Optimization and Implementation of low-background Gamma Spectrometry using HPGe Detector in Environmental Radioactivity Research

Transcription

1 Optimization and Implementation of low-background Gamma Spectrometry using HPGe Detector in Environmental Radioactivity Research Author : Boon Kiat Chew Supervisor: Dr Taw Kuei Chan Co-supervisor : A/P Thomas Osipowicz A Honours Thesis submitted in partial fulfilment of the requirements for the Degree of Bachelor of Science with Honours Department of Physics Faculty of Science National University of Singapore Academic Year 2014/2015

2 Abstract We are interested in the ability of the high-purity Germanium (HPGe) detector in detecting low-level gamma energies. We first did a energy calibration for the HPGe detector using a Eu-152 sample. Using this calibration, we found the efficiency, energy resolution, minimum detectable activity and minimum detectable mass for the detector. We then did a back-calculation to find the activity of another Eu-152 sample. We also took a reading of rice, flour, milk powder and soil.

3 Acknowledgement I would like to thank Dr Chan and A/P Thomas for their invaluable help for this project, where they would often draw time out from their busy schedules for project meetings. I would also like to thank them for their patient guidance and advice, for without which this project would not have been possible. 1

4 Contents Contents 2 1 Motivation Environmental Effects Chernobyl Nuclear Disaster Fukushima Daiichi Nuclear Disaster Radionuclei Nuclear Plans In Asia Theory Gamma radiation Compton Effect Photoelectric Effect Pair Production Processes In Detector Background Experimental Set-up

5 CONTENTS CONTENTS Semiconductor Detectors High Purity Ge Detector Pre-amplifier Amplifier Multichannel Analyser Data Analysis Energy Calibration Background Spectrum Energy Resolution Energy Peak Efficiency Limit Of Detection Back Calculation Samples Food Products Soil Samples Conclusion 63 Appendix A Gamma Vision 65 Appendix B Tables 70 B.1 Energy Calibration B.2 Energy Resolution

6 CONTENTS CONTENTS B.3 Efficiency B.4 Minimum Detectable Activity Bibliography 73 4

7 Chapter 1 Motivation 1.1 Environmental Effects With the usage of nuclear energy, there is a chance that radionuclides can be accidentally released into the environment. Studies have been made to study the effect of nuclear accidents on the environment, and more importantly, to humans. There have been two notable nuclear accidents that had resulted in the release of large amounts of radionuclide into the environment, Chernobyl and Fukushima Daiichi Chernobyl Nuclear Disaster The Chernobyl nuclear disaster in 1986 resulted in large quantities of radioactive particles being released into the atmosphere. resulting in radioactive contamination of the surrounding environment. Environmental contamination 5

8 1.1. ENVIRONMENTAL EFFECTS CHAPTER 1. MOTIVATION can result in direct exposure of radioactivity to humans, or indirect exposure through contaminated food. Studies have been done after the accident to study the effect on the environment and humans. One study by the IAEA in 2005 found that most of the radionuclide released during nuclear accidents have short half-lives. 1 Smaller amounts of the long-lived radionuclide were released. Out of the short-lived radionuclide, radioactive iodine is a cause for concern as it will accumulate in the thyroid after ingestion. For radionuclide with long half-lives, Cs-134 and Cs-137 are important contributors to radioactive contamination. Other radionuclide have deposition levels were too low to cause a problem, or have low transfer ratio of soil-to-plant to cause real problems in agriculture. In urban areas, open surfaces such as roads and roofs became contaminated with radionuclide. However, the water solubility of caesium resulted in high Cs-137 activity around houses, where the rain had transported the Cs-137 from the roofs to the ground. Additionally, cleaning process lead to the secondary contamination of sewage systems. The nuclear incident also lead to contamination of food products. Initially, milk was the main contributor to internal dose due to large amounts of I-131 being released. The radioiodine deposited on plant surfaces were grazed by dairy cow. Radioiodine ingested is absorbed in the gut and is then transferred to the animal s thyroid and milk within a day. During this period, the I-131 activity concentration in 6

9 1.1. ENVIRONMENTAL EFFECTS CHAPTER 1. MOTIVATION milk in affected regions exceeded regional action levels by a few hundred to a few thousand Becquerel per litre. In Russia and Ukraine this lead to significant thyroid dosage to those who consumed milk, especially children. In the long run, milk was contaminated with radiocaesium. Contamination of plant products happen over two phases. Initial contamination was due to the direct deposition of radionuclei onto the plants. After direct contamination, plants uptake radionuclei through contaminated soil, hence continuing to pose a health issue. Cs-137 and Cs-134 where especially problematic due to its solubility in water, was well as it being used by the plant in place of other minerals such as potassium. The highest levels of contamination with radiocaesium was observed in mushrooms, due to their tendency to accumulate mineral nutrients, such radiocaesium. For animals, the radionuclides circulate in the blood after ingestion. Some accumulate in specific organs, for instance, radioiodine accumulates in the thyroid, whereas radiocaesium is distributed throughout the soft tissues. Hence it was found that I-131 and Cs-137 in meat, milk and plant products are the most important contributors to human internal dose. However, due to the long half-life of Cs-137, the activity concentration in these food products have been decreasing slowly. The decrease in activity concentration is about 3 to 7 percent per year. This means that Cs-137 will continue to contribute to human dose for years to come. 7

10 1.1. ENVIRONMENTAL EFFECTS CHAPTER 1. MOTIVATION Another study of foodstuff in Poland has found that I-131, Cs-134 and Cs- 137 were the main contributors to activity in foodstuffs. 2 The contamination with I-131 decreased quickly after June The only considerable concentrations observed were for Cs-134 and Cs-137. It was found that the radiation contamination of fruits and vegetable remains the same after a few years. However, the higher radioactivity still remains in milk and forest mushrooms. This agrees with the finding of the previous report. Figure 1.1: Cs-137 activity in milk. It can be seen that activity is decreasing slowly, and has not returned to pre-accident levels. For cases of transference of radionuclei from soil to grass to animals, milk is an ideal liquid to dissolve the radionuclei. 3 This is because milk contains fat, while also existing as an aqueous solution, Therefore, both fat-soluble and water-soluble contaminant can be found in milk as it offers both environments. This is important as milk is a fundamental food for infants and children, and is consumed by all the age groups. Hence from this we can 8

11 1.1. ENVIRONMENTAL EFFECTS CHAPTER 1. MOTIVATION see that there is a need to identify food that are vulnerable to radioactive contamination, and to have a reference set of data we can refer to in any event of a contamination. Hence preliminary data is necessary for such food products Fukushima Daiichi Nuclear Disaster The Fukushima Daiichi Nuclear Disaster in 2011 resulted in radionuclides in the form of fine particles or souble gas being released into the environment. A study on the soil about thirty kilometres north-west of the power plant shows the radioactivity of the soil samples collected. 4 9

12 1.1. ENVIRONMENTAL EFFECTS CHAPTER 1. MOTIVATION Figure 1.2: A larger amount of short-lived radionuclei were released initally but decayed rapidly, resulting in Cs-134 and Cs-137 being main contributors of activity. As can be seen from the graph, two months after the incident, the majority of the residual deposits were Cs-134 and Cs-137 due to other radionuclides having shorter half-lives. Though Cs-134 and Cs-137 initially only accounted for 9 percent of the total activity, from 20 May 2011 on, they were contributing more than 80 percent of the activity of the residual deposits in Japan. The accident led to the contamination of the aboveground parts of the plants, 10

13 1.1. ENVIRONMENTAL EFFECTS CHAPTER 1. MOTIVATION and consequentially the plant products intended for consumption. The study found that leafy vegetables such as spinach were the most affected by the contamination due to the fallout directly on the leaves. Over the course of the month of March, several vegetable samples from the surrounding districts showed contamination exceeding sales or consumption standards. Until the end of June, caesium activites were still detected, though they are below the sales and consumption standard limits. Figure 1.3: Change in iodine-131 and caesium-134 and caesium-137 contamination in spinach from the Fukushima prefecture Later in the year, other plant foodstuffs showed significant levels of contamination with caesium-134 and caesium-137. These products did not exist at the time of the nuclear accident and thus was not contaminated by the fallout. Instead it is the transfer of caesium from the soil to the roots that is the 11

14 1.2. RADIONUCLEI CHAPTER 1. MOTIVATION main cause of contamination in these products. However, the contamination is always far lower than the initial contamination caused by direct deposition. Furthermore, due to the migration time of radionuclides in soil,this contamination is only significant for radionuclides having a necessary long half-life, such as Cs-134 and Cs Radionuclei From the case studies, it can be seen that the radionuclei that post the greatest health risk are I-131,Cs-134 and Cs-137. I-131 has a half life of 8.02 days, and produce gamma radiation of 364 kev. 5 Cs-137 has a half-life of 30 years and emits gamma radiation of 662 kev. In addition, another radioisoptope of interest is K-40, which have a half-life of 1.28E10 years and produces gamma radiation of 1460 kev. 6 K-40 can be found in most soils, building materials, plants and animals, and is present in % of naturally occurring potassium. The radioactivity from the K-40 in our bodies contributes to about half of our yearly exposure to all sources of radiation. Furthermore, since Cs and K belong to the same periodic group, they are competing elements for transfer from soil to plant, When the transfer of ceasium from soil to plants increase, the corresponding transfer of potassium to plants decreases. The similar chemical behaviour of caesium and potassium makes it important to study K-40 as well. Therefore, we aim to focus on these radionuclei and to test the capability of our gamma detector at these energy range. 12

15 1.3. NUCLEAR PLANS IN ASIA CHAPTER 1. MOTIVATION 1.3 Nuclear Plans In Asia The case studies show us that nuclear accidents overseas can have a direct impact on us, either through direct contamination or through contaminated food. This is increasing relevant as our neighbouring countries are making nuclear plans in the near future. 7 Countries such as Vietnam and Indonesia are actively making preparation for nuclear energy, and have already searched out potential sites for nulcear power plants. Figure 1.4: Proposed Sites of Nuclear Power Plants in Vietnam However, there are some concern raised over the location of the nuclear power plants. Vietnam is known to be vulnerable to the impacts of climate change 13

16 1.3. NUCLEAR PLANS IN ASIA CHAPTER 1. MOTIVATION such as rising sea levels and stronger typhoons. In particular, Ninh Thuan is identified as a disaster-prone coastal province. Furthermore, Vietnam s coast has been subject to tsunamis in the past. Figure 1.5: Proposed Sites of Nuclear Power Plants in Indonesia Similarly, the nuclear plans in Indonesia has draw concern both domestically and internationally due to the frequent occurrence of natural disasters such as earthquakes and tsunamis. Depending on the extent of an accident, sources of food can be contaminated by airborne radiation and radioactive water from affected power plants. Transboundary airborne particles may contaminate agricultural farmlands not just in Vietnam, but other countries such as Thailand. Regional fishing grounds can also be contaminated, especially the South China Sea. This will impact Singapore significantly as we import a large percentage of our food. Furthermore, other South-East Asian countries such as Philippines and 14

17 1.3. NUCLEAR PLANS IN ASIA CHAPTER 1. MOTIVATION Malaysia are also making nuclear plans for the near future. Hence it can be seen that in the near future Singapore could potentially be surrounded by countries that have nuclear power plants. There is a need to be prepared in the event of any nuclear accident. This is especially relevant to us as Singapore imports a large percentage of our food. Hence we need to have the capacity to identify food products that are vulnerable to radioactive contamination. We need to be have the ability to monitor food products over a period of time as radioactive nuclulides such as caesium have long half-lives. As such, we need to test the capability of detectors as well as calibrate them such that prelimary work can be done. This will be the main focus of our project. 15

18 Chapter 2 Theory 2.1 Gamma radiation The activity of a radioactive source is defined by its rate of decay and is given by 8 A = dn dt = λn where: A = activity N = number of radioactive nuclei λ = decay constant The SI unit of activity is the Becquerel (Bq). It is defined as the activity in which one nucleus decays per second and 1 Bq is equivalent to s 1. Measurement of radiation energy is done in electron volts (ev), and is defined as the 16

19 2.1. GAMMA RADIATION CHAPTER 2. THEORY kinetic energy gained by an electron by its acceleration through a potential difference of 1 volt. It is related to the Si unit of energy, Joules (J), by 1 ev = J Gamma radiation is produced by an excited nucleus when it transit from from a higher to a lower energy level. For the case of Cs-137, the caesium can undergoes beta decay to energetic Ba-137, which is unstable and decay to Ba-137 by emitting a gamma ray, shown in the diagram below. Figure 2.1: Emission of Gamma Ray from Cs Gamma rays interact with matter through 3 main processes : Compton scattering, photoelectric absorption and pair production Compton Effect Compton effect occurs when a photon scatters off a nearly free atomic electron. 10 This results in a less energetic photon and a scattered electron carrying the energy lost by the photon. 17

20 2.1. GAMMA RADIATION CHAPTER 2. THEORY Figure Photoelectric Effect For the photoelectric effect, an atom absorbs an incoming photon and emits a atomic electron, known as a photoelectron. The kinetic energy of the photoelectron is given by T e = E γ E B where T e E γ = kinetic energy of photoelectron = energy of incoming photon E B = binding energy of electron Pair Production In pair production, an incoming photon interacts with an atom to create an electron-positron pair. The energy for this process is given by E γ = T + + mc 2 + T + mc 2 18

21 2.1. GAMMA RADIATION CHAPTER 2. THEORY where E γ T + T m = energy of incoming photon = kinetic energy of positron = kinetic energy of electron = mass of electron Since there is a energy threshold of 2mc 2 = MeV, this means that pair production process is only significant at higher energy levels. The figure below shows the three gamma ray interaction processes and the energy range in which they are dominant. Figure

22 2.1. GAMMA RADIATION CHAPTER 2. THEORY Processes In Detector Figure 2.4: Processes during gamma ray detection. (1) Photon Compton scatters and leave crystal before depositing all its energy. (2) Compton scattering followed by total energy deposition though photoelectric absorption. (3) Pair production followed by total absorption. (4) Pair production followed by 1 annihilation photon escaping. (5) Pair production followed by both annihilation photon escaping. The figure above shows some of the processes that can occur when a gamma ray enters a solid detector. Firstly, the photon can Compton scatter a few time within the detector, each time losing some energy and producing a photoelectron. Eventually the photon will experience 2 events: either the photon will wander too close to the edge of the crystal and scatters outside, or it 20

23 2.1. GAMMA RADIATION CHAPTER 2. THEORY will continue to lose energy and is eventually absorbed by photoelectric effect when its energy become low enough. The photoelectrons have a short range in the crystal, and loses energy quickly by creating electron-hole pairs. Alternatively, the photon can result in pair production. This can lead to 3 scenarios. In the first scenario, the positron undergoes annihilation, and the resulting annihilation photons are both absorbed. In the second scenario, one of the annihilation photons leave the detector, and the gamma ray deposits its full energy less 511 kev. In the last scenario, both annihilation photons leave the detector, resulting in energy deposition of the full gamma ray energy less 1022 kev. When a gamma ray deposits all its energy in the detector (through photoelectric effect), it will result in a photopeak. However, if the photon undergoes Compton scattering and escape before being fully absorbed, it will lead to a range of energy which forms the Compton continuum. The maximum energy that can be deposited from this event is when the photon back scatters at 180 degrees, corresponding to the Compton edge. When one or both of the annihilation photon escapes from the detector, it will result in a single escape and double escape peak respectively. 21

24 2.2. BACKGROUND CHAPTER 2. THEORY Figure 2.5: Response of detector to monoenergetic gamma rays 2.2 Background The magnitude of the background determines the minimum detectable radiation level, hence it is important to keep this level as low as possible. This is especially important for our experiment, which involves sources of low activity. Background radiation can be grouped into five categories 11 : 1. The natural radioactivity from the materials of the detector itself. 2. The natural radioactivity of equipments, supports and shielding placed in immediate vicinity of the detector. 3. Radiation from the Earth s surface, construction materials of the laboratory or other far-away structures. 4. Radioactivity of the air surrounding the detector. 22

25 2.2. BACKGROUND CHAPTER 2. THEORY 5. The primary and secondary components of cosmic radiation. The radioactivity of ordinary construction materials is due to the low concentration of naturally occurring elements that exist in the materials as impurity. Significant contributors are potassium, thorium, uranium and the members of the long decays chains of thorium and uranium. Potassium is a widespread component in concrete and other building materials. Natural potassium contains % of radioactive K-40, with a 11 % chance to emits a gamma ray of MeV when it decays. This will contribute to a noticeable peak in the background spectrum. Thorium, uranium and radium are naturally occurring radionuclei with long decay chains. Of the natural decay chains, the dominant contributors to background are from the decay of Rn-222 and Rn-220. The resultant daughter nuclei such as 214-Pb and Bi-214 will result in gamma energy peaks in the background spectra. 23

26 2.2. BACKGROUND CHAPTER 2. THEORY Figure 2.6: Natural Decay Chain of U-238 series, U-235 series and TH-232 series. Furthermore, the lead shield has an intrinsic activity for Pb-210, which is also part of the U-238 series decay chain. Cosmic rays can also interact with the shield and the gamma detector, leading to additional background contribution. Lastly, additional background radiation can also be observed as a result of the primary gamma ray from the source interacting with the structural and shielding materials around the detector. Compton backscattering of the primary gamma rays, the generation of secondary annihilation photos and characteristic X-ray production through pair production or photoelectric absorption can lead to an increased background. 24

27 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY 2.3 Experimental Set-up Semiconductor Detectors Some semiconducting materials such as germanium forms solid crystals where the valence-4 atoms form four covalent bonds with neighbouring atoms. 12 Since all valence electrons takes part in forming bonds, we have a filled valence band and an empty conduction band. Semiconductors differs from insulators by having a small band gap of about 1 ev. At room temperature, a small number of electrons can be thermally excited across the band gap into the conduction band. This will leave behind a vacancy known as a hole. To control the electrical conduction, small amount of substances known as dopants can be added into the semiconductor material. When a valence-5 atom is introduced, four of the electrons form covalent bonds with neighbouring germanium atoms. The fifth electron moves through the lattice, forming a set of donor states just below the conduction band. This material is known as n-type semiconductor due to the excess of negative charge carriers. When valence-3 atoms are introduced, the excess of holes will form acceptor states just below the valence band. This is known as a p-type semiconductor as the primary charge carriers are positively charged holes. 25

28 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Figure 2.7 When the p-type and n-type materials are brought into contact, the electrons from the n-type material will be able to diffuse across the junction into the p-type material. The electrons will then recombine with the holes in the vicinity of the junction, thereby creating a depletion region. The diffusion of electrons from the p-type region will leave behind ionized donor sites. Conversely, the diffusion of holes from the n-type region will leave behind negatively charged acceptor sites. The charges from these sites will create an 26

29 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY electric field will will eventually halt further migration, resulting in a junction diode. When radiation enters the depletion region, it will create electron-hole pairs. The electron will flow in one direction while the hole in another, resulting in a electronic pulse whose amplitude is proportional to the energy of the radiation. These detectors are often operated with a reverse bias voltage. The reverse bias voltage increases the magnitude of the electric field in the depletion region, thus making charge collection more efficient. The reverse bias voltage also increases the dimension of the depletion region, thereby increasing the active volume of the detector. 27

30 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Figure 2.8: Depletion region in the semiconductor junction. However, simple junction detectors are not suitable for more penetrating radiations. Their major limitation is the maximum active volume that can be created. For germanium of normal semiconductor purity, it is difficult to achieve a depletion depth beyond 2 to 3mm, even when applying bias voltage of near break down level. This depletion depth is easily penetrable for medium energy gamma rays (The range of a 100-keV gamma ray is about 28

31 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY 4mm in germanium). A greater thickness is therefore necessary for detectors to be used in gamma spectroscopy. The thickness of the depletion region is given by d = ( ) 1 2ɛV 2 en where V = the reverse bias voltage N = the net impurity concentration in the bulk semiconductor material ɛ e = the dielectric constant = the electronic charge Since there is a limit to the reverse bias voltage we can use, the other method to increase the thickness of the depletion layer is by increasing the purity of the semiconductor High Purity Ge Detector Current refining techniques are able to reduce the impurity levels in germanium such that a depletion region of about 10mm can be obtained with a bias voltage of less than 1000V. However, the active volume in a planar configuration is still insufficient. A larger active volume is necessary for gamma spectroscopy. Our HPGe detector is constructed with a coaxial geometry instead to obtain a much larger active volume. 29

32 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY As shown in the figure below, one electrode is fabricated on the other surface of a long germanium crystal. The core of the crystal is then removed and the other electrode is placed on the inner cylindrical surface. Since the crystal is long in the axial direction, a larger active volume that is suitable for gamma spectroscopy can be obtained. Figure 2.9: Top down view of the coaxial Ge crystal 30

33 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Figure 2.10: Left: Ge crystal with the thick lines signifying the inner and outer electrodes. Middle: The germanium crystal is installed in a cylindrical lead shield. Right: Actual lead shield (about 4 inches thick). 1 A sample is placed in the lead shield, with the lead shield closed off during data taking. Since the HPGe detector has a small band gap, it cannot work at room temperature due to a large thermally-induced leakage current. the HPGe must be cooled so that thermally-induced leakage does not affect it s energy resolution. This is done by placing an insulated dewar beneath the lead shield, and the detector is placed in thermal contact with a reservoir of liquid nitrogen. 31

34 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Figure 2.11: Insulated dewar that allows the detector to be in contact with liquid nitrogen. 14 Shown below is the basic schematic of the detector set-up. Figure 2.12: Schematic of the detector 32

35 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Pre-amplifier The pre-amplifier converts the charge pulse from the detector into a voltage pulse and drives the pulse to the amplifier. Since the weak electronic signal from the detector goes directly to the pre-amplifier, the pre-amplifier is located as close to the detector as possible to minimise capacitance. Hence the pre-amplifier is usually packaged as part of the HPGe system. This has the additional advantage of keeping the input part of the pre-amplifier cool to reduce electronic noise Amplifier The amplifier has two main roles. The first is to amplify the signal coming from the detector. Amplifier Gain In the amplifier, the amplitude of incoming signal is amplified to a certain degree. This is measured in terms of gain, which can be defined as the ratio of the input signal amplitude to the output signal amplitude. Hence a gain of 100 will amplify the amplitude of the incoming signal by 100 times. The amplified output signal will then be sorted into a higher channel number. For our experiment, we are interested in the detection of gamma radiation of I-141, Cs-137 and K-40, which have energy peaks at 0.362, and

36 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY MeV respectively. Hence we need a gain setting that is allows us to have a good view of energies at about 1.5 MeV while still having good resolution at the lower energy levels. By varying the gain, it was found that a gain setting is 100 is large, and the 1.46 MeV cannot be seen. Conversely, a gain setting of 20 is too low and not optimal for the lower energies. A gain setting of 50 is the most suitable as we can still see the energy peak of K-40, while still maintaining reasonable spectrum at lower energies. The second role is to shape the signal received from the detector. In order to ensure that total charge collection occurs, amplifiers are necessary to ensure a decay time for the pulse. The pulses can be quite long and tend to overlap with each another. Furthermore, the time spacing for nuclear decay is random and can lead to each pulse being superimposed on different residual tail. Such a pulse train can be seen in the figure below. Since the amplitude of a pulse measures the charge Q deposited on the detector, the resulting amplitude is no longer a good measurement of Q. 34

37 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Figure 2.13 To avoid this, we need to shape the pulses in a way such that we obtain a pulse train shown in (b). Is this done by the amplifier. The long tails are removed in a way such that the maximum amplitude of the pulses is preserved. Shaping time To find the best shaping time for the experiment, we took readings of a Cs- 137 source at various shaping time. The full width half maximum (FWHM) of the energy peak at 662 kev is found using the software Gamma Vision (refer to Appendix A for more info about Gamma Vision). 35

38 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY Figure 2.14: FWHM (kev) of Cs-137 energy peak at different shaping time (µs) It can be seen that 6 shaping time (6 µs) gives us the best energy resolution, with the highest peak and lowest FWHM. This agrees with the supplier s recommended setting of 6 µs for the shaping time Multichannel Analyser The multichannel analyser (MCA) is used to convert a analog signal (a pulse amplitude) into a digital signal. The basic function of the MCA involves an analog to digital converter (ADC) and the memory. The memory can be illustrated as a vertical stack of addressable location, ranging from the first address (channel number 1) at the bottom, to the maximum address location at the top. During operation, a pulse first passes through the ADC and is then sorted into a memory location corresponding to its amplitude. 36

39 2.3. EXPERIMENTAL SET-UP CHAPTER 2. THEORY This increases the count of that location by one. A spectrum of the count against the channel number can then be obtained and shown on the computer through a software. 37

40 Chapter 3 Data Analysis For this chapter we will be doing an analysis of the results we have obtained from the HPGe detector. 3.1 Energy Calibration We first need to do a energy calibration for the detector. We used a Eu-152 source for the calibration, which is suitable due to its multiple peaks. Using the HPGe detector, a spectrum for Eu-152 was obtained. The count rate (s 1 ) was then plotted against the channel number. 38

41 3.1. ENERGY CALIBRATION CHAPTER 3. DATA ANALYSIS Figure 3.1 The 11 major peaks were then identified form literature values. 14 Using literature values for the different peaks of Eu-152, we did a calibration by plotting the energy values against the channel number. Next we want to find out if we have a good linear fit for our data. This is done by finding the reduced χ 2 value for our linear plot, given by the equation 15 where χ 2 ν = reduced χ 2 χ 2 ν = χ2 ν = 1 ν [ ] N (E i E (Ch i )) 2 i=1 σ 2 i ν y i σ = degrees of freedom = data from theory = standard deviation of the energy peak centroid E (Ch i ) = A Ch + B, the simulated data from our linear fit 39

42 3.1. ENERGY CALIBRATION CHAPTER 3. DATA ANALYSIS To find σ, we first manually fitted a best Gaussian fit to a photopeak centroid µ. Next, we do a manual shift of the Gaussian fit by varying the peak centroid position to µ. The reduced χ 2 value was found for both curves. From literature, we know that a difference of 1 for the two reduced χ 2 will mean that the resultant µ = µ µ is the standard deviation of the peak centroid, σ. It was found that the fitting at 122 kev gave us a large χ 2 value of 122. The resultant χ 2 ν value is 14.2, which suggest that our data is a bad fit. However, when removed from the linear fit, the χν 2 became 0.738, a much more reasonable number. However, it stills differ from the ideal χν 2 value of 1. This could be due to either the over fitting of the data, or the overestimation of σ. Hence for our calibration, we left out the 122 kev peak, and used ten energy peaks instead. 40

43 3.1. ENERGY CALIBRATION CHAPTER 3. DATA ANALYSIS Figure 3.2: Linear fit of Energy (kev) against Channel for 10 Eu-152 energy peaks From the line of best fit we obtain the equation E = Ch Another way to test for the goodness of the fit is to look at the residual of the data. 16 The residual is obtained by taking the theoretical energy peaks minus the simulated (fitted) energy peaks. If the resultant data points are random, it suggest that our fit is a good fit. 41

44 3.1. ENERGY CALIBRATION CHAPTER 3. DATA ANALYSIS Figure 3.3: Plot of residual (kev) against energy (kev) From our data, it seems that the data points are random, and close to zero, taking into account the error. This suggest that our data is a good fit. We next obtain a calibrated spectrum for Eu-152 using this calibration. Figure 3.4: Spectrum of Eu-152, Count rate (s 1 ) against Energy (kev) 42

45 3.2. BACKGROUND SPECTRUM CHAPTER 3. DATA ANALYSIS 3.2 Background Spectrum Next, we want to test the effectiveness of the lead shield of the HPGe detector. To do so, we first obtain a background spectrum of the empty HPGe detector when the lead cover is closed. Using the calibration found previously, a plot of the count rate (s 1 ) against energy (kev) was obtained. Figure 3.5: Spectrum of background with closed lead shield The major peaks are identified and many were found to be of the natural long decay series of U-238 and Th These are natural decay chains that are found in the environment. As discussed previously, theoretically we had expected that these decay series will contribute to background data. Next, a separate reading was done with the top part of the lead shield open. The resulting data was plotted together with the closed cover reading. 43

46 3.3. ENERGY RESOLUTION CHAPTER 3. DATA ANALYSIS Figure 3.6: Plot of Open Cover vs 5 x Closed Cover As the reading of the closed cover reading is very low, the reading is multiplied by five to allow for easier comparison. From the plot, it can be seen that there is a significant reduction in the background activity, especially at the lower energy range. Hence we can conclude that the lead shield is effective, and it reduces background data by one order. With the shielding, the resultant background is two orders smaller than the reading for Eu-152 peaks. 3.3 Energy Resolution We are also interested in knowing the response of the detector to radiation. The figure shows the pulse height distribution that can be produced by the detector. The curve labelled Good resolution shows a possible distribution around a certain energy, while the curve labelled Poor resolution shows an 44

47 3.3. ENERGY RESOLUTION CHAPTER 3. DATA ANALYSIS inferior distribution around that point. Figure 3.7: Example of pulse height distribution (count vs energy) with good and bad resolution. Assuming that the same number of pulses are recorded for each case, the area under the peaks are the same. Though both distributions are centred at te same point, the width of the poor resolution peak is much wider. This indicated that a large fluctuation was recorded from pulse to pulse even though the same energy is deposited in the detector for each event. If the amount of fluctuation is reduced, the corresponding distribution will be made smaller and the peak will approach a delta function. Hence the ability of a measurement to resolve fine details in the incident radiation energy will be improved with decreasing peak width. From our data for Eu-152, the FWHM of the different peaks are recorded using Gamma Vision. 45

48 3.3. ENERGY RESOLUTION CHAPTER 3. DATA ANALYSIS Figure 3.8: FWHM of Eu-152 energy peaks The relative resolution was then found by taking dividing the FWHM with their corresponding energies, Relative Resolution (%) = FWHM E 100 where E is the peak centroid energy. 46

49 3.4. ENERGY PEAK EFFICIENCY CHAPTER 3. DATA ANALYSIS Figure 3.9: Relative Resolution fitted from eleven energy peaks of Eu-152 Other detectors such as Na(Ti) detectors typically have a resolution of 13.5 % at 113 kev (Lu-177), 7.7 % at 662 kev (Cs-137), and 6.07 % at 1277 kev (Na-22). 18 This suggest that our HPGe detector gives a much better resolution. Since we are interested in identifying different energy peaks in a sample of unknown composition, we need to be able to resolve peaks that may be close together. Hence the HPGe detector, with its high resolving power, is suitable for our experiment. 3.4 Energy Peak Efficiency We are interested in knowing how much activity the HPGe will detect out of the actual activity of a source. This is known as the efficiency of the detector. To calculate the efficiency, we took the fraction of the net area of 47

50 3.4. ENERGY PEAK EFFICIENCY CHAPTER 3. DATA ANALYSIS each photopeak of Eu-152 over the actual activity. Efficiency = Detected Activity Actual Activity = N/t A 0 B where N = net area of each peak A 0 = activity of the source B = brunching ratio for different energy peaks t = time taken for measurement Using our Eu-152 sample, the net area of 11 major peaks was obtained from the programme Gamma Vision. The activity was calculated from the initial activity, and the brunching ratio obtained from literature. Figure 3.10: Energy Efficiency across different energy, fitted from 11 peaks of Eu-152 data. From the line of best fit, efficiency = E ( 0.692) 48

51 3.4. ENERGY PEAK EFFICIENCY CHAPTER 3. DATA ANALYSIS From the line of best fit, we obtain the equation efficiency = E ( 0.692). A value of 0.01 for efficiency will indicate that out of every 100 actual activity emitted by the source, 1 of it will be detected by the HPGe detector. The efficiency is expected to decrease as photons of higher energy will have a higher chance of passing through the detector undetected. For coaxial detectors, there are a variety of fits used for the extrapolation of energy efficiency over a wide energy range. One published function that fits the efficiency over a energy of kev is given by 20 where ɛ = energy efficiency a i = the fitted perimeters lnɛ = E 0 = a fixed reference energy N i=1 a i ( ln E E 0 ) i 1 and is shown below graphically. 49

52 3.4. ENERGY PEAK EFFICIENCY CHAPTER 3. DATA ANALYSIS Figure 3.11: Published fitting of Ln Efficiency against Ln Energy for a wide range of energy Compared to our data, our spectrum seems to show agreement with literature, between an energy range of about 100 to 1400 kev. Figure 3.12: Plot of Ln Efficiency against Ln Energy using Eu-152 energy peaks 50

53 3.5. LIMIT OF DETECTION CHAPTER 3. DATA ANALYSIS 3.5 Limit Of Detection We are also interested in knowing the limit of detection for the HPGe detector. Let N T be the number of counts recorded with a sample at a point, and N B be the number of counts recorded in the absence of a sample. The net counts of the sample will then be 21 N S = N T N B N S is then compared to a minimum detectable value to determine whether the sample contains activity at a particular point. If N S is less then this value, then the sample does not contain activity at that point. Conversely, if N S is larger then this value it is assumed that there is some real activity present. In the absence of statistical fluctuations, the minimum detectable value can be set to zero, and any net postive count can be treated as evidence of real activity. However statistical fluctuation is inevitable in any counting measurement. Hence there will be instances where a postive N S will be observed even in regions of no activity. We have to choose a minimum detectable value that is high enough to minimise the likelihood of such false positives, while keeping it low enough to avoid false negatives (missing real activity). Since our counting time is sufficiently long, we can assumed that the total number of counts of N T and N B follow a Gaussian distribution. From Poisson 51

54 3.5. LIMIT OF DETECTION CHAPTER 3. DATA ANALYSIS statistics, the standard deviation in the number of recorded events N is then expected to be σ N = N We are interested in knowing what is the minimum detectable activity of a sample by the HPGe detector. The relation of actual activity of sample and the activity detected by the detector is given by A 0 ɛ = A det where A 0 ɛ = activity of sample = efficiency A det = activity detected by the detector To find the minimal detectable activity, we first found the gross area (using Gamma Vision) of non-peak regions in our background spectra. Since these are regions without peaks, they corresponds to background activity only. This gives us the relation A actual background ɛ = A detected background = N B t where N B =gross area of background 52

55 3.5. LIMIT OF DETECTION CHAPTER 3. DATA ANALYSIS From Poisson statistics, the standard deviation of minimum detectable activity of the sample is then given by σ Aactual background = N B t 1 ɛ By taking 3σ Aactual background, we can then be 99.7 % certain that there is real activity when the activity of the sample is larger than this value. This is thus our minimum detectable activity. Figure 3.13: Minimum detectable activity of a sample in the HPGe detector. From this we are 99.7 % sure that any value above the limit of detection correspond to actual activity. The error bars arise from the uncertainty in the efficiency and the gross area. Next, we are interested in finding the minimum detectable mass of a radionuclei for our detector. This is done by first interpolating from our best fit the 53

56 3.5. LIMIT OF DETECTION CHAPTER 3. DATA ANALYSIS minimum detectable activity of different radionulei. It was found that the minimum detectable activity A min = ( )E 2 + ( )E , where E is energy (kev). Next, we calculated the decay constant λ that was found from the half-lives of the radionuclei. 22 By using the relation, A min = λn min we can then obtain the minimal detectable number of radionuclei N min. Lastly, the number of radionuclei is multiplied by the atomic mass to obtain the minimal detectable mass (kg). We did a calculation for 4 radionuclei, shown in the table below. Radionuclei γ Energy (kev) A min (s 1 ) N min Mass min (kg) I E E-21 Cs E E-18 Co E E E E-19 K E E-10 From our results, this means that if there is a radionuclei mass above the minimal detectable mass, there is a 99.7 % that it is detectable. 54

57 3.6. BACK CALCULATION CHAPTER 3. DATA ANALYSIS 3.6 Back Calculation Now that we know the efficiency of the detector, we can find the activity of an unknown radioactive source. To demonstrate this, we first took another reading of a different Eu-152 source. The efficiency we calculated previously was used to calculate the activity of the source. This values were compared to the actual activity of the source. Figure 3.14: Percentage discrepancy of actual activity from the calculated activity of new Eu-152 sample It can be seen that there is a reasonable percentage discrepancy from the actual activity of the Eu-152 sample. The relatively large percentage discrepancy at about kev range could be due to the relatively large errors in the calculated activity. The efficiency of the new Eu-152 was also found, and plotted against the original efficiency. 55

58 3.6. BACK CALCULATION CHAPTER 3. DATA ANALYSIS Figure 3.15: Plot of the original efficiency with the new efficiency obtained from the second Eu-152 sample. It can be seen that the two values are similar, suggesting good consistency for our detector. 56

59 Chapter 4 Samples 4.1 Food Products We next did a data collection for different food samples. We used a sample of flour, rice and milk powder. For the different spectra, energy peaks of the natural decay chains were identified. Compared to the background, there is some increase in the different energy peaks, shown in the milk sample spectrum below. 57

60 4.1. FOOD PRODUCTS CHAPTER 4. SAMPLES Figure 4.1: Spectrum of milk powder ploted together with background spectrum. However, the most significant difference is the K-40 peak. In all the food product, there was an increase in the K-40 peak. However, the K-40 peak is significantly higher in milk power compared to the rest. Figure 4.2: K-40 peak for background, flour, rice and milk powder spectrum 58

61 4.1. FOOD PRODUCTS CHAPTER 4. SAMPLES This is due to the higher potassium content of milk. Since % of potassium occurs as K-40 naturally, this result is expected. The background count is then subtracted from that of milk powder and the peaks were identified. Figure 4.3: Spectrum for the milk spectrum less the background Hence it is shown the our detector is able to detect small amounts of radioactivity in food samples. These data of food samples can be kept as a reference for the future. Different food samples can be tested regularly as a means of regulation and to check for contamination. Future work can be to take identify and take data from food samples that are radiologically sensitive, such as mushrooms and milk. We could also compare food from different regions to compare the difference in the spectrum. 59

62 4.2. SOIL SAMPLES CHAPTER 4. SAMPLES 4.2 Soil Samples Next, we did a reading of soil collected from MacRitchie reservoir. A spectrum of the soil sample is plotted with the background spectrum. Figure 4.4: Spectrum of soil sample with spectrum of background It can be seen that the readings for the energy peaks are much larger than that of the background. 60

63 4.2. SOIL SAMPLES CHAPTER 4. SAMPLES Figure 4.5: Spectrum of soil sample less the background However, we expected that there will be a peak for Cs-137, due to atmospheric fall-out from the use of nuclear weapons in the past. The reason could be that the soil sample we took is only the top layer, which might be mixed with new compost and fertilizer. A point for future study could be to take a reading for different soil samples from different regions. The data for the soil samples can then be kept as a reference for the future. We would then be able to compare the data of soil to that of an uncontaminated soil sample, either as regulatory checking or in the event of atmospheric contamination. We can further expand the samples in future works to include seawater and tap water as well. Since Singapore is surrounded by the sea, we will be 61

64 4.2. SOIL SAMPLES CHAPTER 4. SAMPLES vulnerable to any contamination of the surrounding seawater. Furthermore, a portion of our drinking water is obtain from the reservoir, which might be vulnerable to atmospheric contamination as well. Hence, a preliminary data reading can be first done for seawater and tapwater. 62

65 Chapter 5 Conclusion In conclusion, we did a energy calibration for the HPGe detector using a Eu- 152 source. Using this calibration, we found the energy resolution, efficiency, minimum detectable activity and minimum detectable mass for the detector. The high resolving power of the HPGe suggest that it is suitable for our experiment, as it is able to differentiate the numerous energy peaks in an sample of unknown composition. We next found the efficiency of the HPGe detector. By knowing the efficiency, we will be able to calculate the actual activity of an unknown sample from the activity detected in the detector. Our efficiency also seems to agree with the trend of the efficiency from literature. Next, we found the minimum detectable activity and minimum detectable mass of radionuclei. The minimum detectable activity means we are 99.7 % certain that any activity detected about this level correspond to actual activity. The minimum detectable mass for different radionuclei means that 63

66 CHAPTER 5. CONCLUSION we are 99.7 % sure that we are able to detect the radionuclei if that amount of it is present in the detector. We then did a back-calculation to find the activity of another Eu-152 sample. The calculated activity has a reasonable percentage discrepancy from the theoretical activity, taking into account the error. The efficiency found from the new Eu-152 source also matches the original one we found initially. Lastly, we took a reading of rice, flour, milk powder and soil. Our data suggest that our detector is able to detect activity in the different samples. For future work, reference data of activity can be taken for food products, soil and seawater from various different regions. 64

67 Appendix A Gamma Vision For our experiment we made use of the program Gamma Vision for our analysis. From the manual, the program does the analysis according to the method shown below

68 APPENDIX A. GAMMA VISION Figure A.1 The background on the low-channel side of the peak is the average of the first three channels of the point of interest (see Fig). The channel number for the background is The middle channel of the three points. The background on the high-channel side of the peak is the average of the last three channels of the point of interest. The channel number for this background point is also the middle channel of the three points. These two points on each side of the peak form the endpoints of the straight-line background.the background is given by the equation B = ( l+2 C i + i=l h i=h 2 C i ) h l

69 APPENDIX A. GAMMA VISION where: B = background l = lowest channel in region of interest h = highest channel in region of interest C i = content of that particular channel 6 = total number of channel used (3 on each side) The gross area of the peak is the sum of the content of each channel between the background, given by: A g = h i=l C l where: A g C i l = the gross area = the data value of channel i = the center channel of the background calculation width at the low energy side of the spectrum h = the center channel of the background calculation width at the high energy side of the spectrum 67

70 APPENDIX A. GAMMA VISION Figure A.2 The adjusted gross area is the sum of all the remaining channels that was marked by the ROI but not used in the background, given by where: A og = h 3 i=l+3 C i A og = adjusted gross area l h C i = the ROI low limit = the ROI high limit = the contents of channel i The net area is the adjusted gross area minus the adjusted calculated background, given by A n = A og B(h l 5) (h l+1) 68

71 APPENDIX A. GAMMA VISION and the uncertainty in net area is given by where: σ An = A ag = the adjusted gross area A og + B ( h l 5 6 ) ( h l 5 ) h l+1 B l h = the background area = the ROI low limit = the ROI high limit The uncertainty in the net area is the square root of the sum of the squares of the uncertainty in the adjusted gross area and the weighted error of the adjusted background. The background uncertainty is weighted by the ratio of the adjusted peak width to the number of channels used to calculate the adjusted background. 69

72 Appendix B Tables B.1 Energy Calibration Figure B.1: Calculated χ 2 from the linear fit of theoretical energy against channel. Sigma was obtain from manual Gaussian fit of energy peaks. 70

73 B.2. ENERGY RESOLUTION APPENDIX B. TABLES B.2 Energy Resolution Figure B.2: FWHM values obtained from Gamma Vision. B.3 Efficiency Figure B.3: Table for efficiency calculation. Initial activity and uncertainty, energy peaks, branching ratio obtained from literature. Net area and uncertainty obtained from Gamma Vision. 71