Why CsI? Care & Feeding Get Ready for Physics. CsI: QWG3 Topical School. B Heltsley. Oct

Transcription

1 QWG3 CsI-1 Why CsI? Care & Feeding Get Ready for Physics CsI: QWG3 Topical School. B Heltsley. Oct

2 Credits Cornell, Syracuse, & Minnesota groups developed/maintained CLEO CsI & SW Cornell: C. O Grady, H. Worden, M. Sapper, B. Gittelman, N. Mistry, J. Kandaswamy, E. Nordberg,C. Bebek, bkh Minnesota: Y. Kubota, S. Lee, V. Frolov, T. Riehle, V. Savinov, S. Li, D. Gong Syracuse: T. Skwarnicki, H. Muramatsu, M. Artuso, S. Stone, And many more that I probably forget CsI: QWG3 Topical School. B Heltsley. Oct

3 History/Caveat I came to Cornell 9/84 from SLAC (lead/steel PWC sandwich e-m & hadron calorimeters) Joined CsI R&D, already in progress for >1 yr 5 GeV e - test SLAC, 180 MeV e + test Cornell, 2 NIM papers Had role in sw/calib/sim ever since Perspective is mine, not necessarily CLEO s! You may have better ideas! CsI: QWG3 Topical School. B Heltsley. Oct

4 Why CsI? Outline Requirements Technology decision Energy resolution CLEO II EMCal design Triggering considerations Down in Front! E. Blucher, et al., NIM A249 (1986) 201. Y. Kubota, et al., NIM A320 (1992) 66. CsI: QWG3 Topical School. B Heltsley. Oct

5 Why CsI: Requirements Early 80 s detector landscape Good tracking + poor calorimetry Mark III, CLEO I, Argus Poor tracking + good calorimetry Crystal Ball, CUSB CLEO II: have both! EMCal must fit outside ~1m radius tracking chamber Magnet coil outside EMCal, so EMCal must be compact, work inside B-field CsI: QWG3 Topical School. B Heltsley. Oct

6 Why CsI: Requirements Neutrals γ: E/E ~ 1-10%, φ~10 mrad from ~30 MeV to ~5 GeV K L, n: direction Charged e ± id: π ± suppression by ~1000 π ±, µ ± separation Veto showers caused by nuclear interactions: satellite showers, splitoffs : need shower shape info CsI: QWG3 Topical School. B Heltsley. Oct

7 Why CsI: Technology decision Lead-PWC sandwich ~>15%/ E (GeV). Old school. Need active absorber. Liquid Argon, Xenon Cryogenics! Container! $$ BGO too dense, too $, too little light NaI (Tl) λ too small (no PMT!), physically delicate (cleaves), hygroscopic (must keep DRY!) Scintillating Glass not dense enough, too little light CsI (Tl) λ good for photodiodes More plastic than NaI, easier handling (dents, not shatters) Less hygroscopic (no ultra-dry environ.) Can live with slow response, barely. CsI: QWG3 Topical School. B Heltsley. Oct

8 Spectral Properties CsI: QWG3 Topical School. B Heltsley. Oct

9 Material Properties Property CsI(Tl) NaI(Tl) BGO Scint. Glass Rad. Len (cm) Int. Len (g/cm 2 ) Rad Len/Int Len Density ( g/cm 3 ) Light Output (S11 PMT) Decay time (µsec) CsI: QWG3 Topical School. B Heltsley. Oct

10 Why CsI: Energy resolution Contributions Incoherent electronic noise (lo E) Coherent electronic noise (lo E) Rear leakage (depends on L, hi E) Transverse leakage (md. E) Response non-uniformity in z Clarity (free of bubbles, etc.) Polish, wrapping (air gaps important!) CsI: QWG3 Topical School. B Heltsley. Oct

11 Length: $ vs σ E (>1 GeV) CsI: QWG3 Topical School. B Heltsley. Oct

12 Gradient in z Positive gradient (more light per MeV at back of crystal) partially compensates for leakage, in a proportional way: if shower develops later, that energy is magnified. Assists at high energy. Low energy prefers ~0 gradient. Need to balance the two considerations. CsI: QWG3 Topical School. B Heltsley. Oct

13 Transverse Leakage Matters most at lo E This plot does not include elec. noise Need noise small enough to be able to sum more crystals Will come back to this for reconstruction CsI: QWG3 Topical School. B Heltsley. Oct

14 Side leakage & # summed 100 MeV 1 GeV 5 GeV Dramatic gains from 3x3 to 5x5 sum, but noise can offset the improvements, esp. coh. noise. CsI: QWG3 Topical School. B Heltsley. Oct

15 Simply Staggering! Vertex pointing is a good thing, but it can cause degradation of uniformity across crystal, & energy resolution, if excessive. CsI: QWG3 Topical School. B Heltsley. Oct

16 Put them all in the pot If noise can be reduced, # crystals summed can be increased, improving resolution from transverse leakage as well. CsI: QWG3 Topical School. B Heltsley. Oct

17 Shower Spread Test beam data: ~83% in central block if hit in center ~5% if 5cm away ~1% if 10cm away Above true at low & high energy Fluctuations around above numbers larger at low energy CsI: QWG3 Topical School. B Heltsley. Oct

18 CLEO II EMCal Design Crystal quality control Physical layout Triggering considerations CsI: QWG3 Topical School. B Heltsley. Oct

19 Crystal Testing CsI: QWG3 Topical School. B Heltsley. Oct

20 Acceptance Criteria Minimum average light output required. Gradients in z of 1% to +11% accepted. Manufacturers learned to meet these criteria with same test setup on-site. CsI: QWG3 Topical School. B Heltsley. Oct

21 Hermetic Coverage Slightly off vertex-pointing to avoid line of sight view of cracks or support fins. 90 o is the exception. CsI: QWG3 Topical School. B Heltsley. Oct

22 Container: Egg-crate Held together with rivets beware of spot-welding Belle container had to be re-assembled from scratch after spot-welds failed during crystal loading. CsI: QWG3 Topical School. B Heltsley. Oct

23 CLEO II Layout 48 in z 128 in φ TOF CsI: QWG3 Topical School. B Heltsley. Oct

24 CLEO III Layout Barrel Unchanged TF removed RICH Inserted DR Lengthened & Narrowed Endplate material thinned Endcap repackaged Pushed back 7 cm CsI: QWG3 Topical School. B Heltsley. Oct

25 Endcap Layout 820 Crystals/Endcap Rectangular cm 3 Some shaved at inner or outer ring to fit container CsI: QWG3 Topical School. B Heltsley. Oct

26 Readout 4 diodes/crystal Local preamps, 4 crystals/card Externally summed Shaped Digitized (usually multiple scales to preserve dynamic range) Sparsified To tape CsI: QWG3 Topical School. B Heltsley. Oct

27 Triggering Considerations CLEO: 4x4 crystal group is atom Need energy component & tracking component Useful to have at least 2 thresholds: min. ion. (~100 MeV), & higher one (~500 MeV) Pure energy, Pure tracking Independence allows beating them against each other to measure effic s on particular topologies Combined track(s) plus energy Tile sharing : overcome edge effects Want sharp turn-on at threshold, even at atom edges & corners CsI: QWG3 Topical School. B Heltsley. Oct

28 Why CsI: Down in Front! Material in front of EMCal reduces γ-efficiency & worsens E, φ resolution Far is worse than close CLEO was never successful in adding back TOF energy to showers to improve resolution Challenge for BES III CsI: QWG3 Topical School. B Heltsley. Oct

29 Conversions in CLEO III γ s convert at R~80 cm (RICH radiator) w/probability ~7% (~2% for R<80cm). When a γ converts at R=80 cm, it is irretrievably lost from mass peaks. CsI: QWG3 Topical School. B Heltsley. Oct

30 Photon recon. efficiency Good Barrel Extended Barrel Bad Barrel RICH Endplate Endcap For 1-photon MC events: rate at which biggest shower is in [-4,+3]σ E Losses mostly from conversions prior to calorimeter. CsI: QWG3 Topical School. B Heltsley. Oct

31 π 0 efficiency: MC study Examine probabilities for MC-generated π 0 to survive various cuts, in events with only 1 π 0 in the MeV momentum range, compared with π 0 s in qqbar events near 10 GeV Single π GeV Evts E γ inside [-4,+3]σ E 76% 70% Tight shower shape cut 96% 84% Loose shower shape cut 96% 96% No (tight) track match 100% 96% No (loose) track match 100% ~90% E γ >30 MeV 86%+ 86%+ CsI: QWG3 Topical School. B Heltsley. Oct

32 Material Lessons Less is more CLEO III drift chamber thinned its endplate: helped a lot! RICH radiator, endplate: hurts Belle phototubes: hurt The more material, the more crucial it is to model it correctly, & prove it! CsI: QWG3 Topical School. B Heltsley. Oct

33 Why CsI: Conclusions CsI does NOT stand for CLEO s Idea). Built upon NaI(Tl) success. CsI(Tl) became a std for e-m cals in mid- 80 s, inspiring BaBar & Belle, Crystal Barrel, KTeV (no Tl: fast), BES III, others Attains few % σ E /E, read out by photodiodes inside a B field, easy to machine & maintain, affordable. Rest of detector design should cooperate with CsI by keeping material low Next: Care & Feeding! CsI: QWG3 Topical School. B Heltsley. Oct