Hadron Blind Detector Outline Hadron Blind Detector HBDat

Hadron Blind Detector 東京大学　小沢恭一郎

Outline • Hadron Blind Detector (HBD)　at PHENIX exp. • R&D at RIKEN (advertisement) • Outlook 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Hadron Blind Detector 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Realized example R. P. Pisani et. al, NIM A 400(1997) 243 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Realized example Total Charge [A. U. ] R. P. Pisani et. al, NIM A 400(1997) 243 0 2005/12/27 4 8 Number of Pad 12 PHENIX ws@RIKEN 小沢恭一郎(東大) 16

Testing Apparatus • Absolute QE of produced Photocathodes Vs Wavelength • Photoelectron signal in CF 4 (CF 4 transmittance) • GEM systematics (gain, stability, etc. ) • Systematic Measurements of Cs. I Coated Triple GEM (pe collection efficiency Vs ED, etc) Photocathode Test Independent Lamp Monitor Mg. F 2 Windows Photocathode or Triple GEM VUV beam Absorption Length ~ 30 cm Beam splitter/Collimator 2005/12/27 Vacuum Vessel (“cross”) PHENIX ws@RIKEN 小沢恭一郎(東大)

QE in CF 4 at High and Low Collection Fields 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) @BNL

光検出器としてのテスト Stainless steel box Pumped to 10 -6 before gas filling @Weizmann Gas gain Fe 55 x-ray UV lamp • Gains in excess of 104 are easily attainable • Voltage for CF 4 is ~140 V higher than for Ar/CO 2 but slopes are similar for both gases • Gain increases by factor ~3 for ΔV = 20 V • Pretty good agreement between gain GEM foils of 3 x 3, 10 x 10 PHENIX and ws@RIKEN 2005/12/27 小沢恭一郎(東大) measured with Fe 55 and UV lamp 2 Measurements: * Fe 55 x-rays * Am 241 source * UV lamp 25 x 25 cm produced at CERN

Cosmic ray tests: Experimental Set-up Cosmic trigger C S 1. S 2. S 4 C: CO 2 radiator • pth 3. 8 Ge. V • 1. 30 m long • rate 1/min Cherenkov response to S 1. S 2. S 4 S 1, S 2 S 1. S 2. S 4. C • 50 mip cm long CF 4 Radiator S 1. S 2. S 4. C “electron” • directly coupled to detector • triple GEM + Cs. I Detector Box • test with Fe 55, UV lamp, 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

response to “electrons” ED = 1 KV/cm (“collection”) Signal = mip + Cherenkov photons ED = -0. 5 KV/cm (“repulsion”) Signal = Cherenkov photons All spectra calibrated into pe using the response to Fe 55 x-rays. 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) @Weizmann

Discharge Probability Small GEMs: 3 x 3 cm 2 • Stability of operation and absence of discharges in the presence of heavily ionizing particles is crucial for the operation of the HBD • Use Am 241 to simulate heavily ionizing particles v In Ar-CO 2, discharges increase sharply when total charge is close to the Raether limit of 108 ΔVGEM HV segmented GEMs 10 x 10 cm 2 v In CF 4 discharges do not depend on the presence of particles. It seems that local defects are responsible for the discharges v CF 4 more robust against discharges than Ar/CO 2 v HBD expected to operate at gains < 104 2005/12/27 PHENIX below ws@RIKEN 小沢恭一郎(東大) i. e. with comfortable margin the discharge threshold ΔVGEM @Weizmann

Ion back-flow Absorber Mesh GEM 1 GEM 2 GEM 3 PCB E=0 Hg lamp Cs. I 1. 5 mm 2 mm Independent of gas Independent of Et p. A Fraction of ion back-flow defined here as: Iphc / IPCB Depends only on EI (at low EI some charge is collected at the bottom face of GEM 3) Ions seem to follow the electric field lines. In all cases, ion back-flow is of order 1!!! 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) @Weizmann

Aging Cs. I photocathode: * In spite of the large ion backflow there is no dramatic deterioration of the Cs. I QE. * For a total irradiation of ~10 m. C/cm 2 , the QE drops by only 20%. (The total charge in 10 years of PHENIX operation is conservatively estimated to 1 m. C/cm 2. ) Stability measurements performed during day 3 (4 m. C/cm 2 ), day 4 (3 m. C/cm 2 ), day 5 (2 m. C/cm 2 ). GEM foils: * During the whole R&D period we never observed aging effects (e. g. loss of gain) on the GEM foils. Total irradiation was well in 2. excess of 10 m. C/cm 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) @Weizmann

Hadron Blind Suppress electrons from ionization Apply Reverse field 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Hadron Blindness (I): UV photons vs. @Weizmann particles At slightly negative ED, photoelectron detection efficiency is preserved 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) whereas charge collection is largely suppressed.

Hadron Blindness (II) : response to @KEK mip ED = 1 KV/cm (“collection”) ED = -0. 5 KV/cm (“repulsion”) Average amplitude dropped by a factor of ~2. 5 and rate dropped by a factor of 12 Strong Hadron Suppression limited by ionization between GEM 1 and GEM 2. 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) Asymmetric operation ED = 0 GEM 1 GE M 2 GEM 3 PCB Suppressed ionization Full charge collection

HBD final design 2 x 21 HV connectors serving 2 x 3 detector modules Gas out Clearance +/- 3 mm Z= 656. 4 mm Removable window 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)FR 4 frame all around the cover

HBD exploded view 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

at Weizmann 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Full Scale Prototype Gas Tightness Test • • We started Nitrogen flow (200 l/h) with a single 50 um mylar window With single mylar we reached 15 -20 ppm water level With double mylar window we reached 5 -6 ppm Bypassing the HBD showed 2 ppm in the gas system On 19. 04. 05 we replaced 50 um window by 127 um mylar window coated with Al With this single window we reached the same 5 -6 ppm On 06. 05 we opened box and put into it several GEMs and resumed the Nitrogen flow I. Ravinovich 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

蒸着・組立 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

@ PHENIX signal electron partner positron needed for rejection e. Cherenkov blobs + e qpair opening angle ~1 m 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Performance in Run 9 Rejection factor Hadron few pe Projected Performance @ Run 10 Np. e. of single electron Single electron ~20 pe Signal significance 1. 4 /nb recorded improves effective statistics by ≥ 35 Luminosity [unit of Run 4 ] 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

R&D @ RIKEN 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Back ups

Scintillation of CF 4 v CF 4 scintillates at 160 nm. Two measurements in the literature: * NIM A 371, 300 (1996): 110 ph/Me. V * NIM A 354, 262 (1995): 200 ph/Me. V v Planned to be measured at BNL 2/2003 v Results of simple simulations: (using 200 ph/Me. V, QE=0. 3, Nch = 250) * signal/noise 10 * shades can reduce the noise by at least a factor of 3. 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

メモ Gas CF 4 CH 4 Ar C 2 H 6 Ar-C 2 H 6 N 0 940 185 255 170 200? E cutoff 11. 5 8. 5 9 7. 8 ? 光量　∝ N 0 / γｔｈ^2 * L Weitzman　HBD　40 p. e. L = 50 cm CNS 14 p. e. 2005/12/27 Γth 28 34 42 22 25 index 1. 000620 1. 000444 1. 000283 1. 001038 1. 000811 Pion: 198. 9 17. 3 electronに当る Electronは、1. 38倍 198. 9*1. 38 = 274. 5 (23. 9) Electron(measured): 284. 1 (24. 7) 差は、チェレンコフ分で、1 p. e. くらい PHENIX ws@RIKEN 小沢恭一郎(東大)

Weizmannでのテスト Detector Box Overall Set-up 50 cm long CF 4 Radiator D 2 UV Lamp Detector box 2005/12/27 2 produced at CERN GEM foils of 3 x 3 and 10 x 10 cm PHENIX ws@RIKEN 小沢恭一郎(東大)

Gain Curve: Triple GEM with Cs. I in CF 4: measured with Fe 55 and with UV lamp • Pretty good agreement between gain measured with Fe 55 and UV lamp. • Gains in excess of 104 are easily attainable. Fe 55 x-ray UV lamp 2005/12/27 • Voltage for CF 4 is ~140 V higher than for Ar/CO 2 but slopes are similar for both gases. • Gain increases by factor PHENIX ws@RIKEN 小沢恭一郎(東大) ~3

Total Charge in Avalanche in Ar-CO 2 and CF 4 measured with Am 241 Charge saturation in CF 4 !!! When the total charge in CF 4 exceeds 4 x 106 a deviation from exponential growth 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) is observed leading to gain saturation when the total charge is ~2 x 107.

Calibrated PMT Cs. I absolute QE § Many measurements of Cs. I QE in 6 -8 e. V range § No data beyond 8. 3 e. V § Measurements extended to 10. 3 e. V confirm ~linear behavior of QE Cs. I on GEM Bandwidth: 6. 2 – 10. 3 e. V PMT and Cs. I have same solid angle QE(Cs. I) = QE(PMT) x I(Cs. I) / [ I(PMT) x C 1 x C 2 ] C 1 optical transparency of mesh (81%) C 2 opacity of GEM foil (83. 3%) All currents are normalized to I(PMT-0) Extrapolation to 11. 5 e. V: N 0 ≈ 820 cm-1 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Discharge Probability • Stability of operation and absence of discharges in the presence of heavily ionizing particles is crucial for the operation of the HBD. • Use Am 241 to simulate heavily ionizing particles. In Ar-CO 2, the discharge threshold is close to the Raether limit (at 108), whereas in CF 4 the discharge threshold seems to depend on GEM quality and occurs at voltages VGEM 560 -600 V vs. ΔVGEM CF 4 more robust against discharges than Ar/CO 2. HBD expected to operate at gains < 104 i. e. with very comfortable margin below the discharge threshold 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) vs. Gain

Charge Collection in Drift Gap : Mean Amplitude 2005/12/27 Rate At ED = 0: - signal drops dramatically as anticipated. - rate also drops dramatically large hadron suppression PHENIX ws@RIKEN 小沢恭一郎(東大)

Length Budget Preamps: 0. 95 % Sockets: 0. 60% 2005/12/27 Total = 0. 92 + 0. 54 + 0. 95 +0. 60 = 3. 01 % PHENIX ws@RIKEN 小沢恭一郎(東大) CDR

Gas Quality Before and After Test Water “peaks” • There were no means to test the gas quality during the test, however we did test the gas quality before and after the test, under more or less the same conditions (i. e. , similar purging period and flow rates). B. Azmoun 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Photocathode B. Azmoun • Photocathode (produced at Stony Brook) was stored in a sealed vessel filled with N 2 (but possibly not the best environment) for two weeks prior to the test, and may have suffered some additional deterioration during its installation into the detector (we have no glove box !!). • The last time the GEM PC was measured, its integrated QE was ~ half that of the nominal value. 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Preliminary Measurements 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Photocathode Production Notes • 1 st USB Chicklet Produced @ Big Mac 2 nd USB Chicklet Produced @ Big Mac 1. Big Mac Vac = 2. 2 x 10 -6 Torr 2. No heating of substrate before, during or after evap. , No pre-evaporation of Cs. I 3. Thickness target = 2000 Angstroms 4. The chicklet was transferred to a paint can while still within Big Mac, in an Ar environment (little contact with air) 5. Chicklet was transferred to BNL within an hour of evap. within Ar filled can 6. Chicklet was mounted to QE testing flange and put under Vac with ~10 min exposure to air (had to solder on HV connector). 7. QE was measured in Vac (after 2 hours of pumping), Spectrometer Vac = 1. 0 E-5 Torr. 1. Big Mac Vac = 2. 1 x 10 -6 Torr. 2. Galvanized steel replaced with polished stainless steel. 3. No Chimney used 4. Pre-evap done ~ 10 min before deposition. 5. 5000 angstroms (instead of 2000) with about a 5% error 6. ~5 min elapsed before chicklet was vacuum sealed in the desiccator. (Desiccator sat in Ar as the PC was placed within it. ) 7. Used glass dessicator to transport PC 8. Used glove box (under N 2 purge) to transport PC from dessicator to test flange: ~2 min. exposure to air total 9. QE was measured in Vac (after 2 hours of pumping), Spectrometer Vac = 1. 0 E-5 Torr. Initial Chicklet Produced @ USB in Andrzej’s Lab 1. Bell Jar Vac ~ 7. 5. 0 x 10 -8 Torr 2. Thickness Target = 5000 Angstroms 3. Heated substrate before, during, and after evap 4. Pre-Evap of Cs. I 5. Placed PC in glass dissicator while in Ar atmosphere, pumped out dissicator, and transported to BNL 6. PC mounted onto test flange in air, total air exposure ~ 7 min. 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大) 7. QE was measured in Vac (after 2 hours of pumping), Spectrometer Vac = 1. 0 E-5 Torr.

Integrated Transmittance 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Big Mac Provides • Large Volume (8’ diameter, 6’ tall) • High Vacuum (few 10 -7 torr) • Lots of Feedthroughs – Lots of blank “do-it-yourself” ports • Mechanical Motions: – 2 tables covering 320 o in f – One tower (up/down & rotate) • Several Large ports – 12” Inside Diameter 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Mechanics Inside • All tables removable • One fixed table • Two rotating tables (320 o) • Target holder does up/down as well as rotation 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

CｓI蒸着 Top View • • Arrange 6 GEMs in ARC facing down. Rotates via Big Mac Tables. Evaporate one-by-one　5000 A (or 2000 A) ~5 min elapsed before chicklet was vacuum sealed in the desiccator　in the vessel. • (Desiccator sat in Ar as the PC was placed within it. ) • Used glass dessicator to transport PC Used glove box (under N 2 purge) to transport PC from dessicator to test flange: ~2 min. exposure to air total 量子効率を測定 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)

Thickness Uniformity Underside of Scaffolding Cs. I Thickness Target: 2000 Angstroms Al Foil Control Substrates for measuring Cs. I thickness r Quartz Crystal Thickness Monitor Quartz crystal thickness monitor 2005/12/27 Au-Ni-Cu clad G-10 Photocathode Substrates PHENIX ws@RIKEN 小沢恭一郎(東大)

Cs. I蒸着(Stony Brookの研究室の例） Objective: To develop and test the technique of Cs. I evaporation at Stony Brook for production of Cs. I coated GEM foils for HBD prototypes and possibly even the final detector. Method • Use high purity Cs. I (Scintillator grade) • High Vacuum (1 E-7 Torr) [diffusion pump w/ N 2 trap] • Thoroughly clean vessel and all components • Bake the Cs. I • Mask substrates • Evaporate very small amount of Cs. I • This vaporizes any contaminants on molybdenum boat and/or outer surface of Cs. I crystals • Vessel walls coated with Cs. I will also act as a “getter” • Warm up substrates before, during and after evaporation • Withholds water and contaminants from condensing onto the substrates before deposition, and onto the Cs. I after deposition 2005/12/27 • Thickness monitor • Quartz crystal oscillator, Al foil control substrate PHENIX ws@RIKEN 小沢恭一郎(東大) • Transportation/ Storage of Photocathodes (vacuum, gas flow)

Dimensions of Evaporator Vessel Scaffolding: Photocathode substrate/Thickness monitors Bell Jar Diameter ~ 43 cm Max. Height ~80 cm mask Electrodes/Cs. I crystal/ Molybdenum Boat 理研（放射線研）との協力で同様なものを進めていく予定。 2005/12/27 PHENIX ws@RIKEN 小沢恭一郎(東大)