Trinuclear Metallacycles: Metallatriangles and Much More Ennio Zangrando, Massimo Casanova, and Enzo Alessio* Dipartimento di Scienze Chimiche, Universita ` di Trieste, Via L. Giorgieri 1, 34127 Trieste, Italy Received January 31, 2008 Contents 1. Introduction 4979 2. Definitions 4980 3. Organization of the Work and Analytical Issues 4981 4. Metallatriangles with Angular Linkers and Metal-Containing Linear Edges 4982 4.1. Angular Linkers with Monodentate Binding Units 4982 4.2. Angular Linkers with Bidentate Binding Units 4985 4.2.1. µ 3 -X-Centered Trinuclear Metallacycles 4986 4.2.2. Oxo-centered Trinuclear Metal Carboxylates and Pyrazolates 4986 4.3. Angular Linkers with Tridentate Binding Units 4987 5. Trinuclear Metallacycles with Metal Corners 4987 5.1. Generic Trinuclear Metallacycles (Miscellaneous) 4987 5.2. Half-Sandwich Metal Corners and Tridentate Linkers 4988 5.3. Cyclohelicates 4991 6. Metallacalix[3]arenes 4994 7. Molecular Triangles with ca. 90° Metal Corners 4997 7.1. General Considerations 4997 7.2. Flexible Linkers 4997 7.3. Angular Linkers 4999 7.4. Trans Strands 4999 7.5. Linear Linkers 5001 7.5.1. Triangles and Squares in Equilibrium 5001 7.5.2. Triangles Plus Squares (Nonequilibrating Mixtures) 5003 7.5.3. Triangles or Squares 5004 7.5.4. Molecular Triangles Formed Exclusively 5005 7.5.5. Molecular Triangles with Pyrazine Edges 5006 7.5.6. Other Short Linkers 5008 8. Concluding Remarks 5008 9. Abbreviations 5010 10. Acknowledgments 5009 11. References 5010 1. Introduction In 1990, Fujita and co-workers described the first example of a rationally designed metallacycle, the molecular square [{Pd(en)(µ-4,4′-bipy)} 4 ](NO 3 ) 8 , prepared by self-assembly of the cis-protected square-planar Pd(II) precursor with two adjacent labile ligands [Pd(en)(ONO 2 ) 2 ] (en ) ethylenedi- amine) with the linear linker 4,4′-bipyridine (4,4′-bipy). 1 Since then, supramolecular chemistry has produced an amazing number of fascinating 2D and 3D metal-mediated molecular architectures, including many macrocycles and cages. Several review articles have thoroughly covered this thriving field in recent years. 2-16 Beside their shared structural beauty, some of these assemblies are finding applications as receptors and molecular vessels for trapping reactive intermediates, 17,18 as well as for stoichiometric 19 and catalytic reactions. 20 The smallest and simplest of the metal-mediated molecular polygons, the trinuclear metallacycles, are typically defined in the papers as rare or seldom found, but a search of the literature demonstrates that, indeed, they are not particularly rare. Instead, it is perhaps fair to say that very often their isolation was totally unexpected, or serendipitous at best, typically from reactions aimed at the preparation of larger metallacycles. In other words, the formation of trinuclear metallacycles is frequently in contrast with the paradigms of the directional-bonding approach as defined by Mirkin and Holliday (originally called the molecular library model by Stang and co-workers 7 ), that rationalizes the metal- mediated construction of supramolecular architectures. 9 Thus, in their apparent simplicity, trinuclear metallacycles pose a number of questions, both practical and theoretical, to the chemists involved in metal-mediated supramolecular chemistry. Despite their apparent rarity, trinuclear metallacycles are a somehow recurrent feature in our scientific career: one of us in previous years has determined the X-ray structures of several trinuclear metallacycles (see below) and very recently we isolated and structurally characterized two rare examples of molecular triangles with octahedral Ru(II) corners and pyrazine edges. 21,22 Thus, intrigued by our above-average familiarity with this so simple, and yet somehow elusive, metallacycle, we decided to perform a thorough search of the pertinent literature: this review article is intended to be a comprehensive, and yet critical, collection of the trinuclear metallacycles described to date. The main focus will be on the geometry of the metallacycles: the synthetic pathway leading to these species will not be used as a discriminating factor. In most cases, they were obtained (either on purpose or not) according to the principles of self-assembly, i.e., by the combination of metal fragments with two labile ligands (either with a trans or, more often, cis geometry) and ditopic linkers. However examples of trinuclear cyclic compounds obtained by different synthetic pathways (e.g., transformation of mononuclear metal complexes or metal-induced transfor- mation of added reagents) will be equally considered. The demonstrated (or more often proposed) functions of these metallacycles (e.g., as receptors) will be mentioned occasion- ally, but not stressed. * Corresponding author e-mail: alessi@units.it. Chem. Rev. 2008, 108, 4979–5013 4979 10.1021/cr8002449 CCC: $71.00  2008 American Chemical Society Published on Web 12/10/2008