Over the last decade, spintronic technology has found its way into consumer electronics, used in devices like cell phones, hard drives, and RAM. Now, researchers at the University of Delaware have opened a door to making spintronics even more ubiquitous. They've demonstrated how to control electron spin in silicon, today's semiconductor of choice (Fig. 1).
"Silicon forms the basis of the microelectronics industry," said Ian Appelbaum, professor of electrical engineering and lead author of the study (Fig. 2). "We can make complex devices like CPUs if we can harness… extra information encoded in spin and apply it in silicon."
Spintronics involves harnessing an electron's spin—a quantum property with either an "up" or "down" orientation—to encode and process data. While electronics uses electrical fields to push electrons (read: information) in a current, spintronics uses magnetic fields to move electrons along.
To work, spintronics requires all electron spins to be oriented in the same direction. To achieve this in silicon, Appelbaum's team injected electrons from aluminum—which, like silicon, is non-magnetic and therefore has the same amount of spin-down electrons as spin-up electrons—through a ferromagnetic alloy of cobalt and iron. Electrons in the ferromagnet have only a spin-down orientation.
When Appelbaum applied voltage to the device, spin-down electrons remained with the ferromagnet while spin-up electrons passed through the silicon and precessed (or gyrated) along the same spin orientation. "At this point, we're just demonstrating potential," Appelbaum said, explaining that previously, nobody knew how to manipulate electron spin in silicon. "We've finally shown one way of doing it."
Researchers have been able to control spin in other semiconductors like gallium arsenide, which is used in cell phones. But those methodologies, Appelbaum said, would not work in silicon. And like other spintronic devices, silicon spintronics could reduce power consumption, reduce circuit scaling limitations, and slash waste heat.
"One limitation to scaling electronic circuits is that you can't get the heat out fast enough," Appelbaum said. "If you can figure out how to isolate spin current, you won't dissipate heat."
Currently, spin current is coupled with charge current. Other researchers at the University of Delaware will be investigating ways to capture a "pure" spin current at the university's up-and-coming Center for Spintronics and Biodetection. The U.S. Department of Energy recently awarded the university a $1.9 million grant for the establishment of the center.
John Xiao, professor of physics and astronomy, will lead the center, which will focus on spintronics in metals, not semiconductors. It is easier to generate a spin current in metals, Xiao said, which is ideal for the center's aim of generating a type of spin "battery."
"We are more concentrated on developing a source… that does not source an electric current, but sources a spin current," Xiao said. This research, he added, can eventually be applied to semiconductor spintronics.
While the "most visible impact of spintronics in consumer electronics today is entirely metals-based," Appelbaum said, semiconductor spintronics are superior because spin lifetime in semiconductors is orders of magnitude longer than in metals.
"That's not to overstate the impact \[of silicon spintronics\] on consumer electronics," he said. "But I think now it's clear that the potential is there and there's a lot of work to be done. We've opened the door, but we're taking just one step in the direction."
Ultimately, Appelbaum hopes to see his research applied to quantum computing. "It may take my lifetime to do it," he says, "but that's where it's going."