1 / 33

MEG II実験:液体キセノン検出器の 物理ラン開始に向けたコミッショニング(3)

MEG II実験:液体キセノン検出器の 物理ラン開始に向けたコミッショニング(3). 家城 佳 for MEG II collaboration 東大素セ. decay. History of LFV upper limits. MEG II experiment searches for cLFV decay,. Sensitivity goal: (10 times better than MEG) BSM prediction : (e.g. SUSY-seesaw).

yettam
Download Presentation

MEG II実験:液体キセノン検出器の 物理ラン開始に向けたコミッショニング(3)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. MEG II実験:液体キセノン検出器の物理ラン開始に向けたコミッショニング(3) 家城 佳 for MEG II collaboration 東大素セ

  2. decay History of LFV upper limits MEG II experiment searches for cLFV decay, . • Sensitivity goal: (10 times better than MEG) • BSM prediction : (e.g. SUSY-seesaw) If is found Discovery new physics!

  3. MEG II experiment Key concepts: - High rate continuous beam at PSI (- High resolution detectors to distinguish from accidental BG 900l LXe detector γ μ+ BG detector or e+ Signal BG(Accidental) e back-to-backsame timing Cylindrical drift chamber + Timing counter All detectors were installed in 2018! e (w/ partial readout electronics)

  4. LXe detector 900L liquid Xe (LXe) scintillatorto detect energy, position and timing of MEG II ~1% resolution at 52.8MeV In MEG II, entrance face is replaced from 216 PMTs (2 inches) to 4092 MPPCs (12x12 mm2) 1 m MEG ~2% 2 inch Light collection uniformity and granularity improved! x2 energy and position resolution improvement expected 12 mm

  5. Commissioning with 17.6 MeV Monochromatic 17.6 MeV from is used for commissioning. Energy resolution at 17.6 MeV: (Data) (MC) 17.6 MeV peak (4cm<conv. depth<8cm) Li target (Li2B4O7) Resolution depends on statistical fluctuation of Npho, EM shower shape etc. Why Data disagree with MC? 14.8 MeV broad peak • Problem in Data? • MC is too good? p beam (1MeV) CW accelerator

  6. Problem in Data?MC is too good?

  7. Example of events Distribution of Npho is compared between high energy event and low energy events. Averaged distribution of Npho(MPPC, projected to 1D) Example of Npho distributions at MPPC low E event high E event normalized to 1 low E high E (4cm<conv. depth<8cm) energy distribution No specific difference can be seen.

  8. Energy reconstruction procedure Charge integration of each channel Charge  Nphoton conversion with calibration constants Calculate sum of Nphoton Convert to energy example of MPPC waveform Example of event in 2018 is a coverage factor assigned area for this PMT Possible reasons of bad E resolution: noise, calibration …

  9. Noise? Effect of coherent noise can be measured in random trigger data.  Fluctuation is 0.6% of 17.6 MeV. Much smaller than 3%.

  10. Calibration? Energy distribution • How much does resolution change if we make mistake in calibration? (4cm<conv. depth<8cm) We compared two cases in MC Correct calibration: Use same calibration constants for all channels Wrong calibration: Apply calibration constants obtained in Data correct calibration wrong calibration Resolution become worse (1.21.3%) with (completely) wrong calibration, but it is not big enough to explain data.

  11. Stability? Nsum(around 17.6MeV peak) vs. run 1 Dec. 08:20 2 Dec. 19:30 Instability of gain, PDE might explain bad resolution. 0.2% difference in 1.5 days Time dependence is found to be too small to explain bad resolution. 3% resolution 2.5 hrs data

  12. Fluctuation of sensor response? There might be a event-by-event fluctuation in response of MPPC or PMTs (crosstalk+afterpulsing probability etc.) To distinguish such effect from fluctuation of sum of Npho, even-odd resolution is calculated like this: Such fluctuation does not seem to be big enough compared to MC to explain bad resolution in Data.

  13. PMT or MPPC? Npho (PMT, Data) Npho (MPPC, Data) We checked Npho sum distribution for both MPPC and PMT. Sigma/Mean: 0.069 Sigma/Mean: 0.052 Resolution looks bad for both MPPC and PMT.(Especially for PMT.) Npho (PMT, MC) Npho (MPPC, MC) Sigma/Mean: 0.036 Sigma/Mean: 0.034

  14. Problem in Data?MC is too good?

  15. Production of the Gamma-ray via narrow resonance reaction and its applications CW energy? Is CW peak really monochromatic? • width is only 12.2 keV. If the target is thin enough and the energy deposit of proton on target is negligibly small, energy fluctuation is negligible. EPJ Web Conf. Volume 142, 01019 (2017)

  16. Physics parameters in MC? Is there any parameters in MC which affect the resolution? We tried changing settings as follows: • Optical photon simulation • Reflection at PMT holders  off/on • LXe Absorption length 400cm  100cm • LXe Rayleigh scattering length 45cm  20cm • Cherenkov light • Angular dependence of MPPC PDE  off/on • Physics model • EM model  precise model option • Photo-nuclear effect  off/on Angle dependence of PDE photon θ Photo-nuclear effect None of these settings helped reproducing resolution of Data.

  17. Intrinsic resolution of LXe? Resolution depends on (at least) four parameters: (K Ni et al 2006 JINST1 P09004) Non-proportional scintillation response to electron may worsen resolution Fraction of escape electron may cause additional fluctuation Technique and Application of Xenon Detectors, pp. 17-27 (2003) Effect is expected to be ~0.5% @ 17.6 MeV Effect on resolution is not known above ~10 MeV.

  18. What do we expect at signal energy (52.8MeV)? Energy dependence of resolution is not fully understood. In MC, resolution simply depends on statistical fluctuation and EM shower shape fluctuation. However, an unknown component which we observed in MEG might still exist. Resolution vs. energy(MEG) Resolution was ~1.2% with MEG prototype LXe detector, so the intrinsic resolution is expected to be ~1% level or less. Energy resolution estimated from radiative muon decay spectrum was not bad (~1%), but we it needs to be confirmed with 55 MeV ()this year.

  19. Summary • Commissioning of MEG II LXe detector was done with 17.6 MeV monochromatic -source. • Energy resolution was measured to be larger than what we expect from simulation. • Noise and calibration problem does not seem to explain the difference. • Both MPPC and PMT shows larger fluctuation • Changing parameter settings in simulation does not help • Unknown component which affect resolution exists since MEG. • Resolution at ~52.8MeV needs to be checked this year with ~55 MeV calibration source.

  20. Backup slides

  21. Non-proportionality of scintillation yield Paper referenced by previous paper: It can be explained from non-proportional response of NaI scintillation yield as a function of energy. For NaI, energy resolution is known to be much worse than what we expect from statistical fluctuation of number of photon.

  22. Estimation of the effect of Non-proportionality Estimation based on measured non-proportional response to electron. Effect is smaller than 1% @ 52.8 MeV

  23. qsum2Weight • Energy resolution is calculated from nsum2. • Is qsum2Weight correct? • qsum2Weight should be defined for each PM like this: (empty area around PM) / (area covered by PM) DB value seems to agree with calculation based on drawing.

  24. Reconstruction? Coverage factor might not be optimal for data if we misunderstand the geometry of detector. We tried changing the coverage factor to find best value. Initial value fixed to 5 Initial value from DB • Best fit value seems to change if I change the initial value. Error is probably large because the resolution does not change significantly by changing parameters. • Values of Data and MC are roughly consistent with each other. • Deviation from DB can be seen for lateral PMTs.Lateral 0 (inner) is probably due to MPPCs which are not read out.I do not understand why other lateral PMTs are different from DB.

  25. Cerenkov radiation Cerenkov radiation is an another source of light in shower. In gem4, only 0.2% of the detected was found to be due to Cerenkov light.  Is it implemented properly? G4Cerenkov only works in the region where the refractive index is defined. Refractive index is defined only in 5-10eV by default in gem4.  Cerenkov light at low/high energy are NOT generated. Theoretical formulae: Cerenkov angle : Number of photons:

  26. Extending to low/high energy Refractive index of LXe can be calculated by using Sellmeier formula. In gem4, however, it was limited in . Sellmeierformula with parameters in gem4 THE JOURNAL OF CHEMICAL PHYSICS 123, 234508 (2005)  It seems fine to use Sellmeier formula in longer wavelength. • I decided to use • At shorter wavelength, absorption length at quartz becomes very short (0.25mm @ 155.5nm) • At long wavelength (> ~800nm?), PMT sensitivity becomes too small.  ~2.4% of generated photons are Cerenkov light. I ran gem4 with Quartz removed because absorption length is not known at long wavelength. However, energy resolution did not change.

  27. Photo-nuclear effect Photo-nuclear effect may affect E resolution? Hadronic cross section is large around 17MeV!

  28. Photo-nuclear effect Photo-nuclear effect is implemented in Geant4 physics list (e.g. QGSP-BERT). Photon cross section Photo-nuclear effect cross section (NIST XCOM database) ~0.3barn ( similar to 131Xe) ~10barn ( IAEA Photonuclear Data Library) Geant4 underestimates the effect of photo-nuclear effect. Anyways, however, it is NOT used in gem4 physics list. I added QGSP-BERT in gem4, but the energy resolution did not change at all. Photo-nuclear cross section is much smaller than gamma conversion cross section.

  29. Physics parameters in MC? I was not able to reproduce resolution in data in any of the settings.

  30. Resolution effect from escape electron Number of escape electron in ionization affects resolution.

  31. Resolution effect from escape electron Size of this effect is not known above 10 MeV.

  32. Resolution vs. energy Resolution (MC) vs. energy Energy resolution 1/sqrt(Nphe) sqrt(black2-red2) = effect of shower shape fluctuation etc.

More Related