Gamma Ray Spectroscopy

Karabo Letsholo : LTSKAR009

February 2024

Abstract
The energy calibration of a NaI(Ti) detector subjected to set electronic configurations based on four known
radioactive samples (60Co, 137Cs, 22Na, and 133Ba) was performed using known energy data from the Nudat
library and counting data collected on USX via Spectech’s multichannel analyzer (MCA). An unknown sample was
then identified on the basis of the energy of its two full-energy photopeaks. The identification of the features of a
typical radioisotope spectrum was carried out.

1 Introduction
1.1 Radioactive Decay as Origins of Radiation
Radioisotopes are isotopes with nuclide that is unstable and undergoes radioactive decay [2]. There are different modes
of decay, but one of the most common is decay by beta emission (which results in the release of a beta particle), of
which beta minus / fast electron emission is favorable for neutron-rich radionuclides and beta plus / particle emission is
favorable to neutron-deficient radionuclides [2]. Beta-minus emitters decay by converting neutrons to protons, thereby
releasing a β − /fast-electron particulate in the process, most of the time decaying to an excited state of the daughter
nucleus which is then followed by the release of a gamma ray while the daughter nucleus de-excites to the ground
state [2]. For beta-plus emitters, a nuclide decays followed by the release of a beta-plus/positron particulate [2]. Beta-
plus emitters have a neutron deficiency and hence decay either by one of electron capture/EC (where an extranuclear
electron is captured to convert the proton into a neutron) or via nuclear de-excitations when a nucleus has a sufficiently
large mass/energy difference exceeding the critical value of 1022.00 keV [2]. Regardless of which mode is used, the
nucleus that is normally left in an excited state de-excites by the release of a characteristic gamma ray of a specified
energy that will interact with the mechanisms of the detector to produce a signal, which with the help of particular
electronics and counting statistics will be collected to help calibrate our detector by energy.

1.2 Scintillation, Detection, Dead time and Electronics

Of the various mechanisms radiation can interact with the scintillation material of the detector, the most common are
Photo-electric Absorption (PE), Compton Scattering (CS), and Pair Production (PS) [2] [1]. Each produces a defining
characteristic of the gamma-ray spectrum collected by counting channels. Photoelectric absorption occurs when an
incoming gamma ray is absorbed by an electron, most commonly the bound electron of an atom, followed by the release
of an X-ray observed as blue light from the detector or scintillation material [2] [4]. This blue light is then used to
produce electrons on the cathode of a photomultiplier tube. The electrons then multiply exponentially by the relation
(overall gain = αδ N ), where δ is the ratio of the number of secondary electrons emitted to the number of primary
electrons and α is the fraction of photoelectrons collected in the photomultiplier tube [1]. These electrons produce a
signal whose decay time is proportional to the detector decay time [1]. The preamplifier uses a capacitor to convert the
electrical charge into voltage, the ORTEC pre-amplifier can input capacitance from the pre-set value of 45.00pF [3].
When analyzing the decaying signals from the detector-preamplifier set up via the oscilloscope, it becomes apparent
that they possess a fast leading ledge, which is dictated by the decay time of the signal carriers (free electrons) of the
detector and the preamplifier RC network; a decay time constant of this signal can be identified and will be helpful
to determine the dead time [4]. To ensure that the detector does not read/detect multiple events simultaneously, we
introduce a separation time that separates two events and is called the dead-time [1]. The decay time of a typical
NaI(Ti) detector is around 0.23 µs [1]. Ideally, the dead time should be around 4 times the pre-amplifier decay time
constant [4].

1
2 Experimentation
Each NaI(Ti) detector had a voltage reading on it. This indicated the maximum voltage that can be used with the
detector; going beyond this value may damage the detector. The detector used for this experiment had a rating of
+600.00V. First, we had to ensure that the high-voltage (HV) source on the Nuclear Instrumentation Module (NIM)
crate had a positive voltage lead on because the detector requires positive voltage. Next, we turned on the NIM-crate,
the HV-source (we had to turn the voltage slowly to avoid damaging the equipment, a warm-up time of 30 minutes
is recommended by the manufacturer to avoid fluctuations in voltage [3], we then set it to +600.00V), and amplifier.
Next, after connecting the HV source to the detector via a BNC cable and connecting the detector’s anode to the
preamplifier’s input whose output is connected to the oscilloscope’s input, the preamplifier’s output was connected
to the amplifier’s unipolar input, and the amplifier’s output is connected to the Multi-Channel Analyzer’s (MCA)
input. We then set the pre-amplifier’s input capacitance to 1000.00 pF and used the oscilloscope to determine our
pre-amplifier’s decay time constant to be 0.25 µs, the shaping time is then 4 × 0.25 = 1 µs. After turning on the
amplifier, we set the shaping time of the amplifier itself at 2 µ s > 1 µ s. We ensured that the mode for USX was
PHA direct in and connected the amplifier to Spectech’s multichannel analyzer, which was connected to a computer
with a Pulse-Height-Analysis (PCA) software USX. We then set the course gain to 200, and began with our first
sample 60Co. We identified spectral features and the two full-energy peaks of 60Co, then used the Nudat database
[5] to determine which peak corresponds to which energy, by assuming an approximately linear relationship between
the energy and channel number, the lower-energy gamma-ray energy corresponded to the low channel number (whose
value and uncertainty were determined using USX as the centroid and FWHM) and similarly for the high-energy
gamma-ray energy. An approximately linear fit with these two data points was made as a rough calibration. We then
took measurements of all the remaining samples (137Cs, 133Ba, 22Na, Unknown Source) and performed analysis of
their spectra using python.

3 Spectral Features
This section is focused on the identification of different features found on one of our samples, mainly the 22Na isotope.
Below is a graph of the spectrum and the identification of different features, as well as a Gaussian fit to the full energy
peak. From figure(1), we can identify the full-energy-peak (corresponding to a peak with a high channel number as a

Figure 1: [1] The spectrum of the Sodium-22 source obtained form the experiment via the USX software with its
channels versus counts data plotted on python.

linear relationship is assumed), situated around channel 518.00 and the annihilation peak (corresponding to the peak
with a low channel number), situated around channel 200.00. The lead (Pb) X-rays may be identified in the lower
channels from 0-40.00 and the backscatter peak is located around channel 90.00. The annihilation peak rests on the
Compton continuum of the full-energy peak, and this Compton continuum is also identified. A zoom of the Gaussian
fit on the full energy peak is illustrated.

2
4 Results
The results of the channel peak centroids, peak uncertainties, energy values and energy uncertainties for each radioiso-
tope are tabulated below in table(1).

Radioisotope Eγ [keV] Ch uCh Radioisotope Eγ [keV] Ch uCh

137Cs 661.657(3) 272.039 10.594 60Co 1173.288(3) 488.250 21.059
22Na 1247.537(7) 518.992 15.951 60Co 1332.492(4) 557.123 16.930
22Na 511.000 208.427 9.939 133Ba 80.998(11) 63.411 15.994
133Ba 356.013(7) 146.593 10.809 133Ba 302.851(5) 123.416 13.046
Unknown1 652.862 +
− 10.400 270.726 10.563 Unknown 1107.770 +
− 22.606 454.939 15.506

Table 1: Table 1 shows the centroid, variance, and energy data acquired from the curve fit procedure from python,
Nudat3, and IAEA databases. The energy values of the unknown sample were computed from the slope and intercept
values of the fit.

In table(1), Ch represents the channel numbers of the centroids of the peaks and Eγ represents the gamma ray
energies with their uncertainties given in the brackets to the n-th decimal place. uCh represents the standard deviation
of the peak centroids. Note that Ch and uCh were acquired from python using guesses from USX. Table(1) includes
an estimate of channel peak centroids and standard deviations from the fitting of the Gaussian curves via Scipy’s
optimization library using curvef it. The Spectech MCA analyzed pulses with voltages in the range [−5V, 5V ] [6], and
using USX to set the Region-Of-Interest(ROI), the curve fit was carried out on each feature resembling a photo-peak.
The unknown sample’s gamma ray energy estimates are given from the parameters (slope and intercept) of the linear
fitting procedure.

5 Energy Calibration
The assumed approximately linear fit was carried out using the channel data from table(1) as the dependent variable and
the gamma ray energies from table(1) as the independent variable. The x-error (xerr) and y-error (yerr) parameters
in the errorbar plot were set to the Nudat[ [5]] energy uncertainties and the standard deviations from each peaks’
Gaussian fit respectively. From that, the best approximation for the parameters were acquired.
−5
A = Slope = 0.405+
− 8.812 × 10 keV−1
B = Intercept = 6.351+
− 0.053

With the energy calibration curve is given by the approximate relation, Ch = B + A · Eγ , by substituting in the best
approximations yielded the following curve for our particular setup,
−5
Ch = 6.351+ +



− 0.053 + 0.405− 8.812 × 10 · Eγ (1)

The graph of the calibration curve is plotted below for our sample peaks from table(1) and Nudat([5]).

Figure 2: [2] The calibration curve eq(1), ignoring uncertainties, of the setup using curve fit and data from table(1).

3
6 Unknown Sample
By utilizing the curve in equation(1), and using curve f it on the ROI around the unknown spectrum’s full-energy
peaks, we found values included in table(1). The gamma ray energy interval for 2 of the unknown sample’s full-energy
peaks are determined and are listed in table(1) using the formula,

(Ch − B) Ch − 6.351
Eγ = =
A 0.406
σCh − B uCh − 6.351
∆Eγ = =
A 0.406
The 2 energy peaks found in the spectrum of the unknown sample are are given in table(1) and the unknown sample’s
counts vs channels spectrum is also given below. From the energy values, using the Nudat([5]) and Bing’s AI, the

Figure 3: [3]The Counts versus channels spectrum plot for the unknown sample with Gaussian fits around the full-
energy peaks.

radioisotope with the energies in the range identified is 207-Bismuth. 207 207
83 Bi decays to 82 Pb by electron capture and
releases 2 primary gamma rays [5]. The first gamma ray has an energy of 569.698(2) keV, an intensity of 97.750(3)%,
and a dose of 0.557(17) MeV/Bq·s [5]. The second gamma ray has an energy of 1063.656(3) keV, an intensity of
74.500(3)%, and a dose of 0.792(3) MeV/Bq·s [5]. With the above information, I conclude that 207-Bismuth is mostly
likely the unknown sample.

7 Conclusions
In this experiment, the NaI(Ti) detector had its energy calibrated for the particular setup in this experiment. An
approximate linear relationship was attained by perfoming a weighted fit using python. The spectrum features of a
typical radioisotope where identified and the unknown sample was identified to likely be 207-Bismuth.

References
[1] Glen F Knoll. “Radiation detection and measurement, NY 2000”. In: John Wiley&Sons ().
[2] Gordon Gilmore, JD Hemingway, and JL Durell. “Practical gamma-ray spectroscopy”. In: Journal of Physics
G-Nuclear and Particle Physics 22.7 (1996), p. 1117.
[3] ORTEC. Model 113 Scintillation Preamplifier Operating and Service Manual.
[4] T Leadbeater. Introduction to gamma ray spectroscopy.
[5] National Nuclear Data Center. NuDat 3 User Guide. Information extracted from the NuDat database. Interactive
Chart of Nuclides. url: https://www.nndc.bnl.gov/nudat/.
[6] SPECTECH. Spectrum Techniques UCS-30 Universal Computer Spectrometer.