First results of the experiment to search for double beta decay of 106Cd with 106CdWO4 crystal scintillator in coincidence with four crystals HPGe detector

An experiment to search for double beta processes in 106Cd by using cadmium tungstate crystal scintillator enriched in 106Cd (106CdWO4) in coincidence with the four crystals HPGe detector GeMulti is in progress at the STELLA facility of the Gran Sasso underground laboratory of INFN (Italy). The 106CdWO4 scintillator is viewed by a low-background photomultiplier tube through a lead tungstate crystal light-guide produced from deeply purified archaeological lead to suppress gamma quanta from the photomultiplier tube. Here we report the first results of the experiment after 3233 hours of the data taking. A few new improved limits on double beta processes in 106Cd are obtained, in particular T1/2(2nuECb+)>8.4e20 yr at 90% C.L.


Introduction
Experiments to search for double beta (2) decay are considered as a promising way to investigate properties of neutrino and search for effects beyond the Standard Model of particles [1,2,3,4,5]. The isotope 106 Cd (energy of decay Q 2 = 2775.39(10) keV [6], isotopic abundance  = 1.25(6)% [7]) is one of the most suitable nuclei to search for 2 processes with decrease of nuclear charge (see e.g. [8] and references therein). A strong motivation to study the double beta "plus" processes was discussed in [9] where a possibility to distinguish between a right-handed currents admixture and the neutrino mass mechanisms has been considered.
A cadmium tungstate crystal scintillator enriched in 106 Cd to 66% ( 106 CdWO 4 ) was developed [10] to search for double beta processes in 106 Cd. The first stage of the experiment realized in the DAMA/R&D set-up at the Gran Sasso underground laboratory was reported in [8].
To increase the experimental sensitivity to the 2 processes with emission of  quanta, the 106 CdWO 4 scintillator was placed inside a low background HPGe detector with four Ge crystals. Here we report first results of the experiment.

Experiment
The 106 CdWO 4 crystal scintillator (mass of 215 g) is viewed through a lead tungstate (PbWO 4 ) crystal lightguide (40 × 83 mm) by 3 inches low radioactive photomultiplier tube (PMT) Hamamatsu R6233MOD (see Fig. 1). The PbWO 4 crystal was developed from deeply purified [11] archaeological lead [12]. The detector is installed in an ultra-low background GeMulti HPGe  spectrometer of the STELLA (SubTErranean Low Level Assay) facilities [13] at the Gran Sasso underground laboratory (LNGS) of the INFN (Italy) on the depth of 3600 m of water equivalent. Four HPGe detectors of the GeMulti set-up are mounted in one cryostat with a well in the centre. The volumes of the HPGe detectors are approximately 225 cm 3 each. The typical energy resolution (FWHM) is 2.0 keV for the 1332 keV  quanta of 60 Co.
An event-by-event data acquisition system is based on two four-channel all digital spectrometers (DGF Pixie-4, XIA, LLC). One device (marked (1) in Fig. 1) is used to provide spectrometric data for the HPGe detectors, while the second Pixie-4 (2) acts as a 14-bit waveform digitizer to acquire signals from the 106 CdWO 4 detector at the rate of 18.8 MSPS over a time window 54.8 s. The second Pixie-4 unit records also trigger signals from the home made unit SST-09, which provides the triggers only if the signal amplitude in the 106 CdWO 4 detector exceeds ~ 0.6 MeV to avoid acquisition of a large amount of data caused by the decays of 113m Cd (Q  = 580 keV) presented in the 106 CdWO 4 crystal [8]. The signals from the timing outputs of the HPGe detectors after summing are fed to the third input of the second Pixie-4 digitizer to select off-line coincidence between the 106 CdWO 4 and HPGe detectors. The detector was calibrated with 22 Na, 60 Co, 137 Cs and 228 Th  sources. The energy resolution of the detector can be described by the function: FWHM = (20.4 × E   1/2 , where FWHM and E  are given in keV. Energy spectrum and distribution of the start positions of the 106 CdWO 4 detector pulses relatively to the HPGe signals accumulated with the 22 Na  source (see Fig. 2) demonstrate presence of coincidences between the 106 CdWO 4 and HPGe detectors under the condition that the energy of events in the HPGe detectors is equal to 511 keV (energy of annihilation gamma quanta), while practically there is no coincidence in the data accumulated with 137 Cs. The measured data is in agreement with the distributions simulated by the EGS4 code [14] (Fig. 2).

Low background measurements
The mean-time pulse-shape discrimination method (see [8]) was used to discriminate () events from  events caused by internal contamination of the crystal by uranium and thorium. The energy spectrum of the () events accumulated in the set-up over 3233 h is shown in Fig. 3. The data confirmed the assumption about surface contamination of the 106 CdWO 4 crystal by 207 Bi [8]. The  peaks of 207 Bi disappeared thanks to the cleaning of the scintillator by potassium free detergent and ultra-pure nitric acid.
The spectrum was fitted by the model built from the energy distributions simulated by EGS4 code. The model includes radioactive contamination of the 106 CdWO 4 crystal [8], PMT [15], PbWO 4 light-guide, copper shield, aluminium wall of the cryostat. The main components of the background are shown in Fig. 3.
Energy spectra accumulated by the HPGe detector are presented in Fig. 4. The counting rate of the HPGe detector with the 106 CdWO 4 counter inside exceeds slightly the background counting rate. Some excess on the level of (30170)% (depending on the energy of gamma quanta) is observed in the peaks of 214 Bi and 214 Pb (daughters of 226 Ra from 238 U family). Simulation of the background is in progress with an aim to identify materials of the 106 CdWO 4 detector contributing to the counting rate excess.  The counting rate of the 106 CdWO 4 detector is substantially suppressed in coincidence with events in the HPGe detector with energy 511 keV (see Fig. 5). The decrease of the background was confirmed also by the Monte Carlo simulation. A counting rate in the coincidence data presented in Fig. 5 (53 counts in the energy interval 50  3000 keV) is in agreement with the calculated background (53 events) using the parameters of the fit of the 106 CdWO 4 detector background without coincidence.

Sensitivity to the 2 processes in 106 Cd
The response functions of the 106 CdWO 4 detector to the 2β processes in 106 Cd were simulated with the help of the EGS4 code. The distributions without coincidence and in coincidence with 511 keV  quanta in the HPGe detector are presented in Fig. 6. There are no peculiarities in the data accumulated with the 106 CdWO 4 detector that could be ascribed to the 2 processes in 106 Cd. Therefore only lower half-life limits can be set by using the formula: where N is the number of 106 Cd nuclei in the 106 CdWO 4 crystal (2.42 × 10 23 ),  is the detection efficiency, t is the time of measurements, and limS is the number of events of the effect searched for, which can be excluded at a given confidence level (C.L.).
To estimate limS values for the double beta processes in 106 Cd, the measured number of events in the background spectrum was compared with the expected background, which was estimated by using the result of the fit of the data accumulated by the 106 CdWO 4 detector without coincidence (see Fig. 3). For instance, the number of events in the coincidence data in the energy interval 500  1200 keV is equal to 13 counts, while the model of background gives 17.6 counts. According to the procedure proposed in [16], we should take 3.7 counts as an effect's limit which can be excluded with 90% C.L. Taking into account the detection efficiency to the two neutrino electron capture with emission of positron in 106 Cd (7.6%), the part of the energy spectrum in the energy interval (67.0%), the selection efficiency of the time and energy cuts used to obtain the coincidence spectrum (totally 99%), one could get a new improved limit on the effect searched for: Similarly, by using the described procedure, the limits on some other double beta processes in 106 Cd were obtained. Some of the excluded distributions of double beta processes in 106 Cd are presented in Fig. 5. All the half-life limits are summarized in Table 1, where results of the most sensitive previous experiments are given for comparison. Table 1. Half-life limits on 2 processes in 106 Cd to the ground state (g.s.) and to the first 0 + 1134 keV excited level of 106 Pd.