Making the switch, this time with an insulator

The growing field of spin electronics - spintronics - tells us that electrons spin like a top, carry angular momentum, and can be controlled as units of power, free of conventional electric current. Nonvolatile magnetic memory based on the "spin torques" of these spinning electrons has been recently commercialized as STT-MRAM (spin transfer torque-magnetic random access memory).

Colorado State University physicists, joining the fundamental pursuit of using electron spins to store and manipulate information, have demonstrated a new approach to doing so, which could prove useful in the application of low-power computer memory. Publishing Sept. 1 in Nature Communications, they've demonstrated a new way to switch - or direction of magnetization - of electrons in a thin film of a barium ferrite, which is a magnetic insulator. Until this point, scientists have only demonstrated this switching behavior - the key to writing information as memory - in metal thin films.

The work was led by Mingzhong Wu, professor in the Department of Physics, with first author Peng Li, a former postdoctoral researcher now at Seagate, and second author Tao Liu, a current postdoc at CSU. The work was performed in collaboration with researchers at University of Alabama, Argonne National Laboratory, University of Notre Dame, and University of Wyoming. Other CSU authors include faculty members Stuart Field and Mario Marconi, and graduate students Houchen Chang and Daniel Richardson.

Switching magnetic moments of electrons in an insulator instead of a metal could prove to be a major breakthrough in spintronics, by allowing a spin current-based memory storage device to be simpler, and also maintain more efficiency per electron. A property known as perpendicular magnetic anisotropy (PMA), key for information storage, in this case originates from the intrinsic magneto-crystalline anisotropy of the insulator, rather than interfacial anisotropy in other cases, Wu said.

"Higher efficiency and lower power than the standard are always the goal in memory applications," Wu said.

Beyond the application for computer , which captivates most spintronics researchers today, the CSU researchers' device does something bigger: It demonstrates the possibility of a new class of materials for spintronics. "What's exciting about this is that it's an enabling technology for exploring an entirely different class of configurations, some of which are theorized to be useful," said Jake Roberts, professor and chair of the Department of Physics.

In the CSU researchers' device, the spin current does the job of assisting magnetic switching. Next, they will attempt to further refine their device for more efficient switching, including using a topological insulator or the photo-spin-voltaic effect to produce spin currents. The photo--voltaic effect was discovered by Wu and colleagues, and reported in Nature Physics.

Explore further

Device architecture that can tune a material's magnetic properties could reduce the power consumption of memories

More information: Peng Li et al, Spin–orbit torque-assisted switching in magnetic insulator thin films with perpendicular magnetic anisotropy, Nature Communications (2016). DOI: 10.1038/ncomms12688
Citation: Making the switch, this time with an insulator (2016, September 1) retrieved 20 August 2019 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Feedback to editors

User comments

Sep 02, 2016
OK, I get it, use QM; however, if the electron is a point charge, i.e. no material center of the spherical field, which is updated strictly based upon position at t, at the speed-of light, relative to the center. That is the entire field due to the charge. A magnetic field requires a closed loop and complete current flow around the loop. Take a look at the incomplete field in your mind.

So this science is about the alignment of magnetic moments based upon orbital paths Not spin. Spin exist only in QM. Demonstrate spin with a point charge, else produce a specimen with surface, may be globular with a generator inside, and you have identified a particle. But no particle exist.

The point charge only responds to a single virtual point, i.e. view point of only the charge which describes it's motion, a distance r away with polarity +/-1, i.e. a superimposed point representing everything at that instant of time. The point only travels through space, there is no spin.

Sep 02, 2016
Maxwell and frequency, field strength, controls, proper models, aligned based upon E(r), r is 4D, using Maxell is all you got anyway. Except for a bag of tricks, and funny names. Please forget GR. Think, the wavelet leaving a point travels through space practically unchanged, you know what it was when emitted, you know the wavelength, frequency. You measure the time it takes for one wavelet to pass so that you can adjust the equipment to receive the message correctly. Your calculation adjust the relative speed of the wavelet as Emitted_Wavelength divided by total time to pass, i.e. c lambda_emitted/lambda_observed. The speed of light is not a constant. lambda_nu is constant.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more