1. Calor 2010Anastasia Freshville1 SuperNEMO Calorimetry Anastasia Freshville (On behalf of the SuperNEMO Collaboration) Calor2010 – 12 th May

2. Calor 2010Anastasia Freshville2 by Matthew Kauer Neutrinoless Double Beta Decay Neutrinoless double beta decay: 2  0 Beyond SM: Total lepton number violated by 2 Most sensitive way to establish Majorana/Dirac nature of neutrino (requires the neutrino to be a Majorana particle) Most sensitive way to measure absolute neutrino mass in a lab environment (for Majorana neutrino) Possible access to neutrino mass hierarchy and Majorana CP- violation phases Link to matter-antimatter asymmetry (leptogenesis) _ ν = ν ΔL = 2! Nuclear matrix element Lepton number violating parameter Phase space (α Q ββ 5 ) Half-life  η can be due to different mechanisms: (η = ν e effective mass), V+A, Majoron, SUSY, H - - or a combination of them. Therefore a detector which can probe different mechanisms and different isotopes is required!

3. Calor 2010Anastasia Freshville3 by Matthew Kauer M mass (g) A isotope molar mass (g/mol)  efficiency k C.L. confidence level N Avogadro’s number t experimental time (y) N Bkg background events (keV -1.g -1.y -1 )  E energy resolution (keV) > AA M. t N Bkg.  E (y). ln2 N k C.L.. 12% FWHM7% FWHM The Importance of Energy Resolution The experimental sensitivity: 2 νββ is always a background to 0 νββ : The better the energy resolution the better the distinction between 2 νββ and 0 νββ there is → require best possible energy resolution!

4. Calor 2010Anastasia Freshville4 SuperNEMO Sensitivity Modular detector with a planar geometry: Use and improve NEMO3’s tracker-calorimeter technique (currently running in LSM, France) 1 module (of 20) consists of: Source foil: 5kg of enriched 82 Se (total of 100kg) → separate from detector hence can measure different foils Tracker: ~ 2000 drift cells in Geiger mode → particle identification (for background suppression) Calorimeter: ~550 scintillator blocks + PMTs → energy and time of flight measurements of particles Sensitivity Simulations: (5 years with 100kg of 82 Se) SuperNEMO goals: T ½ > 10 26 years → <0.04 – 0.1 eV Calorimeter requirements: 1) FWHM = 4% @ 3 MeV (Q ββ of 82 Se) (compared to FWHM = 8% @ 3 MeV for NEMO3) 2) Time resolution: 250ps @ 1 MeV 3) Calibration precision < 1% 4) PMT radio-purity (Depending on Rn emanation: 214 Bi < 0.05Bq/kg, 208 Tl < 0.004Bq)

5. Calor 2010Anastasia Freshville5 Requirements for the SuperNEMO Calorimeter:  Energy resolution:  Photomultiplier tube (PMT) characteristics: Quantum efficiency (QE)  Significant breakthroughs have been made with new bialkali photocathodes (Hamamatsu and Photonis), with a QE range of 35-43% at the peak wavelength. Uniformity Cathode to first dynode collection efficiency Radiopurity  Scintillator light output (collection efficiency): Scintillator material and polishing/depolishing  Low backscattering → low density material Scintillator geometry Reflector material and efficiency Optical coupling quality (gels, light guides etc)  Ease of manufacturing and assembly  Sensible cost (a large surface is required)  Robust (reliable and time withstanding) technology SuperNEMO Calorimeter Requirements Choose Plastic Scintillator: Must improve energy resolution by factor of 2/light output by a factor of 4 from NEMO3! Hamamatsu QE Profile:

6. Calor 2010Anastasia Freshville6 Two Possible Calorimeter Designs Baseline (Block) Design Bar Design minimum 12cm 20-25cm 200cm 10cm Hexagonal PVT Block + 8” PMT Advantages: Excellent energy resolution: 7% FWHM at 1 MeV Well understood technology from NEMO-3 experience Good time resolution Disadvantages: 11,000 block + PMT units required for full detector: More expensive Potentially more radioactivity due to number of channels PVT Bar (2m x 10cm x 2.5cm) + 3” PMT at Each End Advantages: Efficient γ tagging 5,000 - 7,500 bar + PMTs units required for full detector (depending on source foil thickness) Cheaper Potentially less radioactivity Disadvantages: Worse energy resolution: 10% FWHM at 1 MeV Possible ageing problems More difficult to extrapolate from NEMO-3

7. Calor 2010Anastasia Freshville7  Full calorimeter simulations: GENBB event generator Physics simulations with GEANT4 (optical photon transport in scintillator detectors)  The model accounts for wavelength dependence of optical properties, all of which have been experimentally measured, of the: scintillators (self absorption and re-emission) reflective wrappings photomultipliers (QE) optical coupling materials refractive index of optical materials  Simulations verified using data from the currently running NEMO3 detector Calorimeter Simulations Block TypeSimulated ΔE/EMeasured ΔE/E 114.4 ± 1.113.8 ± 0.3 214.0 ± 1.113.5 ± 0.2 314.9 ± 1.116.7 ± 0.2

8. Calor 2010Anastasia Freshville8 Calorimeter Simulations  SuperNEMO calorimeter unit simulated: Scintillator:  Hexagonal PVT (eg. EJ-200) block of 22.5cm radius  Light yield of scintillator: 10,000 photons per MeV Wrapped in:  Aluminised mylar (92% reflectivity coefficient) around the sides and entrance face of scintillator  Teflon wrapping on the top face of scintillator (near PMT) Coupled to:  Super-bialkali 8” Hemispherical PMT (eg. Hamamtsu R5912-MOD)  QE of PMT: 33%  Using the symmetry of the block, each sextant of the hexagonal face was divided into 16 regions to check the uniformity of the energy resolution across the face of the block: Mean energy resolution across sextant: 7.19% FWHM @ 1 MeV Minimum energy resolution: 7.14 ± 0.5% Maximum energy resolution: 7.24 ± 0.5% The simulations and bench test (7.7 ± 0.2% FWHM @ 1 MeV) results agree!

9. Calor 2010Anastasia Freshville9 Baseline Design R&D Energy resolution testing carried out using two different methods for cross checks: The two methods are consistent with each other! A lot of R&D work has been done to obtain the result we see today for the block design: Scintillator material study: liquid, liquid and solid hybrid, solid (PST and PVT) Scintillator wrapping and polishing studies to achieve best possible light output Light-guide and optical gel studies to achieve best possible optical contact: Removing lightguide from setup → breakthrough! Block shape studies → arrived at hexagonal shape with PMT looking directly at the block Radiopurity studies Polished PVT Block 2) 90 Sr Spectrometer 1 MeV electron (magnetic field used to select energy of beam) Simple fitting procedure 90 Sr Energy Distribution 7.6% FWHM at 1 MeV 1) 207 Bi Source Conversion electrons at 482keV and 976keV smeared by 2 Compton edges Complicated fitting procedure 207 Bi ADC Distribution 7.8% FWHM at 1 MeV

10. Calor 2010Anastasia Freshville10 EJ-200 Hexagonal Block (10cm depth) with 8” High QE Photonis PMT and 15μm mylar wrapping on entrance face + teflon wrapping on sides Baseline Design R&D PMT R&D: New PMTs developed to reach a quantum efficiency > 40% and a good linearity (NPE = 3000) with Photonis and Hamamatsu (8 stage dynode tubes developed for improved linearity and timing) Work in progress: a material radiopure enough to use for PMT glass has been identified, allowing us to reach the purity levels required for SuperNEMO (depending on Rn emanation!): 40 K < 0.1 Bq/kg 214 Bi < 0.05 Bq/kg 208 Tl < 0.004 Bq/kg 8” PMTs Best energy resolution result for blocks to date: EJ-200 Hexagonal Block (10cm depth) with 8” High QE Hamamatsu PMT and 15mm mylar wrapping on entance face + teflon wrapping on sides Photonis have announced a stop to all R&D and production in 2009 The problem has been identified: currently working closely with Hamamatsu to improve the collection efficiency Target  E/E = 4% at 3 MeV (Q  of 82 Se) reached!

11. Calor 2010Anastasia Freshville11 Bar Design R&D Looking along the scintillator bar Best energy and time resolution result for bars to date: ΔE/E: at 1 MeV Coincidence time resolution: 395ps Bars give a 50% lower sensitivity if SuperNEMO target backgrounds are reached In February 2010 a decision for a block demonstrator (1 SuperNEMO) module design was made, which will measure the background levels Keep low level R&D for bar design going as a back up option to the block design (in case our main background comes from the PMTs) EJ-200 scintillaor bar: 2m x 10cm (tapered to 6.5cm at ends) x 2.5cm Wrapped in 14.5μm aluminised mylar3” Hamamatsu SBA-select tubes (~ 40% QE)

12. Calor 2010Anastasia Freshville12 Shape 1 + PMT Shape 2 Shape 4 + PMT Shape 4 Block Shape Study  Final details of the block design must be finalised, such as the block shape!  Study the effect of block geometry on the energy resolution with one scintillator (PS: raw polystyrene) block and verify simulations Note: PVT has given us the best results, but use PS block for first study due to ease of machining Machined after each measurement to the new geometry  Measurements are all carried out under the same conditions: Teflon wrapping on the sides and aluminised mylar on entry face Same PMT used for tests (Photonis standard QE (25%) 8”) at the same High Voltage Optical contact used: alcohol Source: 1 MeV 90 Sr electrons (from spectrometer) at the centre of the entrance face

13. Calor 2010Anastasia Freshville13 Block Shape Study Final shape for SuperNEMO: 4 or 5 Use shape 4 or 5: close in energy resolution, but shape 4 is slightly larger therefore there would be fewer channels (-33%) PVT (EJ-200) block shape study currently under way: remember 7% FWHM @ 1MeV for a hexagonal block! FWHM (Simulated) 9.9 ± 0.1% FWHM (Measured) 10.0 ± 0.1% 9.5 ± 0.1% 8.9 ± 0.1% 8.8 ± 0.1%

14. Calor 2010Anastasia Freshville14  Over more than 5 years of data taking for SuperNEMO the gain of 11,000 PMTs must be controlled at a 1% level  Detector response must be linear Any non-linear effect must be controlled at a 1% level up to 3-4 MeV.  Calibration systems being developed: Monthly absolute calibration (ADC → MeV):  207 Bi sources inserted into detector → 2 conversion electron calibration points: 500 keV and 1 MeV  Occasionally use 90 Sr: 2.3 MeV UV-LED based light injection system: gain and linearity monitoring Alpha and light sources embedded in scintillator: gain monitoring Occasionally use 60 Co for absolute time calibration (2 coincident γs of 1.1 MeV and 1.3 MeV) Calibration 207 Bi 482 кeV 976 keV FWHM = 135 keV (13.8%) ADC Number of events

15. Calor 2010Anastasia Freshville15 Calibration Linearity of 8” Hamamatsu PMT: UV-LED based light injection system: 1 MeV 3 MeV Q ββ value of 82 Se Extremely linear! Optical fibre carrying UV light inserted into scintillator to remove wavelength dependence of LED

16. Calor 2010Anastasia Freshville16 Calibration Alpha and light sources embedded in scintillator: Deposit a thin pure α or light source underneath the skin of the scintillator block The benefits of using α and light sources: Absence of additional electronics Long term stability Similar to light produced with electrons Pure α Sources used: 148 Gd (Е α =3.18 MeV) 238 Pu (E α =5.50 MeV, 5.46 MeV) 241 Am(Е α =5.49 MeV, 5.44 MeV) Light Source used: 3x3x1 mm  A micro source of 238 Pu (A O(10Bq)) deposited in the middle of a crystal non-organic scintillator  Surrounded by a teflon “case”  Light source placed underneath the skin of a scintillator block For the SuperNEMO demonstrator module: The first module will be equipped with a “redundant” calibration system: 207 Bi sources for relative calibration “UV-LED light injection” calibration system for linearity and timing Equip some blocks with α source embedding to test how effective it is and whether it is needed

17. Calor 2010Anastasia Freshville17 Conclusions  After 3 years of fruitful R&D an unprecedented energy resolution reached for large volume low Z scintillator: Design energy resolution for block design reached: Design energy resolution for bar design reached: This is a significant improvement from NEMO3:  NEMO3 scintillator block: 4 litres  SuperNEMO scintillator block: 8 litres  First SuperNEMO module/demonstrator to be constructed in 2010-2011, running in 2013 with 7kg of 82 Se  Demonstrator sensitivity: 0.2 - 0.5 eV  Full detector target sensitivity: 0.04 - 0.1 eV Size increase: factor of 2 Energy resolution improvement: factor of 2

18. Calor 2010Anastasia Freshville18 Backup Slides

19. Calor 2010Anastasia Freshville19 Bars: Coincidence Time Resolution Look at time difference between 2 TDC channels → time resolution Coincidence IN OUT TDC Common Start Stop PMT 1 PMT 2 Reset from Flip-Flop PMT 1 PMT 2 Coincidence timing tuned to TDC range Not delayed beyond TDC range EQUAL DELAY Coincidence time resolution: (confirmed by simulations)

20. Calor 2010Anastasia Freshville20 Bars: Coincidence Time Resolution  A sizeable contribution to time resolution is from “intrinsic time resolution” of the 3” Hamamatsu tubes (365ps compared to an average measurement of 650ps for Bar1)  The TTS (time transition spread: the spread of the time the electron avalanche takes to go through the dynode chain) for the 3” SBA- Select tube is 14ns ~3 worse than for an 8” tube (4ns)  Hamamatsu say they have a “plano-concave” tube with a flat window on the outside but a concave photocathode on the inside which reduces the TTS to 4ns (designed for another experiment)  There is no one-to-one correspondence between the TTS and the time resolution BUT We expect that reducing the TTS will help us to improve the time resolution  Tubes are available to purchase

21. Calor 2010Anastasia Freshville21 Square Root of E Dependence  Tested for blocks and bars  Results shown below: 2m bar with mylar wrapping, tested with 3” Hamamatsu PMTs Good /sqrt(E) result for both energy and time resolution!

22. Calor 2010Anastasia Freshville22 Demonstrator Module

23. Calor 2010Anastasia Freshville23 Demonstrator Module Design

24. Calor 2010Anastasia Freshville24 From NEMO3 to SuperNEMO

25. Calor 2010Anastasia Freshville25 SuperNEMO Schedule