Chemical Control of Metal-Metal Interactions in Octahedral Iron Clusters

Abstract

Despite decades of research in both biomimetic and abiological metal cluster synthesis, systematic control of the electronic structure of molecular metal clusters remains a significant challenge. This challenge stems in part from the relatively low stability of the core architecture of many metal clusters, which often renders investigations of the relationship between their chemical and electronic features by chemical substitution intractable. Our lab has recently sought to address this problem by preparing a polynucleating ligand, HLH6 (HLH6 = CH3(CH2NH-(2-NH2)C6H4)3), that possesses a strong bias towards the formation of a single hexanuclear cluster architecture. Upon metallation with iron, this ligand gives the hexanuclear iron cluster (HL)2Fe6, whose pseudo-octahedral [Fe6] core topology is reliably retained upon a variety of derivatization reactions. Accordingly, this cluster offers an intriguing opportunity to investigate how chemical synthesis may be used to rationally control electronic structure in clusters. To begin, we first sought to investigate a novel spectroscopic tool, multiple wavelength anomalous dispersion (MAD), which has been previously used to provide site specific information about the oxidation state of metals in both extended solids and molecules. Investigation of a series of cluster compounds with this technique revealed that the spectra obtained are influenced not just by oxidation state, but also its ligand field. In fact, we demonstrate that in compounds with sufficiently strong metal-ligand and metal-metal interactions, it is these interactions, and not oxidation state, that are the primary determinant of the MAD spectrum for a given cluster. Having completed our spectroscopic investigation, we next turned to synthesis to explore the electronic structure of a series of [Fe6] clusters. We began by preparing a series of all-ferrous [Fe6] clusters featuring both weak and strong field ligands. Binding weak field ligands was shown to systematically switch the cluster between S = 6, S = 8, and S = 10 spin ground states depending on the number of weak field ligands. Strong field ligands, conversely, are shown to drive the cluster towards a low spin ground state. These changes are rationalized both in terms of exchange coupling of localized spins and from a fully delocalized molecular orbital approach. With the influence of exogenous ligands established, we also investigated the influence of the polynucleating ligand itself. Thus, a series of ligand derivatives were prepared by functionalizing a Caryl–H moiety on the ligand backbone with fluoro, 4-fluorophenyl, methyl, and methoxy groups. Metallating with iron then yielded the cluster series (RL)2Fe6; R = F, Ar, Me, and MeO, according to the ligand used in the metallation. While these changes produce only minimal geometric or spectroscopic changes to the cluster, they substantially alter its electrochemical behavior, resulting in a systematic shift in the potential at which electron transfer events occur for each cluster derivative. Additionally, the clusters with electron donating substituents, MeL2Fe6 and MeOL2Fe6, are shown to electrochemically access nine different electron transfer events, implying that a record ten redox states may be accessible in these clusters. Lastly, we sought to provide a more generalized description of the electronic structure of the [Fe6] family of clusters. We therefore reviewed the utility of the fully delocalized molecular orbital approach in describing the spin states of a variety of [Fe6] clusters. This analysis also allowed us to discuss a potential origin of the magnetoanisotropy of these clusters. Preliminary results from a HF-EPR investigation of this mangetioanisotropy indicate that this magnetoanisotropy may be well described by a giant spin model. Lastly, a method for synthetically manipulating the anisotropy is explored.