The group of physicists, based at the University of New South Wales and Purdue University, said they had laid the groundwork for a futuristic quantum computer that might one day function in a nanoscale world and would be orders of magnitude smaller and quicker than today’s silicon-based machines.
In contrast to conventional computers that are based on transistors with distinct “on” and “off” or “1” and “0” states, quantum computers are built from devices called qubits that exploit the quirky properties of quantum mechanics. Unlike a transistor, a qubit can represent a multiplicity of values simultaneously.
That might make it possible to factor large numbers more quickly than with conventional machines, thereby undermining modern data-scrambling systems that are the basis of electronic commerce and data privacy. Quantum computers might also make it possible to simulate molecular structures with great speed, an advance that holds promise for designing new drugs and other materials.
“Their approach is extremely powerful,” said Andreas Heinrich, a physicist at I.B.M. “This is at least a 10-year effort to make very tiny electrical wires and combine them with the placement of a phosphorus atom exactly where they want them.”
Dr. Heinrich said the research was a significant step toward making a functioning quantum computing system. However, whether quantum computing will ever be harnessed for useful tasks remains uncertain, and the researchers noted that their work demonstrated the fundamental limits that today’s computers would be able to shrink to.
“It shows that Moore’s Law can be scaled toward atomic scales in silicon,” said Gerhard Klimeck, a professor of electrical and computer engineering at Purdue and leader of the project there. Moore’s Law refers to technology improvements by the semiconductor industry that have doubled the number of transistors on a silicon chip roughly every 18 months for the past half-century. That has led to accelerating increases in performance and declining prices. “The technologies for classical computing can survive to the atomic scale,” Dr. Klimeck said.
Demonstrations of single-atom transistors date from 2002, but the researchers from Purdue and New South Wales said they had made advances on two fronts: in the precision with which they placed the Lilliputian switch; and in the use of industry-standard techniques to build the circuitry, making it possible to read and write information from the tiniest conceivable switch.
The results were reported on Sunday in the journal Nature Nanotechnology.
Until now, single-atom transistors have been created on a hit-or-miss basis, the scientists said.
“But this device is perfect,” Michelle Simmons, a group leader and director of the ARC Centre for Quantum Computation and Communication at the University of New South Wales, said in a statement. “This is the first time anyone has shown control of a single atom in a substrate with this level of precise accuracy.”
In the 1950s, the physicist Richard P. Feynman predicted a world where there would be “plenty of room at the bottom,” opening new vistas into engineering disciplines that would use individual atoms as bricks and mortar in fields as diverse as computing and biology.
Since then, computer designers have moved ever closer to the smallest components that are possible to fabricate. Now, with the publication of the New South Wales and Purdue research, the scientists said they had shown the fundamental limits to which the components of silicon-based computers would be able to shrink in the future. Currently, the smallest dimension in state-of-the-art computers made by Intel is 22 nanometers — less than 100 atoms in diameter.
If the semiconductor industry remains on its current pace, it might be possible to reach that limit within two decades, Dr. Klimeck noted.
The scientists placed the single phosphorus atom using a device known as a scanning tunneling microscope. They used it to essentially scrape trenches and a small cavity on a surface of silicon covered with a layer of hydrogen atoms. Phosphine gas was then used to deposit a phosphorus atom at a precise location, which was then encased in further layers of silicon atoms.
While offering astounding precision for research, these microscopes are not currently applicable as manufacturing tools to make chips that contain billions or even trillions of transistors. Moreover, the devices now operate at very low temperatures.
Despite these limits, the semiconductor industry has made great progress in finding ways to build circuits that are far smaller than the wavelength of visible light. And recently, equipment makers have begun making it possible to assemble layers in silicon chips a single atom at a time.
The low temperatures at which the experiment was performed led Intel scientists to express caution about the results. “It’s good science, but it’s complicated,” said Mike Mayberry, an Intel vice president who is the director of the company’s components research group. “By cooling it to very low temperatures, they’ve frozen out a lot of effects that might otherwise be there.”
Shrinking conventional computer circuitry offers radical increases in the speed at which computers can solve problems, lowers the power they require and drastically increases the amount of data they can store.
