Self-Assembled Molecular Grid Hosts Ordered Layers of Buckyballs

In, on, and around cells, myriad molecules execute specific tasks. Enzymes catalyze reactions, ribosomes manufacture proteins, receptors transmit signals, and so on. The molecules owe their impressive specialization to their bulk: The larger and more complex a molecule, the more sophisticated its function.

Scientists want to mimic nature’s ability to craft large, functional molecules. And they want to marshal the molecules into smart materials or nanoscale devices. Already, using a technique called molecular self-assembly, one can synthesize intricate cagelike structures a few nanometers across. Now, a group from the University of Nottingham in England has shown that molecular self-assembly can also be used to arrange large molecules on a surface.

Led by physicist Peter Beton and chemist Neil Champness, the Nottingham group has created a honeycomb-shaped grid of organic molecules on a passivated silicon surface. ¹ Other groups have made self-assembled arrays. What makes the Nottingham grid remarkable is the large size of the hexagonal holes. At 3 nanometers across, the holes can accommodate ordered layers of buckyballs.

Neither the molecules that make up the grid nor the buckyballs that fill its holes are especially functional. Even so, the same sort of spontaneous surface structuring could serve as a platform for future nanotechnologies.

Hooking up

The key to self-assembly, as the third little pig might advise, is choosing the right building materials. Two or more unreactant molecular species will normally mix and mingle without forming lasting bonds. But if the species have complementary shapes and charge distributions, the molecules can hook up via van der Waals, hydrogen, or other noncovalent bonds.

The Nottingham team chose two aromatic compounds: the rectangular PTCDI and the triangular melamine. When a PTCDI molecule meets another PTCDI molecule, the pair forms two hydrogen bonds; two melamine molecules do the same. But, as shown in the top figure, when PTCDI meets melamine, the two form three hydrogen bonds. Thanks to the extra bond, PTCDI and melamine preferentially bind to each other. And because of the two species’ geometry, a honeycomb network is the result. The middle figure shows the network’s unit cell.

Picking the right substrate is also important for assembling molecules on a surface. If the substrate is too sticky, molecules won’t circulate and will fail to find complementary bonding partners. The Nottingham researchers settled on a material they’d worked with before as their substrate: silver-passivated silicon(111).

To create the grid, the team first deposited a sparse layer of PTCDI at room temperature and at ultrahigh vacuum (about 5 × 10⁻¹¹ torr). Left to themselves on the surface, PTCDI molecules congregate in chains or islands with no particular pattern. But when the temperature is raised to about 100°C, the PTCDI molecules break free and diffuse over the surface. Depositing melamine molecules at this stage leads to the formation of a regular honeycomb-shaped grid.

To see whether they could fill the grid, the Nottingham researchers wafted buckyballs over the surface. Buckyballs are relatively easy to make and handle. And, as the middle figure shows, a flat cluster of seven closely packed buckyballs is just the right size to fill a hole.

Heptameric buckyball clusters duly populated the holes. The bottom figure shows an STM image of the surface, taken at a coverage of 0.03 monolayers of buckyballs. The PTCDI–melamine grid appears purple and blue; the clusters appear yellow and red.

What occurred at higher coverages is perhaps more intriguing. Once all the holes had been filled with heptameric clusters, the buckyballs formed a gridlike second layer above, and commensurate with, the PTCDI-melamine grid. Because the second layer lacked sticking points for additional buckyballs, the deposition reached a natural stopping point.

Though far from potential applications, the Nottingham group’s work has several interesting implications. The grid, the buckyballs, and the substrate could all be chemically altered to change their properties.

For example, one might be able to increase the size of the holes by inserting a midsection into each PTCDI molecule before assembly. Or by creating a square version of melamine, one could build a square grid. And given its scale, the spontaneously formed grid could have a role in molecular electronics.