AINEWS
A long suspected result in an exotic quantum state of matter, now confirmed by two independent teams, and a candidate ingredient for quantum computers that resist their own errors.
For three decades, the even-denominator ν=1/2 fractional quantum Hall state was the kind of result physicists cited but rarely measured cleanly. Theorists kept predicting that the excitations in this rare state of matter should carry one-quarter of an electron's charge. Every direct attempt kept sliding past. Now two independent groups, at the Weizmann Institute of Science and at EPFL, have both put the same quarter-charge value on the board from separate experimental setups on the same semiconductor platform. The result, posted to arXiv on February 9, 2026, retires a prerequisite measurement that condensed-matter physicists have been chasing since the early 1990s, as popular coverage of the preprint notes. It does not, by itself, deliver a topological qubit.
A "quasiparticle" is not a particle in the usual sense. In a many-electron system at temperatures near absolute zero and in strong magnetic fields, the electrons stop behaving as individuals. Their collective excitations, organized disturbances in the electronic fluid, behave like particles in their own right, with their own charge, mass, and statistics. A fractional charge, then, is not a broken electron. It is a new kind of charge that emerges from the coordinated motion of many electrons. The first such excitations, at the ν=1/3 state, earned the 1998 Nobel Prize in Physics. The state at ν=1/2 is rarer and, until now, far less settled.
ν, the filling factor, is the ratio of electrons to magnetic flux quanta threading the two-dimensional sheet they live in. ν=1/2 means roughly one electron for every two flux quanta, a regime where the simplest theories predict no quantized behavior at all, but where the real material keeps producing a quantized plateau anyway. Odd-denominator states like ν=1/3 and ν=2/3 are well-mapped and well-replicated. Even-denominator states, by contrast, are the long shots. The best-known one, ν=5/2, has been measured for years and remains the leading candidate for hosting the non-Abelian statistics a topological qubit would need. ν=1/2 sits in the same family of exceptions and has been harder still to pin down.
The two new measurements come from nominally identical devices built in the same material system: a 70-nanometer-wide gallium arsenide quantum well, the layered semiconductor used in many LEDs and high-speed chips, in which electrons are squeezed into a thin sheet and exposed to magnetic fields measured in tesla, the unit familiar from MRI magnets. An etch-defined quantum point contact, a narrow constriction etched into the device, partitions the quasiparticles as they flow past. From the shot noise, the random electrical fluctuations that betray the discrete nature of the charge carriers, in the weak-backscattering regime, where most quasiparticles flow through and only a small fraction scatter backward, the authors infer a charge of e/4. The Weizmann and EPFL teams report this same result in separate setups, which is the part of the story that turns an interesting measurement into a confirmed one.
The result does not yet pick between two possibilities the authors' own abstract leaves open. The ν=1/2 state might turn out to be Abelian, meaning its quasiparticles behave like ordinary, if fractional, charges when swapped around each other. It might also turn out to be non-Abelian, meaning the order in which quasiparticles are exchanged would alter the quantum state of the system, the property that would make it useful for a topological qubit. Probing exchange statistics remains future work, and the same gap sits at the heart of the more famous ν=5/2 story.
Topological qubits are a leading candidate for quantum machines that resist certain classes of errors without heroic overhead. The e/4 charge at ν=1/2 is not a qubit. It is a confirmation that the long-predicted excitations exist in a measured, replicated way. From there, a working topological computer still needs years of materials science, device fabrication, and statistics experiments the field has not yet finished. The white whale is on the table. The machine is not.