Drawing a fine line

 作者:麦煊格     |      日期:2019-03-08 04:18:02
By Michael Brooks THE small print could soon become a lot smaller, thanks to nanotubes. Researchers from Michigan State University and the University of Toronto have designed a nanometre-scale “fountain pen” that they say will be able to precisely deposit atoms on surfaces with a kind of pump. Although it is still a design proposal and not a prototype, they say that all the technologies necessary to make their instrument are currently in use. If it can be made to work, the atomic pen could take over from fine line lithography in microchip manufacturing, allowing the dimensions of written features such as connecting wires and transistor contacts to move below today’s micrometre-scale limits. It could also be used for atom-by-atom construction of micromachines. Developments in nanotechnology are hampered by a lack of tools for handling parts on atomic and molecular scales. The best current method of atomic manipulation, which uses a scanning tunnelling microscope (STM) to move single atoms around a surface, is slow and laborious. But David Tomanek and Peter Kral estimate that their atomic pump, which could be preloaded with atoms, might deposit them on a surface at a rate of one every 15 microseconds. The “ink cartridge” of the pump is made from a carbon nanotube—a sheet of carbon atoms rolled into a tube—and can be filled with whatever atoms are required. Two interfering laser beams, one with twice the frequency of the other, control the flow of atoms. The lasers hit the tube and generate electrons from the tube walls by a photoelectric effect. Interference between the laser beams moves the electrons along. As they flow, the electrons exert a force on the trapped atoms, pumping them along. By controlling the relative phase and power of the laser beams, an operator can direct the speed and direction of the flow, spitting an atom out of the tip of the tube whenever it reaches the right position. Building the instrument will be an exacting test, the researchers say. Any constrictions and defects in the nanotubes would have to be smoothed out to avoid blocking the flow of atoms. Moreover, the geometry of the tip will be important in ejecting atoms from the tube efficiently. The best design for this is still unknown. Heating effects may also cause a problem, since very high-powered lasers will be needed to generate a useful current. In their paper, which is due to appear in a forthcoming edition of Physical Review Letters, Tomanek and Kral suggest that atoms held between nanotubes bundled together in a rope-like structure might clog less and provide an easier escape from the tip. This mirrors Austrian work on moving individual atoms down the outside of a wire (This Week, 13 March 1999, p 16). The heating effects could be reduced by using short, 15-microsecond laser pulses: each pulse would still be enough to move an atom the length of a nanotube, pumping it out onto a substrate. Mauricio Terrones, a nanotube expert at the University of Sussex,