Gadolinium nicotinate clusters as potential MRI contrast agents

Abstract

Three multinuclear gadolinium(III) clusters containing the nicotinate (nic) ligand, [Gd₂(nic)₆(H₂O)₄] (1), [Gd₄(μ₃-OH)₄(Hnic)₅(H₂O)₁₂](ClO₄)₈·7H₂O (2), and [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₆]₂(ClO₄)₄·4H₂O (3) were synthesized and characterized. Potential applications of these Gd(III) clusters as high-field magnetic resonance imaging (MRI) contrast agents were evaluated. The longitudinal relaxivities (r₁) of 1–3 in water at 7 T were determined to be 10.45 ± 0.16, 9.28 ± 0.06, and 2.04 ± 0.29 mM⁻¹ s⁻¹ on a per Gd ion basis, respectively. In 1% agarose solution, the r₁ values increased slightly to 10.37 ± 0.31, 10.74 ± 0.11, and 4.60 ± 0.29 mM⁻¹ s⁻¹ for 1–3, respectively. The ability to tune the number of inner-sphere water molecules, cluster sizes, organic ligands, and Gd coordination modes in such Gd(III) clusters provide interesting opportunities to further enhance the MR relaxivities for potential MRI applications.

---

Introduction

Magnetic resonance imaging (MRI) has become one of the most important techniques in modern diagnostic medicine because of its ability to non-invasively produce three-dimensional images of biological specimens with a deep tissue penetration and at a high spatial resolution.¹ In MRI, image contrasts result from differences of water relaxation time and proton density between adjacent tissues.² To enhance the image contrast, Gd(III) chelates³ are typically employed as contrast agents to reduce the longitudinal relaxation times (T₁) of water molecules. Gd(III) contrast agents effectively increase the sensitivity of MRI, allowing reliable diagnosis of diseased states that are typically not discernible in the non-contrast enhanced MR images. The effectiveness of a Gd(III) chelate as a contrast agent is determined by its ability to relax water protons, as expressed by the longitudinal relaxivity (r₁) which is the slope of the linear dependence of 1/T₁ vs. Gd concentration in mM.⁴ Clinically used Gd(III) contrast agents, such as Gd–DPTA,^5,6 are mononuclear compounds with a tightly bound chelating ligand and one to two coordinated water molecules.^7–10 These Gd contrast agents however tend to have modest r₁ relaxivities, particularly at high magnetic field strengths. As a result, significant amounts of Gd chelates need to be administered in order to afford adequate image contrasts between normal and pathological tissues in the clinic.¹¹ Significant research efforts have therefore been devoted to the development of more efficient MRI contrast agents, for example, based on dendrimers^8,12–14 and fullerenes,^15,16 Gd₂O₃ nanoparticles,^17,18 nanoscale metal–organic frameworks,^19–22 and other nanoparticles containing a high density of Gd chelates, in the past few years.

Metal clusters have attracted considerable attention in coordination and materials chemistry, and can be prepared by modular and metal-directed self-assembly methods.^23–29 Many of these large yet molecular species^30–35 possess beautiful structures as well as interesting optical and magnetic properties. Some of these metal clusters have potential applications as MRI contrast agents. For example, Pan and co-workers described that Mn₁₂–acetate [Mn^III/IV₁₂O₁₂(CH₃CO₂)₁₆(H₂O)₄] could be coated with a polymeric shell via ligand-exchange to afford contrast agents with large r₁ and r₂ relaxivities of 5.2 ± 0.12 mM⁻¹ s⁻¹ and 10.7 ± 0.24 at 1.5 T, respectively.³⁶ Duan et al. reported the use of hexanuclear gadolinium-based octahedron (Gd–PT1) as selective MRI contrast agents for detecting glucosamine and observed enhanced r₁ relativity (64.8 mM⁻¹ s⁻¹ on per a Gd ion basis, 400 MHz, 298 K) owing to the rigid facial bridging ligands.²⁵ In this work, we report the synthesis and characterization of three Gd clusters based on the nicotinate (nic) ligand, Gd₂(nic)₆(H₂O)₄ (1), [Gd₄(μ₃-OH)₄(Hnic)₅(H₂O)₁₂](ClO₄)₈·7H₂O (2)³⁷ and [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₆]₂(ClO₄)₄·4H₂O (3). We further examined their MR relaxivities using a 7 T scanner, and studied the influences of the number of Gd ions and coordinated water molecules as well as the size of the clusters on their MR relaxivities (Fig. 1).

Fig. 1 Molecular parameters that influence relaxivity: rotation (τ_R), water exchange (τ_M) and hydration number (q).

Experimental

Materials and general methods

Nicotinic acid, nitric acid, gadolinium oxide, perchloric acid, sodium hydroxide, and gadolinium oxide were obtained from commercial sources. Other solvents were of reagent grade and used without further purification. Elemental analyses were performed on an Elemental Vario EL analyzer. Thermogravimetric analysis (TGA) results were obtained on a SDT Q600 Thermal Analyzer. HR-MS spectra were recorded on a Bruker En Apex ultra 7.0T FT-MS apparatus. X-Ray powder diffraction patterns were obtained on a Rigaku Ultima IV XRD.

Synthesis of Gd₂(nic)₆(H₂O)₄ (1)

The cluster 1 was prepared by a method that was modified from the procedure reported by Hong et al.³⁸ 10 mL of H₂O was added to a mixture of GdCl₃·6H₂O (750 mg, 2 mmol) and nicotinic acid (490 mg, 4 mmol). 650 mg of NaN₃ (10.0 mmol) was added to the solution, followed by the addition of 300 μL of 2 M HNO₃. The mixture was transferred to a Teflon-lined high pressure reactor and then heated at 180 °C for three days. The reactor was slowly cooled at 1 °C per 10 min to room temperature, and the mixture was filtered. Colorless crystals of 1 were obtained after the solution was allowed to stand at room temperature for two days. Yield: 0.1 g (13.4%, based on nicotinic acid). Anal. calcd for C₃₆H₃₂Gd₂N₆O₁₆: C, 38.6%; H, 2.86%; N, 7.50%. Found: C, 37.43%; H, 3.14%; N, 7.27%. IR (KBr): 3505 (m), 1645 (s), 1597 (s), 1573 (w), 1547 (m), 1476 (m), 1433 (s), 1405 (s), 1195 (m), 1166 (w), 1114 (w), 1093 (m), 1028 (m), 976 (w), 864 (m), 853 (m), 759 (vs.), 702 (vs.), 634 (m), 566 (m), 567 (m) cm⁻¹. F.W. = 1119.18.

Synthesis of [Gd₄(μ₃-OH)₄(Hnic)₅(H₂O)₁₂](ClO₄)₈·7H₂O (2)

The cluster 2 was synthesized in the same way as that reported by Kong et al.³⁹ Block-shaped colorless crystals of 2 were obtained. Yield: 0.2 g (8.2% based on nicotinic acid). Anal. calcd for C₃₀Cl₈H₆₇Gd₄N₅O₆₅: C, 14.28%; H, 2.34%; N, 2.53%. Found: C, 14.47%; H, 2.22%; N, 2.91%. IR (KBr): 3388 (m), 2861 (w), 2025 (w), 1643 (s), 1593 (s), 1417 (s), 1352 (w), 1305 (w), 1089 (s), 940 (w), 835 (w), 750 (s), 688 (s), 672 (m), 627 (s), 560 (m) cm⁻¹. F.W. = 2450.47.

Synthesis of [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₆]₂(ClO₄)₄ ·4H₂O (3)

The cluster 3 was synthesized by a method that was modified from the procedure reported by Kong et al.³⁹ Nicotinic acid (0.62 g, 5 mmol) in 5 mL of deionized water was titrated with NaOH (1.0 M) to reach a pH of 8, which was then added dropwise to the solution of 10 mL of Gd(ClO₄)₃ (1.0 M) that have been dropwise added a fresh NaOH (1.0 M) solution to the point of incipient but permanent precipitation at 90 °C. After stirring at 90 °C for one additional hour, the resulting white epinephelos mixture was quickly filtered while hot. The mixture was allowed to stand at room temperature for two days to afford block-shaped colorless crystals. Yield: 0.13 g (8.8% based on nicotinic acid). Anal. calcd for C₇₂H₈₈Gd₈Cl₄N₁₂O₆₄: C, 24.4%; H, 2.48%; N, 4.74%. Found: C, 24.90%; H, 2.32%; N, 4.84%. IR (KBr): 3388 (m), 1617 (s), 1566 (s), 1476 (w), 1409 (s), 1195 (m), 1158 (m), 1097 (m), 1030 (m), 973 (w), 842 (m), 759 (s), 698 (vs.), 636 (m), 563 (w) cm⁻¹. F.W. = 3545.34.

X-Ray crystal structure determination

The X-ray diffraction data of 1 and 3 were collected on Agilent Supernova with Mo-Kα radiation (λ = 0.71073 Å) at 100 K. Absorption corrections were applied by using the multiscan program CrysAlis Red. The structures were solved by direct and Fourier methods and refined by full-matrix least squares based on F₂ using the SHELX and Olex2 software. Detailed crystal data and refinement parameters are listed in Table S1.†

Measurement of r₁ relaxivities

To measure the r₁ relaxivities of 1–3, samples with different Gd(III) concentrations were dispersed in Mill-Q water or in 1% agarose solution. The Gd(III) concentrations were measured by inductively coupled plasma-mass spectrometry (ICP-MS). The samples were scanned (at 300 K) on both 0.5 T and 7 T MRI scanners. T₁-weighted MRI phantom images of the three samples are performed on a 7 T MRI scanner.

Results and discussion

Synthesis

Cluster 1 was synthesized in 13.4% yield by first hydrothermally treating a mixture of nicotinic acid, GdCl₃, and NaN₃ in the presence of a small amount of HNO₃ followed by crystallization at room temperature over a period of two days.³⁸ Clusters 2 and 3 were prepared in low yield (about 8%) by first heating appropriate mixtures at 90 °C, followed by crystallization at room temperature. Such a ligand-controlled hydrolytic approach has been previously established for Gd cluster synthesis.^40,41 The structures and compositions of 1–3 have been established by single-crystal X-ray crystallography, powder X-ray diffraction, and elemental analysis. Cluster 2 was prepared by a one-pot synthesis method whereby an aqueous solution of NaOH was added dropwise into the mixture containing both the nicotinic acid ligand and the Gd salt. In contrast, cluster 3 was obtained by a two-step procedure. The pH of the ligand aqueous solution and the lanthanide perchlorate was adjusted individually to appropriate values before they were mixed. The one-pot synthesis produced the tetranuclear cluster of [Gd₄(μ₃-OH)₄(Hnic)₅(H₂O)₁₂](ClO₄)₈·7H₂O, while the two-step synthesis afforded [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₆]₂(ClO₄)₄·4H₂O. Cluster 3 slightly differs from [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₈]₂(ClO₄)₄·3H₂O which was previously described by Long et al. The pyridyl nitrogen atoms of the nic ligands in cluster 2 was protonated, while the nic ligands in cluster 3 were completely deprotonated.

X-Ray single crystal structures

Single-crystal X-ray diffraction analyses reveal that 1 is a neutral dinuclear compound that crystallizes in the monoclinic space group P2₁/c. In 1, each Gd³⁺ ion is coordinated to eight oxygen atoms (Fig. 2), with two oxygen atoms coming from a chelating nicotinic acid ligand, four oxygen atoms coming from four bridging nicotinic acid ligands, and two oxygen atoms from two water molecules.⁴² Each Gd ion is thus coordinated to two water molecules in 1. In addition, there are two crystallization water molecules for each molecule of 1 in the crystals.

Fig. 2 X-Ray single-crystal structure of [Gd₂(nic)₆ (H₂O)₄] (1) with thermal ellipsoids at the 50% probability level. Hydrogen atoms are removed for clarity. Each Gd(III) ion has two coordinated water molecules.

The structure of 2 was previously reported by Kong et al.³⁷ Briefly, the [Gd₄(μ₃-OH)₄]⁸⁺ cationic core adopts a distorted cubane structure (Fig. 3). Two Gd ions (Gd₁ and Gd₂) have two coordinated water molecules whereas two other Gd ions (Gd₃ and Gd₄) have four coordinated water molecules. All of the pyridyl nitrogen atoms of the nicotinate ligands in 2 are protonated to maintain charge balance.

Fig. 3 X-Ray single-crystal structure of [Gd₄(μ₃-OH)₄(Hnic)₅(H₂O)₁₂](ClO₄)₈·7H₂O (2) with thermal ellipsoids at the 50% probability level. Perchlorate anions, crystallization water molecules, and hydrogen atoms are removed for clarity.

The molecular structure of 3 consists of a dimer of [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₆]²⁺ as shown in Fig. 4. There is an inversion center residing at the center of the dimer. The Gd₄(μ₃-OH)₄ cubane core is encapsulated by six nicotinate ligands. These ligands display two different coordination modes: five of them use their carboxylate groups to bridge two Gd³⁺ ions whereas the sixth one uses its carboxylate group to bridge two Gd³⁺ ions of one Gd₄(μ₃-OH)₄ cubane core and uses its pyridyl N atom to coordinate a Gd³⁺ ion in the other Gd₄(μ₃-OH)₄ cubane core (as shown for Gd₄ and Gd₃ in Fig. 4).

Fig. 4 X-Ray single-crystal structure of [Gd₄ (μ₃-OH)₄(nic)₆(H₂O)₆]₂(ClO₄)₄·4H₂O (3) with thermal ellipsoids at the 50% probability level. Perchlorate anions and hydrogen atoms are removed for clarity.

Gd₁, Gd₂, and Gd₃ centers in 3 are thus 8-coordinate, by binding to three oxygen atoms from the bridging nicotinate ligands, three μ₃-OH groups, and two water molecules. The Gd₄ center is seven-coordinate and binds to three oxygen atoms from the bridging nicotinate ligands, three μ₃-OH groups, and one pyridyl atom of the bridging nicotinate ligand. The structure of 3 is related to, but distinct from an octanuclear Gd complex with the formula of [Gd₄(μ₃-OH)₄(nic)₆(H₂O)₈]₂(ClO₄)₄·3H₂O that was reported earlier.³⁷

Characterization of clusters 1–3 by PXRD, TGA and ESI-MS

We have used PXRD to verify the phase purity of the clusters. As shown in Fig. 5a–c, the experimental PXRD patterns of 1–3 generally match the PXRD patterns that were simulated from their single crystal structures well, indicating that the bulk phase of these clusters are identical to those of the single crystals chosen for X-ray diffraction studies. We note a slight peak shift in PXRD pattern of 3 from the simulated pattern, which can be attributed to different temperature between the single crystal X-ray and PXRD experiment as well as possible partial loss of lattice solvent molecules in the powdery samples for PXRD. Indexing the PXRD pattern of 3 gives similar lattice parameters to those determined from single crystal data (Table S3 and Fig. S7†).

Fig. 5 (a–c) PXRD patterns of cluster 1 (a); cluster 2 (b), and cluster 3 (c). PXRD pattern of the as-synthesized sample (black), simulated from single-crystal structure (red). (d) TGA curves for cluster 1 (black); cluster 2 (red); cluster 3 (blue).

Thermogravimetric analyses (TGA) were also performed on clusters 1–3. As shown in Fig. 5d, the TGA curves for 1–3 display an initial weight loss 6.4% (calcd 7.9%) for 1, 12.5% (calcd 13.9%) for 2, and 8.1% (calcd 9.0%) for 3 at the room temperature to 300 °C temperature range, which correspond to the loss of guest and coordinated water molecules. When the temperature is higher than 300 °C, clusters of 1–3 were rapidly decomposed to form Gd₂O₃ with remaining weight of 32.5% (calcd 32.4%) for 1, 30.1% (calcd 29.6%) for 2, and 40.0% (calcd 40.9%) for 3 at 900 °C.

We have also tried to characterize the clusters using electrospray ionization-mass spectrometry to ensure that the clusters remain intact when they are dissolved in aqueous solutions. As shown in Fig. S1,† the ESI-MS spectrum (positive ion mode) of 1 gave the most intense peak at m/z = 1049.00, which corresponding to the fragment of [Gd₂(nic)₆ + H]⁺. This assignment was further confirmed by the isotopic distribution pattern of the [Gd₂(nic)₆ + H]⁺ fragment. As shown in Fig. 6a and b, the experimental and theoretical isotopic distribution patterns match extremely well.

Fig. 6 The measured mass spectrum (a) and simulated (b) isotopic distributions of 1 (the shorter isotope peaks at 1048.50 m/z belong to the [Gd₄(nic)₁₂+2H] ²⁺ which might have formed during the ionization events of the ESI-MS experiments).

The ESI-MS spectra of 2 and 3 indicate that the clusters have undergone extensive fragmentation during the electrospray ionization process. We have not been able to unambiguously assign the major peaks in their ESI-MS spectra. Instead, we resorted to dynamic light scattering (DLS) to probe the species resulted from the dissolution of 2 and 3 in aqueous solutions.

As shown in Fig. S3 and S4,† DLS data indicate the presence of hydrodynamic diameter peaks centering at 0.9 nm and 1.8 nm for 2 and 3, respectively. These DLS hydrodynamic diameters are consistent with the molecular dimensions of 2 and 3: the short and long dimensions of 2 are 0.6 and 1.7 nm, respectively, and the short and long dimensions of 3 are 0.9 and 2.7 nm, respectively. The DLS data thus support the notion that clusters 2 and 3 remain intact upon dissolution in aqueous solutions.

r₁ Measurements and T₁-weighted MRI phantom images

With seven unpaired electrons in each Gd ion, we expect that the synthesized Gd clusters to exhibit strong ability to relax water protons. We determined the longitudinal relaxivity (r₁) values of 1–3 at both 0.5 T and 7 T. Each type of the clusters was prepared by dissolving them in 1% agarose solution with total Gd concentrations of 0.025, 0.1, 0.2, and 0.4 mM. As shown in Fig. 7a, the T₁-weighted signal intensities in a 7 T scanner increase as the Gd concentration of the clusters increase, indicating that all of the clusters have the ability to enhance MRI contrasts with T₁-weighted pulse sequences. At 7 T, the r₁ values for 1–3 are 10.45 ± 0.16, 9.28 ± 0.06, and 2.04 ± 0.29 mM⁻¹ s⁻¹ in water through the linear fit in Fig. 7b, respectively. In 1% agarose solution, the r₁ values increased slightly to 10.37 ± 0.31, 10.74 ± 0.11, and 4.60 ± 0.29 mM⁻¹ s⁻¹ for 1–3, respectively. The r₁ values of the cluster 1 and 2 in 1% agarose are significantly higher than that of Gd–DTPA (r₁ = 3.30 ± 0.25 mM⁻¹ s⁻¹ at 7 T) which we have determined under the same conditions. As expected, the r₁ values for 1–3 decrease as the field strength increases; at 0.5 T, the r₁ values are 11.00 ± 0.25, 10.08 ± 0.22, and 2.09 ± 0.32 mM⁻¹ s⁻¹ for 1–3 in water, respectively (Fig. S5†). The two Gd clusters of the 1 and 2 show high r₁ relaxivities at 7 T, and they exhibit interesting relaxivity trends due to their different coordination environments. As described in the Solomon–Bloembergen–Morgan equations, the key factors contributing to the inner-sphere relaxivity include the rotational tumbling time (τ_R), the number of inner-sphere water molecules (q), and the residence lifetime of inner-sphere water molecules (τ_M). A larger relaxivity value will result when inner-sphere water have greater access to the paramagnetic metal center (i.e., shorter τ_M) and the inner-sphere relaxivity is linearly proportional to the number of inner-sphere water molecules (q).⁴³ Since the molecular weights of the cluster ions range from 1119 to 3075 for 1–3, their tumbling rates in water should be relatively fast. We do not expect that the differences in tumbling rates of these clusters will significantly influence their r₁ values. Instead, the number of coordinated water molecules and the steric hindrance around the water coordination sites for 1–3 should have significant impacts on their r₁ values. Cluster 2 has the largest average number of coordinated water molecules per Gd (q_ave = 3), whereas cluster 3 has the smallest average number of coordinated water molecules per Gd (q_ave = 1.5). As a result, the r₁ value for 2 is much higher than that of 3. Although 2 has a larger average number of coordinated water molecules per Gd than 1 (q_ave = 2), their r₁ values are essentially identical. X-Ray structure comparisons indicate that the water coordination sites in 1 are more open than those in 2, and as a result, the τ_M value for 2 should be smaller, which compensate for its smaller q_ave value.

Fig. 7 The phantom study (a) and relaxivity plots (b) of the three clusters performed on a 7 T MRI scanner.

Conclusions

We have prepared and characterized three gadolinium nicotinate clusters of different sizes. These Gd clusters of the cluster 1 and 2 have very high relaxivities at 7 T when compared to the clinically used Gd–DTPA. Because of the ability to tune the number of inner-sphere water molecules, cluster sizes, organic ligands, and Gd coordination modes in such Gd(III) clusters, we believe that Gd clusters provide interesting opportunities for designing new multinuclear MRI contrast agents for potential biological and biomedical imaging.

Acknowledgements

The work was supported by the National Natural Science Foundation of P. R. China (21471126), the National Thousand Talents Program of P.R. China, and the 985 Program of the Chemistry and Chemical Engineering disciplines of Xiamen University. We acknowledge Ms Zhiwei Lin for the FT-MS measurement and Ms Ruiyun Huang for administrative help.

Notes and references

1. X. Zhang, X. Jing, T. Liu, G. Han, H. Li and C. Duan, Inorg. Chem., 2012, 51, 2325.
2. L. N. Goswami, L. Ma, S. Chakravarty, Q. Cai, S. S. Jalisatgi and M. F. Hawthorne, Inorg. Chem., 2013, 52, 1694.
3. T. Courant, V. G. Roullin, C. Cadiou, M. Callewaert, M. C. Andry, C. Portefaix, C. Hoeffel, M. C. de Goltstein, M. Port, S. Laurent, L. V. Elst, R. Muller, M. Molinari and F. Chuburu, Angew. Chem., Int. Ed., 2012, 51, 9119.
4. E. J. Werner, A. Datta, C. J. Jocher and K. N. Raymond, Angew. Chem., Int. Ed., 2008, 47, 8568.
5. H. Gries, U. Speck, H.-J. Weinmann, H. P. Niendorf and W. Seifert, DE3640708A1, 1988.
6. C. T. Williams, J. P. Stack, B. Loveday, Y. Watson and I. Isherwood, Br. J. Radiol., 1988, 61, 596.
7. D. T. Puerta, M. Botta, C. J. Jocher, E. J. Werner, S. Avedano, K. N. Raymond and S. M. Cohen, J. Am. Chem. Soc., 2006, 128, 2222.
8. P. Caravan, J. J. Ellison, T. J. McMurry and R. B. Lauffer, Chem. Rev., 1999, 99, 2293.
9. Z. Zhang, A. F. Kolodziej, M. T. Greenfield and P. Caravan, Angew. Chem., Int. Ed., 2011, 50, 2621.
10. J. Xu, D. G. Churchill, M. Botta and K. N. Raymond, Inorg. Chem., 2004, 43, 5492.
11. H. Ersoy and F. J. Rybicki, J. Magn. Reson. Imaging, 2007, 26, 1190.
12. E. Boros, M. Polasek, Z. Zhang and P. Caravan, J. Am. Chem. Soc., 2012, 134, 19858.
13. P. Caravan, G. Parigi, J. M. Chasse, N. J. Cloutier, J. J. Ellison, R. B. Lauffer, C. Luchinat, S. A. McDermid, M. Spiller and T. J. McMurry, Inorg. Chem., 2007, 46, 6632.
14. E. Garanger, S. A. Hilderbrand, J. T. Blois, D. E. Sosnovik, R. Weissleder and L. Josephson, Chem. Commun., 2009, 4444.
15. K. B. Ghiassi, M. M. Olmstead and A. L. Balch, Dalton Trans., 2014, 43, 7346.
16. M. Mikawa, H. Kato, M. Okumura, M. Narazaki, Y. Kanazawa, N. Miwa and H. Shinohara, Bioconjugate Chem., 2001, 12, 510.
17. M. L. Viger, J. Sankaranarayanan, C. de Gracia Lux, M. Chan and A. Almutairi, J. Am. Chem. Soc., 2013, 135, 7847.
18. A.-A. Guay-Bégin, P. Chevallier, L. Faucher, S. Turgeon and M.-A. Fortin, Langmuir, 2011, 28, 774.
19. K. M. L. Taylor, A. Jin and W. Lin, Angew. Chem., Int. Ed., 2008, 47, 7722.
20. J. Della Rocca, D. Liu and W. Lin, Acc. Chem. Res., 2011, 44, 957.
21. M. D. Rowe, C.-C. Chang, D. H. Thamm, S. L. Kraft, J. F. Harmon, A. P. Vogt, B. S. Sumerlin and S. G. Boyes, Langmuir, 2009, 25, 9487.
22. W. Hatakeyama, T. J. Sanchez, M. D. Rowe, N. J. Serkova, M. W. Liberatore and S. G. Boyes, ACS Appl. Mater. Interfaces, 2011, 3, 1502.
23. M. W. Cooke and G. S. Hanan, Chem. Soc. Rev., 2007, 36, 1466.
24. X.-J. Kong, Y. Wu, L.-S. Long, L.-S. Zheng and Z. Zheng, J. Am. Chem. Soc., 2009, 131, 6918.
25. C. He, X. Wu, J. Kong, T. liu, X. Zhang and C. Duan, Chem. Commun., 2012, 48, 9290.
26. D. T. Thielemann, A. T. Wagner, E. Rösch, D. K. Kölmel, J. G. Heck, B. Rudat, M. Neumaier, C. Feldmann, U. Schepers, S. Bräse and P. W. Roesky, J. Am. Chem. Soc., 2013, 135, 7454.
27. Y.-L. Hou, G. Xiong, P.-F. Shi, R.-R. Cheng, J.-Z. Cui and B. Zhao, Chem. Commun., 2013, 49, 6066.
28. L.-X. Chang, G. Xiong, L. Wang, P. Cheng and B. Zhao, Chem. Commun., 2013, 49, 1055.
29. S. K. Langley, B. Moubaraki and K. S. Murray, Inorg. Chem., 2012, 51, 3947.
30. J. Bai, A. V. Virovets and M. Scheer, Science, 2003, 300, 781.
31. T. C. Stamatatos, K. A. Abboud, W. Wernsdorfer and G. Christou, Angew. Chem., Int. Ed., 2007, 46, 884.
32. H.-H. Zou, R. Wang, Z.-L. Chen, D.-C. Liu and F.-P. Liang, Dalton Trans., 2014, 43, 2581.
33. N. M. Randell, M. U. Anwar, M. W. Drover, L. N. Dawe and L. K. Thompson, Inorg. Chem., 2013, 52, 6731.
34. J.-M. Jia, S.-J. Liu, Y. Cui, S.-D. Han, T.-L. Hu and X.-H. Bu, Cryst. Growth Des., 2013, 13, 4631.
35. F.-S. Guo, Y.-C. Chen, L.-L. Mao, W.-Q. Lin, J.-D. Leng, R. Tarasenko, M. Orendáč, J. Prokleška, V. Sechovský and M.-L. Tong, Chem.–Eur. J., 2013, 19, 14876.
36. B. Kim, A. H. Schmieder, A. J. Stacy, T. A. Williams and D. Pan, J. Am. Chem. Soc., 2012, 134, 10377.
37. X.-J. Kong, L.-S. Long, L.-S. Zheng, R. Wang and Z. Zheng, Inorg. Chem., 2009, 48, 3268.
38. M. Wu, F. Jiang, X. Kong, D. Yuan, L. Long, S. A. Al-Thabaiti and M. Hong, Chem. Sci., 2013, 4, 3104.
39. X.-J. Kong, L.-S. Long, L.-S. Zheng, R. Wang and Z. Zheng, Inorg. Chem., 2009, 48, 3268.
40. R. Wang, Z. Zheng, T. Jin and R. J. Staples, Angew. Chem., Int. Ed., 1999, 38, 1813.
41. R. Wang, H. Liu, M. D. Carducci, T. Jin, C. Zheng and Z. Zheng, Inorg. Chem., 2001, 40, 2743.
42. W. Chen and S. Fukuzumi, Inorg. Chem., 2009, 48, 3800.
43. E. L. Que and C. J. Chang, Chem. Soc. Rev., 2010, 39, 51.

---

Footnote

† Electronic supplementary information (ESI) available: X-Ray crystallography, electrospray mass spectrum, IR spectra and DLS spectra. CCDC 1014973 and 1014974. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra07853e

---