EMC in BESIII Experiment Weiguo Li Representing BESIII

EMC in BESIII Experiment Weiguo Li Representing BESIII Collaboration Calor 2010 May 10, 2010 IHEP, Beijing 1

Outline Ø BEPCII /BESIII Ø EMC Design and Construction Ø EMC Performances Ø Summary 2

BEPC II Storage ring: Large angle, double-ring RF SR IP RF Beam energy: 1. 0 -2. 3 Ge. V Luminosity: 1× 1033 cm-2 s-1 Optimum beam energy: 1. 89 Ge. V Energy spread: 5. 16 × 10 -4 No. of bunches: 93 Bunch length: 1. 5 cm Total current: 0. 91 A Achieving high lum. with many bunches 3 And low

BES-III Magnet: 1 T Super conducting current 3400 Amp MDC: small cell & Gas: He/C 3 H 8 (60/40) xy=130 m p/p = 0. 5% @1 Ge. V d. E/dx=6% TOF: T = 100 ps Barrel 110 ps Endcap EMC: Cs. I crystal E/E = 2. 5% @1 Ge. V z = 0. 6 cm/ E Data Acquisition: Event rate = 4 k. Hz Total data volume ~ 50 MB/s Muon ID: 9 layers RPC 8 layers for endcap The detector is hermetic for neutral and charged particle 4 with excellent resolution, PID, and large coverage.

Expected Events productions per year at BEPCII Average Lum: L = 0. 5×Peak Lum. ; One year data taking: T = 107 s Nevent/year = exp L T Resonance Mass(Ge. V) CMS Peak Lum. (1033 cm-2 s-1) Physics Cross Section (nb) #Nevents/year J/ 3. 097 0. 6 3400 10 109 t+t- 3. 670 1. 0 2. 4 12 106 (2 S) 3. 686 1. 0 640 3. 2 109 D 0 D 0 bar 3. 770 1. 0 3. 6 18 106 D +D - 3. 770 1. 0 2. 8 14 106 Ds. Ds 4. 030 0. 6 0. 32 1. 0 106 Ds. Ds 4. 170 0. 6 1. 0 2. 0 106 5

BESIII collaboration: 43 Institutes US (6) Univ. of Hawaii Univ. of Washington Carnegie Mellon Univ. of Minnesota Univ. of Rochester Univ. of Indiana EUROPE (8) Germany: Univ. of Bochum, Univ. of Giessen, GSI Russia: JINR, Dubna; BINP, Novosibirsk Italy: Univ. of Torino，Frascati Lab Netherland：KVI/Univ. of Groningen Souel Nat. Univ. Pakistan (1) Univ. of Punjab ~ 300 collaborators Korea (1) China(26) Japan (1) Tokyo Univ. IHEP, CCAST, Shandong Univ. , Univ. of Sci. and Tech. of China Zhejiang Univ. , Huangshan Coll. Huazhong Normal Univ. , Wuhan Univ. Zhengzhou Univ. , Henan Normal Univ. Peking Univ. , Tsinghua Univ. , Zhongshan Univ. , Nankai Univ. Shanxi Univ. , Sichuan Univ Hunan Univ. , Liaoning Univ. Nanjing Univ. , Nanjing Normal Univ. Guangxi Normal Univ. , Guangxi Univ. Hong Univ. , Hong Kong Chinese Univ. 6 6

BEPCII Construction and Data Taking Dec. 2003, Project approved June 19 2008 first physics collision July 17, 2009, passed government review 7

So far, peak luminosity achieved ~3. 0 *1032 cm-2 s-1 BESIII reached designed performances Till now, data taking 106 M (2 S); 220 M J/ events are obtained; Currently run on psi(3770) with ~ 610 pb-1 so far Peak Lum. in 2010, at 1032 cm-2 s-1 8

March 25 8: 00 – March 26 8: 00 • Delivered collision beam for 19. 9 hours， • Data taking for 16. 8 hours • Online luminosity 12. 8 pb-1 9

Stable data taking, BESIII eff. > 80% Mar. 6 – April 14, 2009 May 24 – June 2, 2009 100 M (2 S) 45 pb-1@3. 65 Ge. V June 12 – Jul. 28, 2009 220 M J/ Jan. 17 – Apr. 12, 2010 450 pb-1 10

MDC, Good performance Wire reso. design： 130 mm Eff. : ~ 98% Beam related backgrounds d. E/dx design: 6% σP=11. 0 Me. V/c

TOF, Top time resolution Barrel Double Layer Time Resolution (ps) Z (cm) Time Resolution （ps） Design Target Bhabha Dim u Barrel Single Layer 100~110 98. 0 95. 3 Barrel Double Layer 80~90 78. 9 76. 3 Endcap 110~120 136. 4 95. 0

BESIII Cs. I(Tl) EMC, Design and Construction • • To measure the energy of electromagnetic particles Barrel: 5280 crystals，Endcap: 960 crystals Energy resolution Crystal: (5. 2 x 5. 2 – 6. 4 x 6. 4) x 28 cm 3 Babar: 2. 67% @1 Ge. V Readout: ~13000 Photodiodes, 1 cm 2 cm, BELLE: 2. 2% @1 Ge. V Energy range： 20 Me. V – 2 Ge. V position resolution: 6 mm@1 Ge. V CLEO: 2. 2% @1 Ge. V Tiled angle: theta ~ 1 -3 o, phi ~ 1. 5 o BESIII: 2. 5%@1 Ge. V Crystal calorimeter without supporting wall between crystals 13

Single crystal unit n n 2 Photodiode + 2 Preamplifier + (1 Amplifier) Photodiode(PD): Hamamatsu S 2744 -08 (1 cm x 2 cm) Preamplifier noise: < 1100 e (~220 ke. V) Shaping time of amplifier: 1 s

Crystal Production Have to check the crystal dimensions, light output, radiation dose sensitivity,

Light output and uniformity along crystal barrel: 5280 pieces • By PMT + 137 Cs • Requirement: LO > 33%; Uniformity < 7% • Quality control : LO > 35%; Uniformity < 7% n Light-output n uniformity

Photo diode ( PDS 2744 -08, 13200) checkout Measure the dark current, capacitance and quantum efficiency of each PD There is a LED-optical fiber system to monitor every crystal during construction and data taking. See Jian Fang’s talk.

Checkout of pre-amplifier , and match two in one crystal to similar gains The difference between the two preamps in the same crystal should be < 3%.

Quality Control of Crystal Radiation Hardness • Radiation hardness: after 1000 rads radiation decrease of light out <20% • 100 rads radiation decrease of light out <9% Sample check Most of crystals’ radiation hardness is good, some crystals unqualified were rejected 17 pieces of 210 samples have not passed Total we rejected 482

Electronics Design parameters Parameter Values Number of channels 6, 240 System clock 20. 8 MHz L 1 trigger latency 6. 4 μs Max single channel hit rate ≤ 1 k. Hz Equivalent noise charge (energy) 0. 16 f. C (200 ke. V) @80 p. F Integral non-linearity ≤ 1% (before corrections) Cross talk ≤ 0. 3% Dynamic range 15 bits Information to trigger Analog sum of 16 channels Gain adjustment range for triggers ≤ 20 % average noise of 384 channels 973 e.

EMC Electronics On crystals By detector 2 10 bit ADC Q-module CR-(RC)2 Main amplifier Preamp L 1 Tigger System 10 bit T/Q Info buffers V M E Range 16 ADC selection Fanout Test Control Use three 20. 8 MHz 10 bit ADCs to cover 15 bits required dynamic range, and provide 6 bits peaking time See Jinfan Chang’s talk on BESIII EMC electronics 21

　　 Barrel EMC assembly

Installation Barrel and endcap barrel weight : 54 ton Moving from stand No gap between crystals installing Barrel EMC Endcap　 assembly

Experience in EMC design and construction • Insure mechanical stability: calculation; matching drilling of crystal support and frame; support of whole EMC at the bottom; so far so good; • Good signal and noise control: insure good connection of cables and careful shielding and grounding; no crystal is lost so far, channels with only one FED from 2 to < 10 now; low noise, ~ 200 ke. V; • Co-operate with BEPCII people to control radiation dose to EMC; dose under control; 24

Radiation Dose Cs. I Crystal Calorimeter is the most expensive part of the detector, According to the design, the allowed radiation dose per year should be less than 200 Rads at crystals, ( at 1000 rads, crystal light should be > 80% of original) Pin diode can withstand more dose, Rad. FET has more Dynamic range and comparatively more stable

Results from pin diodes and Rad. FETs PIN Diodes 6 on east and west sides respectively Phi angles : 30° 90°… 330° 1 -6 : East; 7 -12 : West Are used for tuning the injection and beam orbit

Rad. FETs Barrel Endcap From Rad. FET, so far, for ~two years operation, average dose < 100 Rad, From detector calibration, the average drop of light < 3%.

Crystal radiation damage from the offline calib. Const 4. 10 -4. 25 psipp Machine study 6. 7 -7. 28 Jpsi 3. 7 -3. 27 Psip 1. 18 -3. 30 psipp 5. 25 -6. 2 3. 65 Ge. V 4. 2 -4. 14 Psip 2009 2010 So far acceptable, should be careful at higher beam currents, understand the reasons for some higher light loss ~15% ( radiation damage vs light coupling? ).

Changes from offline calib. Changes from LED

EMC in BESIII trigger Ø Trigger cell, barrel 4 x 4, endcap 15, thres. at 70 -80 Me. V; then form cluster, fully efficient at ~ 200 Me. V; Ø Trigger condition from EMC, Nclus, Etot, Clus. BB Etot_l 50% @ ~ 200 Me. V; 100%@~400 Me. V Etot_m 50% @ ~ 700 Me. V; 100%@~1000 Me. V (neutral events) For Etot_l For Etot_m Efficiency for trigger conditions for event total energy in EMC

Global Trigger tables Endcap bk-bk Charge 1 2 Barrel bk-bk Charge 3 Charge 4 Neutral NLTRK 1 -------- ---- YYY ---------- NLTRK 2 ----- YYY Y Y--- -------- YYY ----- STRK_BB YYY -------- ---------- LTRK_BB -------- Y----Y -------- NBTOF 1 -------- Y------ YYY ----- NBTOF 2 ----- YYY Y Y--- ---------- NETOF 1 YYY -------- ---------- BTOF_BB -------- ----- Y -------- NBCLUS 1 ----- YYY -------- YYY ----- NBCLUS 2 -------- ---------- NECLUS 1 YYY -------- ---------- NCLUS 2 -------- -------- YYY ETOT_L -------- ---- YYY ---------- ETOT_M -------- -------- YYY Y: 1 st data set (2 S); Y: 2 st data set J/ ; Y: 3 rd data set (3770), To reduce the trigger rate at (3770) (by a factor ~3), Charge 2 trigger is not used, still very efficient for hadron events importance of EMC in trigger

J/ data Etot_M is very efficient for neutral events

EMC calibration and monitoring ØBhabha events are used for normalizing the crystal gain Ø Radiative Bhabha and di-photons/ 0 are used for energy scale Ø Correct detector material important for data/MC agreement Ø LED system is used for monitoring the EMC conditions See Liu Chunxiu’s talk on calibration using Bhahba Bian Jianming’s talk on absolute energy calibration Operationally, EMC is on with power all the time, help to monitor the machine operation and make lum. measurement easier.

e 5 x 5 E 5 x 5 vs. Phi of Bhabha event @ boss 6. 5. 1 Lab CMS Data(black) Data(black ) MC(red) Phi In lab, calibrate to the MC expected energy DATA/MC consist with each other both in Lab. and CMS after Bhabha calibration.

Energy peak and resolution in CMS in different runs Energy peak Energy resolution 8447(3. 686 Ge. V) 9680(3. 65 Ge. V) 10138(3. 097 Ge. V) DATA and MC consist very well for Bhabha events, after the calibration with Bhabha

EMC Performances Ø No channel lost so far; Ø Low electronic noise; Ø Energy resolution and position resolution reached design values; Ø Gap effect at the boundary of crystals is small; Ø Timing information is very useful in rejecting background; Ø Energy reconstruction with TOF information, improve performance, especially for low energy showers;

Performance reach/exceed design Barrel energy resolution energy deposit for e+e- design： 2. 5%@1 Ge. V energy resolution for Bhabha events Position resolution for Bhabha design： 6 mm/ E 4. 4 mm@1. 89 Ge. V

Nice features Lowest electronic noise: < 200 Ke. V Photon detection: EMC+TOF Without TOF Using timing info. to reject bks. crystal Air gap center Energy resolution in gaps: minimum dead material

EMC energy resolution after energy correction at the boundary of crystals MC(3. 770 Ge. V) digamma bhabha Before correction 2. 59% 2. 40% After correction 2. 46% 2. 19% Bhabha data 3. 770 Ge. V 3. 686 Ge. V 3. 097 Ge. V Before correction 2. 57% 2. 50% 2. 56% After correction 2. 33% 2. 27% 2. 36% To be used in the physics analyses

’ c 1, 2 J/ l+l- (With TOF) ’ c 1 Data/MC difference Energy scale: 0. 5% Energy resolution: 5% E Etof ’ c 2 E Line shape have good DATA/MC consistency after using TOF energy E with/without TOF Eg with/without TOF Etof The DATA/MC agreement of TOF Energy indicates the calibration of TOF energy work well The tail of the line shape is reduced due to the use of TOF energy

Energy scale and resolution(With TOF energy) see Miao HE’s talk for details of EMC reconstruction Fit result of ’ c 2 J/ Emeasure/Eexp in radative Bhabha (solid-data, circle-MC) Data/MC Fit result of ’ c 1 J/ Difference in Emeasure/Eexp between DT/MC Energy scale ~0. 5% Energy resolution ~5%

Photon efficiency improvement with TOF energy Solid-Without TOF, circle-With TOF Photon efficiency increased significantly when E<0. 8 Ge. V For higher energy, the difference is smaller

Detection efficiency improvement

0 efficiency of ’ 0 0 J/ with/without TOF MC efficiency Mgg (0. 12 -0. 145 Ge. V) MC efficiency improvement ~12% circle: without TOF energy dot: with TOF energy DATA efficiency improvement ~12% circle: without TOF energy dot: with TOF energy 0 efficiency increase about 12% in low energy range

EMC is well understood, so the BESIII physics analyses based on EMC (neutral channels) are published 1 st,

(2 S)→ 0 0 , ( → , 0 → ) ’ 0 0 N c 0 : 17443± 167 N c 2 : 4516± 80 s co c 2 c 2 N c 0 : 2132± 60 N c 2 : 386± 25 c 2 c 0 co c 0 c 2 BESIII 3. 23± 0. 03± 0. 23± 0. 14 0. 88± 0. 02± 0. 06± 0. 04 PDG 08 2. 43± 0. 20 0. 71± 0. 08 CLEO-c 2. 94± 0. 07± 0. 32± 0. 15 0. 68± 0. 03± 0. 07± 0. 05 BESIII 3. 44± 0. 10± 0. 24± 0. 15 0. 65± 0. 04± 0. 05± 0. 03 PDG 08 2. 4± 0. 4 <0. 5 CLEO-c 3. 18± 0. 13± 0. 31± 0. 16 0. 51± 0. 05± 0. 03 BR (10 -3) 0 0 c 0 ’ Phys. Rev. D 81, 052005 (2010) CLEO-arxiv: 0811. 0586

E 1 -tagged ’ 0 hc, hc c 0 recoil mass spectrum in E 1 -tagged analysis Significance = 18. 6 M(hc)=3525. 40± 0. 13 Me. V N(hc)= 3679± 319 G(hc) = 0. 73± 0. 45 Me. V 2/d. o. f = 33. 5/36 Breit-Wigner convoluted with a D-Gaussian resolution + bkg. The mass and width of hc are allowed to float. The background is represented by the 0 recoil mass spectrum in the sideband of the E 1 photon and the normalization is allowed to float. 47

Inclusive ’ 0 hc Inclusive 0 recoil mass spectrum in ’ decay DATA inclusive Significance = 9. 5 N(hc) = 10353± 1097 2/d. o. f = 24. 5/34 The mass and width of hc are fixed to the values obtained from E 1 tagged analysis. The background is parameterized by a 4 th-order Chebychev polynomial, and all of its parameters are allowed to float. 48

Summary of systematic errors The total systematic errors are the square root of the sum of all systematic errors squared, at this stage, the systematic errors are somewhat conservative, can be reduced further 49

Results Phys. Rev. Lett. 104(2010) 132002 Combine the fully inclusive and E 1 -tagged analysis, we get: BESIII CLEO (E 1 -tagged) M(hc) 3525. 40± 0. 13± 0. 18 Me. V 3525. 35± 0. 23± 0. 15 Me. V G(hc) 0. 73± 0. 45± 0. 28 Me. V(<1. 44 Me. V at CL=90%) B( ’ 0 hc) ×B(hc c) (4. 58± 0. 40± 0. 50) × 10 -4 (G(hc) float) (4. 22± 0. 44± 0. 52) × 10 -4 (G(hc) fixed to 0. 9 Me. V) Br( ’ 0 hc ) (8. 4± 1. 3± 1. 0) × 10 -4 No measurement Br(hc c) No measurement (54. 3± 6. 7± 5. 2)% 50

Summary Ø BESIII EMC successfully built with very nice performances -- all channels working; Low noise; nice energy and position resolutions; -- Timing information is useful to reject background -- EMC is essential in BESIII trigger ØReconstructing energy with TOF information improves the performances Ø BESIII EMC has been understood well, physics papers are published mainly with EMC information

Thanks

0 DATA/MC in ’ 0 0 J/ with/without TOF energy p 0 daughter photon energy p 0 shape with/without TOF energy p 0 daughter photon energy in the TOF p 0 shape with TOF energy The tail of p 0 line shape is reduced after adding the TOF energy in to the shower energy

Novosibirsk function A: Normalization factor m 0: Peak t: describe asymmetry tail s: resolution

Assembly of the module Drill holes in the bigger end of crystal Once gluing 80 piece crystal Two pieces of PDs are glued together onto the plastic 1. 5 mm Drill machine Four holes (2. 8 mm )are drilled on the big end of the crystal The position of holes in different circle are different fixing the aluminum base plate 4 screws to Fixed Al base plate and preamp

LED-fiber monitor One crystal has one LED-fiber Check modules quality Monitor Radiation Hardness calibration energy n Scan energy： 10 Me. V-1. 5 Ge. V n Scan rate: 300 Hz n Stability：< 1% 10 min/run Electronics: n Control LED-pulse (10 point) n Scan address (10) n CLK L 1 self-trig before assembly of super module Test each cell by LED-fiber, if light output < 80% of PMT data, it will opened cell to check PD-crystal gluing and preamplifiers and so on.

East endcap Int. Dose of Crystals barrel The Int. dose at west endcap is larger than that at east. West endcap