Graphene switch behaves like phase-change memory
Graphene, a zero-gap semi-metal, can reversibly switch between an insulating and conductive state, according to new work by researchers in Germany. Such a result has never been reported before, either in experiments or theory, and opens up a host of new potential applications.
The graphene switch made by Tim Echtermeyer and Max Lemme of AMO in Aachen and colleagues is not much different in design to a silicon transistor (MOSFET). The researchers deposited graphene, obtained by exfoliating layers of graphite, onto a SiO2 covered silicon wafer. They then added source and drain electrodes to the device.
Next, the scientists covered the graphene with a silicon oxide gate insulator and top-gate electrode. “When we apply high voltages to the top gate, we believe we induce a reaction with water-related species in the gate oxide with the graphene underneath the silicon oxide gate insulator,” explained Echtermeyer.
Normally, the carbon pi-electrons in graphene are delocalized, which is why the material is an excellent conductor. According to the researchers, once the devices are switched off, the water-related species are chemisorbed by the graphene. This causes the carbon pi-electrons to become localized, which destroys the conductivity of the graphene. “Appropriate top-gate voltages or heating the graphene using current pulses desorbs the attached species and restores the conductivity,” said Echtermeyer.
The new carbon-based “resistive” switch could find use in non-volatile memory applications and can be likened to a phase change memory, adds Echtermeyer. However, thanks to the large on/off ratio of >106 the new graphene switch could allow for multilevel memory.
The team now plans to investigate how to better control its device, improve switching speed and cyclability.
“It is remarkable that graphene is fully open to controlled reversible surface modification,” said Echtermeyer. “And it is exciting that molecules can be adsorbed and desorbed while graphene’s electronic properties are conserved – something that may lead to new applications that were not considered feasible until now.” These include non-volatile memory, sensing and multilevel logic.
The work, which was carried out as part of the ALEGRA project, was published in IEEE Electron Device Letters
Aug 29, 2008
Author: Belle Dumé is contributing editor at nanotechweb.org