Researchers close in on single-atom switch
By Paul Kallender
EE Times
(03/07/01 04:50 a.m. PST)
TOKYO — After more than a decade of research using scanning tunnel microscopes (STMs) to precisely manipulate polymers, Japan's Institute of Physical and Chemical Research in now months away from developing Japan's first single-electron tunneling transistor capable of operating at room temperature, according to Masakazu Aono, head of the institute's surface and interface laboratory.
Though the technology may not find its way into semiconductors until the end of the decade, the tunneling transistor will make possible a 10 cm2 microprocessor of terabit density, Aono told EE Times.
"It's hard to imagine the world after a couple of generations of technology. Fifty years ago Bell Labs created the first transistor and at that time no one imagined the present scale of integration," Aono said. "This transistor will be three magnitudes of order smaller than the gigabit limit for MOS as predicted by Moore's Law."
The transistor Aono is developing makes a switch circuit consisting of three, 3-nanometer-wide polydiacetylene "wires" that act as a source and a drain, each separated from a well containing a 10-atom-diameter cluster of 500 silver atoms that acts as a capacitor. On the other side of the separation lies the gate, which sits 4 nm from the capacitor, with the whole unit resting on a graphite substrate. Quantum conductance dictates that the circuit will work by exploiting different Homo (highest occupied molecular orbital) and Lumo (lowest unoccupied molecular orbital) potentials created between the source drain and the capacitor. When carrying a single electron, current can flow through source and drain, said Aono, but 1 volt placed at the gate adds a second electron to the capacitor, thus raising its potential and closing the circuit.
"We can make an atomic switch in a cluster of silver atoms," he said. "The island is so small we are talking about a one-electron effect circuit."
The circuit relies on a series of process technologies built around STM probe manipulation and redevelopment, and the whole project rests on one key breakthrough — the manufacture of a one-molecule-thick conductor out of an insulator at room temperature. This breakthrough occurred late last year when Aono and a fellow researcher, Yuji Okawa, perfected a technique to create the multiple polydiacetylene channels that can support the flow of current. By applying a tiny voltage probe to a single-molecule-layer of compound monodiacetylene film, they found they could repeatedly produce a domino-effect chain that was one molecule wide but up to 300-nm long — in effect, a conductive "wire."
Tied up in chains
Single-molecule-wide chains are typically 100-nm (about 200 molecules) long, said Aono, but his team continually repeat the process continually produced multiple, parallel chains and found that they were stable enough to carry current "infinitely for practical purposes," Aono said.
To complete the circuit, the group need to create a reliable, repeatable technique for building the silver molecule cluster capacitors, Aono said. The researchers used a single-tungsten-atom-tipped STM probe to punch a 0.3-nm pit in the silicon substrate to create the capacitor well. They are now struggling to perfect the precise carving of a pit to contain the capacitor, said Okawa. "It's not so difficult," he said.
Then, in an application reminiscent of a fountain pen, Okawa is working to use an STM probe soaked in ionic Ag2S, to "drip" the atomic-sized silver blobs into the holes to create the cluster capacitor. "It's like a nano-pen with silver ink," quipped Aono.
By creating a network of wires and pits, Aono is confident the two-man team can create a rudimentary circuit board, recreating the birth of semiconductor integration but on a nanometer scale.
The project faces two key hurdles at this time, Okawa said. "It's relatively easy to simply put the silver dots on the graphite substrate, but if I want to put them in exact patterns, it becomes much more difficult," he said. "At the moment I'm creating random patterns, but it's going to be difficult," he said.
If the pair can control the process, it will be possible to scale up to a 1,000 x 1,000-dot "circuit board" within months, Okawa claimed. But first, the pair must find a way to test the circuit's reliability, which behooves the development of a nanoscale circuit tester. With a radius of 50-nm, the STM tip is too big to act like a minute contact to the switch circuit.
"Even if we tilt the probes, the minimum distance between the single atom apexes is 100-nm. That's far too large," said Aono. To overcome this, Okawa and Aono are developing a carbon nanotube tunnel that sticks out of the conventional STM tip rather like an inch-long human hair glued on the end of a short, sharpened pencil. Aono expects to have this combination tip ready in four to 12 weeks. "Then circuit testing will be possible," he said.
Imperial intervention
Twelve years ago Aono proposed using STMs not only to observe atoms but to construct circuits, and he mentioned his idea to Japan's Emperor Akihito, who had just acceded the throne and was making the rounds through Japan's academia.
"The Tenno [emperor] asked me many questions," Aono said. "After I explained my idea, his reply was, 'This is only one finger. How can you make a tool?'
"I told him, 'We can pick up a grain of rice when we wet the tip of our finger. That force is some kind of tool.' "
About that time, IBM Fellow Donald Eigler used an STM to write the letters "IBM" with xenon atoms on a nickel surface, lending credence to Aono's point. Then in 2000, Eigler used the STM to place atoms in rings, creating 20-nm-wide "quantum corrals," and a path to atomic-scale circuits.
With five patents on his STM-nanoscale etching techniques, Aono is confident he can build his own single-atom switch circuit this spring or summer.
As his laboratory uses up the last of its five-year, 400 million-yen ($3.5 million) grant, the institute's surface and interface laboratory has just received approval for three more years of cash, having been selected as one of 12 recipients out of 50 requests for proposals for Japan's government-backed Japan Science and Technology Corp. program. But with a 10-year commercialization schedule in mind, Aono said he has even better news from abroad. UCLA and Cambridge University are already expressing interest in co-development, and so are two undisclosed major U.S.-based semiconductor makers.
"My research doesn't need so much funding [to succeed]," Aono said. "Call it nano-money." |
