A progress report on the development of the CsI-RICH detector for the ALICE experiment at LHC

Abstract

The particle identification in ALICE (A Large Ion Collider Experiment) at LHC will be achieved by two complementary systems based on time of flight measurement, at low $p_t$, and on the Ring Imaging Cherenkov (RICH) technique, at $p_t$ ranging from 2 to 5 GeV/$c$, respectively. The High Momentum PID (HMPID) system will cover $\sim$5\% of the phase space, the single arm detector array beeing composed by seven 1.3$\times$1.3 m$^2$ CsI-RICH modules placed at 4.7 m from the interaction point where a density of about 50 particles/m$^2$ is expected.\\ A 1 m$^2$ prototype, 2/3 of HMPID module size, has been successfully tested at the CERN/PS beam where 18 photoelectrons per event have been obtained with 3 GeV/c pions and 10 mm liquid $\mathrm{C}_6\mathrm{F}_{14}$ radiator. Mechanical problems related to the liquid radiator vessel construction have been solved and the prototype, fully equipped, will be tested at the CERN/SPS to investigate the PID capability in high particle density events.\\ In this report, after an introductory discussion on the requirements for PID in ALICE, the HMPID prototype is described and the main results of beam tests on large area CsI photocathodes, operated in RICH detectors, are given.

Full text

Page 1

EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH
CERN–PPE/97-96
10 July 1997
A PROGRESS REPORT ON THE DEVELOPMENT OF THE CsI-RICH DETECTOR
FOR THE ALICE EXPERIMENT AT LHC
A. Braem, M. Davenport, A. Di Mauro, B. Goret, G. Paic, F. Piuz,
J. Raynaud, J.C. Santiard, S. Stucchi, T.D. Williams.
CERN, Geneva, Switzerland
N. Colonna, D. Di Bari, D. Elia, R. Fini, L. Galantucci, B. Ghidini,
A. Grimaldi, E. Monno, E. Nappi, F. Posa, G. Tomasicchio.
INFN and Dipartimento Interateneo di Fisica dell’Universit´ a di Bari, Italy
The ALICE HMPID group
Abstract
The particle identification in ALICE (A Large Ion Collider Experiment) at LHC will be
achieved by two complementary systems based on time of flight measurement, at low
and on the Ring Imaging Cherenkov (RICH) technique, at
spectively. The High Momentum PID (HMPID) system will cover
the single arm detector array beeing composed by seven 1.3 ?1.3 m
placed at 4.7 m from the interaction point where a density of about 50 particles/m
pected.
A1m
beam where 18photoelectrons per event have been obtained with 3GeV/c pions and 10mm
liquid
tion have been solved and the prototype, fully equipped, will be tested at the CERN/SPS to
investigate the PID capability in high particle density events.
In this report, after an introductory discussion on the requirements for PID in ALICE, the
HMPID prototype is described and the main results of beam tests on large area CsI photo-
cathodes, operated in RICH detectors, are given.
p
t,
p
tranging from 2 to 5 GeV/c, re-
?5% of the phase space,
?CsI-RICH modules
?is ex-
?prototype, 2/3 of HMPIDmodule size, has been successfully tested at the CERN/PS
C
?
F
??radiator. Mechanical problems related to the liquid radiator vessel construc-
Presented by A. Di Mauro at the 7th Pisa Meeting on Advanced Detectors,
May 25-31, 1997 - La Biodola, Isola d’Elba, Italy
Submitted to Nuclear Instruments and Methods in Physics Research A

Page 2

1Introduction
ALICE (A Large Ion Collider Experiment) is the only LHC experiment fully dedicated
to the study of heavy ion collisions and it will address most of the hadronic and leptonic signals
of QGP (Quark Gluon Plasma) formation. In
interaction rates ( ? 10 KHz with about 10% of central events) and high multiplicities (
particles per rapidity unit) are expected. A detailed description of ALICE, whose detector is
shown in fig. 1, can be found in [1, 2].
???Pb-
???Pb collisions, at 2.76 TeV/nucleon, low
? 8000
MUON CHAMBERS
L3 MAGNET
ITS
TPC
TDF-PID
HMPID
ABSORBER
DIPOLE MAGNET
MUON FILTER
PHOS
Figure 1: The ALICE detector layout. It consists of two main parts: a central device, embedded
in the L3 magnet, with the Inner Tracking System (ITS), the TPC, the TOF barrel and the single
arm HMPID and Photon Spectrometer systems, and a forward part with the muon spectrome-
ter and two zero degree calorimeters (not shown in the figure) enabling the trigger on central
collisions.
Particle identification (PID) is an important design feature in ALICE, because of the large
spectra of momenta and masses that will be measured. In the momentum range below 2 GeV/
where the bulk of hadrons is concentrated, the PID will be accomplished by TOF measurement
in a barrel system. Only 3% of the produced hadrons will have
which makes this region accessible only to inclusive measurements. A dedicated High Momen-
tum PID (HMPID) system has been designed to measure inclusive particle ratios and
c,
p
tlarger than 2 GeV/ c, a feature
p
tspectra
1

Page 3

in the range 2?5 GeV/ c. It is a single arm array of seven 1.3 ?1.3 m
ing about 12 m
the tilting of the modules decrease the scatter of the incident angles of the particles (limiting
them to a 0
the radiator and improves the identification capability. The RICH modules are placed at 4.7 m
from the interaction vertex, position corresponding to multiplicities
ing expected background).
To verify the PID capabilities in such conditions of particle density a 1 m
built to be tested at the CERN/SPS. In this report a description of the prototype and main test
beam results on large area CsI-RICH detector are presented.
?CsI-RICH modules cover-
?, namely 5% of full central acceptance (fig. 2). The detector segmentation and
o
?15
ospread); this minimizes Cherenkov photons total internal reflection losses in
? 50 particles/m
?(includ-
?prototype has been
Figure 2: Perspective view of the HMPID system with the 7 RICH modules placed on top of
the barrel frame and tilted according to the distance from the interaction point.
2The HMPID RICH prototype
The RICH detector has a proximity focusing geometry, with 10 mm liquid
diator, suitable for the momentum range of interest, coupled through a 5 mm quartz window
to a multiwire chamber (MWPC, 4 mm anode wire pitch, 2 mm anode-cathode gap) with a
cathode segmented into 8 ?8 mm
film evaporated on the pads converts the Cherenkov photons into single electrons originating
avalanches at the MWPC anode wires. The analog readout of the pads allows accurate localiza-
tion through centroids evaluation. The use of a solid photoconverter permits thin sensitive gaps
in the MWPC, resulting in a considerable reduction of the background from ionizing particles.
This feature is mandatory with the expected multiplicity which makes the pattern recognition
capability a prime requirement for our application. The photodetector can even be used as an
additional tracking plane with good position resolution [1, 3].
C
?
F
??ra-
?pads for 2-D readout (fig. 3). A 500 nm photosensitive CsI
The HMPID requirements make acceptance less crucial than performance, thus allowing
some dead space in the detector plane. Therefore, to optimize the production procedure and to
ease the handling, the photocathode(PC) and the radiator have been segmented into 64
panels and 130?42 cm
The HMPID prototype is 130?85 cm
cessfully tested at the CERN/PS test beam, demonstrating not only the good operation of the
MWPC, having 1.3 m long anode wires and a support line in the middle, but also the feasibility
and the mounting under clean controlled specifications of 64
will be completely equipped, with full size radiator vessels, to be tested at the CERN/SPS.
The designed radiator vessel consists of a NEO-CERAM (NIPPON GLASS, Japan) tray closed
by a fused silicaSUPRASIL window.The two materials haveequal thermal dilationcoefficients
?40 cm
?
?vessels (fig. 4).
?, namely 2/3 of one RICH module size. It has been suc-
?40 cm
?CsI PCs. The prototype
2

Page 4

????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
????????????????????
CsI film
10 mm
5 mm
70 mm
4 mm
freon radiator
quartz window
collection
wire electrode
proximity gap
wire cathode
sense wires
pad cathode
with 500 nm
charged particle
frontend
electronics
Figure 3: Schematic cross section of the HMPID CsI-RICH detector.
and the resulting container is characterized by the stiffness required to cope with the
drostatic pressure. The vessel strain under hydrostatic load, in every position foreseen by the
HMPID detector design, has been measured on a full size vessel and the results have confirmed
the mechanical design expectations.
The ALICE low interaction rates allow multiplexed analog readout based on full custom elec-
tronics. Currently the frontend electronics of the RICH prototype is based on the 16 channel
GASSIPLEX chip [4]. In the final detector version it will be based on the MultiChip Module
(MCM) DIGITPLEX, composed by 4 GASSIPLEX’s , the 10 bit ADC CRIAD and the zero
suppression chip DILOGIC on a board [5]. The specifications are: ECN
input, baseline restoration
in 3?s.
C
?
F
??hy-
? 800
e
?at 20 pF
? 0.5% after 3
?s, data ready after zero suppression for 64 channels
3Test beam results on large area CsI-RICH detectors
The HMPID requirements have directed the choice towards a fast-RICH configuration
with CsI PC, which has been made possible by the CERN/RD26 project activities demonstrat-
ing the feasibility of large area CsI photocathodes with high and stable quantum efficiency [6].
The QE of 29 large PC’s, produced at CERN, has been evaluated, exploiting the Cherenkov
radiation in a RICH detector in test beam. A detailed presentation of the RD26 results can be
found in [7].
Fig.5 shows the CsI QE measured at 170 nm as a function of the PC production number, sum-
marizing the results of 4 years activity. The improvement can be attributed to the definition of
a new procedure to produce the photocathodes. Firstly, the substrate is obtained by chemical
deposition of nickel and gold layers on top of the copper pads, previously mechanically and
chemically polished. Then the CsI is evaporated at 50
oC and 10
??Torr, and the PC is kept
3

Page 5

????????
????????
????????
????????
????????
????????
????????
4 mm spaced
PAD CATHODE SEGMENTATION
wire support line
64 cm
130 cm
42 cm
RADIATOR SEGMENTATION
40 cm
1.3 m sense wires
Figure 4: Segmentation of pad cathode and radiator.
under vacuum at 50
Another important feature is the QE stability: two PC’s, among the best ones, kept under argon
(between beam test periods) and periodically tested, show QE degradation less than 5% in two
years (fig.6).
Fig. 7 shows typical 3 GeV/c pion events obtained at the CERN/PS test beam. Each PC is tested
at several MWPC gains and Cherenkov radiator thickness and the photodetector performance
is evaluated by means of dedicated analysis and simulation programs. The main event quanti-
ties are extracted by the analysis program from a fiducial region of the PC plane where all the
Cherenkov photons hits are expected (figs. 8 and 9).
Then these quantities are reproduced by the Monte Carlo simulation, allowing to estimate CsI
photoelectric yield and the photon feedback contribution. An essential feature, that the simula-
tion program has permitted to observe, is the spatial distributionof feedback photons: due to the
small anode-cathode gap and the pad size, they overlap to the cluster originated by the primary
Cherenkov photon, thus resulting in a very small background.
The main test beam results can be summarized as follow:
1) a very stable, discharge-free, operation of the MWPC at a gain of
detection efficiency of 95%;
2) average number of 18 Cherenkov photoelectrons per ring with 3 GeV/c pion and 10 mm
oC for six hours.
? 10
?and single electron
C
3) single photon angular resolution of 8 mrad.
Furthermore two 80?20 cm
Imaging Cherenkov detector of the CERN/NA44 experiment, showing no degradation of the
?
F
??radiator;
?CsI photocathodes have been successfully used in the Threshold
4

Page 6

Figure 5: CsI QE measured at 170 nm as a function of the PC number. All the photocathodes
have large area, from 30?30 cm
the HMPID prototype.
?to 64?40 cm
?of PC 28 and 29 which have been produced for
5

Page 7

Figure 6: CsI QE measured at 170 nm as a function of PC age.
Figure 7: A 3 GeV/ c pion event, obtained with PC24 and 10 mm thick
the pad cluster MIP surrounded by the Cherenkov photons ring. The ring radius of about 10 cm
has been obtained with a proximity gap of 7 cm. The grey scale on the right, in arbitrary units,
is proportional to the signal measured on the pads.
C
?
F
??radiator, showing
6

Page 8

Figure 8: Average number of pads, pad clusters and photoelectrons as a function of the C
radiator thickness, at a gain of 10
obtained as ratio between the total pulse height and the single electron pulse height, includes
Cherenkov and feedback photons, while the number of resolved clusters is related to Cherenkov
events. The mean number of pads per event depends on the cluster size, which varies roughly
from2to3padsinthe2 ?10 mmradiatorthicknessrange.Theclustersizeincreasewithradiator
thickness can be attributed to photoelectrons overlapping, which enhances the signal induced
on pads and therefore the number of pads beeing over threshold.
?F
?
?
?and with 3 GeV/ c pions. The number of photoelectrons,
7

Page 9

Figure 9: Average number of pads, pad clusters and photoelectrons as a function of the MWPC
high voltage, at a radiator thickness of 4 mm and with 3 GeV/
electrons, obtained as ratio between the total pulse height and the single electron pulse height,
includes Cherenkov and feedback photons, while the number of resolved clusters is related to
Cherenkov events. Finally the mean number of pads per event depends on the cluster size which
varies from from 1.5 pads, at 2050 V (i.e. at a gain of 1.5?10
a gain of 7?10
feedback, as shown by the change of slope of the number of photoelectrons curve.
c pions. The number of photo-
?), to 3 pads, at 2225 V (i.e. at
?). The cluster size increase with chamber gain is due to the onset of photon
8

Page 10

performance during four weeks of high irradiation level in lead ion run [8].
4 Conclusions
The status of the development of the CsI-RICH detector for the HMPID in ALICE at
LHC has reached a very important phase. A prototype, 2/3 of one final RICH module, has been
built, confirming mechanical design expectations and showing also good performance of the
MWPC with large area CsI photocathode. The foreseen test at the CERN/SPS will be of prime
importance to assess the PID capability in events with high particle densities.
References
[1] ALICE Collaboration, Technical Proposal, CERN/LHCC 95-71
[2] ALICE Collaboration, Addendum to the Technical Proposal, CERN/LHCC 96-32
[3] E. Nappi et al., High Momentum PID with RICH detectors in the ALICE experiment at
LHC (to be published in the proceedings of the International Conference on High Energy
Physics ICHEP96 - Warsaw 1996, World Scientific, Singapore)
[4] J. C. Santiard, Gassiplex, a low noise analog signal processor for readout of gaseous detec-
tors, CERN-ECP/94-17
[5] J. C. Santiard, MCM-Digitplex, a compact on-detector analog to digital processor, (pub-
lishedintheProceedings ofthe2nd Workshoponelectronics forLHC experiments,Balaton
96)
[6] RD26 Collaboration, Status Report, CERN/LHCC 96-20
[7] F. Piuz, CsI-photocathode and RICH detector, Nuclear Instruments and Methods A 371
(1996) 96–115
[8] A. Braem et al., A Threshold Imaging Cherenkov Detector with CsI Photocathodes, to be
published in the proceedings of the 7th Pisa Meeting on Advanced Detectors, May 25-31,
1997 - La Biodola, Isola d’Elba, Italy.
9