The 200-Year-Old Shadow That Just Rewired a Photonics Frontier

TL;DR

• A team at Nanyang Technological University (NTU) Singapore, led by Assistant Professor Shen Yijie, has generated optical skyrmions — stable, swirling topological light patterns — using nothing more than a laser aimed at a small circular disc.
• The method exploits the Poisson spot, a bright dot at the centre of a circular object's shadow, first predicted in 1818 as an attempted disproof of wave theory. It turned out to prove it instead.
• The setup replaces the nanofabricated metasurfaces and metamaterials normally required to build optical skyrmions — hardware that can cost hundreds of thousands of dollars per chip and requires clean-room access.
• Bonus finding: the same shadow produces four different classes of skyrmion at once — spin, Stokes, electric-field, and magnetic-field — nested in a single spot of light.
• Published in Optica on 18 June 2026; re-surfaced globally by ScienceDaily on 13 July, driving fresh virality across physics X and r/Physics.
• What this is not: a data-storage breakthrough. What it is: a democratisation event for a field that has been gated by nanofabrication.

Paris, 1818

The French Academy of Sciences ran a competition that year to settle an argument. Was light a wave, as Augustin-Jean Fresnel claimed, or did it travel as tiny corpuscles in straight lines, as Isaac Newton had insisted for more than a century?

Siméon Denis Poisson, a mathematician on the judging panel and a committed particle-theorist, worked through Fresnel's equations and found what he considered a killing objection. If light really behaved as a wave, then the very centre of the shadow cast by a small circular disc — the darkest point of the darkest region — should contain a bright spot of light. This was, Poisson said, absurd. Everyone could see shadows. Nobody had ever seen a bright dot in one.

The chair of the committee, François Arago, went to check. He placed a small disc in a beam of light, looked at the centre of its shadow, and found the bright dot exactly where Fresnel's wave theory predicted.

Fresnel won the prize. The bright dot — an experimental accident that was supposed to demolish wave theory and instead confirmed it — has been known ever since as the Poisson spot, occasionally the Arago spot, and in undergraduate physics labs everywhere as the thing you show students when you want them to actually believe in diffraction.

Two centuries later, the same bright dot in the same shadow has just been shown to contain something Poisson could not possibly have imagined: a stable, mathematically knotted swirl of light.

What the NTU team actually did

Optical skyrmions are among the more exotic constructions in contemporary photonics. Named after Tony Skyrme's 1960s work in particle physics, they are localised, particle-like patterns in which some property of a field — spin, polarisation, electric direction, magnetic direction — twists continuously around a central point. Imagine the spines of a hedgehog, but the spines are arrows describing how light is vibrating, and the whole hedgehog is a few micrometres across.

Skyrmions are prized for a specific reason: they are topologically protected. You cannot smoothly deform them into a uniform field. To unwind one, you have to inject enough energy to break the whole pattern. That stability — the same reason physicists care about them in magnetic materials — makes them a candidate substrate for storing information at densities well beyond anything conventional magnetic media can achieve.

The catch, until now, has been how to make them in light. The standard approach uses metasurfaces: two-dimensional arrays of sub-wavelength nanostructures, patterned by electron-beam lithography, engineered atom by atom to bend light into precisely the required topology. It works. It is also expensive, slow, and requires the kind of nanofabrication facility that a few dozen institutions in the world can afford to run.

The NTU team, in a paper published in Optica on 18 June 2026 (Yao et al., 13(6): 1184), showed that you do not need any of it.

Their setup: a coherent laser. A small circular disc — the kind you can machine in an afternoon. A screen to catch the shadow. In the centre of that shadow, in the Poisson spot itself, the light is already arranged in exactly the swirling topological structure that metasurfaces are painstakingly engineered to produce.

The physics behind why is, in retrospect, elegant. The bright dot at the centre of the shadow is formed by light waves that have bent around every point on the edge of the disc and arrived simultaneously in phase at the central axis. That geometric symmetry — a continuous ring of bent light meeting at a point — turns out to encode the same topological structure that skyrmions require. The disc is doing the work the nanofabrication was doing. It has been doing it, quietly, for 200 years.

The four-in-one detail

The paper's more surprising result is what the researchers found when they looked more carefully at the spot. A single Poisson spot, in their setup, was carrying four distinct kinds of skyrmion simultaneously: a spin skyrmion (in the rotational polarisation of the light), a Stokes skyrmion (in the polarisation state), an electric-field skyrmion, and a magnetic-field skyrmion.

This matters more than it might sound. Building any one of those in isolation used to require its own specialised metasurface. Getting all four to coexist in a single controllable structure would, on the previous methodology, have been a multi-year fabrication project. Here they emerge together, in the same shadow, from the same disc.

For researchers who want to study how different kinds of skyrmion interact — and this is a live open question in the field — the NTU setup has just handed them a laboratory bench that used to be a decade of grant applications away.

What this isn't

It is not a data-storage breakthrough. Optical skyrmions as an information medium remain a research problem, not an engineering one. Reading and writing individual skyrmions at speed, addressing them individually, integrating them with any existing electronic system — all of that is unsolved. The NTU work does not change the timeline to commercial optical-skyrmion storage. It changes the number of labs that can meaningfully work on it.

It is also not, as some coverage has framed it, a rediscovery of forgotten physics. The Poisson spot has never been forgotten. It is in every undergraduate optics course on Earth. What is genuinely new is the recognition that its geometry, taken seriously as a topological object rather than a curiosity, produces the same structures that a whole subfield had been fabricating from scratch.

The right frame is not lost physics found. It is known physics re-read — with an eye that had been trained by the last decade of topological photonics to see what was already there.

Who this actually helps

Small photonics groups in universities that cannot afford clean-room fabrication. Research programmes in countries where nanofabrication capacity is limited — much of Southeast Asia, most of Africa, most of Latin America. Graduate students who now have a viable dissertation-scale experiment on topological light without needing to negotiate access to a shared metasurface facility.

For the broader public, the practical implications sit years away. If skyrmion-based storage eventually reaches products, the density improvements over current solid-state memory could be substantial — but the timelines are decade-scale, and the bottleneck was never metasurface fabrication. It was, and remains, control and readout.

The near-term consequence is quieter and more interesting: a whole area of photonics just became several orders of magnitude cheaper to enter. That kind of moat collapse tends to produce results a funding-body forecast cannot easily see coming.

What's still uncertain

• Fidelity at scale. The paper demonstrates skyrmion generation and characterises the four-in-one structure. It does not yet demonstrate the precision and reproducibility needed to use these skyrmions as information carriers. That work is ahead.
• Dynamic control. Metasurface approaches allow arbitrary skyrmion topology to be programmed into the surface. The Poisson-spot method produces a fixed family of structures set by the disc geometry. Whether you can dynamically switch between skyrmion types — essential for any storage or computing application — is an open question.
• The interaction physics. The four coexisting skyrmion classes may interact in ways the theory has not fully mapped. That is either a limitation or, more likely, a research programme.

Bottom Line

Optical skyrmions have spent the last decade trapped behind a nanofabrication moat. A team in Singapore has just walked around the moat with a laser and a disc, using a piece of physics that has been sitting in undergraduate lab manuals for two hundred years. This will not put skyrmion storage in your laptop next year — but it will put skyrmion research in a lot more laboratories next year, and that is how frontiers actually move.

Sources

• Tier 1: Yao J., Xie X., Meng Y., Sun S., Hu J., Shen Y., Yang Y. "Optical skyrmions in Poisson spots." Optica 13(6): 1184, 18 June 2026. DOI: 10.1364/OPTICA.591840. Peer-reviewed.
• Tier 1/2: NTU Singapore official media release, "Creating complex light patterns using a two-century-old light phenomenon" (June 2026).
• Tier 2: ScienceDaily, "A 200-year-old physics experiment could help build future computers" (13 July 2026).
• Tier 2: Phys.org, "Scientists create optical skyrmions using a two-century-old light phenomenon" (23 June 2026).
• Tier 2: EurekAlert (AAAS), NTU Singapore press feed (23 June 2026).
• Tier 2: The Debrief, "Two-Century-Old Light Phenomenon Could Power the Next Generation of Communications and Data Storage Tech" (June 2026).