A University of Florida-led research team has created one of the first practical magnetic semiconductors made with materials commonly used in high-speed electronics, a notable advance in a hot new field known as "spintronics."
The achievement, described in an article in today's issue of the journal Applied Physics Letters, is a step toward a new breed of computer chips that will couple memory with information processing and photonic capabilities. Such chips, expected to be available within the next few years, will retain data even when the computer is turned off, eliminating the time-consuming process of "booting up" information from hard drive to processor. In the long term, advances in spintronics may usher in vastly more powerful "quantum" computing.
"If we can make semiconductor-based magnetic memories, when you turn your computer on it would be like a TV set -- all the information would be there, and it will not have to reload the operating system," said Stephen Pearton, a UF professor of materials science and engineering and research team member. "Down the road, we may have new electronic and light emitting devices that far exceed the functionality of current silicon chips."
Today's semiconductors work by exploiting the electric charge attached to electrons. But electrons in solids have another fundamental property known as "spin," which makes them act like small magnets. Computer hard discs, which consist of layers of metals such as cobalt and copper, already tap this magnetic property to produce memory storage.
Scientists have long sought to merge this capability with the signal processing capability of microprocessors, a technology called magneto resistive random access memory, or M-RAM.
Pearton said the stumbling block had been that semiconductors displayed magnetism only at impractically low temperatures of hundreds of degrees below zero Fahrenheit. Researchers had been slowly chipping away at that limitation until last year, when a prominent scientist in the field predicted that a certain class of semiconductors would exhibit magnetism at room temperature. His words jump-started new research; at least two groups have proven him correct and achieved room temperature magnetism in recent months, Pearton said.
The rapid progress and excitement in the field has closely tracked similar advances in superconductors of a decade ago, Pearton said. The hitch in the latest research is that the semiconductors achieving magnetism at warm temperatures are made of exotic materials not seen in typical semiconductor production, making them impossible or unlikely candidates for commercialization in computer chips or other applications.
The UF-led team -- five materials engineers and four physicists at UF -- achieved magnetism using gallium phosphide "doped," or infused, with manganese. Gallium phosphide is a commonly used material in the production of many semiconductor devices. In addition, the researchers' method, called molecular beam epitaxy, is standard in the chip and electronics industry, Pearton said.
In the Applied Physics Letters paper, the team reports achieving magnetism at about 100 degrees below zero Fahrenheit. While that represents progress, it's still too cold to be easily commercialized.
But results achieved by the UF team after the paper was submitted and announced at the National Vacuum Conference in Seattle last week show magnetism in the manganese-doped phosphide semiconductor at temperatures that exceed room temperature.
"This makes it less of a leap of faith that practical spin-electronic devices could actually be manufactured," Pearton said.
Besides eliminating long boot-up time, M-RAM chips are likely to require far less power because controlling the current flow by altering electronic spin may require only small voltages, Pearton said. That could significantly extend battery life in hand-held electronic devices such as mobile phones, he said.
In addition, spin-based lasers and other light-emitters could transmit data that is encoded or labeled by the polarization of the light. Many scientists believe researchers one day will learn how to capture and manipulate the spin of individual electrons to create quantum computers thousands of times faster than today's computers.
The other UF researchers involved in the spin research are Mark Overberg and Brent Gila, both postdoctoral associates in materials science and engineering; Jerry Thaler and JiHyun Kim, both doctoral students in engineering; Cammy Abernathy, a professor of materials science and engineering; Nikoleta Theodoropoulou and Kevin McCarthy, both doctoral students in physics; Stephen Arnason, a postdoctoral associate in physics; Art Hebard, a physics professor; and Fan Ren, a professor of chemical engineering. - By Aaron Hoover
[Contact: Stephen Pearton, Aaron Hoover]
06-Nov-2001
