Ultracold Neutron Bottle
A bowl made of magnets and gravity that catches neutrons walking at four meters per second and times how long they live.
Traps free neutrons at energies below ~300 neV (≈ 10 mK temperature) in a magneto-gravitational trap built from a Halbach array of permanent magnets and gravity. Neutrons that lack enough kinetic energy to climb out of the magnetic potential are confined for storage times of hundreds to thousands of seconds. Counting how many remain after each storage time gives the free-neutron lifetime τ_n with world-leading precision.
Ultracold neutrons are produced at a spallation neutron source (LANSCE at Los Alamos) by cooling moderated neutrons below ~300 neV in a cold deuterium moderator; UCNs travel at ~5 m/s. They are guided into a bowl-shaped trap whose interior is tiled with NdFeB permanent magnets in an asymmetric Halbach octupole geometry, producing a strong surface field that repels low-field-seeking neutron states. Gravity caps the bottle: neutrons without enough energy can't climb out the open top. A small holding field (~1 mT) prevents Majorana spin flips at field zeros. After a chosen storage time, a ¹⁰B-coated ZnS scintillator is lowered into the trap; ¹⁰B + n → ⁷Li + α + 2.31 MeV, the alpha lights up the ZnS, and a PMT or SiPM counts each event. UCNτ collected data over five years (2017–2022) and counted ~40 million surviving UCNs to push uncertainty below 0.3 s.
In plain terms
A free neutron is normally impossible to hold; it carries no charge and passes through most materials. But if you cool a neutron until it's slower than you can walk, it bounces off the gradient of a strong magnet like a marble in a bowl. The UCN bottle is a bowl made of magnets. The experiment counts how many neutrons are still bouncing after a few minutes versus 15 minutes, and the ratio tells you exactly how long a neutron lives before it falls apart. The puzzle: a different experiment that catches the bits a neutron breaks into gives a different answer — and that disagreement might mean there's hidden physics inside the neutron.
Also called: UCN trap, magneto-gravitational trap, magnetic neutron bottle.
Why it matters
- The neutron lifetime sets the axial-vector coupling g_A of the weak nuclear force; 0.3 s precision constrains g_A to ~0.01%.
- Big-Bang nucleosynthesis predictions for primordial ⁴He are limited at the ~0.1% level by τ_n; sharpening τ_n sharpens cosmology.
- If the long-standing 'bottle vs. beam' lifetime discrepancy is real, it could indicate neutrons decaying into an unknown dark-sector particle — a rare experimental window onto dark matter that doesn't require a particle accelerator.
- The same UCN bottles host the world's best neutron electric-dipole-moment searches — sensitive probes of CP violation beyond the Standard Model.
Where it lives
- UCNτ at Los Alamos Neutron Science Center (LANSCE), New Mexico — bathtub-shaped Halbach trap; world's most precise bottle measurement
- PanEDM / nEDM at PSI, Switzerland — UCN storage for neutron electric-dipole-moment searches; also measures τ_n
- HOPE magnetic trap at NIST — alternative Halbach octupole UCN trap
- Institute Laue-Langevin (ILL), Grenoble — world's highest UCN density source
Key numbers
- UCNτ best result (2025): τ_n = 877.83 ± 0.29 s (5-year LANL campaign)
- Beam-method world average: ~888.1 ± 2.0 s (Source of bottle-beam tension)
- Bottle–beam discrepancy: ~10 s (≈5σ) (Open puzzle in particle physics)
- UCN energy: <300 neV (≈10 mK; speed ~5 m/s)
- Halbach magnets in UCNτ: 5,310 NdFeB tiles (Bowl volume ~670 L)
Surprising facts
- An ultracold neutron moves at roughly 5 m/s — about walking pace — slow enough that gravity does meaningful work on it, which is why gravity itself caps the trap.
- The most recent UCNτ result is τ_n = 877.83 ± 0.29 s — the most precise single measurement of the free-neutron lifetime ever reported.
- Bottle measurements (~877–878 s) and beam measurements (~888 s) of the same quantity disagree by ~10 seconds, a roughly 5σ tension that has persisted and deepened for two decades.
- If that 5σ gap really came from neutrons decaying into a dark-sector particle, roughly 1% of all neutrons in the early universe would have leaked silently into dark matter rather than decaying into protons.
Open questions
- Bottle–beam neutron-lifetime discrepancy (~10 s, ~5σ): The most precise bottle measurement (UCNτ 2025: 877.83 ± 0.29 s) and the beam-method world average (~888.1 ± 2.0 s) differ by ~10 seconds, far outside combined error bars. This tension has persisted and grown for two decades. Proposed explanations include a dark-sector neutron decay channel, neutron excited-state effects, and unknown systematics in one method. There is currently no consensus, and the speculation about a dark decay channel is exactly that — speculation, not evidence.
A short history
Zeldovich (1959) and Fermi proposed UCN production and trapping. First UCN observations were made by Lushchikov (1969) and Shapiro. Magnetic trapping of UCN was demonstrated in the 1990s. The UCNτ collaboration at Los Alamos, led by Geoffrey Greene and then Robert Picker, ran its first major campaigns in 2017–2018 and published a precision result in PRL 2021 (τ_n = 877.75 ± 0.28 s). Updated analysis of five years of data published in July 2025 gave τ_n = 877.83 ± 0.29 s — the most precise single measurement of the free-neutron lifetime to date.
Sources
- LANL (2025) — UCNτ world-record precision (https://www.lanl.gov/media/news/0731-free-neutron-record)
- PRL (2021) — UCNτ improved measurement (https://link.aps.org/doi/10.1103/PhysRevLett.127.162501)
- CERN Courier (2026) — bottle-beam debate at PSI (https://cerncourier.com/the-beam-bottle-debate-at-psi/)
- Phys. Rev. C (2014) — UCNτ Halbach trap design (https://link.aps.org/doi/10.1103/PhysRevC.89.052501)
- arXiv (2015) — Halbach UCN trap comparison (https://arxiv.org/abs/1503.05226)