Schematic depictions of graphene crystal structure (lattices), conduction band (blue cones and curves), valence band (yellow cones and curves), and Fermi level (dotted lines). Top: Monolayer graphene. Center: Bilayer graphene. Bottom: A bandgap (Δ) is induced in bilayer graphene by an external electric field (arrows).

Previously, in 2006, researchers at the ALS observed a bandgap in bilayer graphene in which one of the layers was chemically doped by adsorbed metal atoms. But such chemical doping is uncontrolled and not compatible with device applications. Researchers then tried to tune the bilayer graphene bandgap by doping the substrate electrically instead of chemically, using a perpendicularly applied, continuously tunable electrical field. But when such a field is applied with a single gate (electrode), the bilayer becomes insulating only at temperatures below 1 K, near absolute zero—suggesting a bandgap value much lower than predicted by theory.

To better understand exactly what was happening electronically, the Berkeley team built a two-gated bilayer device, which allowed them to independently adjust the electronic bandgap and the charge doping. The device was a dual-gated field-effect transistor (FET), a type of transistor that controls the flow of electrons from a source to a drain with electric fields shaped by the gate electrodes. Their nano-FET used a silicon substrate as the bottom gate, with a thin insulating layer of silicon dioxide between it and the stacked graphene layers. A transparent layer of aluminum oxide (sapphire) lay over the graphene bilayer; on top of that was the top gate, made of platinum.

Two-gated bilayer graphene. Left: Optical microscopy image of the bilayer device. Right: Illustration of a cross-sectional side view of the gated device.

Then, rather than measuring the device's electrical resistance, or transport, they decided to measure its optical transmission. The problem with transport measurements is that they are too sensitive to defects. A tiny amount of impurity or defect doping can create a big change in the resistance of the graphene and mask the intrinsic behavior of the material.

Using infrared Beamline 1.4.3, the researchers measured variations in the light absorbed by the gated graphene layers as the electrical fields were tuned by precisely varying the voltage of the gate electrodes. The absorption peak in each spectrum provided a direct measurement of the bandgap at each gate voltage.

The results from the ALS measurements were obtained with relative ease and efficiency, and showed that by independently manipulating the voltage of the two gates, the researchers could control two important parameters, the size of the bandgap and the degree of doping of the graphene bilayer. In essence, they created a virtual semiconductor from a material that is not inherently a semiconductor at all. Moreover, their experiment was conducted at room temperature, requiring no refrigeration of the device.

Left: Allowed transitions between different sub-bands of a graphene bilayer. Center: Gate-induced absorption spectra for different applied displacement fields. Absorption peaks due to transition I are apparent (dashed black lines are guides to the eye). The sharp asymmetric resonance observed near 200 meV is due to phonon resonances with continuum electronic transitions. The broad feature around 400 meV is due to electronic transitions II, III, IV and V. Right: Theoretical prediction of the gate-induced absorption spectra. The fit provides an accurate determination of the gate-tunable bandgap.

The researchers emphasize that these first experiments are only the beginning. The electrical performance of the demonstration device is still limited, and there are many routes to improvement, for example through extra measures to purify the substrate.

Nevertheless, they've demonstrated that we can arbitrarily change the bandgap in bilayer graphene from zero to 250 meV at room temperature, which is remarkable in itself and shows the potential of bilayer graphene for nanoelectronics. This is a narrower bandgap than common semiconductors like silicon or gallium arsenide, and it could enable new kinds of optoelectronic devices for generating, amplifying, and detecting infrared light.

Research conducted by Y. Zhang, T.-T. Tang, and C. Girit (University of California, Berkeley); Z. Hao (Berkeley Lab); M.C. Martin (ALS); and A. Zettl, M.F. Crommie, Y.R. Shen, and F. Wang (University of California, Berkeley and Berkeley Lab).

Research funding: U.S. Department of Energy, Office of Basic Energy Sciences (BES), Alfred P. Sloan Foundation, Miller Institute for Basic Research in Science, and National Science Council of Taiwan. Operation of the ALS is supported by BES.

Publication about this research: Y. Zhang, T.-T. Tang, C. Girit, Z. Hao, M.C. Martin, A. Zettl, M.F. Crommie, Y.R. Shen, and F. Wang, "Direct observation of a widely tunable bandgap in bilayer graphene," Nature 459, 820 (2009).

ALSNews Vol. 301, August 26, 2009