CeSOX: An experimental test of the sterile neutrino hypothesis with Borexino

The third phase of the Borexino experiment that’s referred to as SOX is devoted to test the hypothesis of the existence of one (or more) sterile neutrinos at a short baseline (~5–10m). The experimental measurement will be made with artificial sources namely with a 144Ce–144Pr antineutrino source at the first stage (CeSOX) and possibly with a 51Cr neutrino source at the second one. The fixed 144Ce–144Pr sample will be placed beneath the detector in a special pit and the initial activity will be about 100 – 150 kCi. The start of data taking is scheduled for April 2018. The article gives a short description of the preparation for the first stage and shows the expected sensitivity.

For the rst time the accelerator anomaly had been observed in the LSND experiment. An excess of ν e events was registered in a ν µ at ≈ 3.8σ beam [2,3,4,5]. Two similar experiment, KARMEN and MiniBooNE, reported contradictory results. The former didn't nd the excess in the transition ν µ −→ ν e [6] whereas the latter demostrated the excess signals at 2.8σ and 3.4σ The gallium anomaly or ν e disappearance had been revealed in two radiochemical solar neutrino experiments SAGE and GALLEX during processing of calibration data that were acquired with 51 Cr and 36 Ar articial radioactive sources [9,10,11,12,13,14]. It is reported in [15] that the signicance of the anomaly is ≈ 2.9σ.
After revaluation of reactor antineutrino spectra [16, 17] a decit of counting rate was noticed in almost all reactor neutrino experiments [18] and a recent calculation [15] indicates the eect at ≈ 2.8σ. However the reactor anomaly is strongly weakened by the results of Daya Bay [19].

PF gey overview
The third phase of the Borexino experiment or CeSOX (hort-distance yscillations with Boreino and a gerium source) [20] is dedicated to the search for sterile neutrinos by means of a compact articial antineutrino source 144 Ce− 144 Pr [21] with characteristic dimensions of 15 cm. The Borexino detector is a large ultra low background liquid scintillator detector. The Borexino target mass for CeSOX equals ∼ 240 t and the ducial volume has a radius of 4 m. Along with that the spatial resolution is about 10 cm/ E(MeV). It's planned to place the source right beneath the detector center in a special pit. Such conguration of the experiment will allow to make a measurement at distances 4.5 12.5 m from the source. Thus one will be observed not only a decit of counting rate as it occurs in a standard neutrino disappearance experiment but a spectrum distortion with distance (so-called oscillation waves) as well. This complex approach will provide a clear experimental 2D 1 or 3D 2 pattern [21] in case of the existence of sterile neutrinos with ∆m 2 14 ∈ (0.5, 5.0) eV 2 (3+1 model, NH). A few characteristic 2D patterns obtained from the Monte Carlo simulation are shown in Figure 1.
The inverse β-decay reaction, ν e + p −→ e + + n, is applied for antineutrino detection. The process has a clear signature consisting of two consecutive events with a delay of 258.7 ± 0.8(stat) ± 2.0(sys) µs [23] between them. The former is a prompt event and it's a electron-positron annihilation. The released energy is directly related to the ν e energy as E prompt = E νe − 0.789 MeV. The latter is a delayed event and this is a 2.22 MeV de-excitation γ-ray that appears as a result of a neutron capture on a proton (on a 1 H nucleus).
The inverse β-decay has one signicant disadvantage consisting in the presence of an energy threshold E thr = 1.806 MeV. For selection a radioactive source it means one need to nd a source with the β-decay energy of a few MeV and the life-time of more than a year simultaneously. To satisfy the requirements a two-component antineutrino source was chosen. It consists of a longlived nuclide with a low β-decay energy ( 144 Ce, Q = 318 keV, T 1/2 = 285 days) and a daughter nuclide with a high β-decay energy ( 144 Pr, Q 1 = 2996 keV, Q 2 = 2301 keV, T 1/2 = 17.3 min). Thus only 144 Pr component will be observed in the CeSOX experiment. Herewith two 144 Pr decay branches will be measured namely a non-unique rst-forbidden transition 0 − −→ 0 + (97.9%) with endpoint energy 2996 keV and a unique rst-forbidden transition 0 − −→ 2 + (1.0%) with endpoint energy 2301 keV.
The 144 Ce− 144 Pr source has an activity of (3.7 − 5.5) · 10 15 Bq (100 − 150 kCi). Taking into account the half-life of the source the CeSOX will take data about 1.5 yr. During this period 10 4 events will be acquired. All measurements will be mostly background free. Based on the data of the geo-neutrino study with Borexino [24] it can be argued the background equils ∼ 15 ev/yr and it is negligible for CeSOX.

pigure PF
The expected sensitivity of the SOX experiment to the sterile neutrino oscillation parameters for the 3+1 scenario. The possible results are shown taking into account the following systematic uncertainties: a total uncertainty of the normalization rate of 1.5% and an absolute error of 0.03 on the 144 Pr electron spectrum shape factor b. The anomalies are taken from [22].

QF oure prodution nd trnsporttion
The 144 Ce− 144 Pr antineutrino source is made by extracting cerium from exhausted nuclear fuel. 144 Ce is produced in the form of a chemical compound CeO 2 (powder). Due to the cerium β-decay the compound transforms into Pr 2 O 3 with the release of oxygen O 2 . A special stainless steel capsule was made to withstand high temperature (500 • C) and pressure (up to 6 bar). All additional radioactive backgrounds of the source are almost completely suppressed with a thick tungsten container (minimum thickness 19 cm).
The source will be manufactured by the PA ¾Mayak¿ company in Russia and delivered to Gran Sasso in April 2018. During transportation the source capsule and shielding will be inside an extra container (TN MTR). The root to Gran Sasso includes a way to St. Petersbourg (Russia) by train, then to Le Harve (France) by ship, to Saclay (France) by truck and nally to Gran Sasso (Italy) by truck as well. It's expected that the transportation will take three weeks and the source will lose ∼ 5% of its initial activity.

RF ensitivity nd fetures of the gey experiment
As it is mentioned above the main idea of the experiment is an observation of the 2D or 3D oscillation pattern. The sensitivity of CeSOX is shown in Figure 2. It's clearly seen that the experiment may exclude a large part of the region of the allowed parameters and actually demonstrate the failure of the hypothesis of the sterile neutrino existence at ∼ 3σ. source. This dependence is expressed in the counting rate N (E ν , L, t) where A(t) is an activity of the 144 Ce-144 Pr source, S ν (E ν , b) is a shape of the 144 Pr spectrum, P (t) is a thermal power of the source, < E(b) > is a mean energy per decay. The b factor is a parameter of the weak nite-size correction C(Z, W ) in case of the widespread parameterization: C(Z, W ) ≡ 1+a·W +b/W +c·W 2 , where a, b and c are parameters, Z is a charge of the nucleus, W = E e /m e + 1 is a total energy of the β − particle in units of the electron rest mass. There are two physical quantities, b and P (t), whose inuence on the sensitivity of the measurements is crucial. Figures 3 and 4 illustrate that fact.
pigure QF The impact of the calorimetric measurement uncertainty on the sensitivity of the SOX experiment.
pigure RF The impact of the uncertainty related to the spectral shape of the main 144 Pr β − -decay branch on the sensitivity of the SOX experiment. All curves correspond to 95% CL, unless otherwise explicitly stated.
The thermal power will be measured with two independent calorimeters immediately after delivery of the source to Gran Sasso and also instantly after the end of data-taking. For the precision knowledge of the 144 Pr shape factor ve spectroscopic experiments are ongoing within the Borexino collaboration. The previous shape factor measurements dier by 10%.
Among other factors aecting the sensitivity it can be mentioned the following: spatial and energy uncertainties, the precise knowledge of the Inner Vessel shape, the Monte Carlo simulation quality and the eciency of the selection cuts. To minimize the uncertainties introduced by these factors the new comprehensive calibration campaign is scheduled for January and February 2018. A lot of dierent radioactive sources will be applied: 241 Am-9 Be+Ni (neutrons), 68 Ga-68 Ge (positrons), 40 K (γ), 54 Mn (γ), 65 Zn (γ), 85 Sr (γ) and 222 Rn+ 14 C (α, β and γ) as well.

SF gonlusion
In two years CeSOX may exclude the most part of the region of the allowed parameters. If the oscillation pattern is observed additional measurements with a neutrino source 51 Cr might

eknowledgments
The author acknowledges the continual support to the SOX project coming from INFN and the Laboratori Nazionali del Gran Sasso. A special thank goes also to CEA-Saclay for their crucial role in the development of the 144 Ce− 144 Pr antineutrino source, and to all members of the Borexino collaboration for their active role in the development of the project. The SOX project is funded by the European Research Council project ERC-Adv 320873, P.I. Marco Pallavicini. The research was partly supported by the grants of the Russian Foundation for Basic Research (project 16-29-13014) and the Russian Science Foundation (project 17-12-01009). eferenes