Reactor Neutrinos

A commercial nuclear reactor with 3 GW of thermal power releases about $6\times 10^{20}$ electron antineutrinos per second from the $\beta$-decay of fission fragments. The neutrino luminosity is roughly 6% of the thermal power. This pure, intense, and well-characterized ${\bar{\nu }}_e$ flux has been the foundational tool of neutrino physics since Reines and Cowan’s 1956 experiment at Savannah River.

Spectrum and flux

Reactor antineutrinos come from the cumulative beta decay of many fission-fragment isotopes. Four isotopes dominate: ${}^{235}$U, ${}^{238}$U, ${}^{239}$Pu, and ${}^{241}$Pu. The spectrum extends up to about 10 MeV, with the bulk below 5 MeV. The inverse-beta-decay cross-section scales roughly as $E_{\nu }^2$, so the detected rate peaks near 3–4 MeV.

For decades, the reference flux calculations relied on conversions of electron spectra measured at research reactors. Revised calculations in 2011 (Huber, Mueller) raised the predicted flux by about 6%, producing the apparent reactor antineutrino anomaly — a few-percent deficit of measured rate relative to prediction. The anomaly triggered intensive sterile-neutrino searches and motivated detailed ab-initio modelling. Current understanding attributes most of the apparent deficit to uncertainties in the conversion, and a residual “5-MeV bump” in the measured spectrum is well established but not fully explained.

Short-baseline reactor experiments

In the 2000s, precision measurements of ${\bar{\nu }}_e$ survival at baselines of order 1 km — where the $\Delta m_{31}^2$ oscillation first develops — were used to probe ${\theta }_{13}$: $P({\bar{\nu }}_e\to {\bar{\nu }}_e)\approx 1-{\sin }^{2}2{\theta }_{13}{\sin }^2\left(\frac{\Delta m_{31}^{2}L}{4E}\right)-\text{subleading}$ Daya Bay, RENO, and Double Chooz deployed near and far detectors to cancel reactor flux uncertainties. Daya Bay’s 2012 result established ${\sin }^{2}2{\theta }_{13}=0.092\pm 0.017$, and the current world average has narrowed to ${\sin }^{2}2{\theta }_{13}\approx 0.086\pm 0.004$. A non-zero ${\theta }_{13}$ was the precondition for observable CP violation in the lepton sector.

Long-baseline: KamLAND

KamLAND at the Kamioka site — 1 kt of scintillator, at distances of 138–214 km from Japanese reactors — saw the full development of the $\Delta m_{21}^2$ oscillation and was the first experiment to observe a complete oscillatory pattern in the energy spectrum. Its 2003 and subsequent results resolved the ambiguity among several solutions to the solar neutrino problem in favor of the large-mixing-angle MSW solution.

Medium-baseline: JUNO

JUNO in southern China sits at a 53 km baseline from two new reactor complexes with a combined 35 GW thermal power. This distance is near the first $\Delta m_{21}^2$ oscillation maximum, where the signal is shaped simultaneously by the slow $\Delta m_{21}^2$ modulation and the fast $\Delta m_{31}^2$ modulation. The interference pattern encodes the mass ordering.

JUNO’s 20 kt of scintillator and ~78% photo-coverage target 3% energy resolution at 1 MeV, sufficient to resolve the spectral features. First oscillation-spectrum measurements were released in 2025; full ordering determination is expected over 6 years.

Applied use: remote reactor monitoring

Reactor antineutrinos have been proposed as a verification tool for the nonproliferation of nuclear material. The total rate scales with thermal power, and the ratio of ${}^{239}$Pu to ${}^{235}$U fission events can be extracted from the spectrum shape. Small near-field detectors such as PROSPECT, STEREO, Neutrino-4, and SoLid have demonstrated kilogram-scale measurements at short baselines. This is an area of ongoing applied physics development.

Reactor experiments in the CEvNS era

With CEvNS now experimentally established, reactor antineutrinos at sub-MeV energies are becoming a new frontier. At energies below the IBD threshold (1.8 MeV), CEvNS is the only available channel. Several experiments — CONUS, CONNIE, Dresden-II, and RED-100 — are pursuing CEvNS at commercial reactor sites with germanium, silicon CCD, and liquid-xenon detectors. The signal is small (a few events per day per kilogram at ~10 m from the core) and dominated by nuclear recoils in the keV range, but the science reach — weak mixing angle at low energy, non-standard interactions, sterile neutrino searches — makes it one of the most active applied-physics frontiers.