Gamma Rays

OBJECT:
To understand the various interactions of gamma rays with matter.
To calibrate a gamma ray scintillation spectrometer, using gamma rays
of known energy, and use it to measure the energy of an unknown
gamma ray.
To use positron annihilation radiation to determine the mass of the electron
and to observe correlated gamma rays.
READINGS: The lab manual Experiments in Nuclear Science AN34, 1984
EG&G ORTEC provides an excellent hands-on discussion of the background and
techniques for a number of undergraduate level nuclear experiments. The
equipment described resembles, with some variation, the equipment available in
the laboratory. Additional readings are given at the end of this write-up.
APPARATUS: NaI:Tl scintillator and photomultiplier tube detector with
integrated preamplifier (2), high voltage power supply, Canberra model 2000
power supply, NIM bin, Canberra 2015A amplifier/single channel analyzer
module (2), Rutgers P1075 scaler/timer, Canberra model 1446 coincidence
module, multichannel analyzer PC board in computer, Maestro Analyzer
software, monitor.
BACKGROUND: In this experiment you will study the radioactive decay of a
nucleus by detecting gamma rays emitted consequent to the decay. Gamma ray
detection is a slightly complicated, multi-step process: the gamma ray enters a
NaI:Tl scintillator crystal where it produces a rapidly moving free electron that,
in turn, loses its energy by excitation of the ions in its path as it travels through
the crystal. This excitation energy is given off in various ways, one them being
emission of visible light (fluorescence). Thus a single high energy gamma ray
entering the scintillator produces a flash of low energy photons.
These photons are directed to the photosensitive surface of a photomultiplier
tube, where they eject electrons via the photoelectric effect. The electrons are
collected in the photomultiplier and amplified to yield a current pulse, which is
converted to a voltage pulse whose height is proportional to the number of
photoelectrons and is thus proportional to the number of photons reaching the
tube, which in turn is proportional to the initial energy of the fast electron.
When a radioactive source is placed near the scintillator, the photomultiplier
produces a series of pulses, each corresponding to the decay of a single nucleus.
The amplitude of each pulse is related to the energy of the electron freed by the
gamma ray. These pulses are studied using either a single- or multi-channel

analyzer. A single channel analyzer (SCA) counts on the number of voltage

pulses whose height falls within a given (adjustable) window of values, while a
multi-channel analyzer (MCA) sorts the pulses according to height and the
counts the number in each window to give a spectral (energy) distribution of the
fast electrons. Figure 1 shows a typical MCA spectrum. In order to relate this
spectrum to the nuclear decay, we need to understand how gamma rays interact
with matter.

Figure 1. Na(Tl) Spectrum for Cs137

Interaction mechanisms for gamma rays with matter: When entering a crystal,
gamma rays produce fast charged electrons by three different processes: the
photoelectric effect, the Compton effect (Compton scattering) and pair
production. It is these fast electrons, which give rise to scintillations, not the
gamma ray. The observed spectral distribution will thus depend on the detailed
interaction process of the gamma rays in the crystal.
Consider a beam of mono-energetic gamma rays striking the scintillator. For our
purposes the most important energy loss mechanism is the photoelectric effect.
When a gamma ray strikes an ion in the crystal, it is absorbed and all of its
energy is transferred to one of the bound electrons, which is freed and moves
rapidly through the crystal. Since the energy of the gamma ray (typically about
0.5 MeV) is much greater than the binding energy of the electron of the ion
(typically 10 to 100 eV), the energy of the freed electron can be considered equal
to that of the incoming gamma. (Especially since the energy resolution of the
detector is only about 10%.) Thus the photoelectric effect results in a peak, called
the photopeak, in the photomultiplier spectrum at an energy equal to that of the

gamma

incoming gamma ray.

In Compton scattering, the gamma ray is not absorbed, but rather scattered
through an angle by an electron, which recoils and carries away some of the
gamma rays energy E. (The scattered gamma ray then escapes from the
scintillator; the probability that a gamma ray Compton scatters in a typical size
scintillator is quite small (1% to 10%), which means you are unlikely to detect a
gamma ray that has undergone two Compton scatterings.) The gamma rays
initial wavelength is = hc/E = 1240/E nm, where E is in eV. The change in
wavelength is:
=h/mc (1 - cos ) = 0.00243 (1 - cos ) nm,

(1)

where h is Plancks constant, m is the mass of the electron and c is the speed of
light. From this equation you can see that the energy of the scattered electron,
which is the energy loss of the gamma ray, will vary from zero (when = 00) to a
maximum corresponding to a wavelength shift of 0.00486 nm (when = 1800).
This maximum energy is called the Compton edge. The energy distribution of
Compton scattered electrons is essentially a constant. So the Compton spectrum
produced by a photomultiplier tube is an almost flat plateau from zero energy
up to the Compton edge where it drops off sharply (at a rate limited by the
energy resolution of the tube).
Consider, for example, a 622 keV gamma ray from a Cs137 decay. Its initial
wavelength is 0.00199 nm. If it is Compton scattered through = 1800 the
wavelength becomes 0.00199 + 0.00486 = 0.00685 nm corresponding to an energy
of 181 keV. The lost energy, 622 - 181 = 441 keV, is transferred to the electron
from which it scattered and is observed as a Compton edge at 441keV. For
smaller scattering angles, the transferred energy will be less, extending down to
zero for = 00.
The discussion above refers to gamma rays that are Compton scattered by
electrons within the scintillator. It is also possible for a gamma ray to be
Compton scattered into the scintillator from an interaction outside the
scintillator. In this case the observed signal is from the scattered gamma and not
from the recoiling electron. The scattered gamma ray could then be detected
through the photoelectric effect. For Cs137 the Compton scattered gamma rays
will have energies ranging from 181 keV up to the full 622 keV. However,
because of the geometry of the detector, most of the gamma rays scattered into
the scintillator will have been scattered through a large value of . But cos
varies only slowly with for near 1800, which means [see Eq. (1)] that these
gamma rays will all have energies near 181 keV. The resulting energy peak is
called the backscatter peak. It can be enhanced by placing a sheet of lead around

gamma

the outside of the scintillator.

The third interaction mechanism is pair production. If the incoming gamma ray
energy is above 1.02 MeV = 2mc2, the rest mass of an electron-positron pair, the
gamma ray can spontaneously create an electron-positron pair and be totally
absorbed. If both the electron and positron lose all of their kinetic energy while
still in the scintillator, they would produce a photomultiplier pulse
corresponding to an energy 2mc2 below the gamma ray energy E. (Of course,
either might escape the crystal after partial loss of kinetic energy.)
But the spectrum is actually more complicated since if the positron has been
slowed down and stopped in the crystal, it will annihilate with an electron,
emitted two gamma rays, each of energy mc2. One, or both, of these gamma rays
may be absorbed in the crystal, and thus contribute to the height of the
photomultiplier output pulse. Pair production therefore produces a full
energy peak (E), a one-escape peak (E-mc2) and two escape peak (E-2mc2),
depending upon whether both annihilation photons are absorbed in the
scintillator or one, or both, escapes.
The final question to consider is that of the relative importance of the three
interaction mechanisms, which depend in different ways upon the energy of the
gamma ray. For low energy rays, the photoelectric effect predominates. Since
the photopeak directly yields the energy of the gamma ray, most scintillators are
designed to maximize the photopeak. In the NaI:Tl scintillator you will use, a
small amount of the heavy metal thallium is added for this purpose when the
crystal is grown. (The strength of the photoelectric effect depends strongly on the
number of electrons bound to the ion.) As E increases, the photoelectric
absorption deceases rapidly, while the Compton scattering decreases much more
slowly and predominates above several hundred keV. The absorption coefficient
for pair production rises rapidly above the threshold E = 1.02 MeV and exceeds
the Compton scattering, while photoelectric absorption becomes negligible.
Energy resolution: A NaI:Tl detector has an energy resolution of only about
10%. [More advanced detectors such as high purity germanium detectors have as
much as 30 times greater resolution, but are much more expensive.] When a
beam of mono-energetic gamma rays strikes the scintillator, there is a fluctuation
from gamma ray to gamma ray in the height of the voltage pulse from the
photomultiplier, which shows up as a broadening of the photopeak. The pulse
height variation is chiefly due to statistical fluctuations in the number of
electrons emitted at the cathode of the photomultiplier when a flash of photons
arrives from the scintillator, but is also due to the occasional escape of electrons,
X-rays or gamma rays from the crystal, all of which depend on how large the NaI
crystal is. The fractional full width of a peak at half its maximum height

gamma

(FWHM) is a convenient measure of the resolution of the instrument.

Nuclear decay: You will study the decay of several light nuclei that proceed by
beta decay -- the emission of an electron or positron (and the associated
neutrino). These decays, Figure 2, leave the daughter nucleus in an excited state
that then returns to the ground state by the emission of one or more gamma rays.
The range of the beta particles is so short that they cannot penetrate into the

Figure 2. Principal Decay Modes of Cesium 137, Cobalt 60, and Sodium 22.
scintillator. So you will only be detecting the gamma rays. Table 1 gives the
energies of the gamma rays emitted by the daughter nuclei for the three nuclei
you will be studying. The Ba k X-ray occurs about 4% of the time when instead
of emitting a gamma ray, the Ba nucleus de-excites by transferring its energy to
one of the innermost electrons (K shell) that overlaps the nucleus. This gives the
electron enough energy to escape from the atom. One of the outer electrons then
drops into the hole left by the escaping electron and emits its excess energy as an
X-ray. Although the escaping electron cannot penetrate into the scintillator, the k
X-ray does and is efficiently detected.

gamma

Parent Lifetime Decay Daughter Gamma energy

nucleus (years) mode nucleus
(MeV)
22
+
22
Na
2.605
e
Ne
1.2746
60
60
Co
5.272
e
Ni
1.1732
1.3325
137
137
Cs
30.17
e
Ba
0.6616
0.0322*
*Ba k X-ray
Table 1. Nuclear Decay Data
(from: Handbook of Chemistry and Physics)

Angular Correlations: When the nuclei in a radioactive sample decay the

individual decays are unrelated and the emitted particles (gamma rays) are
emitted isotropically in all directions. We will study one case where the gamma
rays from a sample are correlated -- the annihilation radiation of positrons.
When a Na22 nucleus emits a positron, the positron rapidly slows and then
annihilates, upon encountering an electron. Conservation of momentum
requires that the two gamma rays that are emitted travel in exactly opposite
directions. You will use two scintillation counters and a coincidence module to
study this angular correlation.

Figure 3: Block Diagram for Scintillation Counter Apparatus

gamma

PROCEDURE: Warning. The experiment involves high voltages; do not turn on any
switches or knobs until you have become familiar with the apparatus. The high voltage
applied to the phototube should never exceed 1200 volts. Observe caution at all times
with the scintillation detector since the crystal and photomultiplier are fragile and
expensive.
The apparatus is for the most part pre-wired and very simple to operate. The
block diagram is shown in Fig. 3. The scintillation counter consists of a NaI:Tl
crystal glued to a photomultiplier (PM) tube. The PM tube is powered by an
adjustable high voltage supply (~1000 V). The supply is a vacuum tube model
and should be set to standby for about 1 minute before turning to on. There
is normally no need to adjust the voltage level. More information about this
detector is available in the supplementary material.
The voltage output pulse of the detector is fed into a preamplifier attached to the
PM tube, then to an amplifier in a NIM bin. [A Nuclear Instrument Module
(NIM) bin is a rack with a number of slots for various instruments (amplifier,
scaler, coincidence counter, etc) and a DC power supply for the modules. It
allows great ease and flexibility in setting up an experiment.] After amplification,
the signal is divided and fed into two separate pulse height analyzers, which
measure the gamma ray spectrum in different ways.
Amplifier:
The CANBERRA Model 2015A combines, in one single width
module, a spectroscopy amplifier with gated baseline restoration and a timing
single channel (SCA) pulse height analyzer. You will first use this unit as part of
a multi-channel analyzer (MCA) and then as part of a single channel analyzer
(SCA) and a coincidence circuit.
MCA function. In this part of the experiment the AMP output of the amplifier is
fed into the MCA plug-in card and associated software on the computer. The
MCA software sorts each PM pulse as it arrives from the amplifier according to
pulse height and generates a fine-grained histogram of the number of voltage
pulses from the detector versus pulse amplitude and, therefore, energy deposited
in the scintillator. The amplifier output is proportional to the energy of the
incoming signal with a maximum of 10V. The proportionality factor is controlled
by the GAIN.
Note: If the gain is set too high the output corresponding to a signal of interest may be
outside of the 10V range and the output will be saturated. The calibration procedure
described below should prevent this.

SCA function. The SCA counts the number of events within a predetermined
window of energies. In this mode the amplifier output is fed internally to the
SCA, which has two knobs to adjust the range of pulse heights the SCA will pass.

gamma

The first knob adjusts the lower level E and the second pulse the window width
E, so that pulses with heights between E and E+E will be passed to the
output. The correspondence between energy and scale on the knobs is
determined by the Gain, which you will set during the calibration. The SCA
output is attached to the scaler input.
Scaler. The scaler counts the number of pulses arriving during a set time interval.
To obtain the same data with the SCA as you would get with the MCA, you will
need to step E across the range of possible voltages keeping E constant and
counting for a fixed length of time.
1. Calibration:
a. Check that the apparatus is connected according to the diagram in Fig. 3 for
detector 1. Turn on the NIM bin power supply and switch the high voltage
PM supply to standby. Turn on the computer. Log in as Student using the
course password. The MCA program is called MAESTRO. Maximize the
screen.
b. Place the Na22 source on the Aluminum rod in front of detector 1. Turn the
PM power supply from standby to on and make sure that the setting is
about 1000 V (this would have been set by previous groups). Click on the
Go button (in the top bar) to start accumulating the spectrum. It will take
about five minutes to accumulate enough counts to resolve the spectrum.
Note that the 1.2746 MeV photopeak is the highest energy feature in the
spectrum, but not the strongest. Use the mouse to move the cursor onto the
photopeak to identify its channel number (called Marker on the status bar
at the bottom). Record the channel for the maximum of the photopeak. If it is
in bin 900 skip the next step.
c. Adjust the amplifier gain. Use the mouse to bring the cursor over the
photopeak. The channel number and the number of counts will be listed at
the bottom left of the screen. If the peak is not on channel 900 adjust the gain
accordingly and retake the spectrum. If necessary repeat a few times until the
peak region falls in channel 900 (plus/minus ~3). Once you have set this
gain do not change it; otherwise you will change the calibration.
d. To easily identify the energies of other features in the spectrum, click on the
photopeak and go to Calculate Calibration and enter the energy of the
photopeak in keV. Now, you can click on any other part of the spectrum and
see the energy on the bottom of the screen.
e. Saving the data. Go to File Save As and save your data in the ASCII
SPE format. If you open this file in notepad, you will see a list of the counts in
each bin. You can copy this into a data analysis program such as Origin or
Excel.
f. Now take the spectrum of the Cs137 source. Note the channel number of the
0.6616 MeV photopeak. Is it consistent with your calibration?.
g. Remove all sources to distant storage, and measure the background
spectrum with the MCA.

gamma

2. MCA Decay Spectra. Without changing the calibrated gain setting record the
spectra of the other two isotopes Cs137, Co60 and Na22with the MCA. When taking
data, be sure that you have removed the other sources far away from the scintillator!
When saving the data make sure that you include the collection time (you will
need this later). For your report you will need to determine the (minimum)
activity of each source in curies. To do this you have to measure (for each
source) the appropriate distances to calculate the solid angle subtended by the
scintillator relative to the source.
3. SCA Decay Spectra. Now measure the gamma spectrum of the cesium source
using the SCA. Set the SCA window width at 0.1 V. [Note that there is a switch
on the back of the module that allows you changes full-scale forE from 1 V to
10V.] Leave this setting fixed and step E from zero to a value beyond the Ba137
photopeak, recording at each step the scaler readings for a fixed count time (e.g.
100 s). In the region of the photopeak, where the counting rate changes rapidly
with voltage, use 0.1 V steps in E. In the more slowly varying regions steps of 0.3
V should be adequate. Be sure to record all instrument settings, -- count time,
PM voltage, etc. Plot the pulse height distribution spectrum [counting rate
versus E+(E/2)].
4. correlation: You will now study the correlation of the gamma rays emitted
when the positron from the Na22 source annihilate. You will need two
scintillation counters, two single channel analyzers (SCA), a coincidence module,
and a scaler to observe this correlation. The first step will be to set the windows
of the two SCAs to pass PM pulse heights corresponding to the 511 keV
annihilation gamma rays. When a pulse from the PM tube meets the window
requirements, the SCA outputs a narrow pulse, which is fed into the coincidence
module. The coincidence module outputs a pulse whenever it detects pulses
from both SCAs at the same time. The scaler counts these coincidence pulses for
a fixed length of time. The two scintillators are mounted on arms that pivot
about the location of the Na22 sample. When the two arms are 1800 apart, strong
coincidences will be recorded, but when the angle is decreased the count rate will
drop sharply. (There will still be a number of accidental coincidences for nuclei
that just happen to decay at the same time.)
4A. The first step will be to set up the window for counter 1. With the sodium
source in place, observe the amplifier output with the MCA and with the
oscilloscope (channel 1). [On the MCA spectrum of Na22 the 511 keV gamma ray
is the strongest feature. Note that the positron annihilation occurs in the sodium
source not in the detectors.] Set the scope sweep speed at 5s/cm and the gain
to 1 V/cm with the trigger source channel 1 and the trigger level about 2 V. You
will see a series of superimposed pulses of widely varying height, but with many

gamma

falling in the same band, which results in the scope traces being brightest for that
band. These pulses correspond to the 511 keV gamma rays. Adjust the amplifier
gain so that this band corresponds to a pulse height of about 3 V.
Now connect the SCA output to channel 2, change the trigger source to channel
2, and set the scope to display both channels 1 and 2. [Note that the SCA pulses
displayed in channel 2 are extremely narrow (~ 0.7 s.) which places stringent
limits on whether the coincidence module will consider two gamma rays to be in
coincidence.] Observe how the channel 1 scope trace changes as you change E,
the lower edge of the window, and E, the window width. Set E to about 2.8 V
and E to about 0.4V. [Note that there is a switch on the back of the module that
allows you changes full-scale forE from 1 V to 10V.] These settings will limit
the pulses accepted by the SCA to just those corresponding to the 511 keV peak.
4B. Repeat step A for the second counter. Be sure to shift the cable leading to the
MCA from the first amplifier to the second one.
4C. Now connect the outputs of the two SCAs to channels 1 and 2 of the
coincidence module, set the corresponding switches to coincidence and
connect the output of the coincidence module to the scaler input. Set the scaler to
count for 10 or 100 s. With the two counters set 1800 apart, you should obtain a
strong coincidence count rate, which should drop almost to zero when the angle
is reduced.
4D. Record the number of coincidences (for a fixed counting interval) as a
function of the angle of separation of the detectors. Make a plot for your report.
Explain the width of this curve, i.e. why do you obtain coincidences even when
the angle is not exactly 1800?
4E. For your report you will need to measure the appropriate distances to
calculate the solid angle subtended by both scintillators relative to the source.
5. Co60 correlation (optional): Figure 2 shows that when Co60 decays its emits
two gamma rays. The second gamma ray rapidly follows the first (the excited
state has a mean life of 0.7 x 10-12s) and thus appears to be coincident. Its
direction of emission is correlated with the direction in which the first is emitted.
We will not go into the theory, which is complicated, but the prediction is that
the angular correlation is given by
w( ) 1

cos2 cos4

,
8
24

(2)

gamma

10

where is the angle between the two gamma rays. This is a rather small effect.
The correlation function is 1.00 at 90o and 1.17 at 180o. Thus this a much more
challenging effect to experimentally detect; to detect a 17% effect, you will need
to measure the correlation rate to about 1%. The error in a measurement of N
counts is sqrt(N). Thus to achieve 1% accuracy at a particular value of you will
need to count until you have accumulated about 104 counts. For the source
available in the lab this means you will need to count for 1000 s for each point.
5A. Mount the Co60 source at the center of the spectrometer and move the two
detectors fairly close to the source (to get a large count rate).
5B. Set the windows for the two detectors to pass both of the gamma rays (1.1732
and 1.3325 MeV).
5C. Record the number of coincidences C() as you vary in increments of 10o
between 180o and 90o. Count for 1000 s for each point. Plot the ratio C()/C(90o)
and compare with the theoretical correlation, Eq. (2).
ANALYSIS
Save each spectrum as an in a file and use your preferred software to analyze it.
For each peak subtract the background count and record the net peak count,
counting time, peak energy, FWHM and FWTM (full width at one-tenth of
maximum) widths. On each plot label the photopeak(s) the Compton regions
and all other features of interest. Be sure to study the k X-ray for the cesium
source. Calculate the activity of each source in Curies. Make a plot of the
REPORT:
The report should contain the calibrated spectra for the sodium, cobalt, and
cesium sources with labels identifying each spectral feature. It should contain a
plot of the number of coincidences (for a fixed counting interval) as a function of
the angle of separation of the detectors
1. From your data for cesium calculate the minimum activity in curies. Discuss
why this can only be an estimate of the minimum activity. Make a spectral plot
(counts versus E+E/2) of your SCA data for cesium and compare with the MCA
spectrum.
2. For the cobalt spectrum determine the energies of the two gamma rays and
compare with the accepted value. Discuss how the overlap of the tails affects the
energies you measure. Calculate the theoretical Compton edges for the two
gamma rays and indicate where they fall on your measured spectrum.

gamma

11

3. Resolution: From your MCA spectra calculate the resolution (FWHM) of the
counter for the various photopeaks. Present your results in a table; is the
resolution dependent on the energy of the gamma ray?
4. Discuss all the spectral features you observe on the sodium spectrum. From
the calibrated spectrum determine the rest mass of the electron with an
estimate of your experimental error.
5. Make a plot of the correlation count rate versus angle between counters.
Explain the width of the curve.
6. In the discussion session include answers to the following questions:
a. Why does the photoelectric peak not have zero width?
137

b. Why are there x-rays emitted after the " Cs " decays? To what atomic
transition and to what atom do these x-rays correspond?

gamma

12

REFERENCES:
experimental details
1. AN34: Experiments in Nuclear Science, 3rd Ed., EG & G Ortec. Experiment 1,
Basic identifications in Electronic Measurement Systems, pp 1-7; Experiment 3,
Gamma Spectroscopy Using NaI, pp 15-24; Linear and Logic Signals in EG & G
Ortec NIM Instruments, pp159-160
2. D.W. Preston & E.R. Dietz: The Art of Experimental Physics, Appendix B
Counting and Sorting Particles: The Scintillation Counter, pp376-385
general theory
3. Serway, Moses and Moyer: Modern Physics, pp 431-432
(scintillator/photomultiplier), 389-392 (beta and gamma decay)
4. Beiser: Concepts of Modern Physics, pp 443, 471-473 (gamma decay)
5. Halliday: Introductory Nuclear Physics
6. Siegbahn: Alpha, Beta and Gamma Ray Spectroscopy

JP 2/2000

gamma

13