This is the fourth part of the Sources of Neutrinos series. We turn from natural and reactor sources to the most controlled source available: high-energy neutrino beams produced by particle accelerators.
The basic idea
An accelerator neutrino beam exploits the fact that charged pions and kaons decay to muons and neutrinos. You start with the only ingredient you actually have to provide: a high-energy proton beam. Everything else is downstream of that proton beam interacting with a target.
The beam line typically has five stages:
1. Proton injection. A high-power proton synchrotron produces a tightly bunched beam at energies of tens to hundreds of GeV. Fermilab’s Main Injector runs at 120 GeV; J-PARC’s Main Ring runs at 30 GeV; CERN’s SPS at 400 GeV (used historically for OPERA).
2. Target. The proton beam slams into a long, narrow graphite or beryllium target a few centimeters across. The proton-nucleon collisions produce a hadronic shower dominated by pions and kaons.
3. Focusing. The charged mesons emerge from the target with a range of angles and energies. Two or three magnetic horns — current-carrying conductors shaped like nested cones — produce toroidal magnetic fields that focus mesons of one sign forward and defocus the other. Flipping the horn current selects either positive pions (for a neutrino beam) or negative pions (for an antineutrino beam).
4. Decay tunnel. The focused mesons enter a long evacuated tunnel — typically 200 meters long for modern beams. They decay in flight: $\pi^+ \to \mu^+ + \nu_\mu$ and $K^+ \to \mu^+ + \nu_\mu$ are the dominant processes. The decay products inherit the forward boost of the parent mesons, producing a forward-focused neutrino beam.
5. Absorber and rock. At the end of the decay tunnel, a large absorber stops the surviving hadrons. Beyond it, ~100 meters of rock absorb the muons. Only neutrinos survive — and continue traveling toward the far detector, often hundreds or thousands of kilometers away.
Energy and flavor composition
The resulting beam is predominantly muon neutrinos (or muon antineutrinos in antineutrino-mode running). Typical composition: ~98% $\nu_\mu$, ~1% $\bar\nu_\mu$, ~1% $\nu_e + \bar\nu_e$. The small $\nu_e$ contamination comes mostly from muon decays in the decay tunnel and from kaon three-body decays.
Energy spectrum depends on the proton beam energy and the design. T2K’s beam peaks at about 0.6 GeV (low energy by design). NOvA’s beam peaks at about 2 GeV. DUNE’s beam will peak at 2-3 GeV with a broad spectrum extending to many GeV.
The flavor purity is what makes accelerator beams uniquely powerful: a clean source of $\nu_\mu$ pointed at a detector lets you cleanly measure $\nu_\mu \to \nu_e$ oscillation — the signal channel for CP violation in the lepton sector.
On-axis vs off-axis
You can either point the detector directly at the beam center (on-axis) or offset by a few degrees (off-axis).
On-axis gives a broad spectrum: many energies present, useful for measuring the full oscillation shape. MINOS and MicroBooNE were on-axis.
Off-axis is a clever trick. Pions of different energies decay at different angles, and the kinematics conspire so that an off-axis detector sees a narrower, more nearly monochromatic spectrum. T2K is 2.5° off-axis from the J-PARC beam axis; NOvA is 14.6 mrad off-axis from the NuMI beam axis. Both choose the off-axis angle to put the oscillation maximum exactly at the peak of the neutrino flux.
DUNE will be on-axis (because the experiment wants a wide spectrum for matter-effect-driven mass-ordering), and Hyper-K will continue T2K’s off-axis approach (because Hyper-K’s larger statistics favor precision measurements at a single energy).
The long-baseline experiments
T2K (Tokai to Kamioka, Japan, 295 km baseline). Running since 2010. Producing leading constraints on $\delta_{CP}$, the lepton-sector CP-violating phase.
NOvA (Fermilab to Ash River, Minnesota, 810 km baseline). Running since 2014. Complementary measurement of $\delta_{CP}$ at longer baseline with different matter effects.
MINOS and MINOS+ (Fermilab to Soudan, 735 km). Operated 2005-2016. Established precision atmospheric oscillation parameters and searched for sterile-neutrino disappearance.
K2K (KEK to Super-Kamiokande, 250 km). Operated 1999-2004. The first long-baseline accelerator experiment — confirmed Super-K’s atmospheric oscillation with an accelerator beam.
OPERA (CERN to Gran Sasso, 730 km). Operated 2008-2012, with analysis through 2018. Famous for catching five tau-neutrino appearance events — the first direct observation of $\nu_\mu \to \nu_\tau$ oscillation.
Why long-baseline matters
The 295 km, 810 km, or 1,300 km baselines aren’t arbitrary. They are chosen so that the muon-neutrino-to-electron-neutrino oscillation probability is near its first maximum for the beam energy. That maximum is set by atmospheric oscillation parameters.
The combination of long baseline plus high-energy beam gives access to:
- θ₁₃ appearance signal (now precisely measured by reactor experiments).
- θ₂₃ disappearance from $\nu_\mu \to \nu_\tau$.
- δ_CP through difference between neutrino and antineutrino appearance probabilities.
- Mass ordering through MSW effects on appearance.
DUNE: the next-generation beam
LBNF/DUNE (Long-Baseline Neutrino Facility / Deep Underground Neutrino Experiment) will be the most ambitious accelerator-based neutrino program ever attempted. Fermilab’s existing Main Injector is being upgraded with the new PIP-II proton driver to deliver a 1.2-megawatt beam initially, upgrading to 2.4 MW in the late 2020s. The beam is aimed at a 70-kiloton liquid-argon TPC at Sanford Underground Research Facility in South Dakota, 1,300 km away.
Beam first expected: early 2030s. Full DUNE physics: mid-2030s.
What’s next
Future programs being discussed include:
- Muon Collider (long-term, exploratory): a muon-decay-based “neutrino factory” producing extraordinarily intense, clean neutrino beams. Decades away.
- ESSnuSB: a CERN-based design at the European Spallation Source for long-baseline running. Conceptual.
For the moment, accelerator beams remain a key precision tool. The next part of this series turns to a much rarer but extremely consequential natural source: supernovae.
Frequently asked
How do you make a neutrino beam?
You start with a high-energy proton beam from an accelerator, fire it into a graphite target, and let the proton-nucleon collisions produce pions and kaons. The charged mesons are focused by magnetic horns into a long evacuated decay tunnel. They decay in flight to muons plus neutrinos. The muons are absorbed in rock at the end of the tunnel; the neutrinos continue, forward-focused, toward a distant detector. The horn polarity selects either ν or ν̄ running.
Why focus the beam at all?
Neutrinos can't be directly focused — they have no charge. But the parent pions and kaons are charged, so focusing them with magnetic horns concentrates their decay products into a forward cone. Without focusing, the neutrino flux at a far detector would be reduced by orders of magnitude.
Why use off-axis beams?
Off-axis (e.g., T2K's 2.5° offset from the J-PARC beam direction) produces a narrower, lower-energy neutrino spectrum than on-axis at the same baseline. The narrower spectrum is preferred for precise oscillation measurements because it concentrates the flux at the energy where oscillation effects are maximal. NOvA also runs off-axis.
What are the major accelerator-based neutrino experiments?
T2K (J-PARC to Super-Kamiokande, Japan, 295 km), NOvA (Fermilab to Ash River, Minnesota, 810 km), and historically MINOS, K2K, OPERA. DUNE (Fermilab to Sanford Lab, 1,300 km) is under construction. Hyper-Kamiokande will replace Super-K as the T2K successor experiment's far detector in 2027.
Cite this article 5 formats
APA
Neutrino Times Editorial Team. (2026, February 25). Sources of Neutrinos — Part 4: Accelerator beams and making your own neutrinos. Neutrino Times. https://neutrino-times.com/articles/sources-of-neutrinos-part-4-accelerator/
Chicago
Neutrino Times Editorial Team. "Sources of Neutrinos — Part 4: Accelerator beams and making your own neutrinos." Neutrino Times, February 25, 2026. https://neutrino-times.com/articles/sources-of-neutrinos-part-4-accelerator/.
MLA
Neutrino Times Editorial Team. "Sources of Neutrinos — Part 4: Accelerator beams and making your own neutrinos." Neutrino Times, 25 Feb. 2026, https://neutrino-times.com/articles/sources-of-neutrinos-part-4-accelerator/.
BibTeX
@misc{neutrino-times-sources-of-neutrinos-part-4-accelerator,
author = {Neutrino Times Editorial Team},
title = {Sources of Neutrinos — Part 4: Accelerator beams and making your own neutrinos},
howpublished = {Neutrino Times},
year = {2026},
month = {feb},
url = {https://neutrino-times.com/articles/sources-of-neutrinos-part-4-accelerator/},
note = {Accessed: 2026-02-25}
} RIS
TY - GEN TI - Sources of Neutrinos — Part 4: Accelerator beams and making your own neutrinos AU - Neutrino Times Editorial Team PY - 2026 DA - 2026-02-25 PB - Neutrino Times UR - https://neutrino-times.com/articles/sources-of-neutrinos-part-4-accelerator/ ER -