lect 18 radiation detectors - Engineering Physicsphys352/lect18.pdf · Radiation Detectors I: Gas Detectors Prelude: General Properties •radiation/particles interacting in materials

PHYS 352

Radiation Detectors I: Gas Detectors

Prelude: General Properties

• radiation/particles interacting in materials cause

• ionization

• excitation of atoms/molecules/of electrons in the crystal lattice

• radiation detectors work by using these interactions and either collecting the charges or observing the de-excitation

• three classes of detectors to be examined

• gas

• semiconductor

• scintillation

• older radiation detector technologies include: ionization that nucleates bubbles (e.g. cloud chamber, bubble chamber); ionization tracks that chemically alter emulsion layers (e.g. film, nuclear track etch detectors)

we won’t discuss these older radiation detectors any further

First Thing: Radiation Has to Get Into the Detector

• the radiation you want to detect has to get in the detector

• radiation detectors have a “window” and/or dead layer

• alphas? require very thin window and/or very thin dead layer

• betas? can be a little thicker though energy loss and backscattering can be important

• gammas rays? X-rays? these need to penetrate into the sensitive volume of the detector...and interact

Second: Detector Efficiency

• intrinsic efficiency to charged particles (alphas, betas) is typically ~100%

• would be slightly less than 100% if the window or dead layer is too thick that it absorbed a small fraction of the incident particles

• could be less than 100% if the electrons backscatter off the detector and deposit too small an energy to be noticed

• geometric efficiency

point source

detector sensitivearea

distance r

dΩ4π

=A4πr2

Gamma Interaction Probability

• gamma rays in a gas detector – low interaction probability (because gas has low density)

• often, in a Geiger counter, it’s the metal tube walls that are thick enough to have significant probability for gammas to interact, and then thin enough for the photoelectric effect electrons to make it from the wall into the gas

• gamma ray efficiency in a solid/liquid detector – high density increases detector efficiency; also high Z increases photoelectric effect cross section and pair production cross section

photoelectric effect electron

Gamma Ray Peak Efficiency

• how is efficiency defined for gamma ray detection?

• answer: however you want!

• you can define a peak efficiency

• you can define a total efficiency

137Cs spectrum in NaI detector

ε = # counts in peak# of gammas emitted by source

ε = # all counts in the spectrum# of gammas emitted by source

photopeakCompton edge

what’s this?137Cs has only has one gamma line?










what’s this?137Cs has only has one gamma line?










what’s this?137Cs has only has one gamma line?it’s the backscatter peak: the 137Cs gamma going the opposite direction and scattering backinto the NaI detector

Photon Interactions – Recall

Photon Interactions – Recall

Gas Detectors

• gas detectors were the first radiation detectors (that produce an electrical signal – i.e. transducers); work by direct collection of the ionization electrons (and ions)

• gas is good medium for collecting charge; here’s a generic diagram of a gas detector (gas is inside the cylinder)

• electric field (voltage) needed

• to collect the charge

Radiation Detection Process in General

Steps:1. Photon enters detector2. Photon interacts with Gas *

In this example: photoelectric

N2

* a) If by photoelectric effect ( ): photon absorbed + photoelectron

b) If by compton ( ): scattered photon (lower in energy) + photoelectron

c) If by Pair production ( ): (low probability)for Nitrogen gas

Relative probability depends on cross sections

Ie. vs vs for Nitrogen

example: photoelectric effect produces ionization

no electric field: ion-electron pairs recombinewith electric field: ions and electrons separate

Gas Detectors






N2+

N2+

N2+N2+ N2+

N2+

N2+N2+N2+


Here its photoelectric3. Photoelectron moves and interacts

via collisions. Many many N2’s are ionized. Photoelectron loses all energy



Gas Detectors






N2+

N2+

N2+N2+ N2+

N2+

N2+N2+N2+





Electrons recombine with positive ions “Recombination”

N2+

N2+

N2+

N2+ N2

+

N2+

N2+N2

+N2

+

HV~ 0 v

no E field



Gas Detectors






N2+

N2+

N2+N2+ N2+

N2+

N2+N2+N2+






N2+

N2+

N2+

N2+ N2

+

N2+

N2+N2

+N2

+EHV

~ 300 v

ICurrent Pulse

time

Area of pulse Total Charge Energy of photoelectron = energy of photon


Ionization Produced in a Gas

• for most gases, 1 ion-electron pair is produced per ~30 eV of energy lost

• why is this for most gases? doesn’t it depend on the ionization energy?

• it does depend on the gas, to some extent; but what happens is energy loss also goes to excitation (not all collisions produce ionization directly; some just produce excitation of the atoms/molecules)

• excited atoms/molecules can later produce electron-ion pairs

• thus, the total energy deposited, on average, per ion-electron pair created, exceeds the ionization potential (and exceeds the excitation potential)

• number of primary ion-electron pairs is proportional to the energy loss (the energy deposited) in the gas detector

• e.g. 300 keV energy loss in a gas detector produces 300,000/30 = 10,000 ion-electron pairs ΔN

N=

NN

= 0.01 or 1%

Energy per Electron-Ion Pair

• energy absorbed to create one electron-ion pair W

Gas W [eV]

Ar 26

He 41

H2 37

N2 35

Air 33.7

O2 31

CH4 28

Xe 22

accepted value for STP air

Electron Transport

• two things get in the way of collecting charge

• recombination: ions and electrons find each other and recombine

• electron affinity: some atoms want to capture an electron to complete their outer electron shell (e.g. O2, H2O, CO2, SF6)

• detector gas must be free of electronegative impurities

• high voltage (electric field) used to separate ions and electrons

• drift velocity and mobility

• the electric field accelerates electrons and ions; but collisions with other atoms slow them down to a “drift velocity” vD (like a terminal velocity in free fall)

• mobility μ is defined:

• electron drift velocities = to a few times 106 cm/s (for E = ~1 kV/cm)

µ = vD / E

whereas typical ion drift velocities: few times 103 cm/s

Operating Modes of Gas Detectors

• if voltage is insufficient, recombination occurs before collection

• enough voltage: “all” electron-ion pairs are collected; no effect from further increase in voltage (plateau)

• proportional mode: electrons are accelerated with higher voltage so that secondary ionizations liberate more electrons producing an avalanche or cascade; in the proportional region, the number of electron-ion pairs in the avalanche stays proportional to the number of primary electrons

• saturation when avalanches runaway (plateau)

voltage [V]

Avalanche Multiplication

• in regions of high electric field, electrons are accelerated to high drift velocity; this happens (deliberately) right next to the anode wire

• cylindrical geometry, radial electric field:

• where b radius of cylinder [cm’s]; a radius of anode wire [10’s μm]

• avalanche occurs within a few radii of the anode wire: electrons produce secondary electrons and those get accelerated possibly producing additional electron-ion pairs, and so on

• gain: as high as 106; hence output signal is larger and more easily detected

• proportionality is maintained: signal size will be proportional to total # of electron-ion pairs (which stays proportional to # of primary electrons)

Er =1r

V0ln(b / a)

liquid drop shape forms because of: lateral diffusion and electrons more mobile than positive ions

More Illustrations of Avalanche Multiplication

positive ions drift away from the anode wire after all electrons collected

!"#$%&'(%)#*+!*'$,'-'!'+.+/+%)0*',$#1+"($%',$#'#'!'#(

!"#$%&'(%)#*+!*'$,'-'!'+.+/+%)0*',$#1+"($%',$#'#'!'#(

cloud chamber photoof an avalanche


N2+

N2+

N2+

N2+ N2

+

N2+

N2+N2

+N2

+EHV

~ 300 v

ICurrent Pulse

time

Area of pulse Total Charge Energy of photoelectron = energy of photon ionization along the particle track: all of it drifts to the anode and contributes (with avalanche-multiplied gain) to the signal

Geiger-Müller Mode

• an avalanche has many secondary ion-electron pairs and also many excited molecules; these can de-excite with the emission of UV photons

• those UV photons produce photoelectrons on the cathode or in the gas; these electrons then produce their own avalanches

• there is transverse avalanche propagation when the probability of UV photon emission and new photoelectrons becomes large

• avalanches quickly engulf the entire length of the anode wire instead of taking place in one place (or over a limited extent corresponding to the charged particle track length)

• what cuts this off is that eventually enough ions are produced close to the anode wire that they halt the avalanche process “holding back” the electrons (this is called space charge effects)

• the positive ions then drift toward the cathode slowly (mobility is much lower), ending the signal after a decay time much longer than the rise time

!"#$%&'($)*+,-,*+.$$$$$$$$$$$$$$$$$$$$$$$$$*/#010(1213"($/4#$*/#0/2014$!"2&1#1516'$ .78.$$

!"!"#$%&'%()*+,,%(#-./01%(###2"#$%&'%(#3&4-56('%##$

$906:$.:;+:$<"06"=$#0(>&/=6"$?">&/40(?:$

$@&"$ "A>02"#$ 6/($ ?15">B5"($ 3=1#B>"#$ #B=046$ /$ @1C4("4#$ /D/5/4>&"$ 04$ /##02014$ 21$ 2&"$(">14#/='$014($="2B=4$21$2&"$6=1B4#$(2/2"$E'$"?022046$3&1214(F$C&0>&$35/'($/4$0?31=2/42$=15"$04$2&"$<"06"=$#0(>&/=6":$@&"$"4"=6'$1G$3&1214($0($04$2&"$D0(0E5"$1=$B52=/D015"2$="6014:$@&"$ 3&1214$ >/4$ E"$ /E(1=E"#$ "02&"=$ E'$ /412&"=$ 6/($ ?15">B5"$ 1=$ E'$ 2&"$ >/2&1#"$ C/55$2&=1B6&$ 3&121"5">2=0>$ "GG">2F$ C&0>&$ >="/2"($ /412&"=$ G=""$ "5">2=14:$ @&0($ 4"C$ "5">2=14$ 0($/>>"5"=/2"#$E'$ 2&"$"5">2=0>$ G0"5#$ 04(0#"$ 2&"$#"2">21=$/4#$ 2=066"=($/412&"=$/D/5/4>&":$@&"$>&/04$="/>2014$1G$2&"$/D/5/4>&"$5"/#($21$2&"$<"06"=$#0(>&/=6":$H4$/$<"06"=$#0(>&/=6"F$2&"$=/30#$ 3=13/6/2014$ 1G$ 2&"$ >&/04$ ="/>2014$ G1=?($ ?/4'$ /D/5/4>&"($ 2&=1B6&1B2$ 2&"$?B52035'046$="6014$2&/2$(B==1B4#($2&"$/41#":$@&"="G1="F$2&"$#0(>&/=6"$6=1C($21$"4D"513"$2&"$"420="$/41#"$C0=":$$$@&"$>14#02014$="IB0="#$G1=$2&"$>&/04$="/>2014$0($

0 7# #8#

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

Geiger Counter

• the result: signals have the same saturated amplitude whether one primary electron started the avalanche breakdown or several

• signal size depends only on the voltage applied

• no ability to measure the energy; but, large signal (volts) for even small, single particle interactions leading to efficient, reliable counting

=HV filter capacitorintegrating capacitor so you can read off dial

cathode

e

uv photon anode wire

cathode UV photons emitted by excited atoms in the original avalanche can trigger additional avalanches

in Geiger-Mueller region, electric field E near anode large enough to produce complete “sheath” of low mobility positive ions

positive space charge is created around anode thus reducing E field strength stopping the avalanche process

all pulses are approximately same size and are determined by applied voltage only (order of volts)

+HV

R

C

dead time

RC ~ 100 s

Signal rise-time 1 s

Geiger Counter Circuit

Quenching when positive ions reach cathode they capture free electrons from the cathode surface and are neutralized

until the positive ions drift far enough away from the anode wire, they prevent avalanches and the Geiger counter is “dead”

Continuous Discharge Region

• turn up the voltage even more, the gas continually discharges even without radiation

• if it is rugged enough to take it

• if you can supply required current

• if de-excited atoms/molecules emit

• you basically have a neon light

Electrical Pulse Formation

• the previous discussion gives the impression that the electrical signal formed comes from direct collection of charges (of electrons) on the anode wire

• actually, the electrical pulse is formed by induction mostly due to movement of the ions toward the cathode!

• where C is the capacitance per unit length

• potential energy U of charge q, at radius r: U = qφ(r)

• the charge moves distance dr; the change in potential energy dU is:

• energy stored in the cylinder capacitor is:

Er =1r

V0ln(b / a)

; ϕ(r) = V0 −V0ln(r / a)ln(b / a)

; C =2π ε0ln(b / a)

dU = q dϕ(r)dr

dr12lCVo

2; dU = lCV0dV = q dϕ(r)dr

dr

dV =q

lCV0−V0r

1ln(b / a)

⎛⎝⎜

⎞⎠⎟dr = −

q2π ε0l

drr

- electrostatic configuration charged to some voltage- charge is injected and moves around; this changes the electrostatic configuration and voltage

Electrical Pulse Formation cont’d

• consider ionizing event with avalanche multiplication at a distance r′ from the anode (avalanche dominates over the single, primary electron drift history)

• induced voltage from the electrons is:

• Q is the total charge of all electrons/ions

• induced voltage from the ions is:

• compare relative:

• if the avalanche occurs with r′ only a few wire radii away, then a+r′ is roughly equal to a, within a factor of a few; but, b could be orders of magnitude larger than wire radius a

• e.g. 1 cm radius proportional counter versus 10 μm radius wire

• we see the induced electrical signal is due almost entirely to the motion of the ions!

V − = −(−Q)2π ε0l

drr= −

Q2π ε0l

ln a + ′raa+ ′r

a

∫

V + = −Q

2π ε0ldrr= −

Q2π ε0l

ln ba + ′ra+ ′r

b

∫V −

V + =ln[(a + ′r ) / a]ln[b / (a + ′r )]

More on Pulses from a Proportional Counter

• the total induced voltage is:

• the time development of the pulse thus traces the motion of the ions:

• drift velocity of the positive ions:

• don’t need to write out V(t); just note it gets more negative logarithmically with r(t), with the time constant set by V0, the mobility μ, capacitance (geometry)

V = V − +V + = −Q

2π ε0l[ln(a + ′r

a) + ln( b

a + ′r)] = −

Q2π ε0l

ln ba=−QlC

V + (t) = −Q

2π ε0ldrr= −

Q2π ε0l

ln r(t)a + ′ra+ ′r

r(t )

∫drdt

= µE(r) = µCV02π ε0

1r

r dr = µCV02π ε0

dt

12r2 (t) = µCV0

2π ε0; r(t) = µCV0

π ε0t

⎛⎝⎜

⎞⎠⎟

1/2

Final Story on the Pulse from Proportional Counter

• gets differentiated by the RC high-pass filter

Choice of Gas and Quenching

• want low working voltage to achieve high gain, good proportionality, high rate (fast ion drift velocity); noble gases like Ar, Xe are good...but they have high excitation energies

• excited Ar atoms de-excite with UV photons

• what prevents these UV photons from causing unwanted avalanche breakdown, like in a Geiger counter?

• quench gas is added (molecules like methane CH4, isobutane) that absorb UV photons; a common mixture is 90% Ar and 10% methane (called P10 gas)

• the quench gas molecules also de-excite the main gas molecules (excitations transfer non-radiatively to the quench gas molecule)

• proportional mode, high gain achievable when quench gas is added

• introduces other problems, though, like carbon build up in the counter

Sealed Gas Counters versus Flowing

• sealed counters are common (e.g. Geiger counter)

• if you want good control of operating behaviour of the proportional counter, a sealed counter may leak out, may leach impurities out of the counter walls, air can leak in and bring in oxygen; electronegative impurities, if high enough, prevent charges from drifting

• solution: continuous flowing, fresh gas

Parallel-Plate Geometry?

• works, and used in ionization chambers

• can you think of an example?

• but not used in proportional mode...why not?

• E = V/d and is constant; for usable d, V needs to be huge to get gain

• and if the field is high enough to cause an avalanche from secondary ion-electron pairs, it is high enough throughout the whole chamber

• electrons far from the anode avalanche more than those close to the anode

• introduces position-dependent response...but the whole point of proportional mode is for the signal to be proportional to the energy in a simple way

Parallel-Plate Geometry?

• works, and used in ionization chambers

• can you think of an example?

• but not used in proportional mode...why not?

• E = V/d and is constant; for usable d, V needs to be huge to get gain

• and if the field is high enough to cause an avalanche from secondary ion-electron pairs, it is high enough throughout the whole chamber

• electrons far from the anode avalanche more than those close to the anode

• introduces position-dependent response...but the whole point of proportional mode is for the signal to be proportional to the energy in a simple way

smoke alarms

Multiwire Proportional Chamber

• used in particle physics experiments

• Georges Charpak, 1992 Nobel Prize in Physics

• note to Dan Brown: does not play frisbee!

field lines and equipotential lines in a MWPC

2D Planar Readout MWPC

• stack MWPC layers; allows planar position sensitivity to (radiation-induced) particle tracks

lect 18 radiation detectors - Engineering Physicsphys352/lect18.pdf · Radiation Detectors I: Gas Detectors Prelude: General Properties •radiation/particles interacting in materials

Documents