1. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy1 Studies on High QE PMT Tadashi Nomura (Kyoto U.) Contents –Motivation –Performance of H7422P-40 –Application to Scintillation counter with WLSF readout –Summary

2. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy2 Veto detectors surrounding Decay Region Motivation K   experiment needs hermetic veto system Major K L decay modes –K L   +  -  0 (13%) –K L   e  (39%)  These may fake the signal if charged particles are missed Cause of inefficiency –  - p   0 n (all neutrals and lost) – e + annihilated with materials  Detection before these interactions (i.e. with low energy deposition) reduces the inefficiency Signature = 2  + nothing

3. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy3 Simple Consideration of Inefficiency In case of loss due to  - p   0 n reaction Cross section  (CH),max ~100mb (at the resonance peak: E  ~170MeV) Required inefficiency < 10 -4 Energy deposit ~ 2MeV/cm  Edeposit before P int =10 -4 ~ 40keV 6x10 23 [n/mol] / 104[g/mol] x 8[CH pair/n] x 1[g/cm 3 ] x 100x10 -27 [cm 2 ] x (thickness) = 10 -4  thickness ~ 200  m, Energy deposit ~ 40keV (Note: Resultant  0 can be detected by photon veto detectors and thus inefficiency might be smaller) How many photoelectrons per 40keV can we obtain? High Quantum Efficiency (QE) PMT desired

4. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy4 Example: KOPIO Downstream CPV Beam pipe liner inside pre-radiator / calorimeter Counter should be inside vacuum, but it’s desirable to locate PMT outside  Long WaveLength-Shifting-Fiber (WLSF)  Small light yield  Need High QE PMT

5. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy5 High QE PMT : Hamamatsu H7422 GaAsP photocathode Sensitive area: 5mm in diameter Metal channel dynode structure Price ~1.8k$ H7422P-40 Measured QE using LEDs (relative to bi-alkali PMT) Bi-alkali (H7415) GaAsP (H7422) QE=30-40% for Green Expected QE Calculated from relative QE and bi-alkali catalog value Peak Wavelength of WLSF

6. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy6 Basic Properties of H7422 Stable upto 1MHz gain drops above 200kHz Sensitivity map Structure due to focusing mesh was seen… (~10% dip) Light yield: 100 p.e. Light yield: 1000 p.e. Linearity Linearity not so good (even within 100 p.e.) Gain~6x10 6 Rate dependence Gain~10 6 Improvement might be possible by optimizing base circuit

7. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy7 High QE PMT + Scintillator + WLSF Test configuration –PMT: H7422P-40 –Scintillator: EJ-212 (ELJEN) 3mm-thick, 1m-long –WLSF: Y11(200) (Kuraray) 1mm-diameter –Machined groove, 1cm-pitch, bundle 7 fibers –Wrapped by Aluminized mylar 90 Sr (Edep~580keV) Results: 70 p.e (both) with High QE PMT (x 3~3.5 larger than with Bi-alkali PMT) 70 p.e. / 0.58 MeV  120 p.e. / MeV

8. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy8 WLSF attenuation A S =9.0, S =1.0m A L =12.9, L =6.1m Attenuation in WLSF alone –with LED-excited, viewed by High QE PMT – L =6.1m – A L /A S ratio decreases if measured by bi-alkali PMT (longer wavelength ~ longer attenuation) Attenuation in Scinti + WLSF –Consistent well with WLSF alone  In case of 4m long WLSF (1m in Scinti + 3m outside vacuum), Light yield will be 46% of our test result  120 x 0.46 x 40x10 -3 = 2.2 p.e. / 40keV

9. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy9 Further Effort to Increase Light Yield Use thicker WLS fiber Better acceptance of primary scintillation lights –1.0mm  1.5mm diameter  30% increase expected (by our measurement) Use clear fiber to transport light Longer attenuation –1m WLSF in Scintillator + 3m clear fiber (outside vacuum)  50% improvement expected Connection ~90% x attenuation ~75%  68% cf. 46% attenuation for WLSF Need large area PMT cathode to read a bundle of 7 fibers 5mm  8mm diameter

10. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy10 Summary (1) High Quantum Efficiency PMT Hamamatsu H7422P-40 (GaAsP photocathode) Basic properties –QE : 30-40% for green light ~3 times larger than bi-alkali PMT –Linearity : not so good if we use “default” base circuit –Rate capability : stable upto 1MHz for 100 p.e. light Application to Scintillator + WLSF –120 p.e. / MeV (sum of both end) with 1m-long test counter ~3 times larger than bi-alkali PMT, as expected

11. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy11 Summary (2) Application to KOPIO DS Charged Particle Veto –In case we use 4m long WLS fiber, 2.2 p.e. / 40keV (sum of both end) will be expected –Threshold of 3 p.e. in both end, for example (  6 p.e. in total) inefficiency will be 3x10 -4 (without help by “backup” photon vetoes)  Effort to increase light yield 1mm  1.5mm diameter fiber ( x 1.3 expected ) Use clear fiber to transport long distance Change reflection material (Aluminized mylar  Al evaporation?)

12. May 26-27, 2005Tadashi Nomura (Kyoto U), KRare05 at Frascati, Italy12 Further High QE PMT issues Large sensitive area desired –In order to use thicker fibers  Need negotiation with the vendor (Hamamatsu) Linearity might be improved –Not so good with “default” base circuit  Optimize base circuit Life time of photocathode? –Degradation of GaAsP ?  Long-term test is planned