Fascination with supramolecular chemistry over the last few decades has led to the synthesis of an ever-increasing number of elegant and intricate functional structures with sizes that approach nanoscopic dimensions. Today, it has grown into a mature field of modern science whose interfaces with many disciplines have provided invaluable opportunities for crossing boundaries both inside and between the fields of chemistry, physics, and biology. This chemistry is of continuing interest for synthetic chemists; partly because of the fascinating physical and chemical properties and the complex and varied aesthetically pleasing structures that supramolecules possess. For scientists seeking to design novel molecular materials exhibiting unusual sensing, magnetic, optical, and catalytic properties, and for researchers investigating the structure and function of biomolecules, supramolecular chemistry provides limitless possibilities. Thus, it transcends the traditional divisional boundaries of science and represents a highly interdisciplinary field. In the early 1960s, the discovery of ‘crown ethers’, ‘cryptands’ and ‘spherands’ by Pedersen,1 Lehn,2 and Cram3 respectively, led to the realization that small, complementary molecules can be made to recognize each other through non-covalent interactions such as hydrogen-bonding, charge-charge, donor-acceptor, π-π, van der Waals, etc. Such ‘programmed’ molecules can thus be self-assembled by utilizing these interactions in a definite algorithm to form large supramolecules that have different physicochemical properties than those of the precursor building blocks. Typical systems are designed such that the self-assembly process is kinetically reversible; the individual building blocks gradually funnel towards an ensemble that represents the thermodynamic minimum of the system via numerous association and dissociation steps. By tuning various reaction parameters, the reaction equilibrium can be shifted towards the desired product. As such, self-assembly has a distinct advantage over traditional, stepwise synthetic approaches when accessing large molecules. It is well known that nature has the ability to assemble relatively simple molecular precursors into extremely complex biomolecules, which are vital for life processes. Nature’s building blocks possess specific functionalities in configurations that allow them to interact with one another in a deliberate manner. Protein folding, nucleic acid assembly and tertiary structure, phospholipid membranes, ribosomes, microtubules, etc. are but a selective, representative example of self-assembly in nature that is of critical importance for living organisms. Nature makes use of a variety of weak, non-covalent interactions such as hydrogen–bonding, charge–charge, donor–acceptor, π-π, van der Waals, hydrophilic and hydrophobic, etc. interactions to achieve these highly complex and often symmetrical architectures. In fact, the existence of life is heavily dependent on these phenomena. The aforementioned structures provide inspiration for chemists seeking to exploit the ‘weak interactions’ described above to make scaffolds rivaling the complexity of natural systems. The breadth of supramolecular chemistry has progressively increased with the synthesis of numerous unique supramolecules each year. Based on the interactions used in the assembly process, supramolecular chemistry can be broadly classified in to three main branches: i) those that utilize H-bonding motifs in the supramolecular architectures, ii) processes that primarily use other non-covalent interactions such as ion-ion, ion-dipole, π–π stacking, cation-π, van der Waals and hydrophobic interactions, and iii) those that employ strong and directional metal-ligand bonds for the assembly process. However, as the scale and degree of complexity of desired molecules increases, the assembly of small molecular units into large, discrete supramolecules becomes an increasingly daunting task. This has been due in large part to the inability to completely control the directionality of the weak forces employed in the first two classifications above. Coordination-driven self-assembly, which defines the third approach, affords a greater control over the rational design of 2D and 3D architectures by capitalizing on the predictable nature of the metal-ligand coordination sphere and ligand lability to encode directionality. Thus, this third strategy represents an alternative route to better execute the “bottom-up” synthetic strategy for designing molecules of desired dimensions, ranging from a few cubic angstroms to over a cubic nanometer. For instance, a wide array of 2D systems: rhomboids, squares, rectangles, triangles, etc., and 3D systems: trigonal pyramids, trigonal prisms, cubes, cuboctahedra, double squares, adamantanoids, dodecahedra and a variety of other cages have been reported. As in nature, inherent preferences for particular geometries and binding motifs are ‘encoded’ in certain molecules depending on the metals and functional groups present; these moieties help to control the way in which the building blocks assemble into well-defined, discrete supramolecules.4 Since the early pioneering work by Lehn5 and Sauvage6 on the feasibility and usefulness of coordination-driven self-assembly in the formation of infinite helicates, grids, ladders, racks, knots, rings, catenanes, rotaxanes and related species,7 several groups - Stang,8 Raymond,9 Fujita,10 Mirkin,11 Cotton12 and others13,14 have independently developed and exploited novel coordination-based paradigms for the self-assembly of discrete metallacycles and metallacages with well-defined shapes and sizes. In the last decade, the concepts and perspectives of coordination-driven self-assembly have been delineated and summarized in several insightful reviews covering various aspects of coordinationdriven self-assembly.15 In the last decade, the use of this synthetic strategy has led to metallacages dubbed as “molecular flasks” by Fujita,16 and Raymond and Bergman,17 which due to their ability to encapsulate guest molecules, allowed for the observation of unique chemical phenomena and unusual reactions which cannot be achieved in the conventional gas, liquid or solid phases. Furthermore, these assemblies found applications in supramolecular catalysis18,19 and as nanomaterials as developed by Hupp20 and others.21,22 This review focuses on the journey of early coordination-driven self-assembly paradigms to more complex and discrete 2D and 3D supramolecular ensembles over the last decade. We begin with a discussion of various approaches that have been developed by different groups to assemble finite supramolecular architectures. The subsequent sections contain detailed discussions on the synthesis of discrete 2D and 3D systems, their functionalizations and applications.
