Physicists Put a New Spin on Reading Magnetic States

Carnegie Mellon University scientists have uncovered a fascinating physical phenomenon at the nanoscale that brings next-generation spintronic devices one step closer to reality. The discovery offers a way to read out the information stored in the magnetic layer in a planar device geometry — the holy grail of spintronics technologies.

The groundbreaking work, which was published in Nature Materials in March and has a related patent pending, has the potential to enable the creation of highly energy efficient, ultra-compact magnetic-based computing and data storage technologies.

While magnets have long been used for computer memory, our data-driven world constantly demands better-performance, higher-speed, and lower power. To meet these demands, physicists have turned to spintronics, a field that uses the electron’s spin in conjunction with the electron’s charge to store information in magnetic devices.

In the latest work, scientists working in Carnegie Mellon’s Department of Physics' Lab for Investigating Quantum Materials, Interfaces and Devices (LIQUID) Group, have discovered a new form of magnetoresistance by harnessing spins in a quantum material.

“We’ve utilized a quantum material’s crystal structure symmetry to generate spin current with out-of-plane polarization. And we can use this spin current to read the magnetic state of a nearby magnetic layer in a heterostructure,” said Simranjeet Singh, an associate professor of physics. “At the end of the day in spintronics, everything is about how to read and control the magnetic states.”

In previous work, the LIQUID group showed that running an electrical current through their novel two-dimensional quantum material generated an out-of-plane spin current, which in turn was used to control the magnetization state of a neighboring magnetic material without the need to apply an external magnetic field. This essentially ‘writes’ information onto the magnetic layer by assigning the magnetization an up state (a one) or a down state (a zero).

Now, the researchers have used that same out-of-plane spin current to ‘read’ the information stored on the magnetic layer.

Spin current generated in conventional materials is oriented in-plane. Such current cannot distinguish between up and down magnetic states of perpendicularly polarized magnets, which are highly desired for magnetic memory and spin-logic devices. The newly discovered out-of-plane spin current can — making it possible to electrically read the up and down magnetic states of the perpendicularly polarized magnetic layer.

This breakthrough enables the reading of the magnetic layer through longitudinal resistance measurements. That means a magnetic state can be read using only two electrodes. And that's a big deal, says Singh, because the devices that people are currently exploring for spintronics technology, commonly known as spin-orbit torque switching devices, have three terminals — one to generate the spin current to write the magnetic state and one to read the magnetic state via magnetoresistance relative to a third, common terminal.

“Before our work, there was no known physical mechanism which can allow you to have this spin-orbit torque device based on perpendicularly polarized magnets where you only need two terminals for its complete operation,” Singh said.

The more electrodes you have, the larger the footprint of the device, so exploiting this phenomenon — what the team calls unconventional unidirectional magnetoresistance — can enable the creation of ultra-small next-generation technologies.

“To build an industry competitive magnetic memory device, a small footprint hardware node is desired to implement a dense network of storage elements,” Singh said.

I-Hsuan Kao, a Carnegie Mellon alumnus, current postdoctoral fellow in the LIQUID group and the paper’s first author, developed the two-layer structure used in the current work. The structure, only a few microns in length/width, is made of atomically thin layers of tungsten ditelluride (WTe₂), a topological semimetal that generates the spin current, and of Cr₂Ge₂Te₆ (CGT), a perpendicularly polarized magnet.

To read data written on the CGT layer, the team sent an electrical current through the WTe₂ layer, generating an out-of-plane spin current. When that spin current interacts with the magnet that is in an ‘up’ state, it will register a low electrical resistance. If the magnet is in a ‘down’ state, it will register a high electrical resistance. These so-called unidirectional magnetoresistance measurements, or resistance that depends on the magnetic orientation, are what the LIQUID team used to read the up and down magnetic states of the CGT layer.

The signals themselves are very small, but the atomically clean and precise design of the two-layered structure allowed their detection. The spin currents generated in the WTe₂ layer interact with the CGT layer at the interface between the two, so “you need atomically clean interfaces in order to see some of these phenomena,” said Jyoti Katoch, an associate professor of physics. “I-Hsuan figured out how to make this beautiful interface. There’s nothing in between.”

Kao also showed that the signals can be tuned using an electric field. This tunability can boost the magnetoresistance signals, making them easier to read.

The LIQUID team is working on optimizing their current heterostructure and is exploring other material systems that may also exhibit the unconventional unidirectional magnetoresistance phenomena. The goal is for materials scientists and engineers to use their foundational physics discoveries in these quantum materials to create next-generation technologies.

In addition to Kao, Singh and Katoch, co-authors on the paper "Unconventional unidirectional magnetoresistance in heterostructures of a topological semimetal and a ferromagnet" include Carnegie Mellon undergraduate student Sean Yuan; Junyu Tang and Ran Cheng from the University of California, Riverside; Gabriel Calderon Ortiz, Menglin Zhu and Jinwoo Hwang from The Ohio State University; Rahul Rao from the Air Force Research Laboratory at Wright-Patterson Air Force Base; Jiahan Li and James H. Edgar from Kansas State University; Jiaqiang Yan and David G. Mandrus from Oak Ridge National Laboratory and The University of Tennessee, Knoxville; and Kenji Watanabe and Takashi Taniguchi from the National Institute for Materials Science, Tsukuba, Japan.

Funding for this research was provided in part by the National Science Foundation (ECCS-2208057); the US Office of Naval Research (N00014-23-1-2751); the Center for Emergent Materials at The Ohio State University, an NSF MRSEC (DMR-2011876); the US Department of Energy, Office of Science, Office of Basic Sciences (DE-SC0020323). Singh and Katoch both received Faculty Early Career Awards from the National Science Foundation, which partially supported the current research.