Detailed performance evalua2on of a new 20- inch photomul2plier tube with a Box and Line dynode

Transcription

1 Detailed performance evalua2on of a new 2- inch photomul2plier tube with a Box and Line dynode 215 July 7 Yuji Okajima Yasuhiro Nishimura A, Ryosuke Akutsu A, Yusuke Suda B, Miao Jiang C, Seiko Hirota C, Daisuke Fukuda D, Masahiro Kuze, Masaki Ishitsuka, Masayuki Nakahata A,Masato Shiozawa A,Yoshinari Hayato A,Shoei Nakayama A,Hidekazu Tanaka A, Masashi Yokoyama B, Tsuyoshi Nakaya C,Akihiro Minamino C Akimichi Taketa E,Yoshihiko Kawai F,Takayuki Ohmura F,Masatoshi Suzuki F Department of Physics, Tokyo Ins2tute of Technology, Kamioka Observatory, Ins2tute for Cosmic Ray Research, University of Tokyo A Department of Physics, University of Tokyo B Department of Physics, Kyoto University C Department of Physics, Okayama University D Earthquake Research Ins2tute, University of Tokyo E Hamamatsu Photonics K.K. F 1

2 Mo2va2on of development Super- Kamiokande (SK) is successful experiment with rich physics topics focused on neutrinos and nucleon decay. Update plan, Hyper- Kamiokande is ongoing. 2 2mes larger detector than Super- K. - >Photodetectors (PDs) in Hyper- K are required to be low cost and high performance to be realized. Volume.99 Mton (SK: 5 kton) PDs (Inner detector) 99 (SK: 11146) PDs (Outer veto detector) 25 (SK: 1885) Photo- coverage 2 % (SK: 4%) Target Measurement of neutrino oscilla2on parameters Proton decay discovery Study of astrophysics such as supernova 2

3 Low dark rate Required performance Dark noise affects an accuracy of event reconstruc2on High rate tolerance In order to measure burst events like supernova Quick gain recovery To detect events with two con2nuous signal like muon and its decayed electron 3

4 2 inch φ (5 cm) Photomul2plier tube with Box and Line dynode New PD with different dynode shape from Super- K PMT (R36 with Vene2an blind dynode) has been developed. Vene2an blind dynode can bring sufficient collec2on efficiency, however, photoelectrons might miss 1st dynode and the path is varied. Features Path length of photoelectron is iden2cal - > Beher 2ming resolu2on Wide 1st dynode - > High photoelectron collec2on Simple structure - > Low cost Path through 1st dynode Vene2an Blind Many photoelectron path Iden2cal path Box & Line Wide 1st dynode 4

5 Improvement of Box and Line PMT Box & Line PMT (Hamamatsu R725) is currently used for KamLand experiment in 17- inch φ effec2ve region with 2φ bulb. The new Box and Line PMT for Hyper- Kamiokande was developed to reach 2 inch φ effec2ve area. Developed point Op2mize curvature and dynodes alignment to improve collec2on Added focusing electrode Enlarged detec2on area of first dynode 5

6 Efficiency New Box & Line PMT has high collec2on efficiency and quantum efficiency Quantum efficiency (%) Quantum Efficiency vs. Wavelength High-QE Box&Line PMT High-QE Super-K PMT Normal-QE Super-K PMT 5 R36 (Super- K PMT) (Normal photocathode) Quantum efficiency ~22% ~3% Collec2on efficiency (φ 46 mm (φ 498 mm)) Sensor Efficiency (φ 46 mm (φ 498 mm)) (Quantum Collec2on Eff.) 67% (61%) 95% (85%) Wavelength (nm) Box & Line PMT (High QE photocathode) ~15%(~13%) ~29% (~26%) 2 6

7 Performance Evalua2on 7

8 Pulse shape Owning to the development of new bleeder circuit, ringing effect of new Box and Line PMT is small. (Bleeder B) Fast signal than R36 Bleeder B was used for this performance evalua2on Pulse height (Normalized) R Time [ns] R36 Box and Line PMT (Old circuit) Box & Line PMT (Bleeder A) Box & Line Box PMT and Line PMT (Bleeder (New circuit) B) Rise 2me (1%- 9%) Fall 2me (1%- 9%) Pulse width (FWHM) R ns 13.2 ns 18.8 ns 2 High QE Box & Line PMT (Bleeder A) 2 High QE Box & Line PMT (Bleeder B) 6.2 ns 6.3 ns 1. ns 6.8 ns 15.2ns 13.2 ns 8

9 Gain, Transit 2me Gain ZB828 ZB821 ZB8243 ZB8246 ZB8248 ZB826 Transit 2me pc 2.2pC ( ) HV [V] Based on the R36 seqng at Super- Kamiokande, HV of each PMT is set to the value whose 1 p.e. gain is 2.2 pc. 9

10 Dark rate Number of photodetectors Dark rate HQE of DR HQE PMTs Dark rate [khz] Box & line PMT had showed about 2 khz of dark rate due to its structure. Old Box & Line New Box & Line HQE R36 Dark rate was reduced to ~1 khz By successful improvement of photocathode, arer pulse reduc2on and so on. Comparable level as R36 considering high QE and CE S2ll measured at room quickly, and expected to be lowered by low temperature and stabiliza2on. Dark Rate [khz] 2 1 Dark rate as a func2on of thresholds of improved Box and Line PMTs Dark Rate [khz] HV on Stabiliza2on of dark rate (Threshold: 2mV, 14 ) 1 1 Dark noise of Box and Line PMT 1mV Threshold 1 [mv] (Nominal threshold) PD of blue line /11 6/18 6/25 7/2 Date 1

11 Response at single photoelectron Entries (a.u.) R36 High QE Box & Line PMT Entries (a.u.) 1.8 R36 High QE Box & Line PMT Charge [photoelectron] Time (ns) High QE Box & Line PMT 1p.e. resolu2on σ/peak Peak/Valley Timing Resolu2on σ FWHM R36 53% ns 7.3 ns 35% ns 4.1 ns 1p.e. distribu2on Peak 1)p.e)resoluon Valley Charge 11

12 Time offset correc2on by charge Time derived at fixed threshold depends on the charge of signal. Time offset is corrected by observed charge according to its profile. Time [ns] (with arbitral offset) The peaks of 2ming distribu2on (Before correc2on) Charge [pc] Time [ns] The peaks of 2ming distribu2on (Arer correc2on) Charge [pc] 12

13 Time resolu2on in wide range of charge Time resolu2on is evaluated as a func2on of charge arer correc2ng 2me offset. Timing distribu2on is fihed by Exponen2ally modified Gaussian (EMG) p.e. Timing distribu2on correct χ 2 / ndf / 96 p ±.54 p ± 6.7 p ±.392 p3.528 ±.21 EMG func2on f(x) = e 2 2 (2µ+ 2 µ + 2x) 2 x erfc FWHM [ns] R36 R36 Box and and Line PMT Line PMT σ [ns] p 2 R36 and Line PMT Box and Line PMT FWHM σ Charge [p.e.] Charge [p.e.] Box and Line PMT showed beher 2ming resolu2on than R36. Especially, FWHM of Box and Line PMT is superior to R36 (2/3 at 1 p.e.). 13

14 Gain linearity measurement Two laser diodes (LDs) are used to evaluate the linearity of gain LD A: One for signal with fixed intensity LD B: One for signal with variable intensity Circuit for Linearity check Sync Clock Func)on* FANIN/ Generator FANOUT 1*[Hz] IO*Register 1*[Hz] Delay FANIN/ FANOUT Gate FANIN/ FANOUT Gate Coincidence FANIN/ FANOUT Gate:*1*ns,*1*Hz Gate ADC Variable:*1*[Hz] Op)cal* alenuator Reference* 1*[Hz] Coincidence Delay Delay LD LD The linearity was checked in a wide range of charge. 14

15 Result of gain linearity measurement Measured Charge [p.e.] : Expected Charge [p.e.] A devia2on in high p.e. region is due to satura2on of electrons in laher steps of dynodes. The linearity is kept up to 34 p.e. with gain drop less than 5%. 15

16 Rate tolerance Rate tolerance is necessary for Supernova measurement and so on (Supernova of Betelgeuse will bring O(1) MHz neutrino interac2on at maximum) Circuit for rate tolerance measurement Clock 1*[Hz] In order to evaluate this effect, two laser diode (LDs) were used as light source One for signal with 1 Hz One for background with a variable rate And measure change of signal pulse in different background configura2on. Cync Func)on* FANIN/ Generator FANOUT f*[hz] Change*I/O*of* background*ld IO*Register Delay Gate FANIN/ FANOUT Coincidence 1*Hz Gate Coincidence FANIN/ FANOUT Gate ADC Signal Delay Delay Measure*Charge LD LD Signal:*1*[Hz] Op)cal* a@enuator Background* fk1*[hz] A@enuator 16

17 Baseline shir The coupling condenser of the bleeder circuit can not discharge in high rate current, and resulted in shir of baseline voltage. V +HV Signal Baseline shir Coupling condenser Base Line Shift [mv] Voltage of baseline shir 25 p.e. 5 p.e. 1 p.e frequency [khz] The charge can be rela2vely taken referring to the baseline level before pulsing, same as in Super- K, therefore the shir can be excluded in terms of the rate tolerance at Hyper- K. 17

18 Result of rate tolerance measurement Attenuation Rate With base line shir excluded, the rate tolerance was measured x- axis: pulse frequency 25 p.e. 5 p.e. 1 p.e. AttenuationRate = frequency [khz] Sig.charge w/ bkg Sig.charge w/o bkg Attenuation Rate x- axis: charge current 25 p.e. 5 p.e. 1 p.e µa Gain kept stable, up to 17 μa for 5 p.e (gain drop less than 5%). For 1 p.e. signal, the rate can be up to 87 MHz by corresponding above current. Most of supernova neutrinos less than O(1) MHz at maximum bring low energy neutrinos, that is 1 p.e. level for PMT hits. - > Enough tolerance to measure supernova with constant gain. 18

19 Gain Recovery A quick gain recovery is important to iden2fy a delayed event in Hyper- K, such as μ decay. Gain stability was measured with two con2nuous pulses in various interval 2me. Result Gate The performance is evaluated by comparing delayed signal w/ primary and delayed signal w/o primary There were no any significant changes observed, and stable within.5% (comparable with sta2s2cal fluctua2on). 19

20 Summary For Hyper- Kamiokande, new Box and Line PMT was developed. Measured performance Pulse shape Result Faster and less ringing than R36 Dark rate Comparable level as Normal QE R36, with high CE & QE effect considered Response against 1 p.e. Timing resolu2on Rate tolerance Linearity Gain recovery Beher than R36 Keep un2l 17 μa (1p.e. signal: 87MHz) Keep un2l 34 p.e. Negligible even in nearest pulse The detailed performance evalua2on was done. Arer pulse etc. are shown in the next talk. As a result, it was shown that new PD has beher performance than current Super- K PMT. The performance is confirmed to be high and sufficient for Hyper- K 2