Neutrino Coherent Scattering Detector

A predicted reaction stayed invisible for 43 years; in 2017, a 14-kg crystal finally caught it.

Coherent elastic neutrino-nucleus scattering (CEνNS) is the largest Standard-Model neutrino interaction below ~50 MeV: a neutrino scatters off an entire nucleus, with cross-section enhanced by N² (N = neutron number) because all neutrons contribute coherently. The recoil is a nuclear shudder of 0.1–10 keV. COHERENT (ORNL Spallation Neutron Source) achieved first detection in CsI[Na] in 2017 (6.7σ); subsequent detections used liquid argon and germanium. CONUS+ at Leibstadt (2025) made the first reactor-CEνNS detection in HPGe at 3.7σ. The technique is now a precision probe of the weak mixing angle, sterile neutrinos, and non-standard interactions.

For a 30 MeV neutrino on cesium, the maximum nuclear-recoil energy is T_max = 2E_ν²/M ≈ 0.6 keV — barely detectable. Two technologies are used. (1) CsI[Na] scintillator: a ~14.6 kg crystal with photomultiplier readout and pulse-shape discrimination separates nuclear-recoil-like events from gamma backgrounds; the pulsed SNS proton beam structure enables time-of-flight rejection of beam-uncorrelated events. (2) High-purity germanium (HPGe) diodes at LN₂ temperature with point-contact geometry reach ~160 eVee energy threshold — the lowest in any reactor-CEνNS experiment. The quenching factor Q_f (ratio of measured electron-equivalent to true nuclear-recoil energy) is ~0.2–0.4 in Ge, meaning a 1 keVnr recoil produces only 200–400 eVee — at the edge of electronic noise.

In plain terms

Imagine throwing a ping-pong ball at a bowling ball — the bowling ball barely moves, but it does move. Now imagine the ping-pong ball is so light it goes through almost everything, and it bounces off the whole bowling ball at once, not one pin in it. Because every pin pushes back together, the bowling ball gets a much bigger nudge than from any single pin. That's coherent neutrino scattering. The nudge is so tiny that the detector has to be cooled, shielded under fifty meters of rock, and listened to for years before anyone hears it.

Also called: CEνNS detector, COHERENT detector, coherent elastic neutrino-nucleus scattering sensor.

Why it matters

First experimental confirmation of a 1974 Standard-Model prediction (Freedman) — closing a 43-year gap between theory and observation.
Most direct probe of the weak nuclear charge and a sensitive constraint on non-standard neutrino interactions.
Compact reactor monitors could provide real-time non-proliferation verification of fissile inventory without intrusive measurements.
A future tonne-scale CEνNS array would detect ~1,000 events from a galactic-supernova neutrino burst — the highest-statistics terrestrial supernova signal proposed.

Where it lives

COHERENT at SNS (ORNL) — CsI[Na], LAr, HPGe
CONUS+ at Leibstadt nuclear plant — HPGe (3.7σ first reactor detection, 2025)
RICOCHET at ILL Grenoble — bolometers
RED-100 — liquid xenon (Russia)

Key numbers

First detection (COHERENT CsI 2017): 6.7σ (308 kg·day exposure)
First reactor detection (CONUS+ 2025): 3.7σ (HPGe, 21 m from 3.1 GW reactor)
Threshold (CONUS+ HPGe): ~160 eVee (Lowest for reactor CEνNS)
CEνNS cross-section (Cs, 30 MeV): ~1.8 × 10⁻³⁹ cm²
Original prediction: Freedman, 1974

Surprising facts

The CEνNS cross-section on cesium is enhanced by N² ≈ 6,100 over single-nucleon scattering — yet the reaction was technically invisible for 43 years because the resulting few-hundred-eVee recoil sits below the noise floor of any room-temperature detector.
The COHERENT CsI detector that made the first detection weighs only 14.6 kg — less than a sack of potatoes — and was sited 19.3 m from the SNS mercury target.
CONUS+ detects reactor antineutrinos with energies too low to produce any charged particle — only a silicon nucleus quivering by a fraction of a nanometer — at 3.7σ in a 1.1-kg crystal.

Open questions

Quenching factor uncertainty in germanium: The Ge quenching factor at sub-keV nuclear-recoil energies is poorly constrained — predictions from Lindhard theory and direct measurements differ by 20–50 %. This propagates directly into the inferred CEνNS cross-section and complicates reconciling SNS and reactor results.
Possible spectral shape deviations: COHERENT CsI spectral analyses show modest deviations from Standard-Model predictions that could indicate nuclear-form-factor effects or non-standard interactions. Current statistics do not resolve the question; a longer COHERENT run is in progress.

A short history

Daniel Freedman predicted CEνNS in Phys. Rev. D in 1974, noting it as the largest neutrino cross-section at low energy. Decades of effort to observe it failed because the keV-scale nuclear recoils sat below detector thresholds. The COHERENT collaboration at ORNL, exploiting the SNS's intense pulsed neutrino flux from pion/muon decay-at-rest, made the first detection in CsI[Na] in 2017 (Akimov et al., Science). Subsequent detections in liquid argon (2020) and germanium (2025), and the first reactor detection by CONUS+ in 2025, have established CEνNS as a precision tool.

Sources

COHERENT — Akimov et al., Science 357, 1123 (2017) (https://www.science.org/doi/10.1126/science.aao0990)
Freedman, Phys. Rev. D 9, 1389 (1974) (https://link.aps.org/doi/10.1103/PhysRevD.9.1389)
CONUS+ first reactor CEνNS, Nature (2025) (https://www.nature.com/articles/s41586-025-09322-2)
COHERENT germanium detection, PRL 134 (2025) (https://link.aps.org/doi/10.1103/PhysRevLett.134.231801)