Many naturally occurring systems show us how multi-component supramolecular assemblies can generate useful functional behaviour. In this article the problems and limitations associated with achieving such behaviour in artificial multi-component assemblies is discussed, together with two examples of functions in artificial supramolecular assemblies based on (i) host–guest chemistry in cavities of cages, and (ii) light-harvesting in multi-chromophore arrays. Important challenges for the future are summarised.

Covalently linked chromophore–quencher complexes are widespread in the area of transition-metal photochemistry and as models for photosynthesis. This review surveys recent examples of supramolecular complexes in which interacting chromophore and quencher fragments are instead held together by non-covalent interactions such as hydrogen bonding, aromatic π-stacking, hydrophobic interactions and labile metal–ligand coordinate bonds. The use of these methods to assemble multi-component photo-active complexes has led to the preparation of many highly sophisticated systems for energy transfer or charge separation which would not be accessible by ‘conventional’ synthetic methodology.

This article reviews recent work in the area of non-innocent behaviour in polynuclear metal complexes. Non-innocence, which occurs when metal-based and ligand-based redox orbitals are similar in energy, has been known since the first dithiolene complexes of the Ni triad. Our recent work in this field is with complexes of two distinct types: polynuclear complexes of Ru(II) with dioxolene-type bridging ligands; and dinuclear complexes based on tris(pyrazolyl)borato-Mo(V) or -Mo(I) units linked by bis-phenolate or bis-pyridyl bridging ligands. Detailed redox and UV/Vis/NIR spectroelectrochemical studies on these complexes have been carried out. An important point which emerges is that non-innocent behaviour in dinuclear complexes is an essential prerequisite for strong metal–metal electronic coupling across extended bridging ligands. Many of the complexes studied show intense charge-transfer transitions in the near-IR region of the spectrum, and the use of these in prototypical optical devices is briefly discussed.

Reaction of simple bis-bidentate ligands, containing two chelating pyrazolyl-pyridine units connected to a central aromatic spacer, with six-coordinate transition metal dications results in self-assembly of an extensive series of polyhedral cage complexes. These include M₄L₆ tetrahedra, M₈L₁₂ cubes, M₁₂L₁₈ truncated tetrahedra and M₁₆L₂₄ tetra-capped truncated tetrahedra. In all cases the metal : ligand ratio is 2 : 3, reflecting the combination of six-coordinate metal ions with tetradentate ligands. The resulting structures are based on those polyhedra which have a 2 : 3 ratio of vertices to faces, with a metal ion at each vertex and bridging ligand spanning each edge. The cages display a range of interesting properties such as an anion-based template effect in the smaller examples; host–guest chemistry associated with the central cavity; aromatic stacking around the periphery between electron-poor and electron-rich ligand fragments which appears to contribute substantially to their stability; and modified fluorescence properties arising from the aromatic stacking of fluorophores such as naphthyl and anthracenyl groups built into the ligand backbone. Even more complex structural types are available using a mixture of face-capping (tris-bidentate) and edge-bridging (bis-bidentate) ligands, such as examples of M₁₂ cuboctahedra which select a combination of two types of ligand during the self-assembly process.