Skip to content
PaperFren

How does an electric current flip a thick antiferromagnet?

Open paper intelligence

Current flips thick Mn3Sn layers not by spin torque alone but by heating them to their ordering temperature and letting a spin current set the interface as they cool, which then templates the whole layer.

Source

Setting of the magnetic structure of chiral kagome antiferromagnets by a seeded spin-orbit torque

Pal B, Hazra BK, Göbel B, et al. · Science advances · 2022

doi.org/10.1126/sciadv.abo5930Read the full paper ↗33 citationscc by

Study at a glance

Design
Other — Current-pulse switching experiments on sputtered Mn3Sn/W Hall-bar devices read out by anomalous Hall resistance, varying pulse length, fall time, temperature, field and film thickness, plus atomistic LLG simulations
N
No single N; devices with five Mn3Sn thicknesses (30 to 100 nm) under many pulse and field conditions
Population
Thin films of the chiral kagome antiferromagnet Mn3Sn with a tungsten spin-current layer
Outcome
Anomalous Hall resistance change, critical switching current density and switching ratio

Structured fields used in claim comparison tables when every cited study has a complete layer.

What they did

The researchers grew Mn3Sn films 30 to 100 nm thick with a tungsten overlayer, patterned them into Hall bars, and switched their magnetic state with current pulses while reading the anomalous Hall resistance. They used a protocol that reads the device during the pulse to track its temperature, varied pulse length, fall time, set temperature, magnetic field and film thickness, and ran devices without the tungsten layer. Atomistic simulations with a spin-orbit-torque term were used for comparison.

What they found

Switching always happened when the device reached about 435 K, close to the Néel temperature, regardless of pulse length, starting temperature or field. The critical current fell with thicker films while the critical power stayed constant, and pure spin-orbit-torque simulations needed far larger currents than measured, both inconsistent with a torque-only mechanism. Pulses with an abrupt 750 ps fall time gave much weaker switching than pulses with a slow fall, and adding a bias current let 10 ns pulses switch as effectively as 100 ms pulses.

The limits

What it doesn't show

The seeded mechanism is an interpretation that fits the data; the seed layer at the interface is not imaged directly. Device temperature is inferred from resistance and finite-element modelling rather than measured independently. Switching only reaches about 40% of the field-induced Hall signal, so the whole layer is not fully reversed. Heating to over 400 K costs power, and the authors only suggest, rather than demonstrate, materials that would reduce it.

Key terms

Antiferromagnet
A material whose atomic magnetic moments cancel out overall, so it has almost no net magnetization or stray field.
Spin-orbit torque
A torque on magnetic moments from a spin current generated when charge current flows in a heavy metal such as tungsten.
Néel temperature
The temperature above which an antiferromagnet loses its magnetic order.
Anomalous Hall effect
A sideways voltage in a magnetic conductor that depends on its magnetic state, used here to read which state the device is in.
Exchange bias
A shift in a ferromagnet's switching caused by an adjacent antiferromagnet whose interface order was set during cooling.

Flashcards

1 / 10

0 of 10 answers reviewed

Research intelligence for this paper

See its role on concept claims, tensions it is part of, placement history, and related discoveries.

Open paper intelligence

Quiz yourself

1 / 5

At about what temperature did switching consistently occur?

Common questions

Why can't ordinary spin-orbit torque explain the switching?

The spin current only penetrates about a nanometre, yet layers up to 100 nm switch, and the needed current falls rather than rises with thickness.

Why does the fall time of the pulse matter?

The spin current must still be flowing while the film cools back through its ordering temperature; if it stops too early, the seed is not set.

How did they reach nanosecond switching?

They combined short pulses with a steady bias current that supplied the spin current during cooling.

More on Magnetism and spintronics