Controlling magnetic skyrmion motion without an electric current

A magnetic field gradient is an efficient way of manipulating skyrmions

Researchers from the University of Oxford have used resonant elastic X-ray scattering (REXS) at Diamond for the first demonstration of using a field gradient to control the motion of magnetic skyrmions. Their work, published in Nature Communications, overcomes the limitations inherent in using electrical currents to control skyrmion motion and opens up new avenues of research in non-conductive (as well as conductive) skyrmion-carrying materials. Ultimately, their results bring us a step closer to high-density, low-power data storage devices based on racetrack memory. 


Figure 1: A 3D printed model of a Bloch-type skyrmion.

The future of magnetic memory

The search is on for technologies that will bring us the next-generation computers on which societies of the future will depend. We need high-density, low-power devices for data storage applications, and magnetic random access memory (RAM) is one promising candidate. One experimental technology, racetrack memory, stores data as a series of magnetic domains. Particle-like magnetic skyrmions can act as the information carriers, and research has focused on using electrical currents to ‘push’ the domains past a read/write element, via spin transfer torque (STT).

Driving and controlling the skyrmion motion is the single most important task in this style of memory device, but STT has its limitations. Although an ultra-low current density is enough to control the movement of the skyrmions, it only allows for low-speed motion. In order to drive skyrmions at useful speeds, a higher current density is needed. Unfortunately, that electric current will eventually dissipate as heat, and in most skyrmion-carrying materials temperature stability and homogeneity are sensitive parameters. And, of course, STT-based movement is only possible in conductive materials.

Researchers from the University of Oxford have now demonstrated a novel method for controlling skyrmion motion using a non-contacting field gradient, which removes the need for a driving current and is applicable to all skyrmion materials.

Figure 2: Dr Shilei Zhang explaining helical magnetic order.

Making magnetic skyrmions move

The team conducted their experiments in Cu2OSeO3, a well-established skyrmion system. The material is an insulator, so its skyrmions cannot be manipulated using STT. It also has a relatively large crystal lattice constant, which allows for the observation of skyrmions using resonant elastic X-ray scattering (REXS), a technique ideally suited to studying slow dynamics in situ. REXS is a soft X-ray technique that is sensitive to magnetic structure. It uses polarised X-rays, and is only available at synchrotrons. These experiments were carried out on I10, the Beamline for Advanced Dichroism Experiments (BLADE) using a rotating field gradient, but the results are applicable to the linear field gradients that would be used in memory devices.

The individual skyrmions In Cu2OSeO3 appear as spin swirls, and assemble into closed-packed, hexagonally ordered magnetic lattices. These magnetic structures are highly mobile. In the radial field gradient, the skyrmions were found to rotate collectively, following a given velocity-radius relationship. The rotating ensemble disintegrates into a shell-like structure of discrete circular racetracks. When the field direction is reversed, the rotation reverses as well. These results show that field gradients offer effective fine control of skyrmion motion, with lower energy requirements than STT.

Next steps
We are still years away from commercial devices using racetrack memory. Scientists are still working on ways to build effective read/write heads, and new materials will be needed. The materials currently being used require low temperatures, but the search is on for higher temperature replacements. The next steps for the Oxford University team are to fabricate a racetrack memory device, using a field gradient control system. However, they’ll also be spending some time investigating some interesting physical effects they found during the course of their experiments.

Lead author Dr Shilei Zhang explains: “In our experiments the skyrmions followed spiral trajectories, rotating faster at the centre and slowing towards the edge of the system. The skyrmions also repel each other, and in the model this leads to a quasi-static rotation scenario. However, we observed the rotation for 6 days without it reaching an equilibrium state. It’s an interesting phenomenon to investigate, and it may lead to a design for circular racetrack memory.”

Experimental REXS time-lapse video of skyrmion lattice rotation

Video 1: Real-time snapshots of the REXS pattern showing the rotation of the skyrmion lattice domains in a field gradient. Each ‘snapshot’ is a CCD camera image exposed for 2 ms, taken every 5 s. The 
~16-min-long movie shows the skyrmion rotation in a field gradient ~25 min after the beginning of the rotation.

To find out more about the I10 beamline, or to discuss potential applications, please contact Principal Beamline Scientist Dr Paul Steadman:

Related publication:

Zhang S et al. Manipulation of Skyrmion Motion by Magnetic Field Gradient. Nature Communications, doi:10.1038/s41467-018-04563-4 (2018).