AFM-Based 'Etch-a-Sketch' Draws Nanoscale Features

Alexander E. Braun, Senior Editor -- Semiconductor International, 3/20/2008 8:00:00 AM

Using an atomic force microscope (AFM), researchers at the University of Pittsburgh have devised a way to draw, erase and rewrite — much in the way of the venerable but still popular Etch-a-Sketch toy — conductive <4 nm lines and 2-nm-diameter dots. The AFM draws the lines on an insulating layer, a 1.2 nm layer of lanthanum aluminum oxide on a strontium titanium oxide.

Jeremy Levy, a professor of physics and astronomy at the university, and his coworkers used oxide materials because they are similar to semiconductors and their properties are determined at the atomic scale. “The idea is to have one material, such as SrTiO₃, which tends to be thick, with another, such as LaAlO₃, which is very thin, ~1.2 nm, which is grown on top of the first,” Levy said. At the interface between these two materials, both of which are insulators or wide bandgap semiconductors, it is possible to get conduction or metallic behavior. “It was recently discovered that it is possible to switch reversibly between an insulating and a conducting state; this fact became the basis of our work. What we have been able to do is demonstrate that it is possible to locally switch this interface between the insulating and metallic state, which can be significant for a number of applications.”

Using an Etch-a-Sketch-like AFM process, researchers have achieved the capability to write, erase and rewrite conducting nanoscale lines and dots. (Source: University of Pittsburgh)

The research group focused on very simple features: mostly lines and dots — 1-D and 0-D objects. They were able to write conducting lines 3.3 nm wide and erase them, demonstrating that the development can be a reversible process. “We were also able to show very strong field effects in these wires,” Levy said. “We found that when applying a voltage over them using a conductive probe, which was also used to create these lines that we could, very slowly, turn off the conduction. This is somewhat similar to a FET, but it is semi-permanent; we can apply a voltage and modulate the conduction and when the probe is removed, the feature remains in that state. It shows extreme flexibility because it tends to be non-volatile.” Levi and his colleagues have been able to produce structures that could form the basis for transistors and have made isolated islands whose size approached 1 nm, which could be used in ultrahigh-density memory applications.

Presently, the researchers are focused on testing a theory that indicates that when a voltage is applied using a very sharp probe between the top surface and the interface at the low end, the resulting field is quite large and removes oxygen from the surface, oxizing the materials. “If you remove a certain fraction of oxygen atoms, then, theoretically, such an interface between insulating metallic states will change,” Levy said. “We have not yet proven that we are actually removing oxygen, so we still need to experiment further to exactly pinpoint what kind of change is taking place. However, we think that this is a likely source of this modification, and if you have oxygen or some other atoms present, the process may be reversed. Most of these surfaces have either molecular oxygen or OH groups, possibly from water, which have adsorbed to the surface.”

This theory, if proven, could open the possibility of the integration of logic and storage. Currently, magnetic storage or even flash memory are used, but these are different platforms from silicon-based computing architecture. If it is possible to not only combine those two functionalities in a single material, but also to scale beyond what has been done for silicon, it might be possible to enter a new regime in which scaling is done down to a size where single electrons play an important role. “The fact that the charge is quantized in units of the electron charge can lead to effects and devices that take advantage of this, such as single-electron transistors,” Levy said. “However, these are things that typically work only at very low temperatures, ~1 K, and it remains yet to be seen whether we can produce such a device capable of operating at room temperature.” If this were to happen, then it might make it possible to store information by using single electrons. Because the sensitivity to change is so tremendous, it should also be possible to produce an extremely sensitive sensor that can measure very small fractions of a single charge using single-electron transistors. This would be invaluable for biomedical applications.

“On a more exotic side, there have been reports of interfacial magnetism; however, so far this occurs at very low temperatures,” Levy said. “It may be that one can have these effects at higher temperatures, but that still remains to be determined. Superconductivity is another state that can be turned on and off at the same interface, but, again, that is a low temperatures.”

At the moment, despite the potential of these preliminary results, the University of Pittsburgh researchers are forced to take relatively small steps, hindered by insufficient funding.