Heat Creates Spins Without Moving Them

A new in magnetic materials shows that heat alone can create spin information—the fundamental property that could power future low-energy computing—without the energy losses that plague current electronic devices. Researchers have demonstrated what they call the "magnon thermal Edelstein effect," where a temperature gradient across a special magnetic material generates spin accumulation directly, bypassing the need for electric currents that cause heating and energy waste. This finding opens possibilities for developing computing technologies that use heat instead of electricity to process information, potentially leading to more efficient devices.

The key finding is that in certain antiferromagnetic materials with specific structural properties, a temperature gradient along one direction can produce an accumulation of magnons—quantized spin waves—with a specific polarization at the interface. Unlike previous s that required generating spin currents, this effect creates spin accumulation directly, similar to how a temperature difference can create pressure in a gas without moving the gas as a whole. The researchers found this occurs in materials with Dzyaloshinskii-Moriya interaction (a type of spin-orbit coupling) and non-uniaxial magnetic anisotropy, which together create a spin-momentum locking that enables the effect.

Ology involved theoretical modeling of an ultrathin antiferromagnetic nanostrip in contact with a heavy metal like platinum. The team used a simplified one-dimensional spin chain model with specific magnetic interactions and solved it using linearized Holstein-Primakoff transformation and Bogoliubov transformation to understand magnon excitations. They then calculated how the spin accumulation converts into detectable electrical signals through the inverse spin Hall effect in the adjacent metal layer. The approach focused on incoherent thermal magnons rather than coherent ones, making the effect more practical for real-world applications.

Analysis shows several important patterns. As shown in Figure 2(a), the injected spin current density depends monotonically on the Dzyaloshinskii-Moriya interaction strength but non-monotonically on the hard-axis anisotropy—increasing initially with anisotropy then decreasing when anisotropy becomes too strong. Figure 2(b) reveals that the effect increases monotonically with temperature and, counterintuitively, is an even function of applied magnetic fields along the Néel vector direction. The researchers estimate that with typical material parameters, a temperature gradient of 1K/μm could generate a detectable voltage of approximately 100 nanovolts at room temperature, well within measurable range.

This matters because it offers a new pathway for spin generation that avoids the Joule heating problem inherent in electronic approaches. In practical terms, it means future computing devices could potentially use thermal gradients instead of electric currents to create and manipulate spin information, leading to significantly lower energy consumption. The effect could be particularly valuable for developing all-magnon transistors and other spin-based devices where minimizing energy loss is critical. Since magnons are charge-neutral quasiparticles, they can transport spin information without moving physical charges, eliminating resistive heating entirely.

Limitations of the research include the simplified assumption of constant magnon relaxation time independent of momentum, spin, and temperature, which may not hold in real materials. The modeling also assumes an idealized one-dimensional system, whereas real materials have three-dimensional structures and more complex interactions. Additionally, the actual output voltage depends on several temperature-dependent parameters in the detection metal that were not fully characterized. The researchers note that their work provides the physical mechanism but leaves detailed material optimization and experimental verification for future studies.