 |
A
simplified animation that illustrates how magnetic switching occurs in
ferromagnetic and antiferromagnetic materials. Credit: Scott Schrage |
University Communication |
Most scientists would recoil at the thought of being called a "spin doctor." The lab coat, on the other hand, fits Evgeny Tsymbal, Ding-Fu Shao, and their colleagues.
Physicists
at the University of Nebraska–Lincoln have pushed the boundaries of
spintronics, a next-generation type of data storage and processing that is
poised to complement the digital electronics that have dominated the high-tech
world for decades.
However,
nanoscale hurdles loom ahead of that future, their small ness belying their
difficulty. The physicists may be on their approach to overcoming an extremely
difficult problem: finding order among disorder and data amid apparent
disarray, thanks to a $20 million National Science Foundation award. Beyond
that stumbling block, there are two prizes: density and speed, which, in
retrospect, might make today's technologies appear gluttonous and sloth-like.
Digit spinners
By
measuring the charge of electrons passing through their circuits, electronics
read and communicate binary—1s and 0s. Spintronics distinguishes itself by
determining the spin of an electron, which is a magnetism-related feature that
essentially points up or down. Electronics-only devices can store and process
significantly more data, at much faster speeds, and with much less power than their
binary-fluent equivalents.
To date,
ferromagnets, the type with a permanent magnetic field best known for pinning
photographs to refrigerators, have been used in most electronic and spintronic
memory. Every atom's spin in ferromagnets points in the same direction, which
can be changed by introducing an external magnetic field.
These
characteristics make them attractive in tunnel junctions, which consist of two
ferromagnets sandwiched around an insulating barrier, with electrons
"tunnelling" across the barrier to flow between them. When an
electron's spin aligns with the ferromagnet's spin orientation, the electron
finds low resistance and has a higher chance of tunnelling through. When those
spins don't line up, the chances are stacked against you, limiting the overall
flow of electricity. The magnetoresistance effect, or the difference between
those two states, can be viewed as a 1 vs. 0.
Because
ferromagnets are useful, their cousins, antiferromagnets, are even more so.
Antiferromagnets have alternating columns of atoms with spins pointing in
opposite directions, resulting in a net magnetic field of almost zero. Because
there is no magnetic field, there is no risk of a tunnel junction interfering
with its neighbor's magnetic state, allowing engineers to pack more
data-storage elements into a device without fear of data corruption.
Antiferromagnets
will be used again if next-generation gadgets require speed, according to
Tsymbal. In nanoseconds, the spins of a ferromagnet can be swapped. That looks
fast until you realise that semiconductors can switch at a pace of picoseconds
(a picosecond is equivalent to 31,710 years), which is nearly 1,000 times
faster than a ferromagnet. Antiferromagnets, on the other hand, can keep up,
preparing them for a prominent role in far speedier gadgets.
One minor point: encoding or decoding data in antiferromagnets is akin to trying to write with a dried-up pen or deciphering a toddler's scribbles.
Tsymbal, a George Holmes University Professor of physics and astronomy, remarked, "The difficulty—and it's a huge difficulty—is how to write and read information."
When it
comes to actually capturing data, the same antiferromagnetic feature that
functions as a benefit in one context—the lack of a net magnetic field that
prevents data corruption—becomes a drawback, according to Tsymbal. In a
ferromagnet, writing a 1 or 0 is as simple as flipping its spin orientation, or
magnetization, with another magnetic field. In an antiferromagnet, this is
impossible.
While
reading the spin state of a ferromagnet is simple, distinguishing between the
spin states of an antiferromagnet—up-down vs. down-up—is more difficult because
neither provides a net magnetism that would cause noticeable variations in
electron flow. These characteristics have combined to stymie efforts to construct
antiferromagnetic tunnel junctions that can be used in real devices.
"This is one of the issues," Tsymbal explained. "However, I believe we have proposed a really good solution to this problem."
Telling up from down
An
antiferromagnetic tunnel junction should, in theory, work similarly to a
ferromagnetic one. An antiferromagnetic variant focuses on adjusting the
so-called Néel vector: the axis along which spins are pointed one way or the
other, rather than switching the overall magnetization of a ferromagnet to
govern the flow of electrons.
However,
only certain antiferromagnets are capable of detecting spin-related changes in
electron flow, which are caused by a mismatch in the Néel vectors at either end
of the tunnel junction. What's the deal with those antiferromagnets? Momentum-specific
pathways through which electrons will mostly flow in one of two directions: up
or down.
Ruthenium
oxide was identified as such an antiferromagnet by Tsymbal, Shao, and
colleagues. Another substance, titanium dioxide, was identified as the barrier
past which electrons can tunnel. The atoms of the two oxides have the identical
crystalline structure, resulting in a flawless match that allows electrons to
keep their momentum—and their momentum-dependent spin—as they pass between
them.
The Husker
team demonstrated that by factoring those momenta into calculations of the
resulting electric current, it is able to discern between the channels and, as
a result, their reactions to varying Néel vectors. According to the
researchers' estimates, the channel-specific magnetoresistance effect is
similar in size to that of ferromagnetic tunnel junctions, making it a
particularly attractive method of writing and reading spintronic data.
Tsymbal is
collaborating with Chang-Beom Eom of the University of Wisconsin-Madison and
other experimentalists who can manufacture and test the antiferromagnetic
tunnel junction, as the theoretician has done in the past. He and his
colleagues at the Nebraska Center for Materials and Nanoscience are also
looking into other materials that have similar but not identical properties to
ruthenium oxide.
"This feature is found in a small number of antiferromagnets, but it does exist," Tsymbal added. "We'll be looking at these materials in the future as well."
0 Comments