Controlling the formation of nanostructures at the molecular scale is a great technological challenge in the design of electronic circuits and devices. Molecular manipulation is a tricky business and, although there is scope for greater miniaturization of electronic components, future developments will require an even closer integration between chemistry, biology, physics and engineering. The most promising approaches so far — using scanning tunnelling microscopes (STM) or self-assembly processes — are either too slow or do not allow accurate monitoring of the growth of these structures.

Elsewhere in this issue (page 48), Robert Wolkow and colleagues describe how they have created one-dimensional organic structures on a silicon surface, by combining STM and self-directed growth techniques. The picture on the left shows an STM image of a straight line of styrene molecules (C8H8; red) attached to a silicon surface. This is an ideal substrate for the assembly of straight ‘wires’ because it is anisotropic; that is, the growth process proceeds in one direction, which depends on the crystallographic orientation of the surface.

How are these molecular wires produced? The first step is to expose a clean surface of silicon to atomic hydrogen in an ultra-high vacuum to obtain a surface terminated by hydrogen atoms. A hydrogen atom is then removed from the surface with the tip of an STM to create a single silicon dangling bond, which will then spontaneously react with any available styrene to form a silicon–carbon bond. To compensate for this newly reactive carbon species, it is likely that a hydrogen atom is removed from a neighbouring surface site, leading to a chain reaction. Theoretical calculations have shown that anisotropic surfaces similar to the silicon substrate are energetically favourable to this sort of chain reaction.

The end result is a line of styrene molecules bonded to the surface. The longest chains created were 13 nm long. The phenyl rings of the styrene molecules were separated by 0.4 nm (see image), which corresponds to the intermolecular distance between two hydrogen atoms on this surface. Although it is not possible to tell whether the rings are actually parallel, some degree of intermolecular coupling between adjacent rings would be necessary for these chains to function as molecular wires.

Figure 1
figure 1

The phenyl rings of the styrene molecules were seperated by 0.4nm

Full size image

As the authors point out, the process could be applied to other organic molecules, such as alkenes and alkynes, which may be used to design conducting wires. Many factors need to be explored to help unravel the exact growth mechanism, but this approach is already sufficiently under control for one to imagine the rapid formation and connection of many nanostructures in parallel.