Physicists at CERN have found evidence of a particle transformation so rare that it takes massive amounts of collision data to detect. The LHCb experiment has seen evidence of a neutral B meson decaying into a pair of muons. This process is predicted to occur only once in every 10 billion decays of this particle.
It is not yet a discovery by particle physics standards, but it represents a step toward measuring a process that has eluded experimenters for over 40 years. Its rarity also makes it a good place to look for any sign that nature might behave differently from the Standard Model.
A particle decay that happens just once in 10 billion cases
The process is written as B⁰ → μ⁺ μ⁻, a neutral B meson decaying to a positive muon and a negative muon. The muon is essentially an electron but much heavier. The predicted branching fraction is 1.03 × 10⁻¹⁰.
This minuscule rate is why the signal is so hard to detect. The LHC produces a vast number of particles in a proton collision, but the vast majority of B mesons decay in other ways. Researchers had to analyze large numbers of collision events to find the candidates that showed this decay pattern. They also had to separate the genuine candidates from all the other processes, which can give rise to a similar detector signal.
The research has been going on for over 40 years. A combined LHCb and CMS analysis reported initial evidence of this decay in 2015, but it did not establish a discovery. Subsequent analyses have not established the decay. The latest analysis provides more data and improves the case, but it’s still not quite at the level required for a discovery.
Better detectors and machine learning help isolate the signal
The new result is based on collision data taken in 2024 and 2025 by LHCb, one of the experiments at CERN’s Large Hadron Collider. The LHCb detector has been upgraded to handle collisions at a much higher rate. The team reports that the upgraded detector collects data five times faster than that its predecessor did between 2010 and 2018, while also improving its ability to identify particles.
This is important because the decay of interest is buried among much more common events. Researchers have used machine-learning techniques to separate the possible B⁰ decays from background events. The software is used to classify the events so that the physicists can then assess whether the remaining candidates are consistent with the expected signature.
The statistical tests and the various checks are still required to establish whether a small excess is in fact a real signal or just a statistical fluctuation. The new data on its own produces a statistical significance of 3.0 standard deviations.
This increases to 3.2 standard deviations when the researchers combine that result with data from LHCb’s first two running periods. The analysis was conducted by an international team of 19 researchers and co-led by Marco Santimaria.
Evidence is not yet a confirmed discovery
Particle physicists normally require a result to reach five standard deviations (5 sigma) before they call it a discovery. A 3.2 sigma result is therefore termed evidence and not a discovery. This is equivalent to about a one-in-1,500 chance that background processes alone could produce a signal of that strength, assuming the analysis and the statistical model are correct.
This behavior is in accordance with the Standard Model, the theory that describes the fundamental particles and three of the four fundamental forces. The Standard Model does not include gravity, and there is ongoing work to test its predictions for any deviations.
The same analysis also produced the most precise measurements to date of another related process, the decay of a strange B meson, Bₛ⁰ → μ⁺μ⁻. More data from future LHC running and the planned upgrades are expected to improve the precision of both measurements. The result advances a search that has been going on for decades, but confirmation still requires further data.
Sources: INFN, ScienceDaily, arXiv, Physics World
Affiliate links on Cult of Mac may earn us and our partners a commission.