Neutrino CP violation: what comparing neutrinos with antineutrinos can show

CP compares matched probabilities

The CP question compares transition probabilities, not raw counts of selected events. Charge conjugation swaps particles and antiparticles; parity reverses spatial coordinates. In neutrino oscillation, CP symmetry links a flavor transition to the matching antineutrino transition under the same conditions.

With three mixed active neutrinos, the usual oscillation matrix can contain a Dirac CP phase. In the standard convention, a phase of zero or half a full turn conserves CP. Any other value can produce an asymmetry in vacuum, provided the relevant mixing angles and mass splittings are nonzero.

The asymmetry comes from interference. The amplitudes for the different mass states pick up different phases in flight and then add up in the detected flavor. Changing the complex mixing coefficients changes how they interfere. No single event carries a label saying that CP symmetry was broken.

Sources: [1] · [2]

Matter in the Earth adds its own asymmetry

Reversing the beam polarity swaps neutrinos for antineutrinos, but the Earth does not turn into an anti-Earth. Ordinary matter contains electrons, and electron-flavor neutrinos pick up an extra contribution from charged-current forward scattering. For antineutrinos the same term has the opposite sign.

Matter alone can make neutrino and antineutrino appearance probabilities differ, even if the vacuum phase conserves CP. How large the effect is, and how it varies with energy, depends on the path, the density and the mass ordering. Calling every asymmetry seen on Earth intrinsic CP violation would mistake an effect of the medium for a property of the mixing matrix.

An analysis therefore fits travel through matter together with the source and detector models. Different baselines and spectra help tell the explanations apart. Disappearance data and reactor constraints also matter, because uncertain mixing parameters can mimic part of a change in appearance.

Sources: [1] · [2] · [3]

An experiment compares reconstructed samples

The horns favor parents of one charge, but the beam is never perfectly pure. Neutrinos and antineutrinos also have different cross sections, and event selection shifts the mix that is kept. These effects must be predicted before any difference is attributed to what happened in flight.

A fit in energy bins links flux, interactions, reconstruction and oscillation to the observed samples. Backgrounds, wrong-sign components and events reconstructed at the wrong energy all enter that fit. Near detectors constrain the model but do not directly measure the oscillation phase at the far site.

The same data may fit several combinations of phase, ordering and mixing parameters. A reported interval also depends on the statistical method and the external constraints used. Bayesian credible regions and frequentist confidence regions answer different questions, so each should keep its original label.

Sources: [2] · [4] · [3]

The CP result depends on the mass ordering

The first joint oscillation analysis by T2K and NOvA was published in October 2025. It combined specific samples that had already been published, not every event either experiment may have recorded since. Fitting the two together was a major step: two accelerator programs compared within one statistical framework.

The analysis found no strong preference for either mass ordering. If the ordering is inverted, it excluded CP-conserving phase values at the reported three-standard-deviation credible level. If the ordering is normal, the allowed region included CP-conserving values. The CP statement cannot be quoted without that condition.

So this is a conditional exclusion, not an unconditional discovery of leptonic CP violation. Earlier T2K limits on the phase and forecasts for DUNE or Hyper-K each come with their own data and assumptions. A promising preference, a conditional exclusion and a completed discovery are different claims.

Sources: [2] · [4] · [5] · [6]

From CP violation to the matter in the Universe takes extra steps

The Universe’s excess of matter motivates the search, but an oscillation phase does not by itself produce matter. Leptogenesis models first create an imbalance of leptons, and processes in the early Universe can convert part of it into an imbalance of baryons.

The simplest standard thermal scenarios need heavy states, lepton-number violation and departures from equilibrium. The masses and phases of those heavy states are generally not fixed by the low-energy oscillation phase. Linking the two requires a specific model and further assumptions.

A measured CP asymmetry would tell us something important about how flavors evolve. On its own, it would not show how the excess of matter arose or decide whether neutrinos are Majorana particles. Those are related questions that need other measurements; they do not follow automatically from the same fit.

Sources: [7] · [1] · [8]

Try it in the Lab

Primary sources & revision

1. Particle Data Group · Neutrino Masses, Mixing, and Oscillations (2024)
2. T2K & NOvA · Joint neutrino oscillation analysis (22 October 2025)
3. NOvA Collaboration · Bayesian constraints on oscillation parameters (2023/2024)
4. T2K Collaboration · Constraint on matter–antimatter symmetry violation (2020)
5. DUNE Collaboration · Long-baseline oscillation physics potential (2020)
6. Hyper-Kamiokande Collaboration · Design report (2018)
7. Davidson, Nardi & Nir · Leptogenesis (2008)
8. Dvali, Maiezza, Senjanović & Tello · Neutrino mass versus new physics (2023)

First published 2026-10-11; last revised 2026-10-13. Original explanatory text and diagrams by Sandbox Physics. Illustrations are schematic; no experimental event records are reproduced here. This is an independent educational publication, not an official Nobel or experiment collaboration publication.