19 April 2006Fabrizio Cei1 The Liquid Xenon Calorimeter of the MEG Experiment Fabrizio Cei INFN and Universita’ di Pisa Incontri di Fisica delle Alte Energie – IFAE 2006 Pavia, April 2006

19 April 2006Fabrizio Cei2 Outline Brief overview of the MEG experiment;Brief overview of the MEG experiment; The Liquid Xenon scintillation calorimeter;The Liquid Xenon scintillation calorimeter; The Calorimeter (Large) Prototype;The Calorimeter (Large) Prototype; Measured and expected performances;Measured and expected performances; Status of calorimeter preparation;Status of calorimeter preparation; Conclusions.Conclusions.

19 April 2006Fabrizio Cei3 Overview of the MEG experiment

19 April 2006Fabrizio Cei4 The MEG Experiment at PSI Search for Lepton Flavour Violating decay  e  The Paul Scherrer Institute  The most powerful continuous machine in the world;  Proton energy 590 MeV;  Power 1.1 MW;  Nominal operation current 1.8 mA.

19 April 2006Fabrizio Cei5 The MEG Collaboration ~ 40 FTEs

19 April 2006Fabrizio Cei6 The   e  decay – 1) Forbidden in the Standard Model of electroweak interactions because of the conservation of lepton family numbers.Forbidden in the Standard Model of electroweak interactions because of the conservation of lepton family numbers. With massive neutrinos (we know that m > 0 !) and mixing,   e  is allowed but at a negligible level (relative probability ~ )With massive neutrinos (we know that m > 0 !) and mixing,   e  is allowed but at a negligible level (relative probability ~ )

19 April 2006Fabrizio Cei7 The   e  decay – 2) All SM extensions enhance the rate through mixing in the high energy sector.All SM extensions enhance the rate through mixing in the high energy sector. SUSY ≈ Predictions in the range  SM Background negligible  SM Background negligible  clear evidence for physics beyond the standard model

19 April 2006Fabrizio Cei8 Historical perspective BR < 0.1 Present limit 1.2 x Improvements in physics linked with improvements in technology.

19 April 2006Fabrizio Cei9  e  : Signal and Background e+ +  e+ + e+ +  e+ +   e  = 180° E e = E  = 52.8 MeV T e = T  Signal   e  Background physical   e  e+ +  e+ + e+ +  e+ +  accidental   e    e  ee   eZ  eZ  e+ + e+ +e+ + e+ +  Accidental background dominant in the signal region

19 April 2006Fabrizio Cei10  e  : required performances Exp./LabYear  E e /E e (%)  E  /E  (%)  t e  (ns)  e  (mrad) Stop rate (muons s -1 ) Duty cyc.(%) BR (90% CL) SIN x x TRIUMF x x LANL x x Crystal Box x 10 5 (6..9) 4.9 x MEGA x 10 8 (6..7) 1.2 x MEG x x The sensitivity is limited by the accidental background BR (   e  allowed, but needed FWHM Need of a DC beam Some of them already fulfilled !

19 April 2006Fabrizio Cei11 MEG Experiment Layout Muon beam stopped in a 150  m target.Muon beam stopped in a 150  m target. Liquid Xenon calorimeter (800 l,  850 PMTs) for photon detection using scintillation light: fast response (~ 20 ns) and high light yield (~ 0.8 NaI).Liquid Xenon calorimeter (800 l,  850 PMTs) for photon detection using scintillation light: fast response (~ 20 ns) and high light yield (~ 0.8 NaI). (Thin wall) solenoidal spectrometer & drift chambers for positron momentum measurement.(Thin wall) solenoidal spectrometer & drift chambers for positron momentum measurement. Scintillation counters for positron timing.Scintillation counters for positron timing. Easy signal selection for  + decaying at rest e +  +  E e = E  = 52.8 MeV  e  = 180°

19 April 2006Fabrizio Cei12 Energy release in LXe Positron track Hits on TC One MC event

19 April 2006Fabrizio Cei13 The Liquid Xenon Calorimeter

19 April 2006Fabrizio Cei14 The Calorimeter – 1) Measurement of  energy, direction and timing Liquid Xenon properties Experimentalcheck

19 April 2006Fabrizio Cei15 The Calorimeter – 2)  Homogeneous 0.8 m 3 volume of Liquid Xenon - pulse tube refrigerator - pulse tube refrigerator - 67 cm < r < 108 cm - 67 cm < r < 108 cm - |cos(  )| < 0.35; |  | < 60 o    10 % - |cos(  )| < 0.35; |  | < 60 o    10 %  Only scintillation light;  Read by 846 PMTs (Hamamatsu): - 2 inches diameter; - 2 inches diameter; - Maximum photocathodic coverage on the - Maximum photocathodic coverage on the photon entrance face  43 %; photon entrance face  43 %; - Immersed in Liquid Xenon; - Immersed in Liquid Xenon; - Low temperature (165 o K); - Low temperature (165 o K); - Quartz window for matching with - Quartz window for matching with scintillation light wavelength (178 nm); scintillation light wavelength (178 nm);  Thin entrance wall;  Waveform 2 GHz for pile-up rejection. pile-up rejection.

19 April 2006Fabrizio Cei16 PMT R & D history - 1) The MEG calorimeter will work in an intense  background environment. Photons will be produced by several sources (muon radiative decay, bremsstrahlung, positron annihilation, neutron capture in Xenon and materials surrounding the detector …). We estimated that the light level due to the background would correspond to a few  A anodic current for a PMT gain G = At so high rates one expects undesired behaviours because of  increasing of photocathode resistivity at low temperatures;  changes in amplification when the average anodic current becomes comparable with the PMT base current (50  A). becomes comparable with the PMT base current (50  A).

19 April 2006Fabrizio Cei17 PMT R & D history – 2) 1 st generation: R6041Q 2 nd generation: R9288TB Photocathode: Rb-Cs-Sb K-Cs-Sb Material to reduce surface resistivity: Mn layer Al strips 165 o K: ~ 5 % ~ 15 % Photon high rate simulated by high frequency (tens of kHz) LED pulsing ON ONOFF OFF Overlinearity (hidden in the previous plot by Q.E. drop)

19 April 2006Fabrizio Cei18 PMT R & D history – 3) 3 rd generation: R9869Q Photocathode: K-Cs-Sb Material to reduce surface resistivity: Al strips (doubled) 165 o K: ~ 15 % Doubling the Al strips produces a better stabilization of resistivity at low temperatures. Insertion of Zener diodes in the last two stages of base amplification chain removes overlinearity Insertion of Zener diodes in the last two stages of base amplification chain removes overlinearity.

19 April 2006Fabrizio Cei19 Xenon purity – 1) Energy resolution strongly depends on scintillation light absorption: - reduced number of photoelectrons; - reduced number of photoelectrons; - loss of uniformity (combined with Rayleigh scattering). - loss of uniformity (combined with Rayleigh scattering). Xenon almost transparent to its own scintillation light, but possible contaminants can be very opaque …

19 April 2006Fabrizio Cei20 Xenon purity – 2) We developed a purification system to reduce impurities below ppb. Xenon is circulated in liquid phase (100 l/hour by means of a Barber-Nicols cryogenic fluid pump) and water vapor is removed by a purifier cartridge filled with molecular sieves. Purification performances Purification system tested and improved by means of the Calorimeter Prototype. (old system, new one is much faster)

19 April 2006Fabrizio Cei21 The Calorimeter Prototype (Large Prototype)

19 April 2006Fabrizio Cei22 The Large Prototype - 1)  Presently the largest Liquid Xenon calorimeter in the world: Xenon calorimeter in the world: 40 x 40 x 50 cm 3 ; 40 x 40 x 50 cm 3 ; ~ 70 liters of Liquid Xenon ~ 70 liters of Liquid Xenon  228 PMTs (types R6041 & R9288, (types R6041 & R9288, not the newest ones); not the newest ones);  Measurements: - cryogenic and long term - cryogenic and long term operation; operation; - absorption length; - absorption length; - energy, timing and position - energy, timing and position resolution. resolution.  Operating conditions similar to that of final detector. that of final detector.

19 April 2006Fabrizio Cei23 The Large Prototype – 2) LEDs  -source  - sources and LEDs for PMT calibration and monitoring Seen from inside

19 April 2006Fabrizio Cei24 The Large Prototype - 3)  Home-made Polonium alpha –sources mounted on alpha –sources mounted on 50 micron tungsten wires 50 micron tungsten wires (to be replaced by commercial (to be replaced by commercial Am sources, specifically Am sources, specifically developed by SORAD Ltd.); developed by SORAD Ltd.);  ~ 50 Bq for each source;  First application of this type of sources; preprint submitted of sources; preprint submitted to NIM. to NIM.  Already used for Q.E. determination. determination.

19 April 2006Fabrizio Cei25 Measured Performances of the Calorimeter

19 April 2006Fabrizio Cei26 Measurement of absorption length By using alpha sources (on walls and on wires) is possible to give a lower limit of the Xenon absorption length abs and an estimate of the light yield.  abs > 95 cm (95 % C.L.)  Light Yield ~ scintillation photons/MeV (0.9 NaI)

19 April 2006Fabrizio Cei27 Measurement of energy resolution Charge exchange reaction  - p   0 n    Liquid Hydrogen target to  Liquid Hydrogen target to maximize photon flux; maximize photon flux;   0 Frame:  monochromatic spectrum LAB Frame:  flat spectrum; LAB Frame:  flat spectrum;  Back-to-back configuration: E  = 55, 83 MeV; E  = 55, 83 MeV;  Even a modest collimation (  5 o ) guarantees a sufficient guarantees a sufficient monochromaticity (  E  0.3 MeV); monochromaticity (  E  0.3 MeV);  Need of an opposite side detector (a NaI array with LYSO preshower). (a NaI array with LYSO preshower).

19 April 2006Fabrizio Cei28 Experimental setup H 2 target LYSO Eff  14 % NaI beam S1 LP Eff (S1&& LP)  88 %

19 April 2006Fabrizio Cei29 Energy spectra in NaI & LP 83 MeV  55 MeV correlation 129 MeV line from  - p  n  (LXe sensitive to 9 MeV neutrons)

19 April 2006Fabrizio Cei30 Energy 55 MeV Event selection:  LP && S1 && (NaI + LYSO);  83 MeV line in NaI + LYSO  65 MeV < E NaI + LYSO < 95 MeV;  65 MeV < E NaI + LYSO < 95 MeV;  No saturated PMTs;  Collimator: r < 4 cm. FWHM:  E/E = (4.9  0.4) % Collimator PMTs

19 April 2006Fabrizio Cei31 Measurement of timing resolution Measurement of timing resolution high gain normal gain 110 psec 103 psec LXe – LYSO timing 55 MeV   t LYSO Beam   t LYSO Beam Normal gain 110 Ɵ 64 Ɵ 61 = 65 High gain 103 Ɵ 64 Ɵ 61 = 53 FWHM = 153 ps 125 ps  z  1  2 cm

19 April 2006Fabrizio Cei32 Status of Calorimeter Preparation

19 April 2006Fabrizio Cei33 PMTs: 850 PMTs underPMTs: 850 PMTs under testing in PSI and in Pisa testing in PSI and in Pisa LXe facility (3  4/day); LXe facility (3  4/day); Cryostat underCryostat under construction; delivery construction; delivery at PSI in spring 2006; at PSI in spring 2006; Gas System: almost readyGas System: almost ready in  E5 area at PSI. in  E5 area at PSI.

19 April 2006Fabrizio Cei34 PMT mounting

19 April 2006Fabrizio Cei35 Conclusions

19 April 2006Fabrizio Cei36  The MEG experiment is expected to start engineering runs in 2006; engineering runs in 2006;  Experimental tests with sub-detectors showed that many of the needed resolutions were already fulfilled; of the needed resolutions were already fulfilled;  For LXe calorimeter, we obtained an absorption length abs > 100 cm, an energy resolution  E/E 100 cm, an energy resolution  E/E < 5 55 MeV and a timing resolution of ~ 150 ps 55 MeV and a timing resolution of ~ 150 ps FWHM;  We successfully used alpha sources mounted on wires for calibration and monitoring of the detector for calibration and monitoring of the detector (the first application of these sources); (the first application of these sources);  The calorimeter building and the PMT testing and calibration are in advanced state. calibration are in advanced state.

19 April 2006Fabrizio Cei37 The MEG web page Please, visit it !

19 April 2006Fabrizio Cei38 Backup slides

19 April 2006Fabrizio Cei39 Comparison with LHC MEG “Supersymmetric parameter space accessible by LHC” W. Buchmueller, DESY, priv. comm. MEGA R&D Engineering Data Plans Data taking from 2007 on to reach sensitivity (90% CL) Obtain a “significant” result before the LHC era Eventually reach during LHC era Plans Data taking from 2007 on to reach sensitivity (90% CL) Obtain a “significant” result before the LHC era Eventually reach during LHC era

19 April 2006Fabrizio Cei40 Connection with -oscillations –1) MEG Experimental bound Largely favoured and confirmed by Kamland Additional contribution to slepton mixing from V 21, matrix element responsible for solar neutrino deficit. (J. Hisano & N. Nomura, Phys. Rev. D59 (1999) ) All solar  experiments combined tan(  ) = 30 O tan(  ) = 0 O AfterKamland

19 April 2006Fabrizio Cei41 Connection with -oscillations –2) Correlation between BR (  e  & sin 2 (  13 ) (still unknown !) In these models, BR (  e  is one of the most sensitive tool to measure sin 2 (  13 ). Sensitivity of future long-baseline experiments A. Masiero et al., hep-ph/

19 April 2006Fabrizio Cei42 MC Simulation  Geometry: full simulation of calorimeter structure, internal and external vessels, PMT holders and honeycomb;  Scintillation light tracking: - decay curve and wavelength spectrum of LXe scintillation; - absorption and Rayleigh scattering in Liquid Xenon; - Fresnel and total reflection on PMT quartz window and - Fresnel and total reflection on PMT quartz window and PMT holders; PMT holders; - PMT quartz window transmittance; - PMT quartz window transmittance;  Outputs: - Energy deposit, position and timing in Liquid Xenon; - Energy deposit, position and timing in Liquid Xenon; - Waveform output: hit timing of scintillation photons for - Waveform output: hit timing of scintillation photons for each PMT with digitizer binning. each PMT with digitizer binning. ( ～ 8 x MeV). ( ～ 8 x MeV).

19 April 2006Fabrizio Cei43 Calibration techniques - 1) The required performances of the detector demand multiple and complementary calibration and monitoring methods.  Alpha sources: Q.E. determination, LXe optical properties, energy scale stability, permanently installed within the calorimeter. energy scale stability, permanently installed within the calorimeter. But: no good energy reference, useless for timing; But: no good energy reference, useless for timing;   -lines from neutron capture in Ni (E  = 9 MeV): absolute energy scale, light yield stability, usable very frequently. scale, light yield stability, usable very frequently. But: useless for timing & Q.E., low sensitivity to optical properties; But: useless for timing & Q.E., low sensitivity to optical properties;  Charge exchange reaction:  - p   0 n (followed by  0 decay in two gamma’s): energy scale determination, absolute timing and position gamma’s): energy scale determination, absolute timing and position calibration, simultaneous calibration of the whole apparatus. calibration, simultaneous calibration of the whole apparatus. But: difficult to use frequently, hardware modifications needed, But: difficult to use frequently, hardware modifications needed, useless for Q.E., low sensitivity to optical properties; useless for Q.E., low sensitivity to optical properties;

19 April 2006Fabrizio Cei44 Calibration techniques - 2)  Cockroft-Walton Proton accelerator based on - Resonant cross section at E p = 440 keV (  peak = 6 mbarn,   15 keV); - Main method: simultaneous calibration of the whole apparatus, useful for absolute energy scale and monitoring, frequently usable. useful for absolute energy scale and monitoring, frequently usable.

19 April 2006Fabrizio Cei45 Alpha sources in LXe Measured Simulated 100  m thick tungsten wire 50  m thick gold plate clipped around the wire

19 April 2006Fabrizio Cei46 Radioactive Background in LP   -trigger with 5  10 6 gain;  Geometrical cuts to exclude  -sources;  Energy scale:  -source 208 Tl (2.59±0.06) MeV 208 Tl (2.59±0.06) MeV 40 K (1.42±0.06) MeV 40 K (1.42±0.06) MeV Other lines ?? Other lines ??  uniform on the front face;  few 10 min (with non- dedicated trigger);  nice calibration for low energy  ’s. 40 K (1.461 MeV) 208 Tl (2.614 MeV) Never seen before !

19 April 2006Fabrizio Cei47 Measurement of position resolution (40 MeV gamma beam with 1 mm collimator) Reconstruction by (localized) weighted average method

19 April 2006Fabrizio Cei48 Sensitivity T = s R  =  + /s Detector parameters  /4  = 0.09  e 0.9  e  0.9  sel (0.9) 3 = 0.7  sel  (0.9) 3 = 0.7   0.6    0.6 Signal N sig = BR T R m  /4   e  sel   Single Event Sensitivity SES = 1/(T R   /4   e  sel   )  Correlated Background BR corr  Accidental Background BR acc  R    E e (  E   2 (   ) 2  t e   Upper 90 % C.L. BR (   e  )  Discovery 4 events (P = ) correspond to BR = FWHM