex compared to aq. solutions of the free ligand and trifluoroethanol is caused by its paramagnetism, which shortens the T₁ relaxation time of water protons (r₁(complex) = 0.12 mm⁻¹ s⁻¹, 4.7 T, 25 °C). As each sample has a different ¹⁹F NMR chemical shift (δ_F −26 ppm, −68 ppm and −77 ppm for the complex, free ligand and trifluoroethanol, respectively), each signal can be excited separately. In the case of the fast sequence, only a negligible signal of the free ligand was detected, as virtually no diamagnetic sample relaxation occurs during the sequence time-scale. On the contrary, in the experiment employing the long sequence, no signal of the paramagnetic sample was found as its magnetization relaxes before the start of acquisition.

Fig. 4 MRI study of phantoms containing trifluoroethanol, free H₄te2p-tfe₂ and the trans-[Ni(te2p-tfe₂)]²⁻ complex (c_F = 0.004 M in all samples), B = 4.7 T, 25 °C, home-made ¹H/¹⁹F surface single loop coil. (A) ¹H MRI scan, gradient echo sequence, flip angle 30°, TE = 3.7 ms, TR = 100 ms, matrix 256 × 256. (B) Overlay of ¹H MRI with ¹⁹F MRI; ¹⁹F MRI was optimized for the complex; acquired at δ = −26 ppm, gradient echo sequence, TE = 1.3 ms, TR = 3 ms, matrix 32 × 32 interpolated to 256 × 256. (C) Overlay of ¹H MRI with ¹⁹F MRI; ¹⁹F MRI was optimized for the ligand; acquired at δ = –77 ppm, turbospin echo sequence, TE = 40 ms, TR = 2000 ms, matrix 32 × 32 interpolated to 256 × 256.

The samples of free ligand 1 and cis-[Ni(1)(H₂O)₂](ClO₄)₂ show fully concordant behaviour (Fig. S9†).

In conclusion, transition metal ion complexes of fluorine-containing ligands can be considered a new class of ¹⁹F MRI contrast agents, as shown in the case of Ni(II). The presence of strongly complexing and electron donating phosphonates enhances the kinetic inertness of the studied complexes and compensates the disadvantageous coordination properties of fluorine-containing ligands. Relaxation parameters of the trans-[Ni(te2p-tfe₂)]²⁻ complex with fluorine atoms located about 5 Å from the Ni(II) centre are highly suitable for ¹⁹F MRI hot-spot imaging employing fast pulse sequences. As the Ni(II) complexes with coordinated water molecules exhibit useful water proton T₁-relaxivity, properly designed compounds could be potentially used as dual ¹H/¹⁹F MRI contrast agents.²⁹

The work was supported by the Czech Science Foundation (P207-11-1437) and by the project of the Ministry of Health, Czech Republic, for development of research organization IN00023001 (Institutional support, Institute for Clinical and Experimental Medicine). We thank Z. Böhmová and J. Hraníček for potentiometric and AAS measurements, respectively.

References

1. The Chemistry of Contrast Agents in Medical Magnetic Resonance Imaging, ed. A. Merbach, L. Helm and É. Tóth, Wiley, Hoboken, 2nd edn, 2013.
2. P. Caravan, J. J. Ellison, T. J. McMurry and R. B. Lauffer, Chem. Rev., 1999, 99, 2293–2352; P. Hermann, J. Kotek, V. Kubíček and I. Lukeš, Dalton Trans., 2008, 3027–3047; C. F. G. C. Geraldes and S. Laurent, Contrast Media Mol. Imaging, 2009, 4, 1–23.
3. J. Ruiz-Cabello, B. P. Barnett, P. A. Bottomley and J. W. M. Bulte, NMR Biomed., 2011, 24, 114–129.
4. J. C. Knight, P. G. Edwards and S. J. Paisey, RSC Adv., 2011, 1, 1415–1425.
5. J.-X. Yu, R. R. Hallac, S. Chiguru and R. P. Mason, Prog. Nucl. Magn. Reson. Spectrosc., 2013, 70, 25–49.
6. I. Tirotta, V. Dichiarante, C. Pigliacelli, G. Cavallo, G. Terraneo and F. B. Bombelli, Chem. Rev., 2015, 115, 1106–1129.
7. S. Temme, F. Boenner, J. Schrader and U. Floegel, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2012, 4, 329–343.
8. E. T. Ahrens and J. Zhong, NMR Biomed., 2013, 26, 860–871.
9. Y. B. Yu, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2013, 5, 646–661.
10. D. Bartusik and B. Tomanek, Adv. Drug Delivery Rev., 2013, 65, 1056–1064.
11. H. Amiri, M. Srinivas, A. Veltien, M. J. van Uden, I. J. M. de Vries and A. Heerschap, Eur. J. Radiol., 2015, 25, 726–735.
12. (a) P. K. Senanayake, A. M. Kenwright, D. Parker and S. K. van der Hoorn, Chem. Commun., 2007, 2923–2925; (b) A. M. Kenwright, I. Kuprov, E. De Luca, D. Parker, S. U. Pandya, P. K. Senanayake and D. G. Smith, Chem. Commun., 2008, 2514–2516; (c) K. H. Chalmers, E. De Luca, N. H. M. Hogg, A. M. Kenwright, I. Kuprov, D. Parker, M. Botta, J. I. Wilson and A. M. Blamire, Chem. – Eur. J., 2010, 16, 134–148; (d) K. H. Chalmers, M. Botta and D. Parker, Dalton Trans., 2011, 40, 904–913; (e) P. Harvey, I. Kuprov and D. Parker, Eur. J. Inorg. Chem., 2012, 2015–2022; (f) P. Harvey, K. H. Chalmers, E. De Luca, A. Mishra and D. Parker, Chem. – Eur. J., 2012, 18, 8748–8757; (g) E. De Luca, P. Harvey, K. H. Chalmers, A. Mishra, P. K. Senanayake, J. I. Wilson, M. Botta, M. Fekete, A. M. Blamire and D. Parker, J. Biol. Inorg. Chem., 2014, 19, 215–227.
13. K. H. Chalmers, A. M. Kenwright, D. Parker and A. M. Blamire, Magn. Reson. Med., 2011, 66, 931–936.
14. E. Belorizky, P. H. Fries, L. Helm, J. Kowalewski, D. Kruk, R. R. Sharp and P.-O. Westlund, J. Chem. Phys., 2008, 128, 052315.
15. A. O. Olatunde, S. J. Dorazio, J. A. Spernyak and J. R. Morrow, J. Am. Chem. Soc., 2012, 134, 18503–18505.
16. W. Curran and R. Angier, J. Org. Chem., 1966, 31, 3867–3868.
17. X. Sun, M. Wuest, Z. Kovács, A. D. Sherry, R. Motekaitis, Z. Wang, A. E. Martell, M. J. Welch and C. J. Anderson, J. Biol. Inorg. Chem., 2003, 8, 217–225.
18. C. E. McKenna, M. T. Higa, N. H. Cheung and M.-C. McKenna, Tetrahedron Lett., 1977, 18, 155–158.
19. B. Bosnich, C. K. Poon and M. Tobe, Inorg. Chem., 1965, 4, 1102–1108.
20. (a) K. R. Adam, I. M. Atkinson and L. F. Lindoy, Inorg. Chem., 1997, 36, 480–481; (b) M. Zimmer, Coord. Chem. Rev., 2001, 212, 133–163.
21. M. A. Donnelly and M. Zimmer, Inorg. Chem., 1999, 38, 1650–1658.
22. J. Kotek, P. Vojtíšek, I. Císařová, P. Hermann and I. Lukeš, Collect. Czech. Chem. Commun., 2001, 66, 363–381.
23. J. Kotek, P. Hermann, I. Císařová, J. Rohovec and I. Lukeš, Inorg. Chim. Acta, 2001, 317, 324–330.
24. J. Havlíčková, H. Medová, T. Vitha, J. Kotek, I. Císařová and P. Hermann, Dalton Trans., 2008, 5378–5386.
25. J. Kotek, P. Vojtíšek, I. Císařová, P. Hermann, P. Jurečka, J. Rohovec and I. Lukeš, Collect. Czech. Chem. Commun., 2000, 65, 1289–1316.
26. I. Svobodová, J. Havlíčková, J. Plutnar, P. Lubal, J. Kotek and P. Hermann, Eur. J. Inorg. Chem., 2009, 3577–3592.
27. I. Svobodová, P. Lubal, J. Plutnar, J. Havlíčková, J. Kotek, P. Hermann and I. Lukeš, Dalton Trans., 2006, 5184–5197.
28. J. Kotek, P. Lubal, P. Hermann, I. Císařová, I. Lukeš, T. Godula, I. Svobodová, P. Táborský and J. Havel, Chem. – Eur. J., 2003, 9, 233–248.
29. M. Wolters, S. G. Mohades, T. M. Hackeng, M. J. Post, M. E. Kooi and W. H. Backes, Invest. Radiol., 2013, 48, 341–350.

---

Footnote

† Electronic supplementary information (ESI) available: Synthesis of studied compounds, single-crystal RTG diffraction data, details on potentiometric study, study of acid-assisted dissociation of prepared complexes, relaxation study, ¹H/¹⁹F MRI visualization. CCDC 1430237–1430242. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5dt04138d

---