After more than 40 years of research, physicists have finally found convincing evidence of an event that had remained elusive until now.
At CERN, near Geneva, scientists collide particles traveling at extremely high speeds. These collisions briefly produce other particles, which almost immediately transform. Detectors record the traces left by the products of these transformations. But some are so rare that an immense number of collisions must be examined to identify them.

Views of sector 3-4 of the LHC tunnel.
Image: CERN
This is the case for a particle called the neutral B meson, or B0. It contains two tiny constituents, quarks, which can disappear together during a transformation. In the process being studied, this disappearance gives rise to two other particles, muons. These resemble the electrons found in atoms, but are much heavier.
The phenomenon is exceptional: according to current calculations, it occurs about once in every ten billion transformations of this particle. In other words, even a device capable of producing enormous numbers of B mesons must record gigantic quantities of data in the hope of finding a few events of this kind.
The LHCb experiment has made progress in this search. Its teams analyzed collisions recorded in 2024 and 2025, then combined them with older data. They found a signal strong enough to describe as solid scientific evidence. For the first time, an experiment provides this level of evidence for this specific transformation.

View of the LHCb experiment.
Image: CERN
Why did it take so long? Muons can appear in many other, far more frequent reactions. The rare pairs being sought must therefore be distinguished from a multitude of ordinary traces. Improvements to the detector and computing methods have made it possible to better eliminate these false candidates. Since its upgrade, LHCb has also been recording data much more quickly.
The result reaches a statistical confidence level of 3.2 sigma after combining the different measurements. That is enough to speak of preliminary evidence, but not yet of a definitively established discovery. Physicists generally require a level of 5 sigma before reaching that final milestone.
The significance goes beyond the event's mere rarity. Researchers have a theory that predicts the behavior of known particles: the Standard Model. By measuring these transformations, they test whether nature really follows their calculations. A persistent difference could reveal the influence of previously unknown particles, even ones impossible to produce directly in the accelerator.
For now, the results remain compatible with the predictions. The team must therefore collect more collisions to confirm the signal and measure its frequency more precisely. The high-luminosity accelerator, expected to begin operating from 2030 onward, and a further upgrade of LHCb planned for the mid-2030s should provide far more data.