Results from GERDA Phase I and status of the upgrade to Phase II. Stefan Schönert (TU München) on behalf of the GERDA collabora;on

Transcription

1 Results from GERDA Phase I and status of the upgrade to Phase II Stefan Schönert (TU München) on behalf of the GERDA collabora;on 1

2 The GERDA LNGS plasac µ- veto clean room with lock and glove box for detector handling Eur. Phys. J. C (2013) 73:2330 arxiv: muon & cryogenic infrastructure cryostat, Ø4m, with internal Cu shield control rooms Ge-detector array (enriched in 76 Ge) water plant & radon monitor water tank, Ø10m, part of muon-veto detector 2

3 GERDA Phases Goal: background free opera;on T 0ν 1/2 scales with exposure claim 3

4 GERDA Phases Goal: background free opera;on T 0ν 1/2 scales with exposure claim Phase I: Use refurbished HdM & IGEX (18 kg) BI 0.01 cts / (kev kg yr) Sensi;vity acer 20 kg yr 4

5 GERDA Phases Goal: background free opera;on T 0ν 1/2 scales with exposure Phase II: Add new enr. BEGe detectors (+20 kg) BI cts / (kev kg yr) Sensi;vity acer 100 kg yr claim Phase I: Use refurbished HdM & IGEX (18 kg) BI 0.01 cts / (kev kg yr) Sensi;vity acer 20 kg yr 5

6 The GERDA construc;on S. Schönert (TUM): GERDA Phase I results and status of Phase II Neutrino

7 Nov 2011: deployment of 3- string & start of Phase I physics runs 8 refurbished enriched diodes from HdM & IGEX 86% isotopically enriched in Ge kg total mass plus 1 natural Ge diode from GTF 2 diodes shut off because leakage current high: total enriched enriched detector mass 14.6 kg 7

8 Overview of data taking Total exposure for 0νββ analysis: 21.6 kg yr (bi- )weekly calibra;on runs ( spikes ) Eur. Phys. J. C (2013) 73:2330 arxiv: νββ background paper Data blinding: All events in Q ββ ±20 kev removed in Tier 1 2 copies of raw data kept for processing acer unblinding 1 st physics: 2νββ analysis (5.04 kg yr) 8

9 Measurement of T 2ν 1/2 (76 Ge) Acer only 5.04 kg yr exposure: T 2ν 1/2 (76 Ge) = ( ) 1021 yr 9

10 Overview of data taking Total exposure for 0νββ analysis: 21.6 kg yr (bi- )weekly calibra;on runs ( spikes ) Eur. Phys. J. C (2013) 73:2330 arxiv: νββ background paper Data blinding: All events in Q ββ ±20 kev removed in Tier 1 2 copies of raw data kept for processing acer unblinding 1 st physics: 2νββ analysis (5.04 kg yr) Inser;on of 5 Phase II enr BEGe 10

11 Calibra;on: stability of HPGe detectors Peak posi;on stability of kev calibra;on line: coax: 1.5 kev / BEGe: 1.0 kev (FWHM) Eur. Phys. J. C (2013) 73:2330 Summing all runs: shift [kev] shift of kev position GERDA ANG 2 ANG 3 ANG 4 ANG 5 RG 1 RG 2 counts all coaxial detectors 42 K background line GERDA energy: ± 0.1 kev FWHM: 4.47 ± 0.12 kev -1-2 no. entries kev (FWHM) FWHM=1.29 kev mean=-0.11 kev shift 2 [kev] Nov/11 Jan/12 Mar/12 May/12 Jul/12 Aug/12 Oct/12 Dec/12 Mar/13 May/ energy [kev] Energy resolu;on of coax detectors at 2039 kev Mean energy resolu;on at Q ββ =2039 kev: Coax: 4.8 kev (FHWM) BEGe: 3.2 kev (FWHM) 11

12 Physics run: energy spectra arxiv: Eur. Phys. J. C (2014) 74: Ar 39 Ar natural 12

13 Physics run: energy spectra as func;on of ;me (blinded) blinded S. Schönert S. Schönert (TUM): (TUM): GERDA GERDA results Phase on I results 0νββ decay and status search of Phase ArtFest, II Mai Neutrino 31,

14 Physics run: background rate as func;on of ;me Coax- detector data set split in Gold and Silver (30 d) 14

15 Physics run: background decomposi;on arxiv: Eur. Phys. J. C (2014) 74:2764 Coaxial detector Thick n + contact Thin p+ contact At high energies: dominated by (unsupported) 210 Po (T 1/2 =138 d) also contribu;ons of 226 Ra and progenies located on thin p+ contact; confirmed by pulse shape analysis 15

16 Physics run: background decomposi;on arxiv: Eur. Phys. J. C (2014) 74:2764 Coaxial detector Thick n + contact Thin p+ contact At intermediate energies: Coax: no single dominant component at Q ββ for coax detectors BEGe: 42 K at n+ surface dominate because of thinner dead layer 16

17 Physics run: background decomposi;on arxiv: Eur. Phys. J. C (2014) 74:2764 At low energies: 2νββ decays (>0.6 MeV) 39 Ar (<0.6 MeV) Q ββ (blinded) 17

18 Physics run: background model and predic;on of BI at Q ββ Eur. Phys. J. C (2014) 74:2764 blinded No background peaks expected around Q ββ expected BI at Q ββ ( ) x 10-3 cts/(kev kg yr) depending on assump;ons for loca;on of sources Spectrum can be modeled with flat background (red line) in kev excluding known peaks at 2104 and 2119 kev StaAsAcal uncertainty of BI from interpola;on coincides numerically with systemaac uncertainty from model Predic;on for 30 kev blinded side wings: Min./Max Mod: / observed.: 13 18

19 Pulse shape discrimina;on: method and cuts fixed prior unblinding coax BEGe 19

20 Unblinding: full data set (21.6 kg yr) counts/kev 3 GERDA gold- coax: 17.9 kg yr; silver- coax: 1.3 kg yr; bege: 2.4 kg yr Q ββ ±5keV blinded window PRL 111 (2013) before PSD acer PSD 1 counts/(2 kev) kev 2039 kev ββ Q background interpolation 2190 kev Bi 2204 kev Full data set: energy [kev] 7 events obs. in blinded window vs. 5.1 expected for bgd only 3 events survive PSD cut vs. 2.5 expected for bgd only 20

21 counts/kev counts/(2 kev) Parameter Det./Set Value Uncertainty <ε> w/o PSD Coax Profile likelihood fit to 3 data set (21.6 kg yr) Frequen;st and Bayesian limits & median sensi;vi;es 3 GERDA % lower limit (T 0ν 1/2 ) kev energy [kev] Systema;cs: 2039 kev BEGe Energy res. Golden 4.83 kev 0.19 kev Silver 4.63 kev 0.14 kev BEGe 3.24 kev 0.14 kev Energy scale (kev) N.A. 0.2 kev ε PSD Coax / ββ Q background interpolation BEGe kev Bi 2204 kev 214 PRL 111 (2013) Frequen;st limit: 90% lower limit derived from profile likelihood fit to 3 data sets (constraint to physical 1/T range; excluding known γ- lines from bgd model at 2104±5 and 2119±5 kev) Best fit: N 0ν =0 No excess of signal counts above the background 90% C.L. lower limit: Limit on half- life corresponds to N 0ν <3.5 cts Median sensi;vity (90% C.L.): > yr Bayesian: Flat prior for 1/T Posterior distribu;on for T 0ν 1/2 Best fit: N 0ν =0 90% credibile interval: Median sensi;vity: (90% C.I.): > yr Systema;cs folded: limit weakened by 1.5% 21

22 Comparison with Phys. Le. B (2004) 0νββ claim in 76 Ge Claim: T 0ν 1/2 = Klapdor- Kleingrothaus et al. NIM A 522 (2004) PLB 586 (2004) 90% lower limit (T 0ν 1/2 ) PRL 111 (2013) H0: background only H1: claimed signal plus background è Claim refuted with high probability independent of NME and lepton number viola;ng mechanism p- value from profile likelihood P(N=0 H1 ) = (0.006 if 1/T unconstrained) Bayes factor: P(H1)/P(H0 )=

23 Comparison with the Xenon DBD experiments Sensi;vity GERDA Phase I EXO sensi;vity 23

24 Reducing background further for GERDA Phase II Phase II: 10-3 cts/(kg kev yr) 1/10 Phase I: 10-2 cts/(kg kev yr) Major hardware upgrade: Novel detectors with advanced pulse shape discrimina;on (BEGe s) Improved detector assembly & electronics Liquid argon instrumenta;on (veto system) 24

25 Thick window BEGe s with advanced pulse shape performance p+ n+ Signal shape provides clear topology for event-by-event signal ID / bgd discrimination: SSE/MSE discrimination Surface events: n+ slow pulses p+: amplified current pulses signal- like bgd- like (MSE γ s) DEP: 90% 0νββ-like γ-bgd: 11% bgd- like (p+ surface α, β) bgd- like (n+ surface β, 42 K) 25

26 Phase II enriched BEGe detector characteris;cs 1.74 kev Background index due cosmogenic ac;va;on(w/o PSD): 60 Co: 68 Ge: 30 detector slices with 20.8 kg total Energy resolu;on in vacuum cryostats <1.9 kev 1.3 MeV 26

27 Phase II enriched BEGe detector assembly and VFE electronics Phase II detector assembly Silicon support BEGe- pair Wire bonded contacts silicon copper mock- up flex cables with JFET and feedback resistor/capacitor 27

28 Suppression of bgd by PSD & liquid argon scin;lla;on veto Calibra;on with 228 Th source near source LArGe test LNGS far source 1 MeV LAr spiked with 42 Ar( 42 K) 2 MeV Bare BEGe (scaled) With nylon MS PMT cut 89% acc. PSD + PMT cut 73% acc. PSD + PMT cut 28

29 Liquid argon instrumenta;on for Phase II 29

30 30

31 Summary & outlook GERDA Phase I design goals reached: Background index acer PSD: 0.01 cts / (kev kg yr) Exposure 21.6 kg yr No 0νββ- signal observed at Q ββ = 2039 kev; best fit: N 0ν =0 Claim strongly disfavored (independent of NME and of leading term) Limit on half- life: GERDA: GERDA+IGEX+HdM: T 0ν 1/2 > yr (90% C.L.) T 0ν 1/2 > yr (90% C.L.) (<m ee > < ev) Results reached acer only 21.6 kg yr exposure because of unprecedented low background: bgd expecta;ons acer analysis cuts and correc;ng for efficiencies: cts /(mol yr FWHM) Analysis in pipeline: 2νββ to excited state (poster #136), 0νββ exc. state, Majoron TransiAon to Phase II ongoing: Increase of target mass (+20 kg; total 40 kg of Ge detectors) New custom made BEGe detectors with enhanced pulse shape discrimina;on Liquid argon instrumenta;on Background 10-3 cts /(kev kg yr) Explore T 0ν 1/2 (0ν) values in the 1026 yr range 31