New results, BEST-Experiment, sterile neutrinos?

  • A
  • Thread starter malawi_glenn
  • Start date
  • Tags
  • #1


Science Advisor
Homework Helper
Gold Member
2022 Award
TL;DR Summary
BEST Experiment - gallium anomaly - sterile neutrinos
(not sure if this is more suitable in the beyond standard model and/or this has been discussed previously in a recent time, I did some searching and could not find any)

"Search for electron-neutrino transitions to sterile states in the BEST experiment" got published yesterday

Preprint versions:

Popular science description:

Now the gallium anomaly has also been confirmed by BEST. Theoretical error in the nuclear model calculation or solid discovery? What do you guys think of this?
  • #2
A nuclear process explanation makes more sense to explain most of the results attributed to sterile neutrinos, and I suspect that this one will end up the same way. Previous sterile neutrino hints were mutually inconsistent. A recent paper discussing the reactor anomalies and 5 MeV bump on the basis of nuclear fuel when previous work had attributed to sterile neutrinos is below.

We investigate the possible origins of the reactor antineutrino anomalies in the framework of a summation model (SM) where missing β transitions are simulated by a phenomenological Gamow-Teller β-decay strength model.
We show that the general trends of the discrepancies between the measured antineutrinos energy spectra and the Huber-Mueller model can be reproduced both in norm and shape.
Using the exact electron-antineutrino correspondence of the SM model, we predict similar distortions in the fission-electron spectra, suggesting a norm bias for the 235U ILL electron spectrum as being at the origin of the "Reactor Antineutrino Anomaly" and a shape bias in the measured electron spectra of 235U and Pu isotopes as being at the origin of the "5 MeV bump".
A. Letourneau, et al., "Anomalies in reactor antineutrino spectra in light of a new summation model with parameterized missing transitions" arXiv:2205.14954 (May 30, 2022).

A more complete explanation which also frames the problem is found in the introduction portion of the body text:
The reactor antineutrino anomalies are a several-years long standing problem in neutrino physics. They refer to an observed ∼6% deficit in the detected rate, known as "Reactor Antineutrino Anomaly" (RAA), and a ∼10% excess of event in the 4-6 MeV range, known as the “5- MeV bump", when comparing experimental data to the prediction of the state-of-the-art Huber-Mueller (HM) model. The RAA was first put in evidence by comparing to short baseline reactor experiments, and confirmed by all recent high precision reactor antineutrino experiments at distances of 300-500 m from the reactor and below 30 m. The “5-MeV bump” is observed in all the above-cited high precision reactor antineutrino experiments also with slightly different amplitudes and shapes.

At present, no consensus has been reached concerning the origins of these anomalies. The RAA was first interpreted as the possibility of the existence of a hypothetical sterile neutrino state, mixing with the active electronic flavor. The best fit parameters for this sterile state to absorb the anomaly was found around 1 eV^2 for the oscillation frequency (∆m^2) and 0.14 for the amplitude sin^2(2θ). This best fit region of oscillation parameters is now rejected to high C.L. by several experiments that have tested the sterile neutrino hypothesis in a model independent way.

On the other hand, the Daya Bay and RENO experiments have studied the dependence of the antineutrino yield to the fuel-composition. They concluded that a ∼8% bias of the 235U Inverse Beta Decay (IBD) yield could be solely responsible for the RAA. This result is slightly in tension with experiments at research reactors with pure 235U fuel showing a (5 ± 1.3) % deficit, not allowing to conc [sic] But the hypothesis of a normalization bias on 235U spectrum is reinforced by the recent measurement of the 235U to 239Pu electron energy spectra ratio reporting a constant ∼5% disagreement with respect to the HM prediction.

Regarding the shape anomaly, extensive studies have been conducted to find explanations in the prediction modeling but none of them have succeeded to bring satisfactory solutions.

The Huber-Mueller model is based on an improved method to convert the cumulative β spectra measured at ILL with the BILL spectrometer into antineutrino spectra. In this method, if experimental biases exist on the measured β spectra, they would be transferred to the converted antineutrino spectra and could be at the origins of the anomalies. The method itself is not guaranteed unbiased due to the contribution of first forbidden transitions.

The present contribution proposes to use the exact electron-antineutrino correspondence of a refined summation model (SM) to test the consistency of the electron and antineutrino spectra predicted by the HM model and to search for biases in the original β spectra used as reference to construct the HM model.
The body text of the paper concludes as follows:
In summary, we have presented a phenomenological Gamow-Teller strength model able to simulate β-decay transition-intensities for fission fragments and to correct for Pandemonium effect and missing transitions in the ENSDF database.
Despite the simplicity of the model, the main features and divergences observed in antineutrino experiments compared to the Huber-Mueller model can be reproduced by a summation model with tuned input parameters. It highlights the importance of missing transitions in the modeling of antineutrino fission spectra. Using the exact correspondence between electron and antineutrino in the summation approach, we have seen that equivalent deviations are expected on the electron side.
The conclusions of this study suggest that the reactor antineutrino anomalies could find their origins in a norm bias for the measured 235U spectrum after 12h of irradiation and a shape bias for all measured electron spectra. Although these conclusions are supported by independent measurements, the origin of the biases are still unclear at this stage.
Some biases on the neutron cross sections used to normalize the beta spectra could cover part of the RAA and part of the shape anomaly could be included in the envelope of systematic of the BILL spectrometer efficiency. This work tends to confirm the need for improving the accuracy of β fission spectra both on the experimental and theoretical sides.
The abstract of the new BEST paper is as follows:

Screen Shot 2022-07-12 at 5.23.28 PM.png

No other experimental hints of an oscillation to a sterile neutrino this massive, and with such a large mixing angle have been noticed before in other experiments, which seems unthinkable if the sterile neutrino hypothesis were correct.

The sterile neutrino implied by the RAA data had best fit parameters for this sterile state to absorb the anomaly was found around 1 eV2 for the oscillation frequency (∆m2) (i.e. less than a third what the BEST result estimates) and 0.14 for the amplitude sin2(2θ) (about a third of the mixing angle).

Also note that the upward uncertainty in ∆m2 in their sterile neutrino model is a flabbergasting + ∞.

Taken literally, this means that the BEST experiment believes that there is a 68% chance that their sterile neutrino has a mass eigenstate somewhere between 1.0 eV2 and the mass of the Milky Way galaxy in excess of the most massive of the three active neutrino mass eigenstates (which is experimentally constrained to be between 0.06 eV to 0.90 eV).

What heck does that mean?

I would argue strenuously that you can not simultaneously claim a 4 sigma significance signal and have an infinity sign in one the key error bars expressing your conclusion.

The 95% (two sigma) confidence interval for ∆m2 includes all values from 0 to ∞; likewise, the 95% (two sigma) confidence interval for sin2(2θ) is 0.08 (about 8.2º) to 0.72 (about 29º), with a best fit value of 0.42 (about 20.2º), which is also all over the map.

And, if you have really want to claim a near discovery of something like a sterile neutrino with a given mass and a given mixing angle in a 3+1 neutrino scenario like the one they are modeling, it isn't good enough that your one experiment is consistent with the BEST experiment's best fit to its data. You also have to be able to credibly claim that this is consistent with every other experiment that has ever been modeled in a 3+1 neutrino scenario. Yet, the BEST experiment's best fit numbers are flatly contradicted by every other attempt using different data that has ever been modeled in a 3+1 neutrino scenario, and given the large mixing angle, that isn't something that other neutrino experiments could have missed for lack of precision.

Also, while Gallium and Germanium are hardly "exotic", we have far less operational experience with these elements than with Uranium and Plutonium, which have had full time engineers and scientists at nuclear power plants and in weapons programs all over the world for the last eighty years devoted to understanding their nuclear decay properties and generating massive amounts of data and expertise in the process to model it. So, it would hardly be surprising if the nuclear model for Gallium and Germanium were not quite as accurate as it is for Uranium and Plutonium.


BEST has grossly understated the systemic uncertainty in their nuclear model, and that uncertainty swamps their claimed near discovery of a sterile neutrino.
Last edited:
  • Like
Likes vanhees71 and malawi_glenn

Suggested for: New results, BEST-Experiment, sterile neutrinos?