Synthesis and Magnetic Properties of a Copper Cube: [Cu4(OH)4(C16H18N2)4]4+ (ClO4)4 C3H6O [C16H18N2=(E)-1,6-[Di(pyridin-4-yl)hex-3-ene].

The synthesis of a Cu4(OH)4 cube which is coordinated by four molecules of the dipyridyl ligand 1,6-[di(pyridin-4-yl)hex-3-ene] is reported. This compound has a trans double bond which restricts the conformational freedom of the ligand and favours coordination within a unique copper cube. The structure was solved by an X-Ray single crystal structure determination and low temperature magnetic susceptibility measurements examined its magnetic properties. The cube classification corresponds to the type I classification of Mergehenn and Haase and the short/long distribution of Cu ⋅⋅⋅ Cu separations in the cube as defined by Ruiz. The magnetic susceptibility measurements show paramagnetic behaviour down to 50 K but below this the copper cube shows weak ferromagnetic exchange interactions. The low temperature magnetic susceptibility characteristics are examined in detail then modelled and compared to other similar Cu4O4 copper cubes.

Introduction

Clusters of transition metal‐ions have potential applications in transistors, magnetic memory, molecular wires, logic gates and other molecular devices.1 Multi‐nuclear copper (II) complexes are of interest for molecular devices and in the active sites of copper oxidases.2, 3, 4 A common arrangement corresponds to cubane like structures which contain a Cu4O4 core. The oxygen bridges are hydroxo or alkoxo groups. Two different classifications have been proposed which are either based on the distribution of long Cu−O bond lengths in the cube, by Mergehenn and Haase (type I, II and III),5 or on the number of short and long Cu−Cu distances by Ruiz (2+4, 4+2 or 6+0 classes)6, 7 (Figure 1). Compounds of type I or class 2+4 contain 2 short and 4 long Cu−Cu distances. Complexes of type II or class 4+2 have 4 short and 2 long Cu−Cu distances. The third class are 6+0 compounds which contain six similar Cu−Cu distances.

Figure 1

The classification of Cu4O4 cubanes. The bold lines are short Cu−O bond lengths and the dashed lines are long Cu−O bond lengths .5, 6, 7

The specific geometry of the Cu4O4 core can lead to ferromagnetic as well as antiferromagnetic interactions.8 Small variations in their structure leads to tunable magnetic behaviour.9, 10, 11 Much effort has been made to understand the magnetic properties with the Cu4O4 structure and bond lengths.12, 13, 14 Other interesting Cu cubane custers are also known.15, 16, 17, 18, 19 In this paper a new Cu4(OH)4 cube is described and analysed.

Experimental Section

General

IR spectra were recorded on a diamond anvil spectrophotometer. UV/Vis spectra were recorded using a Perkin‐Elmer Lambda spectrometer with EtOH as the solvent. The crystal field stabilisation energy (CFSE) parameter in KJ mol−1=0.119/λmax (meters). 0.119=Na×h×c (Avogadro's number×Planck's constant×speed of light).

[Cu4(OH)4(C16H18N2)4]4+ (ClO4)4 C3H6O 1

(E)‐1,6‐[di(pyridin‐4‐yl)hex‐3‐ene] 2 (Figure 2)20, 21 (50 mg, 0.21 mmol) was dissolved in EtOH (3 ml) and acetone (3 ml) then layered onto CuClO4 . 6H2O (78 mg, 0.21 mmol) in water (7 ml) in an 8×2.5 cm sample vial. After 24 h nucleation of blue crystals had started around the sides and the layers were gently mixed. Deep blue crystals were carefully collected (43 mg, 47.4 %) after two days and briefly air dried. The crystals were collected from the sides of the sample vial after pouring off the supernatant solvent into a beaker and washing the sample vial with fresh solvent of the same composition used in the synthesis. The crystals were also collected from the supernatant solvent by decanting off the solvent from the beaker and adding more fresh solvent to decant. IR (Diamond anvil): =3502 (w), 2934 (w), 1710 (w), 1619 (s), 1559 (w), 1506 (w), 1431 (s), 1229 (w), 1062 (s), 975 (s), 930 (w), 828 (s), 621 (s), 589 (s) and 528 (s) cm−1; After storage for 4 months: =2938 (w), 1619 (s), 1558 (w), 1508 (w), 1433 (s), 1230 (w), 1109 (vs), 1050 (vs), 968 (w), 930 (w), 812 (s), 620 (vs), 585 (s), 523 (s) and 493 (w); UV/Vis (EtOH): λmax (ϵ)=420–380 (2594), 362 (2939), 257 nm (6256 mol−1dm3cm−1); UV/Vis (DMF): λmax (ϵ)=690 nm (117.6 mol−1 dm3 cm−1); elemental analysis calcd (%) for C67H79Cl4Cu4N8O21: C 46.50, H 4.57, N 6.48; found: C 46.22, H 4.60, N 6.46.

Figure 2

Drawing of (E)‐1,6‐[di(pyridin‐4‐yl)hex‐3‐ene] 2.

Intensity data for 1 were collected at T=100 K using a Rigaku AFC11 CCD diffractometer (Mo Kα radiation, λ=0.71073 Å) and the structure was easily solved by direct methods and completed and optimised by least‐squares refinement against |F|2 using SHELXL‐2014.22 The O‐bonded H atom was located in a difference map and refined as riding in its as‐found relative position. The C‐bound H atoms were geometrically placed (C−H=0.95–0.99 Å) and refined as riding atoms. The methyl groups were allowed to rotate, but not to tip, to best fit the electron density. The constraint U iso(H)=1.2U eq(C) or 1.5U eq (methyl C) was applied in all cases. The O atoms of one of the perchlorate anions are statistically disordered over two sets of sites.

1 C76H100Cl4Cu4N8O24, M r=1905.59, blue block, 0.16×0.10×0.08 mm, tetragonal, space group P 21 c (No. 114), Z=2, a=14.69480 (10) Å, c=19.0833 (2) Å, V=4120.79 (7) Å3. Number of measured and unique reflections=38188 and 4679, respectively (−17≤h≤19, −19≤k≤19, −24≤l≤25; 2θmax=54.9°; R Int=0.033). Final R(F)=0.028, wR(F 2)=0.064 for 263 parameters and 4616 reflections with I >2σ(I) (corresponding R‐values based on all 4679 reflections=0.028 and 0.065, respectively). Flack absolute structure parameter=0.000 (4), CCDC deposition number 1891223.

Synthesis of Copper Cube 1

The ligand 1,6‐[di(pyridin‐4‐yl)hex‐3‐ene]20, 21 was dissolved in a mixture of acetone and ethanol then layered onto a solution of Cu(ClO4)2 dissolved in water. After four days blue crystals were growing around the sides of the sample vial and some were on the base. The solution, which had a slight haze, was decanted carefully and the sample vial was washed with water. The decanted solution, in a beaker, was also decanted again leaving more blue crystals which were washed with water. The air dried crystals were combined and characterised by IR, UV/Vis, microanalysis, an X‐ray single crystal structure determination, magnetic susceptibility measurements and epr spectroscopy.

Results and Discussion

The IR spectrum of the copper cube showed a weak carbonyl stretch at 1710 cm−1 owing to the presence of acetone (1715 cm−1) in the lattice (Figure S1 and S2). The absorption for a carbonyl group would normally be strong but here it may be weak because by mass acetone is only 3 % of the compound. However, after storage for a few months this peak vanished presumably because the lattice desolvated. The UV/Vis spectrum was weak owing to poor solubility of the copper cube in EtOH (Figure S3). It showed an absorption maximum at 362 nm with a broad shoulder at 380–420 nm but the spectrum was to weak to see the long wavelength blue absorption. However the blue crystals dissolved in DMF and showed a broad absorption maximum at 690 nm (Figure S4) with a crystal field stabilisation energy (CFSE) of 172.5 KJ/mol which is comparable with that for other Cu(II) ions.23 Over time they appeared to decompose as the solution turned hazy. The asymmetric unit of 1 consists of one Cu2+ ion, one (E)‐1,6‐di(pyridin‐4‐yl)hex‐3‐ene20, 21 2 (C16H18N2) ligand, one μ3‐OH− ion, two perchlorate ions (both with Cl site symmetry 2 and one featuring statistical disorder of its O atoms over two sets of sites) and one acetone solvent molecule of crystallisation. The symmetry elements in the tetragonal crystal [at ( , , ) for the asymmetric fragment] generate tetra‐nuclear [Cu4(OH)4(C16H18N2)4]4+ ‘cubane’ clusters accompanied by one ClO4 − counter ion and one C3H6O molecule per copper ion, for an overall chemical formula of [Cu4(OH)4(C16H18N2)4](ClO4)4 ⋅ 4(C3H6O) (Figure 3). The dihedral angle between the N1 and N2 pyridine ring is 28.33 (14)° and both the Ca−Cm−Cm−CH=(a=aromatic, m=methylene) fragments in the linking chain have gauche conformations [torsion angles=−60.2 (4) and 71.1 (3)°] to result in an approximate overall U shape for this species. The ligand has a trans double bond which lowers the entropy and restricts conformational freedom. This is expected to favour coordination within a unique copper cube. A similar crystallisation of 1,6‐di(pyridin‐4‐yl)hexane with Cu(NO3)2 ⋅ 6H2O gave two infinite crystalline phases24 suggesting that the double bond is important in favouring the macrocycle formation with Cu(II) salts and in reducing the potential for disorder of the ligand in the lattice.21 The flexible hexyl spacer (C6H12) can exhibit disorder in the lattice of co‐ordination networks.21

Figure 3

The molecular structure of the [Cu4(OH)4(C16H18N2)4]4+ cluster in 1 showing 50 % displacement ellipsoids (perchlorate ions and acetone molecule omitted for clarity). Symmetry code: (i) y, 1‐x, 1‐z.

The copper co‐ordination geometry is well described as a CuN2O3 square‐based pyramid with the two ligand N atoms [Cu−N=1.992 (3) and 2.032 (2) Å; N−Cu−N=88.84 (10)°] in a cis disposition in the basal plane. The geometry index or Addison tau parameter for Cu1 in the copper cube is 0.12 which indicates a slightly distorted square pyramidal geometry (tau=0 for a regular square pyramid and tau=1 for a regular trigonal bipyramid).25 The apical Cu1−O1 bond [2.336 (2) Å] is notably longer than the Cu1−O1 basal bonds [1.9493 (19) and 1.9803 (19) Å]. The Cu−O bond‐length distribution within the Cu4(OH)4 cube (Figure 4) corresponds to the type I classification of Mergehenn and Haase5 and the short/long distribution of Cu ⋅⋅⋅ Cu separations in the cube as defined by Ruiz6 [two at 3.0028 (4) Å and four at 3.2599 (5) Å] is completely consistent with this. A much longer Cu1−O2 contact [2.877 (3) Å] to a perchlorate O atom would complete an extremely distorted CuN2O4 octahedron for the metal‐ion but given that its bond valence is only 0.05, we do not regard it as a significant bond.26

Figure 4

Detail of the central Cu4(OH)4 cube and attached ligand N atoms in 1. Symmetry codes: (i) y, 1‐x, 1‐z; (ii) 1‐x, 1‐y, z; (iii) 1‐y, x, 1‐z.

In the crystal of 1 the [Cu4(OH)4(C16H18N2)4]4+ cations form a ‘windmill’ arrangement when viewed down [001] (Figure 5). An O−H ⋅⋅⋅ O hydrogen bond from the cubane hydroxyl group to the O atom of the acetone solvent molecule (H ⋅⋅⋅ O=2.04 Å, O−H ⋅⋅⋅ O=166°) is a notable feature of the packing, which is consolidated by weak C−H ⋅⋅⋅ O interactions.

Figure 5

The windmill packing in 1 viewed down .[001]

Magnetic susceptibility measurements were performed on polycrystalline sample of 1 in the range 300–2 K with an applied field of 0.1 T (Figure 6 Top). The χMT product remains almost constant down to 50 K with a value of around 1.5 emu.K.mol−1. This value corresponds for paramagnetic spin‐only behaviour of four Cu(II) (S=1/2) isolated centres. Below this temperature the χMT product gradually starts to increase and takes a value of 1.56 at 3 K. This behaviour could be caused by the presence of weak ferromagnetic exchange interactions within the cluster. The inset of Figure 6 shows the isothermal magnetisation (M) measured up to 5 T at temperatures of 2 and 5 K. The experimental data can be modeled using an isotropic spin‐spin interaction by the following Heisenberg‐Dirac‐van Vleck Hamiltonian:

Figure 6

Top: temperature dependence of the product of the magnetic susceptibility times the temperature at 0.1 T; Inset: isothermal magnetisations measured at T=2 and 5 K; Empty circles: experimental data; Full lines: simulated curves using best simultaneous fit; Bottom: surface error plot of χMT vs T as a function of J1 and J2 revealing a banana minimum; Full lines: isoerror lines from the innermost with R=4.0×10−6 to the most external (R=8.192×10−3) and the value of each line is double the next more internal.

where J1 and J2 are the isotropic exchange parameter among the nearest and most distant copper, respectively, and S1=S2=S3=S4=1/2. Simultaneous fit of both the susceptibility and magnetisation data, performed with the Magpack program,27, 28 yields J1=0.096, J2=0.062 cm−1 and g=2.00 (R=3.4×10−6) values for the Cu(II) tetramer. Solutions of similar quality are obtained for other parameters and this fit is not very sensible to the J1/J2 ratio. The dependence of the fit with respect to them is shown in a two‐dimensional plot of the error factor R on the values J1 and J2 (Figure 6 Bottom). The minimum error region (R=4.0×10−6) has a banana shape where both J's can be interconverted and with limits at (J1, J2)=(−0.12, 0.17) and (0.28, −0.025) cm−1.

EPR measurements were performed on a polycrystalline sample of 1 at 100 K (Figure 7). From crossing zero and small shoulder at low field: gper=2.06 and gpar=2.35 approx. (g average=2.16). The exchange parameters are not affected by the EPR result and this scale is comparable with the result from the susceptibility measurements.

Figure 7

EPR spectroscopy measurements on compound 1 at 100 K.

Conclusions

These parameters are in the same range found for other families of polynuclear complexes containing a Cu4O4 core and show similar magnetostructural correlation studies.12 In these studies, the J1 interaction can be both ferromagnetic and antiferromagnetic, whereas the J2 interaction must be weakly ferromagnetic for cubane type 2+4 similar to complex 1. In our case, both interactions have Cu−O−Cu angle very close to 98.5°. This angle makes the border between the ferromagnetic (angles smaller than 98.5°) and antiferromagnetic (angles larger than 98.5°) interactions, and this may justify the very low absolute values of the interactions obtained from the fit of complex 1. The square based pyramid coordination geometry of each copper atom shows an elongated apical Cu−O bond which is an example of Jahn‐Teller distortion.29 Of particular interest is the partially unsaturated dipyridyl ligand, containg a trans double bond, which will favour the formation of a unique complex and reduce the potential for disorder in the lattice. Ligands of this type may play an important role in advancing the science of transition metal‐ion supramolecular cluster synthesis by stabilising clusters and catalysing their formation.

Supplementary

Figure S1–S4: IR and UV/Vis spectra of compound 1

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.