Functionalized phosphonates as building units for multi-dimensional homo- and heterometallic 3d–4f inorganic–organic hybrid-materials

Abstract

Using the multifunctional ligand H₄L (2,2′-bipyridinyl-5,5′-diphosphonic acid), a new family of inorganic–organic hybrid-materials was prepared. The ligand shows a very high flexibility regarding the coordination mode, leading to a large structural diversity. The compounds 1a, 1b ([M(H₂L)(H₂O)₄]·2.5H₂O; M = Co²⁺ (a), Ni²⁺ (b)), 2 ([Gd₂(H₂H′L)₂(H₂H′₂L)(H₂O)₆]Cl₄·14H₂O), 3a, 3b, 3c ([MCo(III)(H₂L)₃(H₂O)₂]·6.5H₂O; M = Gd³⁺ (a), Dy³⁺ (b) and Tb³⁺ (c)), and 4 ([GdNi(II)(H₂L)₃(H₂O)₃]NaCl·6H₂O) were isolated and characterized with single crystal X-ray diffraction. Depending on the used metal ions and on the stoichiometry, either discrete entities (0D), extended 2D layers or 3D frameworks were obtained. In contrast to the general approach in phosphonate chemistry, the compounds were prepared without hydrothermal synthesis, but under ambient pressure. Variable temperature magnetic measurements were carried out to determine the magnetic properties.

---

Introduction

The field of inorganic–organic hybrid-materials^1,2 shows an enormous potential in a broad spectrum of applications, such as adsorption/desorption, catalysis, functionalization of surfaces, nanomaterials or photoluminescent devices.^3–12 Additionally, the incorporation of open-shell transition metal ions or rare earth elements into these hybrids yielded a large number of magnetic materials, which exhibit several highly relevant properties from the field of molecular magnetism, e.g. long range magnetic ordering, single chain magnetism, large magnetocaloric effect or slow relaxation of magnetisation.^13–21

Whereas the predominant type of inorganic–organic hybrid-materials is of homometallic nature,^22,23 the first heterometallic systems by Gatteschi et al. date back to 1985.²⁴ However, they are not as extensively studied as homometallic compounds, mostly due to their much smaller number.²⁵ A reason for the lack of heterometallic complexes is the much larger effort their synthesis bear. The challenge in the design of heterometallic compounds originates in the competition of different metal ions for the same or similar coordination sites.²⁶ One approach to overcome this problem is to make use of the desire of different sorts of metal ions for different coordination environments.^27,28

We followed this strategy by employing a bipyridine based ligand, which contains additionally two phosphonate functionalities. While the bipyridine site shows a well-documented readiness for the coordination of transition metal ions,²⁹ the phosphonate groups offer perfect coordination condition for lanthanide ions due to the strong oxophilicity of the rare earth elements.^30–33 And although the phosphonate group is known for its highly divers modes of coordination, it's coordination can be fine-tuned by altering its charge by changing the pH value, making it very valuable for the task.³⁴

We therefore employed the ligand 2,2′-bipyridinyl-5,5′-diphosphonic acid (H₄L or H₃H′L, see Scheme 1), first synthesized in 1998 by Penicaud et al.³⁵ It contains both functionalities mentioned above, which show a certain distinction in their coordination capabilities. To the best of our knowledge, there is no report of any complex in which the phosphonate group and the bipyridine pocket coordinate at the same time. So far the ligand has solely been used for the preparation of diamagnetic Ru(II) and Zr(IV) complexes, which, however, have not been characterized by single crystal diffraction.^36–38

Scheme 1 Ligand H₄L or H₃H′L.

Herein we report the synthesis and full characterisation of seven novel compounds incorporating 3d transition metal ions and 4f elements. We were able to prepare 0D and 2D homometallic structures as well as heterometallic 2D and 3D structures with Co, Ni, Gd, Tb and Dy. We obtained single crystal diffraction data for all compounds (Table 1) and investigated the magnetic properties of all the more-dimensional complexes.

Table 1 Crystallographic data for compounds 2–4. Measurements were conducted at 173 K

	2	3a	3b	3c	4
Empirical formula	C30H42Cl4Gd2N6O38P6	C30H27.5CoGdN6O26.5P6	C30H25CoDyN6O26.5P6	C30H27CoN6O26P6Tb	C30H22ClGdN6NaNiO27P6
Formula weight	1736.81	1298.08	1300.81	1291.24	1358.75
Crystal system	Triclinic	Monoclinic	Monoclinic	Monoclinic	Monoclinic
Space group	P1̄	C2/c	C2/c	C2/c	C2
a/Å	10.362(3)	25.1648(18)	25.1186(12)	25.203(2)	22.6611(11)
b/Å	10.819(3)	17.1628(12)	17.0759(9)	17.0993(14)	16.9747(8)
c/Å	14.686(4)	23.0284(18)	22.9551(14)	22.983(2)	15.2845(7)
α/°	104.138(7)	90	90	90	90
β/°	100.605(7)	116.436(2)	116.0600(10)	116.064(2)	106.794(2)
γ/°	98.755(7)	90	90	90	90
V/Å^3	1535.5(7)	8905.9(11)	8845.0(8)	8897.4(13)	5628.7(5)
Z	1	8	8	8	4
ρ calc/g cm^−3	1.878	1.936	1.954	1.928	1.603
μ/mm^−1	2.572	2.166	2.371	2.265	1.811
Θ/°	3.968–55.91	2.982–55.892	2.99–55.906	3.946–55.91	3.046–55.7
F(000)	854	5140	5136	5112	2680
Data/restraints/parameters	7328/6/406	10 663/54/726	10 611/96/752	10 680/42/688	13 356/115/792
GOF (F^2)	1.032	0.975	0.893	1.032	0.947
R 1, wR2 (I ≥ 2σ(I))	0.0577, 0.1433	0.0470, 0.1159	0.0524, 0.1029	0.0690, 0.1679	0.0678, 0.1582
R 1, wR2 (all data)	0.0879, 0.1663	0.0718, 0.1263	0.0996, 0.1163	0.1132, 0.1933	0.1152, 0.1783

---

Results and discussion

Synthesis

One of the most common and applied methods in phosphonate chemistry is hydrothermal synthesis.^30,39 Unfortunately, these reaction conditions involve the loss of selectivity towards the 4f metals. We decided therefore to employ a different strategy and induced a growth of crystals by either changing the polarity of the solvent or else evaporation of the solvent at elevated temperatures to form the thermodynamically more stable compound. The outcome of the reaction mixture depends on three factors: the ligand/metal ratio, the pH-value and the order of the added metal salts. By varying them, we could observe different connectivity and had hence influence onto the dimensionality up to a certain degree.

Crystal structures

Compound 1a and 1b ([M(H₂L)(H₂O)₄]·2.5H₂O; M = Co²⁺ (a), Ni²⁺ (b)) are isostructural and comprise one metal centre, one twofold deprotonated ligand (H₂L)²⁻ and 6.5 water molecules. The detailed crystal structure can be found in the ESI.† In both cases, the central ion shows an octahedral coordination geometry in which four of the water molecules are included. As expected, the ligand coordinates in an η₂-coordination mode. Despite the large number of possible coordination sites, the complexes are of molecular structure. This emphasizes our earlier statement, that transition metal–phosphonate bonds are rather formed during hydrothermal synthesis, at least under moderate pH conditions. The nitrogen–metal bond lengths of 2.123 Å and 2.139 Å for 1a and the oxygen–phosphonate bond lengths within the phosphonates indicate an oxidation state of +2 for the metal ions.⁴⁰ The positive charge is compensated by the single deprotonated phosphonate groups. Individual complexes are connected via hydrogen bonds mediated by water molecules (see Fig. S3 in ESI†).

Compound 2 ([Gd₂(H₂H′L)₂(H₂H′₂L)(H₂O)₆]Cl₄]·14H₂O) forms a two dimensional network. The structure was solved in the space group P1̄. The asymmetric unit contains 1.5 ligand molecules with different grades of protonation, one eightfold coordinated gadolinium(III)-ion, ten water molecules from which three are coordinating and two chloride counter ions. A comparison of the oxygen–phosphorus bond lengths reveals that they are either around 1.50 Å or else 1.55 Å. The latter one is the typical bond length for a protonated oxygen atom. Therefore, all phosphonate groups are assumed to be only single deprotonated. Additionally, the nitrogen atoms of the split ligand (H₂H′₂L) are protonated, whereas (H₂H′L)⁻ shows only one protonated nitrogen, leaving it with a negative charge. Both types of ligands differ as well in their coordination mode. The phosphonate of H₂H₂′L and the phosphonate of (H₂H′L)⁻ which is closer to the protonated nitrogen atom are each bridging two lanthanides with two oxygen in a μ₂-mode.

The remaining phosphonate coordinates to one gadolinium ion via one oxygen atom, which gives in total five metal–phosphonate connections. The remaining three coordination sites of the square antiprismatic sphere are occupied by water molecules. Due to the μ₂-briding mode of the individual phosphonate groups, the central ions show a one dimensional I¹ chainlike structure along the a-axis, as shown in Fig. 1(c), with a mean distance between two lanthanide ions of 5.212 Å. In Fig. 1(b), the O¹ connection by the bipyridine acting as a tether along the c-axis is depicted. The spacing between the chains is on average 11.504 Å. The layers are connected via hydrogen bonds forming between the counter ions along the b-direction.

Fig. 1 (a) Primitive cell of compound 2 with broken off bonds. (b) Top view shows the O¹ tethering of the bipyridine backbone. (c) Graphical representation of the I¹ gadolinium chains. Colour code: brown – gadolinium, pink – phosphorus, red – oxygen, dark blue – nitrogen, grey – carbon. Hydrogen atoms and solvent molecules have been omitted for clarity.

Compounds 3a, 3b and 3c ([MCo(III)(H₂L)₃(H₂O)₂]·6.5H₂O; M = Gd³⁺ (a), Dy³⁺ (b) and Tb³⁺ (c), Fig. 2(a)) are two-dimensional heterometallic coordination polymers. The structures were solved in the space group C2/c. All three complexes are isostructural and differ only in their 4f element. The lanthanides are coordinated in a for lanthanides rather rare sevenfold coordinated capped trigonal prismatic geometry.⁴¹ As expected, the coordination sphere is formed by seven oxygen atoms belonging to different groups. Five of the oxygen atoms belong to five separate phosphonate groups, the remaining two origin from water molecules, leaving one non-coordinating phosphonate group. Similar to compound 2, all phosphonates are single protonated as indicated by the P–O bond lengths. One trigonal plane and the cap of the coordination polyhedral is built from four phosphonate oxygen atoms, while the other plane includes both water molecules and one ligand molecule. A comparison of the oxygen–lanthanide bond lengths shows, that the average distance to the water molecules are quiet shorter than the bonds to the ligand (2.291 Å and 2.451 Å, respectively). The second metal is chelated in an η₂-coordination mode comparable to compound 1a and 1b, but by three bipyridine coordination sites. They form (Co(H₂L)₃)³⁻ building units which connect the incorporated lanthanide ions. The N–Co bond lengths differ significantly from the lengths in the previously described complexes. They range from 1.916(8) Å to 1.950(9) Å with a mean value of 1.933 Å. The origin is found in the oxidation state of the cobalt ion, which oxidizes during the reaction from Co(II) to Co(III). The small differences in bond lengths and the cis-angles being close to 90° indicate a good approximation to an ideal octahedral coordination sphere. Fig. 2(b) represents a side view along the b-axis of the two-dimensional layers. It is clearly visible, that only five out of six phosphonate groups are coordinating. The remaining group protrudes above and beneath the layer, connecting it by hydrogen bonds between the phosphonate and crystal water to neighbouring arrays. A closer examination of the arrangement within the layer in Fig. 2(c) shows, that the cobalt octahedron form some sort of one-dimensional chain along the b-axis. The distance between two terbium polyhedron is 5.514 Å, which is comparably close to the chain in compound 2. However, they are connected by hydrogen bonds between the phosphonates and show therefore no direct coordinative or covalent contact.

Fig. 2 (a) Asymmetric unit of compound 3c with broken off bonds. (b) Side view of a layer illustrating the non-coordinating phosphonate groups. (c) Top view of the crystal structure showing the cobalt and terbium chains. Colour code: brown – gadolinium, dark red – cobalt, pink – phosphorus, red – oxygen, dark blue – nitrogen, grey – carbon. Hydrogen atoms and solvent molecules have been omitted for clarity.

Additionally, the arrangement of the metal centres should rather be described as dimers then chains due to their discontinuous nature: the spacing to the next terbium in the opposite direction is with 11.885 Å much longer.

The asymmetric unit of compound 4 (Fig. 3(a)) has the sum formula [GdNi(II)(H₂L)₃(H₂O)₃]NaCl·6H₂O. The crystal structure can be solved in the space group C2 and forms a three-dimensional coordination polymer by the μ₂-bridging of the gadolinium centres via the ligands. The lanthanide ion has, similar to compounds 3a, 3b and 3c, a coordination number of seven and is in a capped trigonal prismatic coordination environment. But in contrast to the three cobalt complexes, the nickel compound shows three lanthanide–water contacts and only four lanthanide–phosphonate connections, leaving two non-coordinating phosphonate groups.

Fig. 3 (a) Cation of the asymmetric unit of compound 4 with broken off bonds. (b) View along the b-axis showing the connection of the crooked layers by the Ni(H₂L)₃ units. (c) The view along the c-axis reveals the three-dimensional structure of the compound and the oval motive it is forming. Colour code: brown polyhedron – gadolinium, light blue polyhedron – nickel, pink polyhedron – phosphorus, rose – sodium, red – oxygen, dark blue – nitrogen, grey – carbon. Hydrogen atoms and solvent molecules have been omitted for clarity.

The different connectivity of the phosphonates arises from the different initial pH value due to the use of the chloride metal salt instead of the acetate metal salt and the larger amount of hydrochloric acid which was employed to dissolve the precipitate. As a consequence, the number of phosphonate–metal contacts is reduced. Both trigonal planes of the capped trigonal prism equally consist of two phosphonate groups and one water molecule. The cap is built from a water molecule as well. Again, the nickel ions are coordinated octahedral by the bipyridine coordination sites of the ligand and form building blocks, which connect the gadolinium ions. Due to the different coordination mode of the phosphonates, the motive of a paired chain as in 3a–3c is not present in this compound. Instead, the two different metal ions Ni²⁺ and Gd³⁺ occur alternating. In Fig. 3(c), a view along the c-axis of the crystal structure is shown. It is a representative picture of the three-dimensional structure of this compound and shows the oval pores the structure is forming in this direction. The representation reveals, that the sodium ions are occupying the space within the pores.

Magnetic properties

Magnetic measurements at variable temperatures have been carried out for all more-dimensional compounds. All measurements regarding χ_MT where conducted under an applied field of 1 kOe in a temperature range between 2 K and 300 K. For compound 2, a χ_MT value of 7.76 cm³ K mol⁻¹ was observed at room temperature (Fig. 4), which is in good agreement with the expected values for two uncoupled Gd³⁺ ions (7.88 cm³ K mol⁻¹, ⁸S_7/2, S = 7/2, L = 0, g = 2). Upon decreasing the temperature, the value remains almost constant until 10 K, where it increases abruptly. The increase of the magnetic moment can be attributed to several causes. First, it can be the result of a weak ferromagnetic exchange coupling, occurring between neighbouring gadolinium ions, mediated by the phosphonate groups. In order to quantify these findings, the temperature dependent magnetic molar susceptibility was fitted with the program PHI using an isotropic exchange interaction Hamiltonian Ĥ_EX = −2JŜ₁·Ŝ₂.⁴²

Fig. 4 Temperature dependent measurement of the molar magnetic susceptibility per gadolinium ion for compound 2 in a temperature range between 2 K and 300 K with an applied field of 1 kOe. Black circles represent the experimental data, the red line the simulation with the parameter g = 1.992 and J = 0.0076 cm⁻¹.

For the simulation, the structure was considered as a dimer. The largest congruence of the data was obtained with the parameters J = 0.0076 cm⁻¹ and g = 1.992, where the positive sign of the coupling constant indicates that the coupling is indeed of weak ferromagnetic nature. It should be noted, that such small values are more of a qualitative nature. Another reason for the increasing susceptibility at low temperatures could be dipolar coupling. Such magnetic dipole–dipole interaction can occur, when two spin carriers are in very close proximity to each other, which is the case for the gadolinium ions along the I¹ chains in this structure.⁴³ And finally, such an increase of χ_MT could arise from a phase transition the compound undergoes at low temperatures.^44,45

Fig. 5 shows the magnetic characterization for compounds 3a, 3b and 3cvia measurements of the magnetic molar susceptibility. The χ_MT value at room temperature for the isotropic gadolinium complex is with 7.90 cm³ K mol⁻¹ close to the expected value for an uncoupled gadolinium ion (7.88 cm³ K mol⁻¹, ⁸S_7/2, S = 7/2, L = 0, g = 2). The susceptibility remains almost constant during the decline of the temperature until 13 K, where it decreases due to small zero field splitting or saturation effects. For quantification, the magnetic data were fitted using the programme PHI.⁴² The obtained parameters are g = 1.980 and a temperature independent paramagnetism (TIP) of 3.75 × 10⁻⁴. Fig. 5 shows the magnetic susceptibility for 3b and 3c as well. At room temperature, the values of χ_MT are 14.66 cm³ K mol⁻¹ and 11.45 cm³ K mol⁻¹, respectively. This is in good agreement with the expected values of 14.17 cm³ K mol⁻¹ for 3b and 11.82 cm³ K mol⁻¹ for 3c for uncoupled lanthanides: Tb³⁺ (⁷F₆, S = 6, L = 3, g = 3/2) and Dy³⁺ (⁶H_15/2, S = 5/2, L = 5, g = 4/3). Upon cooling, a decrease of the χ_MT value is observed, which is attributed to a progressive depopulation of the Stark levels split due to the ligand field.^43,46,47

Fig. 5 Temperature dependent measurement of the molar magnetic susceptibility for compounds 3a (yellow circles), 3b (red circles) and 3c (blue circles). The red line represents the fitted of 3a.

For the field dependent magnetization data at different temperatures for compound 3b and 3c see the ESI.† The obtained values for the magnetization are quiet lower than the theoretical saturation values one would expect for an isolated Dy³⁺ ion (10Nβ) and Tb³⁺ ion (9Nβ). The reason is found again in the magnetic anisotropy with a lower effective spin, and the splitting of the Stark level by the ligand field.^14,48,49

Compound 4 ([GdNi(II)(H₂L)₃(H₂O)₃]NaCl·6H₂O) shows at room temperature a χ_MT value of 9.04 cm³ K mol⁻¹ (see Fig. 6), which is in good agreement with the expected values for an uncoupled Gd³⁺-ion (7.88 cm³ K mol⁻¹, ⁸S_7/2, L = 0, S = 7/2, g = 2) and an uncoupled Ni²⁺-ion (1.00 cm³ K mol⁻¹, ³A_2g, L = 0, S = 1, g = 2). Upon cooling, the value remains almost constant and decreases only slightly to 8.76 cm³ K mol⁻¹ at 6 K, where it declines to a final value of 8.51 cm³ K mol⁻¹ at 2 K, most probably due to zero field splitting. The field dependent magnetization was measured in a temperature range from 2 K up to 10 K. At low fields, M rapidly increases until a maximum of 8.97Nβ and 6.78Nβ is slowly reached at 70 kOe (2 K and 10 K, respectively). Again, the values are in good agreement with the expected values at 2 K and 10 K for an uncoupled Gd³⁺ ion (6.98Nβ and 5.73Nβ) and an uncoupled Ni²⁺ ion (1.98Nβ and 1.10Nβ). The data were fitted using the program PHI. The best accordance of the data was obtained with the fitting parameters g(Gd³⁺) = 1.980, g(Ni²⁺) = 2.0438 and a TIP of 9.469 × 10⁻⁴. Due to the similarity of the coordinative environment between compound 3a and 4, the g(Gd³⁺) value from 3a was used. The magnetic data of compound 4 and compounds 3a, 3b and 3c confirm the oxidation states of the nickel and cobalt ions being +II and +III, respectively. In a strongly split ligand field, Co(III) is found almost exclusively in the diamagnetic low-spin state due to the large ligand field stabilisation energy arising from the electronic configuration t_2g⁶ e_g⁰.

Fig. 6 Representation of the magnetic properties of compound 4. Left: Variable temperature measurement of the molar susceptibility χ_M. Plot shows the product χ_MT vs. T. Circles represent the experimental data, the line the fit obtained by PHI. Right: Magnetization at different temperatures in an applied external magnetic field between 1 kOe and 70 kOe.

The χ_MT value of compound 3a reflects the absence of another spin carrier additional to the gadolinium ion. The oxidation during the reaction is promoted by two circumstances: first, by the elevated temperature during the crystallization. And second, due to the electronic configuration of t_2g⁶ e_g¹ in an octahedral coordination environment caused by a strong ligand field, such as from the 2,2′-bipyridine, the Co(II) ion seeks for a Jahn–Teller distortion to remove the degeneracy of the e_g-orbitals.^50,51 This is hindered by the rigid ligand backbone, which disallows the elongation along a Jahn–Teller axis. The electronic configuration of an octahedral coordinated Ni(II) ion (t_2g⁶ e_g²) is much more preferential compared to a Ni(III) ion (t_2g⁶ e_g¹), which would be forced by the Jahn–Teller effect into a for the system not realizable distorted octahedral coordination geometry.

Conclusions

In conclusion, we were able to identify and employ a highly versatile ligand system for the synthesis of several more-dimensional 3d and 4f metal complexes. By taking advantage of the different coordination properties, which are offered by the functionalities of the ligand, we prepared several compounds of homometallic and heterometallic nature. In compound 2, the metal ions are connected by phosphonate groups and form one-dimensional chains. They are tethered via the ligands backbone, which space them by a distance of 11.504 Å. The lanthanides in the heterometallic compounds 3a, 3b, 3c and 4 are in a capped trigonal prismatic coordination geometry and show a rare coordination number of seven. Magnetic measurements were carried out for all the more-dimensional complexes. Compound 2 revealed an increase of the magnetic molar susceptibility at low temperatures, which can be attributed to weak ferromagnetic exchange interactions. Compounds 3a, 3b, 3c and 4 showed that, depending on the metal, an in situ oxidation during the synthesis was observed due to the possible occurrence of a Jahn–Teller effect. Studies of the magnetization at various temperatures of compounds 3b and 3c, which contain anisotropic lanthanides, showed a split of the Stark level at low temperatures caused by the ligand field. Further syntheses involving the use of various metal/metal/ligand ratios and different reaction conditions are underway in our laboratory to widen the diversity of complexes which can be prepared by employing the ligand system and improve on the magnetic properties.

Experimental

General methods and materials

All chemicals were purchased from Alfa Aesar, Deutero, Fisher Chemicals, Sigma–Aldrich and Acros Organics and used without further purification. The NMR spectra were recorded at room temperature by using a Bruker DRX 400 spectrometer and analysed with the program MestReNova.⁵² Magnetic susceptibility data was collected with a Quantum Design SQUID magnetometer MPMSXL in a temperature range of 2–300 K with an applied field of 10 kOe. Elemental analyses (C, H, N and S) were measured at the micro analytical laboratories of the Johannes Gutenberg University Mainz. X-ray diffraction data were collected at 173 K with a Bruker SMART diffractometer and a STOE IPDS 2 T at the Johannes Gutenberg University Mainz. The structures were solved with ShelXT and refined with ShelXL 2013 with the program Olex2. CCDC 1477300 (for 1a), 1477301 (for 1b), 1477302 (for 2), 1477303 (for 3a), 1477304 (for 3b), 1477305 (for 3c), 1477306 (for 3a) and 1477307 (for SI1) contain the supplementary crystallographic data for this paper.

Ligand synthesis. The ligand was prepared as reported in the literature:³⁵ with a selective 5,5′-dibromination of 2,2′-bipyridine hydrobromide,⁵³ followed by a palladium catalysed phosphonation, the respective phosphonic dialkyl ester 5,5′-bis(diethylphosphonato)-2,2′-bipyridine was obtained as precursor. The desired ligand H₄L (2,2′-bipyridinyl-5,5′-diphosphonic acid) was accessible by hydrolyzation using TMS-Br according to the literature.⁵⁴ It was isolated as H₃H′L with H being a phosphonic acid proton and H′ being a nitrogen bound proton as depicted in Scheme 1. The crystal structure of H₃H′L can be found in the ESI (Fig. S1†).

Complex synthesis. For the preparation of compound 1a, 0.2 mmol H₃H′L (63.23 mg) were added to a mixture of 3 mL H₂O and 40 μL of concentrated aqueous ammonia. After filtration, a solution of 0.2 mmol Co(Ac)₂·4H₂O (49.82 mg) in 3 mL H₂O was added dropwise and under stirring. An immediate change of colour from pale pink to an intense orange was visible. The viol was left to stand for the solvent to evaporate slowly. Single crystals of orange colour suitable for single crystal XRD (see ESI Fig. S2†) grew after several days. Yield: 41.57 mg, 46.7%, C₁₀H₁₁N₂O_7.5P₂Co ([Co(H₂L)]·1.5H₂O) (400.08): calcd C 30.02, H 2.77, N 7.00; found C 30.30, H 3.02, N 6.96.

Compound 1b (Fig. S4†) was prepared according to the procedure of 1a by using 0.2 mmol NiCl₂·6H₂O (47.59 mg). Upon addition of ligand and metal, a colour change from pale green to light blue was observed. Yield: 37.72 mg, 42.39%, C₁₀H₁₅N₂O_9.5P₂Ni ([Ni(H₂L)]·3.5H₂O) (435.87): calcd C 27.56, H 3.47, N 6.43; found C 27.74, H 3.43, N 6.47.

Compound 2 was prepared by adding solution of 0.1 mmol Gd(NO₃)₃·5H₂O (43.33 mg) in 3 mL H₂O to a mixture of 0.15 mmol H₃H′L (47.42 mg) in 3 mL H₂O. Immediately, a white solid precipitated. The precipitate was separated from the solution by centrifugation. After the emulsification in 4 mL water, concentrated HCl (0.5 mL) was added to dissolve the solid, giving a clear solution. Acetonitrile (2 mL) was added and the viol was left to stand for six days, giving colourless crystals suitable for single crystal XRD. Yield: 81.51 mg, 58.13% C₃₀H_38.5N₆P₆O_23.25Gd₂Cl₄ ([Gd₂(H₂H′L)₂(H₂H′₂L)]Cl₄]·5.25H₂O) (1497.31): calcd C 24.07, H 2.59, N 5.61; found C 23.81, H 2.74, N 5.91.

Compound 3a, 3b and 3c were prepared by using the following procedure: 0.3 mmol H₃H′L (94.84 mg) were dissolved in 3 mL H₂O and 40 μL of concentrated ammonia were added. After filtration, 0.1 mmol Co(Ac)₂·4H₂O (24.91 mg) dissolved in 4 mL H₂O were added, whereupon the solution turned orange. The addition of 0.1 mmol M(NO₃)₃·xH₂O (3a: M = Gd, X = 5, m = 43.33 mg; 3b: M = D, X = 1, m = 36.65 mg; 3c: M = Tb, X = 1, m = 36.30 mg) dissolved in 4 mL H₂O resulted in an immediate precipitation of a white solid, which was dissolved by the addition of 0.5 mL concentrated HCl. The viol was placed in an oven set to 80 °C for 18 h, after which isostructural single crystals suitable for single crystal XRD of orange colour with similar morphology were obtained. 3a: Yield: 50.17 mg, 43.3%. C₃₀H₄₀N₆P₆O₂₆CoGd ([GdCo(H₂L)₃]·8H₂O) (1302.69): calcd C 27.66, H 3.09, N 6.45; found C 27.64, H 3.20, N 6.45. 3b: Yield: 54.41 mg, 46.8%, C₃₀H₃₄N₆P₆O₂₃CoDy ([DyCo(H₂L)₃]·5H₂O) (1253.89): calcd C 28.74, H 2.73, N 6.70; found C 28.60, H 2.65, N 6.74. 3c: Yield: 44.35 mg, 38.2%, C₃₀H₃₈N₆P₆O₂₅CoTb ([DyCo(H₂L)₃]·7H₂O) (1286.35): calcd C 28.01, H 2.98, N 6.53; found C 28.06, H 2.93, N 6.59.

For the preparation of compound 4, 0.15 mmol H₃H′L (47.42 mg) were dissolved in 2 mL H₂O and 0.32 mmol of sodium hydroxide were added. After filtration, 0.05 mmol NiCl₂·6H₂O (11.90 mg) dissolved in 2 mL H₂O were added. The addition of 0.05 mmol Gd(NO₃)₃·5H₂O dissolved in 2 mL H₂O resulted in an immediate precipitation of a white solid, which was dissolved by the addition of 0.4 mL concentrated HCl. The viol was placed in an oven set to 80 °C, where the solvent evaporated and the solution concentrated until a remaining volume of approximately 1.5 mL and then left to stand at room temperature. Single crystals of pale orange suitable for single crystal XRD could be obtained after one day. Yield: 26.58 mg, 43.7%. C₃₀H₃₅N₆P₆O₂₃NiGdNaCl ([GdNi(H₂L)₃]·5H₂O) (1307.85): calcd C 27.55, H 2.70, N 6.43; found C 27.39, H 2.88, N 6.51.

Acknowledgements

Dr Dieter Schollmeyer and Regine Jung-Pothmann are kindly acknowledged for the support with the crystallographic data.

Notes and references

1. J. C. Tan and A. K. Cheetham, Chem. Soc. Rev., 2011, 40, 1059–1080.
2. L. D. Carlos, R. A. Ferreira, V. De Zea Bermudez, B. Julian-Lopez and P. Escribano, Chem. Soc. Rev., 2011, 40, 536–549.
3. M. T. Wharmby, J. P. S. Mowat, S. P. Thompson and P. A. Wright, J. Am. Chem. Soc., 2011, 133, 1266–1269.
4. A. Kondo, T. Satomi, K. Azuma, R. Takeda and K. Maeda, Dalton Trans., 2015, 44, 12717–12725.
5. S. Jones, J. M. Vargas, S. Pellizzeri, C. J. O'Connor and J. Zubieta, Inorg. Chim. Acta, 2013, 395, 44–57.
6. M. Pramanik, M. Nandi, H. Uyama and A. Bhaumik, Catal. Sci. Technol., 2012, 2, 613–620.
7. H. Tan, W. Chen, D. Liu, X. Feng, Y. Li, A. Yan and E. Wang, Dalton Trans., 2011, 40, 8414–8418.
8. L. D. Carlos, R. A. S. Ferreira, V. D. Bermudez, B. Julian-Lopez and P. Escribano, Chem. Soc. Rev., 2011, 40, 536–549.
9. K. J. Gagnon, H. P. Perry and A. Clearfield, Chem. Rev., 2012, 112, 1034–1054.
10. K. Maeda, Microporous Mesoporous Mater., 2004, 73, 47–55.
11. K. Maeda, Y. Kiyozumi and F. Mizukami, J. Phys. Chem. B, 1997, 101, 4402–4412.
12. Y.-S. Ma, H. Li, J.-J. Wang, S.-S. Bao, R. Cao, Y.-Z. Li, J. Ma and L.-M. Zheng, Chem. – Eur. J., 2007, 13, 4759–4769.
13. L. Rosado Piquer and E. C. Sañudo, Dalton Trans., 2015, 44, 8771–8780.
14. J. Tang, I. Hewitt, N. T. Madhu, G. Chastanet, W. Wernsdorfer, C. E. Anson, C. Benelli, R. Sessoli and A. K. Powell, Angew. Chem., Int. Ed., 2006, 45, 1729–1733.
15. G. Novitchi, G. Pilet, L. Ungur, V. V. Moshchalkov, W. Wernsdorfer, L. F. Chibotaru, D. Luneau and A. K. Powell, Chem. Sci., 2012, 3, 1169–1176.
16. J. A. Sheikh, A. Adhikary, H. S. Jena, S. Biswas and S. Konar, Inorg. Chem., 2014, 53, 1606–1613.
17. Y. Z. Zheng, M. Evangelisti and R. E. P. Winpenny, Angew. Chem., Int. Ed., 2011, 50, 3692–3695.
18. S. Hu, L. Yun, Y. Z. Zheng, Y. H. Lan, A. K. Powell and M. L. Tong, Dalton Trans., 2009, 1897–1900.
19. M. Andruh, J.-P. Costes, C. Diaz and S. Gao, Inorg. Chem., 2009, 48, 3342–3359.
20. J. A. Sheikh, S. Goswami and S. Konar, Dalton Trans., 2014, 43, 14577–14585.
21. Y. Bing, N. Xu, W. Shi, K. Liu and P. Cheng, Chem. – Asian J., 2013, 8, 1412–1418.
22. M. Kurmoo, Chem. Soc. Rev., 2009, 38, 1353–1379.
23. L. J. Murray, M. Dinca and J. R. Long, Chem. Soc. Rev., 2009, 38, 1294–1314.
24. A. Bencini, C. Benelli, A. Caneschi, R. L. Carlin, A. Dei and D. Gatteschi, J. Am. Chem. Soc., 1985, 107, 8128–8136.
25. Z.-G. Gu and S. C. Sevov, J. Mater. Chem., 2009, 19, 8442–8447.
26. B. Zhao, P. Cheng, Y. Dai, C. Cheng, D. Z. Liao, S. P. Yan, Z. H. Jiang and G. L. Wang, Angew. Chem., Int. Ed., 2003, 42, 934–936.
27. T. K. Prasad, M. V. Rajasekharan and J. P. Costes, Angew. Chem., Int. Ed., 2007, 46, 2851–2854.
28. A. C. Rizzi, R. Calvo, R. Baggio, M. T. Garland, O. Peña and M. Perec, Inorg. Chem., 2002, 41, 5609–5614.
29. R. A. Palmer and T. S. Piper, Inorg. Chem., 1966, 5, 864–878.
30. V. Lopez-Diaz, T. M. Smith Pellizerri, M. D. Lijewski, K. Ruhlandt and J. Zubieta, Inorg. Chim. Acta, 2016, 441, 109–116.
31. E. M. Pineda, F. Tuna, Y.-Z. Zheng, S. J. Teat, R. E. P. Winpenny, J. Schnack and E. J. L. McInnes, Inorg. Chem., 2014, 53, 3032–3038.
32. M. Bartholomä, H. Chueng, S. Pellizzeri, K. Ellis-Guardiola, S. Jones and J. Zubieta, Inorg. Chim. Acta, 2012, 389, 90–98.
33. E. Fernández-Zapico, J. Montejo-Bernardo, A. Fernández-González, J. R. García and S. García-Granda, J. Solid State Chem., 2015, 225, 285–296.
34. G. Guerrero, J. G. Alauzun, M. Granier, D. Laurencin and P. H. Mutin, Dalton Trans., 2013, 42, 12569–12585.
35. V. Penicaud, F. Odobel and B. Bujoli, Tetrahedron Lett., 1998, 39, 3689–3692.
36. F. Odobel, B. Bujoli and D. Massiot, Chem. Mater., 2001, 13, 163–173.
37. I. Gillaizeau-Gauthier, F. Odobel, M. Alebbi, R. Argazzi, E. Costa, C. A. Bignozzi, P. Qu and G. J. Meyer, Inorg. Chem., 2001, 40, 6073–6079.
38. H. Zabri, I. Gillaizeau, C. A. Bignozzi, S. Caramori, M.-F. Charlot, J. Cano-Boquera and F. Odobel, Inorg. Chem., 2003, 42, 6655–6666.
39. S. Khanra, S. Konar, A. Clearfield, M. Helliwell, E. J. L. McInnes, E. Tolis, F. Tuna and R. E. P. Winpenny, Inorg. Chem., 2009, 48, 5338–5349.
40. A. Rajput and R. Mukherjee, Coord. Chem. Rev., 2013, 257, 350–368.
41. Comprehensive coordination chemistry II: from biology to nanotechnology, ed. J. A. McCleverty and T. J. Meyer, Elsevier Pergamon, Amsterdam, Boston, 2004.
42. N. F. Chilton, R. P. Anderson, L. D. Turner, A. Soncini and K. S. Murray, J. Comput. Chem., 2013, 34, 1164–1175.
43. W. Haase, Ber. Bunsen-Ges. Phys. Chem., 1994, 98, 1208–1208.
44. A.-M. Pütz, L. M. Carrella and E. Rentschler, Dalton Trans., 2013, 42, 16194–16199.
45. C. L. Calvin, Magnetochemistry, Springer Verlag, New York, Heidelberg, Tokyo, 1986.
46. Y.-L. Hou, G. Xiong, B. Shen, B. Zhao, Z. Chen and J.-Z. Cui, Dalton Trans., 2013, 42, 3587–3596.
47. F. Gao, Y.-Y. Li, C.-M. Liu, Y.-Z. Li and J.-L. Zuo, Dalton Trans., 2013, 42, 11043–11046.
48. F. Yang, Q. Zhou, G. Zeng, G. Li, L. Gao, Z. Shi and S. Feng, Dalton Trans., 2013, 43, 1238–1245.
49. S. Osa, T. Kido, N. Matsumoto, N. Re, A. Pochaba and J. Mrozinski, J. Am. Chem. Soc., 2004, 126, 420–421.
50. H. A. Jahn and E. Teller, Proc. R. Soc. London, Ser. A, 1937, 161, 220–235.
51. H. A. Goodwin, in Spin Crossover in Transition Metal Compounds I, ed. P. Gütlich and H. A. Goodwin, Springer, Berlin Heidelberg, 2004, pp. 59–90.
52. J. C. Cobas and F. J. Sardina, Concepts Magn. Reson., Part A, 2003, 19A, 80–96.
53. A. B. Zdravkov and N. N. Khimich, Russ. J. Org. Chem., 2006, 42, 1200–1202.
54. C. E. McKenna, M. T. Higa, N. H. Cheung and M.-C. McKenna, Tetrahedron Lett., 1977, 18, 155–158.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 1477300–1477307. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6dt02023b

---