French Team Demonstrates Terahertz Transistor

The creation of a nanotransistor that generates a terahertz signal by means of plasma waves was reported by researchers in France in the 29 March issue of Applied Physics Letters . Such devices could fill a gap between 0.3 and 3 terahertz, for which no compact solid-state sources are available, says Wojciech Knap of the University of Montpellier, leader of the multinational team that published the report.

The terahertz gap occupies the region between infrared radiation and microwaves. Recently, that gap has come into the limelight because of some especially interesting properties it has. For one, radiation in this range can traverse matter just as X-rays do, but because this radiation is nonionizing, it does not have the harmful biological effects of X-rays. For another, many complex biomolecules are invisible at other wavelengths but they absorb terahertz waves, making it possible, for example, to detect spores or microorganisms dangerous to humans.

For its experiment, the French team used a high electron-mobility transistor with an indium gallium arsenide channel and a T-shaped 60-nanometer gate, which controls the flow of electrons through the channel. The transistor was fabricated using electron-beam lithography at the Institute of Electronics, Microelectronics and Nanotechnology at the University of Science and Technology at Lille, one of the most advanced nanotechnology institutions in Europe, according to Knap.

The channel contains a two-dimensional electron gas that behaves like a plasma (a gas consisting of charged particles) locked up in a cavity. The cavity for the gas is created when the 60-nm-wide gate is shorted to the source. A dc voltage applied to the drain terminal (the source is grounded) creates a current in the channel. Increasing the drain voltage increases the source-drain current, and if this current exceeds a certain value, instabilities arise in the 2-D plasma, giving rise to plasma waves.

The researchers observed the onset of plasma waves at 0.2 volt. Increasing the source-drain voltage above this threshold causes the emission to increase in intensity, while its maximum frequency shifts from a fraction of a terahertz to 1 THz for a source-drain voltage of 0.8 V.

These waves bounce back and forth between the boundaries of the gated 60-nm-long channel. As in a laser cavity, the waves increase the amplitude of the plasma wave upon successive passages in the cavity. The oscillating charge density waves lead to electromagnetic radiation that leaves the device by the metallic drain and source contacts, which act as imperfect antennas. The radiation will be more intense with the incorporation of half-wave antennas of about 150micrometers, which are now being manufactured in Lille.

The generation of these terahertz waves in nanometer devices was predicted by Mikhail Dyakonov of the University of Montpellier and Michael Shur, now at Rensselaer Polytechnic Institute (RPI) in Troy, N.Y. They correctly forecast that the frequency of the waves would increase with the decreasing size of the cavity, and they argued that these waves, also called plasmons, are density waves that can be compared directly to the waves generated by air passing through an organ pipe or flute.

For a long time, the terahertz emission could not be demonstrated because small enough devices that had sufficient electron density and mobility were not available. But progress in nanotechnology came to the rescue. "We had to wait 10 years before we could verify this phenomenon experimentally," says Knap, whose group included researchers at Lille, RPI, the University of Warsaw, and the Institute of Radio Engineering and Electronics in Saratov, Russia.

To shield the spectrometric detector from external thermal radiation, both the transistor and the spectrometer for recording the still extremely weak emission were placed in a copper waveguide cooled with liquid helium to 4.2 Kelvin and sealed from the environment. But the transistor should also work at room temperature, and by putting large numbers of nanotransistors on a single chip, the emissivity would be increased.

With an output of several watts per square centimeter, the transistor should be able to compete with existing terahertz sources, Knap claims. Currently, his team is experimenting with silicon nanotransistors, but because these devices have a lower electron mobility, the gated area in the channel would have to be much smaller, probably less than 30 nm, says Knap.

Paul Planken, who researches microscopy applications at terahertz frequencies at the Delft University of Technology in the Netherlands, says that we must now wait for an independent confirmation of the terahertz emission, as well as for a demonstration showing that the transistor emits at room temperatures. "If others can confirm this, this would be great--we would have a new terahertz source," Planken says. Michael Pepper, a physicist at the University of Cambridge in the United Kingdom, agrees that if the effect can be observed at room temperature, interesting applications could be developed.