Cages on a plane: a structural matrix for molecular ‘sheets’

A family of heterometallic Anderson-type 'wheels' of general formula [MM(hmp)12]4+ (MIII = Cr or Al and MII = Ni or Zn, Hhmp = 2-pyridinemethanol) has been extended to include MIII = Cr or Al and MII = Co, Fe, Mn or Cu, affording five new species of formulae [Cr2Co5(hmp)12](ClO4)4 (1), [Cr2Fe5(hmp)12](ClO4)4 (2), [Cr2Mn5(hmp)12](ClO4)4 (3), [Cr2Cu5(hmp)12](ClO4)2(NO3)2 (4) and [Al2Co5(hmp)12](ClO4)4 (5). As per previous family members, the metallic skeleton common to the cations of 1-5 describes a centred hexagon with the two MIII sites disordered around the outer wheel, with the exception of compound 4 where the CuII sites are localised. A structurally related, but enlarged planar disc possessing a [MMII] hexagon capped on each edge by a CuII ion can be formed, but only when MIII = Al and MII = Cu. In [AlCu(OH)12(hmp)12](ClO4)6(NO3)2 (6) the Anderson moiety contains a central, symmetry-imposed octahedral CuII ion surrounded by a wheel of AlIII ions. Solid-state dc susceptibility and magnetisation measurements reveal the presence of competing exchange interactions in 1-5, and very weak antiferromagnetic exchange between the CuII ions in 6 which may be intra- and/or intermolecular in nature.


hange effects
in heterometallic 3d-4d complexes, 4 and geometric spin frustration in antiferromagnetically coupled cages 5 and 2-3D materials (e.g. the kagomé lattice) possessing high symmetry. 6n 3d transition metal chemistry the molecular triangle is most commonly found in one of two structure types: (a) the oxo-centred planar triangle [M 3 O] n+ , as personified by the basic metal carboxylates, 7 where all four atoms lie on (or nearly on) the same plane, or (b) the [M 3 O 4 ] n+ partial cubane where the metal ions and O-atoms lie on different planes, i.e. a cube missing one metal vertex.The latter moiety also often acts as the building block for the creation of large and (occasionally) very large molecules whose structures conform to molecular 'sheets', i.e. the metallic skeleton of the complex grows in 2D.From a structural/synthetic perspective this is simple to understand as a series of O-bridged, edge-and vertex-sharing metal triangles (Fig. 1).For example, two edgesharing triangles form tetranuclear [M 4 O 2 ] n+ or [M 4 O 6 ] n+ butterflies or partial cubanes (Fig. 1a and b), with detailed magneto-structural correlations developed for Fe 8 and Mn. 9 Such triangles and butterflies/partial cubanes are by far the most common building blocks seen in large cages containing multiple 3d M II=III n ions (n > 4).Continued edge-sharing growth in just one dimension/ direction from triangle to butterfly/partial cubane to larger species results in the formation of molecular rods (Fig. 1c), a pertinent example being the use of tripodal alcohol ligands to direct the formation of Mn 6 , Mn 7 , Mn 8 , Mn 12 complexes. 10rowth in two dimensions/directions leads to planar disc-like complexes (Fig. 1d-i), the most common of which is the Anderson-type wheel.This structure describes a centred hexagon, with homometalic, 11 heterometallic, 12 homovalent 13 and heterovalent 14 examples known.Larger complexes are somewhat unusual, but are all characterised by beautiful structural aesthetics, the presence of the Anderson moiety at the core of their metallic skeletons, and interesting physical properties.For example, [Ni 10 ] (Fig. 1e) is a rare example of a large nuclearity Ni single-molecule magnet (SMM), 15 mixed-valent [Co 13/14 ] cages (Fig. 1f and g) display ferromagnetic exchange interactions between the Co II ions, 16 [Fe 17 /19 ] is an example of a trapped/molecular mineral phase with S ≥ 33/2, 17 two [Mn 19 ] cages possess a similar brucite-like core (Fig. 1h), one displaying intramolecular ferrimagnetic exchange and long range magnetic order, 18a and the other being a very rar example of a Mn-alkoxide, while [Co 24 ] was the first polynuclear Co II species to exhibit slow relaxation of the magnetization (Fig. 1i). 19It is also interesting to note a common thread in the synthesis of each of these species: the use of alkoxide-based bridging ligands.

We recently reported a small family of Anderson-type complexes of general formula [M III 2 M II 5 (hmp) 12 ] 4+ (M III = Cr or Al and M II = Ni or Zn, Hhmp = 2-pyridinemethanol) in which the two M III sites were disordered around the outer wheel. 20The relative ease of synthesis of these species and their stability in both the solid and solution state suggested that more family members could be

ade simply by changing
the identity of both the M III and M II ions.Herein we report expansion of this family to include M II = Cu, Co, Mn and Fe, and M III = Al and Cr, alongside the serendipitous self-assembly of the related, but larger complex [Al III 6 Cu II 7 (OH) 12 (hmp) 12 ](ClO 4 ) 6 (NO 3 ) 2 .


Experimental


Materials and physical measurements

All chemicals were procured from commercial suppliers and used as received (reagent grade Fe(ClO 4 ) 2 •6H 2 O (0.363 g, 1 mmol) and Cr(ClO 4 ) 3 •6H 2 O (0.229 g, 0.5 mmol) were dissolved with NaOMe (0.162 g, 3 mmol) in MeOH (24 ml) to give a dark red solution.Upon full dissolution, Hhmp (0.285 ml, 3 mmol) was added dropwise and the reaction left overnight with continuous stirring.12 ml samples of the resulting dark brown solution were heated

ernight with continuous
stirring.12 ml samples of the resulting dark purple/red solution were heated in Teflon-lined autoclaves at 100 °C for 12 hours.After slowly cooling to room temperature the reaction vessels were allowed to sit undisturbed for 24 hours yielding pale purple hexagonal crystals suitable for X-ray diffrac-


X-ray crystallography

Single crystal X-ray diffraction data for samples 1-6 were collected using a Rigaku Oxford Diffraction SuperNova diffractometer with MoK α (1 & 5-6) or CuK α (2-4) radiation.

Experimental details are given in Table S1 in the ESI.† An Oxford Cryosystems Cryostream 700+ low temperature device was used to maintain a crystal temperature of 120.0 K for all experiments.2][23] All nonhydrogen atoms were refined using anisotropic displacement parameters.H atoms were placed in calculated positions geometrically and refined using the riding model except for some in compound 6 which were refined freely.CCDC 1855222-1855227.†


Magnetic data

Magnetic susceptibility and magnetisation measurements were performed on powdered, polycrystalline samples of 1-6 in the T = 2-300 K and B = 0-7 T temperature and field ranges on a Quantum Design MPMS XL SQUID magnetometer equipped with a 7 T dc magnet.Hexadecane was employed to prevent potential torquing of the crystallites.Diamagnetic corrections were applied to all data using Pascal's constants.


EPR spectroscopy

EPR spectra of 6 were measured at Q-band on a Bruker EMX spectrometer.


Results and discussion


Structural description

There are two unique structure types present in 1-6 compounds 1-5 possess the [M 7 ] Anderson-type structure, while 6 is an [M 13 ] cluster containing an Anderson core capped on each of its six edges by an additional metal ion.

Crystallographic details for all complexes are given in Table S1, † with pertinent bond lengths and angles provided in Tables 1-3.

Table 1 Pertinent structural parameters for the M central -M outer dialkoxo bridge in 1-5.We begin with a generic description of complexes 1-5.Complexes 1-3 and 5 are isostructural, crystallising in the trigonal space group R3 ˉwith the asymmetric unit (ASU) containing only the central metal ion, one outer metal ion, two hmp − ligands and two ClO 4 − anions.The structure (Fig. 2 and 3) is that of a centred metal hexagon in which the two M III ions are disordered around the outer [M 6 ] wheel.There are therefore two distinct metal sites in the [M III 2 M II 5 ] cluster, the central metal ion is always an M II ion (Co (1, 5), Fe (2), Mn (3)), which is bridged to the outer metal ions by six symmetry equivalent µ 3 -OR groups from six hmp − ligands.The central ion thus has a symmetry imposed, octahedral (D3 d ) [M II O 6 ] coordination sphere.The outer metal ions are all also symmetry equivalent, crystallographic disorder resulting in the M III ions being equally distributed around all six positions, each with a 2/3 M II , 1/3 M III occupancy, with an average charge of +2.33.This was modelled as a 5 : 2 substitutional disorder ratio of metal centres by splitting the unique site into two separate parts with identical, constrained co-ordinates and anisotropic displacement parameters, and by fixing the occupancies such that they sum to give a 5 : 2 ratio of M II to M III .The disorder gives three distinct structural isomers with the M III ions occupying outer ring positions 1,2 1,3 or 1,4 in a ratio of 2 : 2 : 1 (Fig. 3).
r = M-O bond length, ϕ = M-O-M bridging angle M-M [Å] r [Å] ϕ [°]1
Around the ring, the metal ions are connected

y one µ-OR (h
p − ) group on the 'outside' of the wheel and one µ 3 -OR (hmp − ) group on the 'inside' of the wheel.Two terminally bonded N-atoms from the hmp − ligands complete the octahedral coordination spheres on each metal ion.A total of twelve hmp − ligands therefore 'frame' the metal-oxygen core, six sitting above and six sitting below the metal ion plane.Charge balance is maintained through the presence of four Compound 4 (Fig. 5) crystallises in the monoclinic space group I2/a, with half the molecular formula in the ASU.The structure is analogous to that seen for 1-3 and 5 but with the important exception that the two Cr III sites in the outer wheel are now not disordered, instead being localised in the 1,4 positions, i.e. trans to each other.The reason for this, and the lowering of crystallographic symmetry, is not clear but may be associated with the presence of Jahn-Teller (JT) distortions at the four peripheral Cu II      ) of ∼3 Å.Note that the closest intermolecular Cu⋯Cu distance is ∼8.5 Å (see magnetism and EPR sections below).


Magnetometry

Dc magnetic susceptibility (χ M ) measurements were carried out on powdered polycrystalline samples of compounds 1-6 in a B = 0.1 T applied magnetic field over the temperature range T = 2-300 K, and are plotted as the χ M T product versus T in Fig. 7  and 8.

For complexes 1-5 the experimental room temperature values of χ M T are close to the Curie constants expected for five   and two non-interacting M II and M III ions, respectively; 1: 19.6 cm 3 K mol −1 (expected 16.2 cm 3 K mol −1 , g Cr = 2.00, g Co = 2.30); 2: 17.7 cm 3 K mol −1 (expected 18.2 cm 3 K mol −1 , g Cr = 2.00, g Fe = 2.20); 3: 25.4 cm 3 K mol −1 (expected 25.6 cm 3 K mol −1 , g Cr = g Mn = 2.00); 4: 6.1 cm 3 K mol −1 (expected 6.0 cm 3 K mol −1 , g Cr = 2.00, g Cu = 2.20); 5: 13.7 (expected 12.4 cm 3 K mol −1 , g Cr = 2.00, g Co = 2.30).The temperature dependence of χ M T for all five complexes down to approximately T ≈ 25 K is rather similar, all decreasing slowly with decreasing temperature.For complex 1 the value of χ M T then plateaus at a value of 17.0 cm 3 K mol −1 , before decreasing to a value of 14.2 cm 3 K mol −1 at 2 K.For complexes 3 and 5 the value of χ M T increases to maximum values of 19.7 and 14.4 cm 3 K mol −1 , respectively.For complexes 2 and 4 the value of χ M T continues to decrease, reaching T = 2 K values of 7.8 and 0.5 cm 3 K mol −1 , respectively.The behaviour in each case is therefore consistent with the presence of competing exchange interactions, as observed and quantified for the structurally analogous [Cr 2 Ni 5 (hmp) 12 ] 4+ family of complexes. 20The positional disorder of the Cr III ions and resulting different isomers, the large number of different exchange interactions and, in the case of complexes, 1, 2, 5, the zero-field splitting effects of the M II ions precludes any detailed/quantitative analysis of the susceptibility data.Magnetisation (M) versus field data, collected for 1-5 in the T = 2-7 K and B = 0.5-7 T temperature and field ranges (Fig. S6-S10 †) are consistent with this picture, in each case M rising rapidly with increasing B without reaching saturation.

The dc susceptibility and magnetisation data for complex 6 is shown in Fig. 8.The high temperature χ M T value of 3.06 cm 3 K mol −1 is close to that expected for seven non-interacting (s = 1 2 mol −1 ).This value remains constant in the T = 400-25 K temperature regime, before falling to a value of 1.7 cm 3 K mol −1 at T = 2 K.This is consistent with the presence of very weak antiferromagnetic exchange interactions between the Cu II

ons, as would be
xpected from the presence of a 3-atom (Cu-O-M-O-Cu) bridge between neighbouring paramagnetic sites. 24The data is invariant in measurements performed at different field strengths (Fig. S11 †).The χ M T and magnetisation data were fitted simultaneously using isotropic spin-Hamiltonian (1) and the exchange interaction scheme depicted in Fig. 9, where the indices i and j refer to the interacting Cu II ions, µ B is the Bohr magneton, B is the applied magnetic field, g is the g-factor of the Cu II ions (fixed from the EPR with g ∥ = 2.21 and g ⊥ = 2.06), Ŝ is a spin operator and J is the isotropic exchange interaction.Using this model, the best fit parameter was found to be J = −0.47cm −1 .This value is similar to that previously observed
Ĥ ¼ μ B B X i g i Ŝi À 2 X iÁj,i J ij Ŝi Ŝj
Given the very small value of J, fitting was also attempted using a model in which intermolecular interactions (see the EPR section below) were also included via a mean-field approach, but all solutions remained inferior to that given above.


EPR spectroscopy

EPR spectra of a powdered sample of complex 6, measured at Q-band (ca.34 GHz; Fig. 10), are consist

t with tetra
onal Cu(II) centres with near axially-symmetric g-values with "g ∥ " = 2.06 and "g ⊥ " = 2.21.There is no resolution of any fine structure and the spectra change little with variable temperature (beyond simple Curie behaviour), consistent with any intramolecular exchange interactions being very weak.However, there is no resolution of 63,65 Cu hyperfine structure, hence the Cu ions are not magnetically dilute.At the g ∥ region, where the hyperfine interaction would be at its largest for tetragonal Cu(II), the (Lorentzian) linewidth (ca. 4 mT) is much narrower than the expected spread of the hyperfine multiplet (50-60 mT for A ∥ = 0.015-0.02cm −1 ): this is characteristic of an exchange narrowing regime where the intermolecular interactions in the lattice are comparable to the hyperfine interaction.Hence, care

hould be taken in inte
preting the bulk magnetic properties of 6 from