An atom leaving a surface does not necessarily depart empty-handed. If an electron moves between a magnetic solid and an adsorbed atom during desorption, it may carry a preferred spin orientation—the quantum-mechanical property often described as a tiny magnetic direction. Measuring that transfer directly has been difficult because the atom is already moving into the gas phase.
Kanta Asakawa, an assistant professor at Yokohama National University, worked with former graduate student Naoki Tanabe and Professor Atsushi Hatakeyama at Tokyo University of Agriculture and Technology, and Oki Watanabe, Shuji Kamada, Keisuke Hara and Associate Professor Kaori Niki at Chiba University. Their experiment followed rubidium atoms as ultraviolet pulses released them from a magnetized cobalt film.
The departing atoms were spin polarized. When the researchers reversed the cobalt film’s magnetization, the sign of the measured rubidium polarization also reversed. Under comparable control conditions using an MgO substrate without cobalt, the signal stayed within experimental uncertainty.
Why Watch an Atom Leave?
Spin-dependent processes have been proposed in reactions including oxygen evolution and ammonia synthesis. One possible step is spin transfer from a magnetic catalyst surface to an adsorbed reactant. The participating electron may change the adsorbate’s magnetic state and influence which reaction pathway is accessible.
That idea is difficult to isolate in a working catalyst, where charge transfer, bonding, heat and many reaction intermediates overlap. The Japanese team therefore used a simpler model interface: rubidium atoms adsorbed on a spin-polarized face-centered-cubic cobalt (110) surface.
Rubidium donates its outer electron when adsorbed on cobalt. The researchers reasoned that ultraviolet excitation could send an electron back from cobalt to rubidium, neutralizing the atom and allowing it to leave. If that returning electron were spin polarized, the departing rubidium atom should retain a measurable trace.
Circularly Polarized Light as the Detector
The team grew a 10-nanometre epitaxial cobalt film on an MgO(110) substrate, magnetized it and exposed it to rubidium in ultrahigh vacuum. Pulses of 355-nanometre ultraviolet light triggered desorption at room temperature. The decreasing average departure speed as coverage increased supported a nonthermal, electronically driven process rather than ordinary heating.
A probe laser tuned to a resonant transition of rubidium-85 passed one millimetre above the sample. Departing atoms absorbed that light. The researchers alternated right- and left-circular polarization and compared the absorption, exploiting the fact that the two helicities interact differently with spin-polarized atoms.
Reversing the cobalt magnetization reversed the observed asymmetry. The reported average magnetic quantum number was −0.024 ± 0.003 in one configuration and +0.024 ± 0.004 in the other. The effect was small but reproduced with a second sample, and the estimated stray magnetic field was far too weak to explain it.
Measurement and Mechanism Are Different Claims
The optical measurement directly supports the conclusion that rubidium atoms leaving the cobalt surface were spin polarized in a direction controlled by the film’s magnetization. Density-functional calculations supplied the proposed mechanism: ultraviolet light excites spin-polarized cobalt 3d electrons, which transfer to adsorbed rubidium, neutralize it and permit nonthermal desorption.
The calculated direction of polarization agreed with the experiment, but the paper does not claim that every microscopic step was directly observed. The authors say the electron and spin dynamics during desorption remain speculative, that their calculation does not rule out other mechanisms and that wavelength-dependent studies are needed.
The apparatus detected only rubidium-85 atoms in one hyperfine ground state, F=3. The spin polarization of F=2 atoms was not measured. The authors also could not perform X-ray photoelectron spectroscopy during desorption and considered multilayer rubidium likely; atoms leaving upper layers may dilute the measured polarization because they do not receive electrons directly from cobalt.
| Principal institutions | Yokohama National University, Tokyo University of Agriculture and Technology, and Chiba University |
|---|---|
| Surface system | Rubidium adsorbed on a magnetized 10-nm fcc-Co(110) film grown on MgO(110) |
| Desorption trigger | 355-nm pulsed ultraviolet light in ultrahigh vacuum at room temperature |
| Detection | Spin-selective absorption of resonant right- and left-circularly polarized probe light |
| Key control | Reversing cobalt magnetization reversed the rubidium signal; MgO without cobalt did not reproduce it |
| Paper | “Light-induced spin-polarized desorption of Rb atoms from Co surfaces” |
| Journal | Physical Review B 114, L111412, published August 24, 2026 |
| DOI | 10.1103/56hb-jcdl |
A New Probe, Not Yet a Spin Catalyst
Tokyo University of Agriculture and Technology describes the result as the first observation of spin transfer from a solid surface to an adsorbed atom. That precedence claim belongs to the participating institutions and authors; Japan.co.jp did not independently review the complete surface-science literature to establish priority.
The durable contribution may be the measurement technique. It converts the spin state of a departing, optically detectable atom into a difference in circularly polarized light absorption. Comparisons across magnetic surfaces, adsorbates, wavelengths and coverages could test when spin survives charge transfer and when it relaxes.
Its scope is also clear. The method requires an atom that can be detected optically after desorption, and rubidium on cobalt is a model system rather than a water-electrolysis or ammonia-synthesis reaction. Using the result to design a more efficient catalyst will require experiments that connect measured spin transfer to reaction rates, selectivity, energy use and stability.
What Needs Testing Next
- Time-resolved electron and spin dynamics during desorption
- Measurements across ultraviolet wavelengths and surface coverages
- The unmeasured rubidium-85 F=2 hyperfine population
- Other adsorbates and magnetic surfaces that remain optically detectable
- A direct link between spin transfer and catalytic performance
The finding does not prove that spin transfer makes a catalyst faster. It does something more fundamental: it shows that spin information can cross a solid–atom boundary during a light-driven departure and can be read after the atom has left. That gives researchers a new way to test the microscopic step that future spin-catalysis theories depend upon.
- Tokyo University of Agriculture and Technology: spin transfer from a solid surface to an adsorbed atom (joint announcement, institutions and study summary; Japanese)
- Joint university press-release PDF (experimental method, terminology, limitations, funding and paper details; Japanese)
- Yokohama National University joint release (official affiliation and institutional account; Japanese)
- Chiba University joint release (research group, researcher profile and terminology; Japanese)
- Physical Review B research paper (authors, affiliations, methods, results, limitations and DOI; English)
Japan.co.jp reviewed the three universities’ joint materials, official researcher profiles and the Physical Review B paper available through August 30, 2026. Names, readings, titles and specialist terms follow Japanese primary sources; Roman spellings and affiliations follow the paper. The priority and spin-catalysis claims remain attributed, and the proposed microscopic pathway is distinguished from the direct optical measurement. No direct quotations are included.