Click chemistry as a route to the synthesis of structurally new and magnetically interesting coordination clusters: a {Ni} complex with a trapezoidal prismatic topology.

The synthesis of a new {Ni 8 } cluster bearing tetrazolate- and azido-bridging ligands, and supported by chelating α -methyl-2-pyridine-methanol (mpmH) groups, is herein described. The reported compound has a unique trapezoidal prismatic topology, resulting from an unexpected in-situ click reaction between the MeCN reaction solvent and the N 3- ions under mild, room-temperature conditions. Such a click chemistry approach to the preparation of 0-D compounds is relatively unexplored and represents a fruitful strategy for the synthesis of new coordination clusters and molecule-based magnetic materials.

Although click chemistry has been used in almost all key areas of synthetic organic chemistry, 3 there are only few polynuclear metal complexes resulting from ligands that were in-situ derived by click reactions. 4 In fact, a convenient synthetic route based on a metal-ion assisted [2+3] cycloaddition of nitriles with azides was explored by Sharpless and coworkers, 5 and developed by Xiong 6 and others 7 to prepare coordination polymers via in-situ generated 5-substituted 1H-tetrazolate bridging ligands. Endon (EO) bridging azides have a special position in polynuclear 3d-metal cluster chemistry and molecular magnetism due to their binding versatility and flexibility, and their known capability to promote ferromagnetic exchange interactions between the metal ions they bridge, respectively. 8 In addition, tetrazoles have been found to adopt at least nine distinct types of coordination modes with metal ions in the construction of metal-organic frameworks. 6 Therefore, tetrazoles have attracted increasing attention in molecular chemistry and crystal engineering due to the excellent coordination ability of the four nitrogen atoms of the functional group, acting as either a multidentate or a bridging building block in various supramolecular assemblies. 6,9 Interestingly, all previously reported coordination compounds (clusters and polymers) resulting from click reactions and bearing in-situ generated tetrazolate ligands have been prepared from hydro/solvothermal methods. 6,7,9,10 The only exception, to our knowledge, is a {Co II 10} cluster reported by Zhang and Sato, 11 which has been prepared and crystallized under conventional synthetic techniques.
To facilitate the formation and crystallization of 0-D cluster compounds, the employment of chelating and/or bridging organic ligand(s) appears as a necessary synthetic parameter which would prevent the extensive aggregation/polymerization of metal ions that would otherwise lead to multidimensional coordination polymers. Thus, we have recently initiated a program targeting the synthesis of new and potentially chiral pyridyl-alkoxide based chelating/bridging ligands and their use in the formation of high-nuclearity 3d-metal cluster compounds with interesting magnetic properties, such as high-spin molecules and single-molecule magnets (SMMs). To this end, the employment of α-methyl-2-pyridine-methanol (mpmH; Scheme 1, left) in Mn carboxylate chemistry has afforded a{Mn31} SMM with an unprecedented nanosized structure and a large energy barrier of ∼60 K for reversal of the magnetization. 12 As a part of our ongoing studies on understanding the chemistry-structure-magnetism relationships within a library of coordination 3d-metal clusters based on mpmH, we have herein been able to synthesize -under mild conditionsand characterize a structurally unique {Ni8} cluster (1), which possesses a trapezoidal prismatic topology resulting from the coordination of both end-on bridging azides (N3 -) and 5methyltetrazolate (mtz -) groups. The mtzligands have unexpectedly appeared in the structure of 1 due to the in-situ, metal-assisted click reaction between N3ions and the MeCN reaction solvent (Scheme 1, right). The co-presence of both bridging azido and mtzgroups in 1 resulted in the appearance of both ferro-and antiferromagnetic exchange interactions between the Ni II centers, respectively.
Scheme 1 Structural formula and abbreviation of the chelate ligand α-methyl-2-pyridine-methanol (mpmH) used in this work (left), and the click reaction that leads to the formation of the crystallographically established μ4-bridging mtzgroups in 1 (right).
Each Ni II ion in 1 is bridged to its neighboring metal ions through a μ-1,1 end-on N3and a diazine-part of the mtzligands; the latter groups adopt an overall η 1 :η 1 :η 1 :η 1 :μ4 mode, each of them linking four Ni II atoms in a nearly planar conformation (mean deviation of Ni atoms from the Ni4 plane is 0.016 Å). The μ4-bridging mode of the mtzligands in 1 accommodates the maximum number of metal ions that this group can potentially bind to, thus representing the second example of a 0-D molecular compound (the first in Ni II chemistry) bearing μ4-mtzligands. 11 The intramolecular Ni•••Ni separations and Ni-(μ-N3)-Ni angles span the range 3.417-6.784 Å and 108.0-114.3°, respectively, whereas the Ni-N-N-Ni torsion angles are within the range 1.0-15.2°. Peripheral ligation about the [Ni8(μ-N3)8(μ4-mtz)4] 4+ core (Fig. 2, top) is provided by a total of eight N,O-bidentate chelating mpm -/mpmH groups, each of them capping a different Ni II ion. The charge neutrality of the {Ni8} cluster requires four of the bidentate chelating mpmligands to be protonated, i.e. in the mpmH form. Indeed, this agrees with the short O1•••O2 contact (2.509 Å) between two adjacent mpm -/mpm(H) ligands, thus implying a relatively strong H-bonding interaction (dashed lines in Fig. 1, bottom) and consequently the presence of a H-atom between the pair of these alkoxido groups (an O1•••H•••O2 homo-synthon). All Ni II ions in 1 are six-coordinate with near-octahedral geometries. The crystal packing of 1•2.2H2O revealed that the lattice H2O molecules occupy the voids between adjacent {Ni8} clusters (Fig. S1). The shortest Ni···Ni distance between neighboring {Ni8} clusters in the crystal is 8.422(1) Å.
Although there are many {Ni8} complexes reported in the literature with an open-shell structure, such as cubes 14 and rings, 15 the trapezoidal prismatic topology of complex 1 appears to be a new addition in Ni II cluster chemistry. Given the structural novelty of 1, there are some geometrical features that deserve further discussion. A trapezoidal prism is a three-dimensional figure that consists of two trapezoids on opposite faces connected by four rectangles. A trapezoidal prism has 6 faces, 8 vertices and 12 edges. The two trapezoids within complex 1 (Fig.  2, bottom)   Based on detailed DFT calculations, Noodleman, Sharpless and coworkers have proposed several different mechanisms of tetrazole formation, including concerted cycloaddition and stepwise addition of neutral or anionic azide species to various nitriles at elevated temperatures. 16 Given the reaction conditions employed for the synthesis of 1, it is very likely that an anionic, metal-assisted cycloaddition would be the predominant mechanism of the mtzformation. This involves either a direct [2+3] cycloaddition or a two-step sequence wherein the azide first nucleophilically attacks the relatively electron-rich MeCN, followed by ring closure. 16,17 Variable-temperature (2.0-300 K range), direct-current (dc) magnetic susceptibility measurements were performed on a freshly-prepared microcrystalline sample of 1•2H2O; a dc field of 0.3 T was applied from 30 to 300 K and a weak dc field of 0.03 T was used from 2 to 30 K to avoid saturation effects. The data are shown as χΜT vs. T plot in Fig. 3. The χΜT product slightly increases from a value of 9.93 cm 3 mol -1 K at 300 K to 10.43 cm 3 mol -1 K at 80 K, followed by a rapid decrease to a value of 1.49 cm 3 mol -1 K at 2.0 K. The value of the χΜT product at 300 K is slightly higher than the value of 9.68 cm 3 mol -1 K (calculated with g=2.2, as usual in Ni II cluster compounds 14,15 ) expected for eight non-interacting Ni II (S = 1) ions. The observed magnetic behavior is consistent with the presence of both ferro-and antiferromagnetic exchange interactions between the eight Ni II ions at the corresponding high-and low-T regions. The small χΜT value at 2.0 K and its tendency for heading to zero suggests that the antiferromagnetic component eventually dominates, thus fostering the stabilization of an S = 0 ground state for the octanuclear complex 1. Magnetization (M) vs. field (H) measurements were also performed for 1 at 2 K and the corresponding plot (Fig. 3, inset) shows a nearly linear increase up to a nonsaturated value of ~6 NμB at 5 T. This is consistent with an S = 0 ground state and a progressive population of close in energy low-lying spin states with S > 0 even at low T. There are four different types of superexchange interactions between the Ni II ions in complex 1, which comprise the following magnetic pathways (Fig. 4) (1). An excellent simultaneous fit of the magnetic susceptibility and magnetization data was obtained, and the resulting best-fit parameters were: J1 = +11.9 cm -1 , J2 = +0.7 cm -1 , J3 = -1.2 cm -1 , J4 = -1.9 cm -1 and g = 2.18 [R(χMT) = 3.8·10 -5 and R(M) = 2.0·10 -3 ]. The obtained J constants agree with magnetostructural correlations previously reported for end-on azido-and diazine-bridged Ni II complexes (vide infra), and they are consistent with the expected ferromagnetic interactions promoted by the N3 -/2,3-tetrazolate combination (J1 and J2) 18 and the weak antiferromagnetic interactions mediated solely by the long 1,4-or 1,3-tetrazolate pathways (J3 and J4), 15,19 thus leading to an overall S = 0 ground state. The small J2 value, compared to J1, could be attributed to the co-presence of one EO-N3 -(ferromagnetic coupler) and two diazine NN-bridges (antiferromagnetic couplers) from the mtzgroups. 15,18,19 (1) Good fits of the experimental data can be also obtained with several sets of J values (see the corresponding discussion in ESI †), and therefore it is poorly reliable to assign absolute values for each one of them. In light of the results derived from the different fits, the maximum ferromagnetic interaction should be attributed to the EO-azide/2,3-tetrazolate bridge, with a J value close to +11-12 cm -1 , and weak interactions promoted by the remaining bridges. For most divalent 3d-metal complexes with EO bridging N3ligands, the angle for switching from ferro-to antiferromagnetic coupling is typically > 104°. 8 Despite the fairly large average Ni-N-Ni angle of 110.6° in 1, and the copresence of planar N,N-bridging mtzligand(s) with small torsion angles, which are known antiferromagnetic couplers, the ferromagnetic component induced by EO-N3is still significant. This is not surprising as it has been previously observed in several examples of {Ni2} complexes bridged by μ-1,1 end-on N3and pyrazolate-type bridging ligands with very large Ni-Nazide-Ni angles (>115°). 18,20 This is also in line with DFT calculations, which predicted that the coupling in Ni II chemistry should be always ferromagnetic for all ranges of Ni-N-Ni angles, and experimentally evidenced by the ferromagnetically coupled polyoxometallate-based dinickelazide compound with a very large Ni-(μ-N3)-Ni angle of 129.3°. 20 In summary, we have herein shown that click chemistry is a useful synthetic route not only in organic chemistry and related fields but also for the preparation of new coordination clusters with unprecedented structural motifs and magnetic properties related to the combination of bridging azides and in-situ generated tetrazoles. The reported octanuclear Ni II cluster was obtained under mild, room-temperature synthetic conditions and exhibits a fascinating trapezoidal prismatic topology, resulting from the copresence of end-on N3and μ4-bridging mtzgroups; the latter serve to hold the vertices of the two {Ni4} trapezoids together and additionally link the two opposite trapezoids to the resulting prismatic conformation. We are currently trying to isolate the chiral forms of the reported {Ni8} compound by utilizing the R-and S-mpmH derivative, and subsequently study the magnetic and electric properties of the resulting chiral species. Work in progress also includes the in-depth investigation of click chemistry as a means of preparing magnetic coordination clusters with unique structural motifs, and the use of racemic and potentially chiral pyridine alkoxide ligands with other 3d-metal ions, as well as 4f-metals and 3d/4f-metal combinations. These results will be reported in the full paper of this work and other upcoming publications.

Conflicts of interest
There are no conflicts to declare.