Complex molecular architectures are usually constructed by connecting a variety of building blocks together in a stepwise manner. But sometimes, complex structures emerge from the self-assembly of a single constituent. Writing in Nature, Datta et al.1 show how polycatenanes chains made of interlocked nanometre-scale rings can be formed by the remarkable self-assembly of a simple molecular building block.
Read the paper: Self-assembled poly-catenanes from supramolecular toroidal building blocks
Catenanes are molecules in which two or more molecular rings are entangled like the links of a chain2; indeed, their name derives from catena, the Latin word for chain. The rings are not connected by a covalent bond, but instead form a different kind of linkage called a mechanical bond, in which the connected rings can move freely around each other. This dynamic property makes them useful as components of artificial molecular machines3. Many catenanes reported so far consist of only two rings. The construction of molecular chains made of several rings is a major challenge for synthesis, and has been achieved only in the past few years, for small molecular rings (with a radius of approximately 1 nanometre)4.
The construction of larger systems is limited by the efficiency of the catenation step, in which a preassembled toroid precursor forms a ring that interlinks through another toroid; moreover, a large number of covalent bonds must be formed in the preassembled structure. Synthetic routes that involve non-covalent assembly techniques are therefore preferred5. In supramolecular polymerization, for example6, simple molecular building blocks self-assemble in a single step through non-covalent interactions to form large-scale structures of varied geometries. Unfortunately, the gain in size of the assemblies that are made in this way often comes at a price: chemists have less control of the final constructs architecture, compared with a multistep covalent strategy.
Datta and co-workers have combined aspects of covalent and non-covalent strategies to form their complex polycatenane structures. The authors started with a monomer composed of a polar head and a non-polar tail, separated by a rigid section consisting of benzene rings (Fig. 1). Six of these monomers can self-assemble in an appropriate solvent to form a star-shaped rosette. The polar heads form a hexagonal core that is held together by hydrogen bonds in much the same way that DNA helices are held together by hydrogen bonds in nucleotide base pairs and the rigid sections point outwards from the core like arms.
Figure 1 | Self-assembling polycatenanes. Datta et al.1 report that molecules consisting of a polar head, a rigid central region and a non-polar tail self-assemble into rosettes, which then stack together into fibres that form various nanoscale structures, such as toroids, helicoids and random coils. The authors devised a protocol in which they separate the toroids, and use them to seed the formation of new catenated rings thereby producing polycatenane structures in which up to 22 toroids are connected in linear chains or branched systems.
Once formed, the rosettes self-assemble by stacking on top of each other a process driven by the formation of interactions (known as interactions) between the rigid regions of neighbouring rosettes. Because each rosette added to the stack is slightly offset from its predecessor, the resulting assembly grows with an intrinsic curvature that produces various geometries: random coils, helicoids and toroids7. The type of geometry that forms depends on the rate of cooling of the initial monomer solution. Slow cooling (about 1 kelvin per minute) favours the formation of helical fibres; faster cooling (about 10Kmin1) generates random coils; and abrupt cooling adds toroids into the mix.
Datta and colleagues noticed that rapid cooling also produced traces of catenanes consisting of two interlocking toroids. This suggested that individual toroids could act as secondary sites from which another ring could grow, thereby forming the catenated dimers. The authors took advantage of this fortuitous process to devise a protocol for making large, self-assembled polycatenanes by using a solution of toroids as seeds for catenation.
The authors rapidly cooled a solution of the monomer in a solvent mixture that was chosen to facilitate toroid formation, and thereby produced a solution in which approximately half of the monomer molecules were incorporated into toroids; the remaining monomers self-assembled into randomly coiled linear structures. Because the toroids are more stable to heat than are their linear counterparts, the authors could selectively disassemble the coils back into monomers by heating the solution. Subsequent slow cooling promoted the formation of long, helical supramolecular assemblies from the monomers, leaving the toroids intact. Datta et al. then filtered the mixture to remove the long helical structures, thus producing a solution that predominantly contained toroids.
Molecular machines swap rings
Finally, the authors produced polycatenanes by adding monomers to the solution of toroids, which seeded the formation of new catenated rings, as had been hoped. The non-polar tails were originally incorporated into the monomers to improve monomer solubility, but Datta and colleagues found that they also have a crucial role in the seeding process: unfavourable interactions between the tails and the solvent makes it more likely that rosette self-assembly will initiate on the surface of existing toroids.
Atomic force microscopy revealed that polycatenanes of various sizes form in the reactions, and that the toroids have a radius of 12.5nm. The authors found that addition of monomers in small portions favours the initiation of self-assembly processes that lead to catenation and were thus able to produce linear and branched polycatenanes containing up to 22 rings. This is close to the number previously achieved using covalent assembly (up to 26 rings in linear polycatenanes)4, and further demonstrates the effectiveness of Datta and colleagues approach for synthesizing complex, non-covalent structures.
The authors protocol also shows that a multistep approach, borrowed from the covalent-synthesis playbook, can be used to produce large and complex self-assembled architectures in a controllable way. This is an important step in the development of non-covalent synthesis, and it can be expected that their protocol will inspire the field to tackle more-challenging targets5. It would be interesting to see, for example, whether the monomer can be adapted to obtain catenated rings of various sizes, or whether hetero-catenanes can be made, in which the seed toroid consists of a different type of monomer from the one from which the catenated macrocycles are assembled.
It remains to be seen how the mechanical and dynamic properties of the self-assembled polycatenanes compare with those of their smaller covalent counterparts. A main appeal of covalently assembled catenanes is that, if the relative motion and position of the rings can be controlled, it opens up potential applications for molecular machines. The possibility of achieving the same level of control over large, self-assembled structures would bring us a little closer to what nature achieves with cellular machinery.