Back to top Back to top

About the Stereo Experiment

Light Sterile Neutrinos

During the last decades, numerous experiments have proved neutrinos to be massive particles through the observation of a phenomenon called neutrino oscillations. Neutrino flavour eigenstates can be described as a superposition of a set of mass eigenstates that evolve with different complex phases (see Figure 1). This generates a lepton content change in a given initial neutrino state as it propagates through space, thus generating an effective change of flavour.

Neutrinomischung Figure 1: Pictorical representation of flavour mixing of neutrinos. The different combinations of flavour give raise to different mass eigenstates (and vice versa). To the left, normal hierarchy scheme; to the right, inverted hierarchy scheme. © M. Wurm, JGU

Although neutrino oscillation is a well understood phenomenon, some experimental observations are currently challenging the theory. The most prominent is known as the reactor antineutrino anomaly (RAA). In 2011, the re-evaluation of the νe flux in combination with most up to date corrections and nuclear data led to a higher expected flux of νe from nuclear reactors [1]. The re-analysis of previous reactor neutrino experiments resulted in a deficit of 2.7 σ significance (see Figure 2). Two other well-known experimental anomalies in the field are: the LSND anomaly with a νe short-baseline appearance in a νμ beam [2] and the gallium neutrino anomaly with a short-baseline disappearance of νe during the calibration process of GALLEX [3] and SAGE [4] experiments using chromium and argon radioactive sources [3].

Reactor Antineutrino Anomaly Figure 2: Electron antineutrino survival probability considering 3 active neutrino flavours (dashed line) and the addition of a sterile neutrino (solid line). Points correspond to the observed neutrino fluxes for different experiments with reference to the expected value at their distance to the reactor core. © T. Lasserre, CEA

These anomalies may reveal that the current understanding of neutrino oscillations is incomplete. However, additional oscillations to other active light neutrino states would be in disagreement with the measurement of the Z boson decay width performed at LEP [5]. For that reason, an oscillation into a non-weakly interacting light neutrino state has become a popular explanation for the aforementioned anomalies. In that case, the survival probability for a νe with an energy Eνe after a propagation over a distance L can be approximated by

P (νeνe) = 1 - sin2(2θnew) sin2 (1.27 Δm2new L ⁄ E νe [eV]2 [m] ⁄ [MeV]) (1)

The addition of a sterile neutrino can be worked out within the framework of one of the most popular mechanisms to generate neutrino masses, the type I seesaw. This mechanism introduces a set of three right-handed neutrinos to match with the three active ones and by adding a minimal extension it could explain the existence of a sterile state [6]. In such a framework, neutrino masses would be suppressed via the seesaw mechanism, provided that right-handed neutrinos exist within the electroweak mass scale. Thus, an eV-scale sterile neutrino together with sizable active-sterile mixing is accommodated without the need of artificially inserting small mass scales or Yukawa couplings.

The best fit parameters for this new oscillation able to match the anomalies listed above are sin2(2θnew) ∼ 0.10 and Δm2new ∼ 1 eV2 [7], which favours the existence of light sterile neutrinos at the eV scale. The Stereo experiment aims to test such a hypothesis [8] by measuring with high precision the evolution of the νe energy spectrum at multiple baselines from the source.

The Stereo Detector and its Detection Principle

Stereo Detektor Figure 3: Schematic of the design of the Stereo detector. © L. Scola, CEA

In order to probe the existence of light sterile neutrinos, the strategy of Stereo is to measure with high precision the evolution of the antineutrino energy spectrum at short distances. This is the reason why the detector is centred at about 10 m away from the compact 58 MW research reactor at the Institut Laue-Langevin (ILL) in Grenoble, France.

Figure 4: Oscillated neutrino spectra at the closest (blue) and the furthest (red) cell from the core. In black the non-oscillated spectrum is shown. © D. Lhuillier, CEA

The pattern arising from active to sterile neutrino oscillations depends both on the neutrino energy and the distance travelled from the core to the detection vertex (Eq. 1). The distance between the oscillation extrema is about 2 m for sterile reactor neutrinos. Therefore, the 2.2 m long inner detector is segmented in six identical target cells accounting for six different baselines oriented along the neutrino travel path (see Figure 3). Under this setup, Stereo can measure the relative distortions of the neutrino energy spectrum and thus detect the oscillations (see Figure 4).

The antineutrinos emitted by the reactor at the ILL are observed by the Stereo detector via Inverse Beta Decay (IBD)

νe + p → e+ + n (2)

The experimental signature of the IBD is based on a delayed coincidence between the positron signal and the neutron capture (see Figure 5). For this reason, the six target cells composing the neutrino target are filled with organic liquid scintillator (LS) doped with a gadolinium (Gd) based complex to enhance the neutron capture cross-section.

Inverse Beta Decay
Figure 5: Neutrino detection process in Stereo. The incident neutrino interacts with a hydrogen nucleus of the liquid scintillator to form a positron (e+) and a neutron (n) which, in most cases, will be captured by a gadolinium nucleus. © D. Lhuillier, CEA

On the one hand, the positron ionises the medium right before it annihilates and releases two gammas. This process happens very rapidly and constitutes the prompt event. The energy spectra of the antineutrino can be extracted from the kinetic energy of the positron

Ee+ = Eνe - 0.8 MeV (3)

On the other hand, the neutron quickly thermalises in the LS and diffuses until it is captured on a nucleus, defining thus the signal with a few μs delay. The capture can occur naturally on hydrogen nuclei emitting a characteristic low-energy γ of 2.2 MeV. Doping the liquid with gadolinium allows to reduce the diffusion time and, most importantly, to detect the neutron capture by an 8 MeV γ-cascade well above the natural background.

The neutrino Target is surrounded by the Gamma-Catcher, a Gd-free LS volume which is used to improve both the energy resolution and the detection efficiency by catching escaping gammas. The produced scintillation light is collected by a set of 48 photomultiplier tubes (PMTs) on top of the detector. These photosensors are immersed in an oil-like liquid on top of an acrylic buffer layer.

As a consequence of the surrounding neutron beam lines, which lead from the reactor to neighbouring experiment installations, the Stereo experiment is exposed to neutrons and gammas. For that reason a heavy shielding made of B4C, lead, steel, iron and borated polyethylene surrounds the detector. In addition, a water-Cherenkov veto is placed above the detector to tag cosmic muons.

Stereo Sensitivity
Figure 6: Exclusion and sensitivity contours of the Stereo experiment after 2 years of data taking, in the parameter space (Δm2new, sin2(2θnew)), covering the reactor anti-neutrino anomaly. © M. Vialat, ILL

The Stereo experiment measures ∼400 neutrinos per day. In 2018, it was able to probe at 90% C.L. a large fraction of the parameter space favoured by the reactor antineutrino anomaly [9]. Exploiting a factor 2.7 enlarged dataset, the Stereo experiment was able to further increase the accessible parameter space in 2019 [10] (Figure 6). The detector is taking data since November 2016 and currently continues to increase statistics.


[1] P. Huber, Phys. Rev. C84, 024617 (2011)
[2] LSND Collaboration, Phys.Rev.D 64 112007 (2001)
[3] C. Giunti & M. Laveder, Phys.Rev. C83, 065504 (2011)
[4] SAGE, J. N. Abdurashitov et al., Phys. Rev. C73, 045805 (2006)
[5] The ALEPH Collaboration, Physics Reports 427 (2006) 257
[6] K. N. Abazajian et al., arXiv:1204.5379 [hep-ph]
[7] Stereo proposal (2013)
[8] C. Giunti et al., Phys. Rev. D88, 073008 (2013)
[9] H. Almazán et al. (Stereo Collaboration), Phys. Rev. Lett. 121, 161801
[10] H. Almazán et al. (Stereo Collaboration), arxiv:1912.06582