R D on Cs I Tl APD Optimisation of

R&D on Cs. I(Tl) + APD Optimisation of the energy resolution M. Gascón I. Durán

R 3 B calorimeter Main characteristics (from R 3 B Tech Proposal) : Total absorption efficiency 80 % (Eg=15 Me. V lab) Very large crystals Eg sum s(Esum)/ <10% (same as DE/E) g Multiplicity s(Ng)/ DE/E for g < 10% Moliere Radius 2 -3 % DE/E for p (up to 300 Me. V) Calorimeter for p Scintillation properties Detector granularity 3% Large crystals Dynamic range Large dynamic range for detecting p, g Inner radius 35 cm ( house Si trackers) R&D on crystals + readout systems Design based on CAD + GEANT 4 full simulations see http: //www. usc. es/~genp -> calorimeter -> reports and talks

Calorimeter sectors Due to the Lorentz boost we identify three sectors: 130 45 15 8 8 ECM = 10 Me. V β = 0. 82 Demonstrator meeting, Madrid May-09

Crystals !? R 3 B/EXL meeting Milan Oct-06

Crystals M. Mozynski et al. , NIM A 485(2002)504 (API LAAPD)

St Gobain XP 3102 Standard 10 -stage, 25 mm tube, borosilicate glass, bi-alkali XP 1918 Standard 10 -stage, 19 mm tube, fused silica, UV bi-alkali XP 1901 Standard 10 -stage, 19 mm tube, lima glass, Green bi-alkali Hamamatsu Cs. I (Tl) Photonis Material we used for test 10 x 10, 10 x 50 , 10 x 100, 10 x 20 x 170 mm Cs. I (pure) 10 x 10, 10 x 50 , 10 x 100 mm La. Cl 3(Ce) 10 x 10, 10 x 50 , 10 x 100, 10 x 20 x 170 mm S 8665 -55 (CMS) 5 x 5 mm S 86641010(PANDA) 10 x 10 mm S 5345+ API 19 mm 2 UV extended (200 -1000 nm) R 3 B/EXL meeting Milan Oct-06

R&d on crystals and APDs Crystals from St. Gobain (France) tested coupled to LAAPD from Hamamatsu (Japan) and API (USA)

Data-taking setups Standard pulse-height analysis : - preamplifiers: Canberra 2001 A, Ortec R 142 - Shapping amplifier: Canberra 2022 - Multichannel analyzer: Ortec Maestro or Amptek MCA-8000 Pulse-shape analysis : - Oscilloscope Tektronix TDS 3054 B - Off-line Digital filtering: Trapezoidal finite-impulse algorithm Not used for APDs but providing the best results for PMT! R 3 B/EXL meeting Milan Oct-06

Experimental setup

Test at Genp-Lab (USC)

Test at Genp-Lab (USC)

APD gain vs. Bias Voltage 5 cm length The energy resoltution shows a minimum when The APD Bias Voltage=380 The energy resoltution depends on the APD gain and the Bias Voltage ~380 V Dgain/DV= 2. 84 %

Dependence on shapping time For small crystals 4 ms is a good compromise between energy-resolution and pile-up

Dependence on the acquisition time Best results achieved for acquisition times between 30 -60 s These curves conatin two effects: statistics and bias voltage drift. Could be solved by controling Temperature and Bias Voltage

Dependence on Amplifier Gain Energy resolution improves by increasing the amplifier gain

Best results

Study of the non-uniformity Energy resolution dependence on the first interaction point Non-uniformity ~ 8. 4 %, to compare with < 3% measured by St. Gobain Light collection uniformity CMS CAL P. Sempere Ph. D Thesis

Cs. I(Tl) Resolution Energy resolution improves for increasing energy Cs. I(Tl) with a 60 Co source (1171 ke. V) and (1332 ke. V) R 3 B/EXL meeting Milan Oct-06

Some conclusions • Several test (wrapping, optical coupling, amplifier gain, shapping time) have been performed to optimise the energy resolution of Cs. I(Tl) • APDs from API showed the best energy resolution for Cs. I(Tl) crystals • APDs S 8664 -1010 from Hamamatsu showed a very good energy resolution almost independent on the crysal size • The main drawback of APDs vs PMT is the strong dependence of the gain with Temperature and Bias Voltage How to improve the resolution: -Better stabilisation of both temperature and bias-voltage. - Better optical coupling --> Scionix - Surface coating --> Ukranian Institute for Crystals - New generation of scintillating crystals

T and V dependence Both Crystals and APDs response depends on Tº T must be kept very stable ! ~0. 1 o otherwise, T should be accurately measured in order to keep gain constant by controlling the bias voltage

Characterization of S 8664 -1010 APD T. Ikagawa et al. , NIM A 538(2995) 640 T stabilisation ~0. 1ºC keep the gain variarion ~ 0. 3%

Temperature stabilisation Refrigerator 4 -5 º C radiator Ceramics Cold focus ~ 10ºC T controller Resistor warm focus T probe T stable within 0. 1ºC Humidity control box < 30% (1%) 22ºC Allow pulse shape analysis with Tektronix TDS 3054 B

Further tests through prototype • Good results are only achieved when there is a perfect matching between the crystal exit surface and the APD entrance area Type 2 View of the Barrel average dimensions 3 x 1 x 13 cm Crystal type-2 Bi-frustrum shaped Waiting for an agreement with St. Gobain and IMPLanzhou(China) --> Price drop! Exit face fits to LAAPD S 8664 -2010 under negotiation First prototypes done at Lanzhou (China

Study of St. Gobain vs IMP Lanzhou crystals Tests performed with API APD Crystal dimensions 1 x 1 x 1 cm Very similar behaviour

Resolution achieved IMP Lanzhou St. Gobain

R 3 B Type Crystal

Uniformity of the crystal Wrapped with ESR (from 3 M), tested with 60 Co