The NA62 experiment at CERN reports a new measurement of an ultra‑rare kaon decay, K+ → π+ ν ν̄. Combining data taken from 2016 through 2024, the collaboration finds a branching ratio — the probability that a K+ will decay this way — of (9.6 +1.9 −1.8) × 10⁻¹¹. This result is compatible with the Standard Model prediction and has a precision better than 20%.

The decay K+ → π+ ν ν̄ is called a “golden” mode in flavor physics because it is both extremely rare and theoretically clean. The Standard Model predicts its branching ratio at the level of 10⁻¹⁰ with small intrinsic theory uncertainty (about 3%), while the remaining uncertainty comes mostly from the values of parameters in the quark-mixing (CKM) matrix. Because the process is sensitive to heavy virtual particles, a precise measurement can test models of new physics up to mass scales around 100 TeV.

NA62 studies this decay using a beam of fast kaons that decay in flight. A 75 GeV secondary beam, produced by 400 GeV protons hitting a beryllium target, contains about 6% K+. The experiment combines several detectors: a kaon tagger (KTAG), a silicon beam spectrometer (GTK), a ring‑imaging Cherenkov detector (RICH) for particle ID, electromagnetic and hadronic calorimeters (including a liquid krypton calorimeter, LKr), and a set of photon veto detectors (LAV, IRC, SAC). Since 2023 the experiment lowered the beam intensity to about 75% of the design maximum and replaced the KTAG gas to reduce material in the beam. Timing is excellent (around 100 picoseconds) and particle identification suppresses muons being misidentified as pions by about a factor of 10⁷, while the photon veto system rejects neutral‑pion backgrounds by about a factor of 10⁸.

The analysis identifies signal candidates by measuring the kaon and pion momenta and computing the “missing mass squared,” which isolates the cases where unseen neutrinos carry away the rest of the energy. NA62 uses the common K+ → π+ π0 decay as a normalization channel to count how many kaons were effectively observed. The 2023–2024 dataset alone is expected to contain about 22.9 ± 1.1 Standard Model K+ → π+ ν ν̄ events, with a total expected background of 11.9 +2.9 −2.3 events. The largest single background source in that dataset comes from decays or interactions upstream of the fiducial decay region.