"We have to stress that we did not just make a single graphene nanoribbon but a large number of GNRs in parallel," said team leader Christos Dimitrakopoulos of IBM. "These arrays cover about 50% of the finished device channel area, which means that integrated circuits based on GNRs with the required high current densities are now possible."

The researchers say that they were also able to produce GNRs with well controlled dimensions and that had smooth edges. Edge roughness is very important in tiny structures like nanoribbons because it can seriously degrade graphene's exceptional electronic transport properties. The process they developed to make the GNR arrays is a hybrid one consisting of two main complementary steps: a top-down e-beam lithography step (that can also be performed using standard photolithography with an appropriate mask) and a bottom-up self-assembly step involving a block copolymer template comprising alternating lamellae of the polymers PS and PMMA.

Addressing graphene's shortcoming

Although graphene has numerous unique mechanical and electronic properties (including high electron mobility and extremely high strength), it does suffer from an important drawback, which is that it has no gap between its valence and conduction bands. Such a bandgap is essential for electronics applications because it allows a material to switch the flow of electrons on and off. However, one way of introducing a bandgap into graphene is to make extremely narrow ribbons of the material – as in this work.

"By fabricating dense arrays of 10 nm wide GNRs, which are theoretically predicted to have a bandgap of about 0.2 eV (and doing so with manufacturing-friendly processes), we have taken an important step towards addressing this shortcoming of graphene," Dimitrakopoulos told nanotechweb.org. "Our process is now available to anybody for fabricating and characterizing devices based on such GNRs."

Although much work still needs to be done before GNRs find their way into applications, the next logical step would be to try to make and characterize GNR devices, such as field-effect transistors, using our methodology, adds Dimitrakopoulos. "It will also be important to study the structural quality of GNR edges and experiment with different edge-terminating atoms. Increasing the bandgap by further reducing the width of the GNRs would also be good thing to aim for."

The current work is reported in ACS Nano.