1. Planned Transregional Collaborative Research Center TR27: Neutrinos and Beyond Project A4: Development of segmented germanium detectors for the investigation of neutrinoless double beta decay I.Abt, K. Ackermann, M. Altmann, A. Caldwell, D. Kollar, M. Jelen, K. Kröninger, J. Liu, X. Liu, S. Mayer, B. Majorovits, R. Richter, F. Stelzer, S. Vogt MPI für Physik (Werner-Heisenberg-Institut), München Neutrinoless Double Beta Decay and the GERDA experiment Cryoliquid Cryo-tank / vacuum Water / Muon-Veto (Č) Germanium detectors Clean room / lock Neutrinoless double beta decay (0νββ) can occur as an extremely rare second order weak process if the neutrino is a Majorana particle. The half-life of the process is a function of the neutrino masses, their mixing angles and CP- phases. An observation of the 0νββ- process would not only give information about the absolute scale of neutrino masses but also reveal the nature of the neutrino. Schematic diagram of neutrinoless double beta decay. The nucleus changes its charge by two units. If the neutrino is a Majorana particle, the two emitted neutrinos can annihilate. Only two electrons remain in the final state. Today‘s best limit on the half-life of the 0νββ-process of 76 Ge is measured by the Heidelberg-Moscow and IGEX experiments with T 1/2 > 1.9·10 25 years and T 1/2 > 1.6·10 25 years (at 90% C.L.), respectively. The GERmanium Detector Array, GERDA, is a new experiment that will search for neutrinoless double beta decay of the germanium isotope 76 Ge. Its main design feature is to submerge and operate high purity germanium detectors, enriched in 76 Ge to a level of 86%, directly in a cryogenic liquid (nitrogen or argon). The latter serves as coolant and shield from external radiation simultaneously. The cryostat is placed inside a buffer of ultra-pure water which serves as additional shielding and will be instrumented as For the envisioned background index of 10 -3 counts/(kg keV y) and an exposure of 100 kg years the lower limit on the half-life is T 1/2 >1.35·10 26 years which corresponds to a limit on the effective neutrino mass of m ee <200 meV/c 2. Cherencov detector in order to veto cosmic muons. With this setup a background index of 10 -3 counts/(kg keV y) is expected. The GERDA experiment will be installed in the Hall A of the Gran Sasso National Laboratory, LNGS, in Italy. Sensitivity of the GERDA experiment using a Bayesian analysis technique. The lower limit for an exposure of 100 kg years and the ensvisioned background index of 10 -3 counts/(kg keV y) is T 1/2 >1.35·10 26 years. Baseline design of the GERDA experiment. Time Charge Mirror charges Signal electrode Background identification using segmented detectors The largest background contribution is expected to come from events with photons in the final state. An identifi- cation of those events is therefore crucial for the understanding and suppression of the background. Distribution of the energy spread for different background processes. R 90 is defined as the radius which contains 90% of the energy within an event. Class II–IV events have photons in the final state and can be distinguished from Class I and V events which have electrons and alpha-particles in the final state, respectively. Photons in the relevant energy region lead to a spatially more extended energy deposit inside the detector than electrons. Segmented detectors can be used to distinguish those two event classes by requiring coincidences between seg- ments. Charge [a.u.] Time [ns] 1 2 1 2 Mirror charge trajectory DEP event efficiency Tl-208 2.6 MeV event efficiency Single segment events Left: The GERDA Phase II detectors are forseen to have a 6-fold segmentation in the azimuthal angle φ and a 3-fold segmentation in the height z. Right: Cross- sectional view onto the GERDA detector array. The signal process has two electrons in the final state which mostly deposit energy in only one segment. In the decay of Co-60 two photons are emitted which scatter and possibly deposit energy in more than one segment. A cut on the segment multiplicity can be used in order to identify events with photons in the final state. 0νββCo-60 A Phase II prototype detector has been produced and is currently being investigated at the MPI. Measurements with a phase II prototype detector Th-228 spectrum taken with the GERDA Phase II prototype detector. The black line represents the core spectrum, the red line represent the core spectrum for events with only one segment being hit. The double escape peak (left) is not suppressed while the gamma-line from the Thorium decay chain (right) is strongly suppressed by anti- coincidence require- ments between seg- ments. A variety of measurements is performed and compared with Monte Carlo taking background into account. The comparison yields an agreement on the 10% level. Detector responses and pulse shape analysis In addition to segmentation, the pulse shapes of the core and each segment can be used to further identify the energy distribution within the detectors. A thorough understanding of the signal development is needed in order to interpret the measured signals. The induced charge seen at the upper right electrode. Deposited energy translates into electron-hole pairs which separate and drift towards the inner and outer electrodes. Energy deposited at position 1 will lead to a gain in the net charge seen by the preamplifier (top pulse shape) whereas at position 2 only mirror charges are induced in the electrode under consideration and no net charge is gained. Pulses from all electrodes can be used in order to gain information on the topology of the charge deposition within the detector. Radius [cm] Risetime [ns] z [cm] Asymmetry top-bottom Angle [°] Asymmetry left-right Top: An 18-fold segmented detector with dimensions similiar to the prototype detector was simulated. The charges are drifted towards the electrodes and the induced charges calculated using Ramo‘s theorem. The raw signals are convoluted with a transfer function and sampled in order to account for the electronics. Left: Spatial information gained by the analysis of pulse shapes. The risetime between 10% and 90% of the signal amplitude is correlated to the radius at which the energy is deposited. A minimum at 2 cm is observed since the electrons and holes have a minimum drift. The asymmetry of the mirror charge amplitudes of the left and right neighboring segments is correlated to the angle at which the energy is deposited within the segment under study. For segments in the (z-)middle the top-bottom asymmetry of the pulse amplitudes is correlated with the z-position of the energy deposition. Photon identification Pulse shape data was taken with the Phase II prototype detector. A feasibility study for three different approaches for the analysis of pulse shapes (of single- segment events) was performed. The results are in agreement and give an additional suppression of ≈50% given a signal efficiency of ≈80%. The double escape peak of Tl-208 is due to events with a localized energy deposition whereas the Tl-208 peak at 2.6 MeV is mostly populated by multiply scattered photon events. The efficiencies for both samples obtained with a neural network analysis are shown. Single-segment spectrum Total spectrum (μm)