c Pleiades Publishing, Inc., 2006. ISSN 1063-7788, Physics of Atomic Nuclei, 2006, Vol. 69, No. 12, pp. 2096–2100.


Proceedings of the 5th International Conference on NONACCELERATOR NEW PHYSICS Double-Beta Decay and Rare Processes

The SuperNEMO Project* F. Piquemal** (on behalf of the NEMO Collaboration) CEN Bordeaux-Gradignan CNRS/IN2P3 and Universite´ Bordeaux I, France Received November 23, 2005

Abstract—Neutrinoless double-beta decay (ββ(0ν)) is the most sensitive process in the search for leptonic number violation and its discovery would prove that the neutrino is a Majorana particle. From the experience of the NEMO-3 detector construction and data analysis, the NEMO Collaboration proposes a three-year R&D program in order to design a detector (SuperNEMO) sensitive to a ββ(0ν) period of few 1026 yr coupling track reconstruction and calorimeter. PACS numbers : 29.30.-h DOI: 10.1134/S1063778806120131

1. INTRODUCTION Neutrinoless double-beta decay (ββ(0ν)) measurement, which is forbidden by the Standard Model owing to lepton number violation by two units ((A, Z) → (A, Z + 2) + 2e− ), would probe the Majorana nature of neutrinos (neutrino and antineutrino are the same particle). Experimental signature of such a process is the observation of two electrons, for which the total kinetic energy sum is exactly the transition energy and measurable physical observable is the period of this process. Different experimental techniques can be used. First, there are experiments using a pure calorimeter, as Ge semiconductor or bolometer detectors, for which a ββ source is used as the detector and the only measured quantity is the total deposited energy. Second are “tracko–calo” detectors, which not only measure the energy but also use tracking reconstruction techniques to directly identify the two emitted electrons. The “tracko–calo” experiments have less good energy resolution compared to a pure calorimeter but are able to identify the two electron emissions and thus to reject backgrounds with greater efficiency than calorimeters do. These experiments also measure individual energy for each of the two electrons and their angular distribution, which allows one to distinguish ββ(0ν) decay with light-neutrino exchange or with right current interaction. And finally, ∗ **

The text was submitted by the author in English. E-mail: [email protected]

using a detector independent from the source foil allows one to measure different ββ isotopes. Using the experience acquired with the series of NEMO detectors, the NEMO Collaboration has started a 3-yr R&D program to design a detector sensitive to 2 × 1026 yr for 100 kg of 82 Se, corresponding to an effective neutrino mass of 50 meV. Below are described neutrino physics stakes associated with this project, what we learned with the construction and data collection of the NEMO-3 detector, what the R&D programs are to realize, an example of possible geometry for SuperNEMO, and the associated expected performances. 2. NEUTRINO PHYSICS AND DOUBLE-BETA DECAY The neutrino oscillation discovery showed the massive property of neutrino particles and the need for an extension of the Standard Model. Nevertheless, oscillations are only sensitive to the neutrino mass-squared differences. Thus, we do not know the absolute scale of the neutrino mass nor the mass hierarchy (quasi-degenerate, normal hierarchy or inverse hierarchy). Also, numerous properties of neutrino particles are not yet discovered, as its nature (is the neutrino a Dirac or Majorana particle?), the existence of electric or magnetic momentum, right-current interactions or CP violation in the neutrino sector, etc. In most of supersymmetric or grand-unification models, light and heavy Majorana neutrinos are needed, which using a seesaw mechanism allow one to justify the small value of light-neutrino masses (heavy-neutrino masses are of the order of grandunification scale). The consequence of the existence

2096

THE SUPERNEMO PROJECT

of a Majorana would be the nonconservation of the leptonic number, needed condition neutrinos with CP violation to explain matter creation through leptogenesis (by heavy Majorana neutrino decays). Studies of the nature and mass scale of neutrinos can be realized by searching for neutrinoless double-beta decay ββ(0ν) ((A, Z) → (A, Z + 2) + 2e− ). This process may arise if a β transition toward an intermediate daughter nucleus (Z + 1) is forbidden or strongly suppressed. The ββ(0ν) mode, which is forbidden by the Standard Model owing to lepton number violation by two units, involves a vertex changing two neutrons in two protons with the emission of two electrons and nothing else. This process is possible only if neutrinos are Majorana massive particles. The half-life of the ββ(0ν) decay process is connected to the effective neutrino mass by the relation T −1 = G0ν |M0ν |2 mν 2 , where G0ν is a phase space factor proportional to Q5ββ and also function of
Z; M0ν is the nuclear matrix eleUei 2 mi is the effective neutrino ment; and mν  = mass, where Uei are coefficients of the neutrino mixing matrix and mi are mass eigenvalues. The effective neutrino mass is in turn related to the oscillation parameters by the relation [1]     2 Uei mi  = | cos2 θ13 (m1 cos2 θ12 mν  =  + m2 e2iα sin2 θ12 ) + m3 e2iβ sin2 θ13 |, where θij are the mixing angles between i and j eigenstates, and α and β are Majorana phases. Effective neutrino mass measurements should help to resolve the scheme of neutrino mass scale, which is unknown owing to its dependence on the still not yet measured value of the ∆m23 sign. There are other possibilities for virtual particle exchange in the neutrinoless double-beta-decay process: right-handed V + A weak current interaction mediated by the WR boson, Majoron emission (the Majoron is a Goldstone boson which appears in theory after B − L global symmetry breaking), or through R-parity violation in supersymmetric models. The experimental signature of the different ββ modes is the total energy sum of the two emitted electrons. The ββ(0ν)-decay signal should produce an accumulation of events at energy (Qββ ) when the permitted ββ(2ν) process presents a continuous spectrum with energies between 0 and Qββ . The process with a Majoron particle should also present a continuous spectrum but with different shape than the ββ(2ν)-decay one. PHYSICS OF ATOMIC NUCLEI Vol. 69

No. 12

2097

3. WHAT WAS LEARNED FROM THE NEMO-3 DATA COLLECTION? A detailed presentation of the NEMO-3 results can be found in this volume [2]. The NEMO-3 detector has been running in the LSM in nearly optimal conditions since mid-February 2003 and data analyses have shown that the performance of the detector is as expected. The first NEMO-3 results have shown the background reductions at the expected levels, for both internal and external backgrounds. The main result from this data collection is that all the sources of backgrounds are well identified and understood. Using all the background contributions and after the radon suppression, the sensitivity of the effective neutrino mass after five years of data taking should be 0.2–0.35 eV for 100 Mo and 0.65–1.8 eV for 82 Se. As a summary, the collaboration has shown with the NEMO-3 detector its ability to identify all the sources of backgrounds associated with a tracko– calo detector, its ability to build a very low background detector, the quality of the technical choices, its capability of using chemical methods to purify ββ isotopes removing 214 Bi and 208 Tl contaminants, its ability to remove radon from air, the detector ability to measure internal contaminations of the source foils, its ability to control the external backgrounds (natural radioactivity, neutrons, and muons), and its expertise to develop ultralow background HPGe detectors as well as radon detectors which are sensitive to a 1-mBq/m3 radon activity. The data taking and analysis have shown that all sources of background are identified. Taking into account all the different background contributions, the sensitivity on the effective neutrino mass after five years of data taking should be 0.2–0.35 eV for 100 Mo and 0.65–1.8 eV for 82 Se [3]. 4. THE SUPERNEMO PROJECT The NEMO Collaboration began in December 2003 to study the feasibility of a NEMO-3 technique extrapolation to a detector with a mass of at least 100 kg to reach a sensitivity of 50 meV on the effective neutrino mass for the transitions toward the fundamental or excited states. Thus an “expression of interest” was written. The SuperNEMO detector will use the NEMO-3 technical choices: a thin source between two tracking volumes surrounded by a calorimeter. The main points to improve compared to NEMO-3 are the energy resolution, the ββ(0ν)-detection efficiency, the source radiopurity, and the background rejection. All these points are studied in detailed below.

2006

2098

PIQUEMAL

Top view

Side view

4m 1m

5m Possible design of a module of SuperNEMO in top and side views. The source at the center is placed between the tracking volume surrounded by the calorimeter.

4.1. Design of the SuperNEMO Detector The SuperNEMO detector will be composed of several identical modules (for example, 20 modules with 5 kg of 82 Se each), which will allow one to begin the data collection before the end of the last module construction and mounting. Each module will consist of one source foil, one tracking volume, and one calorimeter, these three parts being independent to make easier the construction, mounting, and eventual changes. The detector will be plane and not cylindrical as NEMO-3 was. The mass/volume ratio is not favored, but this solution is easier from a mechanical point of view; it will be easier to build and allow more flexibility on the mounting. Cabling and tightness seal against radon will also be easier. A possible design of the detector, with around 5 kg of 82 Se, is shown in the figure. The thin source planes are at the center. Wire chambers and calorimeter walls will be 5 × 4 m, with dimensions greater than the source size to increase detection efficiency. The width of the tracking detector will be 1 m. Using a solution with 20 × 20 × 2 cm scintillator block size coupled with 8-in. PMTs, 1200 PMTs and 1 t of scintillator should be necessary per module. In a design where the the calorimeter will be made of scintillator bars read at the end by photomultipliers, the number of PMTs would be 120. The gamma calorimeter, consisting of 20-cm-wide plastic or liquid scintillators coupled to PMTs, will surround the detector to detect γ rays and thus to allow background measurements and rejection. With such a configuration, 20 modules should be necessary for a total isotope mass of 100 kg. Thus, the total number of channels should be 60 000 channels for the drift chambers (9000 electronics channels) and 24 000 channels for the electron calorimeter with scintillator of 20 × 20 cm (2400 for a design with scintillator bars)

4.2. Choice of the Isotope Despite the recent progress in the calculation methods for the nuclear matrix elements, these calculations are always too uncertain and nuclear theory is not yet able to predict the best candidate, which means the isotope having the most favorable nuclear matrix elements. A good criterion for isotope selection is the Qββ value with respect to backgrounds. The natural isotopic abundance is another useful criterion because in general the higher the abundance the easier the enrichment process. And finally, the last criterion is the phase-space factor associated with the isotope. The ββ(2ν) background in the energy-transition region has to be reduced too, and it is possible in choosing an isotope with a period T1/2 (2ν) greater than 1020 yr. As explained before, the 2.615-meV γ ray produced in the decay of 208 Tl is consistently a troublesome source of background, and it is important to select ββ candidates with a Qββ value above this transition or to ensure an ultrahigh radiopurity level of the ββ candidates for this contaminant. With these criteria, the best isotope is 82 Se, which has a ββ(2ν) period of ∼ 1020 yr and an energy transition Qββ = 2.998 meV. Using the actual enrichment factories in Russia, the production of 100 kg of 82 Se is feasible in three years for a reasonable cost. In 2005, the ILIAS European program will fund the production of 2 kg to study radiopurity of 82 Se at each production step (before and after enrichment). An R&D program, also funded by ILIAS, for the 82 Se chemical purification is in progress. 4.3. Energy Resolution The ββ(2ν)-decay contribution to ββ(0ν) backgrounds is related directly to its period and to the detector energy resolution. The global resolution, which

PHYSICS OF ATOMIC NUCLEI

Vol. 69 No. 12

2006

THE SUPERNEMO PROJECT

has to be better than 8% (FWHM) at 3 meV to make negligible this ββ(2ν) contribution, depends on the calorimeter resolution and energy losses in both source foils and tracking volume. The SuperNEMO calorimeter part will use plastic or liquid scintillators coupled to PMTs. The goal is to reach a calorimeter resolution better than 4% (FWHM) at 3 meV. An ideal scintillator must have a fast response for time-of-flight measurements and a very low Z value to avoid electron backscattering, and it must also be very radiopure. No actual scintillator satisfies all the criteria. Despite their low scintillation light production, plastic scintillators are a good compromise. To reach resolutions better than 4% at 3 meV, it is necessary to improve the scintillator light efficiency, the block wrapping to increase light collection efficiency, and the PMT efficiency. Concerning plastic scintillators, an R&D program began with a collaboration between the Institute for Scintillation Materials from Kharkov, JINR in Dubna, LAL in Orsay, and CENBG in Bordeaux. NEMO-3 plastic scintillators were produced by the ISM Kharkov and JINR Dubna laboratories, which both have already shown their expertise in production techniques and are working at present to improve the properties of scintillators based on polystyrene. The first results are very encouraging: with small volumes of scintillators coupled to a Photonis XP5312B PMT, resolutions of 7% at 1 meV have been reached. The optimization of the scintillator shape and wrapping will be studied with simulations using special codes to follow γ rays and next with prototype measurements. An another study concerns the possibility to collect the scintillation photons coming from scintillator blocks by means of optical fibers coupled to PMTs. This system should allow reading several scintillators with the same PMT, which is interesting to decrease the number of channels, for the detector’s compactness, and also to decrease the flux of external gammas if the PMTs are not directly coupled to scintillators. A part of the R&D also concerns the photomultipliers. This development is taking place at CENBG in collaboration with the Photonis company to make low-background phototubes with good energy and time resolutions.

4.4. Radiopurity of ββ Source Foils and Detector Materials To reach a 2 × yr sensitivity with 100 kg of 82 Se, 214 Bi and 208 Tl contamination levels of the source foils have to be lower than 10 and 2 µBq/kg, respectively, which means a reduction by factors of 10 compared to NEMO-3 specifications. 1026

PHYSICS OF ATOMIC NUCLEI Vol. 69

No. 12

2099

Chemical purification techniques used for NEMO-3 100 Mo sources [4] can be applied to the 82 Se isotope. For 100 Mo, extraction factors obtained after chemical purification of the enriched powder for the heads of natural decay families were greater than 15 for uranium/radium series (214 Bi) and greater than 8 for thorium series (208 Tl). The extraction factors are constants whatever radium and thorium concentrations are. Thus, on condition of using really pure chemical reagents, it is possible to reach expected radiopurity levels after several cycles. An R&D program began for chemical purification of selenium involving INEEL, Mount Holyoke College, and Bordeaux gamma-spectroscopy measurements. To measure contaminations levels as low as 10 µBq/kg for 214 Bi and 2 µBq/kg for 208 Tl, it is necessary to develop new detectors. HPGe detectors are needed in order to measure the radiopurity at a level of 10 µBq/kg for 208 Tl to check the efficiency of the chemistry. To measure the required contamination level, a dedicated tracko–calo detector will be developed. The principle of this detector will be the detection of the Bi–Po effect for the 208 Tl with the emission of a β particle followed by an alpha particle with a period of 300 ns. The level of radon in the air surrounding the detector should be reduced to 0.1 mBq/m3 , using either a protection for the detector against the radon in the air or a purification of the air before its injection inside the laboratory.

4.5. Design of the Source Foils Using a thickness of 40 mg/cm2 , the energy resolution due to multiple scattering in the source is approximately 6% (FWHM) at 3 meV. A lower thickness will also increase the probability for α particles to leave the source foils. Thus, it will be easier to reject events with α particles, which correspond to the 214 Bi and 208 Tl decay chains. A possibility could be the use of an “active” source (see below). In any case, an R&D phase is needed to obtain 82 Se source foils 20 mg/cm2 thick, either metallic foils (rolling or deposit after chemical purification) or selenium powder in a mixture with glue in a sandwich between two irradiated Mylar foils.

4.6. Drift Cells It is also necessary to improve the tracking volume transparency, either decreasing the wire diameter or using a new wire composition. NEMO-3 wires are in stainless steel with a 50-µm diameter. The possibility to replace stainless steel wires by carbon wires is

2006

2100

PIQUEMAL

considered. The transparency of the detector would be increased by using thiner wires. A very good radiopurity of the wires is needed, in particular, for the wire near the source. 5. CONCLUSIONS Neutrinoless double-beta decay is the most sensitive process to search for lepton number violation. A ββ(0ν)-signal discovery should allow one to determine the neutrino nature, Dirac or Majorana particle, and also to learn about the hierarchy of the neutrino mass states. Study of the ββ(0ν) process is also a way to search for right currents or Majoron and a means to test some parameters of the supersymmetry theories. Using the knowledge acquired with the NEMO-3 detector, the NEMO Collaboration plans to build a “tracko–calo” detector able to reach in the next 10 years a sensitivity of 50 meV. To reach this goal, it will be necessary to use at least 100 kg of 82 Se. The SuperNEMO detector will use the same principles of detection as those used for NEMO-3 by improving them through a 3-yr R&D program. This program will be mainly focused on improving the calorimeter energy resolution, decreasing the source foil thickness, increasing the tracking volume transparency, and finally a better chemical purification of the sources. The ultimate goal of the R&D is to design a detector sensitive to a period of few 1026 yr and sensitive to few tens of meV. The SuperNEMO detector will be competitive compared to the other experimental projects (CUORE,

Majorana, GERDA, EXO, and MOON), which also plan to reach in the next 10 years a 50-meV sensitivity. The originality of this project is due to the unique ability of this detector to identify the two emitted electrons, which is not possible for pure calorimeters. Several experiments with different isotopes are needed owing to the large uncertainties on the nuclear matrix elements. From this point of view, the ability of the SuperNEMO detector to accommodate several ββ isotopes and to change the source foils is a big advantage in the case of improvements in the nuclearmatrix-element theories, which could give the best ββ candidate or confirm a signal from calorimeters. After two years of R&D, the construction of the first module could start and the 20 modules could be in operation in 2010 for results in 2015. This detector could be placed in the LSM if a new cavity is built with dimensions of at least 70 × 15 × 15 m. REFERENCES 1. F. Feruglio et al., Nucl. Phys. B 637, 345 (2002); 659, 359 (2003). 2. R. Arnold et al., The NEMO Contribution to the NANP-05 Proceedings. 3. Conference on Neutrino Physics and Astrophysics “Neutrino 2004,” Paris, France, 2004; http://neutrino2004.in2p3.fr/slides/ thursday/sarazin.ppt. 4. R. Arnold et al., Nucl. Instrum. Methods Phys. Res. A 474, 93 (2001).

PHYSICS OF ATOMIC NUCLEI

Vol. 69 No. 12

2006