1. Prototype Large-Angle Photon Veto Detectors for the NA62 Experiment at CERN Mauro Raggi (LNF) On behalf of the NA62 Photon Veto Working Group Crete 08 9 th International Conference on Applications of Nuclear Techniques Crete 8-14 June 2008

2. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20082 NA62 photon veto working group A. Antonelli, E. Capitolo, P.S. Cooper*, L. Iannotti, M. Moulson, M. Raggi, T. Spadaro Laboratori Nazionali di Frascati F. Ambrosino, M. Napolitano, V. Palladino, A. Romano, G. Saracino, P. Massarotti INFN and University of Naples R. Fantechi, G. Lamanna, S. Venditti INFN and University of Pisa E. Leonardi, M. Serra, P. Valente, F. Perfetto INFN and University of Rome “La Sapienza” * Visitor from Fermilab

3. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20083 Outline oThe NA62 experiment at CERN Large Angle Veto (LAV) requirements oPhoton veto technologies Scintillating Fiber prototype CKM scintillator tail prototype Opal Lead glass counters oPhoton veto prototypes performance Results of the test at the DAFNE BTF oFuture plans oConclusions

4. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20084 NA62: A search for K +  p + nn oBR(K + →   ) SM ~ 10 -10 minimal theoretical uncertainty oBR(K + →     ) ~ 21% need rejection factor 10 -12 Rejection is based on the detection of 2  coming form  0 decay oMaximum detection inefficiency for single g 10 -4

5. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20085 Large Angle Veto requirements o>20 X 0 to ensure shower containment oVery low detection inefficiency for  O(10 -4 ) down to 50-100 MeV oTime resolution better than 1 ns oEnergy resolution 10% at 1 GeV oOperation in high vacuum region: O(10 -6 ) mbar 13 rings along >120 m decay region

6. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20086 Possible technology choices oLead + Fiber (constructed @ LNF) oLead + Scintillator Tile (loan form CKM) oLead glass crystals (OPAL dismounted) Lead + FiberLead + TileLead glass crystals

7. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20087 Scintillating Fibers + Lead KLOE-type lead/scintillating fiber calorimeter 1-mm diameter scintillating fibers 0.5-mm thick lead foils Full-scale prototype of smallest rings 1/3 of final radial thickness Inner/outer radius: 60 cm - 72.5 cm Inner/outer length: 309.5 cm - 348.8 cm Readout granularity: 18 cells, 4.2  4.2 cm 2 16.8 cm (4 cells)8.2 cm (2 cells) ~13X 0 ~10X 0 3 cells All fibers (same as KLOE)Fibers + 1-mm Pb wires

8. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20088 Optical coupling of fibers to PMT oEnds of prototype milled for good coupling oLucite light guides with Winston cones oMilled and polished to 4.2  4.2 cm 2 o28-mm blue-extended Hamamatsu R6427 PMTs with E2624 bases Fiber pattern

9. Mauro Raggi, LNFCrete 08, Crete 8-14 June 20089 Lead + Scintillator Tiles This solution, designed by CKM Collaboration (FNAL), consist of a sampling calorimeter made by 80 layers that alternate 1 mm lead with 5 mm scintillator tiles. The ring, made by 16 sectors, was 16 X 0 deep. The scintillating layers were read out by Wave Length Shifters Fibers optically coupled to 4 PMT per sector. WLS Scintillator tiles Sector 2 Sector 1 Sector 2 Readout scheme of the CKM prototype PMT1 PMT2 PMT4 PMT3 PMT5 PMT7 PMT6 PMT8

10. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200810 Lead Glass solution This solution is based on the reuse of the Lead glass blocks form the barrel of OPAL electromagnetic calorimeter. To reach the inefficiency requests at least three modules (20 X 0 ) have to be involved in detection of each g

11. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200811 The Frascati Beam-Test Facility attenuation target dipole 1 slits dipole 2 to DAFNE Electron test beam parameters: Energy range100 to 750 MeV Selection resolution  E/E ~ 1% Mean multiplicity1 to 10 9 Bunch length10 ns Repetition rate50 Hz DAFNE LINAC e - line  line E 1e =483 MeV Typical single electron energy spectrum in e  line

12. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200812 The tagging system F1F1 F2F2 H1H1 H2H2 90 cm P1P1 P2P2 e - beam Finger counters (F 1, F 2 ) active - 6x8.5x1 cm 3 scintillator paddles Hole counters (H 1, H 2 ) veto - 33x10x1 cm 3 (e.g. H 2 ) paddle with 1.4-cm hole in center Beam profile meters (P 1, P 2 ) 1e F 1 ·F 2 Events E(MeV) F1F2H1H2F1F2H1H2

13. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200813 Summary of summer 2007 data taking Energy (MeV) P (1 e ) (%)  tag  1e  (%) Tagged events Fiber (KLOE) 20331.32.570k 35033.09.2210k 48333.314.4370k Tile (CKM) 20329.53.765k 35031.88.8220k 48329.017.6370k Lead glass (OPAL) 20330.23.925k 48326.017.190k Analysis nearly complete Analysis still preliminary

14. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200814 Analysis technique E A+B (MeV) Events 1-  = (7.3 +4.1 -3.3 )  10 -5 E A+B < 50 MeV Threshold E A+B (MeV) - Measured value n FC 68.27% CL (1 - e) Total events:8657175 1e fully tagged:68829 1.Calibrate the calorimeter and obtain raw energy distribution 2.Ask for full tag F 1 ·F 2 ·H 1 ·H 2 and count events with <50 MeV 3.Apply Feldman and Cousin recipe to evaluate efficiency limit vs Thr

15. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200815 Results fiber prototype Energy Resolution Time resolution (prel.) Beam Energy (MeV) Tagged Events F 1 F 2H 1 H 2 Event with E<50 MeV 1-  (Inefficiency) E<50 MeV 2036882957.3 +4.1 -3.3 x 10 -5 35020738531.4 +1.1 -0.9 x 10 -5 48337163312.7 +4.7 -1.7 x 10 -6

16. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200816 Results for CKM tiles prototype Beam Energy (MeV) Tagged Events F 1 F 2H 1 H 2 Event with E< 50MeV 1-  (Inefficiency) E< 50MeV 203 6516523.1 +3.5 -3.3 x 10 -5 35022116231.4 +1.0 -0.9 x 10 -5 483 19241215.2 +9.1 -3.3 x 10 -6

17. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200817 Results for the lead glass blocks oOnly four blocks tested oBeam impinging the center of block 1 oConfiguration different from the final assembly Beam Energy (MeV) Tagged Events F 1 F 2H 1 H 2 Event with E< 50MeV 1-   (Inefficiency) E< 50MeV 203 2506931.2 +0.9 -0.8 x 10 -5 483 9151111.1 +1.9 -0.7 x 10 -5 OPAL PbGl at the BTF

18. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200818 Summer ‘07 results comparison Fiber/KLOE Tile/CKM ▲ PbGl/OPAL Tile (CKM) and lead glass (OPAL) results are preliminary Efficiencies for electron detection similar for all 3 technologies (1 -  ) E meas < 50 MeV Assumes no false tags 68.27% FC CLs E beam (MeV )

19. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200819 2 nd lead glass test in February ‘08 oBlocks assembled in a way similar to the final one o25 block in 5 rows of five blocks each oDifferent impinging point tested: Central : on the center of block 33 On edge: on the edge of block 33 Beam Energy (MeV) Tagged Events Event with E<50 MeV 1-  (Inefficiency) Central 4712270314.4 +7.6 -2.8 x10 -5 On edge 471971111. +1.75 -0.65 x10 -4 -2.8 Preliminary not all the available data analyzed already

20. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200820 Roadmap to the final detector oWe decided to use lead glass as baseline Lead glass crystal quality test procedure established Validation of in vacuum electronic heating Measurement of blocks outgassing rates Mechanical design of vacuum tight ring started oSector prototype test in October @ CERN 5x5 crystal prototype with the final geometry and vacuum solution applied will be constructed Test in the NA62 decay region in vacuum oEntire ring prototype (160 crystals) To be completed at LNF before end of 2008

21. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200821 Final ring project oUnit element made by 4 blocks oEach circle made by 8 units of 4 blocks oEach circle is rotated by 1/5 block oEntire ring made by 5 circles (160 blocks)

22. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200822 Conclusions oTest beam have shown for all prototypes efficiencies for low energy electrons <10 -4 oThe efficiencies are better than required ones oMany quality test performed on blocks oThe Lead Glass blocks have been chosen as the baseline technology for the veto construction oA full lead glass ring prototype will be built in the upcoming months to finally validate performance in a direct test with the NA62 beam during 2009

23. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200823 Spares slides

24. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200824 E beam (MeV) Side A Side B ▲ Side A+B  E/E beam (%) E meas (MeV) Not easy to test quantitatively: BTF energy selection s E = 1% (included in error bars) Run-by-run energy scale fluctuations ~5% (not included in error bars) Perform run-by-run energy-scale calibration for resolution and efficiency measurements Linearity basically satisfactory at low energy Need readout with extended dynamic range in real experiment Linearity of fiber prototype

25. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200825 Energy resolution Perform Gaussian fits to 1e - peak (  1.5s) after run-by-run calibration Side A Side B ▲ Side A+B Example fits Side A+B 203 MeV 350 MeV 483 MeV Resolution s E /E (%) E (MeV) Using information from both sides: EE E 5.1% E (GeV)   4.4% =

26. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200826 Timing resolution - a first look Side A Side B ▲ Side A+B Resolution  t (ps) E (MeV) Study various differences: Slewing correction applied to both fingers and sides:        t  t  t  tt  t  t We obtain:  TAG = 147 ps  AB = 158 ps - common fluctuation Overall timing resolution (preliminary) with

27. Mauro Raggi, LNFCrete 08, Crete 8-14 June 200827 Quality test of the Lead glass blocks oOn 32 Lead glass blocks Absolute PMT gain Photoelectron yield The quality of the crystals in term of PeY seems uniform (0.4±0.08) photo electrons x MeV