PORTLAND, Ore. IBM Corp. researchers have demonstrated that graphene, an atoms-thin layer of crystalline carbon, acts as a receiver of optical signals.
In the claiming the world's first graphene photodetector, scientists at IBM's Thomas J. Watson Research Center (Yorktown Heights, N.Y.) predicted it could operate at up to terahertz frequencies. The 40-GHz graphene photodetector demonstration matched IBM's earlier demonstration of a nanotube optical emitter, the second component needed for bi-directional carbon chip-based optical transceivers.
Graphene is formed as a single atomic layer of carbon atoms that, unlike metals, have no bandgap. Optical materials ordinarily require a bandgap so that photons can be absorbed (received by detectors) or given off (transmitted by emitters). Both emitters and detectors are required for an optical transceiver.
IBM previously solved the carbon chip emitter problem by injecting electrons and holes into opposite ends of a light-emitting nanotube. It now claims to have solved the carbon chip receiver problem using nanoscale p-n junctions formed by the electric fields surrounding its metal contacts in a graphene field-effect transistor.
"Graphene can form the basis of a very fast photodetector that has unique properties, despite the fact that graphene doesn't have a bandgap," said IBM Fellow Phaedon Avouris, who oversees its carbon-based materials efforts. "When you shine light on graphene, you generate electron-hole pairs, which ordinarily just recombine. But near electrode contacts, there are built-in fields generated, and these fields separate the electron-hole pairs generating a current."
Photodetectors work by separating the electron-hole pairs generated by the energy of incident light. Avouris claimed that graphene has several unique qualities, such as indifference to wavelength. This differs from most optical materials, which work well only at particular wavelengths. IBM claims other advantages include a zero-source drain bias, low dark current, high efficiency (up to 16 percent) and speeds approaching the terahertz range.
"We looked at the frequency response of these photocurrents by modulating the incoming light source at very, very high frequencies--up to 40 GHz--then watched to see how quickly the photodetector could follow the modulation of the light," said IBM researcher Fengnian Xia. "We found that it was following exactly all the way to the limits of our instrumentation."
IBM said its current instrumentation limited its measurements to 40 GHz. A commercial device made with an expensive material like palladium electrodes would probably limit speeds to about 600 GHz.