Chemical Communications Blog

Non-covalent interactions dictate the assembly of many of nature’s most elegant structures. Similarly, supramolecular chemists have long been intrigued by the challenge of designing functional structures that spontaneously self-assemble from simpler fragments which mutually recognise each other.

A popular self-assembly approach is to produce coordination compounds from transition metal salts with rigid organic ligands. Directional bonding around transition metal centres allows the production of predictable and controllable shapes. Michael Schmittel’s group at the University of Siegen have been exploring a newer approach. They prepared two assemblies, a 2-component triangle T1 and a 3-component rectangle R1. The transition metal “corner” arrangements in T1 and R1 are disfavoured, so if the two assemblies are combined the components re-shuffle to form a more favourable assembly- the 5-component triangle T2. The transformation occurs at room temperature, and can be completed in just 1 hour in the presence of a catalyst, which accelerates the re-shuffling by labilising the metal-ligand bonds.

Unlike previous examples, the conditions needed for the transformation are very mild. The authors compare the process to gene shuffling, the combination of dissimilar genes to form new genetic material. The strategy could be considered a first step towards the evolution of supramolecular architectures, and a great route to more complex supramolecular assemblies with higher information content.

The full communication can be downloaded here (free to access for a limited period).

Cally Haynes

The X-ray crystal structures of two azide-functionalised catenanes

Entropy is not a friend of the macrocyclic chemist because creating large cyclic molecules by tying together several building blocks creates a high thermodynamic hurdle to cross. Nature has been forming these types of molecules for many millions of years and has provided inspiration for chemists to overcome these synthetic challenges. Perhaps purely for the academic interest, work began on investigating if it was possible to interlock two of these cyclic molecules together and form what is known as a catenane. Imagine a pot of spaghetti cooking and think of the strands tumbling around in the boiling water. Now imagine trying to tie all the ends together at the same time. Most will form individual rings but statistically a small number of those rings will be linked together. Conceptually this sounds easy but the reality is not so straight forward. A secondary problem is that only very small amounts of the linked rings are formed which means most of the starting material is wasted.

As interest in these curious molecules grew their potential application as switches for molecular electronics and sensors became apparent. However, if these applications were ever to be realised the cost of their formation would need to be feasible on a commercial scale. Macrocyclic chemists dream of yields over 50%, yet most chemists would be embarrassed to report a yield like this! However, we must remember that these reactions are inherently unfavourable so to achieve respectable yields we need to ‘stack the deck in our favour’ and try to make the interlocking of the rings more favourable.

Deceptively simple but incredibly effective ways of using what are commonly thought of as weak interactions have been used to hold the two fragments together, making the subsequent formation of the interlocked rings (rather than two separate rings) much more likely.  In a recent report in Chemical Communications, Fraser Stoddart and co-workers at Northwestern University, Texas A&M University and the King Abdullah City for Science and Technology report a significant development in catenane synthesis whereby yields of almost 90% are possible. As almost quantitative conversions are now being reported the commercial application of these molecules, which has for so long been discussed, moves another step closer.

Rapid thermally assisted donor–acceptor catenation
Albert C. Fahrenbach, Karel J. Hartlieb, Chi-Hau Sue, Carson J. Bruns, Gokhan Barin, Subhadeep Basu, Mark A. Olson, Youssry Y. Botros, Abdulaziz Bagabas, Nezar H. Khdary and J. Fraser Stoddart
Chem. Commun., 2012,48, 9141-9143
DOI: 10.1039/C2CC34427K

Researchers in Texas are using rotaxane formation to sterically protect or “insulate” molecular wires.

Molecular wires, in which an unsaturated linker separates two or more redox active metal sites, are of great research interest. These structures allow phenomena such as electron delocalisation or transport between the two redox sites. John Gladysz’s group at Texas A&M University have an ongoing interest in dimetallic polyynediyl complexes, in which two metal centres are linked by conjugated polyynediyl linkers that they now hope to “insulate” to reduce interactions between wires and the external environment. A previous approach used long alkyl bis-phosphine, which wrapped around the wire in a double helix to complex both metal centres. However, this gave two enantiomers, which interconverted rapidly in solution via uncoiling of the protective ligands.

The Gladysz group are now reporting a straightforward solution to this problem. They found that by synthesising their bis-platinum wire in the presence of a 33-membered macrocycle, they could incorporate the wire as the thread of a rotaxane complex. This provides a more robust protection for the wire which is unaffected by dynamic processes.

This work shows a fantastic application of rotaxane chemistry for protection of a molecular wire. What’s more, the synthesis of this rotaxane is adaptable, and the Gladysz group are working on exciting new and improved systems including longer polyynediyl linkers and redox inactive macrocycles to improve the properties of the insulated wires.

The full communication can be downloaded here.

Cally Haynes