3rd INT. WORKSHOP ON HIGH-PT PHYSICS AT LHC
March, 16-19, 2008 Tokay, Hungary
Gas Cherenkov detector for high momentum
charged particle identification in the
ALICE experiment at LHC
Giacomo Volpe
Istituto Nazionale di Fisica Nucleare,
Sezione di Bari, Italy.
1
EMCal
High energy g
TRD
Electron ID,
Tracking
ALICE experiment
HMPID
pioni is designed toRICH
ALICE
study, PID
the physics
@ high pof
T
strongly interacting matter and the quarkgluon plasma (QGP)TOF
in nucleus-nucleus
PID @ intermediate p
collisions (sNN = 5.5 TeV) at the LHC. TheT
p-p physics will be study as well as
TPC
Main Tracking, reference data for the nucleus-nucleus
PID with dE/dx analysis.
ITS
Vertexing, low pt tracking
and PID with dE/dx
PHOS
g,p0 -ID
pioni
L3
Magnet
B=0.2-0.5 T
MUON
m-ID
+
T0,V0, PMD,FMD and ZDC
2
Forward rapidity region
ALICE experiment
ALICE has a unique capability, among the LHC experiments, of charged
particle identification, due to the exploiting of different types of detectors:
 ITS + TPC : low pT identification (up to p = 600 MeV/c).
 TOF : covers intermediate pT region.
 TRD : electrons identification.
 HMPID : high pT region (1÷5 GeV/c).
p/K
TPC + ITS
(dE/dx)
K/p
e /p
p/K
TOF
HMPID
(RICH)
TRD
K/p
p/K
0
1
2
3
K/p
4
5 p (GeV/c)
e /p
1
High-pT Physics at LHC, 17 March 2008
10
G. Volpe
100 p (GeV/c)
3
ALICE PID upgrade
RICH results: At RHIC has been observed a large enhancement of baryons and
antibaryons relative to pions at intermediate pT ≈ 2 - 5 GeV/c, while the neutral
pions and inclusive charged hadrons are strongly suppressed at those pT.
• The key issue is to understand what is the mechanism of the hadronization and
the influence of this mechanism on the spectra of baryons and mesons.
4
High-pT Physics at LHC, 17 March 2008
G. Volpe
ALICE PID upgrade
• The baryon puzzle observed at RICH can be interpreted with the “partons
recombination” or “coalescence” mechanism.
• In the recombination scenario quark-antiquark pair close in the phase space
can form a meson at hadronization, while three (anti)quark can form an
(anti)baryon.
At LHC where the density of jets is
very high, a new phenomenon
originates where the recombination
of shower partons in neighboring
jets can make a significant
contribution. It is foreseen that the
baryon enhancement will be present
in a momentum range higher than at
RHIC, pT = 10 ÷ 20 GeV/c. (ref.
Rudolph C. Hwa, C. B. Yang,
arXiv:nucl-th/0603053 v2, 21 Jun
2006)
5
ALICE PID upgrade
Other authors using different
arguments foresee also change in
meson-baryon ratio for pT > 10
GeV/c.
Jet quenching can leave signatures
not only in the longitudinal and
transverse jet energy and multiplicity
distributions, but also in the
hadrochemical composition of the
jet fragments.
S. Sapeta and U. A. Wiedemann,
arXiv:0707.3494 [hep-ph], July
2007.
6
ALICE PID upgrade
EMCal
• The use of the Electromagnetic
Calorimeter
opens
interesting
possibility to distinguish quark and
gluon jets in gamma - jet events and
subsequently the study of the
probability of fragmentation in pions,
kaons or protons.
TPC
g
beam
axis
hadrons
momentum ?
• Regardless of the theoretical interpretations it seems important to have the
possibility to measure the meson-baryon ratio up to momenta well above the
current limits of ALICE for a track-by-track identification.
7
High-pT Physics at LHC, 17 March 2008
G. Volpe
VHMPID
 ALICE-HMPID collaboration is studying the possibility to built a new detector
to identify charged particles with momentum p > 10 GeV/c  VHMPID (Very
High Momentum Particle Identification Detector).
 Energy loss or Time of Flight measurements don’t allow to identify track-bytrack in such momentum range.
 Since the given space in the ALICE detector and the physics requirements it
seems inevitable to use gas Cherenkov counters.
 To use a gas Cherenkov detector in a magnetic field environment brings about
the following key problems: the choice of radiator gas, the photon detection and
the detector geometry.
 A combination of a gas with low value of refractive index, with the proven
concept of large area CsI photocathodes, has been considered.
 Depending on the particle momentum values, with VHMPID will be possible to
have PID by means pattern recognition method or by threshold counters technique.
 Simulation results will be presented.
8
High-pT Physics at LHC, 17 March 2008
G. Volpe
Radiator gas
• CF4 (n ≈ 1.0005, gth ≈ 31.6) has the
drawback to produce scintillation photons
(Nph ≈ 1200/MeV), that increase the
background.
• C4F10 (n ≈ 1.0015, gth ≈ 18.9)
• C5F12 (n ≈ 1.002, gth ≈ 15.84) this gas has
been used in the DELPHI RICH detector.
9
VHMPID
Photon detector
• Pad-segmented CsI photocathode is combined with a MWPC with the same
structure and characteristic of that used in the HMPID detector.
• The gas used is CH4, the pads size is 0.8×0.84 cm2 (wire pitch 4.2 mm), and the
average single electron pulse height is of 34 ADC channels (1 ADC = 0.17 fC ≈
1000 e-) at 2050 V.
• The chamber is separated from the radiator by a window (4 mm of thickness).
10
High-pT Physics at LHC, 17 March 2008
G. Volpe
Photon detector
An other option for the photon detector could be a GEM-like detector combined with a
CsI photocathode (higher gain, photons feedback suppression).
Principles of operation
V. Peskov studies
11
High-pT Physics at LHC, 17 March 2008
G. Volpe
VHMPID
• The simulation has been executed using AliRoot, the official simulation framework of
the ALICE experiment;
• Different geometries has been taken into account;
• C5F12 as radiator;
• CaF2 window.
Material photon transmittances
and CsI photocathode quantum
efficiency
12
High-pT Physics at LHC, 17 March 2008
G. Volpe
Studied setup
Proximity-focusing like setup: Charged particles cross the radiator
producing Cherenkov photons. On the chamber both charged particle and
photon signals are present. Signal topology depends only on the track
momentum.
13
High-pT Physics at LHC, 17 March 2008
G. Volpe
Studied setup
Focusing setup: the focusing properties of a spherical mirror of radius
R = 240 cm, are exploited. The photons emitted in the radiator are
focused in a plane that is located at R/2 from the mirror center, where
the photon detector is placed.
14
High-pT Physics at LHC, 17 March 2008
G. Volpe
Photon blob topology – proximity focusing setup
Nph(b = 1) ≈ (1.4 eV-1cm-1)·(3 eV)·(180 cm) ≈ 760, but
The number of photons detected is much less because of the absorption
in the radiator gas and CsI quantum efficiency.
<N> ≈ 43
<N> ≈ 91
15 GeV/c
3 GeV/c
15
High-pT Physics at LHC, 17 March 2008
G. Volpe
Photon blob diamater
• An algorithm to calculate the blob diameter has been implemented.
• The pad with the largest values of the charge corresponds with the impact particle
point. I consider R that contains the 98% of the total charged pads. The values in the
figure refers to mean and RMS of a sample of 100 events.
16
High-pT Physics at LHC, 17 March 2008
G. Volpe
Photon ring topology: focusing setup
Nph(b = 1) ≈ (1.4 eV-1cm-1)·(3 eV)·(120 cm) ≈
500
b≈1
17
High-pT Physics at LHC, 17 March 2008
G. Volpe
18
High-pT Physics at LHC, 17 March 2008
G. Volpe
19
High-pT Physics at LHC, 17 March 2008
G. Volpe
Track inclination angle = 10°
Orthogonal track displaced
40 cm from the detector
center.
20
High-pT Physics at LHC, 17 March 2008
G. Volpe
Study of the detector response – proximity focusing
like setup
Cherenkov
photons
Parameters used
Blob radius
Blob pads
number
Photodetector
Charged
particle
High-pT Physics at LHC, 17 March 2008
21
G. Volpe
Study of the detector response
background subtraction algorithm
Background produced by Pb-Pb collision event
Charged
particles
It considers pads with charge larger than 200 ADC channels
It checks if that pad is a local maximum in pads charge values
If the pad considered is a local maximum, it cuts that pad and the adjacent ones
(the pads corresponding to the track to identify not are taken into account by this
procedure)
22
High-pT Physics at LHC, 17 March 2008
G. Volpe
23
High-pT Physics at LHC, 17 March 2008
G. Volpe
Pb-Pb collision events, considering the presence of
MIPs background
24
High-pT Physics at LHC, 17 March 2008
G. Volpe
Study of the detector response: focusing setup
• In the case of focusing setup the determination of Cherenkov emission angle is
possible.
• Pattern recognition algorithm is needed to retrieve the emission angle.
• A back-tracing algorithm has been implemented to retrieve the Cherenkov emission
angle. It calculates the angle starting from the photon hit point coordinates, on the
photon detector.
Radiator vessel
Photodetector
Charged particle
q
Mirror
120 cm
High-pT Physics at LHC, 17 March 2008
G. Volpe
25
Simulation results: Cherenkov angle
bb≈ ≈
1
1
26
High-pT Physics at LHC, 17 March 2008
G. Volpe
Simulation results: Cherenkov angle
27
High-pT Physics at LHC, 17 March 2008
G. Volpe
Simulation results: Pb-Pb background
HIJING generator
dNch/dh = 4000 at
mid rapidity
28
High-pT Physics at LHC, 17 March 2008
G. Volpe
Background subtraction algorithm
Hough transform method
• The Hough Transform Method (HTM) is an efficient implementation of a generalized template
matching strategy for detecting complex patterns in binary images.
• In the case of the Cherenkov pattern recognition, the starting point of the analysis is a
bidimensional map with the impact point (xp, yp) of the charged particles, hitting the detector
plane with known incidence angles (θp, φp), and the coordinates (x, y) of hits due to both
Cherenkov photons and background sources.
• A “Hough counting space” is constructed for each charged particle, according to the following
transform:
(x, y) → ((xp, yp, θp, φp) , hc)
• (xp, yp, θp, φp) is provided by the tracking of the charged particle, so the transform will reduce
the problem to a solution in a one-dimensional mapping space.
• A hc bin with a certain width is defined.
• The Cherenkov angle θc of the particle is provided by the average of the hc values that fall in
the bin with the largest number of entries.
29
Simulation results: Cherenkov angle
• Hough transform is used to discriminate the signal from the background.
b≈1
30
High-pT Physics at LHC, 17 March 2008
G. Volpe
Simulation results: Cherenkov angle
31
High-pT Physics at LHC, 17 March 2008
G. Volpe
Momentum range for p, K and p identification
in Pb-Pb collisions environment.
p
Momentum
C5F12
Particle Id.
0
?
1
p
2.5 < p < 8 GeV/c
0
K, p
8 < p < 14 GeV/c
1
p, K
8 < p < 14 GeV/c
0
p
15 < p < 26 GeV/c
1
p
< 2.5 GeV/c
K
2.5< p < 8 GeV/c
p
32
VHMPID in the ALICE apparatus
PHOS
VHMPID
modules
High-pT Physics at LHC, 17 March 2008
G. Volpe
33
Studied setup
• The available space in the ALICE apparatus is not too much. The goal is to decrease
much as possible the detector dimension.
• C5F12 has a boiling point Tb = 28° C at 1 atm, implying a difficult use of it in ALICE
setup, where the internal temperature could be more or less the same (heating plant is
needed).
• A setup with radiator length of 80 cm, C4F10 as radiator and SiO2 window has been also
investigated.
34
High-pT Physics at LHC, 17 March 2008
G. Volpe
Simulation results: Cherenkov angle
b≈1
35
High-pT Physics at LHC, 17 March 2008
G. Volpe
Simulation results: Cherenkov angle
36
High-pT Physics at LHC, 17 March 2008
G. Volpe
Momentum range for p, K and p identification
in Pb-Pb collisions environment.
p
Momentum
C4F10
Particle Id.
0
?
1
p
3 < p < 9 GeV/c
0
K, p
9 < p < 14 GeV/c
1
p, K
9 < p < 17 GeV/c
0
p
17 < p < 24 GeV/c
1
p
< 3 GeV/c
K
3< p < 9 GeV/c
p
37
Conclusions & Outlook
• The focusing setup, since its smaller dimension, is the setup that the collaboration
will develop.
• The goal is to have a small detector performing good PID. The 80 cm setup will be
better investigated.
• To enrich the sample with interesting event, triggering option has been also
considered, using a dedicated trigger (see L. Boldizsar talk) and/or photons in the
EMCal.
38
High-pT Physics at LHC, 17 March 2008
G. Volpe
Backup
39
Study of the detector response
p m
N pads ( p)  N b 1  2  sin (q ch )  N b 1  2  (1  2 2 )
n 1
n 1
n p
n
2
2
n
2
2
2
40