In 1939, a Soviet physicist predicted that a recoiling atomic nucleus would eject one of its own electrons. Last January, scientists finally saw it happen - and it may be the key to finding dark matter.
Arkady Migdal was 28 years old and working through a theoretical problem in quantum mechanics in 1939 when he arrived at a prediction so subtle - and so difficult to test - that it would sit in textbooks for nearly nine decades without anyone confirming whether it was actually true. The prediction: when an atomic nucleus is struck and recoils suddenly, the electrons orbiting it cannot respond fast enough. In that moment of nuclear displacement, the quantum state of the atom is disrupted, and an electron can be torn free. A tiny, measurable ionization signal - from an event too small to detect by any other means. In January 2026, a team of Chinese physicists reported in Nature that they had seen it. After 87 years, the ghost signal finally appeared. And the implications stretch far beyond confirming an old prediction.
The experiment was conducted by researchers at the University of the Chinese Academy of Sciences and published in Nature in January 2026. The team set out to produce the Migdal effect in a controlled setting: firing a neutron beam at a gas-filled detector, using the neutrons to cause nuclear recoils in the gas atoms, and then looking for the accompanying electron ejections that Migdal had predicted would follow.
The challenge was instrumentation. Nuclear recoils from neutron collisions happen all the time in particle physics experiments. The Migdal electron - the quantum bonus that should accompany the recoil - is a far fainter signal, easily swamped by background noise. To isolate it, the researchers developed an ultra-sensitive detection system that combined a micro-pattern gas detector with a pixelated readout chip: essentially a camera capable of photographing the moment a single electron is set free, while simultaneously recording the nuclear recoil that triggered it. Both signals had to appear together, in the right spatial relationship, to qualify as a candidate event.
Out of nearly one million recorded events, the team identified six candidate events that met all the criteria for the Migdal effect - a correlated pair of a nuclear recoil and an ejected electron, appearing simultaneously in the correct geometry. Statistical analysis of these six events yielded a significance of five standard deviations, which is the physics community's threshold for a confirmed discovery.
The result was published in Nature alongside commentary from the MIGDAL Collaboration, a separate international group that has been building toward the same detection at the Rutherford Appleton Laboratory in the UK using a different experimental approach. The two groups converging on the same result from different methods is a strong signal that the detection is real. For the first time in 87 years, Migdal's 1939 calculation has a data point.
To understand why Migdal's prediction matters, you need to understand something about how atoms work - specifically, the relationship between the nucleus and its surrounding electrons. The electrons do not simply orbit a stationary nucleus like planets around a sun. Their probability distributions - their quantum wavefunctions - are shaped by the nuclear charge they feel. The electrons "know" where the nucleus is, in the sense that their behavior is determined by it.
Now consider what happens when that nucleus is suddenly struck by a passing particle and kicks sideways. From the electrons' perspective, the ground has shifted without warning. Quantum mechanics has a framework for thinking about this: the sudden approximation. If the nucleus moves fast enough that the electrons have no time to adjust their wavefunctions - no time to "follow" the nucleus to its new position - then the electrons find themselves in a superposition of states relative to the new nuclear position. Most of the time, they settle back into the ground state without incident. But occasionally, with a small but calculable probability, the disruption is enough to promote an electron into an excited state - or kick it out of the atom entirely. That ejected electron is the Migdal effect.
Migdal worked this out in 1939 using what he called the "sudden approximation" - a technique he developed specifically for this problem and which subsequently became a standard tool in quantum mechanics textbooks worldwide. The mathematics produces a clear prediction: for a nucleus of a given mass undergoing a given recoil, the probability of the accompanying electron ejection can be calculated. It is small - but it is not zero.
"Migdal predicted a signal so faint it would take 87 years to build a camera sensitive enough to see it. That we found it at all says as much about how far detector technology has come as it does about the physics."
- Lisa Pedrosa, lisapedrosa.comThe reason it took 87 years to confirm is not that anyone doubted the prediction in principle. Most quantum physicists assumed it must be correct - the underlying physics is well-established, and the sudden approximation is a reliable framework. The problem was purely technical: the Migdal electron is a very faint signal in a very noisy environment, and distinguishing it from background required detector sensitivity that simply did not exist until very recently.
Dark matter is the dominant form of matter in the universe. It constitutes roughly 85 percent of all matter and 27 percent of the total energy content of the cosmos. We know it exists because of what it does - bending light, shaping galaxies, providing the gravitational scaffolding on which all visible structure hangs. We have never directly detected a dark matter particle, and after decades of searching, the reason is becoming clearer: we may have been looking in the wrong mass range.
The most heavily searched candidates were WIMPs - Weakly Interacting Massive Particles - with masses in the range of 10 to 1,000 times the proton mass (10-1,000 GeV). Experiments like LUX-ZEPLIN and XENONnT have set extraordinarily sensitive limits in this range, and as of 2026, the classic WIMP parameter space is largely ruled out. Attention has shifted to lighter candidates: dark matter particles with masses below 1 GeV - in the same mass range as ordinary matter particles like electrons and protons.
This is where the Migdal effect becomes critical. A light dark matter particle - say, one with a mass of 10 MeV, roughly 10 times the electron mass - is simply too light to produce a detectable nuclear recoil in conventional detectors. When such a particle strikes a nucleus, the recoil energy is so small - well below the detector threshold - that the event registers as nothing. The dark matter passes through undetected.
The Migdal effect changes this. Even if the nuclear recoil itself is below threshold, the accompanying Migdal electron - which is a separate, independently detectable signal - may exceed the threshold. The dark matter interaction that was invisible in nuclear recoil space becomes visible in electronic recoil space. The Migdal electron is a quantum amplifier, converting an undetectable nuclear whisper into something a detector can hear.
This has been understood theoretically since around 2018, when physicists began working out the cross-sections and detection rates for Migdal-enhanced dark matter searches. The problem was that those calculations relied on the Migdal effect being real - and until January 2026, that had never been experimentally confirmed. Using an unconfirmed effect as the basis for a detection claim would not be scientifically credible. Now it is confirmed. The window is open.
The confirmation of the Migdal effect is significant not just as a validation of quantum mechanics - though that alone would be noteworthy. Its primary impact is that it unlocks an experimental strategy. Dark matter experiments can now be designed and operated with the Migdal channel explicitly in view, knowing that the signal they are looking for is real and that their detector physics will behave as calculated.
Several experiments are already positioned to exploit this. The MIGDAL Collaboration at Rutherford Appleton Laboratory was built specifically for this purpose: a Optical Time Projection Chamber filled with CF4 gas, capable of imaging both the nuclear recoil and the Migdal electron track in three dimensions. Their approach should eventually reach sensitivities deep into the sub-GeV mass range. The XENONnT experiment, which uses liquid xenon and has set the current best limits on WIMP-nucleon interactions, can also search for Migdal events - and with 7.8 tonne-years of data now published, they are beginning to do so. SuperCDMS at SNOLAB, reaching its target operating temperature in late 2025, brings cryogenic semiconductor detectors with quantum sensors into play for light dark matter searches.
The deeper question is whether dark matter in the sub-GeV range actually exists. The answer is not known. What is now known is that if it does, and if it interacts with ordinary matter at sufficient rates, the Migdal effect gives us a new handle on detecting it - one that did not exist before January 2026, or rather, one that could not be trusted to exist before then. The 87-year wait for Migdal's confirmation has turned out to be, in a certain light, a 87-year wait for the next generation of dark matter searches to have a solid experimental foundation.
Arkady Migdal died in Princeton in 1991, a celebrated Soviet physicist who lived long enough to see his 1939 approximation become part of the standard quantum mechanics curriculum. He did not live to see it confirmed. But the confirmation has come - in six events out of a million, at five standard deviations of significance, published in Nature - and the search for the universe's invisible majority has a new tool to work with.
© 2026 Lisa Pedrosa · lisapedrosa.com
All articles cited to primary institutional or peer-reviewed sources
Buy me a coffee