Electromagnetic Calorimeters

Calorimetry I
M. Krammer: Detektoren, SS 05 Kalorimeter 2
6.1 Allgemeine Grundlagen
Funktionsprinzip – 1
! In der Hochenergiephysik versteht man unter einem Kalorimeter einen
Detektor, welcher die zu analysierenden Teilchen vollständig absorbiert. Da-
durch kann die Einfallsenergie des betreffenden Teilchens gemessen werden.
! Die allermeisten Kalorimeter sind überdies positionssensitiv ausgeführt, um
die Energiedeposition ortsabhängig zu messen und sie beim gleichzeitigen
Durchgang von mehreren Teilchen den individuellen Teilchen zuzuordnen.
! Ein einfallendes Teilchen initiiert innerhalb des Kalorimeters einen Teilchen-
schauer (eine Teilchenkaskade) aus Sekundärteilchen und gibt so sukzessive
seine ganze Energie and diesen Schauer ab.
Die Zusammensetzung und die Ausdehnung eines solchen Schauers hängen
von der Art des einfallenden Teilchens ab (e
±
, Photon oder Hadron).
Bild rechts: Grobes Schema
eines Teilchenschauers in
einem (homogenen) Kalorimeter
Introduction
Calorimeter:
Detector for energy measurement via total absorption of particles ...
Also: most calorimeters are position sensitive to measure energy depositions
depending on their location ...
Principle of operation:
detector volume
incident particle
particle cascade (shower)
Incoming particle initiates particle shower ...
Shower Composition and shower dimensions depend on
particle type and detector material ...
Energy deposited in form of: heat, ionization,
excitation of atoms, Cherenkov light ...
Different calorimeter types use different kinds of
these signals to measure total energy ...
Important:
Signal ~ total deposited energy
[Proportionality factor determined by calibration]
Schematic of
calorimeter principle
Introduction
Energy vs. momentum measurement:
σ
E
E

1

E
Calorimeter:
[see below]
Gas detector:
[see above]
σ
p
p
∼ p
e.g. ATLAS:
σ
E
E

0.1

E
σ
p
p
≈ 5 · 10
−4
· p
t
e.g. ATLAS:
i.e. !E/E = 1% @ 100 GeV i.e. !p/p = 5% @ 100 GeV
At very high energies one has to switch to calorimeters because their
resolution improves while those of a magnetic spectrometer decreases with E ...
Shower depth:
Calorimeter:
[see below]
L ∼ ln
E
E
c
[Ec: critical energy]
Shower depth nearly energy independent
i.e. calorimeters can be compact ...
Compare with magnetic spectrometer:
Detector size has to grow quadratically to maintain resolution
σ
p
/p ∼
p
/L
2
Introduction
Further calorimeter features:
Calorimeters can be built as 4!-detectors, i.e. they can detect
particles over almost the full solid angle
Magnetic spectrometer: anisotropy due to magnetic field; remember:
Calorimeters can provide fast timing signal (1 to 10 ns); can
be used for triggering [e.g. ATLAS L1 Calorimeter Trigger]
(
σ
p
/p)
2
= (
σ
p
t/p
t
)
2
+ (
σ
θ
/sin θ)
2
large
for small !
Calorimeters can measure the energy of both, charged and neutral particles,
if they interact via electromagnetic or strong forces [e.g.: ", #, $
0
, ...]
Magnetic spectrometer: only charged particles!
Segmentation in depth allows separation of hadrons (p,n,!
±
), from
particles which only interact electromagnetically (",e) ...
...
µ = nσ = ρ
N
A
A
· σ
pair
Electromagnetic Showers
Reminder:
X0
Dominant processes
at high energies ...
Photons : Pair production
Electrons : Bremsstrahlung
Pair production:
dE
dx
=
E
X
0
dE
dx
= 4αN
A
Z
2
A
r
2
e
· E ln
183
Z
1
3
σ
pair

7
9 ￿

4 αr
2
e
Z
2
ln
183
Z
1
3 ￿

=
7
9
ρ
X
0
Bremsstrahlung:
E = E
0
e
−x/X
0
[X0: radiation length]
[in cm or g/cm
2
]
Absorption
coefficient:
After passage of one X0 electron
has only (1/e)
th
of its primary energy ...
[i.e. 37%]
!
=
7
9
A
N
A
X
0
Electromagnetic Showers
Transverse size of EM shower given by
radiation length via Molière radius
[see also later]
R
M
=
21 MeV
E
c
X
0
RM : Moliere radius
Ec : Critical Energy [Rossi]
X0 : Radiation length
Critical Energy [see above]:
Further basics:
20 27. Passage of particles through matter
0
0.4
0.8
1.2
0 0.25 0.5 0.75 1
y = k/E
Bremsstrahlung
(
X
0
N
A
/
A
)

y
d
σ
L
P
M
/
d
y
10 GeV
1 TeV
10 TeV
100 TeV
1 PeV
10 PeV
100 GeV
Figure 27.11: The normalized bremsstrahlung cross section k dσ
LPM
/dk in
lead versus the fractional photon energy y = k/E. The vertical axis has units
of photons per radiation length.
2 5 10 20 50 100 200
Copper
X
0
= 12.86 g cm
−2
E
c
= 19.63 MeV
d
E
/
d
x

×

X
0

(
M
e
V
)
Electron energy (MeV)
10
20
30
50
70
100
200
40
Brems = ionization
Ionization
Rossi:
Ionization per X
0
= electron energy
T
o
t
a
l
B
r
e
m
s

E
E
x
a
c
t
b
r
e
m
s
s
t
r
a
h
l
u
n
g
Figure 27.12: Two definitions of the critical energy E
c
.
incomplete, and near y = 0, where the infrared divergence is removed by
the interference of bremsstrahlung amplitudes from nearby scattering centers
February 2, 2010 15:55
dE
dx
(E
c
) ￿ ￿ ￿ ￿

Brems
=
dE
dx
(E
c
) ￿ ￿ ￿ ￿

Ion
E
Sol/Liq
c
=
610 MeV
Z + 1.24
Approximations:
dE
dx ￿ ￿ ￿ ￿

Brems
=
E
X
0
dE
dx ￿ ￿ ￿ ￿

Ion

E
c
X
0
= const.
with:
& ￿

dE
dx ￿

Brems ￿￿

dE
dx ￿

Ion

Z · E
800 MeV
E
Gas
c
=
710 MeV
Z + 0.92
Electromagnetic Showers
X
0
[cm] E
c
[MeV] R
M
[cm]
Pb 0.56 7.2 1.6
Scintillator (Sz) 34.7 80 9.1
Fe 1.76 21 1.8
Ar (liquid) 14 31 9.5
BGO 1.12 10.1 2.3
Sz/Pb 3.1 12.6 5.2
PB glass (SF5) 2.4 11.8 4.3
Typical values for X
0
, E
c
and R
M
of materials
used in calorimeter
A
b
b
i
l
d
u
n
g
8
.
2
:
E
n
t
w
i
c
k
l
u
n
g
e
i
n
e
s
e
l
e
k
t
r
o
m
a
g
n
e
t
i
s
c
h
e
n
S
c
h
a
u
e
r
s
(
M
o
n
t
e
C
a
r
l
o
S
i
m
u
l
a
t
i
o
n
)

N
u
r
d
i
e
P
r
o
z
e
s
s
e
γ
+
K

K
+
e
+
+
e

e
+
K

K
+
e
+
γ
w
e
r
d
e
n
b
e
r
¨u
c
k
s
i
c
h
t
i
g
t
(
K
=
K
e
r
n
)
.

A
u
f
d
e
r
S
t
r
e
c
k
e
X
0
v
e
r
l
i
e
r
t
d
a
s
e

d
u
r
c
h
B
r
e
m
s
s
t
r
a
h
l
u
n
g
d
i
e
H
¨a
l
f
t
e
s
e
i
n
e
r
E
n
e
r
g
i
e
E
1
=
E
0
2

D
a
s
P
h
o
t
o
n
m
a
t
e
r
i
a
l
i
s
i
e
r
t
n
a
c
h
X
0
,
d
i
e
E
n
e
r
g
i
e
v
o
n
P
o
s
i
t
r
o
n
u
n
d
E
l
e
k
t
r
o
n
b
e
t
r
¨a
g
t
E
±
=
E
1
2

F
¨u
r
E
>

t
r
i
t
t
k
e
i
n
E
n
e
r
g
i
e
v
e
r
l
u
s
t
d
u
r
c
h
I
o
n
i
s
a
t
i
o
n
/
A
n
r
e
g
u
n
g
a
u
f
.

F
¨u
r
E

v
e
r
l
i
e
r
e
n
d
i
e
E
l
e
k
t
r
o
n
e
n
E
n
e
r
g
i
e
n
u
r
d
u
r
c
h
I
o
n
i
s
a
t
i
o
n
/
A
n
r
e
g
u
n
g
.
F
o
l
g
e
n
d
e
G
r
¨o
ß
e
n
s
i
n
d
b
e
i
d
e
r
B
e
s
c
h
r
e
i
b
u
n
g
e
i
n
e
s
S
c
h
a
u
e
r
s
v
o
n
I
n
t
e
r
e
s
s
e
:

Z
a
h
l
d
e
r
T
e
i
l
c
h
e
n
i
m
S
c
h
a
u
e
r

L
a
g
e
d
e
s
S
c
h
a
u
e
r
m
a
x
i
m
u
m
s

L
o
n
g
i
t
u
d
i
n
a
l
v
e
r
t
e
i
l
u
n
g
d
e
s
S
c
h
a
u
e
r
s
i
m
R
a
u
m

T
r
a
n
s
v
e
r
s
a
l
e
B
r
e
i
t
e
d
e
s
S
c
h
a
u
e
r
s
W
i
r
m
e
s
s
e
n
d
i
e
l
o
n
g
i
t
u
d
i
n
a
l
e
n
K
o
m
p
o
n
e
n
t
e
n
d
e
s
S
c
h
a
u
e
r
s
i
n
S
t
r
a
h
l
u
n
g
s
l
¨a
n
g
e
n
:
t
=
x
X
0
N
a
c
h
D
u
r
c
h
l
a
u
f
e
n
d
e
r
S
c
h
i
c
h
t
d
i
c
k
e
t
b
e
t
r
¨a
g
t
i
n
u
n
s
e
r
e
m
e
i
n
f
a
c
h
e
n
M
o
d
e
l
l
d
i
e
Z
a
h
l
d
e
r
s
c
h
n
e
l
-
l
e
n
T
e
i
l
c
h
e
n
N
(
t
)
=
2
t
,
1
5
6
Simple shower model:
[from Heitler]
Only two dominant interactions:
Pair production and Bremsstrahlung ...
" + Nucleus ! Nucleus + e
+
+ e
#
[Photons absorbed via pair production]
e + Nucleus ! Nucleus + e + "
[Energy loss of electrons via Bremsstrahlung]
Electromagnetic Shower
[Monte Carlo Simulation]
Shower development governed by X
0
...
After a distance X0 electrons remain with
only (1/e)
th
of their primary energy ...
Photon produces e
+
e
!
-pair after 9/7X0 " X0 ...
Analytic Shower Model
E! = Ee " E0/2
Simplification:
Ee " E0/2
[Ee looses half the energy]
[Energy shared by e
+
/e

]
Assume:
E > Ec : no energy loss by ionization/excitation
E < Ec : energy loss only via ionization/excitation
Use
... with initial particle energy E0
t =
x
X
0
N(t) = 2
t
E =
E
0
N(t)
= E
0
· 2
−t
t = log
2
(
E
0
/E)
N(E
0
, E
c
) = N
max
= 2
t
max
=
E
0
E
c
N(E
0
, E
1
) = 2
t
1
= 2
log
2
(
E
0/E
1
)
=
E
0
E
1
N(E
0
, E
1
) ∝ E
0
t
max
∝ ln(
E
0
/E
c
)
!
Simple shower model:
[continued]
Shower characterized by:
Number of particles in shower
Location of shower maximum
Longitudinal shower distribution
Transverse shower distribution
Longitudinal components;
measured in radiation length ...
Number of shower particles
after depth t:
... use:
Energy per particle
after depth t:
Total number of shower particles
with energy E1:
Number of shower particles
at shower maximum:
Shower maximum at:
8.1 Electromagnetic calorimeters 231
8.1 Electromagnetic calorimeters
8.1.1 Electron–photon cascades
The dominating interaction processes for spectroscopy in the MeV energy
range are the photoelectric and Compton effect for photons and ionisa-
tion and excitation for charged particles. At high energies (higher than
100 MeV) electrons lose their energy almost exclusively by bremsstrahlung
while photons lose their energy by electron–positron pair production [1]
(see Sect. 1.2).
The radiation losses of electrons with energy E can be described by the
simplified formula:

dE
dx

rad
=
E
X
0
, (8.1)
where X
0
is the radiation length. The probability of electron–positron
pair production by photons can be expressed as
dw
dx
=
1
λ
prod
e
−x/λ
prod
, λ
prod
=
9
7
X
0
. (8.2)
A convenient measure to consider shower development is the distance
normalised in radiation lengths, t = x/X
0
.
The most important properties of electron cascades can be understood
in a very simplified model [2, 3]. Let E
0
be the energy of a photon incident
on a bulk of material (Fig. 8.1).
After one radiation length the photon produces an e
+
e

pair; electrons
and positrons emit after another radiation length one bremsstrahlung
photon each, which again are transformed into electron–positron pairs. Let
us assume that the energy is symmetrically shared between the particles at
0
E
0
/
2
E
0
E
0
/
4 E
0
/
8 E
0
/
16
1 2 3 4 5 6 7 8 t [X
0
]
Fig. 8.1. Sketch of a simple model for shower parametrisation.
Analytic Shower Model
Sketch of simple
shower development
≈ (1 +t
0
) ·
E
0
E
c
X
0
∝ E
0
T =
E
0
E
c
· X
0
· F F < 1
T = X
0
t
max−1 ￿

µ=0
2
µ
+t
0
· N
max
· X
0
= X
0
· (2
t
max
−1) +t
0
·
E
0
E
c
X
0
= X
0
· (2
log
2
E
0/E
c
−1) +t
0
·
E
0
E
c
X
0
Analytic Shower Model
Simple shower model:
[continued]
Longitudinal shower distribution increases only logarithmically with the
primary energy of the incident particle ...
Some numbers:   Ec % 10 MeV,   E0 = 1 GeV   ! tmax = ln 100 % 4.5; Nmax = 100
    E0 = 100 GeV  ! tmax = ln 10000 % 9.2; Nmax =10000
Relevant for energy measurement (e.g. via scintillation light):
total integrated track length of all charged particles ...
with t0: range of electron with energy Ec
[given in units of X0]
Energy proportional
to track length ...
As only electrons
contribute ...
[ with ] ￿
θ
2 ￿
≈ (
m
/E)
2
=
1

2 ￿ ￿

θ
2 ￿

3d

19.2 MeV/c
p ￿

x
X
0
R
M
= ￿θ￿
x=X
0
· X
0

21MeV
E
C
X
0
!
Analytic Shower Model
Transverse shower development ...
16 27. Passage of particles through matter
Eq. (27.14) describes scattering from a single material, while the usual problem
involves the multiple scattering of a particle traversing many different layers and
mixtures. Since it is from a fit to a Moli`ere distribution, it is incorrect to add the
individual θ
0
contributions in quadrature; the result is systematically too small. It
is much more accurate to apply Eq. (27.14) once, after finding x and X
0
for the
combined scatterer.
Lynch and Dahl have extended this phenomenological approach, fitting
Gaussian distributions to a variable fraction of the Moli`ere distribution for
arbitrary scatterers [35], and achieve accuracies of 2% or better.
x
s
plane
y
plane
Ψ
plane
θ
plane
x/2
Figure 27.9: Quantities used to describe multiple Coulomb scattering. The
particle is incident in the plane of the figure.
The nonprojected (space) and projected (plane) angular distributions are given
approximately by [33]
1
2π θ
2
0
exp






θ
2
space

2
0





dΩ , (27.15)
1

2π θ
0
exp






θ
2
plane

2
0






plane
, (27.16)
where θ is the deflection angle. In this approximation, θ
2
space
≈ (θ
2
plane,x
+ θ
2
plane,y
),
where the x and y axes are orthogonal to the direction of motion, and
dΩ ≈ dθ
plane,x

plane,y
. Deflections into θ
plane,x
and θ
plane,y
are independent and
identically distributed.
February 2, 2010 15:55
Opening angle
for bremsstrahlung and pair production
Multiple scattering
deflection angle in 2-dimensional plane ... ￿
θ
2
k ￿
=
k ￿

m=1
θ
2
m
= k￿θ
2 ￿

Small contribution as me/Ec = 0.05
Multiple
coulomb scattering
In 3-dimensions extra factor &2:
[% = 1] ￿ ￿

θ
2 ￿

13.6 MeV/c
p ￿

x
X
0
[% = 1]
Assuming the approximate range of electrons
to be X0 yields lateral extension: R =〈&〉X0 ...
Molière Radius;
characterizes lateral shower spread ...
& ￿
θ
2
k ￿
=
k ￿

m=1
θ
2
m
= k￿θ
2 ￿

Insertion – Multiple Scattering
b
' = 2b/v
Atom
Atomic number: Z
}
M,v
θ ≈
∆p
t
p ￿


∆p
t
p
=
2Zze
2
b
1
pv
∆p
t
=
2Zze
2
bv
Reminder:
[Derivation of energy loss ...]
pt
As & ~ Z ! main influence from nucleus;
contribution due to electrons negligible ...
Proof:
&k
Here, the term '&i&j vanishes as successive interactions
are statistically independent; to calculate ‹&k› one needs to average ...
Coulomb
scattering
Multiple
Coulomb scattering ￿

θ
2
k
= ￿

k ￿

m=1 ￿

θ
m ￿

2
=
k ￿

m=1 ￿

θ
2
m
+2 ￿

i￿=j ￿

θ
i ￿

θ
j
=
k ￿

m=1 ￿

θ
2
m ￿
θ
2
N ￿
= N · ￿θ
2 ￿
= N · ￿


0
P(b) [θ(b)]
2
db
= 8π nx
Z
2
z
2
e
4
p
2
v
2
ln
b
max
b
min
N(b) = 2π b db dx · n
N = ￿

N(b)db
= E
2
s ￿

1
pv ￿

2
z
2
x
X
0
X
0
=
1
4αnZ
2
r
2
e
ln(
183
/Z
1
/3
)
r
e
=
e
2
m
e
c
2
P(b) db =
N(b)
N
=
1
N
· 2π b db dx · n

N
N ￿

b
max
b
min ￿

x
0
2π b · n ￿

2Zze
2
bpv ￿

2
db dx
E
s
= ￿


α
m
e
c
2
! bmax/bmin ~ Z
–!
=

α ￿

m
e
c
2 ￿

2 1
p
2
v
2
· xz
2
· ￿

Z
2 ￿

e
2
m
e
c
2 ￿

2
nα ln
ζ
Z
1
/3 ￿

Insertion – Multiple Scattering
Probability for a single collision
with impact parameter b:
n : particle density
dx, x : layer thickness
db, b : impact parameter
N(b) : average number of collisions
N : total number of collisions
with
Estimation of bmin, bmax:
Atomic radius for bmax  :  bmax  = aB !Z
–"
Nuclear size for bmin  :  bmin  ~ A
"
~ Z
"
Also:
, ￿
θ￿ =
21.2 MeV
E
e ￿

x
X
0 ￿
θ
2 ￿
= E
2
s
·
1
p
2
v
2
· z
2
·
x
X
0
!
!
R
M
=
21 MeV
E
c
X
0
E
s
= ￿


α
(m
e
c
2
) = 21.2 MeV
Analytic Shower Model
Deflection angle:
[Molière-Theory]
with
Lateral shower spread:
Main contribution must come from low energy electrons as〈&〉~ 1/Ee, i.e. for electrons with E = Ec ...
Assuming the approximate range of electrons to be X0 yields〈&〉% 21 MeV/Ee ! lateral extension: R =〈&〉X0 ...
&
R
Lateral extension: R = x!tan& % x! &, if & small ...
x
Molière Radius:
Transverse shower development ...
[continued]
Lateral shower spread
characterized by RM !
[Scale Energy]
On average 90% of the shower energy contained
in cylinder with radius RM around shower axis ...
[% = 1, c = 1, z = 1]
dE
dt
= E
0
· β ·
(βt)
α−1
e
−βt
Γ(α)
t
max
=
α −1
β
= ln ￿

E
0
E
c ￿

+C
eγ !
Electromagnetic Shower Profile
Longitudinal profile
dE
dt
= E
0
t
α
e
−βt
8.1 Electromagnetic calorimeters 235
0
0.1
0.01
1
10
100
10 5 15 20 25 30 35
t [X
0
]
d
E
/
d
t

[
M
e
V
/
X
0
]
lead
iron
aluminium
0
500 MeV
1000 MeV
2000 MeV
5000 MeV
0
200
400
600
10 5 15 20
t [X
0
]
d
E
/
d
t

[
M
e
V
/
X
0
]
Fig. 8.4. Longitudinal shower development of electromagnetic cascades. Top:
approximation by Formula (8.7). Bottom: Monte Carlo simulation with EGS4 for
10 GeV electron showers in aluminium, iron and lead [11].
Figure 8.6 shows the longitudinal and lateral development of a 6 GeV
electron cascade in a lead calorimeter (based on [12, 13]). The lateral width
of an electromagnetic shower increases with increasing longitudinal shower
depth. The largest part of the energy is deposited in a relatively narrow
shower core. Generally speaking, about 95% of the shower energy is con-
tained in a cylinder around the shower axis whose radius is R(95%) = 2R
M
almost independently of the energy of the incident particle. The depen-
dence of the containment radius on the material is taken into account by
the critical energy and radiation length appearing in Eq. (8.11).
Parametrization:
[Longo 1975]
(,% : free parameters
t
#
  :  at small depth number of
secondaries increases ...
e
–$t
 :  at larger depth absorption
dominates ...
Numbers for E = 2 GeV (approximate):
( = 2, % = 0.5, tmax = (/%
More exact
[Longo 1985]
[": Gamma function]
with:
[#-induced]
[e-induced]
C

= −0.5
C

= −1.0
1
!1
r/R
M
e

e
+
, e

γ
t
Abbildung 8.3: Schematische Entwicklung eines elektromagnetischen Schauers
/X
0
z
Abbildung 8.4: Longitudinalverteilung der Energiedeposition in einem elektromagnetischen
Schauer f¨ ur zwei Prim¨arenergien der Elektronen
/ r R
M
Abbildung 8.5: Transversalverteilung der Energie in einem elektromagnetischen Schauer in
unterschiedlichen Tiefen gemessen
159
dE
dr
= αe

r
/R
M
+βe

r

min
Electromagnetic Shower Profile
Transverse profile
Parametrization:
(,% : free parameters
RM  : Molière radius
)min : range of low energetic
photons ...
energy deposit
[arbitrary unites]
r/RM r/RM
Inner part: coulomb scattering ...
Electrons and positrons move away
from shower axis due to multiple scattering ...
Outer part: low energy photons ...
Photons (and electrons) produced in isotropic
processes (Compton scattering, photo-electric effect) move away from
shower axis; predominant beyond shower maximum, particularly in high-Z absorber media...
Shower gets wider at larger depth ...
M. Krammer: Detektoren, SS 05
Kalorimeter
36
6.2.1 Elektromagnetische Schauer
Schauerprofil – 2
Quelle: C. Grupen, Teilchendetektoren, B.I. Wissenschaftsverlag, 1993
Longitudinale und transversale Schauerentwicklung einer durch 6 GeV/c Elektronen
ausgelösten elektromagnetischen Kaskade in einem Absorber aus Blei.
Bild links: lineare Skala. – Bild rechts: hablogarithmische Skala
Electromagnetic Shower Profile
M. Krammer: Detektoren, SS 05
Kalorimeter
36
6.2.1 Elektromagnetische Schauer
Schauerprofil – 2
Quelle: C. Grupen, Teilchendetektoren, B.I. Wissenschaftsverlag, 1993
Longitudinale und transversale Schauerentwicklung einer durch 6 GeV/c Elektronen
ausgelösten elektromagnetischen Kaskade in einem Absorber aus Blei.
Bild links: lineare Skala. – Bild rechts: hablogarithmische Skala
lateral shower width [X0]




lo
n
g
it
u
d
in
a
l
s
h
o
w
e
r

d
e
p
t
h

[X
0
]
energy deposit
[arbitrary unites]
lateral shower width [X0]




lo
n
g
it
u
d
in
a
l
s
h
o
w
e
r

d
e
p
t
h

[X
0
]
energy deposit
[arbitrary unites]
Longitudinal and transversal shower profile
for a 6 GeV electron in lead absorber ...
[left: linear scale; right: logarithmic scale]
Electromagnetic shower profiles (longitudinal)
6
Longitudinal Shower Shape
Depth [cm]
E
n
e
r
g
y

d
e
p
o
s
i
t

p
e
r

c
m

[
%
]
Depth [X
0
]
Energy deposit of electrons as a function of depth in a
block of copper; integrals normalized to same value
[EGS4* calculation]
Depth of shower maximum increases
logarithmically with energy
*EGS = Electron Gamma Shower
t
max
∝ ln(
E
0
/E
c
)
Scaling is NOT perfect
9
Pb Z = 82
Fe Z = 26
Al Z = 13
Longitudinal Shower Shape
Depth [X
0
]
E
n
e
r
g
y

d
e
p
o
s
i
t

p
e
r

c
m

[
%
]
Lead
Iron
Aluminum
Energy deposit of electrons as a function
of depth for different materials
[EGS4* calculation]
10 GeV electrons
Approximate scaling ....
σ ∝ Z
5
, E
−3
σ ∝ Z, E
−1
σ increases with E, Z
asymptotic at ∼ 1 GeV
E
c

1
Z
Longitudinal Shower Shape
Electrons Photons
Z
Photons:
Photo-electric effect ...
Compton scattering ...
Pair production ...
Electrons:
Critical energy ...
In high Z materials
particle multiplication ...
... down to lower energies
! longer showers
[with respect to X0]
Transversal Shower Shape
Lateral profile
16
Radial distributions of the energy deposited
by 10 GeV electron showers in Copper
[Results of EGS4 simulations]
Transverse profile
at different shower depths ....
Distance from shower axis [R
M
]
Molière Radii
E
n
e
r
g
y

d
e
p
o
s
i
t

[
a
.
u
.
]
Up to shower maximum broadening
mainly due to multiple scattering ...
Beyond shower maximum broadening
mainly due to low energy photons ...
R
M
=
21 MeV
E
c
X
0
Characterized by RM:
[90% shower energy within RM]
Lateral profile
15
Transversal Shower Shape
Radial energy deposit profiles for
10 GeV electrons showering in Al, Cu and Pb
[Results of EGS4 calculations]
10
1
0.1
0.01
0 1 2 3 4 5
E
n
e
r
g
y

d
e
p
o
s
i
t

[
%
]
Distance from shower axis [R
M
]
Most striking difference seen in
slope of 'tail' or 'halo' ...
Halo
Material dependence:
Scaling almost perfect at low radii ...
Slope considerably steeper for high-Z
material due to smaller mean free path
for low-energy photons ...
Remark:
Even though calorimeters are intended
to measure GeV, TeV their performance
is determined by low energy particles ...
E
c
=
550 MeV
Z
t
max
= ln
E
E
c
R(95%) = 2R
M
R(90%) = R
M
− 1.0
{
Some Useful 'Rules of Thumbs'
X
0
=
180A
Z
2
g
cm
2
− 0.5
− 1.0
[Attention: Definition of Rossi used]
Radiation length:
Critical energy:
Shower maximum:
e

induced shower
" induced shower
Longitudinal
energy containment:
Transverse
Energy containment:
Problem:
Calculate how much Pb, Fe or Cu
is needed to stop a 10 GeV electron.
Pb  : Z = 82 , A = 207, *= 11.34 g/cm
3
Fe  : Z = 26 , A = 56, *= 7.87 g/cm
3
Cu  : Z = 29 , A = 63, *= 8.92 g/cm
3
L(95%) = t
max
+ 0.08Z + 9.6 [X
0
]

Homogeneous Calorimeters
In a homogeneous calorimeter the whole detector volume is filled by a
high-density material which simultaneously serves as absorber as well
as as active medium ...
Advantage: homogenous calorimeters provide optimal energy resolution
Disadvantage: very expensive
Homogenous calorimeters are exclusively used for electromagnetic
calorimeter, i.e. energy measurement of electrons and photons
Signal Material
Scintillation light BGO, BaF2, CeF3, ...
Cherenkov light Lead Glass
Ionization signal Liquid nobel gases (Ar, Kr, Xe)



Homogeneous Calorimeters
Example: CMS Crystal Calorimeter
Homogeneous Calorimeters
Chapter 4
Electromagnetic Calorimeter
4.1 Description of the ECAL
In this section, the layout, the crystals and the photodetectors of the Electromagnetic Calor-
imeter (ECAL) are described. The section ends with a description of the preshower detector
which sits in front of the endcap crystals. Two important changes have occurred to the ge-
ometry and configuration since the ECAL TDR [5]. In the endcap the basic mechanical unit,
the “supercrystal,” which was originally envisaged to hold 6×6 crystals, is now a 5×5 unit.
The lateral dimensions of the endcap crystals have been increased such that the supercrystal
remains little changed in size. This choice took advantage of the crystal producer’s abil-
ity to produce larger crystals, to reduce the channel count. Secondly, the option of a barrel
preshower detector, envisaged for high-luminosity running only, has been dropped. This
simplification allows more space to the tracker, but requires that the longitudinal vertices of
H →γγ events be found with the reconstructed charged particle tracks in the event.
4.1.1 The ECAL layout and geometry
The nominal geometry of the ECAL (the engineering specification) is simulated in detail in
the GEANT4/OSCAR model. There are 36 identical supermodules, 18 in each half barrel, each
covering 20

in φ. The barrel is closed at each end by an endcap. In front of most of the
fiducial region of each endcap is a preshower device. Figure 4.1 shows a transverse section
through ECAL.
y
z
Preshower (ES)
Barrel ECAL (EB)
Endcap
=
1
.
6
5
3
=
1
.
4
7
9
=
2.6
= 3.0
ECAL (EE)
Figure 4.1: Transverse section through the ECAL, showing geometrical configuration.
146
4.1. Description of the ECAL 147
The barrel part of the ECAL covers the pseudorapidity range |η| < 1.479. The barrel granu-
larity is 360-fold in φ and (2×85)-fold in η, resulting in a total of 61 200 crystals.The truncated-
pyramid shaped crystals are mounted in a quasi-projective geometry so that their axes make
a small angle (3
o
) with the respect to the vector from the nominal interaction vertex, in both
the φ and η projections. The crystal cross-section corresponds to approximately 0.0174 ×
0.0174

in η-φ or 22×22 mm
2
at the front face of crystal, and 26×26 mm
2
at the rear face. The
crystal length is 230 mm corresponding to 25.8 X0.
The centres of the front faces of the crystals in the supermodules are at a radius 1.29 m.
The crystals are contained in a thin-walled glass-fibre alveola structures (“submodules,” as
shown in Fig. CP 5) with 5 pairs of crystals (left and right reflections of a single shape) per
submodule. The η extent of the submodule corresponds to a trigger tower. To reduce the
number of different type of crystals, the crystals in each submodule have the same shape.
There are 17 pairs of shapes. The submodules are assembled into modules and there are
4 modules in each supermodule separated by aluminium webs. The arrangement of the 4
modules in a supermodule can be seen in the photograph shown in Fig. 4.2.
Figure 4.2: Photograph of supermodule, showing modules.
The thermal screen and neutron moderator in front of the crystals are described in the model,
as well as an approximate modelling of the electronics, thermal regulation system and me-
chanical structure behind the crystals.
The endcaps cover the rapidity range 1.479 < |η| < 3.0. The longitudinal distance between
the interaction point and the endcap envelop is 3144 mm in the simulation. This location
takes account of the estimated shift toward the interaction point by 2.6 cm when the 4 T mag-
netic field is switched on. The endcap consists of identically shaped crystals grouped in
mechanical units of 5×5 crystals (supercrystals, or SCs) consisting of a carbon-fibre alveola
structure. Each endcap is divided into 2 halves, or “Dees” (Fig. CP 6). Each Dee comprises
3662 crystals. These are contained in 138 standard SCs and 18 special partial supercrystals
on the inner and outer circumference. The crystals and SCs are arranged in a rectangular
Scintillator : PBW0
4
[Lead Tungsten]
Photosensor : APDs [Avalanche Photodiodes]
Number of crystals: ~ 70000
Light output: 4.5 photons/MeV
Example: CMS Crystal Calorimeter
Sampling Calorimeters
Simple shower model
!
Consider only Bremsstrahlung and (symmetric) pair
production
!
Assume X
0
! !
pair
!
After t X
0
:
!
N(t) = 2
t
!
E(t)/particle = E
0
/2
t
!
Process continues until E(t)<E
c
!
E(t
max
) = E
0
/2
tmax
= E
c
!
t
max
= ln(E
0
/E
c
)/ln2
!
N
max
" E
0
/E
c
5
Alternating layers of absorber and
active material [sandwich calorimeter]
Absorber materials:
[high density]
Principle:
Iron (Fe)
Lead (Pb)
Uranium (U)
[For compensation ...]
Active materials:
Plastic scintillator
Silicon detectors
Liquid ionization chamber
Gas detectors
passive absorber
shower (cascade of secondaries)
active layers
incoming particle
Scheme of a
sandwich calorimeter
Electromagnetic shower
Sampling Calorimeters
Advantages:
By separating passive and active layers the different layer materials
can be optimally adapted to the corresponding requirements ...
By freely choosing high-density material for the absorbers one can
built very compact calorimeters ...
Sampling calorimeters are simpler with more passive material and
thus cheaper than homogeneous calorimeters ...

Disadvantages:
Only part of the deposited particle energy is actually detected in the
active layers; typically a few percent [for gas detectors even only ~10
-5
] ...
Due to this sampling-fluctuations typically result in a reduced energy
resolution for sampling calorimeters ...

Sampling Calorimeters
Absorber
Scintillator
Light guide

Photo detector
Scintillator
(blue light)
Wavelength shifter

electrodes
Absorber as
Charge amplifier
HV
Argon
Electrodes
Analogue
signal
Scintillators as active layer;
signal readout via photo multipliers
Scintillators as active
layer; wave length shifter
to convert light
Active medium: LAr; absorber
embedded in liquid serve as electrods
Ionization chambers
between absorber
plates
Possible setups
Sampling Calorimeters
Example:
ATLAS Liquid Argon Calorimeter
Sampling Calorimeters
Example:
H1 SpaCal
[Spaghetti Calorimeter]
Lead matrix ...
[Technical drawing]
4 SpaCal Supermodules
Lead-Fibre Matrix
[Front view]
Sampling Calorimeters
!"#!$#%!!& '!
()*)+,- /0-1234- (5-67)+8 9:;<
!
:;<=:9 642)1-
>45?34232 53 @9(A 5-6 :9BC
/84 :;<=:9 9D@9/ ?,60*4
EF5?4
@434+3,F 2*5G
HIH ??
%
J562
E9 ;(=:2
KL'!! µMN+85--4*
L"!O +85--4*2
% 4
P
28,74F2Q %! R4S
17


Figure 4-2: Cross section of the barrel part of the detection system (top left) and basic construction
elements of the electromagnetic calorimeter, with at the end (bottom left) the basic detector element,
called detector slab.

4.2 The electromagnetic calorimeter concept
All these considerations led to choose high Z number and density material Ior the radiator going
naturally Ior the tungsten, which has a Moliere radius r
M
oI 9mm and a radiation length (X
0
) oI 3.5 mm. For
the active part oI the device, silicon PIN diodes seem perIect apart Irom their cost. The individual e.m.
energy resolution needed and the Iuture cost evolution will drive the area oI silicon and thereIore the
number oI layers. Many studies have been perIormed with 40 layers, as given in the TESLA TDR |1|, while
a prototype under construction will have only 30 layers. The pad size comes Irom a compromise between
1. the density oI extraction lines oI the readout,
2. the electronics cost with the total number oI channels,
3. the eIIective Moliere radius oI the calorimeter.

A pad size oI 1x1cm
2
has been adopted Ior the Iirst studies and Iirst prototype, but the second generation
prototype as well as the recent studies use 5x5 mm·.


The basic detection units are detector slabs (Figure 3.2) which consist oI a stiIIening H-structure Irom
carbon Iiber material embedding a tungsten sheet overlaid by silicon pad sensors. The entire slab is shielded
by an aluminum envelope and is slit into the module. As shown in Figure 4-3, Ior the Iinal project, the only
possible geometry design locates the VFE chip inside the detector. That will be possible only iI

1. the power dissipation is low enough,
2. the duty cycle reduce suIIiciently the power dissipation. II not, a cooling system has to be designed in
this thin device,
3. A 500 GeV electromagnetic shower crossing the VFE chip does not disturb it too strongly.

All these points are part oI the CALICE R&D program. Ensuring the behavior oI the Iirst ECAL prototype
in the test beam, Ior the prototype, the VFE chips are located outside, on the external part oI the PCB.
Tungsten
frame
Detector
slabs
max.: 1.6 m
Tungsten
layer
Sensors
+ r/o electonics
‘Alveolar Structure’
Example: CALICE Electromagnetic
Calorimeter
Calorimetric Systems
FIGURE 5.16. Structure of a detector slab.
The tungsten plate in the center (light grey) is
enclosed by silicon sensor planes (light red) and
PCBs (green) on both sides. The FE chip is
integrated in the PCB.
FIGURE 5.17. The mechanical structure for
the ECAL proposed by US institutes. Tung-
sten planes (grey) of 200 kg weight are joined
to a module by rods (black). Hexagonal sensor
planes (green) are placed in between the tung-
sten plates.
shower maximum using 12 mm
2
pixels. More tests are needed to understand e.g. channel-
by-channel variations.
The goal of the R&D is to fabricate a full-depth electromagnetic calorimeter prototype
module. This will consist of 30 longitudinal layers, each consisting of an about 15 cm diam-
eter silicon detector outfitted with a KPiX chip sandwiched between 2.5 mm thick tungsten
radiator layers. The module will be fully characterized for electromagnetic response and res-
olution in an electron beam, probably at SLAC in 2007. A first round of 10 silicon detectors,
made from a 6 inch wafer, is purchased and tested in the laboratory. Several prototypes of
the KPiX chip are successfully tested. A second, improved version, is under preparation. The
light cable for signal transport inside the gap is being designed and preparations for bump
bonding are underway.
FIGURE 5.18. A silicon sensor pro-
totype with hexagon pads.
FIGURE 5.19. The cross section of the ECAL. The silicon sensors
interspersed between tungsten plates and readout by the KPiX FE
chip bump bonded to metallic traces connected to each sensor cell.
ILC-Detector Concept Report 77
18

Figure 4-3: Schematic view of a possible detector slab. Green is the PCB in which the VFE are
included, the silicon matrices (orange) are glued on PCB. The tungsten sheet (light blue) is wrapped
in carbon fibre (brown). An overall protection sheet in aluminium (grey) surrounds everything.

4.3 Current status of the project
The agenda oI the project concerns also the construction oI a second generation oI prototype, this
time as close as possible to the Iinal detector. This second generation is partially Iunded by the EU program
EUDET. The test with hadron beams is Ioreseen to be done together with the diIIerent HCAL options oI
CALICE.



Figure 4-4 Left: Schematic view of the first generation prototype. Right: ECAL prototype at CERN
test beam area in 2006.

4.4 The prototype in test beam
4.4.1 General description
The ECAL prototype mimics the Iinal project with tungsten wrapped in carbon Iiber, with 30 layers
and a pad size oI 1x1cm
2
but the very Iront end (VFE) electronics is located outside the device, on the same
PCB on which the silicon waIers are glued. The schematic view oI the prototype is shown in Figure 4-4,
with diIIerent shade oI grey Ior the three stacks, each one with diIIerent tungsten thickness. This choice
Tungsten
Silicon Pads
[5 x 5 mm
2
]
PCB
Read-out chip
1.4 mm
~ 3 mm
~ 3 mm
Aluminum protection
Heat shield
PCB with r/o-chips
Tungsten layer
Silicon matrices
Detector
slab
[embeddable?]
Sampling Calorimeters
Example: CALICE Electromagnetic
Calorimeter
71

Figure 12-3: Distributions of total ECAL energy for 30 GeV electrons (left), with a Gaussian fit
superimposed; resolution of the ECAL for electrons (right), with data and Monte Carlo compared.

layers weighted by tungsten thickness
0 10 20 30 40 50 60
e
n
e
r
g
y

w
e
i
g
h
t
e
d

h
i
t
s

/

e
v
e
n
t

[
G
e
V
]
0
0.5
1
1.5
2
2.5
Entries 30
Mean 17.78
RMS 9.518
=1.5GeV
beam
run 230243, E
=3GeV
beam
run 230248, E
=6GeV
beam
run 230252, E
=6GeV
beam
run 300670, E
=10GeV
beam
run 300672, E
=20GeV
beam
run 300676, E
=30GeV
beam
run 300207, E

energy (GeV)
0 1 2 3 4 5 6 7
R
a
d
i
u
s

(
m
m
)
10
15
20
25
30
35
40
center
90% containment
95% containment

Figure 12-4 Left: Longitudinal shower profile from runs at CERN and at DESY with several energies
measured in the CALICE ECAL; the data are represented by points with statistical uncertainties
and the GEANT4 Monte Carlo simulation by the histogram. Parametrisations of the shower profiles
by the form c t
!
e
(-! t)
, where t is the calorimeter depth in radiation lengths, are also shown. Right:
Radius for containment of 90º or 95º of the energy of electron showers from 1-6 GeV.

12.3 The AHCAL data
The analogue HCAL (AHCAL) was tested at CERN in 2006 with 15/23 oI its intended 38 layers oI
scintillating tiles during the Iirst/second period oI data taking. A calorimeter with roughly 3.8 interaction
lengths was exposed to the beam by placing active layers in alternating slots oI the steel structure. Seven
additional modules have been used to Iurther complete the instrumentation at the beginning oI the absorber
stack Ior the second period without increasing the overall depth. Data were taken both with and without the
ECAL (!1"
0
) in Iront. Many hadronic showers should thereIore be contained, albeit with a lower sampling
Iraction than will be achieved ultimately. The AHCAL analysis strategy is to use muon and electron data,
taken without the ECAL, to understand the calorimeter calibration and perIormance and to tune the
simulation iI necessary. The hadron data taken with and without the ECAL can then be studied with greater
conIidence to extract the perIormance and shower properties. These data are oI great interest in comparing
with simulations, Ior which the underlying interaction model is much less well understood than Ior
electromagnetic processes.

The calibration procedure Ior the AHCAL is more complex than Ior the ECAL. As well as using muon
beams to establish the MIP calibration tile-by-tile, the SiPM photosensors exhibit a signiIicantly non-linear
response, and careIul calibration oI this eIIect is needed, based both on lab measurements and a light
injection system. It is also important to include these eIIects into the Monte Carlo simulations.
Sampling Calorimeters
71

Figure 12-3: Distributions of total ECAL energy for 30 GeV electrons (left), with a Gaussian fit
superimposed; resolution of the ECAL for electrons (right), with data and Monte Carlo compared.

layers weighted by tungsten thickness
0 10 20 30 40 50 60
e
n
e
r
g
y
w
e
ig
h
te
d
h
its
/ e
v
e
n
t [G
e
V
]
0
0.5
1
1.5
2
2.5
Entries 30
Mean 17.78
RMS 9.518
=1.5GeV
beam
run 230243, E
=3GeV
beam
run 230248, E
=6GeV
beam
run 230252, E
=6GeV
beam
run 300670, E
=10GeV
beam
run 300672, E
=20GeV
beam
run 300676, E
=30GeV
beam
run 300207, E
energy (GeV)
0 1 2 3 4 5 6 7
R
a
d
iu
s
(
m
m
)
10
15
20
25
30
35
40
center
90% containment
95% containment

Figure 12-4 Left: Longitudinal shower profile from runs at CERN and at DESY with several energies
measured in the CALICE ECAL; the data are represented by points with statistical uncertainties
and the GEANT4 Monte Carlo simulation by the histogram. Parametrisations of the shower profiles
by the form c t
!
e
(-! t)
, where t is the calorimeter depth in radiation lengths, are also shown. Right:
Radius for containment of 90º or 95º of the energy of electron showers from 1-6 GeV.

12.3 The AHCAL data
The analogue HCAL (AHCAL) was tested at CERN in 2006 with 15/23 oI its intended 38 layers oI
scintillating tiles during the Iirst/second period oI data taking. A calorimeter with roughly 3.8 interaction
lengths was exposed to the beam by placing active layers in alternating slots oI the steel structure. Seven
additional modules have been used to Iurther complete the instrumentation at the beginning oI the absorber
stack Ior the second period without increasing the overall depth. Data were taken both with and without the
ECAL (!1"0) in Iront. Many hadronic showers should thereIore be contained, albeit with a lower sampling
Iraction than will be achieved ultimately. The AHCAL analysis strategy is to use muon and electron data,
taken without the ECAL, to understand the calorimeter calibration and perIormance and to tune the
simulation iI necessary. The hadron data taken with and without the ECAL can then be studied with greater
conIidence to extract the perIormance and shower properties. These data are oI great interest in comparing
with simulations, Ior which the underlying interaction model is much less well understood than Ior
electromagnetic processes.

The calibration procedure Ior the AHCAL is more complex than Ior the ECAL. As well as using muon
beams to establish the MIP calibration tile-by-tile, the SiPM photosensors exhibit a signiIicantly non-linear
response, and careIul calibration oI this eIIect is needed, based both on lab measurements and a light
injection system. It is also important to include these eIIects into the Monte Carlo simulations.
Energy Resolution
Ebeam = 1 GeV
Ebeam = 45 GeV
< 20%/$E
ECAL ‘Physics’ Prototype
[CALICE]
Front End
Electronics
Structure 4.2
Structure 2.8
Structure 1.4
Active area
Event
Display
[Pads: 10x10 mm
2
]
Example: CALICE Electromagnetic
Calorimeter
Homogeneous vs. Sampling Calorimeters
28. Detectors at accelerators 57
Table 28.8: Resolution of typical electromagnetic calorimeters. E is in GeV.
Technology (Experiment) Depth Energy resolution Date
NaI(Tl) (Crystal Ball) 20X
0
2.7%/E
1/4
1983
Bi
4
Ge
3
O
12
(BGO) (L3) 22X
0
2%/

E ⊕0.7% 1993
CsI (KTeV) 27X
0
2%/

E ⊕0.45% 1996
CsI(Tl) (BaBar) 16–18X
0
2.3%/E
1/4
⊕1.4% 1999
CsI(Tl) (BELLE) 16X
0
1.7% for E
γ
> 3.5 GeV 1998
PbWO
4
(PWO) (CMS) 25X
0
3%/

E ⊕0.5%⊕0.2/E 1997
Lead glass (OPAL) 20.5X
0
5%/

E 1990
Liquid Kr (NA48) 27X
0
3.2%/

E⊕ 0.42%⊕0.09/E 1998
Scintillator/depleted U 20–30X
0
18%/

E 1988
(ZEUS)
Scintillator/Pb (CDF) 18X
0
13.5%/

E 1988
Scintillator fiber/Pb 15X
0
5.7%/

E ⊕0.6% 1995
spaghetti (KLOE)
Liquid Ar/Pb (NA31) 27X
0
7.5%/

E ⊕0.5%⊕0.1/E 1988
Liquid Ar/Pb (SLD) 21X
0
8%/

E 1993
Liquid Ar/Pb (H1) 20–30X
0
12%/

E ⊕1% 1998
Liquid Ar/depl. U (DØ) 20.5X
0
16%/

E ⊕0.3%⊕0.3/E 1993
Liquid Ar/Pb accordion 25X
0
10%/

E ⊕0.4%⊕0.3/E 1996
(ATLAS)
traditional sandwich structure is the LAr-tube design shown in Fig. 28.22(a).
A relatively new variant is the use of Cerenkov light in hadron calorimetry. Such
a calorimeter is sensitive to e
±
’s in the EM showers plus a few relativistic pions. An
example is the radiation-hard forward calorimeter in CMS, with iron absorber and quartz
fiber readout by PMT’s.
Ideally, the calorimeter is segmented in φ and θ (or η = −lntan(θ/2)). Fine
segmentation, while desirable, is limited by cost, readout complexity, practical geometry,
and the transverse size of the cascades. An example, a wedge of the ATLAS central barrel
calorimeter, is shown in Fig. 28.22(b).
In an inelastic hadronic collision a significant fraction f
em
of the energy is removed
from further hadronic interaction by the production of secondary π
0
’s and η’s, whose
decay photons generate high-energy electromagnetic (EM) cascades. Charged secondaries
January 29, 2010 16:23
H
o
m
o
g
e
n
e
o
u
s
S
a
m
p
l
i
n
g
Resolution of typical
electromagnetic calorimeter
[E is in GeV]