CERN Academic Training 97/98Particle Detectors
Christian Joram I/1
Introduction
Introduction
Gaseous and solid state trackingdevices
Scintillators, photo detectors, scint.fibers, nuclear emulsions, ‘electronicbubble chamber’
Calorimetry
Particle identification
Many topics will remain uncovered...
Design of detector systems Electronics, trigger and data acquisition Non-particle physics (medical, biological etc.)
applications Detector construction Detector operation (monitoring, calibration,
alignment)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/2
Introduction
Literature on particle detectorsLiterature on particle detectors
Text books C. Grupen, Particle Detectors, Cambridge
University Press, 1996 W. R. Leo, Techniques for Nuclear and Particle
Physics Experiments, 2nd edition, Springer, 1994 R.S. Gilmore, Single particle detection and
measurement, Taylor&Francis, 1992 W. Blum, L. Rolandi, Particle Detection with Drift
Chambers, Springer, 1994 K. Kleinknecht, Detektoren für Teilchenstrahlung,
3rd edition, Teubner, 1992
Review articles Experimental techniques in high energy physics,
T. Ferbel (editor), World Scientific, 1991. Instrumentation in High Energy Physics, F. Sauli
(editor), World Scientific, 1992. Many excellent articles can be found in Ann. Rev.
Nucl. Part. Sci.
Other sources Particle Data Book (Phys. Rev. D, Vol. 54, 1996) R. Bock, A. Vasilescu, Particle Data Briefbook
http://www.cern.ch/Physics/ParticleDetector/BriefBook/
Proceedings of detector conferences (ViennaWCC, Elba, IEEE)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/3
Introduction
A W+W- decay in DELPHI
e+e- (√s=161.2 GeV) → W+W- → qqqq → 4 hadronic jets_ _
≈ 10
m
≈ 10 m
CERN Academic Training 97/98Particle Detectors
Christian Joram I/4
Introduction
A W+W- decay in ALEPH
e+e- (√s=181 GeV) → W+W- → qqµνµ → 2 hadronic jets + µ + missing momentum
_
CERN Academic Training 97/98Particle Detectors
Christian Joram I/5
Introduction
Reconstructed B-mesons in theDELPHI micro vertex detector
τB ≈ 1.6 ps l = cτγ ≈ 500 µm⋅γ
Primary Vertex
Primary Vertex
CERN Academic Training 97/98Particle Detectors
Christian Joram I/6
Introduction
Particle identification methods
CERN Academic Training 97/98Particle Detectors
Christian Joram I/7
Introduction
≈ 13
m
CERN Academic Training 97/98Particle Detectors
Christian Joram I/8
Introduction
A simulated event in ATLAS (CMS)H → 4µ
pp collision at √s = 14 TeVσinel. ≈ 70 mbInterested in processes
with σ ≈ 10−100 fb
≈ 23 overlapping minimum bias events / BC≈ 1900 charged + 1600 neutral particles / BC
L = 1034 cm-2 s-1, bunch
spacing 25 ns
CERN Academic Training 97/98Particle Detectors
Christian Joram I/9
Introduction
The ‘ideal’ particle detector for highenergy physics experiments
High energy collisions ( )
→ production of a multitude of particles (charged,neutral, photons)
The ‘ideal’ detector should provide….
coverage of full solid angle (no cracks, finesegmentation
detect, track and identify all particles (mass, charge)
measurement of momentum and/or energy fast response, no dead time
practical limitations (technology, space, budget)
Particles are detected via their interaction withmatter.
Many different physical principles are involved(mainly of electromagnetic nature).
Eventually we will observe...
ionizationionization and excitationexcitation of matter.
pppp,ep,,ee
CERN Academic Training 97/98Particle Detectors
Christian Joram I/10
CERN Academic Training 97/98Particle Detectors
Christian Joram I/11
Charge Coupled Devices (CCD)
Momentum measurement
Multiple scattering
Bethe-Bloch formula / Landau tails
Ionization of gases
Wire chambers
Some derivatives of wire chambers
Drift and diffusion in gases
Drift chambers
Micro gas detectors
Chamber aging
Gas detector simulation
Silicon drift detectors
Radiation hardness
Silicon detectorsstrips/pixels
?
Introduction
CERN Academic Training 97/98Particle Detectors
Christian Joram I/12
Momentum measurement
CERN Academic Training 97/98Particle Detectors
Christian Joram I/13
Momentum measurement
Momentum measurement
the sagitta s is determined by 3 measurements witherror σ(x):
for N equidistant measurements, one obtains(R.L. Gluckstern, NIM 24 (1963) 381)
ex: pT=1GeV/c, L=1m, B=1T, σ(x)=200µm, N=10
L
s
B
y
x
T
TT
T
T
T
p
BLs
BLpp
p
BLL
Bp
qBp
22
83.0
82cos1
3.0sin
3.022sin
2
m)(T3.0)cGeV(
2
23
23.
3.0
8)()()(
BL
px
s
x
s
s
p
p Tmeas
T
T
)4/(720
3.0
)(2
.
NBL
px
p
p Tmeas
T
T (for N ≥ ≈10)
%5.0
.
meas
T
T
p
p
231
2xx
xs
(s ≈ 3.75 cm)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/14
Multiple Scattering
Scattering
An incoming particle with charge z interacts with atarget of nuclear charge Z. The cross-section for thise.m. process is
Average scattering angle Cross-section for infnite !
Multiple ScatteringSufficiently thick material layer → the particle willundergo multiple scattering.
2sin
14
4
22
p
cmzZr
d
d ee
00
Rutherford formula
d/d
L
plane
rplane
RMSRMSplaneplane space
2
120
CERN Academic Training 97/98Particle Detectors
Christian Joram I/15
Multiple Scattering
Approximation
X0 is radiation length of the medium(accuracy ≤ 11% for 10-3 < L/X0 < 100)
The lateral displacement can also be approximated bya Gaussian distribution with width
000 ln038.01
6.13X
L
X
Lz
cp
MeV
P
plane0
Gaussian
sin-4(/2)
20
2
0 2exp
2
1
planeplaneP
0 LLr RMSplane
RMSplane
CERN Academic Training 97/98Particle Detectors
Christian Joram I/16
Momentum measurement
Back to momentum measurements:contribution from multiple scattering
ex: Ar (X0=110m), L=1m, B=1T
0
0
0
1045.0
3.0
0136.0)(
10136.0sin
LXBBL
X
L
p
p
p
p
X
L
pppp
T
MSMS
T
planeRMS
MS
(p)/p
(p)/p
(p)/p
p
MS
meas.total error
independentof p !
%5.0)(
MS
Tp
p
CERN Academic Training 97/98Particle Detectors
Christian Joram I/17
Momentum Measurement
Momentum measurement in experiments withsolenoid magnet:
polar angle has to be determined from a straight linefit x=x(z).
N equidistant points with error σ(z)
+ multiple scattering contribution….
In practical cases:
In summary:
B
B x
zy
x
yz
sinppT
NBL
px
p
pmeas
1)(2
.
))1(/()1(12)(. NNN
L
zmeas
T
T
p
p
p
p
normally small
CERN Academic Training 97/98Particle Detectors
Christian Joram I/18
Interaction of charged particles
Detection of charged particlesHow do they loose energy in matter ?
Discrete collisions with the atomic electrons of theabsorber material.
Collisions with nuclei not important (me<<mN).
If are big enough ionization.
A photon in a medium has to follow the dispersionrelation
For soft collisions + energy and momentumconservation → real photons:
222
2 022 nck
n
ck
nc
11
cos
cos
nvk
kvkv
density electron :N
ddE
dNE
dx
dE 0
e-
k ,
0,mv
→ Cherenkov photonsif n1
k ,
CERN Academic Training 97/98Particle Detectors
Christian Joram I/19
Interaction of charged particles
Average differential energy loss
Many different approaches (classical, semi-classical) exist...
e.g. Allison+Cobb (Ann.Rev. of Nucl. and Part. Phys, 30 (1988), 253),Photo Absorption Ionization (PAI) modelincludes Cherenkov + transition radiation
PAI model relates dσ/dE to empirical atomicphotoabsorption cross-sections σγ(E)
Ionisation only → Bethe - Bloch formula
1
optical absorptive X-ray
Cherenkovradiation
ionisation transition radiation
regime:
effect:
Re
Im
dxdE
schematically !
CERN Academic Training 97/98Particle Detectors
Christian Joram I/20
Bethe-Bloch formula
dE/dx in [MeV g-1 cm2]
dE/dx depends only on β, independent of m Formula takes into account energy transfers
Bethe-Bloch formula only valid for “heavy”particles (m≥mµ).
Electrons and positrons need specialtreatment (mproj=mtarget), in additionBremsstrahlung!
2
2ln
14 2max
2
222
21
2222
TI
cm
A
ZzcmrN
dx
dE eeeA
eV10 with
potential excitationmean :
00
max
IZII
ITdEI
2121 cm MeV g..dxdE
Z/A does notdiffer muchfor the variouselements,except forHydrogen!
(rough approximation, Ifitted for each element)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/21
solid line: Allison and Cobb, 1980dashed line: Sternheimer (1954)data from 1978 (Lehraus et al.)
Bethe-Bloch formula
dE/dx first falls ∝ 1/β2 (more precise β-5/3),kinematic factor
then minimum at βγ ≈ 4 (minimum ionizing particles,MIP)
then again rising due to ln γ2 term, relativistic rise,attributed to relativistic expansion of transverse E-
field → contributions from more distant collisions.
relativistic rise cancelled at high γ by “densityeffect”, polarization of medium screens more distant
atoms. Parameterized by δ (material dependent) →Fermi plateau
many other small corrections
Measured andcalculated dE/dx
2
2ln
14 2max
2
222
21
2222
TI
cm
A
ZzcmrN
dx
dE eeeA
CERN Academic Training 97/98Particle Detectors
Christian Joram I/22
Landau tails
Real detectors (limited granularity) do not measure<dE/dx>, but the energy ∆E deposited in a layer offinite thickness δx.∆E is the result of a certain number i of collisions withenergy transfers Ei, with cross-section dσ/dE
Problem: Relate F(∆E,δx) to dσ/dESolutions: Williams(1929), Landau(1944),
Allison and Cobb(1980) and others.
ionization by close collisions
-electrons
exci
tati
on
ion
izat
ion
by
dis
tan
t co
llisi
on
bel
ow
exc
itat
ion
th
resh
old
energy transfer/photon exchange
rel.
pro
bab
ility
Schematically !(after Gilmore)
Monte-Carlo approach:1.) divide absorber layer into thin slices. Probability
for any energy transfer in the slice is low.2.) Choose energy transfer randomly from above
distribution.3.) Sum over all slices.
CERN Academic Training 97/98Particle Detectors
Christian Joram I/23
Landau tails
For thin layers (and low density materials):→ Few collisions, some with high energy transfer.→ Energy loss distributions show large fluctuations
towards high losses: ”Landau tails”
data: Harr is et al. (1977)dotted curve: Landau(1944)dashed curve: Maccabee &
Papwor th(1969)(Landau’s method)
solid curve: Allison and Cobb
Argon + 7% CH4
Argon + 7% CH4
For thick layers and high density materials:→ Many collisions.→ Central Limit Theorem → Gaussian shape
distributions.
CERN Academic Training 97/98Particle Detectors
Christian Joram I/24
Ionization of gases
Primary and total ionization
Fast charged particles ionize the atoms of a gas.
Often the resulting primary electron will have enoughkinetic energy to ionize other atoms.
total number of createdelectron-ion pairs.∆E = total energy lossWi = effective <energy loss>/pairprimarytotal
iitotal
nn
W
xdxdE
W
En
43
Number ofprimaryelectron/ion pairsin frequentlyused (detector)gases.
(Lohse and Witzeling,Instrumentation In HighEnergy Physics, WorldScientific,1992)
41
37
36
3126
28
33
22
Wi (eV)
Primary ionization Total ionisation
CERN Academic Training 97/98Particle Detectors
Christian Joram I/25
Ionization of gases
2 remarks:
1. The actual number of primary electron-ion pairs isPoisson distributed
The detection efficiency is therefore limited to
For thin layers εdet can then be significantly lowerthan 1.
Example Ar: d = 1mm → nprimary = 2.5. → εdet = 0.92
2. The total number of electron-ion pairs is subject to
large, non-poissonian, fluctuations (Landau tails).
Detectors which are able to identify primary
clusters, may provide a better energy resolution.
→ cluster counting (see lecture Particle ID).
!)(
m
enmP
nm
neP 1)0(1det
CERN Academic Training 97/98Particle Detectors
Christian Joram I/26
Proportional Counter
≈ 100 electron-ion pairs are not easy to detect!Noise of amplifier ≈1000 e- (ENC) !We need to increase the number of e-ion pairs.
Gas amplificationConsider cylindrical field geometry (simplest case):
Electrons drift towards the anode wire (≈ stop and go!More details in next lecture!).Close to the anode wire the field is sufficiently high(some kV/cm), so that e- gain enough energy for furtherionization → exponential increase of number of e--ionpairs.
a
b
r
E
1/r
a
cathode
anode
gas
Ethreshold
a
rCVrV
r
CVrE
ln2
)(
12
0
0
0
0
C = capacitance / unit length
CERN Academic Training 97/98Particle Detectors
Christian Joram I/27
Proportional Counter
0CVkeM
xrxE ennenn 00 or α: First Townsend coefficient
(e--ion pairs/cm)
(O. Allkofer, Spark chambers, Theimig München, 1969)
1 λ: mean free path
kNFor low e- energies ε
C
r
a
drrn
nM exp
0Gain
(F. Sauli, CERN 77-09)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/28
Proportional Counter
Choice of gas:
Dense noble gases. Energy dissipation mainly byionization! High specific ionization.
De-excitation of noble gases only possible viaemission of photons, e.g. 11.6 eV for Argon.
This is above ionization threshold of metals, e.g.Copper 7.7 eV.
Ar *11.6 eV
Cu
e-
cathode
→ new avalanches →permanent discharges !
(MAGBOLTZ)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/29
Proportional Counter
Solution: Add poly-atomic gases as quenchers.Absorption of photons in a large energy range (manyvibrational and rotational energy levels).
E
r
Energy dissipationby collisions ordissociation intosmaller molecules.
Methane: absorption band 7.9 - 14.5 eV
(MAGBOLTZ)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/30
Signalformation
Avalanche formation within a few wire radii and withint < 1 ns!
Signal induction both on anode and cathode due tomoving charges (both electrons and ions).
Proportional Counter
drdr
dV
lCV
Qdv
0
Electrons collected by anode wire, i.e. dr issmall (few µm). Electrons contribute only verylittle to detected signal (few %).
Ions have to driftback to cathode,i.e. dr is big.Signal durationlimited by total iondrift time !
Need electronic signal differentiation to limit dead time.
(F. Sauli, CERN 77-09)
(F. Sauli, CERN 77-09)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/31
Operation modes: ionization mode: full charge collection, but no
charge multiplication.
Proportional mode: above threshold voltagemultiplication starts. Detected signal proportionalto original ionization → energy measurement(dE/dx). Secondary avalanches have to bequenched. Gain 104 - 105.
Limited Proportional → Saturated → Streamermode: Strong photo-emission. Secondaryavalanches, merging with original avalanche.Requires strong quenchers or pulsed HV. Highgain (1010), large
.
Operation modes of chambers
Geiger mode:Massive photoemission. Fulllength of anodewire affected. Stopdischarge bycutting down HV.Strong quenchersneeded as well.
signals → simpleelectronics.
CERN Academic Training 97/98Particle Detectors
Christian Joram I/32
Multi wire proportional chambers (MWPC)(G. Charpak et al. 1968, Nobel prize 1992)
Capacitive coupling of non-screened parallel wires?Negative signals on all wires? Compensated bypositive signal induction from ion avalanche.
Typical parameters: L=5mm,d=1mm,awire=20µm.
Address of fired wire(s) give only 1-dimensionalinformation. Secondary coordinate ….
Multi wire proportional chambers
field lines and equipotentials around anode wires
Normally digital readout:spatial resolution limited to
12
dx
(d=1mm, σx=300 µµm)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/33
Secondary coordinate
Charge division. Resistive wires (Carbon,2kΩ/m).
Timing difference (DELPHI Outer detector, OPAL vertex detector)
1 wire plane + 2 segmented cathode planes
Multi wire proportional chambers
y
L
QBQA
track
%4.0 toup
L
y
Q
L
y
BA
B
Crossed wire planes. Ghost hits.Restricted to low multiplicities. Alsostereo planes (crossing under small
angle).
yL
track
CFD CFDT
)(OPAL4)(
100)(
cmy
psT
Analog readout ofcathode planes.→ σ ≈ 100 µm
CERN Academic Training 97/98Particle Detectors
Christian Joram I/34
Some ‘derivatives’
Multistep avalanche chambers
Derivatives of proportional chambers
amplification 1 G=103
amplification 2G=103
conversion
transfer
low field
high field
high field
low field
wire meshes
≥ 2 parallel plane avalanche chambers (PPAC),separated by transfer fields, operated at low gain.
Achieve high total gain (106-108) by reducing feedbackphoton problem.
Electron transparency of wire meshes depends on fieldratio Eout/Ein. (Eout/Ein = 0.1→ T ≈ 0.1)
Charge amplification by Penning process at rel. lowfield:
Application: Single photo electron detection.
A* + C → A + C+ + e-
CERN Academic Training 97/98Particle Detectors
Christian Joram I/35
Thin gap chambers (TGC)
Derivatives of proportional chambers
G10 (support)
cathode pads ground plane
graphite
3.2
mm
2 mm
4kV50 m
Operation in saturated mode. Signal amplitudelimited by by the resistivity of the graphite layer(≈ 40kΩ/).
Fast (2 ns risetime), large signals (gain 106), robust
Application: OPAL pole tip hadron calorimeter.G. Mikenberg, NIM A 265 (1988) 223
ATLAS muon endcap trigger, Y.Arai et al. NIM A 367 (1995) 398
Gas:CO2/n-pentane
(≈ 50/50)
CERN Academic Training 97/98Particle Detectors
Christian Joram I/36
Derivatives of proportional chambers
Resistive plate chambers (RPC)2
mm bakelite
(melamine phenolic laminate)
pickup strips
10 kV
spacer
Gas: C2F4H2, (C2F5H) + few % isobutane(ATLAS, A. Di Ciaccio, NIM A 384 (1996) 222)
Time dispersion ≈ 1..2 ns → suited as trigger chamberRate capability ≈ 1 kHz / cm2
15 kV
Double andmultigapgeometries →improve timingand efficiency
Problem: Operation close to streamer mode.