IBM Demonstrates a Competitive Graphene Infrared Detector

Earlier this year, researchers at IBM’s Nanoscale Science and Technology group revealed some of the fundamental photoconductivity mechanisms of graphene.

The IBM researchers demonstrated that graphene can either be positive or negative depending on its gate bias. The positive is due to a photovoltaic effect and the negative is due to a bolometric effect.

The bolometric effect involves photo-generated carriers that, while propagating across graphene, emit quanta of lattice vibrations called phonons and thereby transfer their energy into the lattice. Heating up the lattice implies enhancing the electron-phonon scattering process and reducing the carrier’s mobility. The IBM researchers discovered this effect was dominant in the photo response of graphene and is what leads to the photocurrent flowing in the opposite direction of the source-drain current.

In new research, which was published both in Nature Communications (“Photocurrent in graphene harnessed by tunable intrinsic plasmons”) and Nature Photonics (“Damping pathways of mid-infrared plasmons in graphene nanostructures”), the IBM team has begun to explore ways to amplify this bolometric effect in graphene.

The research team, which includes Hugen Yan, Tony Low, Wenjuan Zhu, YanqingWu, Marcus Freitag, Xuesong Li, Francisco Guinea, Phaedon Avouris, and Fengnian Xia, began by first studying the fundamental property of plasmons in graphene metamaterials by purely optical methods, revealing important information about its dispersion and damping mechanisms. This knowledge guided them in their design of graphene photodetectors, leading to the first demonstration of a graphene infrared detector driven by intrinsic plasmons.

Graphene’s high mobility and zero gap nature gives it fast optoelectronic response and detection in an extremely broad spectral range from the visible over the infrared and into the terahertz range.

In the visible and near-IR, semiconductors are more efficient in detecting light than graphene because they can have matched bandgaps to a particular spectral window, and because a single layer of graphene absorbs only a small fraction of the incoming light. So it is very unlikely that we will some day be able to buy a cell-phone or camera with a graphene photodetector in it.

However, at lower energies, for example in the mid-IR or terahertz regime, graphene could be much more competitive and provide a unique technology solution. Currently, superconducting transition-edge detectors and bolometers are state of the art in these regimes, and these detectors are very expensive. The absorption in a single layer of graphene can be as high as 40 percent in the terahertz, and the window of high absorption can be moved into the mid-IR by patterning the graphene and harvesting graphene plasmons.

The graphene-based photodetectors, which utilize their intrinsic plasmons, have been demonstrated to yield an order of magnitude improvement in the device’s photo-responsivity in comparison to its non-plasmonic counterpart.

The graphene used in the photodetectors were first grown by CVD on copper foil. Copper was then dissolved in etchant, and finally graphene was transferred to a silicon/silicon oxide chip. The researchers built the graphene photodetector itself by patterning graphene into superlattices of graphene nanoribbons using e-beam lithography. The ribbons widths range from 80 to 200 nm and lateral confinement in ribbons provides the necessary momentum to couple with the graphene plasmons. It is then illuminated with a chopped CO2 infrared laser beam.

The researchers believe that graphene plasmonics could potentially provide a natural platform for a range of technologies in the infrared regime such as light detection and modulation, optical communications, photovoltaics, and spectroscopy.

With this basic understanding of how graphene plasmon disperses, damps, and generates photocurrent, the IBM team is now more confident about this line of research. The merging of graphene plasmonics with optoelectronics is a field that has essentially just began so there remain fundamental and technological issues to resolve.

Image: Tony Low

In the image, plasmon dispersion in graphene on silicon dioxide substrate, reveals coupling with long-lived substrate phonons. They can be excited by patterning graphene into nanoribbons.

Related Stories


IEEE Spectrum’s nanotechnology blog, featuring news and analysis about the development, applications, and future of science and technology at the nanoscale.


Dexter Johnson
Madrid, Spain