Although nearly a dozen reporters have interviewed electrical engineering Professor Stephen Chou since his single-electron transistor was unveiled in Science magazine last week, he insists that his newfound fame hasn’t gone to his head.
“That’s not my goal. My goal is to develop something new,” he said.
Invented in 1947 at Bell Labs, transistors are the fundamental devices of modern electronics. The integrated circuits currently used to make computers can have up to 100 million transistors.
A modern transistor has a positively charged end called the drain, and a negatively charged end called the source. The current between the drain and the source is controlled by a voltage applied to the midpoint of the transistor, known as the gate.
The single-electron transistor uses a floating gate that consists of a football-shaped dot of polysilicon only 7 nanometers (7 billionths of a meter) in diameter. To illustrate the minuscule sizes involved, 700,000 such gates would fit in the period at the end of this sentence. The transistor can be turned on by adding a single electron to the dot.
Chou believes that his transistor will be especially appropriate for the fast memory circuits computers use for random access memory (RAM).
“A transistor used for calculation is very different from a transistor used for memory function,” Chou said.
A transistor that is turned on can be used to represent a “one” in binary code while a transistor that is turned off might represent a “zero.” Computers use binary code to represent both instructions and numbers.
Current transistors require about 10,000 electrons to operate. The single-electron transistor will allow nearly 1,000 times more transistors to fit on a silicon chip without increasing the chip’s size.
“A smaller memory cell also leads to faster (computation) speed and lower power consumption,” Chou said.
Because an electron must be either present or not present, there is less opportunity for interference, or noise, to create an ambiguous result in a single-electron transistor than in a conventional transistor.
Chou said that the University is still debating whether or not to file a patent for his invention.
“If you are looking for an international patent, you should file before you publish,” Chou said. “We filed the invention disclosure to the University before the result came out. But the University is taking a long time to decide, and we cannot wait. For a U.S. patent, you have one year after the first publication of the work.”
Lingjie Guo, an electrical engineering graduate student who worked on the project, said the initial stages of the project were the most difficult.
“After you have this idea, you don’t know if it will work or not — it might totally fail,” Guo said.
Despite the success of his work , Guo doesn’t intend to rest on his laurels. “I think I have found a way to create a tunable quantum dot transistor,” he said.
A tunable transistor is one in which properties can be altered electrically. “Eventually, we may be able to change the actual size of the quantum dot,” he said.
Guo said that it may be some time before industry is able to take advantage of the transistor technology now coming out of the lab.”Right now, people don’t see the end of conventional transistors,” he said. “But in not too long — in the 21st century — people will get into the nanometer range.”