Investigation of the 2Ν2β-Spectrum of Cd with the COBRA Experiment
Total Page:16
File Type:pdf, Size:1020Kb
Investigation of the 2ν2β-spectrum of 116Cd with the COBRA experiment Master-Arbeit zur Erlangung des Hochschulgrades Master of Science im Master-Studiengang Physik vorgelegt von Julia K¨uttler geboren am 31.05.1995 in Rodewisch Institut f¨urKern- und Teilchenphysik Fakult¨atPhysik Technische Universit¨atDresden 2019 Eingereicht am 13. M¨arz2019 1. Gutachter: Prof. Dr. Kai Zuber 2. Gutachter: Prof. Dr. Arno Straessner Betreuer: Stefan Zatschler 3 Abstract The COBRA experiment, which is located at the LNGS underground facility in Italy, searches for extremely rare decay modes of several nuclides that are intrinsically abundant in the semi- conductor detectors made of CdZnTe. In this thesis, the 2ν2β-decay of 116Cd, which has a half-life of 2.62·1019 yr [1], will be investigated. To measure such a rare decay, large efforts regarding the background reduction and offline data analysis have to be spent. To improve the discrimination between signal-like and background-like events, a new pulse shape analysis method is implemented and optimized for the data analysis routine. A better signal acceptance compared to the previous used data cuts can be achieved. To further reduce the background level, a data partitioning was accomplished. Run periods and detectors with a higher rate could be identified and excluded from the final data set. This decreases the count rate in the ROI by a factor of 2.8. The exclusion of a certain depth region of the detectors leads to a further rate reduction by a factor of three. The identified background enriched run periods, detectors and detector regions can prospectively be used as a starting point towards a background model for the COBRA demonstrator. Kurzdarstellung Das im Untergrundlabor LNGS in Italien befindiche COBRA Experiment sucht nach extrem seltenen Zerfallsmoden von Nukliden, die im Detektormaterial CdZnTe intrinsich vorhande- nen sind. In dieser Masterarbeit soll der 2ν2β-Zerfall von 116Cd, der eine Halbwertszeit von 2,62·1019 yr [1] hat, untersucht werden. Um solch einen seltenen Zerfall zu messen, m¨ussen besondere Maßnahmen zur Untergrundunterdr¨uckung und bei der Datenauswertung vorgenom- men werden. Um Untergrundereignisse besser von Signalereignissen zu unterscheiden, wurde eine neue Pulsformanalysenmethode in die Datenauswertung implementiert und optimiert. Damit kann eine bessere Signalakzeptanz erreicht werden als mit den bisherigen Auswahlkri- terien. Um den Messuntergrund noch weiter zu reduzieren, wurde eine Datenpartitionierung durchgef¨uhrt.Messperioden und Detektoren mit einer h¨oherenRate wurden identifiziert und von der weiteren Analyse bez¨uglich des 2ν2β-Spektrums ausgeschlossen. In dem f¨urdiese Analyse verwendeten Energiebereich konnte so die Z¨ahlrateum einen Faktor von 2,8 reduziert werden. Der Ausschluss eines bestimmten Tiefenbereiches der Detektoren f¨uhrtezu einer weit- eren Reduktion der Rate um einen Faktor drei. Die ausgeschlossenen Messperioden, Detektoren und Detektorbereiche k¨onnenals Ausgangspunkt f¨urein zuk¨unftigesUntergrundmodell f¨urdas COBRA Experiment dienen. Contents 1 Introduction 7 2 Double beta decay and neutrino physics 9 2.1 Open questions in neutrino physics . .9 2.2 Beta decay . 12 2.3 Double beta decay . 16 2.4 Neutrinoless double beta decay . 20 2.5 Experimental aspects of double beta decays . 23 3 The COBRA experiment 27 3.1 Detector setup . 28 3.1.1 Shielding . 29 3.1.2 Electronic readout system . 31 3.1.3 Experimental infrastructure . 31 3.2 CdZnTe detectors . 32 3.2.1 CdZnTe as a semiconductor . 32 3.2.2 Principle of CPG detectors . 33 4 Pulse shape analysis method A/E 39 4.1 Background identification with PSA . 39 4.2 Currently used data cleaning and PSA methods . 40 4.2.1 LSE-cut . 40 4.2.2 MSE cut . 42 4.2.3 Motivation for a new PSA cut . 43 4.3 A/E parameter . 44 4.3.1 Definition . 44 4.3.2 Experimental data set . 44 4.3.3 A/E implementation in MAnTiCORE . 46 4.4 Optimization procedure . 47 4.4.1 A/E calculation method and smoothing window size . 51 4.5 A/E calibration . 52 4.5.1 Motivation . 52 4.5.2 Calibration procedure . 54 4.5.3 Calibration results . 57 6 Contents 5 Data partitioning 61 5.1 Potential background sources . 61 5.2 Identification of background enriched data periods . 63 5.2.1 Hot pixel . 64 5.2.2 Bad run period . 65 5.2.3 Bad detector . 68 5.2.4 Fiducial volume . 68 5.3 Results of the data partitioning and the A/E-cut . 71 6 Summary and outlook 77 7 Bibliography 79 Appendices 86 List of Figures 87 List of Tables 89 List of Acronyms 91 1 Introduction The neutrino is the particle of the Standard Model of particle physics which properties are least known. The mass of the neutrino as well as its nature, Dirac or Majorana particle, are still unknown after Pauli proposed it more than 85 years ago [2]. Experiments that search for the neutrinoless double beta (0ν2β) decay can help answering some open questions of neutrino physics. The COBRA experiment is one of these experiments. It is located at the underground labora- tory LNGS in Italy and searches for 0ν2β-decay modes of several nuclides intrinsically abundant in the detector material CdZnTe. Besides the ultra-rare 0ν2β-decay, COBRA allows to study several other extremely rare decays such as the neutrino accompanied double beta (2ν2β) de- cay. One nuclide that can undergo this decay is 116Cd. Due to the half-life and the natural abundance of about 7.5% in Cadmium, the decay rate is in a region that can be accessed with the COBRA experiment. An observation of the spectrum with COBRA would be the first mea- surement with a semiconductor detector and thus could verify the measurements from other experiments that used different detection approaches. A consistent half-life can then be used to improve nuclear models to calculate the involved nuclear matrix elements M 2ν as well as to deepen the understanding of the nuclear structure. The first observation of the 2ν2β-decay for 116Cd by a direct counting method was achieved by the ELEGANT-V experiment in 1995 [3]. It used natural and enriched Cd foils between drift chambers for trajectory reconstruction, plastic scintillators for energy measurement and sodium iodide scintillators as background veto. With an exposure of 0.02 kg·yr in terms of 116 +0.9 19 Cd mass, a half-life of 2.6−0.5 · 10 yr was measured [3]. In the same year an exper- iment located in the Solotvina underground laboratory measured the decay with CdWO4 116 +0.5 +0.9 19 crystal scintillators enriched in Cd. A half-life of 2.7−0.4(stat.)−0.6(syst.) · 10 yr was derived [4]. One year later, the NEMO-2 tracking detector measured the decay with a half-life of (3.75 ± 0.35(stat) ± 0.21(syst)) · 1019 yr [5] (later corrected to (2.9 ± 0.3(stat) ± 0.2(syst)) · 1019 yr [6]). It used source foils surrounded by tracking detectors and a calorimeter made of scintillators. Up to now some of these experiments improved their results (Solotvina experiment [7] and NEMO-3 [8]). The Aurora experiment at the LNGS also used CdWO4 and measured a half-life of (2.62 ± 0.14) · 1019 yr [1]. In figure 1.1 the results of the different experiments are compared to each other. With the calculation of the number of 2ν2β-decays that can be measured with COBRA, one gets a rough approximation of the sensitivity and the required background level: ln2 N116 N = c · · Cd = 3.65 · 103 cts/ (kg · yr) , (1.1) T1/2 mCZT 8 1 Introduction Figure 1.1: 2ν2β half-lives of 116Cd measured with ELEGANT-V [3], NEMO-2 [5], Solotv- ina [4][9][7], NEMO-3 [8] and Aurora [1]. A re-evaluated NEMO-2 result [6] is labelled as (NEMO-2)∗. Picture taken from [1]. 19 116 with T1/2 = 2.62 · 10 yr from the Aurora experiment [1] and the number of Cd atoms per CdZnTe mass according to: N116 N Cd = a · s · A = 235.3 g/mol, (1.2) mCZT hMCd0.9Zn0.1Tei where the natural abundance a = 0.0749 of 116Cd and the stoichiometric factor s = 0.9 are used. NA denotes the Avogadro constant and hMCd0.9Zn0.1Tei the molar mass of CdZnTe in the given stoichiometric composition. The factor c takes the detection efficiency as well as the signal acceptance of the analysis cuts into account. 3650 counts per year and kg of CdZnTe require a background level below 1 cts/(kg·keV·yr). In this thesis, two possibilities to reach the goal of observing the 2ν2β-spectrum will be in- vestigated. After an introduction into the physics of double beta decay (chapter 2) and the COBRA experiment (chapter 3), the first approach - a new pulse shape analysis method, called A/E - will be investigated (chapter 4). With this, better efficiencies regarding the background rejection and signal acceptance should be achieved. This would enlarge the number of events that can be detected. In chapter 5 a so called data partitioning will be done. This includes a scan for run periods and detectors that have a higher background index, and the subsequent removal of these periods from the data set. The remaining data are than background reduced and can be used for the analysis. 2 Double beta decay and neutrino physics 2.1 Open questions in neutrino physics 1 Within the Standard Model of particle physics, neutrinos are massless spin- 2 particles (fermions). Since they have no electrical or color charge, neutrinos are only observable via weak interac- tions. There exist three neutrino flavors: νe, νµ and ντ [2].