Graphene-based transistors may be on the horizon

Graphene is today's miracle material. It is strong, it is flexible, and it conducts electricity. It possibly also wears its underpants on the outside of its tights and rescues kids from burning buildings.

The conductivity part is perhaps the most exciting. Graphene allows electrons to move ballistically, meaning they don't face any resistance, but require an external voltage in order to move. That's actually a blessing and a curse. Modern electonics relies on materials that can be switched from being good conductors to poor conductors by a control voltage. If graphene naturally conducts and that can't be changed, it might not be all that useful for electronics. A couple of papers in Nature Physics report that all is not lost and, with the right structuring, three layers of graphene allows it to be turned from insulator to conductor via a control voltage.

All of this comes about because graphene is not a metal-like conductor. In metals, there are lots of free electrons floating around occupying a continuum of states. These reach right down into the energy regions where one would expect electrons to stay bound to individual atoms. Graphene, on the other hand, is more like a semiconductor—its conducting electrons fall into a discrete range of states, and the lowest energy of these just happens to coincide with that of the highest energy of a bound electron.

Because these two energy states are just barely touching, there is always the hope that you can manipulate the graphene to shift them apart. This shift would create a bandgap, an energy gap between electrons bound to atoms and those that are free to move around. If you can do that shifting on demand, then you have created a graphene switch and the road is open for graphene based electronics and companies like Samsung  could make their shareholders very happy. But how to do it?

It turns out that, although a single layer of graphene has no bandgap, two layers do. Unfortunately, it can't be switched on and off. Researchers were disappointed that three layers of graphene didn't improve the situation.

The natural stacking of graphene provides a mirror symmetry. When you lay the first layer of hexagons down, the second layer will be offset somewhat from the first layer. The third layer, however, directly overlays the first layer. If you then apply a voltage across the layers, whatever effect the first layer has on the inner layer is exactly countered by the top layer. The result is that, yes you have a bandgap, but you can't control it.

Now, a large number of researchers from six or seven institutions have published two papers demonstrating that, if you change the way the graphene stacks, you obtain a voltage-controlled bandgap. In this work, the third layer of graphene does not overlap the original layer, but is offset even further. This breaks the mirror symmetry so that a voltage applied across the sheets will alter their conductivity.

The two groups of researchers showed this in slightly different ways. One group observed the photoconductivity of their graphene sheets as a function of wavelength and applied voltage. They showed that the oddly stacked three layer graphene sheets would generate a larger current for particular colors. That is, the light was exciting electrons out of bound states and into conducting states, indicating the presence of a bandgap. Furthermore, this color changed depending on the applied voltage, indicating that the bandgap was changing with the voltage.

The second group used a more traditional approach, where they measured the conductivity of the graphene sheets as a function of voltage across the sheet. This involved making graphene transistors and lifting the graphene away from the substrate, so, technically, it was a more challenging experiment. However, it is also an experiment immune from claims that the bandgap dependence comes from interactions between the graphene and its substrate. They also went further and looked at how the current depends on temperature and applied magnetic field.

Between these two papers, a fairly complete understanding of the bandgap behavior in three layer graphene has been obtained, leaving only the challenge of making the stuff. Graphene is fairly easy to make, and making structures out of graphene is also fairly easy. But making graphene layers with specific properties is proving to be quite a challenge, so I suspect that this is where the research must be focused before graphene electronics will leave the lab.