85
Part Two · Quantum fields and particles

Neutrino physics and oscillations

From 1968, in a tank of 600 tonnes of dry-cleaning fluid a mile underground in a South Dakota gold mine, Ray Davis counted argon atoms made by neutrinos from the Sun: a few a week. He found about a third as many as the Sun should make. For thirty years the gap was blamed on the solar model or the experiment. It was neither. Neutrinos change type as they travel, from the kind Davis could detect into kinds he couldn’t. And a particle can change only if it has mass, which the Standard Model said neutrinos did not.

On the plate: the chance that a muon neutrino is still a muon neutrino, against the distance it has flown.

Two flavours, two masses

The neutrinos made in weak interactions, νe,νμ,ντ\nu_e, \nu_\mu, \nu_\tau, are not the states of definite mass ν1,ν2,ν3\nu_1, \nu_2, \nu_3; they are mixtures, like the quarks of Chapter 83. With two flavours and mixing angle θ\theta, the mass states travel with slightly different phases (Chapter 31’s beats), and the flavour oscillates:

P(να→νβ)=sin⁡22θ sin⁡2 ⁣(1.267 Δm2LE)P(\nu_\alpha\to\nu_\beta) = \sin^22\theta\,\sin^2\!\left(1.267\,\frac{\Delta m^2L}{E}\right)

with Δm2\Delta m^2 in eV², LL in km and EE in GeV. Set the baseline to 295 km and the energy to 0.60 GeV, with Δm2=2.5×10−3\Delta m^2 = 2.5\times10^{-3} eV². T2K in Japan uses exactly these numbers, 295 km and 0.6 GeV, to sit at the first oscillation maximum, at L/E≈496L/E \approx 496 km/GeV. Established

The discoveries

In 1998 Super-Kamiokande, a tank of 50,000 tonnes of water in Japan, found that muon neutrinos made by cosmic rays in the atmosphere arrived depleted when they came up through the Earth, 13,000 km, but not when they came down from overhead, 15 km. In 2002 the Sudbury Neutrino Observatory in Canada, using heavy water, measured both the electron neutrinos from the Sun and the total of all flavours. The total matched the solar model exactly; the electron part was about a third. Nothing was missing; it had changed flavour. Davis and Masatoshi Koshiba shared the 2002 Nobel Prize; Takaaki Kajita and Arthur McDonald the 2015 one. Established

Matter changes the mixing

Electron neutrinos passing through matter scatter off electrons by W exchange, which the other flavours can’t do. That adds an effective potential and changes the mixing. Lincoln Wolfenstein, Stanislav Mikheyev and Alexei Smirnov showed that at a particular density the effective mixing becomes maximal, the MSW resonance, even if the vacuum angle is small. Increase the matter effect to 0.50 (in units of Δm2/2E\Delta m^2/2E). In the Sun’s core, high-energy solar neutrinos pass through this resonance, which is why the fraction surviving depends on energy, as measured by Borexino and SNO.

What we know, and what we don’t

Three mixing angles are measured; JUNO in China reported in 2025 the most precise solar angle yet, sin⁡2θ12=0.309±0.009\sin^2\theta_{12} = 0.309 \pm 0.009. Two mass-squared differences are known. The absolute mass is not: KATRIN, weighing the electron neutrino through the endpoint of tritium beta decay, set an upper limit of 0.45 eV in 2025, and cosmology suggests a sum below about 0.1 eV. Open questions: whether ν3\nu_3 is the heaviest or lightest (the mass ordering), whether neutrinos violate CP, and whether a neutrino is its own antiparticle (Majorana), testable by searching for neutrinoless double beta decay. Debated for each.

What you now have

Flavour and mass states, the two-flavour oscillation formula and the first maximum at 496 km/GeV, the solar and atmospheric discoveries, the MSW matter effect, current measurements, and the open questions of ordering, CP and Majorana nature.

· · ·

Three questions to carry with you.

Why do oscillations prove neutrinos have mass?tap
Oscillation needs mass states with different masses; with all masses zero the phases never differ.
What did SNO measure that settled the solar problem?tap
The total flux of all flavours, which matched the solar model.
What would neutrinoless double beta decay show?tap
That neutrinos are Majorana particles, their own antiparticles, and that lepton number is violated.

And one to work out. For Δm2=2.5×10−3\Delta m^2 = 2.5\times10^{-3} eV², at what L/EL/E (km/GeV) is the first oscillation maximum, where 1.267 Δm2L/E=π/21.267\,\Delta m^2L/E = \pi/2?

L/E =km/GeV

Sources: R. Davis, D. S. Harmer and K. C. Hoffman, Phys. Rev. Lett. 20, 1205 (1968). Y. Fukuda et al. (Super-Kamiokande), Phys. Rev. Lett. 81, 1562 (1998). Q. R. Ahmad et al. (SNO), Phys. Rev. Lett. 89, 011301 (2002). L. Wolfenstein, Phys. Rev. D 17, 2369 (1978); S. P. Mikheyev and A. Yu. Smirnov, Sov. J. Nucl. Phys. 42, 913 (1985). KATRIN Collaboration, Science 388, 180 (2025). JUNO Collaboration, first results (2025 preprint). Checked by code: first maximum 495.9 km/GeV; T2K phase 1.557 ≈ π/2.

Next · Chapter 86
Accelerator physics

Why does LEP lose 3.5 GeV per turn but the LHC only about 6 keV ✓?

Turn the page