Graphene was only discovered in 2004 but from the beginning has showed promise for use in fast and ultra-low power devices thanks to the fact that electrons move through the material at extremely high speeds. Often touted to replace silicon as the material of choice in future electronics, there are, nevertheless, some serious challenges that need to be overcome before real-world devices see the light of day.

One of the main problems is that graphene is a zero-gap semiconductor and cannot be used for digital electronics applications. To open a gap of just 1 eV in the material would mean making graphene ribbons smaller than 2 nm across with single atom precision, something that is beyond current technology.

Attractive alternative
Bilayer graphene offers an attractive alternative to single-layer graphene because a tuneable energy gap can be induced in the 2D sheet by applying a vertical electric field. However, this still only produces an energy gap of a few hundred meV, which rules out using bilayer graphene in conventional FETs. The main drawback here is the large band-to-band tunnelling current, which prevents the transistor from being properly switched off.

Using numerical simulations, Gianluca Fiori and Giuseppe Iannaccone of the University of Pisa have now shown that this problem can actually be turned into an advantage when making TFETs thanks to the material's extremely low subthreshold swing of 20 mV/decade. Such a low subthreshold swing allows a large on-off current ratio close to 104 with a supply voltage of just 100 mV. "It is possible to use devices made of bilayer graphene for ultra-low power applications, which will avoid all of the power consumption problems that are currently limiting the scale-up of next-generation devices," Fiori and Iannaccone told nanotechweb.org.

The graphene bilayers can also be easily grown on dielectric layers and do not need to be laterally defined, which means that they can be fabricated using state-of-the-art lithography.

Fiori and Iannaccone admit that their analysis has been performed on ideal structures that have no impurities, defects or electron interactions with lattice vibrations (phonons) and say that they will study more realistic structures in the future. "We are eager to collaborate with experimental groups to do this," they added.

The research, which was funded by the EC FP7 GRAND project, the ESF FoNE program (DEWINT project) and by the EC FP7 NANOSIL NoE, was reported on arXiv.