Catania

1. 4thInternational Workshop on Very Large Volume Neutrino Telescope 13-15 October 2009 Athens, Greece Catania Performances of four super bialkali large area photomultipliers with respect to standard R7081 Emanuele Leonora on behalf of NEMO collaboration leonora.emanuele@ct.infn.it

2. Hamamatsu Super BiAlkali (SBA) photomultipliers Hamamatsu has realized a Super BiAlkali large area PMT (10 inch. 10 stages ) on the basis of the R7081 standard PMT ( standard QE ≈ 25% @ 400 nm) SBA photocathode has a quantum efficiency (QE) ≈ 35% @ 400 nm Standard and SBA PMTs have the same mechanical design and internal structures. Difference only in the photocathode Dimensions of R7081 (Hamamatsu datasheet) Quantum efficiency of a SBA and a standard R7081 (Hamamatsu internal communication) leonora.emanuele@ct.infn.it

3. Super BiAlkali potentiality The KM3Net consortium has simulated the effects of the quantum efficiency improvement on the neutrino detector effective area Top: Neutrino effective areas as a function of the neutrino energy for the same detector equipped with PMTs with three different photocathode QE. Circles = standard PMT ( QE max value 23% at 390nm) Squares = Super Bialkali ( QE max value 35% at 380nm ) Triangles = Ultra Bialkali ( QE max value 43% at 390 nm). The line reports the “reference effective neutrino ares”. Bottom: Ratio between the super bialkali and ultra bialkali with respect to the standard PMT The effects of the improved quantum efficiency were relevant at lower energy At 1 Tev, the neutrino effective area increased by a factor 1.5 using super bialkali photocathodes R.Coniglione et al. NIM A 602 (2009) 98–100 leonora.emanuele@ct.infn.it

4. Measurements performed The performances of 4 SBA PMTs were measured and compared with those measured on a batch of 72 standard R7081 PMTs. • Main measurement conditions: • Room temperature and atmospheric pressure • Attenuated laser source ( s.p.e. conditions, 410nm, 60 ps width, 15 KHz) • All PMTs at the same gain condition ( 5 x 107 ) • All PMTs powered by ISEG active base PH7081 sel. • Whole photocathode illuminated • Measurements performed : • increase of the quantum efficiency of the SBA PMTs with respect to standard PMTs • super bialkali secondary effects: • dark count rate / decay time • time and charge resolutions • fraction of spurious pulses leonora.emanuele@ct.infn.it

5. The SBA quantum efficiency increase The quantum efficiency increase with respect to standard PMT was measured by the comparison of the detection efficiency for each SBA PMT detection efficiency = n° of PMT detected events / n° of emitted laser pulses Detected events calculated on the transit time spectrum acquired with 1/3 s.p.e. threshold. Laser pulses emitted was measured by a PMT monitor. Comparison of the TT spectrum of a SBA and standard Considering a standard QE of 25%, the measured QE improvement in the SBA complies with Hamamatsu specification Sketch of the testing facility to illuminate the whole photocathode leonora.emanuele@ct.infn.it

6. Overall characteristics of the SBA PMTs • The overall characteristics were measured for each SBA PMT and compared with those of the standard • Dark count rate (threshold 1/3 s.p.e) • Transit Time Spread ( threshold 1/3 s.p.e. ) acquired by TAC FAST (25 ps per channel) • P/V ratio and sigma charge resolution, acquired by QDC Silena (0.17 pC per channel) Resume of the overall characteristics of the 4 SBA PMTs with respect to standard The SBA Time and Charge resolution were not worse than standard The SBA dark count rate is higher than standard PMTs (SBA mean value = 2900 Hz) leonora.emanuele@ct.infn.it

7. SBA decay time measurement The mean dark count (DC) decay time after a short light exposure was measured for 2 SBA PMTs and 2 standard PMTs Decay time: time required to achieve a stable minimum value of dark count rate after light exposure • Test procedure: • Tested PMTs were exposed to the ambient light of a neon lamp (3 seconds) without power supply • After 1 hour of darkness, powered ON at nominal voltage • The behaviour of the DC rate from the instant of power ON was measured Mean dark count rate behaviour vs. time from power ON Other published studies on SBA PMT confirm this behaviour SBA PMTs have a decay time of over 10 hours Standard PMTs have a decay time of around 6 hours leonora.emanuele@ct.infn.it

8. Spurious pulses definition and typical time distributions Spurious pulses are noise pulses time-correlated with the main PMT response • Pre-pulses (P.P.), appear 10-80 ns before the main pulse. Due to direct photo-effect on the first dynode. • - Late pulse (L.P.), appear 10-80 ns after the main pulse, in the place of it. Due to photoelectron backscattering on the first dynode. • Type 1 after pulse (AP1), appear 10-80 ns after the main pulse. Due to luminous reaction on the electrodes. • - Type 2 after pulse (AP2), arrive in the 80ns-16us after the main. Due to the ionization of the residual gas in the PMT. Time distributions of pre- and late pulses after pulse 2 after pulse 1 Arrival time of spurious pulses with respect to main pulses Time distributions of the AP1 and AP2 leonora.emanuele@ct.infn.it

9. The fractions of the spurious pulses The ratio of the number of spurious pulses divided by the number of the true pulses was measured for each SBA PMT Resume of the SBA fraction of spurious pulses compared with standard Comparison with standard PMTs : The fractions of pre- and late pulse were not worse The fraction of type 1 after pulses was increased The fraction of type 2 after pulses was considerably increased (mean SBA AP2 fraction = 11% ) This effect is confirmed by published studies on other SBA PMT leonora.emanuele@ct.infn.it

10. Type 2 after pulse as function of gain The fraction of the type 2 after pulse was studied as function of the gain, decreasing the HV Voltage As expected, the increase of the QE with respect to standard PMTs did not vary with the gain Fraction of the after pulse type 2 as function of the gain The type 2 after pulse fraction decrease almost linearly with the gain decreasing This trend was also measured in standard PMTs, and was confirmed by other studies This effect suggests the use of the photomultipliers with a gain lower than 5 x 107to reduce AP2 fraction leonora.emanuele@ct.infn.it

11. Type 2 after pulse time distribution The delay time of after pulse 2 from main pulse is correlated to atomic mass of residual gas  information about impurity inside the PMT The mean time distribution of the SBA PMTs was compared to the mean distribution of the standard PMTs • The shapes of the distributions were the same • The fractions of events within the peaks with respect to the number of total events did not vary Comparison of the AP2 mean time distributions for SBA and standard The super bialkali photocathode did not introduce other types of impurity with respect to the standard leonora.emanuele@ct.infn.it

12. SBA photocathode uniformity measurements The SBA uniformity was measured and compared to that of the standard by scanning the photocathode surface: • Dark box (2 x 1.7 x 1.5) m • Light source movement system, controlled by LabView code • 410 nm Laser Source pulsed in s.p.e. condition • 5mm light spot diameter, normal to PMT surface • Light monitor PMT, to measure the number • of laser pulses emitted Sketch of the light source movement system Picture of the dark box, movement system and a tested PMT leonora.emanuele@ct.infn.it

13. Standard and SBA local detection efficiency The photon detection efficiency (PDE) was measured on 324 points uniformly distributed on the photocathode for a SBA and a standard PMT Measure of the ratio of the detected events (1/3 s.p.e. threshold) on the number of emitted light pulses Local detection efficiency along vertical axis. Each plot is normalized to own max value Dashed line: declared active area for a 10 inch. PMT The SBA photocathode conserves the uniformity of the standard leonora.emanuele@ct.infn.it

14. Summary • The performances of 4 new large area PMTs (10 inch.) with super bialkali photocathodes was measured and compared with 72 standard PMTs R7081: • The quantum efficiency increase with respect to standard PMTs was between 32% - 40% (@ 410 nm) • The time and charge resolutions were not worse than that of the standard • The dark count rate was higher than standard (SBA mean value ≈ 2900 Hz, standard mean value ≈ 1400 Hz ) • The decay time was around 10 hours, higher than that of the standard (6 hours) • The fraction of type 2 after pulse was almost 3 times greater than that of the standard (SBA mean AP2 fraction = 11% , standard = 4% ) • The fraction of type 2 after pulse linearly decreased with gain decreasing • The quantum efficiency improvement did not vary with the gain • The super bialkali photocathode did not introduce other types of impurity inside the PMT with respect to standard • The SBA photocathode uniformity was not worse than that of the standard PMT However, the increase in the quantum efficiency overcomes the measured noise secondary effects leonora.emanuele@ct.infn.it

15. additional slides leonora.emanuele@ct.infn.it

16. How is it possible to boost the QE ? • Use of highly purified materials for the photo cathode (change from 99.999 to 99.9999 or even of higher purity ) • ( will provide less scattering length for e- , low recombination probability) • Tuning of the photo cathode thickness • Tuning of the material composition • Tuning of the anti-reflective layer • Tuning of the Cs layer thickness leonora.emanuele@ct.infn.it

17. Method to measure after pulses leonora.emanuele@ct.infn.it