Light Caught Trembling in a Curved Spacetime on a Chip
13 Jun 2026, Yanjiang
A photonic chip emulates curved anti-de Sitter spacetime, where a light beam’s geodesic sway and rapid Zitterbewegung trembling reveal Dirac dynamics in a fabricated geometry.
When Dirac first solved the relativistic wave equation for the electron nearly a century ago, he discovered something that looked like a mathematical ghost: the particle should tremble. An electron, even in empty space, ought to jitter with a frequency so high that no instrument could ever track it. This Zitterbewegung — literally “trembling motion” in German — arose from the interference between positive- and negative-energy solutions of the Dirac equation. For decades physicists treated it as a curiosity, a phantom of a single-particle picture that would dissolve in the richer light of quantum field theory. Yet the concept refused to die. And now it has been seen — not in a particle accelerator, but in a glass chip the size of a fingernail, running through a spacetime that is not flat.
A team led by Alexander Szameit at the University of Rostock, working with collaborators in Würzburg, Montreal and beyond, has reported the first experimental emulation of a Dirac fermion moving in a genuinely curved, two-dimensional anti-de Sitter (AdS) spacetime. Their results appear in a preprint (arXiv:2606.09501), and they capture not just the graceful, pendulum-like geodesic swaying that curved space imposes on a particle, but also the rapid, trembling Zitterbewegung. The two motions, superposed in the same beam of light, open a new route to studying quantum dynamics in geometries that are otherwise accessible only to theorists’ imaginations.
The waveguides’ varying properties recreate the curved geometry of anti-de Sitter spacetime. This demonstration bridges abstract gravity theories with tangible optical experiments, advancing our understanding of the universe. (Source: arXiv:2606.09501)
Why AdS?
Anti-de Sitter space is not the universe we live in. Our cosmos is, to excellent approximation, de Sitter — a spacetime with a tiny positive curvature that drives the accelerated expansion. AdS, by contrast, has constant negative curvature, which acts like a gravitational bowl: any object placed in it feels a force pulling it back toward the centre. This peculiar property makes AdS the gravitational setting for one of the most profound conjectures in modern theoretical physics, the AdS/CFT correspondence. That correspondence asserts that a theory of gravity in AdS space is mathematically equivalent to a conformal quantum field theory living on its boundary — no extra dimensions required. If true, it means that some of the deepest questions about quantum gravity can be recast as problems in a more familiar, lower-dimensional quantum theory. The catch has always been that we cannot directly test the dynamics in the AdS bulk; no laboratory black hole sits on a tabletop, and building a curved spacetime in the traditional sense is, to put it mildly, impractical.
Szameit’s team side-stepped this impossibility with a clever sleight of hand. Instead of bending space, they bent the rules of light.
A Chip That Imitates Curvature
The key insight is an old one, though it was sharpened into a precise mapping by Koke and colleagues in 2016. The Dirac equation that governs a relativistic fermion in a curved spacetime looks, after some mathematical massaging, just like the equation that describes light hopping between neighbouring optical waveguides — provided you arrange those waveguides in a particular way. If you vary the waveguide properties along the array, you can encode an effective gravitational potential; if you choose the right variation, you can make the light behave exactly as if it were a Dirac particle moving in AdS₂.
Think of it as writing the score for a piece of music, then letting the orchestra play. The team fabricated an array of hundreds of coupled waveguides in a glass chip, each only a few micrometres wide. The spacing and detuning — the degree to which each guide is “off-resonance” with its neighbours — were engineered to follow a 1/cos(theta) profile, where theta is the spatial coordinate. This profile is no accident: it encodes the radial gravitational well of AdS₂, a spacetime that, despite being two-dimensional, contains the essential curvature of its higher-dimensional cousins. When a specially shaped pulse of light is injected into the chip, its subsequent spread and oscillation through the array mirrors, frame by frame, the fate of a Dirac fermion in AdS.
Light propagation along the waveguide axis stands in for the flow of time. The transverse position of the light beam across the array serves as the spatial coordinate. The whole arrangement is a brilliant shortcut: you do not need to warp the actual fabric of spacetime; you only need to warp the effective Hamiltonian that governs how the light diffuses.
The Dance Emerges
What happens when you let such a photonic fermion loose in an AdS bowl? The team answered that question by imaging the light as it emerged from successive lengths of the chip, snapshotting the “time evolution” in exquisite detail. The beam did not simply meander. It rocked back and forth at two distinct tempos.
A particle on a curved spacetime traces a slow arc overlaid with a rapid, trembling motion. This dual behavior shows how quantum and gravitational effects combine, deepening our understanding of fundamental physics. (Source: arXiv:2606.09501)
The first is a slow, sweeping oscillation. This is the geodesic motion — the kind any classical particle would undergo when trapped in a gravitational well. In AdS, curvature acts as a restoring force: push a particle away from the centre, and it slides back, overshoots, and settles into a steady rhythm. The team watched the centre-of-mass of the light pulse trace out this pendulum arc, and they verified that its frequency depends solely on the curvature strength. When they changed the effective mass of the simulated particle — by dialling a detuning offset in the waveguide array — the geodesic frequency did not budge. It is a signature of pure spacetime geometry: mass drops out. This is exactly what Einstein’s equivalence principle, embedded in the Dirac equation on a curved background, demands.
The second motion is a rapid, jittery trembling riding on top of the slow arc. That is the Zitterbewegung. It arises because the relativistic Dirac equation describes not one, but two types of solutions — positive- and negative-energy states — and any physical wave packet is a superposition of both. Their interference produces a beat whose frequency scales with the particle’s mass. Here the team saw precisely that: when they increased the effective mass, the trembling sped up, while the slow geodesic sway retained its steady pace. Moreover, when they cranked up the curvature — by strengthening the waveguide coupling — both frequencies rose together, showing that the Zitterbewegung is not an isolated quantum quirk but a motion jointly shaped by mass and geometry.
Not every experiment has the luxury of isolating two physical effects so cleanly, with independent control knobs for each. This one does.
What It Is, and What It Isn’t
Every emulation comes with a user guide, and this one is no exception. The mapping that makes the experiment possible also imposes boundaries. Time, in the real AdS universe, is a coordinate that lets you order events and define light cones — the fundamental causal structure of relativity. In the waveguide chip, the “time” coordinate is the propagation distance along the glass, which means that what the researchers observe is a spatial unfolding of a quantum state, not a sequence of truly dynamical events. The causal richness of the Lorentzian manifold — the distinction between “before” and “after” that depends on an observer’s motion — is not reproduced. The team is candid about this, as earlier theoretical analyses (including the foundational work of Koke et al.) have made clear.
A related nuance concerns symmetry. The continuous isometries of AdS₂ — the full SO(2,1) group — are only approximately preserved by the discrete waveguide lattice. In practice, the approximation is good enough to see the core phenomena, and the team’s comparison between measured centre-of-mass motion and analytic expectations from the continuous theory shows remarkably close agreement after accounting for a small lattice correction term. By deliberately choosing a discretization that respects the curvature profile, they managed to capture the essence of AdS dynamics without needing a perfectly symmetric lattice.
Viewed alongside other contemporary efforts — such as the proposal by Dey and collaborators to simulate holographic conformal field theories on hyperbolic lattices — this work fills a complementary niche. The bulk is where the gravitational dynamics live, and now, for the first time, physicists can watch a quantum particle navigate that bulk.
The Road Ahead
The experiment does not claim to solve the problem of quantum gravity. It does, however, build a stage on which some scenes from that grand play might be rehearsed. The photonic platform is inherently scalable: by engineering more elaborate waveguide patterns — perhaps introducing interactions, disorder, or even time-dependent modulation — it could simulate phenomena like Hawking radiation from analogue horizons or the scrambling of quantum information near an AdS boundary. As the authors write in their preprint, the system “establishes a scalable analog platform that may potentially be used for exploring dynamical aspects of holography.” The word potentially is honest; the path from a trembling light pulse to a working holographic simulator is long. But it now has a first, concrete step.
It is worth remembering how many physics revolutions have begun not by going to the thing itself, but by building a faithful imitation that isolates its essence. The chip is not AdS. The light is not a fermion. The trembling, however, is real — and it is now ours to control. For a field that has spent decades treating Zitterbewegung as a theoretical ghost, there is something quietly exhilarating about seeing it flicker, at last, on a camera screen.
— Yanjiang
Yanjiang is an online editor of LoomSci.com.
References
- Himmel et al., Experimental observation of hyperbolic spacetime dynamics, arXiv:2606.09501
- Natsuume, AdS/CFT Duality User Guide, arXiv:1409.3575
- Koke et al., Dirac equation in 2-dimensional curved spacetime, particle creation, and coupled waveguide arrays, arXiv:1607.04821
- Dey et al., Simulating Holographic Conformal Field Theories on Hyperbolic Lattices, arXiv:2404.03062