Metal complexes containing non-innocent ligands for functional materials

Abstract

The existence of complexes of that display non-innocence has been of interest in the field of coordination chemistry since the investigations of square-planar dithiolene complexes of the late transition metals in the 1960s. The ligands used in these systems are termed “non-innocent” when bound to a number of the late transition metals, because the orbital energy levels are similar to those of the central metal ion. This allows there to be significant electron delocalisation over the complex with the potential for the complexes to exist in a number of electronic states due to the combined electrochemical activity. In 1966, Jørgensen classified innocence as ligands that “allow oxidation states of the central atoms to be defined”, thus by this logic non-innocent ligands are defined as complexes where the precise oxidation states of the ligand and metal are ambiguously assigned. However it should be noted that no ligand is inherently non-innocent, but rather the ligand may behave in a non-innocent manner under the right circumstances. The qualification of non-innocence should therefore only be applied to combinations of metal and ligand that result in the aforementioned properties. In this thesis, the term “non-innocent” will be used to define ligands that often possess non-innocent behaviour when complexed to the metal centres they are bound to. A general form of ligand that displays non-innocent behaviour is that of the 1,2-bidentate moiety with an unsaturated carbon backbone. The chelating donor groups (X) are either O, NH, S, or a combination of the three. The central transition metal is generally a late metal that favours a square-planar geometry, because the planarity of the complex is crucial for electron delocalisation within the molecule and molecular interactions in the solid material. When the metal is nickel or platinum for example, their square-planar complexes with such ligands have shown threemembered electron-transfer series. Specific examples of ligands that have been shown to display non-innocent behaviour are those of catechol (1,2-dihydroxybenzene) and 1,2-diaminobenzene, where the unsaturated backbone is provided by a phenyl group. The electronic nature of these compounds has been extensively investigated by the groups of Pierpont and Lever, with focus on their redox and magnetic properties. The combined metal and ligand redox activity results in interesting magnetic behaviour, with potential for magnetic exchange interactions between a paramagnetic metal centre and the radical ligand or between two radical ligands mediated by a diamagnetic metal centre. This research has been advanced by Wieghardt and co-workers who have performed experimental and theoretical examination of non-innocent complexes of 1,2-substituted phenyl chelates, where the donor group is a combination of O and NH. These studies have focused on the understanding the nature of the metal-ligand interactions to apply to biological systems, such as those observed at the active site of enzymes that act upon molecules with similar moieties to the non-innocent ligands. Compounds of catechol may be referred to as dioxolenes in analogy to the sulfur-based dithiolenes. The deprotonated, dianionic form of catechol is known as catecholate (cat), which can be readily oxidised to the monoanionic o-semiquinone (SQ) and neutral o-benzoquinone (Q) forms. It has been seen that catecholate compounds can be described by localised electronic states with defined oxidation states, unlike many of the dithiolene class of molecules. However these states can exist in equilibrium with each other when the metal and ligand orbitals are close in energy, with differences in the charge and spin definition in what has been described as “valence tautomerism”. Therefore, although the complexes may not be seen as non-innocent by definition that their oxidation states are not ambiguous, it is still a useful description due to the potential for easily accessible charge states. Metal dithiolene complexes, where the metal is coordinated by one or more ligands with two S-donor atoms linked by a conjugated backbone, are one of the best researched of the non-innocent class of molecules. The square-planar bis-dithiolenes of the late transition metals have interesting magnetic, optical and electrical properties arising from the delocalised nature of the constituent metal and ligand orbitals, which has enabled their use for a wide range of applications such as non-linear optics, transistors and near-infrared switches. Of particular interest is the ability to fine tune the electrical properties to fit the application by changing the substituents on the core dithiolene moiety. For example, Anthopoulos has shown that by lowering the energy of the lowest unoccupied molecular orbital (LUMO), stable n-channel conductivity can be observed in field-effect transistors (FETs). Materials based on square-planar non-innocent complexes have been tested in FETs, and been seen to display field-effect mobilities as high as 10˗2 cm2 V˗1 s˗1 as with Ni bis(o-diiminobenzo-semiquinonate) complexes. Most of these molecules are based on conjugated, chelating ligands such as 1,2-diaminobenzene and the dithiolene class of molecules. Field-effects have also been observed in square-planar Pt complexes, where the conductivity is thought to arise from beneficial Pt-Pt bonds in addition to the π-stacking between molecules. Despite the similarity to the diaminobenzene and dithiolene counterpart, there are no reports of catechol-based materials displaying field-effect properties in the literature. Catechol compounds are well-researched in the field of biological chemistry due to the prevalence of the catechol moiety and enzymes with which it interacts in nature. However they have not been examined far beyond their simple coordination chemistry or chemical characterisation.