

Researchers have identified an unintended magnetic layer that can distort measurements of Mn₃Sn, a promising material for next-generation electronics. The discovery, made by scientists from Martin Luther University Halle-Wittenberg (MLU) and the Max Planck Institute for Microstructure Physics, highlights how processes used to prepare samples can alter the material's properties. The findings were published in the journal Advanced Materials.
Mn₃Sn, a compound of manganese and tin, belongs to a class of chiral antiferromagnets in which the magnetic moments of neighbouring atoms are arranged at 120-degree angles. Although their magnetic moments cancel each other out, antiferromagnets are resistant to electromagnetic interference and can switch states up to 1,000 times faster than conventional magnets. These characteristics make them candidates for faster and more energy-efficient electronic components.
To investigate the material, the team produced nanometre-thick Mn₃Sn layers and combined them with platinum or a nickel-iron alloy. They used ultrashort laser pulses to excite the samples and measured the resulting electrical voltages within picoseconds, or trillionths of a second.
The researchers found that heating during sample preparation caused atoms at the interface between Mn₃Sn and platinum to form a new ferromagnetic compound. The magnetic signals generated by this interface layer were more than 100 times stronger than those of the antiferromagnet, completely masking its properties.
"Before that, we had essentially only been measuring a layer that was an unintended result of the sample preparation process," said Paul Marschall, the study's first author. The team resolved the problem by allowing the Mn₃Sn layer to cure fully before depositing platinum, preventing the formation of the unwanted interfacial compound.
With properly prepared samples, the researchers measured Mn₃Sn's actual performance. They found that spin currents flowing from the nickel-iron layer into Mn₃Sn were converted into electrical voltage at an efficiency of around 10% of that typically achieved by platinum.
Professor Georg Wolterdorf from the Institute of Physics at MLU said: “Interfaces are not just a peripheral aspect of spintronics, they are often the actual component. If we want to harness chiral antiferromagnets for technological applications, we must first understand and be able to control what happens within these few atomic layers.”