Antiferromagnet Memory Could Survive Magnetic Fields

Antiferromagnet Memory could change how computers store information, because it keeps data safe in places where ordinary magnetic storage can fail. Researchers at Northwestern University have shown that a new kind of magnetic device can both write and read information at room temperature, using materials whose internal magnetism cancels itself out. According to Northwestern Engineering, the work offers a path toward denser, tougher memory for artificial intelligence, high-speed communications, and everyday gadgets that sit close to magnets.

Most computer memory built on magnetism today relies on ferromagnets, familiar materials in which electron spins line up in the same direction. That alignment creates a net magnetization, with north and south poles, and it also creates a weakness: a strong enough field nearby can disturb what is stored. Antiferromagnets behave differently. Neighbouring spins point in opposite directions and cancel one another, so the material has no net magnetization even though its microscopic order remains highly organised. A memory built from such a material would not be wiped by a passing magnet in the same way.

The catch, for decades, was control. Engineers understood how to write and read information in ferromagnets, but the cancelling spins in antiferromagnets seemed to rule out the same tricks. Conventional wisdom held that the standard effects behind modern magnetic memory should not work in these materials, because they appeared to depend on spins having a net alignment. The new research challenges that view directly, with a device that does both jobs electrically and at ordinary temperatures.

How the new device writes and reads

The team, led by Pedram Khalili, built a nanoscale device from a pair of antiferromagnetic layers separated by a thin insulating barrier. The researchers call the structure an all-antiferromagnetic tunnel junction. One capability is a large tunnelling magnetoresistance effect, which gives an electrical way to tell magnetic states apart and therefore to read stored information. The other capability is the newly observed effect that lets a current passing through one antiferromagnet exert torque on another and switch its state, which provides a way to write information.

Khalili said the combination matters because either capability alone would leave a device incomplete. A reader without a writer cannot store new data, and a writer without a reader cannot retrieve it. In the new device, both operations happen electrically, without needing an external magnetic field to flip states. That simplicity is part of why the result has drawn attention beyond physics laboratories: it moves antiferromagnetic memory from a theoretical curiosity toward a practical building block.

Explaining the switching required new physics. Rather than relying on net spin alignment, the effect is governed by the magnetic octupole, a quantity that describes the more complex arrangement of spins in a noncollinear antiferromagnet. Khalili said the torque had not even been theoretically predicted before the team observed it, and the paper supplies a fresh theoretical framework for why it happens. According to the university, that framework gives other researchers a foundation for designing devices that exploit the ultrafast dynamics of antiferromagnets. In short, Antiferromagnet Memory in this design is not a single trick but a pair of complementary operations working together.

Why it matters for readers and devices

The most immediate promise is tougher, denser memory. Khalili said the effect could lead to higher-density, more reliable memories, which are very important for applications like artificial intelligence. Systems that train and run large models need memory that is fast, compact, and dependable, and a memory that shrugs off stray fields would be easier to place inside crowded hardware. The same resilience could help wearables: the lead researcher pointed to smartwatches, where magnetic wristbands, beads, or connectors create nearby fields that could threaten conventional magnetic memory.

Speed is the second promise, though it remains a future goal rather than a demonstrated product. Ferromagnetic dynamics typically occur at gigahertz frequencies, while antiferromagnetic dynamics can extend into the terahertz range, a far faster regime. That gap points toward high-frequency communications, faster computing, and new kinds of detectors and oscillators. The team is clear that the new work does not yet demonstrate terahertz operation, and turning a laboratory junction into a commercial chip will require further materials testing and integration with conventional semiconductor electronics.

Comparison with the past helps put the claim in perspective. Earlier discoveries in magnetic storage, including giant magnetoresistance and tunnelling magnetoresistance, opened a field that shaped research for decades. Khalili described the new finding as foundational in the same sense, because it answers a long-standing question: whether similar effects could be realised with antiferromagnets at all. The answer, in this work, is yes, and the team plans to test other noncollinear materials to see whether they produce stronger effects. For readers, the takeaway is patience with excitement: a laboratory breakthrough that could one day make phones, watches, and servers faster and sturdier, but not a product you can buy this year. Related science coverage includes a look at dusty ice found inside a visitor from another star, how squid rewrite proteins without touching their DNA, and why new machine proofs still need human checking.

What happens next will decide how far the idea travels. The researchers want to learn whether their junctions can directly access terahertz dynamics, and whether combinations with ordinary electronics can deliver devices that resist magnetic fields, operate faster, or increase computing throughput. The work builds on nearly a decade of investigation in the lead researcher's laboratory into electrically controlling antiferromagnets, including earlier studies of different manipulation mechanisms and progressively better ways to detect magnetic states. If that line of work continues, Antiferromagnet Memory may move from a striking laboratory result to a quiet upgrade inside the machines people carry every day, noticed only because those machines fail less often and keep up more easily. The full account from the university is available at Northwestern Engineering.