Some quantum effects are so clean in theory that they take decades to meet in experiment. The Aharonov-Bohm effect is one of them. Predicted in 1959, it says that a charged particle traveling around a region of magnetic flux acquires a shift in its quantum phase — even if it never passes through the magnetic field itself. The particle feels no force along its path, yet the interference of its possible routes records the flux it encircled. Electrons in metal rings showed the effect in the 1980s, and it has since become a standard probe of quantum coherence in mesoscopic physics.
Its optical cousin has been harder to pin down. Excitons are electrically neutral, so they should be indifferent to magnetic flux. But an exciton is a dipole — an electron and a hole held apart — and a dipole moving around a closed loop encloses flux between its two charges. Theory has long predicted that the two partners would accumulate opposite phases, leaving a net, magnetic-field-dependent interference on the light the exciton emits. That is the optical Aharonov-Bohm effect: no charge, no force, but a phase written into the photoluminescence anyway.
Now, in a paper published October 22 in Nature Communications, a Brazilian-led team reports what appears to be the first observation of that effect in a moiré system. Working with twisted bilayer molybdenum disulfide aligned near 60 degrees, the researchers found that magneto-photoluminescence from spin-polarized excitonic complexes oscillates periodically with magnetic field — intensity, not energy, rising and falling in a pattern the authors interpret as optical Aharonov-Bohm interference. The frequencies of the oscillations form a harmonic hierarchy tied to the moiré-cell area, which is the fingerprint that the phase is geometric: it measures flux through the nanoscale pockets where the excitons sit.
The paper — “Evidence of optical Aharonov-Bohm interference in spin-polarized moiré excitonic complexes,” DOI 10.1038/s41467-026-78018-6 — is led by Andreij C. Gadelha of the Universidade Federal de São Carlos (UFSCar) and the Centro Brasileiro de Pesquisas Físicas (CBPF), with Marcio D. Teodoro of UFSCar and Andrey Chaves of the Universidade Federal do Ceará (UFC) and the University of Antwerp as corresponding authors alongside Gadelha. The author list spans four Brazilian institutions — UFSCar, CBPF, UFC, and the Universidade Federal de Minas Gerais — plus the National Institute for Materials Science in Tsukuba, Japan, whose Kenji Watanabe and Takashi Taniguchi supply the high-purity hexagonal boron nitride crystals that have become the encapsulation standard for clean two-dimensional devices. Received in February and accepted in September, the work arrives barely a day before this edition goes to press.
The wording of the title deserves attention. The authors claim evidence of the effect, not its definitive observation, and that caution is appropriate for a first report of a phenomenon this subtle. What follows is what they found, why the interpretation is plausible, and where the caveats live.
The Aharonov-Bohm effect, in brief
The original effect is a statement about the vector potential. In classical electromagnetism, only the electric and magnetic fields are physically real; the potentials are a mathematical convenience. Aharonov and Bohm showed that quantum mechanics disagrees. A charged particle that travels two paths around a region containing magnetic flux picks up a relative phase of 2πΦ/Φ₀, where Φ is the enclosed flux and Φ₀ = h/e is the flux quantum. The particle never enters the field region. The phase is still there, and it shifts the interference pattern.
For electrons in rings, that phase produces resistance oscillations periodic in the flux — the familiar h/e periodicity that mesoscopic physicists have measured for forty years. The flux quantum sets the scale: one period corresponds to one flux quantum threading the loop. That is what makes the effect an interferometer. If you know the field, the oscillation period tells you the area of the loop.
Why a neutral exciton can care about flux
An exciton carries no net charge, so the textbook version of the argument seems not to apply. The electron would pick up a phase going one way around the flux, the hole would pick up the opposite phase going the same way, and the two would cancel. But that cancellation assumes the electron and hole travel the same loop. In a confined geometry where they do not — where the electron’s orbit and the hole’s orbit enclose different areas — the cancellation is incomplete. What survives is proportional to the area between the two orbits, times the magnetic field: a genuine Aharonov-Bohm phase for a neutral composite.
This is the idea behind the optical Aharonov-Bohm effect as it was originally proposed for semiconductor nanostructures. In a type-II quantum dot, for instance, the hole sits inside the dot while the electron is confined to a ring around it; the exciton’s ground-state energy then oscillates with magnetic field as flux threads the annulus between them. Theorists worked out these predictions in the 1990s and 2000s; clean experimental signatures proved elusive. The paper’s abstract frames the history plainly: the optical analogue “has been predicted but not yet observed in moiré materials.”
Moiré materials change the experimental situation in a useful way. Instead of a single etched ring, a twisted bilayer provides a built-in, periodic array of nanoscale traps — the moiré pockets — each one a natural interferometer loop, all of them nearly identical. Near 60 degrees of twist, the two MoS2 layers sit in the so-called H-type (2H-like) stacking, and the moiré superlattice defines a regular landscape of confinement sites. An excitonic dipole trapped in such a pocket is, for the purposes of this experiment, an exciton in a nanoscale cavity. Sweep a magnetic field, and the flux through the pocket grows; the exciton’s quantum phase should cycle; the emitted light should record the interference.
The experiment
The team’s probe is magneto-photoluminescence: excite the sample with a laser, collect the emitted light, and watch what happens as a perpendicular magnetic field is ramped. In an ordinary semiconductor, a magnetic field shifts exciton energies (the Zeeman effect) and splits spin states. What the authors report is different. The intensity of the exciton emission — and of the biexciton peak, the four-particle complex — oscillates periodically with field. Bright, dim, bright, dim, in a regular rhythm.
Intensity oscillations are the more interesting observable here. An energy oscillation would mean the field is tuning the exciton’s level. An intensity oscillation means the field is tuning how efficiently the exciton converts into a photon — the interference is modulating the optical transition itself, brightening and darkening the emission as the accumulated phase cycles through constructive and destructive values. That is the behavior you expect if the exciton’s wavefunction is acquiring a flux-dependent phase inside its trap.
The decisive detail is the harmonic structure. The oscillations are not a single sine wave; they appear at multiple frequencies in a harmonic hierarchy, and those frequencies correspond to the moiré-cell area. In other words, the fundamental period matches one flux quantum through an area the size of a moiré unit cell, and the harmonics match multiples of that area. This is the interferometric signature: the loop whose area the measurement extracts is the moiré pocket itself. If the oscillations came from some unrelated magnetic effect — Landau quantization of free carriers, say, or an artifact of the magnet optics — there would be no reason for their periods to lock to the superlattice geometry. The twist angle sets the moiré period, the moiré period sets the cell area, and the cell area sets the oscillation frequency. That chain is what makes the Aharonov-Bohm reading compelling.
Why spin polarization is in the title
The complexes the team studies are spin-polarized — a detail that matters for coherence. A magnetic field Zeeman-splits the exciton spin states, and at sufficient field the emission is dominated by one spin branch. That selectivity isolates a single coherent channel — the interference is not washed out by averaging over branches with different phases — and confirms the emitters are well-defined excitonic complexes confined in the moiré potential.
The mention of biexcitons is significant in the same way. A biexciton is a bound pair of excitons, a four-particle complex, and its appearance in the oscillation data means the interference survives in a correlated few-body state, not just in the simplest single exciton. That extends the claim from “an exciton shows this phase” to “excitonic complexes show this phase,” which is the stronger statement the title makes.
What the field gains
Suppose the interpretation holds — and the usual confirmations, replications, and alternative-explanation checks will decide that over the next year. What does the field get?
First, it gets a proof that moiré-trapped excitons behave as coherent quantum dipoles sensitive to magnetic flux. Until now, the evidence for moiré confinement has been spectroscopic: multiple sharp emission lines, polarization patterns, saturation behavior — all consistent with excitons sitting in moiré wells, but none of them a direct measurement of the trapped wavefunction’s phase coherence. An Aharonov-Bohm oscillation is exactly that: a phase measurement. It says the exciton is not just energetically localized but quantum-mechanically coherent across the extent of its pocket.
Second, it gets an interferometric probe of the twist-angle landscape. The oscillation period encodes the enclosed area, which encodes the local moiré period, which encodes the local twist angle. Twist disorder — the fraction-of-a-degree variation across real samples that every moiré experimentalist fights — could in principle be mapped by reading the oscillation frequency point by point. That turns the effect from a curiosity into a characterization tool, a way to image the superlattice’s uniformity through the excitons’ own interference.
Third, it opens a door to moiré quantum optics. If moiré pockets are coherent interferometers whose phase is magnetically tunable, they become controllable quantum elements: flux-tuned emitters, phase-sensitive detectors of local magnetic environments, building blocks for schemes in which the interference itself carries information. The moiré lattice is not just a trap array; it is an array of interferometers.
Reading the caveats
The careful reader will notice how much weight rests on interpretation. Periodic magneto-optical oscillations can have more than one parentage, and the history of mesoscopic physics includes signals that looked like Aharonov-Bohm interference and turned out to be something else. The authors’ case rests on the harmonic hierarchy and its lock to the moiré-cell area, which is the right kind of evidence — a geometric fingerprint rather than a single wiggle — but it is still one sample system, one material, one stacking configuration. Replication in other twisted homobilayers and heterobilayers, at other angles, will matter. So will companion theory: quantitative modeling of which complexes produce which harmonics, and whether any competing mechanism can mimic the pattern.
There is also the question the title’s modesty raises. “Evidence of” is the language of a field that has been burned before by overclaiming first observations. It is the right language here. The result is a strong candidate for the first optical Aharonov-Bohm effect in a moiré material; it is not yet the closed case.
What to watch next
The immediate follow-ups are easy to list. Other groups will look for the same intensity oscillations in twisted MoS2 at nearby angles and in related homobilayers — WSe2, MoSe2 — to test whether the harmonic hierarchy tracks the moiré period as the model demands. Time-resolved measurements could separate the interference dynamics from slower processes. And theorists will want to compute, from realistic moiré potentials, the expected oscillation spectrum for specific excitonic complexes, tightening the comparison between the measured harmonics and the geometry of the traps.
Longer term, the interesting question is control. An interferometer you can tune with a magnetic field is a component. An array of them, defined by a twist angle you choose during fabrication, is a platform. The distance from this paper to a device is large — laboratory magnetic fields and delicate samples do not become technology overnight. But every quantum technology begins as a coherence measurement. This one measures the phase of light-emitting dipoles trapped in a lattice made of twist. That is new, and it is worth watching.
Sources and further reading
The primary source is A. C. Gadelha et al., “Evidence of optical Aharonov-Bohm interference in spin-polarized moiré excitonic complexes,” Nature Communications (published October 22, 2026; received February 4, accepted September 7, 2026), DOI 10.1038/s41467-026-78018-6 — open access, with the full author list and affiliations (UFSCar, CBPF, UFC, UFMG, NIMS, University of Antwerp) on the article page. For the original Aharonov-Bohm effect, see Y. Aharonov and D. Bohm, “Significance of electromagnetic potentials in the quantum theory,” Physical Review 115, 485 (1959), and review literature on mesoscopic Aharonov-Bohm oscillations in metal and semiconductor rings. For the optical Aharonov-Bohm effect as theoretically proposed, see the literature on excitons in type-II quantum dots and quantum rings from the 1990s–2000s, where the energy-oscillation version of the effect was worked out. For moiré excitons, confinement, and twist-angle metrology, see the companion article on moiré excitons in this publication and review articles in Nature Reviews Materials, Nature Photonics, and Reviews of Modern Physics on moiré materials and moiré excitonics. Paper-specific claims above — the near-60° twist, the exciton and biexciton intensity oscillations, the harmonic hierarchy — follow the authors’ abstract and the journal’s description; quantitative details beyond those await the full text and are deliberately not stated here.