18 April 2012—Seeking to surpass the capabilities of silicon electronics, graphene researchers keep creating graphene-based transistors that operate at ever higher frequencies. But another quality important in analog radio-frequency (RF) devices—intrinsic voltage gain—has lagged. Now researchers in Germany and Italy have shown that by layering one sheet of graphene on top of another, they can make a significant improvement in the gain of graphene transistors.
“If you don’t have large voltage gain, the transistor more or less is useless,” says Gianluca Fiori, a researcher in the Nanoscale Device Simulation Laboratory at the University of Pisa, in Italy. He likens a transistor with a high cutoff frequency but low voltage gain to an aerodynamic sports car with a weak engine.
Graphene, an arrangement of carbon atoms in a single atomic layer, has remarkable electronic properties, which researchers have used to build analog devices with frequencies of up to 300 gigahertz. IBM, for example, has built a broadband RF mixer, a fundamental component of radios. The hope is that such devices will outperform their silicon cousins.
Faster and faster graphene transistors are continually coming out laboratories at IBM, the University of California, and elsewhere. “Every two or three months, they beat the record they have established before,” Fiori says. “There has not been so much attention paid to get larger voltage gains.”
But by using two layers of graphene to build the channel that carries charge between the source and drain electrodes in a field-effect transistor (FET), Fiori's team achieved a voltage gain of 35, a sixfold improvement over a single-layer FET. The results were published in March in the journal Nano Letters.
The problem with devices made using just one layer of graphene, explains Fiori, is that they have low output resistance, which translates into poor voltage gain. One way engineers increase the frequency of transistors is to shrink their dimensions, but in a shorter channel the resistance drops further, making gain more of a problem. Adding the second layer increases the resistance without harming the transistor’s maximum frequency, Fiori says.
The bilayer device, constructed by Fiori’s colleagues at the nanofabrication company AMO, in Aachen, Germany, had a channel that was 4 micrometers long. But simulations show that the gain effect will remain even if the device is shrunk down to 40 nanometers. At that length, bilayer graphene should produce a voltage gain of about 10. “Even if you shrink the device, you still get larger output characteristics than you see in a monolayer,” Fiori says.
Tomas Palacios, associate professor of electrical engineering and computer science in the Advanced Semiconductor Materials and Devices group at MIT, calls the work “an important step in the direction of improving the performance of graphene amplifiers.” High frequency is necessary for such devices, Palacios says, but “it is also important to have voltage gain.”
Other researchers have been using bilayer graphene in an attempt to address graphene’s natural lack of a bandgap. Without a bandgap—a range of energy states where electrons cannot exist—current flows unimpeded, and the on-off distinction needed for digital logic can’t be achieved. Fiori says the bandgap introduced in the device he worked on, about 100 millielectron volts, is not large enough for digital logic, but it's good enough for analog applications such as amplifiers, he says.
About the Author
Neil Savage, based in Lowell, Mass., writes about strange semiconductors and amazing optoelectronics. In the April 2012 issue he reported on molybdenum disulfide, a potential rival to graphene in nanoelectronics.