TESTS OF THE DO CALORIMETER RESPONSE IN 2-150 GEV BEAMS*

Kaushik De t Department of Physics University of Michigan Ann Arbor, MI 48109

Abstract At the heart of the DO detector, which recently started its maiden data run at the p~ , is a finely segmented hermetic large angle liquid argon calorimeter. We present here results from the latest test beam studies of the calorimeter in 1991. Modules from the central calorimeter, end calorimeter and the inter-cryostat detector were included in this run. New results on resolution, uniformity and linearity will be presented with and pion beams of various energies. Special emphasis will be placed on first results from the innovative technique of using scintillator sampling in the in- termediate rapidity region to improve uniformity and hermeticity.

INTRODUCTION study pp collisions at 1.8 TeV. The three ma- jor components of the detector include a cen- The DO experiment at the Fermilab Teva- tral tracking detector surrounded by a liquid tron uses a large angle hermetic detector to argon calorimeter, which in turn is enclosed in a magnetic tracking detector. A cut-out *Presented for the DO Collaboration: Universi- view of the calorimeter and central detector is dad de los Andes (Colombia), University of Aft- shown in Figure 1. sons, Brookhaven National Laboratory, Brown Uni- versity, University of California at Riverside, CBPF (Brasil), CINVESTAV (Mexico), Columbia Univer- sity, Delhi University (India), Fermilab, Florida State University, University of Hawaii, University of Illi- nois at Chicago, Indiana University, Iowa State Uni- versity, Lawrence Berkeley Laboratory, University of Maryland, University of Michigan, Michigan State University, Moscow State University (Russia), New York University, Northeastern University, Northern Illinois University, Northwestern University, Univer- sity of Notre Dame, Panjab University (India), IHEP- Protvino (Russia), Purdue University, Rice University, University of Rochester, CEN-Saclay (France), SUNY at Stony Brook, Superconducting Supercollider Lab- oratory, Tats Institute of Fundamental Research (In- Figure 1. Cut-out view of the DO calorimeter. dia), University of Texas at Arlington, Texas A & M University. tThis work was supported in part by the U.S. De- The DO calorimeter system is designed for partment of Energy. high resolution electromagnetic energy men-

9 1993American Institute of Physics 1685 1686 Tests of the DO Calorimeter Response

surement, good resolution jet energy measure- uranium plates with some steel and copper ab- ment, large angular coverage with fine seg- sorbers in the outer hadronic sections. Ioniza- mentation, and uniformity of response with tion from showering particles is collected in the very few cracks. The liquid argon calorime- liquid argon gaps. A single layer of scintillator ter modules are contained in three separate sampling is also used between the cryostats to cryostats. The central calorimeter covers the augment uniformity of response. pseudo-rapidity ()/ _= -lntan(0/2)) range of Longitudinally, the calorimeter is divided -1.2 < ~/< 1.2. The end calorimeters provide into electromagnetic, fine hadronic and coarse coverage between 0.7

DESCRIPTION OF THE DO CALORIMETER CENTRAL CALORIMETER RESPONSE

The DO calorimeter is built with a psuedo- We parametrize the energy response of the projective tower geometry pointing to the calorimeter as: nominal pp coLlision point. Each cell in the calorimeter has transverse dimensions of 0.1 in 2=C~+__+ps, N,p-T ~? and 0.1 radians in r Enhanced electron po- sition resolution is achieved through 0.05 x 0.05 where E and q are the mean and sigma from transverse segmentation in the third depth a gaussian fit to the measured calorimeter en- layer at electromagnetic shower maximum. ergy distribution in GeV at a beam momen- Alternating layers of absorbers and readout tum of p GeV. The parameters C and S are gaps sample particle showers as they traverse fitted s over a range of particle energies as the projective towers in the calorimeter. The shown in Figure 2. The noise term N is de- total interaction length ranges between 8-10. termined from the width of the pedestal dis- The primary absorber material are 3-6 mm tributions for the same cells used in measuring K. De 1687

in one layer of the coarse hadronic 0.4 .... I .... I .... I modules along with the response to random Constant = 0.0445 u Pion~s Sampling= 0.470 pedestal events. The minimum ionizing muon 0.3 Noise = Z.O peak is clearly separated from the random Constant ffi 0.0000 noise, with a most probable value of 41.0 -t- 0.7 ectrens SamplLng= 0.148 Noise ffi 0.42 ADC counts from a Landau fit. The full range of the digitizing system is greater than 32,000 ADC counts. 0.!

- _ ~ ,~"~, ""?"':'"':'*:"':"'v":--'.:.--;.-..:..-.~ , LOW ENERGY BEAMS 0.0 0 S0 100 t60 It is important to study the calorimeter re- Beam Momentum sponse to low energy particles since a large fraction of the energy of jets created at the Figure 2. Calorimeter response for and pions. coUider come from low energy particles. For the calorimeter energy. From the fit, we find the Load II run, a special tertiary low energy that the constant term for electrons is negligi- beam was created. Preliminary results T from ble, while for pions it is very small, C -- 0.045 the study of the calorimeter response in low GeV. The sampling term, S, which represents energy electron and pion beams show excellent the intrinsic resolution of the DO calorime- lineafity and resolution with particle energies ter, is 14.8% for electrons and 47.0% for pi- down to 2 GeV. The low energy response as a ons. All measurements were made in the cen- function of energy is shown in Figure 4. The tral calorimeter at 7] -- 0.05. Cuts using infor- mation from detectors outside the calorimeter .''''I .... I .... I .... I .... I .... were applied to clean up the sample of beam = o Electrons particles. 0 0 The DO calorimeter system has an excel- lent dynamic range with very little noise. In

2~0 . , .... I .... I .... 1000 4) Coarse Hadronic Uodule 500 S~O Fitted value = 41.0~-0.7

1rio I 4 $ e 10 Beam Momentum

100 Figure 4. Linearity of calorimeter response with low energy beam. 60

. | .... deviation of the measured energy from a linear 0 ' s 100 100 ~0 fit is less than 2% and is shown in Figure 5. ADC Counts THE INTER-CRYOSTAT REGION Figure 3. Calorimeter response to 15.9 GeV muons compared to pedestal distribution. At intermediate angles, 0.8 <[ ~ [< 1.4, Figure 3 we show the response e from 15.9 GeV showers propagate through both the central 1688 Tests of the DO Calorimeter Response

O.N .... i .... I .... I .... I .... I .... ~,~ : .... I .... I .... I .... 4ra 0 Electrons i fl - Total Energy !'~-~ CCL 1 .. w/o ICD & MG- r ~ i.. f L 100GoV,~ ! 0.00 | S Ira : S ' """~L -O.e~ , i ,-- ~ , - ! i--,t'~-~ t .... O 60 I ra lata

-0.04. ADC Counts

..,I .... [ .... I .... I .... I,...... | .... _ -0.~ Ira-- .... t .... 0 2 4 6 8 10 12 - Total Energy Beam Momentum .. w/o ICD & UG- IO0 GeV n .L ~? ffi 1.25 Figure 5. Deviation from 5nearity.

j , . . . . o and end calorimeters. Energy resolution in o 60 Iw this region is degraded by the presence of ADC Counts support wails, end plates of calorimeter mod- ules and cryostat wails, comprising 1-3 inter- Figure 6. Improvement in measured calorimeter en- ergy with the addition of the ICD and massless gaps. action lengths. In order to improve energy measurement, three additional layers of sam- resulting energy distribution is not gausslan pling are interspersed among the 'dead' mate- and can result in non-unlform calorimeter re- rial to sample shower development. Two of sponse. With the energy from the ICD and the the sampling layers are inside the cryostats massless gaps added in, the average measured and are called 'massless gaps.' The third energy is Gaussian and the energy resolution layer, attached to the wall of the end calorime- is dramatically improved. Improvement in en- ters, use scintillator modules, and is known ergy resolution at ~/ = 1.25 is also shown in as the inter-cryostat detector s (ICD). These Figure 6. We show in Figure 7 the overall ira- additional layers are treated as standard sam- piing layers in D(3 with the support plates and .... I .... I .... I .... I .... I'''" 4

cryostat walls acting as absorber plates. 0 !00 The improvement in energy measurement 80 and resolution vary as a function of ~ in the inter-cryostat region due to the changing pro- aO 1... file of absorber material. Significant improve- 4O X CAL sans ICD/MG ~ 1 ments are observed around ~/ -- 1.2 as one o Total CAL

approaches the edge of the central electro- 20 (Error bars show o') magnetic modules. In Figure 6, we show the ,,,.I .... I .... I .... I .... I... 0 improvement in the energy measured in the 0 0.28 O~ 0.715 1 1.~ !.6 calorimeter modules with the addition of the Psuedorapldity, massless gaps and the ICD at 17 --- 1.15 for Figure 7. Calorimeter energy measurement versus the a 100 OeV ~r beam. Without the additional pseudo-rapidity of the incident pion. sampling, almost 50% of the pion energy is not measured for a majority of the events. The provement in pion energy measurement across K. De 1689

a wide range of 7/'s. As seen from these figures, REFERENCES the use of the ICD and the massless gaps rep- resent a significant innovation in uniform and 1. S. Abachi et al., "Beam tests of the DO hermetic liquid argon calorimetry. uranium liquid argon end calorimeters," Fermilab/PUB/92-162 (1992), (to be pub- NEUTRAL BEAMS fished in Nucl. Inst. & Meth.), and refer- ences therein. Tertiary beams of photons and neutral pi- 2. H. Aihara et al., "Design, construction ons were created during the Load II run us- and performance of the Electromagnetic ing thin layers of converters. In Figure 8, we Module of the DO End Calorimeter," LBL- show the energy distribution of tertiary pho- 31378 (1992), (to be published in Nucl. ton beams compared to secondary electrons. Inst. & Meth.). The beam profiles are fairly well matched. De- tailed analysis of the neutral beam data is 3. N. Amos, "The performance of the DO End underway? Cap Hadronic Calorimeter Modules," Pro- ceedings of the 1st Int. Conf. on Calorime- ~ 15o .... I .... I .... I .... try in High Energy Physics, World Scien- - Electron tific (1991) 84. :: .... 4. A. L. Spadafora, "Test Beam Results from the DO Liquid Argon End Calorimeter Electromagnetic Module," in Proceedings of the FiRh Pisa Meeting on Advanced De- tectors, North Holland (1992) 279. 5. J. Borders, "Hadronic Resolution of the Central Calorimeter," DO Note #1502 s l (1992). ADC Counts 6. W.G.D. Dharmaratna & J. Robertson, "DO Central Calorimeter Response to Figure 8. Comparison of photon and electron beams at 25 GeV. Muons - Test Beam Load 2," DO Note #1501 (1992). CONCLUSION 7. B. Hirosky, "Low Energy Results from the DO 1991 Test Beam," DO Note #1499 The DO calorimeter shows excellent reso- (1992). lution and uniformity from test beam studies performed over a wide range of energies for a 8. K. De & T.L. Geld, "A First Look at Com- variety of particles. Good linearity was ob- parative Hadronic Energy Resolution in served extending to low energy beams. The the ICD Region from Test Beam Data," innovative implementation of the ICD and the DO Note #1272 (1991). massless gaps was shown to preserve hermetic 9. M. Tartaglia, "The e/r and 7r~ Ratios and uniform coverage in the intermediate re- Measured, and Monochromatic 7 and 7r~ gion between the cryostats which contain the Beams Explored in the DO Test Calorime- liquid argon calorimeter modules. Further ter," (to be published in Nuclear Physics studies are in progress, including the energy B). response to neutral beams.