Electron spins in magnetic material are’split’ by a mechanism.

A schematic depiction of virtual electron–positron pairs appearing at random near an electron (at lower left). Credit: RJHall/Wikipedia

By manipulating the correct material at the proper angle, Cornell researchers discovered a mechanism for switching the magnetization in tiny layers of a ferromagnet—a technology that could eventually lead to the development of more energy-efficient magnetic memory devices.

The researchers published an article titled “Tilted Spin Current Generated by the Collinear Antiferromagnet Ruthenium Dioxide” in Nature Electronics on May 5. Arnab Bose, a postdoctoral researcher, and doctorate students Nathaniel Schreiber and Rakshit Jain are co-lead authors on the paper.

For decades, physicists have attempted to reorient electron spins in magnetic materials using magnetic fields. However, researchers such as Dan Ralph, the F.R. Newman Professor of Physics at the College of Arts and Sciences and the paper’s principal author, have looked instead at electron spin currents, which exist when electrons’ spins are generally oriented in one direction.

When these spin currents come into contact with a tiny magnetic layer, they transmit their angular momentum and generate enough torque to completely reverse the magnetization. (Changing the magnetic orientation is how information is stored in magnetic memory devices.)

Ralph’s group has been investigating techniques to manipulate the spin direction of spin currents generated with antiferromagnetic materials. Antiferromagnets have no net magnetism because every other electron spin points in the opposite way.

“Essentially, antiferromagnetic order can reduce the symmetries of the samples sufficiently to allow for the existence of unconventional spin current orientations,” Ralph explained. “The mechanism of antiferromagnets appears to provide a method for obtaining fairly high spin currents as well.”

The researchers had been working with the antiferromagnet ruthenium dioxide and observing how its spin currents shifted the magnetization in a thin layer of Permalloy, a nickel-iron magnetic alloy that is a soft ferromagnet. They examined the torque’s effects at a number of magnetic field angles in order to map out its many components.

“We initially had no idea what we were witnessing. It was entirely unlike anything we had seen previously, and it took us a long time to figure out what it was “As Jain stated. “Also, these materials are difficult to integrate into memory devices, and our hope is to discover alternative materials that exhibit similar behavior but are more easily integrated.”

Finally, the researchers uncovered a process known as “momentum-dependent spin splitting” that is unique to ruthenium oxide and other antiferromagnets of the same class.

“For a long time, it was assumed that spin up and spin down electrons behaved identically in antiferromagnets. This class of elements is very novel “Ralph remarked. “The spin up and spin down electrical states are fundamentally distinct. Once you apply electric fields, you immediately get the ability to generate large spin currents due to the way spin up and spin down electrons react differently. Thus, by accelerating one of them faster than the other, you can generate a strong spin current.”

This process had previously been proposed but not documented. When the antiferromagnet’s crystal structure is properly orientated within devices, the process allows the spin current to be tilted at an angle that enables more efficient magnetic switching than other spin-orbit interactions.

Ralph’s team is now looking for ways to fabricate antiferromagnets in which they can control the domain structure—that is, the regions in which the electrons’ magnetic moments align in the same direction—and investigate each domain independently, which is difficult because domains are typically intermingled.

Eventually, the researchers’ technique could result in advancements in magnetic random-access memory-based technology.

“The goal is to create extremely efficient, extremely dense, and nonvolatile magnetic memory systems that outperform current silicon memory devices,” Ralph explained. “This would enable a significant shift in the way computers manage memory, as you’d have something with virtually endless endurance, high density, and high speed, and the information would persist even if the power was switched off. Nowadays, no memory performs that function.”

Further information: Daniel Ralph, Tilted spin current generated by the collinear antiferromagnet ruthenium dioxide, Nature Electronics (2022). DOI: 10.1038/s41928-022-00744-8. www.nature.com/articles/s41928-022-00744-8

Journal information: Nature Electronics

Source: Cornell University