A decanuclear [Dy III6 Zn II4 ] cluster: a {Zn II4 } rectangle surrounding an octahedral {Dy III6 } single molecule magnet

An octahedral {Dy III 6 } cage within a diamagnetic {Zn II 4 } rectangle is reported, with magnetic relaxation studies revealing single-mole-cule magnet behaviour for the complex under zero external dc ﬁ eld with U e ﬀ = 43 K and τ o = 1 × 10 − 5 s.

The field of lanthanide-based molecular magnetic materials has witnessed a vast growth over the last few years, due to the deeper understanding of the fundamental parameters that govern the magnetic behaviour of the 4f centres. Especially for single molecule magnets (SMMs) or single-ion magnets (SIMs), i.e. molecules that are able to retain their magnetism once magnetized at low temperatures upon removal of the external magnetic field, 1 employment of 4f centres has led to exceptional magnetic properties, due to both the magnitude of the spin, as well as the spin-orbit coupling based magnetic anisotropy that the 4f species possess. 2 Furthermore, for Dybased SMMs/SIMs the uniaxial magnetic anisotropy of the Dy atoms plays a crucial role in enhancing the energy barrier, U eff , needed for the re-orientation of the magnetisation, and thus molecular nanomagnets with U eff values over 1000 K, and blocking temperatures as high as 60 K have been reported, 3 with the best SMM/SIM being the recently reported organometallic complex [(Cp iPr5 )Dy(Cp*)] + (Cp iPr5 = penta-iso-propylcyclopentadienyl) with an U eff value of ∼2210 K and a blocking temperature of ∼80 K. 4 On the other hand, regarding mixedmetal 3d-4f heterometallic complexes as SMM candidates, the situation seems to differ; despite the ongoing research for the construction and characterization of 3d-4f SMMs, the energy barrier for such species remains rather low compared to 4f species, with [Co 2 Dy(L Br ) 2 (H 2 O)]NO 3 (L Br = 2,2′,2″-(((nitrilotris (ethane-2,1diyl))tris(azanediyl))tris(methylene))tris-(4-bromophenol)) holding the record of U eff = 600 K, 5 i.e. ∼1/4 of the highest U eff reported so far for the best lanthanide SIM/SMM. Such a discrepancy may be attributed to either weak (and most commonly antiferromagnetic) interactions between the 3d and 4f metal centres, leading to low lying split sublevels, or to the perturbation of the 4f local magnetic field due to random transversal secondary magnetic fields created by the nearby paramagnetic metal ions, resulting in quantum tunnelling of the magnetization and, thus, faster magnetic relaxation. 6 Recently we embarked on a project of constructing heterometallic Zn II -Ln III clusters and investigating their magnetic properties, as a means of: (i) protecting the 4f-centres from the influence of 3d paramagnetic centres and (ii) employing a diamagnetic 3d metal atom for structural stability and diversity of the products. 7 Herein we report our latest finding regarding a decanuclear [Zn II 4 Dy III 6 ] complex upon employment of the Schiff-base ligand 2-(β-naphthalideneamino)-2-hydroxymethyl-1-propanol, H 3 L (Scheme 1). The reaction of Zn(OAc) 2  Complex 1 crystallizes in the triclinic P1 space group (Fig. 1, top). Its metallic core consists of a diamagnetic {Zn II 4 } rectangular unit of ∼6.12 × 6.42 Å dimensions, surrounding a central magnetic {Dy III 6 } slightly "squeezed" octahedron ( Fig. 1,  bottom). The basal Dy III ions (Dy2, Dy2′, Dy3, Dy3′) are located ∼3.49 and ∼5.64 Å apart, while the axial Dy III centres (Dy1, Dy1′) are found ∼1.73 Å above and below the basal plane.
DC magnetic susceptibility measurements were performed on 1 in the 2-300 K temperature range under an applied magnetic field of 0.1 T, and the results are plotted as χ M T vs. T in Fig. 3, with the isothermal magnetisation (M vs. H) curves shown in the inset. The room-temperature χ M T value of 81.2 cm 3 mol −1 K is slightly smaller than the theoretical value of 85.0 cm 3 mol −1 K expected for six Dy III ions (S = 5/2, L = 5, J = 15/2, 6 H 15/2 , g J = 4/3). Upon cooling, the χ M T product remains practically unchanged until ∼150 K, below which steadily decreases reaching 41.05 cm 3 mol −1 K at 2 K, possibly suggesting the presence of weak antiferromagnetic interactions within the cluster (a Curie-Weiss analysis gave θ = −2.9 K) and/or depopulation of the Dy III Stark sub-levels. The isothermal magnetization versus B plots increase rapidly, reaching values of 29.2 and 28.5N μB at 5 T, for T = 2 and 5 K, respectively, significantly lower than the theoretical value of ∼60N μB for a hexanuclear [Dy III 6 ] complex; this is mainly attributed to the presence of magnetic anisotropy and to the depopulation of the Stark sublevels. 9  Given: (i) the large remaining magnetic moment, even at 2 K, and (ii) the square antiprismatic geometry of the Dy2 and Dy3 centres (and their symmetry related) that promotes suitable charge distribution with axial anisotropy on the Dy III centres, 10 we investigated the AC dynamic magnetic properties of 1. The measurements show temperature and frequency ( f ) dependent fully-formed in-phase, χ′ M , and out-of-phase, χ″ M , signals, under zero-applied DC field and a 3.5 AC field oscillating at various frequencies, in the temperature range between ca. 3 and 15 K (Fig. 4). This behaviour seems to indicate slow relaxation of the magnetization, which is typical for a SMM.
The main feature is a cusp in the in-phase component that occurs at approximately 10.0 K for f = 17 Hz, accompanied by a cusp in the out-of-phase component at somewhat lower temperature (Fig. 4). Effective time constants can be obtained from the reciprocal angular frequency at maximum absorption. As typical of a thermal activation process, the zero-field relaxation times so obtained can be approximated by the Arrhenius law τ = τ o exp(U eff /k B T ), where τ = (2πf ) −1 , τ o is an attempt frequency and U eff an effective energy barrier. The results are presented in the inset of Fig. 4, affording U eff = 43 K and τ o = 1 × 10 −5 s. The AC susceptibility reveals that the magnetic relaxation is particularly complex in 1, since at least another (secondary) relaxation pathway can be spotted at somewhat lower temperature than that of the main feature. In order to visualize the orientation of the anisotropy axis for each Dy III ion present in 1, we employed the electrostatic model reported by Chilton et al., which is based on electrostatic energy minimization for the prediction of the ground state magnetic anisotropy axis, 11 assuming that in the absence of high symmetry the groundstate of Dy III ions is a doublet along the anisotropy axis with m J = ± 15/2. 12 Following this approach, the ground state magnetic anisotropy axes in 1 were found to form three pairs from the symmetry-related Dy centres (Fig. 5); for Dy1(Dy1′) the axis is tilted towards the μ 4 -O1 oxide atom, and towards O2 belonging to a monoatomic methoxide bridge and O2F belonging to bridging acetate group. For Dy2(Dy2′) the axis is pointing towards the μ 4 -O1 oxide atom and towards O2B(O2B′) from a deprotonated aromatic hydroxyl group, while for Dy3(Dy3′) the axis is tilted towards the μ 4 -O1′oxide atom and the O2D′ atom of a bridging acetate group.

Conclusions
In conclusion, in this work we present the synthesis and characterisation of a novel decanuclear [Dy III 6 Zn II 4 ] cluster, upon employment of the Schiff-base ligand 2-(β-naphthalideneamino)-2-hydroxymethyl-1-propanol, H 3 L, with the cluster's topology describing a {Zn II 4 } rectangle surrounding a {Dy III 6 } octahedron. Furthermore, investigation of the magnetic properties, revealed possible SMM behaviour with U eff = 43 K and τ o = 1 × 10 −5 s. To the best of our knowledge, complex 1 represents the first example of a decanuclear [Dy III 6 Zn 4 ] single molecule magnet. Synthetic efforts are currently underway in order to isolate more Zn-4f clusters, as a means of investigating the effect of the diamagnetic ions on the magnetic behaviour of the 4f centres.

Conflicts of interest
The authors wish to declare no conflicts of interest.