Exploratory studies into 3d/4f cluster formation with fully bridge-substituted calix[4]arenes

Author Contributions: A.F., E.K.B. and S.J.D. authors conceived and designed the experiments; A.F. performed the experiments; A.F., L.J.M. and S.J.T. acquired data; A.F. and S.D. analyzed the data; A.F., E.K.B. and S.J.D. wrote the paper. Calix[4]arenes are extremely versatile ligands that are capable of supporting the formation of a wide variety of polymetallic clusters of paramagnetic metal ions. One can exert influence over cluster formation through alteration of the calix[4]arene framework and subsequent ‘expansion’ of the lower-rim polyphenolic binding site. The present contribution investigates cluster formation with calix[4]arenes substituted at all four methylene bridge positions with furan moieties. Two known cluster types have been isolated with this ligand, the structures of which lend insight into factors that may ultimately preclude the formation of mixed-metal species.


Introduction
The ability to influence the formation of paramagnetic metal ion clusters from multicomponent systems is a challenging synthetic goal that holds great potential when considering the possibility to control or fine tune physical properties such as molecular magnetism (1-5). We have found that methylene-bridged calix [4]arenes (cyclic tetraphenols, collectively termed C [4]s hereafter, Figure 1A), are remarkably versatile ligands for cluster synthesis under ambient conditions, affording a wide range of structural topologies as a result (6)(7)(8)(9)(10)(11)(12). This is primarily due to their possessing a tetraphenolic pocket that readily binds 3d or 4f metal ions when deprotonated with a suitable base. These phenolato groups bridge to neighbouring ions within a cluster (for example see Figure 1B) and impart a degree of directionality in the assembly process; this can also be considered as structural capping behaviour (vide infra). C [4]-supported clusters can also be synthesised under solvothermal (13,14) or air-sensitive conditions (15)(16)(17), but discussion of this expansive chemistry is far beyond the scope of this contribution; in these cases the resulting structures differ markedly from those reported herein. Thia-, sulfonyl-and sulfinyl-bridged C [4]s have also been used in cluster-forming chemistry, but lower-rim coordination chemistry is drastically different to that of the methylene-bridged analogues, leading to vastly different structure types and physical properties. Again, this is far outwith the scope of this manuscript, but the reader is directed to recent reviews on the subject (18)(19)(20).  [4]arene (R = H) and p-t Bu-calix [4]arene (R = t Bu), collectively termed C [4]s. Concerted hydrogen bonding interactions at the tetraphenolic lower-rim are shown as dashed lines between OH groups. B) Section of the single crystal X-ray structure of a C[4]-supported {Mn III 2Mn II 2(OH)2} cluster showing wing-tip Mn III binding in the tetraphenolato pocket, as well as phenolato and hydroxide bridging to a body Mn II ion (7,11). C) Schematic of dihomooxacalix [4]arene showing the distortion to the lower-rim tetraphenolato binding pocket through introduction of one ethereal bridge, the result of which is a trapezoidal binding pocket. D) Schematic of tetrahomodioxacalix [4] SMMs (1) that possess a common structural butterfly-like {Mn III 2Mn II 2(µ3-OH)2} core, half of which is shown in Figure 1B (7,11). The oxidation state distribution is rare (21), is the reverse of that found in the majority of structural analogues in the literature (22)(23), and is driven by the C [4] tetraphenolic pocket which preferentially binds the wing- showing that this is a versatile cluster topology for C [4]. Reactions containing both 3d and 4f ions give several different heterometallic clusters depending on the stoichiometries and metal salts employed (6,8,10,24), an interesting example of which is the ability to systematically interchange the body ions in the [Mn III 2Mn II 2(µ3-OH) 2 In addition to carrying out such exploratory work with C [4], we have also reported on aspects of binding site alteration and the subsequent effect this has on cluster formation / composition (25)(26)(27). Our first effort in this regard was to employ oxacalixarenes to systematically vary the size of the tetraphenolato pocket (26)(27). The introduction of one or two ethereal bridges between neighbouring aromatic rings 'expands' the C [4] framework, giving trapezoidal ( Fig. 1c) or rectangular (Fig. 1d) binding pockets respectively. The use of these oxacalixarenes in cluster formation gave markedly different results, an excellent example of which is the isolation of a Ln5 cluster that can be considered an analogue of the aforementioned Ln6 species. In this case one of the equatorial ions from the octahedron has been omitted due to the targeted 'expansion' of the lower-rim binding pocket; this Ln III ion omission coincides directly with the position of the ethereal bridge. We also recently tethered C [4]s with alkyl chains of varying length through synthetic modification at one methylene bridge position (25). In doing so we found cluster formation with these bis-C [4]s to be challenging, except in cases where the tether was long enough to allow the constituent C [4]s to form previously reported cluster topologies (e.g. 1).
Here we present initial results from a new investigation that aims to explore the limits of cluster formation with C [4]s substituted at all methylene bridge positions. A search of the literature shows that there are relatively few fully substituted C [4]s that can be readily synthesised with all substituted groups equatorial, but one candidate that represented a viable starting point for this work is 5,11,17,23-tetra-tert-butyl-2,8,14,20tetra(2-methylfuranyl)-25,26,27,28-tetramethoxycalix [4]arene (4, Scheme 1) (28).

Results and Discussion
The use of compound 4 in cluster formation (via our typically employed ambient reaction conditions) first required de-protection of lower-rim methoxy groups to afford the required tetraphenolic C [4] (6). Crystals of 6 were in a triclinic cell and structure solution was carried out in the space group P-1. The asymmetric unit (ASU) was found to contain half of the compound formula and, as expected, symmetry expansion revealed formation of the common butterfly cluster topology as shown in Figure 2. Inspection of Figure 2A shows that the furans substituted at the methylene bridge positions are all equatorial as expected, and that they are sufficiently far from the cluster core to prevent interference with solvent ligation at the body Mn II ions (ligated solvent is shown in Figure S1); bond lengths and angles relating to the cluster are similar to those of the C[4]-supported analogue, 1 (7,11). The positioning of the furan groups around the cluster periphery is naturally dictated by the coordination chemistry, and as can be seen in Figure 2B, the two equivalents of 5 are arranged in an offset manner due to the cluster core topology.
With the orientation / positioning of the furan groups in mind, we expanded our investigation to include the formation of 3d-4f clusters to establish whether: 1) it would be possible to isolate mixed-metal clusters, 2) different stoichiometries would lead to 3d   Figure 3. Inspection of the structure shows that this overall cluster topology conforms to the previously reported lanthanide octahedra discussed in the introductory section, but in this case there is yet another variation in the nature of anions in and around the cluster core. Figure 3B shows the central core with tetra-anions of H45 omitted for clarity. As can be seen, the inner part of the core contains two µ4-oxides that bridge Tb1, its symmetry equivalent (s.e., Tb1*), Tb2 and Tb3. This is a common feature to all three Ln III octahedra (12,3), as is the presence of two peripheral bridging formate anions (e.g.
see Figure 3B,  The magnetic properties of clusters 6 -8 were not investigated, as all are very closely related to previously published topologies and can be expected to be very similar.

Conclusions
To conclude, we have shown that a C [4] substituted at all methylene bridge positions can be de-protected at the lower-rim and successfully used in the synthesis of 3d or 4f clusters.
The three species isolated conform to two known types reported for analogous C[4]supported cluster chemistry, and the fact that 3d-4f clusters were not isolable suggests that the presence of furan moieties at the C [4] bridge positions hinders mixed-metal cluster formation. This may be due to the increased coordination number / presence of additional ligands around the cluster periphery, but further investigation is required in order to confirm this hypothesis. This will be the subject of future work in the area of fully bridge-substituted C[4]-supported cluster chemistry and results will be reported in due course, including magnetic properties of clusters isolated if these are found to deviate from existing topologies.

Experimental
Compound 4 was synthesised according to literature procedures and purity was confirmed by 1 H NMR prior to use (28)  and wR2 was 0.1354 (all data).