Researchers at the University of Oxford, Diamond Light Source and ShanghaiTech University developed a new X-ray technique that captures the complete three-dimensional dynamics of coupled magnon modes with unprecedented precision, and for the first time reconstructs their vectorial eigenfunctions in reflection geometry. The findings have been published in Nature Nanotechnology.
Many physical systems are governed by waves, from ripples on water to light and sound. In magnetic materials, the equivalent are collective oscillations of electron spins, known as magnons, which underpin technologies from data storage to next-generation spin-based computing. While the frequencies of these waves are routinely measured, their full motion — how they evolve, rotate, and interact — has until now remained largely hidden.
The new method, called X-ray Magnetic Vector Chronoscopy (XMVC), changes that. By exploiting the pulsed nature of synchrotron X-rays at Diamond Light Source's I10 beamline, the technique stroboscopically films the magnetic dynamics — capturing snapshots of spin motion across an entire oscillation cycle at intervals of a few tens of trillionths of a second. Crucially, by tuning the X-ray energy to specific absorption peaks of chemical elements, the team could isolate and track the motion of individual magnetic layers within a device independently.
The approach is analogous to how a CT and MRI scan builds up a full three-dimensional image from many two-dimensional slices. In the same way, the team reconstructed the full magnetic state of the system from multiple experimental measurements, rather than relying on theoretical assumptions.
'What matters is not just the frequency, but the full motion. With XMVC, we can now directly see how spins behave in space and time — and that opens a new window on the physics of magnetic materials,' says Professor Shilei Zhang of ShanghaiTech University, who is the senior author of the study.
Applying XMVC to a synthetic antiferromagnetic multilayer — a stack of two ultrathin magnetic films separated by a ruthenium spacer — the researchers reconstructed the complete three-dimensional trajectories of the spins in each layer. This revealed not only how strongly the spins were oscillating (amplitude), but also how the motion in one layer related to the other (phase relationships), whether they moved in step, out of step, or anywhere in between. It also showed the precise way energy was exchanged and lost between the two layers. Previously, these properties could only be inferred through indirect modelling.
‘Until now we could measure the frequencies of these magnetic oscillations, but not the motion itself,’ comments Professor Thorsten Hesjedal, a coauthor of the study. ‘By tuning the X-rays to specific elements, XMVC allows us to observe—layer by layer—exactly how the spins move and how energy is exchanged between the two layers. This direct insight is crucial for designing the next generation of spintronic and magnonic devices.’
The ability to directly visualise spin dynamics has important implications for emerging technologies. Fields such as spintronics and quantum information processing depend critically on controlling how magnetic excitations propagate, couple, and dissipate. XMVC provides, for the first time, a direct experimental window into these processes at the microscopic level—opening new opportunities to design and engineer magnetic systems with unprecedented precision.
More broadly, this work highlights a shift in synchrotron science: moving beyond measuring signals to reconstructing the full behaviour of physical systems in space and time.
The work was carried out in collaboration with researchers at the Beijing National Laboratory for Condensed Matter Physics and the Chinese Academy of Sciences. Ethan L. Arnold, a graduate student in the Department of Physics, University of Oxford, contributed to the experimental work through a Diamond–EPSRC studentship.
Reconstruction of the magnon eigenfunctions by X-ray magnetic vector chronoscopy, H Jin et al, Nature Nanotechnology, 27 May 2026