You are on page 1of 35

TITLE FUNDAMENTS - 1

PRINCIPLES OF GAS DETECTORS


Fabio Sauli
TERA Foundation
CERN
CH-1211 Geneva Switzerland

Part 1: Fundaments
Part 2: Detectors
fabio.sauli@cern.ch
http://fabio.home.cern.ch/fabio/
http://gdd.web.cern.ch/GDD/

F. Sauli - Gas Detectors - KEK March 14, 2009


BASIC BIBLIOGRAPHY FUNDAMENTS - 2

D.H. Wilkinson: Ionization Chambers and Counters (Cambridge Univ. Press, 1950)
S.A. Korff: Electron and Nuclear Counters (Van Nostrand, 1955)
P. Rice-Evans: Spark, Streamer, Proportional and Drift Chambers (Richelieu, 1974)
F. Sauli: Principles of Operation of Multiwire Proportional and Drift Chambers (CERN 77-09, 1977)
Th. Ferbel, Editor: Techniques and Concepts of High-energy Physics (Plenum, 1983)
R.C. Fernow: Introduction to Experimental Particle Physics (Cambridge Univ. Press, 1986)
W.R. Leo: Techniques for Nuclear and Particle Physics Experiments (Springer, 1987)
C. Fabjan and J. Pilcher, ed.: Instrumentation in Elementary Particle Physics (World Scientific, 1988)
C.F.G. Delaney and E.C. Finch: Radiation Detectors (Clarendon Press, 1992)
R. Gilmore: Single Particle Detection and Measurement (Taylor and Francis, 1992)
F. Sauli, ed.: Instrumentation in High Energy Physics (World Scientific, 1992)
K. Grupen: Particle Detectors (Cambridge Monographs on Part. Phys. 1996)
K. Kleinknecht: Detectors for Particle Radiation (Cambridge Univ. Press 1998)
G.F. Knoll: Radiation Detection and Measurements, 3d Ed. (Wiley, 2000)
W. Blum, W. Riegler and L. Rolandi: Particle Detection with Drift Chambers, 2d Ed. (Springer 2008)

F. Sauli - Gas Detectors - KEK March 14, 2009


ENERGY LOSS FUNDAMENTS - 3

DIFFERENTIAL ENERGY LOSS OF CHARGED PARTICES (Z=1) IN MATERIALS:

Expressed in MeV g-1 cm2, the differential


energy loss is equal within a factor of two for
all materials (except H2):

 (g cm 2 )   (g cm 3 ) l(cm)
dE 1 dE  : density

d  dx

dE
~ 1.5 MeV g 1 cm2
d


 SEE:
Review of Particle Physics
Physics Letters B 667(2008)1-1340

http://pdgLive.lbl.gov

F. Sauli - Gas Detectors - KEK March 14, 2009


PHYSICAL PROPERTIES OF GASES FUNDAMENTS - 4

DIFFERENTIAL ENERGY LOSS, PRIMARY AND TOTAL IONIZATION FOR MINIMUM


IONIZING, Z=1 PARTICLES
NORMAL TEMPERATURE AND PRESSURE (NTP: 20°C, 1 ATMOSPHERE)

GAS Density EX EI WI dE/dx NP NT


g cm-2 eV eV eV keV cm-1 cm-1 cm-1
Ne 0.839 10-3 16.7 21.6 30 1.45 13 50
Ar 1.66 10-3 11.6 15.7 25 2.65 25 106
Xe 5.495 10-3 8.4 12.1 22 6.87 41 312
CH4 0.667 10-3 8.8 12.6 30 1.61 37 54
C2H6 1.26 8.2 11.5 26 2.91 48 112
i-C4H10 2.49 10-3 10.6 26 5.67 90 220
CO2 1.84 10-3 7 13.8 34 3.35 35 100
CF4 3.78 10-3 10 16 54 6.38 63 120

Z : atomic number ; A : atomic mass;  : density


Ex, Ei : first excitation and ionization potentials
wi: average energy per ion pair
nP , nT : primary and total ion pairs per cm

(From various sources)

F. Sauli - Gas Detectors - KEK March 14, 2009


PRIMARY IONIZATION FUNDAMENTS - 5

ELECTRON-ION PAIR PRODUCTION


Coulomb interactions between the electric field of the particle and of the
molecules of the medium produce electron-ion pairs.
Multiple ionizing collisions follow Poisson’s statistics:
n k n n: average
P  e
k
n

k! k: actual number
Detection efficiency:   1  P0n  1  en
Minimum ionizing particles in argon NTP (nP: 25 cm-1)
 s (mm)  (%)
1 91.8
2 99.3
Distribution of the electron closest to an electrode:
A1n (x)  n enx
A1n (t)  n enwt
w: drift velocity (~ 5 cm µs-1)

Limit in time resolution of


proportional counters:
arrival at anode wire of the
closest electron.

F. Sauli - Gas Detectors - KEK March 14, 2009


SECONDARY AND TOTAL IONIZATION FUNDAMENTS - 6

CLUSTER SIZE PROBABILITY IN ARGON

Primary electrons can


further ionize the medium
producing local electron-
ion clusters. Occasionally,
the primary electron has
enough energy to produce
a long trail (delta
electron).

E
Total number of ion pairs: nT  E: energy loss
wi
wi : average energy per ion pair

The average ionization energy is about the same in all gases


 energy and type of particles.
and does not depend from

For minimum ionizing particles in Argon:


E = 2.4 keV/cm wi = 26 eV
H. Fischle et al,
nT Nucl. Instr. and Meth. A301 (1991) 202
nT ≈ 90 ion pairs/cm 3
nP
F. Sauli - Gas Detectors - KEK March 14, 2009
ELECTRONS RANGE IN MATERIALS FUNDAMENTS - 7

Due to multiple scattering and ionizing


collisions, the penetration of electrons in
materials is shorter than the integrated
range along the path; the practical range is
the extrapolated thickness of material
absorbing all the electrons.

Practical range

Integrated path

Fit to experimental data (light elements):


r : practical range in µg cm-2
r = 10 E 1.7
E : electron energy in keV
r
R R : range in cm
  : density in µg cm-3
H. Kanter, Phys. Rev. 121(1961)461

F. Sauli - Gas Detectors - KEK March 14, 2009


APPROXIMATE EXPRESSION FOR ELECTRON RANGE FUNDAMENTS - 8

PRACTICAL ELECTRON RANGE IN GASES AT NTP

A 2 keV delta electron in argon STP has a


practical range of ~ 200 µm.

The asymmetry in released charge affects the


localization accuracy in detectors exploiting the
measurement of the center of gravity (Time
projection Chambers):

180 µm

REAL COG

2 keV

F. Sauli - Gas Detectors - KEK March 14, 2009


HEED FUNDAMENTS - 9

CALCULATION OF PRIMARY IONIZATION AND ELECTRON RANGE

PRIMARY CLUSTERS PER cm (STP):


ELECTRON RANGE IN ARGON (STP):

HEED

2 keV

140 µm

I. B. Smirnov, Nucl. Instr. and Meth. A554(2005)474

HEED: http://consult.cern.ch/writeup/heed/

F. Sauli - Gas Detectors - KEK March 14, 2009


IONIZATION STASTISTICS - 1 FUNDAMENTS - 10

WIDE ENERGY LOSS SPREAD (LANDAU DISTRIBUTION)


The statistics of the energy loss, with wide fluctuations and a long tail (due to delta electrons) requires
statistical analysis of hundreds of samples for determination of the average (as done in Time Projection
Chambers)

I. B. Smirnov, Nucl. Instr. and Meth. A554(2005)474

F. Sauli - Gas Detectors - KEK March 14, 2009


IONIZATION STATISTICS - 2 FUNDAMENTS - 11

The presence of long range delta electrons can substantially


affect the localization accuracy:

DRIFT
Coordinate deduced from drift time:

~5%
CENTER OF GRAVITY
Coordinate from cathode induced charge

F. Sauli, Nucl. Instr. and Meth. 156 (1978) 147

G. Charpak et al, Nucl. Instr. and Meth. 167 (1979) 455

F. Sauli - Gas Detectors - KEK March 14, 2009


ELECTRONS DRIFT AND DIFFUSION FUNDAMENTS - 12

Drift velocity and diffusion of electrons vary in a wide range, depending the gas mixture:
2k x k : characteristic energy 2KT x
x  Thermal limit: x 
e E x: drift distance e E
E: electric field
2k P x
The diffusion at equal E/P depends on the inverse square root of pressure: x 
e E P

DRIFT VELOCITY: DIFFUSION:




1.5 mm

250 µm

F. Sauli - Gas Detectors - KEK March 14, 2009


TRANSPORT THEORY OF ELECTRON DRIFT FUNDAMENTS - 13

Charge transport processes are determined by electron-molecule cross sections:

MAGBOLTZ:
Montecarlo program to compute electron
drift and diffusion
S. Biagi, Nucl. Instr. and Meth. A421(1999)234

http://rjd.web.cern.ch/rjd/cgi-bin/cross

F. Sauli - Gas Detectors - KEK March 14, 2009


MIXTURES FUNDAMENTS - 14

Addition to a noble gas of even small percentages of a molecular gas has dominant effect on the
electron cross section:

CO2 100

CO210
Ar 100

CO2 2

F. Sauli - Gas Detectors - KEK March 14, 2009


DRIFT VELOCITY FUNDAMENTS - 15

ELECTRON DRIFT VELOCITY IN ARGON-METHANE MIXTURES:

(Computed with MAGBOLTZ)

F. Sauli - Gas Detectors - KEK March 14, 2009


LONGITUDINAL AND TRANSVERSE DIFFUSION FUNDAMENTS - 16

At low electric fields, the diffusion is symmetric. At T


moderate to high fields however the longitudinal diffusion E Field L Drift
(in the direction of drift) is reduced.

TRANSVERSE DIFFUSION: LONGITUDINAL DIFFUSION:

In drift chambers, the dispersive factor is the longitudinal diffusion (measured time in the direction of
the electric field)
In time projection chambers, the dispersive factor is the transverse diffusion (center of gravity of
charge induced on pad rows)

F. Sauli - Gas Detectors - KEK March 14, 2009


MAGNETIC FIELD FUNDAMENTS - 17


B
The drifting electrons swarm is rotated by an angle B in
the plane perpendicular to E and B; the magnetic drift

velocity is wB ≤ w0 E
r r 
E B tan  B    
wB
 
B E 
 wB   : mean collision time
E B 1  2  2
wB 
  eB /m Larmor frequency

r r
E B 
 s wB  w0
 E
 T
Drift velocity
s w
L B
L   0
unchanged
r 0
T  The transverse
B 1  2 2 diffusion is reduced

 

EB
  e
  

 

ExB
  
   
2 2 B( E  B )

w E
2 2 
      Friction force theory
 m 1     B B 2


F. Sauli - Gas Detectors - KEK March 14, 2009



TRANSVERSE DIFFUSION IN MAGNETIC FIELD FUNDAMENTS - 18

IN SOME GASES THE TRANSVERSE DIFFUSION IS STRONGLY REDUCED


Improves the precision of the projected coordinate measurement in Time Projection Chambers

r r
200 V/cm
E B

600 µm



50 µm

F. Sauli - Gas Detectors - KEK March 14, 2009


ELECTRON ATTACHMENT FUNDAMENTS - 19

Electrons are lost by radiative or non-radiative capture to resulting in the formation of negative ions: e
+ a -> A- (+h). The attachment cross section is gas and energy-dependent, therefore strongly depends
on the gas composition and electric field. For equal amount of oxygen contamination, capture losses are
much more severe in “cold” gases. In the example, a 5% loss is observed for 20 cm drift for 15 ppm of
oxygen in A-CO2 or 800 ppm in Ar-CH4.

ELECTRONS SURVIVING AFTER 20 CM


DRIFT (E = 200 V/cm):
OXYGEN ATTACHMENT COEFFICIENT:

F. Sauli - Gas Detectors - KEK March 14, 2009


EXCITATION AND CHARGE MULTIPLICATION FUNDAMENTS - 20

Electrons on the high side of the energy distribution reach the excitation and ionization levels,
inducing inelastic collisions.
CROSS SECTIONS AT HIGH ELECTRIC ELECTRONS ENERGY DISTRIBUTION IN
FIELDS: ARGON AT INCREASING FIELDS:

Ex=10.6 eV Ei=15.7 eV

EXCITATION 11.6 eV

IONIZATION
15.7 eV

F. Sauli - Gas Detectors - KEK March 14, 2009


INELASTIC COLLISION PROCESSES IN MIXTURES FUNDAMENTS - 21

Radiative processes with the emission of a


short wavelength photon can induce various
kinds of secondary effects, as internal
reconversion to charge on of molecules with
low ionization potential or emission of
photoelectron at cathodes. Addition to noble
gases of molecular additives reduce these
effects directly, quenching the emissions, or
by absorption.

MAJOR PROCESSES:

Radiative recombination: A++ B -> AB + h J.Meek and J. D. Cragg, Electrical


Radiative capture: e + M -> M + h
-
Breakdown of Gases (Clarendon, 1953)
Dissociative capture: e + AB -> AB -> A + B
- -

Three-body collision: e + A = B -> A- + B


Excimer formation and decay: A* + A -> A*2 -> A + A + h
Penning effect: A*+B -> A + B* + e [Ei(B) < Ex(A)]

F. Sauli - Gas Detectors - KEK March 14, 2009


PHOTON EMISSION SPECTRA IN NOBLE GASES FUNDAMENTS - 22

The emission spectra after excitation and dimers formation of noble gases are peaked in the far
ultraviolet. The low ionization potential vapors used in Cherenkov ring imaging detectors, as
Triethylamine (TEA) and Tetrakis-dimethylamino ethylene (TMAE), added to noble gases, act as
internal wavelength shifters and result in the emission of photons at longer wavelengths:
Relative light yield Energy (eV )
15 10 5 4 3 2
1.0

0.8
Kr
TEA
TMAE
0.6

Ar
0.4
Xe

0.2

0
100 200 300 400 500 600
Wavelength (nm)

IMAGING CHAMBERS
SCINTILLATING PROPORTIONAL COUNTERS

F. Sauli - Gas Detectors - KEK March 14, 2009


COLLISIONAL IONIZATION: TOWNSEND COEFFICIENT FUNDAMENTS - 23

Electrons acquiring enough energy from the field can have ionizing collisions with molecules, resulting in
with creation of an electron-ion pair.
1
Mean free path for ionization:  N: molecules/cm3
N
1
First Townsend coefficient:  Ionizing collisions/cm

TOWNSEND COEFFICIENT FOR NOBLE TOWNSEND COEFFICIENT FOR Ar-CH4:


GASES:

(MAGBOLTZ)

F. Sauli - Gas Detectors - KEK March 14, 2009


AVALANCHE MULTIPLICATION IN UNIFORM FIELD FUNDAMENTS - 24

VISUALIZATION OF AVALANCHES
COMBINING A CLOUD CHAMBER
l WITH AN AVALANCHE CHAMBER:
E x
Ions

Electrons

At each mean free path for ionization, electrons create an


electron-ion pair; results an exponential increase of charge, with
fast electrons on the front and slow ions left behind.
Incremental increase of the number of electrons in the
avalanche:
dn  n  dx n( x)  n0e x

Multiplication factor or Gain:


n x
M ( x)   e
n0

Maximum Avalanche size before discharge (Raether limit):


QMAX ≈ 107 e H. Raether, Electron Avalanches and
Breakdown in Gases (Butterworth 1964)

F. Sauli - Gas Detectors - KEK March 14, 2009


SIGNAL INDUCTION ON ELECTRODES FUNDAMENTS - 25

The multiplying and moving charges in the


avalanche induce signals on the electrodes.
The incremental charge induction due to electrons
after a path s:
ds
dq  en0e s
s0

Integrating over s:
 en0 s en0 s en0 w t
q (s)  (e 1)  e  e
s0 s0 s0

and the corresponding current :


 dq en 0 w  w  t en 0 w  t
i (t)   e   e
dt s0 T

The current signal induced by ions is given by:


 en0  wt w*t  Fast electron signal
i (t )   e e  0  t  T 
T   1 1 1
*
  
en0  s w* t   w w w
 

i (t )   e  e  T  t T
T   Slow ion tail

J. Townsend, Electrons in Gases(Hutchinson 1947)

F. Sauli - Gas Detectors - KEK March 14, 2009


AVALANCHE STATISTICS IN UNIFORM FIELDS FUNDAMENTS - 26

In constant electric field, the probability of an avalanche started by a single electron to have a size N is
given by Furry’s law:
N
1 N SIZE DISTRIBUTIONS FOR AVALANCHES
P(N )  e
N STARTED BY 1, 2,... 10 ELECTRONS:
N  e s : average multiplication factor on
the gap s
 N 
P ,n
The maximum probability is for N=0 (no N 
multiplication!).
The Furry distribution has a variance equal to 
the average:
N
1
N

For an avalanche started by n electrons:


N
 n1  N N
N  e N
P(n, N )   
N  (n 1)!

H. Genz, Nucl. Instr. and Meth. 112(1973)83


F. Sauli - Gas Detectors - KEK March 14, 2009


AVALANCHE SIZE DISTRIBUTION FUNDAMENTS - 27

At large gains (high fields) the AVALANCHE SIZE


avalanche distribution is described POLYA DISTRIBUTIONS: DISTRIBUTIONS AT
by a Polya function: INCREASING FIELDS:
(k  1) k 1 k (k 1)z z  N
P(z)  z e
(k  1) N

for k integer (k  1)  (k)!


The relative varianceof the Polya
distribution is:
 
2
1 1 1
 N     bb
 N  N 1  k N

For k=0 the distributions reduce to


a Furry law.

• The shape of the single electron avalanche distribution has


a major relevance in determining the energy resolution of
proportional counters
• A peaked single electron pulse height distribution provides
efficient detection (RICH)

H. Sclumbohm,
Zeit. Physik 151(1958)563

F. Sauli - Gas Detectors - KEK March 14, 2009


PROPORTIONAL COUNTER FUNDAMENTS - 28

Thin anode wire of radius a, coaxial with a cylindrical cathode of radius b

Electric field: CV0 1


E(r) 
2 0 r
CV0 r
Potential: V (r)  ln
20 a
AVALANCHE REGION
V (b)  V0 V (a)  0
2 0

ELECTRIC FIELD
C capacitance per unit length
ln b 
a
 

DRIFT AND COLLECTION REGION

Cathode radius b

THRESHOLD FIELD FOR


MULTIPLICATION

Anode radius a

a
DISTANCE FROM CENTER

F. Sauli - Gas Detectors - KEK March 14, 2009


PROPORTIONAL COUNTER: AVALANCHE DEVELOPMENT FUNDAMENTS - 29

Electrons approach the anode; on reaching


a critical value of field strength, they start + + +
- + +
an avalanche multiplication, continuing -
until the front reaches the wire. Ions are left
behind in a characteristic drop shape. + + + +
The extent by which the avalanche
surrounds the wire depends on gas,
geometry and gain.

ln M Streamer
GAIN CHARACTERISTICS: Saturation Breakdown
At increasing fields, to a region of Multiplication
charge collection (ionization chamber) Collection
follows a region of multiplication with
the detected charge proportional to the Attachment
initial ionization. At higher voltage
follow a region of limited
proportionality, saturation (with the n1
output charge independent from initial n2
IONIZATION PROPORTIONAL
ionization), streamer formation and
CHAMBER COUNTER
breakdown.
Voltage

F. Sauli - Gas Detectors - KEK March 14, 2009


PROPORTIONAL COUNTERS: INDUCED SIGNALS FUNDAMENTS - 30

For an avalanche starting at a small distance  from the anode, the electron and ion contributions to the
induced charge are: INDUCED CHARGE:
 Q a   dV QC a  
q   dr   ln q(t)
V0 a dr 2 0 a
Q b dV QC b
q   dr   ln
V0 a   dr 2 0 a  

Total induced signal on anode:


QC b Q
q  q  q   ln  Q (+Q on cathode)
2 0 a
0 100 200 300 400 500
Ratio of electron and ion signals: t (µs) T+
q ln(a   )  ln a FAST SIGNAL DIFFERENTIATION:

 ~1% for typical geometry
q lnb  ln(a   ) 50 ns
q(t)
Time development of the signal on anode: 100 ns
QC   CV0  QC  t  300 ns
q(t )   ln
1 t   ln1 
2 0  2 
 20a  2  0  t 0 
Total ions drift time:
 0 (b2  a2 )
T  q(T  )  Q t(ns)
 CV0 0 100 200 300 400 500

F. Sauli - Gas Detectors - KEK March 14, 2009


PROPORTIONAL COUNTERS: ENERGY RESOLUTION FUNDAMENTS - 31

The energy resolution is a convolution of ionization


statistics, avalanche spread and electronics noise:
2 2 2 2
 E   N   M   el 

RESOLUTION %
          
 E   N   M  M  TOTAL
For soft X-rays:   FN
2
N F: Fano factor
Gain variance: AVALANCHE
 M 2 1  A 2 A Single electron
     avalanche variance
IONIZATION
 M  
N  A  A NOISE
A
for a Polya avalanche distribution b GAIN
2 A
 E  1 PULSE HEIGHT SPECTRUM FOR 5.9 keV
  (F  b)
 E  N X-RAYS IN P10 (Ar-CH4 90-10):

GAS 
F(calc) F(exp)

 Ar 0.17 0.19 1.3 


 0.22 fwhm E  9%
Ar-CH4 0.17 0.19 5.9 E
Xe <0.17
Ne+0.5%Ar 0.05

 fwhm 1.3 keV


For 5.9 keV X-rays (N~220):
E
 7% for b=1
E

F. Sauli - Gas Detectors - KEK March 14, 2009


SCINTILLATING PROPORTIONAL COUNTERS FUNDAMENTS - 32

In noble gases, at moderate electric fields before multiplication, there is a large emission of scintillation
photons. In proportional scintillation counters the detection of these photons eliminate the dispersion due
to the avalanches and achieve the best energy resolution (close to the statistical)
 E 2 F CHARGE AND LIGHT YIELD VS VOLTAGE:
  
 E  N

SPHERICAL ANODE COUNTER:


Xe 99.95%
 1030 torr

CHARGE

LIGHT

A.J.P.L. Policarpo et al, Nucl. Instr. and Meth. 102(1972)337

F. Sauli - Gas Detectors - KEK March 14, 2009


ENERGY RESOLUTION OF SCINTILLATION COUNTERS FUNDAMENTS - 33

55
Fe X-RAYS (5.898 keV): FWHM E
R  8.5%  3.6%
E E

Xe 99.95% Primary statistics limit:


1030 torr 
E F
  2.8%
E N

241
Am ENERGY SPECTRUM:
FWHM 500 eV 

A.J.P.L. Policarpo et al,


Nucl. Instr. and Meth. 102(1972)337

H. E. Palmer, IEEE Trans. Nucl. Sci.NS-22(1975)100

Fluorescence analysis
X-Ray Spectroscopy
F. Sauli - Gas Detectors - KEK March 14, 2009
IMAGING CHAMBERS FUNDAMENTS - 34

The light emission in avalanches has been


exploited to detect tracks with simple optical COSMIC RAY ACTIVITY IN A 10x10x10
recorders (solid state cameras). cm3 SENSITIVE VOLUME:

The UV light emission in the avalanches is


converted into the visible using an internal
wavelength shifter (TMAE gas) or a thin WLS
on the semi-transparent anode.

DRIFT VOLUME AVALANCHE M. Suzuki et al, Nucl. Instr. and Meth. A263(1988)237
MULTIPLICATION

F. Sauli - Gas Detectors - KEK March 14, 2009


3-D OPTICAL IMAGING CHAMBER FUNDAMENTS - 35

IMAGES OF NUCLEAR DECAYS:

Optical imaging chamber with


recording of the projected image
using a CCD camera, and the time
profile of the emitted light with a
photomultiplier.
Simultaneous recording of
projection and time development of
the emission permits a 3-D
reconstruction of tracks.

Tested with radioactive ion beams


stopping in the gas volume.

K. Miernik et al, Nucl. Instr. and Meth. A581(2007)194

F. Sauli - Gas Detectors - KEK March 14, 2009

You might also like