Ohio State researchers have discovered an effect in a magnetic semiconductor that might someday lead to improved data processing and more energy-efficient computers, according to a recent study published in Nature Materials.

OSU researchers showed that an electron’s spin can be harnessed to create a small amount of electricity by making one side of the semiconductor warmer than the other.

This phenomenon could lead to improved data storage in computers, said Joseph Heremans, a professor of mechanical engineering and physics.

Computer data is stored as the presence or absence of an electron charge. The problem is that heat generated in the process limits how fast a computer can process data, he said.

A technology under development uses the magnetic spin of an electron to store computer data, the ones and zeros, as “spin up” or “spin down.” These spin-based electronic, or spintronic, devices generate little heat, which will allow computers to process at faster speeds, Heremans said.

“We don’t have (semiconductor) spintronic devices yet, but it’s something worth pursuing,” said Roberto Myers, an assistant professor of materials science engineering and physics.

Although Heremans and Myers were the first to observe the effect in a semiconductor, they did not discover the phenomenon.

A group of Japanese researchers did in 2008.

Eiji Saitoh from Keio University in Yokohama, Japan, and his research group discovered the direction of an electron’s magnetic spin could be flipped when one end of a magnetic metal rod was warmer than the other. This change in spin was detected as a small electrical voltage. They called this the “spin-Seebeck effect.”

The OSU researchers, in collaboration with scientists from the University of California Santa Barbara, were the first to confirm the Japanese group’s findings, to show the effect in a semiconductor, and to offer a glimpse into how this effect might work.

Rather than a metal rod, the OSU researchers used a thin film of magnetic material called gallium manganese arsenide. The manganese provides the magnetism in the sample, but only at certain temperatures, Myers said.

“We could just shift the temperature a little bit and see the effects turn on and off,” he said.

They also demonstrated that the spin-Seebeck effect was not affected when the film was cut in half. This showed the effect did not come from a flow of electrons through the material as some scientists had thought. Rather, it is the temperature difference across the film that creates the effect, Heremans said.

Christopher Jaworski, a fourth-year graduate student in mechanical engineering who works in Heremans’ laboratory, spent many nights in the lab taking precise temperature measurements during the past year.

The measurements were small, in the nanovolt range — where nine billion nanovolts make up a nine volt battery.

“A lot of the measurements we would run at night to make sure no one was around because people walking around could interfere with … the actual measurement,” he said.

Because each measurement was made several times, a single sample could take 20 days to complete, Jaworski said.

The researchers faced a few challenges.

The film was hard to make. The manganese did not easily dissolve into the material. And because any impurities would interfere with film behavior, it had to be made under extremely pure conditions, Myers said.

Another challenge was the voltage signals were extremely small, and Jaworski had to develop ways to detect it, Heremans said.

The gallium manganese arsenide films are magnetic only at temperatures below freezing. This limits the practical applications where this material could be used, Myers said.

“It’s precisely because it’s a bad material for devices — that it works at low temperatures — that makes it interesting scientifically,” he said.

Although there are more questions than answers at this point, Heremans said, the emerging field of physics might lead to smaller, more efficient computers.