Metal–organic frameworks based on polynuclear lanthanide complexes and octahedral rhenium clusters

Yulia M. Litvinovaa, Yakov M. Gayfulin*a, Jan van Leusenb, Denis G. Samsonenkoac, Vladimir A. Lazarenkod, Yan V. Zubavichuse, Paul Kögerlerb and Yuri V. Mironovac
aNikolaev Institute of Inorganic Chemistry SB RAS, Lavrentiev ave. 3, 630090 Novosibirsk, Russia. E-mail:
bInstitute of Inorganic Chemistry, RWTH Aachen University, Landoltweg 1, 52074 Aachen, Germany
cNovosibirsk State University, Pirogova str. 2, 630090 Novosibirsk, Russia
dNational Research Center “Kurchatov Institute”, Kurchatov Square 1, 123182 Moscow, Russia
eFederal Research Center Boreskov Institute of Catalysis, Lavrentiev Ave. 5, 630090 Novosibirsk, Russia

Received 27th March 2019 , Accepted 6th May 2019

First published on 6th May 2019

The octahedral chalcocyanide cluster anion [Re6S8(CN)6]4– can serve as a metallolinker for the self-assembly of metal–organic frameworks (MOFs) comprising polynuclear lanthanide complexes, which have been illustrated for a series of nine compounds isolated as [{Ln43-OH)4(ina)4(H2O)n}{Re6S8(CN)6}]·mH2O (Ln = Pr – Er, ina = isonicotinate). Their solid-state framework structures are based on {Ln43-OH)4}8+ cubane fragments interlinked by isonicotinate and rhenium cyanide clusters. The coordination environment of the lanthanide ions, the coordination numbers and even the number of bridging isonicotinate moieties changes systematically within the lanthanide row, however, the framework topology remains unchanged, confirming the structure-directing role of the rhenium cluster. Structures, magnetochemical properties, luminescence and thermal stability of the new MOFs have been thoroughly analyzed.


Polynuclear assemblies formed by oxo-bridged lanthanide ions are among the most intriguing compounds of rare-earth elements due to their diverse structural motifs and potentially applicable properties, including molecular magnetism,1–4 magnetic cooling,5–7 luminescence,8–13 and biological imaging.14,15 The chemistry of these compounds is being intensively studied, and many discrete species with their lanthanide nuclearity ranging from 3 to 104 have been reported.16,17 Most of these clusters are only stable within a narrow pH range, and their limited chemical stability and insufficient structural rigidity restricts the scope of potential applications. Utilizing such lanthanide clusters as secondary building units (SBUs) in the construction of metal–organic frameworks (MOFs) is one of the most promising approaches for exploiting the functional properties of lanthanide clusters as the components of a stable crystalline matrix.18–24 The flexible coordination sphere of lanthanide ions and the various geometries of their complexes render it possible to synthesize a number of MOFs based on lanthanide clusters (LnCMOFs) customized for a specific practical task. Great efforts have been applied in recent years in order to enable a more targeted synthesis of such compounds. Particularly, it was demonstrated that rationalization of the synthesis and emission properties of the LnCMOFs could be achieved by the use of pre-designed organic linkers containing α-amino groups.9

Metal–organic frameworks often combine properties caused by specific framework architecture, e.g. permanent porosity, selective gas sorption and chemical stability, with those ones related to the properties of discrete building blocks or secondary building units (SBUs), e.g. luminescence and magnetism. Combination of several of these properties in a stable crystalline material is often desirable. For the last few years we are interested in the design of MOFs based on rhenium clusters with {Re63-Q)8}n (Q = S, Se, Te) cores. In contrast to polynuclear lanthanide complexes, the metal atoms within the octahedral {Re6} fragments are bound by covalent bonds, which leads to a remarkable chemical stability, structural rigidity and interesting luminescent and redox properties of compounds comprising these clusters. Particularly, the cyanide cluster anions [Re6Q8(CN)6]4–/3− (Q = S, Se, Te) have a set of properties that make them a promising units for the use in functional coordination polymers.25,26 The presence of ambidentate CN ligands turns the cluster anion into a convenient reaction center that readily coordinates metal ions.27–29 Like all rhenium clusters having 24 cluster valence electrons (CVE), the [Re6Q8(CN)6]4– anions display the red luminescence, while the non-luminescent oxidized forms [Re6Q8(CN)6]3– having 23 CVE display paramagnetic behavior.30,31 A combination of rhenium clusters and lanthanide ions as the components of a MOFs crystal structures makes possible the synergy between the luminescent, magnetic and redox properties of these centers. For example, the fact that the oxidation of the cluster anions leads to a quenching of their luminescence makes the cluster anion a potential “switch” for the properties of coordination polymers. This possibility has been demonstrated in the recent work.32

The goal of the study of metal–organic coordination polymers based on luminescent clusters and lanthanide cations is creation of the materials that have the synergy of the functional properties of individual building blocks. However, investigation of the compounds of this type is in initial stage, and search for the new knowledge about the mechanisms of their formation is relevant. In this work we illustrate how the cluster anion [Re6S8(CN)6]4– can be used as a structure-forming SBU for the effective synthesis of MOFs based on lanthanide clusters and isonicotinate (ina). To achieve this goal, we employ modified traditional synthetic approach based on the ligand-controlled hydrolysis or alcoholysis of Ln3+ ions in a pH range from 5.5 to 7 in the presence of multidentate O-donor ligands.33 This modified technique is based on the hydrolysis of the amide group of isonicotinamide, followed by LnCMOF self-assembly. Reaction between Ln3+ ions in the presence of [Re6S8(CN)6]4– anion and isonicotinamide in hydrothermal conditions led to nine charge-neutral, hydrolytically stable frameworks consisting of cubane clusters {Ln43-OH)4}8+ interconnected by isonicotinate linkers and, in addition, by the rhenium cyanide clusters. These compounds with a common formula [{Ln43-OH)4(ina)4(H2O)n}{Re6S8(CN)6}]·mH2O (Ln3+ = Pr3+ (1), Nd3+ (2), Sm3+ (3), Eu3+ (4), Gd3+ (5), Tb3+ (6), Dy3+ (7), Ho3+ (8), Er3+ (9)) represent the first MOFs based on tetrahedral lanthanide complexes and cyanometallate building blocks. The frameworks demonstrate a combination of red luminescence of rhenium cluster and magnetic properties of {Ln4(μ-OH)4}8+ complexes. All nine frameworks crystallize in the space group P21/n, however, the coordination environments of the lanthanide ions, their coordination numbers and even the number of bridging isonicotinate moieties changes systematically within the lanthanide row, confirming the structure-directing role of the rhenium cluster.


Materials and spectroscopic studies

The starting cluster salt Cs3K[Re6Se8(CN)6]·2H2O was prepared as described.34 Other reagents and solvents were used as purchased. Distilled H2O with pH ≈ 5.8 was used in all syntheses.

Elemental analysis was made on a EuroVector EA3000 analyzer. IR spectra in KBr pellets in the range 4000–375 cm−1 were recorded on a Bruker Scimitar FTS 2000 spectrometer. Energy dispersive spectroscopy (EDS) was performed on a Hitachi TM-3000 electron microscope equipped with a Bruker Nano EDS analyzer. Thermal analysis in the temperature range 30–700 °C was carried out using a Netzsch TG 209 F1 Iris instrument. The experiments were performed under He flow (80 cm3 min−1) at 10 K min−1 heating rate; the sample mass was ∼6 mg. Powder X-ray diffraction patterns were collected at room temperature on a Philips PW1820/1710 diffractometer (Cu Kα radiation, graphite monochromator, silicon plate used as an external standard). Room temperature excitation and emission spectra of powdered samples were recorded with a Horiba Jobin Yvon Fluorolog 3 photoluminescence spectrometer equipped with a 450 W ozone-free Xe-lamp, cooled PC177CE-010 photon detection module with a PMT R2658 and double grating excitation and emission monochromators. Excitation and emission spectra were corrected for source intensity (lamp and grating) and emission spectral response (detector and grating) by standard correction curves.

Single crystal diffraction studies

Single crystals of compounds 1–9 were collected directly from the reaction mixtures. Diffraction data for single crystals 1, 2, 3, 6, 7, 8 were obtained on an Agilent Xcalibur diffractometer equipped with a CCD AtlasS2 detector (MoKα, graphite monochromator, ω scans). Integration, absorption correction, and determination of unit cell parameters were performed using the CrysAlisPro program package.35 Diffraction data for single crystals 4, 5, 9 were collected on the ‘Belok’ beamline (λ = 0.8026 Å, φ-scans) of the National Research Center ‘Kurchatov Institute’ (Moscow, Russian Federation) using a Rayonix SX165 CCD detector. The data were indexed, integrated and scaled using the utility iMOSFLM in CCP4 program.36 Absorption correction was applied using the Scala program.37 The structures were solved by dual space algorithm (SHELXT)38 and refined by the full-matrix least squares technique (SHELXL)39 in the anisotropic approximation (except hydrogen atoms). Positions of hydrogen atoms of organic ligands were calculated geometrically and refined in the riding model. Hydrogen atoms of the water molecules were not located. The crystallographic data and details of the structure refinements are summarized in Table S1. Selected bond distances are listed in Table S2. CCDC 1904278–1904286 contain the crystallographic data for compounds 1–9, respectively.

Magnetochemical analysis

The magnetic data of 1–9 were collected using a Quantum Design MPMS-5XL SQUID magnetometer. The polycrystalline samples were compacted and immobilized into cylindrical PTFE capsules. The dc data were recorded as a function of the magnetic field (0.1–5.0 T at 2.0 K) and the temperature (2–290 K at 0.1 T). In addition, ac data (Bac = 3 G) were collected as function of the temperature (1.9–50 K) and frequency (3–997 Hz) at zero, 0.1 and 0.3 T dc bias field. All data were corrected for the diamagnetic contributions of the sample holders and the compounds (χm,dia/10−3 cm3 mol−1, 1: −1.43, 2: −1.45, 3: −1.45, 4: −1.45, 5: −1.46, 6: −1.47, 7: −1.45, 8: −1.47, 9: −1.48).

Synthesis of [{Pr43-OH)4(μ-ina)(ina)3(H2O)6}{Re6S8(CN)6}]·6H2O (1)

A solution of isonicotinamide (10 mg, 0.082 mmol) in 1 mL of H2O was mixed with a solution of Pr(NO3)3·6H2O (30 mg, 0.069 mmol) in 1 mL of H2O in a screw-capped glass vial. Then a solution of Cs3K[Re6Se8(CN)6]·2H2O (25 mg, 0.012 mmol) in 4 mL of H2O was added forming clear orange mixture. The vial was placed in a preheated furnace at 160 °C for 48 hours, then cooled at natural rate. Orange crystals of compound 1 were formed on the walls of the vial together with colorless amorphous powder. The reaction mixture was sonicated and mother liquor with the suspension of an amorphous powder was decanted. Crystals were washed with H2O and dried in air. Yield: 25 mg (72%). FT-IR (νmax, cm−1): ν(CN) 2131 (vs); ν(OH) 3402 (s), 3601 (vs); ina: 1595 (vs), 1548 (vs), 1494 (m), 1409 (vs), 1226 (w), 1155 (w) 1058 (m), 1004 (w), 864 (w), 769 (s), 711 (m), 682 (s), 547 (w), 418 (w). Anal. calcd for C30H44N10O24Pr4Re6S8: C, 12.57; H, 1.54; N, 4.88. Found: C, 12.75; H, 1.45; N, 5.03. EDS: Pr[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 3.9[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]7.6.

Synthesis of [{Nd43-OH)4(μ-ina)(ina)3(H2O)5.5(C6H6N2O)0.5}{Re6S8(CN)6}]·4H2O (2)

Compound 2 was synthesized using the same procedure as for 1 except using Nd(NO3)3·6H2O (30 mg, 0.068 mmol) instead of Pr(NO3)3·6H2O. Yield: 22 mg (63%). FT-IR (νmax, cm−1): ν(CN) 2131(vs); ν(OH) 3385 (s), 3601 (vs); ina: 1597 (vs), 1548 (vs), 1494 (m), 1409 (vs), 1528 (w), 1155 (w), 1060 (m), 1006 (w), 864 (w), 769 (w) 711 (m), 682 (s), 557 (w). Anal. calcd for C33H42N11Nd4O22Re6S8: C, 13.68; H, 1.46; N, 5.32. Found: C, 13.49; H, 1.38; N, 5.16. EDS: Nd[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 4.1[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.6.

Synthesis of [{Sm43-OH)4(μ-ina)(ina)3(H2O)6}{Re6S8(CN)6}]·6H2O (3)

Compound 3 was synthesized using the same procedure as for 1 except using Sm(NO3)3·6H2O (30 mg, 0.067 mmol) instead of Pr(NO3)3·6H2O. Yield: 23 mg (66%). FT-IR (νmax, cm−1): ν(CN) 2133 (vs); ν(OH) 3410 (vs), 3591 (vs); ina: 1398 (vs), 1548 (vs), 1494 (m), 1403 (vs), 1228 (w), 1157 (w), 1038 (m), 1006 (w), 862 (w), 769 (s), 711 (m), 680 (s), 559 (w), 418 (w). Anal. calcd for C30H44N10O24Re6S8Sm4: C, 12.40; H, 1.53; N, 4.82. Found: C, 12.77; H, 1.21; N, 4.89. EDS: Re[thin space (1/6-em)]:[thin space (1/6-em)]S[thin space (1/6-em)]:[thin space (1/6-em)]Sm = 6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.1[thin space (1/6-em)]:[thin space (1/6-em)]4.0.

Synthesis of [{Eu43-OH)4(μ-ina)(ina)3(H2O)6}{Re6S8(CN)6}]·5H2O (4)

Compound 4 was synthesized using the same procedure as for 1 except using Eu(NO3)3·6H2O (30 mg, 0.067 mmol) instead of Pr(NO3)3·6H2O. Yield: 24 mg (63%). FT-IR (νmax, cm−1): ν(CN) 2133 (vs); ν(OH) 3402 (vs), 3589 (vs); ina: 1602 (vs), 1548 (vs), 1494 (m), 1411 (vs), 1228 (w), 1157 (w), 1060 (m), 1006 (w), 864 (w), 771 (s), 711 (m), 682 (s), 561 (w). Anal. calcd for C30H42Eu4N10O23Re6S8: C, 12.46; H, 1.46; N, 4.84. Found: C, 12.90; H, 1.17; N, 5.06. EDS: Eu[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 4.0[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]7.5.

Synthesis of [{Gd43-OH)4(μ-ina)(ina)3(H2O)6}{Re6S8(CN)6}]·5H2O (5)

Compound 5 was synthesized using the same procedure as for 1 except using of Gd(NO3)3·6H2O (30 mg, 0.066 mmol) instead of Pr(NO3)3·6H2O. Yield: 22 mg (63%). FT-IR (νmax, cm−1): ν(CN) 2133 (vs); ν(OH) 3591 (vs), 3412 (vs); ina: 1600 (vs), 1550 (vs), 1494 (m), 1411 (vs), 1228 (w), 1157 (w), 1058 (m), 1008 (w), 864 (w), 771 (s), 711 (m), 682 (s), 367 (w). Anal. calcd for C30H42Gd4N10O23Re6S8: C, 12.37; H, 1.45; N, 4.80. Found: C, 12.82; H, 1.25; N, 4.85. EDS: Gd[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 3.9[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.4.

Synthesis of [{Tb43-OH)4(μ-ina)(ina)3(H2O)5}{Re6S8(CN)6}]·7H2O (6)

Compound 6 was synthesized using the same procedure as for 1 except using Tb(NO3)3·6H2O (30 mg, 0.066 mmol) instead of Pr(NO3)3·6H2O. Yield: 20 mg (57%). FT-IR (νmax, cm−1): ν(CN) 2131 (vs); ν(OH) 3589 (vs), 3398 (vs); ina: 1598 (vs), 1548 (vs), 1496 (m), 1411 (vs), 1228 (w), 1157 (w), 1060 (m), 1006 (w), 864 (w), 771 (s), 711 (m), 682 (s), 567 (w). Anal. calcd for C30H44N10O24Re6S8Tb4: C, 12.26; H, 1.50; N, 4.77. Found: C, 12.68; H, 1.20; N, 4.80. EDS: Gd[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 3.9[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.4.

Synthesis of [{Dy43-OH)4(μ-ina)1.6(ina)2.4(H2O)4}{Re6S8(CN)6}]·5H2O (7)

Compound 7 was synthesized using the same procedure as for 1 except using Dy(NO3)3·6H2O (30 mg, 0.066 mmol) instead of Pr(NO3)3·6H2O. Yield: 16 mg (46%). FT-IR (νmax, cm−1): ν(CN) 2129 (vs); ν(OH) 3589 (vs) 3332 (vs); ina: 1598 (vs), 1548 (vs), 1496 (m), 1411 (vs), 1128 (w), 1155 (m), 1060 (w), 1008 (w), 864 (w), 773 (s), 711 (m), 682 (s), 567 (w). Anal. calcd for C30H38Dy4N10O21Re6S8: C, 12.43; H, 1.32; N, 4.83. Found: C, 12.38; H, 1.34; N, 4.86. EDS: Gd[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 3.9[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.4.

Synthesis of [{Ho43-OH)4(μ-ina)2(ina)2(H2O)4}{Re6S8(CN)6}]·7H2O (8)

Compound 8 was synthesized using the same procedure as for 1 except using Ho(NO3)3·6H2O (30 mg, 0.065 mmol) instead of Pr(NO3)3·6H2O. Yield: 18 mg (51%). FT-IR (νmax, cm−1): ν(CN) 2127 (vs); ν(OH) 3587 (vs), 3406 (vs); ina: 1600 (vs), 1548 (vs), 1496 (m), 1411 (vs), 1228 (w), 1155 (w), 1060 (m), 1006 (w), 864 (w), 775 (s), 711 (m), 684 (s), 567 (w). Anal. calcd for C30H42Ho4N10O23Re6S8: C, 12.24; H, 1.44; N, 4.76. Found: C, 12.63; H, 1.18; N, 4.72. EDS: Ho[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 4.2[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.4.

Synthesis of [{Er43-OH)4(μ-ina)2(ina)2(H2O)4}{Re6S8(CN)6}]·7H2O (9)

Compound 9 was synthesized using the same procedure as for 1 except using Er(NO3)3·6H2O (30 mg, 0.065 mmol) instead of Pr(NO3)3·6H2O. Yield: 24 mg (68%). FT-IR (νmax, cm−1): ν(CN) 2129 (vs); ν(OH) 3589 (vs), 3400 (s); ina: 1600 (vs), 1548 (vs), 1496 (m), 1411 (vs), 1228 (w), 1153 (w), 1060 (m), 1008 (w), 864 (w), 775 (s), 711 (m), 684 (s), 570 (w). Anal. calcd for C30H42Er4N10O23Re6S8: C, 12.20; H, 1.43; N, 4.74. Found: C, 12.62; H, 1.18; N, 4.79. EDS: Er[thin space (1/6-em)]:[thin space (1/6-em)]Re[thin space (1/6-em)]:[thin space (1/6-em)]S = 4.2[thin space (1/6-em)]:[thin space (1/6-em)]6.0[thin space (1/6-em)]:[thin space (1/6-em)]8.3.

Results and discussion


Several coordination polymers based on {Ln43-OH)4}8+ oxo-clusters and bridging carboxylates are currently known in the literature, however, only few of them adopt framework structures.18 The vast majority of LnCMOFs were synthesized by slow hydrolysis of the soluble lanthanide salt in the presence of organic acid. In addition to the complexity and the need to optimize the synthesis conditions for each of the lanthanides separately, the essential disadvantage of this approach is the low yield of crystalline products, usually in the range of 15–35%.

In search of an effective method that would allow to synthesize MOFs based on lanthanide ions and rhenium clusters in preparative amounts, we tried to control the formation of crystalline products by the use of an amide precursor. Amides are known to hydrolyze to form the corresponding carboxylic acids in both acidic and alkaline environments.40 Until now, we could not find published examples utilizing this strategy for the synthesis of LnCMOFs.

Hydrothermal reaction of lanthanide nitrates, isonicotinamide and a salt of the [Re6S8(CN)6]4– cluster anion led to formation of crystalline compounds 1–9 containing bridging isonicotinate. Compounds 1–9 were obtained in high yields (up to 80% based on {Re6} fragment). This suggests that hydrolysis of amides under controlled temperature and pH generally represents a promising method for the preparation of lanthanide MOFs with carboxylate linkers. None of the compounds, with the exception of 2, contains isonicotinamide molecules. In the structure of 2 an isonicotinamide molecule partially replaces the aqua-ligand at the Nd3 atom acting as a terminal ligand (Fig. S1).

Crystal structures

Compounds 1–9 crystallize in the monoclinic space group P21/n. The praseodymium derivative 1 exhibits the framework structure composed of the [Re6S8(CN)6]4– cluster anion and a cationic cubane oxo-cluster {Pr43-OH)4(μ-ina)(ina)3(H2O)6}4+ (Fig. 1). The rhenium cluster displays its typical geometry, i.e. an octahedral metal core surrounded by eight face-centered μ3-S ligands. Each Re atom is additionally coordinated by an apical cyanide ligand, resulting in a unit with idealized Oh symmetry. The corresponding Re–Re, Re–S and Re–C bond distances vary in narrow ranges, and their average values are 2.595(7), 2.401(5) and 2.12(1) Å, respectively. These values remain almost constant for all nine compounds (Table S1). The lanthanide cluster fragment consists of four Pr3+ ions located at the vertices of a distorted tetrahedron with Pr⋯Pr distances of 3.7513(5)–4.0479(6) Å. Four μ3-OH groups cap the trigonal faces of the tetrahedron, forming the cubane unit {Pr43-OH)4}8+. Similar cubane fragments are a typical structural motif in discrete lanthanide oxo-hydroxo clusters. The carboxylate groups of isonicotinate (ina) ligands bridge four Pr–Pr edges, thereby completing the lanthanide SBU (Fig. 2). The coordination environment of each Pr atom includes also one or two O atoms of H2O ligands and one or two N atoms of cyanide groups. In this way, the coordination numbers of Pr1 and Pr2 are 9, while Pr3 and Pr4 have coordination numbers of 8 (Table S3).
image file: c9qi00339h-f1.tif
Fig. 1 Asymmetric unit of the compound 1 with atom numbering scheme, showing thermal ellipsoids drawn at 50% probability level.

image file: c9qi00339h-f2.tif
Fig. 2 Structure of the {Pr43-OH)4(μ-ina)(ina)3(H2O)6}4+ unit in 1. The central distorted Pr43-O)4 cubane fragment is highlighted in transparent yellow. A* denotes a symmetry-equivalent ina ligand (symmetry code: −x + 0.5; y −0.5; –z + 1.5). Dashed bonds indicate Pr–NCN bonds to adjacent [Re6S8(CN)6]4– cluster anions. Color code: Pr: green, O: red, C: gray, N: blue spheres. Hydrogen atoms omitted for clarity.

The cubane units are connected by bridging ina ligands forming a zigzag chain formulated as {Pr43-OH)4(μ-ina)2/2(ina)3(H2O)6}n4n+ (Fig. 3a). The COO group of the bridging isonicicotinate forms three Pr–O bonds (2.564(5), 2.572(6) and 2.732(5) Å); the Pr–N bond amounts to 2.754(6) Å.

image file: c9qi00339h-f3.tif
Fig. 3 (a) Segment of the polymeric zigzag chain {Pr43-OH)4(μ-ina)(ina)3(H2O)6}n4n+ in 1 in the ab plane; color code as in Fig. 2. (b) Simplified presentation of the same chain segment showing the connection of the {Pr43-OH)4}8+ fragments by the octahedral [Re6S8(CN)6]4– anions via Pr–N–C–Re bridges (Re: dark gray, S: yellow spheres, cyanide groups emphasized by purple C–N bonds). Terminal aqua ligands of Pr3+ ions, hydrogen atoms and ina groups not directly interconnecting the shown {Pr4} clusters are omitted for clarity. View along c. (c) Same structure in a perpendicular view along a direction, showing the alternating stacking of layers of {Re6} and {Pr4} units.

The final step in the formal construction of the framework concerns the linking of the polymeric chains, located in the structure parallel to the b axis, with the rhenium cluster anions (Fig. 3b and c). The six cyanide groups of the [Re6S8(CN)6]4– anion coordinate to six Pr atoms of three different chains (Pr–NCN: 2.598(6)–2.714(6) Å, with an average value of 2.62(6) Å). Therefore, the rhenium cluster act as metallolinker joining the 1D chains into a 3D framework. Note that the literature contains only a few examples of the synthesis of LnCMOFs using metallolinkers,41–45 and no examples of the use of cyanometallates.

The above-described structure principle of the frameworks and the same Ln coordination numbers remain unchanged in the structures of compounds 1–4 (see Table S3 and Fig. S1–S3, S9). There is no significant change in the geometry of the {Ln43-OH)4}8+ fragments or coordination mode of the ina ligands in the structures of these compounds. One can see only the typical narrowing of the average Ln–Ln, Ln–O and Ln–N distances in line with the decreasing ionic radii of the Ln cations. A decrease of the coordination number was found first in the structure of compound 5 (Ln = Gd).

In this structure, the bridging isonicotinate is disordered between two positions, and its COO group form three or two Gd–O bonds as shown in Fig. S4 and S9d. In the latter case, removal of one terminal H2O ligand lowers the Gd coordination numbers to eight (Table S3). The structure of compound 6 (Ln = Tb) also displays disordered bridging ina ligands, which form one or two Tb–O bonds (Fig. S5 and S10a). In all cases, the Tb coordination number is eight.

More noticeable changes were found in the structure of 7 (Ln = Dy). In this structure, a part of terminal isonicotinate ligands adopt a bridging position, replacing the terminal water molecule on Dy2 with its nitrogen (B′ and B positions, respectively, in Fig. S6). Thus, each {Dy43-OH)4}8+ fragment is surrounded by six ina ligands, four of which are bridging. Note that this change in the linking between the lanthanide fragments does not affect the geometry of the polymer chains (Fig. S10) or the frameworks topology. In 7, this switching occurs for approximately 60% of ina ligands in B position, while 8 and 9 exclusively contain configuration bridging “B” ligands. Each Ln3+ cation here remains eight-coordinate.

As a consequence, the structures of 1–9 demonstrate an increase in the “degree of polymerization” between polynuclear fragments {Ln43-OH)4}8+, caused by a decrease in the ionic radius and coordination number of lanthanide cations (Fig. 4). It is important to note that significant changes in the coordination of isonicotinate do not lead to a change in the overall structure of the frameworks. This can be explained by the structure-directing role of the [Re6S8(CN)6]4– cluster anion, whose coordination environment remains unchanged. In a polyhedral representation, it is clear that each cluster [Re6S8(CN)6]4– fits the pseudo-octahedral environment of the {Pr43-OH)4}8+ cubane fragments, and vice versa (Fig. S11). This packing of cluster units can be simplified to the face-centered cubic NaCl structure (Fm[3 with combining macron]m).

image file: c9qi00339h-f4.tif
Fig. 4 Three ways of binding between {Ln43-OH)4}8+ fragments found in the structures of compounds 1–9; color code as in Fig. 2 except the coloring of disordered ina linker in red and blue in b. (a) One bridging ina ligand per {Ln4} unit form three Ln–O bonds in the structures 1–4. (b) One disordered bridging ina ligand per {Ln4} unit form one or two Ln–O bonds in the structures 5, 6. (c) Two bridging ina ligands per {Ln4} unit in the structures 7–9 (c).

Powder X-ray diffraction and thermal stability

According to the PXRD patterns (Fig. S12), compounds 1–9 are phase pure and air-stable. However, grinding of the samples is accompanied by their amorphization, possibly due to the local heating and loss of solvate water molecules. Thermogravimetric analysis showed that solvate H2O molecules leave the structure in the temperature range 25–150 °C, while H2O ligands coordinated to the Ln atoms can be removed at about 150–260 °C (Table S4, Fig. S13–S17). After loss of solvate H2O molecules, the PXRD patterns of all compounds 1–9 become amorphous.

Luminescent properties

Similarly to the previously investigated MOFs based on luminescent octahedral rhenium clusters and rare-earth cations,32 compound 1–9 have not shown sharp emission lines typical for lanthanide ions. The effective quenching of the emission by the vibrational modes of the OH groups of coordinated H2O ligands is the most likely reason for this. We can suppose that obtaining of coordination polymers having emission of both lanthanide ions and cluster anions is possible in non-aqueous media. Luminescence of the [Re6S8(CN)6]4– cluster anion is presented as the broad emission in the range of 600–900 nm with the maximum at about 725 nm (Fig. S18). The absolute quantum yields do not reach 0.01, which is notably less than characteristic quantum yield of the ionic compounds based on the [Re6S8(CN)6]4– cluster.46

Magnetochemical analysis

Magnetic susceptibility data are presented as χmT vs. T plots at 0.1 T and Mm vs. B plots at 2.0 K in Fig. 5. The χmT values at 290 K are 6.03 (1), 6.18 (2), 1.36 (3), 6.14 (4), 31.31 (5), 47.26 (6), 53.21 (7), 53.58 (8) and 44.48 cm3 K mol−1 (9). All these values are within the range expected47 for the four non-interacting respective Ln3+ centers: 5.78–6.48 (Ln = Pr), 5.78–6.23 (Nd), ca. 1.32 (Sm), ca. 6.15 (Eu), 30.43–31.45 (Gd), 47.06–48.04 (Tb), 52.04–56.20 (Dy), 53.07–55.15 (Ho) and 44.20–45.14 cm3 K mol−1 (Er), respectively. Upon cooling the compounds, the general features of the χmT vs. T curves are similar for all compounds, with exception of 3 and 4. For the other compounds, with decreasing temperature χmT is almost constant down to 100 K, subsequently slightly decreases and finally drops off for temperatures below 30 K reaching 0.29 (1), 1.91 (2), 0.19 (3), 0.21 (4), 16.70 (5), 26.42 (6), 28.29 (7), 17.74 (8), 18.90 cm3 K mol−1 (9) at 2.0 K. The slight decrease is mainly due to the thermal depopulation of the mJ substates introduced by single-ion effects characteristic for 4f ions. Specifically the ground terms of 2J + 1 multiplicity are split (and mixed) due to interelectronic repulsion, the electrostatic ligand field and spin–orbit coupling, which results in anisotropic magnetic centers.
image file: c9qi00339h-f5.tif
Fig. 5 Magnetic data of compounds 1–9: (a) χmT vs. T at 0.1 T (color code: see b); (b) molar magnetization Mm vs. applied field B at 2.0 K (symbols: experimental data, solid lines: least-squares fit to data of 5 using an effective spin model (S = 7/2 per Gd3+ center).

This is apparent from a comparison of the χmT vs. T curve of 5 with those of the other compounds, since Gd3+ centers behave as spin-like centers, while all other do not. In this temperature range (100–30 K), there are also small contributions of weak and predominantly antiferromagnetic exchange interactions that are the main cause of the χmT drop-off below 30 K, to which also the Zeeman effect at 0.1 T and T < 10 K contributes, albeit to a minor degree. The data of compounds 3 and 4 show different features due to the 2J + 1 ground multiplets of Sm3+ and Eu3+, respectively, distinctly mixing with some excited terms of almost similar energy, and the Eu3+ center moreover characterized by an mJ = 0 ground state.

The magnetization plots (Mm vs. B, Fig. 5b) at 2.0 K are in agreement with the absence of, or with the presence of only weak, exchange interactions: the magnetization curves rapidly increase up to an applied field of 0.5–1.0 T, and noticeably reduce this increase at higher fields. At 5 T the values of Mm are 1.1 (1), 4.5 (2), 0.5 (3), 0.4 (4), 26.6 (5), 19.6 (6), 21.0 (7), 20.4 (8), and 21.2NAμB (9). The value for 5 (Gd3+) is close to the saturation value of 27.9NAμB anticipated for four non-interacting Gd3+ centers (4 × geff·S NAμB, S = 7/2), while for the other compounds the values reach about half of 4 × gJ·J NAμB (6–9) or less (1–2).

Due to absence of significant exchange interactions, the magnetization of the (nearly) isotropic, spin-like Gd3+ centers reaches almost saturation at 5.0 T and 2.0 K, while the other, anisotropic centers show the magnetization of a powder sample that is a statistical mean of all magnetic axes yielding the above-stated rule of thumb for the later lanthanides. In particular, this rule fails for compounds 3 and 4 containing Sm3+ and Eu3+ centers, respectively, due to the afore-mentioned reasons regarding mixing or an mJ = 0 ground state.

We estimated the potential exchange interactions between the four centers of the {Ln4} units by an analysis of 5 as a representative and, from a magnetochemical point of view, simple system. Although the other centers are anisotropic, the sign and thus the nature of interaction (ferro- vs. antiferromagnetic coupling) are most likely to be similar for all compounds. For 5, we assume an effective Hamiltonian accounting for four exchange-coupled isotropic spin-like S = 7/2 centers (‘−2J’ convention) and the Zeeman effect, and employ the effective spin option of the computational framework CONDON.48,49 The four centers form a distorted tetrahedron with each edge approximately bridged by two oxides (belonging to different ligands). We started with a model assuming all six exchange interaction pathways to be identical (model A), and further refined this model based on the distances from structural information (model B, see Fig. S19). Simultaneously fitting the χmT vs. T and Mm vs. B data yields the parameters shown in Table 1. The effective g-factor geff is slightly below the free electron factor ge due to the well-known mixing of 6P7/2 (and other terms) into the Gd3+ ground term 8S7/2 caused by spin–orbit coupling.50,51 No fit to any of the two exchange pathway models (nor to others, which we tried but do not present) prefers a positive parameter J, i.e. ferromagnetic exchange interaction. All interactions are very weak and antiferromagnetic. Since model B yields a slightly better fit quality than model A (as quantified by SQ, the relative root mean square error), the solid lines shown in Fig. 5 represent the parameters of model B. While the sum of all Jij in both models are similar, the distribution of model B reveals a hierarchy that roughly goes inversely proportional with the distance.

Table 1 Parameters of the least-squares fits according to model A and B; Heisenberg exchange Hamiltonian using the ‘−2J’ convention (see Fig. S19 for more details)
  Model A Model B
geff 1.987 ± 0.001 1.990 ± 0.005
J24/cm−1 (= J34) −0.034 ± 0.002 −0.069 ± 0.002
J12/cm−1 (= J13) = J24 −0.024 ± 0.004
J14/cm−1 (= J23) = J24 −0.012 ± 0.002
SQ 1.3% 0.9%

We also measured the ac susceptibility of the compounds in a dynamic magnetic field at three different static bias fields (zero, 0.1 and 0.3 T). Most compounds (1–5, 8, 9) showed no relevant out-of-phase signals in the ranges 3–997 Hz and 1.9–50 K under these conditions. Compounds 6 and 7 showed similar signals, which were for 6, however, of very small magnitude. Therefore, we exemplarily present the data of 7, although a quantitative analysis cannot be performed as evident from Fig. 6 and Fig. S20–S21. Besides the magnitude being rather small in the measured ranges, neither exhibit maxima in their χ′′m vs. f plots (but absolute maxima at the border of the probed frequency range), nor do the χ′′m vs. χm plots show distinct curvatures. Thus, we cannot apply the formula 2πfmaxτ = 1 using the χ′′m vs. f plots, nor can we use a generalized Debye expression to fit the χ′′m vs. χm plots and derive physically reasonable relaxation parameters.

image file: c9qi00339h-f6.tif
Fig. 6 Magnetic ac data of 7 at 0.3 T static bias field: (a) out-of-phase molar susceptibility χ′′m vs. frequency f (lines are guides for the eyes); (b) out-of-phase χ′′m vs. in-phase molar susceptibility χm (symbol/color codes in (a) for both plots).

We, thus, only summarize the following qualitative observations: first, the low frequency χ′′m values increase with the static bias field, but there is also a temperature dependence of this effect. Second, the maxima in the χ′′m vs. χm plots also depend on the static bias field: with higher bias field, the maximum χ′′m value is (slightly) shifted to higher temperatures.


In summary, we have presented the first metal–organic frameworks based on polynuclear lanthanide fragments and cyanometallate building blocks. Compounds 1–9, isolated as [{Ln43-OH)4(ina)4(H2O)n}{Re6S8(CN)6}]·mH2O (Ln = Pr – Er, ina = isonicotinate), form via the self-assembly of Ln3+ cations, the [Re6S8(CN)6]4– cluster anion and isonicotinate under hydrothermal conditions. Our synthesis strategy combined two extensions of classic synthesis approaches based on controlled hydrolysis of lanthanide ions. First, employing isonicotinamide as the precursor of the isonicotinate linker is crucial to the formation of crystalline products 1–9. This suggests that hydrolysis of amides under controlled temperature and pH generally is a promising method for preparation of the lanthanide MOFs with carboxylate linkers. Second, the pre-synthesized cluster anion [Re6S8(CN)6]4– was successfully used as secondary building unit. This octahedral cluster can act not only as a highly symmetric structural fragment, but also as a functional SBU with a red luminescence and an ability to be reversibly oxidized. Here, it acts as structure-directing metallolinker, effectively joining the 1D chains formed by {Ln43-OH)4}8+ cubane fragments and ina moieties. As a result, changes in the coordination environments of the lanthanide ions, their coordination numbers and the number of bridging ina moieties do not translate into a change of the framework topology. The frameworks demonstrate a combination of properties related to the individual building units, e.g. red luminescence of rhenium cluster and magnetic properties of {Ln4(μ-OH)4}8+ complexes. Magnetochemical analysis revealed that paramagnetic centers in compounds 1–9 interact weakly antiferromagnetic. Compounds 6 and 7 (Ln = Tb and Dy, respectively) exhibit small out-of-phase ac susceptibility components at lowest temperatures (down to 2 K). Given the prototypal character of the above-mentioned synthetic strategies, in future studies we aim to expand the scope of reagents in order to construct a wider variety of functional lanthanide framework architectures.

Conflicts of interest

There are no conflicts to declare.


This work is supported by the Russian Foundation for Basic Research (project 18-29-04007). Y. V. Z. is indebted to Ministry of Science and Higher Education of the Russian Federation (project AAAA-A19-119020890025-3). The authors are grateful to Dr K. A. Brylev (NIIC SB RAS) for the luminescence measurements.

Notes and references

  1. T. Han, Y.-S. Ding and Y.-Z. Zheng, in Recent Development in Clusters of Rare Earths and Actinides: Chemistry and Materials, ed. Z. Zheng, Springer Berlin Heidelberg, Berlin, Heidelberg, 2017, pp. 209–314,  DOI:10.1007/430_2016_9.
  2. S.-D. Han, X.-H. Miao, S.-J. Liu and X.-H. Bu, Inorg. Chem. Front., 2014, 1, 549–552 RSC.
  3. D. N. Woodruff, R. E. P. Winpenny and R. A. Layfield, Chem. Rev., 2013, 113, 5110–5148 CrossRef CAS PubMed.
  4. P. Zhang, Y.-N. Guo and J. Tang, Coord. Chem. Rev., 2013, 257, 1728–1763 CrossRef CAS.
  5. X.-Y. Zheng, X.-J. Kong, Z. Zheng, L.-S. Long and L.-S. Zheng, Acc. Chem. Res., 2018, 51, 517–525 CrossRef CAS PubMed.
  6. Y.-C. Chen, J.-L. Liu and M.-L. Tong, in Recent Development in Clusters of Rare Earths and Actinides: Chemistry and Materials, ed. Z. Zheng, Springer Berlin Heidelberg, Berlin, Heidelberg, 2017, pp. 189–207,  DOI:10.1007/430_2015_5001.
  7. J.-L. Liu, Y.-C. Chen, F.-S. Guo and M.-L. Tong, Coord. Chem. Rev., 2014, 281, 26–49 CrossRef CAS.
  8. M. Pan, W.-M. Liao, S.-Y. Yin, S.-S. Sun and C.-Y. Su, Chem. Rev., 2018, 118, 8889–8935 CrossRef CAS PubMed.
  9. T.-Y. Luo, C. Liu, S. V. Eliseeva, P. F. Muldoon, S. Petoud and N. L. Rosi, J. Am. Chem. Soc., 2017, 139, 9333–9340 CrossRef CAS PubMed.
  10. X. P. Yang, S. Q. Wang, C. R. Wang, S. M. Huang and R. A. Jones, in Recent Development in Clusters of Rare Earths and Actinides: Chemistry and Materials, ed. Z. Zheng, 2017, vol. 173, pp. 155–187 Search PubMed.
  11. Y.-Y. Zhou, B. Geng, Z.-W. Zhang, Q. Guan, J.-L. Lu and Q.-B. Bo, Inorg. Chem., 2016, 55, 2037–2047 CrossRef CAS PubMed.
  12. F. Le Natur, G. Calvez, J.-P. Guégan, L. Le Pollès, X. Trivelli, K. Bernot, C. Daiguebonne, C. Neaime, K. Costuas, F. Grasset and O. Guillou, Inorg. Chem., 2015, 54, 6043–6054 CrossRef CAS PubMed.
  13. 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, 7454–7457 CrossRef CAS PubMed.
  14. J. Wahsner, E. M. Gale, A. Rodríguez-Rodríguez and P. Caravan, Chem. Rev., 2019, 119, 957–1057 CrossRef CAS PubMed.
  15. M. C. Heffern, L. M. Matosziuk and T. J. Meade, Chem. Rev., 2014, 114, 4496–4539 CrossRef CAS PubMed.
  16. X.-Y. Zheng, J. Xie, X.-J. Kong, L.-S. Long and L.-S. Zheng, Coord. Chem. Rev., 2019, 378, 222–236 CrossRef CAS.
  17. G. Calvez, F. Le Natur, C. Daiguebonne, K. Bernot, Y. Suffren and O. Guillou, Coord. Chem. Rev., 2017, 340, 134–153 CrossRef CAS.
  18. L. Chen, F. Jiang, K. Zhou, M. Wu and M. Hong, in Lanthanide Metal-Organic Frameworks, ed. P. Cheng, Springer Berlin Heidelberg, Berlin, Heidelberg, 2015, pp. 145–183,  DOI:10.1007/430_2014_161.
  19. L. Chen, F. L. Jiang, K. Zhou, M. Y. Wu and M. C. Hong, in Lanthanide Metal-Organic Frameworks, ed. P. Cheng, 2015, vol. 163, pp. 145–183 Search PubMed.
  20. F. Le Natur, G. Calvez, C. Daiguebonne, O. Guillou, K. Bernot, J. Ledoux, L. Le Pollès and C. Roiland, Inorg. Chem., 2013, 52, 6720–6730 CrossRef CAS PubMed.
  21. Y. Chen and S. Q. Ma, Rev. Inorg. Chem., 2012, 32, 81–100 CAS.
  22. G. Calvez, C. Daiguebonne and O. Guillou, Inorg. Chem., 2011, 50, 2851–2858 CrossRef CAS PubMed.
  23. F. Shi, L. Cunha-Silva, T. Trindade, F. Paz and J. Rocha, Cryst. Growth Des., 2009, 9, 2098–2109 CrossRef CAS.
  24. O. A. Gerasko, E. A. Mainicheva, M. I. Naumova, M. Neumaier, M. M. Kappes, S. Lebedkin, D. Fenske and V. P. Fedin, Inorg. Chem., 2008, 47, 8869–8880 CrossRef CAS PubMed.
  25. L. G. Beauvais, M. P. Shores and J. R. Long, Chem. Mater., 1998, 10, 3783–3786 CrossRef CAS.
  26. Y. V. Mironov, J. A. Cody, T. E. AlbrechtSchmitt and J. A. Ibers, J. Am. Chem. Soc., 1997, 119, 493–498 CrossRef CAS.
  27. E. V. Alexandrov, A. V. Virovets, V. A. Blatov and E. V. Peresypkina, Chem. Rev., 2015, 115, 12286–12319 CrossRef CAS PubMed.
  28. Y. Kim, S.-M. Park, W. Nam and S.-J. Kim, Chem. Commun., 2001, 1470–1471 RSC.
  29. M. S. Tarasenko, N. G. Naumov, D. Y. Naumov, S.-J. Kim and V. E. Fedorov, Polyhedron, 2008, 27, 2357–2364 CrossRef CAS.
  30. Y. Molard, A. Ledneva, M. Amela-Cortes, V. Cırcu, N. G. Naumov, C. Meriadec, F. Artzner and S. Cordier, Chem. Mater., 2011, 23, 5122–5130 CrossRef CAS.
  31. M. Amela-Cortes, S. Cordier, N. G. Naumov, C. Meriadec, F. Artzner and Y. Molard, J. Mater. Chem. C, 2014, 2, 9813–9823 RSC.
  32. Y. M. Litvinova, Y. M. Gayfulin, K. A. Kovalenko, D. G. Samsonenko, J. van Leusen, I. V. Korolkov, V. P. Fedin and Y. V. Mironov, Inorg. Chem., 2018, 57, 2072–2084 CrossRef CAS PubMed.
  33. Z. Zheng, Chem. Commun., 2001, 2521–2529 RSC.
  34. N. G. Naumov, S. B. Artemkina, A. V. Virovets and V. E. Fedorov, Solid State Sci., 1999, 1, 473–481 CrossRef CAS.
  35. CrysAlisPro, Rigaku Oxford Diffraction, 2015 Search PubMed.
  36. T. G. G. Battye, L. Kontogiannis, O. Johnson, H. R. Powell and A. G. W. Leslie, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2011, 67, 271–281 CrossRef CAS PubMed.
  37. P. Evans, Acta Crystallogr., Sect. D: Biol. Crystallogr., 2006, 62, 72–82 CrossRef PubMed.
  38. G. Sheldrick, Acta Crystallogr., Sect. A: Found. Adv., 2015, 71, 3–8 CrossRef PubMed.
  39. G. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71, 3–8 Search PubMed.
  40. R. S. Brown, A. J. Bennet and H. Slebocka-Tilk, Acc. Chem. Res., 1992, 25, 481–488 CrossRef CAS.
  41. P.-F. Shi, C.-S. Cao, C.-M. Wang and B. Zhao, Inorg. Chem., 2017, 56, 9169–9176 CrossRef CAS PubMed.
  42. P. S. Koroteev, N. N. Efimov, Z. V. Dobrokhotova, A. B. Ilyukhin, A. V. Gavrikov and V. M. Novotortsev, Russ. J. Coord. Chem., 2015, 41, 149–161 CrossRef CAS.
  43. T. Li, X.-J. Hong, X. Liu, R. Chen, Q.-G. Zhan, X. Xu and Y.-P. Cai, CrystEngComm, 2014, 16, 3883–3889 RSC.
  44. L. Wang, W. T. Yang, F. Y. Yi, H. Wang, Z. G. Xie, J. K. Tang and Z. M. Sun, Chem. Commun., 2013, 49, 7911–7913 RSC.
  45. P. M. Cantos, S. J. A. Pope and C. L. Cahill, CrystEngComm, 2013, 15, 9039–9051 RSC.
  46. T. G. Gray, C. M. Rudzinski, E. E. Meyer, R. H. Holm and D. G. Nocera, J. Am. Chem. Soc., 2003, 125, 4755–4770 CrossRef CAS PubMed.
  47. H. Lueken, Magnetochemie, Teubner Verlag, Stuttgart, 1999 Search PubMed.
  48. M. Speldrich, J. van Leusen and P. Kögerler, J. Comput. Chem., 2018, 39, 2133–2145 CrossRef CAS PubMed.
  49. M. Speldrich, H. Schilder, H. Lueken and P. Kögerler, Isr. J. Chem., 2011, 51, 215–227 CrossRef CAS.
  50. A. J. Shuskus, Phys. Rev., 1962, 127, 2022–2024 CrossRef CAS.
  51. J. Sytsma, K. M. Murdoch, N. M. Edelstein, L. A. Boatner and M. M. Abraham, Phys. Rev. B: Condens. Matter Mater. Phys., 1995, 52, 12668–12676 CrossRef CAS.


Electronic supplementary information (ESI) available: Structural and magnetic data, TGA graphs, PXRD patterns. CCDC 1904278–1904286. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9qi00339h

This journal is © the Partner Organisations 2019