A New Window Into Spin Waves: How Magnon Momentum Microscopy Could Reshape Computing

For decades, engineers have chased faster, cooler alternatives to charge-based electronics. Spin waves—collective oscillations of electron spins that ripple through magnetic materials like waves on a pond—have long been a tempting candidate for carrying and processing information. But actually seeing what these nanoscale waves do, especially when they interact, has remained a stubborn blind spot.

That blind spot has now shrunk dramatically. An international team led by the Max Born Institute in Berlin, working with Helmholtz-Zentrum Berlin, the University of Naples Federico II, and EPFL, has unveiled magnon momentum microscopy (MMM), a soft-X-ray technique that images spin waves directly in two-dimensional momentum space . Their findings, published in Nature Physics on June 5, 2026, not only deliver a sharper picture of magnon behavior but also capture a fundamentally new nonlinear regime that could change how we build wave-based computers .

Why momentum matters

Most existing methods for studying spin waves—such as micro-focused Brillouin light scattering or scanning NV-center magnetometry—are optimized for real-space imaging. While powerful, these techniques struggle to resolve the fastest, shortest-wavelength magnons, which are the ones most relevant for dense, high-speed computing .

MMM flips the script. Instead of looking at where spin waves are in a sample, it captures their wave vectors—essentially a snapshot of all the directions and wavelengths present at once. To do this, the technique uses resonant soft X-rays tuned to the magnetic absorption edges of the material. When the X-rays scatter off propagating spin waves, the waves act as a dynamic diffraction grating, producing sharp +1st and −1st order peaks on a detector. The positions of these peaks directly encode the magnon wavelength, propagation direction, and amplitude across the entire two-dimensional plane in a single measurement .

The payoff is immediate. MMM can access sub-100-nm magnon wavelengths down to a few nanometers, corresponding to THz frequencies, with high photon efficiency and without requiring complex sample nanopatterning. It works across a wide range of magnetic materials and excitation geometries, making it a versatile general tool rather than a one-off specialty experiment .

The YIG discovery: seeing four-magnon scattering in action

The team put MMM to the test on yttrium iron garnet (YIG), a magnetic insulator prized for its extremely low magnetic damping. At low excitation powers, magnons in YIG behaved as expected, propagating in a single well-defined direction. But when the researchers cranked up the power, something unexpected appeared on the detector .

The magnon population suddenly redistributed across momentum space, forming a distinctive elliptical ring pattern in the two-dimensional reciprocal plane. Rather than traveling in one neat line, energy scattered in many directions at once.

This pattern is the experimental signature of a four-magnon scattering process, a type of parametric instability. In simple terms, two initially excited propagating magnons collide and annihilate, spawning two entirely new magnons with different wave vectors. Because the process conserves energy and momentum, the newly created magnons can fan out across a wide range of directions, limited only by the available states in the material's dispersion relation .

Four-magnon scattering was previously known for spatially uniform (k=0) modes, but this is the first direct observation of the generalized case involving propagating magnons at finite wave vectors—a far more relevant regime for realistic devices where information must travel from point A to point B .

What this means for wave-based computing

The magnonics community has long envisioned using spin waves as information carriers in place of electrical currents. Since spin waves can propagate in insulators, magnonic circuits could theoretically operate with radically lower energy dissipation than conventional CMOS electronics—no moving charges, no Joule heating. Furthermore, the short wavelengths accessible with MMM correspond to THz frequencies, roughly 100× faster than today's gigahertz-scale CPU clock speeds .

But building useful computing primitives requires more than just propagating waves. It demands the ability to manipulate them—to steer, split, combine, and switch them the way transistors switch current. This is where nonlinear interactions become essential.

MMM gives researchers, for the first time, a practical tool to directly observe and quantify processes like four-magnon scattering. The ability to harvest these parametric instabilities could transform them from laboratory curiosities into functional computing elements: frequency converters, power limiters, logic gates based on interference, and even physical primitives for neuromorphic or reservoir computing schemes .

From imaging technique to enabling platform

Because MMM is material-agnostic and compatible with a broad range of excitation schemes—including electrical, optical, and acoustic driving—it can be extended well beyond YIG . The team anticipates extensions to antiferromagnetic materials, which host even faster spin dynamics, and to ultrafast pump-probe configurations that can capture transient nonlinear processes on femtosecond timescales .

The broader significance of this work lies not in a single discovery, but in lowering a barrier that has held back the entire field. Until now, the fastest magnons with the shortest wavelengths—precisely the ones needed for competitive chip-scale devices—have been largely invisible to the experimentalists trying to study them. MMM removes that invisibility, giving engineers a diagnostic toolkit that can actually keep pace with the physics they hope to exploit.