Isaac Newton wouldn't approve of what Sandia National Laboratories is doing to the tried-and-true transistor.
Conventional transistors are based on classical physics principles, but basic science researchers at Sandia Labs in Albuquerque, New Mexico, recently demonstrated a new transistor that makes use of quantum theory - specifically, it uses a technique called quantum mechanical tunneling of electrons.
"It's pretty hard to make high-frequency transistors. This is true even after the federal government has pumped millions into funding university research. So considering that context, we're pretty happy with our results," says Jerry Simmons, head of the nanoelectronics group at Sandia.
Called the Double Electron Layer Tunneling Transistor, or DELTT, the three-terminal quantum transistor is considered a breakthrough in transistor design because quantum tunneling leapfrogs current transistor technology in terms of its extremely high speed. The transistor will be capable of blistering speeds by relying on a single quantum transition between two electron states, something traditional transistors are unable to do.
Quantum tunneling can also allow the same integrated circuit functions to be achieved with far fewer transistors. And because the transistor only needs existing nanotechnology to build its atomically precise two-dimensional layers of electrons - a new level of miniaturization for transistors - the DELTT is expected one day to be reliably manufactured in large numbers using existing semiconductor production facilities, a challenge long thought to be a limitation with quantum transistor design. Manufacturing the quantum transistor is years in the future, however.
"Right now it's a research device. We've only demonstrated it in the past year," says Simmons. "I don't want to get in there worrying about production yield - it's my function to demonstrate that it's worthwhile to move it into the next stage."
Simmons expects the device to work at room temperature by the end of next year, but even then it may be several more years before the technology is ready for production. "A real chip needs a million working transistors, and thus extremely high yields. We're on our way, but there still are obstacles to commercialization, and I can't say when it will actually be manufactured," he says.
Regardless of its manufacturing timeline, transistor researchers who have examined the fundamental research behind the chip say Sandia has produced a major advance in electronic design.
"This is a premier piece of research," says Paul Berger, associate professor of electrical and computer engineering at the University of Delaware. "I've seen some of these things before, but nothing with this kind of complexity and sophistication. We're going to have to look for alternative ways of computation, and this quantum transistor is certainly noteworthy and needs to be addressed."
Future applications for the quantum transistor include cellular phones and eventually desktop CPUs, Simmons says. Ultimately, Simmons hopes the transistor will find itself in three broad categories of devices: ultra-low-power static memory elements, which will require exceptionally low currents; ultra-high-speed logical processing elements, which will require very small feature sizes and somewhat higher currents; and far-infrared detectors.
Sandia's interest in transistor design is one part of the government laboratory's larger charter, which is to assure the safety, security, and reliability of nuclear weapons. As part of this mission, the 7,500-person lab also conducts research into basic science. While at some point the quantum transistor is likely to find itself attached to a nuclear warhead, the national defense laboratories are also increasingly participating in technology transfer programs with civilian industries.
The Sandia team will present their research advance at the International Electronic Devices Meeting in Washington, DC, on Monday, and will publish various aspects of their research in the upcoming issue of the journal Applied Physics Letters.
