Which channel number is the life span of the U-verse

<--- page back      Environmental projects next page ->


Make your own gamma spectrometer page 2

On this page I give a short description of various experiments that I carried out with my first self-made spectrometer.

1.) Attempts to identify radioactive nuclides in ore samples

The self-made spectrometer gives me the opportunity to create so-called "radiological fingerprints" from various radioactive material samples. Radiological fingerprints are spectrograms that I measured with the self-made spectrometer. Such radiological fingerprints, which were measured here with the same experimental set-up, can be compared very well with one another without having to know more precisely about the properties of the probe. I am assuming that different samples which contain substances with similar radioactive radiation properties produce similar radiological fingerprints.

In contrast to a "gamma spectrum", which should only reflect the radioactive properties of the material sample examined, a "radiological fingerprint" reflects not only the radioactive properties of the material sample examined, but also peculiarities (e.g. energy resolution, control effects, efficiency) of the material used Probe again.

In this terminology, a radiological fingerprint is identical to the pulse height spectrum measured by a unique measuring device at a specific radio source.

In other words :
A "radiological fingerprint" that comes from a certain radio source is practically identical to the pulse height spectrum which was measured with a unique measuring device at this certain radio source.

A free demo program [1] , which identifies radioactive nuclides from a pulse height spectrum, is of course also available on the WWW .... "Identify-Demo" for [DOWNLOAD]

The following picture (Fig. 01a) shows a pulse height spectrum of radium-226, which was measured with a Ge detector.[2]

Bild01a: Pulse height spectrum of radium-226, measured with a Ge detector

In the pulse height spectrum, the pulse height measured at the detector output (channel number) is shown in a spectrum. (Frequency of events as a function of the pulse height). If a gamma source is used, a gamma pulse height spectrum is created. With an alpha detector in connection with an alpha source, an alpha pulse height spectrum can be generated. In various references I have viewed, the abscissa of the pulse height spectrum is simply labeled with the photon energy instead of the channel number. If there is no efficiency calibration, I find the pulse height spectrum labeled in this way often referred to as the "gamma spectrum" in the literature. This can unfavorably lead to misinterpretations and misunderstandings. To avoid this, I differentiate between the pulse height spectrum and the gamma spectrum and first eliminate effects from the background and scatter effects from the pulse height spectrum. Then the pulse height spectrum is mapped (rectified) onto the gamma spectrum by means of a suitable calibration (energy calibration and efficiency calibration). As a rule, I now offset the y-axis with the energy-dependent detector efficiency and only speak of a gamma spectrum as soon as this "efficiency calibration" has been carried out. In this way, the measurement result approximates the quantity of interest.

In the gamma spectrum, the photon energy flowing in a sample volume or flowing over a surface of the sample is shown in an organized manner to form a spectrum. (e.g. gamma flux as a function of photon energy). As far as I know, an ideal gamma spectrum cannot be measured because of unavoidable approximations in the calibration and effects from random measurement errors, or it can only be measured approximately with a great deal of technical effort. The gamma spectrum from 100 Bq of radium-226 activity with the gamma flux as the variable of interest is shown in the next picture 01b:

Fig. 01b: Gamma spectrum of radium-226 (with the gamma flux as the quantity of interest)

Often the activity is the quantity of interest. The gamma spectrum corresponding to Image01b (but now with the activity as the variable of interest) is shown in the next Image01c:

Fig. 01c: Gamma spectrum of radium-226(with the activity as the quantity of interest)

The activity of the parent nuclide is now in equilibrium with the activities of its daughter nuclides (decay products) on the same level. It can take up to 180 years until this state of equilibrium is reached with all radium daughter nuclides. In practice, however, an almost stationary radioactive equilibrium is reached after about 30 days. [3])
From my early experiments with the Ram63 probe (RFT scintillator), however, for the time being I only consider and compare pulse height spectra for the sake of simplicity.

If the material properties and design of the measuring probe used are basically suitable as a scintillator, and if pulse height spectra (radiological fingerprints) are compared with one another, which are without exception generated using a unique test setup from different radio sources and excluding systematic measurement errors, then the material properties and the design of the probe have no influence Correctness of logical conclusions which emerge from comparisons of pulse height spectra (radiological fingerprints).

A systematic measurement error can arise, for example, when a third-party radio source was inadvertently left in the vicinity of the experimental set-up and interspersed with the measurement result. It goes without saying that every experimenting person must carefully ensure that systematic measurement errors are ruled out.

Different experimenters (hereinafter also referred to as "searchers") who follow the terminology described here in their approach must therefore come to exactly the same logical conclusions based on their measurement results, even with measuring devices of different types.

For interested seekers, I will now briefly describe the one I currently use. used experimental set-up. With this I measured all the radiological fingerprints (pulse height spectra) shown on this page.

Image 1: My experimental setup for measuring a radiological fingerprint

First of all, I would like to examine two "pitchblende" samples with this measuring device. Pitchblende, also known as uraninite, is known to have a very high uranium content. The "radiological fingerprint" generated by the pitch diaphragm should therefore j
like enemies of uranium. Five game balls with uranium glaze first give me a radiological fingerprint of uranium. (Backgr. Is eliminated in all spectrograms shown on this page.)

Fig. 2: Example pulse height spectrum (radiological fingerprint) of uranium glaze,
measured with RFT scintillator (experimental setup shown in Figure 1)

I have two small ore samples available to me. This is
1.) a small 0.5 gram pitch black sample (mouse eye)
2.) about 5.4 grams of ore brownish-gray in color (bröggerite)

For the sake of simplicity, I refer to both ore samples as "pitchblende" because these samples were offered and sold to me under this name. First, I take the radiological fingerprints of these samples. Then more fingerprints of samples with known radioactive nuclides. Radiological fingerprints of known nuclides can then be compared with those of the pitch blends. The composition of the pitchblende samples can possibly be determined on the basis of matching features in the various pulse height spectra.

Image 3: Examples of some material samples that have now been further investigated
1.) 0.5 grams of pitchblende (mouse eye);
2.) 5.4 grams of pitchblende (bröggerite);
3.) 5 balls with uranium glaze;
4.) A thorium mantle

Image 4 shows the radiological fingerprints of these four material samples.

Energy calibration:

It is also worth mentioningI have an approximateEnergy calibration theabscissa I have now performed linearly based on spectral comparisons with americium and potassium chloride. The MCA used has a total of 256 channels and the device detects gamma radiation between almost zero and around 2.56 MeV.

However, due to the qualitative material properties of the scintillator used, the spectrograms shown here are not very sharp gamma energy spectra.

Therefore, nuclides cannot be identified on the basis of sharp spectral lines in the spectrogram. I would therefore like to take a different approach to the identification of nuclides. The method presented here for identifying radioactive nuclides in material samples does not require any energy calibration.

But I have adjusted the abscissa approximately to energy in keV for an improved clarity of the displayed spectrograms.

Channel number:

The channel number can easily be calculated back here. This results from the energy value readable on the abscissa, divided by ten. Example: Channel number 146 was converted to 1460keV for display in the spectrogram.

Fig. 4: Pulse height spectrum of four different material samples measured with an RFT scintillator

The radiological fingerprints of the two "pitchblende" samples examined are similar to the one I measured on a thorium mantle. Because pitchblende is considered to be rich in uranium, I expected a match with the uranium glaze. However, the radiological fingerprints from the "pitch blanks" hardly resemble the one I measured on the five game balls with a uranium glaze. Presumably, the pitch diaphragm examined is a thorium-rich pitch diaphragm (also
Called Bröggerit) with lower uranium content.

Because Backgr. has already been eliminated, the still clearly recognizable background in the pulse height spectrum of the pitch diaphragms in Figure 4 also allows me to infer scattering events that come from high-energy gamma radiation. This cannot be seen in the spectrogram of the uranium glaze.

Next, I was interested in the radium spectrogram. I put an old Kienzle pocket watch with luminous hands under the probe. The red curve in Figure 5 shows the result.

Fig. 5: Pulse height spectra from the luminous hands of an old pocket watch (red) in comparison with the other materials,
measured with an RFT scintillator.

The material samples radiate on the one hand with very different intensities and on the other hand, the duration of the measurement varies. Spectrograms, which represent absolute measured values, are therefore very difficult to judge with one another visually for correspondence. It is more favorable if each spectrogram is normalized to the respective mean value of the pulse heights over all channels. I did that once and presented the results for four of the material samples in Figure 6.

Fig. 6: Spectrograms of four different material samples normalized to the respective mean values,
measured with an RFT scintillator

Now it can even be seen that the spectrogram of the large piece of pitchblende (Bröggerite) largely corresponds to the spectrogram of the luminous hands of the pocket watch, as well as an extensive agreement with the thorium spectrogram Proportion of thorium is included. The practically congruent correspondence of the spectrogram from the pocket watch with the thorium spectrogram is also very noticeable here and makes me think. According to the dealer, the luminous hands of the pocket watch contain radium as a radioactive substance. However, dealer information can also be very uncertain, especially if these have been handed down.

The sensible use of reference and test sources for nuclide identification has, among other things, limits on the trustworthiness of the material properties of the product used as test source.

Because I use a probe from the RAM63 scintillation counter for my measurements, but have no precise knowledge of the scintillator used in this probe, I give the scintillator the designation "RFT" here to represent, for example, "NaJ (Tl)" or "plastic" .

As far as I know, the RFT scintillator is an unspecified luminescent material that is embedded in a plastic compound. Although the fluorescent material has certainly suffered due to its age and it is probably not a NaJ crystal, the quality of the spectrograms generated with the RFT scintillator is close to NaJ (Tl) spectrograms. In any case, the RFT scintillator seems to me suitable for simple examinations and to gain initial experience in dealing with scintillators. Therefore I will carry out further experiments with this scintillator.

Fig. 6a: RFT scintillator

For the time being, I would like to carry out the last measurement attempt with thorium-containing tungsten electrodes and compare them directly with the spectrogram of the thorium mantle. The result is shown in the next picture7:

Fig. 7: Thorium mantle and thorium tungsten electrode, spectrograms of thorium are compared,
measured with an RFT scintillator

Figure 7 shows very nicely how this comparative method works. The two samples come from completely different suppliers and were purchased at different times. Completely independent of the properties of the scintillator used, the method provides very good comparative results. The spectrogram of the thoriated tungsten electrode (red curve Fig. 7) and the spectrogram of the thorium mantle (blue curve Fig. 7) fit almost identically on top of each other, apart from the signal noise. Of course, this corresponds to my expectations, because both the thoriated tungsten electrode and the thorium mantle contain the same radionuclide thorium, according to information from two independent dealers. Thus the spectrogram shows Fig. 7
very likely the radiological fingerprint (spectrogram) of thorium and is therefore suitable for the first entry in my small private library with radiological fingerprints of certain nuclides.

Comparisons by eye are very tedious. Therefore I would like to use a correlation coefficient as a numerical measure of the degree of agreement. This gives me numerical comparison possibilities by computer.

For comparison purposes, I carried out further measurements and summarized the results in a graphic below. Here you can also see the measurement result on a rod-shaped Cobald-60 preparation, which I was able to measure with my spectrometer together with Professor Will at the LMU Munich.

Fig. 8: Pulse height spectra of Cobald-60, potassium chloride and
Lutetium (III) oxide, measured with my RFT scintillator

Now I would like to use a simplified sample calculation to estimate whether the measurement result shown in Figure 8 is correct. For this sample calculation, I subject the peak heights of the Cobald-60 preparation and the potassium chloride preparation shown in Figure 8 to a detailed consideration below. The following considerations relate to the illustration in Fig.8:

A.1 be the activity of the cobalt-60 preparation; a stick-shaped typical school preparation (approx. 11kBq)
A.2 be the activity of the potassium chloride preparation; 2kg potassium chloride divided into two PE containers (approx. 36kBq)
A.3  be the activity of the lutetium preparation; 10g Lu2O3 in two glass containers (approx. 450 Bq)

Figure 8 shows the absolute measurement results. The peak maxima shown above the background allow relative conclusions to be drawn about activities A.1 and A2 of the preparations. In comparison, the activities of the preparations can be calculated from known key data. An activity of 370KBq or 10µCi was originally stated for the Co60 preparation at the date of purchase, but in the meantime about 5 half-lives have passed since the date of purchase. The current activity A1 of the Co60 preparation during the measurement is therefore about 11KBq. This is calculated from the activity known at 370KBq on the date of purchase and the number of half-lives h = 5 that have elapsed since the date of purchase

A.1= 370KBq / (2 ^ h) = 370KBq / (2 ^ 5) = 370KBq / 32 = 11KBq.

From the Cobald-60 decay scheme I can see that activity A1 200% of Co60 appears as a gamma flow outside the preparation. So the total gamma flux of the Co-60 preparation is about

Φγ1 = dN1/ dt = A.1* 200% = 11KBq * 200% = 22K [1 / s]

Fig. 8a: Decay pathways of Cobald-60

The total activity A2 of 2Kg KCl is calculated as A2= 2Kg * 18KBq / Kg = 36KBq activity.

From the disintegration pattern of potassium-40 I can see that the activity A2 11% of K40 appears as a gamma flow outside the preparation:

Fig. 8b: Decay pathways of potassium-40

The total gamma flux (gamma emission) of the KCl preparation is therefore approximately

Φγ2 = dN2/ dt = A2* 11% = 36KBq * 11% = 3.96K [1 / s].

The peak heights of the Co60 preparation and the KCl preparation shown in Figure 8 can be set in relation to each other. Assuming that the detector detects approximately the same proportions of the total flow in both measurements, the peak ratio is approximately the same
Φγ1/ Φγ2 the gamma flows of both preparations are the same. Because the gamma energy of Co60 and K40 are very close to each other, it is not necessary here in the practical treatment to include an effectiveness characteristic of the detector.

The expected ratio of the gamma fluxes of both preparations is therefore:

Φγ1/Φγ= dN1 / dt / dN2 / dt = 22K [1 / s] / 3.96K [1 / s] = 5.6

The measured ratio of the gamma fluxes of both preparations can be read from the ratio of the corresponding peak areas in the diagram in Fig. 8. Instead of the peak areas, I put the peak heights in relation to each other. The Cobald-60 preparation appears here at the peak maximum with a counter reading of about Z1 = 500, while the corresponding peak height for the potassium chloride preparation is about Z2 = 100 (see Fig. 8).

The measured ratio of the gamma fluxes of both preparations results here from the ratio of the measured peak heights to:

Z1 / Z2 = 500/100 = 5

The practical measurement result "5" largely agrees with the theoretical expectation "5.6". I explain the difference between practice and theory with the various simplifications I made in this treatment. On the one hand, there are differences in the geometric arrangement of the two preparations. This has not been taken into account in this simplified calculation. Furthermore, I have only roughly estimated the activity of the Co60 preparation due to the fact that the date of purchase is not exactly known to me. Furthermore, the ratio of the peak heights in the case of the poorly resolved two Co60 peaks and the also imprecisely readable potassium peak height does not exactly reflect the ratio of the actual peak areas. In terms of magnitude, however, the measurement result appears coherent. Because the result of this sample calculation seems to be correct to me, I would no longer like to deal with other influences such as the influence of geometric relationships and an influence of the peak resolution.


I leave the measurement result of a Lutetium preparation, also shown in Figure 8, as a further possibility for comparison. I therefore measured the Lutetium preparation as a comparison option after I was told that Lutetium was suitable for the energy calibration of a gamma spectrometer. Naturally occurring lutetium has a specific activity of about 52Bq / gram and contains about 2.6% of the radioactive isotope Lu-176 [1]. The lutetium (III) oxide (Lu2O3) then has a specific activity of around 45Bq / gram.

I found the following additional information in this context:
Lutetium provides gamma energy spectral lines at 55 KeV, 88 KeV, 202 KeV, 307 KeV
which can be used, among other things, for the energy calibration of radiation measuring devices.
Source of supply: Highly pure elements for research, laboratory and teaching
[1] Further information: Test Adapters Based on Natural Lutetium

Fig. 8c: Decay scheme [1] of Lu-176

In my spectrogram (Fig. 8) of the lutetium preparation, a clear increase of around 200 keV to 300 keV can be seen (black curve in Fig. 8). Then another increase in the vicinity of 500 KeV can be seen, which contrasts slightly with the background. This agrees very well with a description of Lutetium, which I have taken from the literature [1].

Fig. 8d: Spectrum of Lutetium-176 measured with NaI (Tl) detector

I recently performed the energy calibration in the lower energy range for my spectrometer using a lutetium preparation.

The next picture9 was made available to me by Mr. Öller and shows the pulse height spectrum of lutetium and americium in comparison.

Fig. 9: pulse height spectrum of lutetium and americium; Graphic by Andreas Öller
The spectrograms of lutetium and americium in Fig. 9 have been measured according to Mr. Öller's self-made spectrometer. Backgr. However, if it has not been eliminated in this spectrum, Fig. 9, various other nuclides from the background also shine into the result. The peak at around channel number 505 probably comes from the obligatory potassium. Up to about channel number 253, the peaks that originate from the preparations can be seen quite well. Information on the creation Fig. 9 and equipment: Pulse height spectrum of lutetium and americium with Harshaw 2x2 "Integrated Nai Type 8SPA8 / S2 detector in connection with a Davidson Multichannel Analyzer 2056-C + Spectrum Amp Option with the following settings: HV = 800V Gain: 512 = 40x Ch = 1024 LLD = 0 ULD = 1023; Measurement times (life time) and activities have been handed down to me as follows: 40 minutes on the lutetium preparation and 7 minutes on the americium preparation and the activity of the americium preparation with 2µC.

Energy calibration:
The energy calibration can be carried out by means of measurements on preparations with known radioactive nuclides. Often the known gamma energy of the potassium-40 isotope that is always present in the environment is sufficient at 1460 keV. The analog pulse height (channel number) coming from the detector usually corresponds very linearly to the photon energy, so that two gamma lines are usually sufficient as calibration points to obtain the channel number-energy assignment over a large energy range.

Under the idealized assumption that the photon energy = zero, the measuring apparatus also delivers the signal voltage = zero and the calibration is proportional. An energy calibration can then easily be carried out using the obligatory K40 isotope, which is always present in the environment.

A calibration can be proportionally adjusted electrically by adjusting the gain on the MCA preamplifier. First, the high voltage on the photomultiplier is set so that the maximum photon energy to be measured falls just below the highest channel number. You can also readjust proportionally by changing the reference voltage on an AD converter on the MCA. If the pulse height coming from the detector does not correspond linearly to the photon energy, fine adjustment can also be made using software. For this purpose, an approximation function is determined from several calibration points, which implements the channel number-energy assignment. A brief description of my approach to energy calibration using the approximation function is shown in the next pictures (Fig. 9a to Fig. 9e).

First, I measure spectrograms of reference preparations whose nuclides and gamma spectral lines are well known to me. I am happy to use a large amount of potassium chloride as the first reference preparation.

The spectrogram of KCl marks the channel number on which about 1460 KeV of photon energy fall. As an example, I am again using measurement data that was made available to me by Mr. Andreas Öller.

Fig. 9a: Pulse height spectrum from a KCl preparation
measured with NaI (Tl) detector

Another very useful calibration option is a lutetium preparation. I am using 10 grams of lutetium (III) oxide divided into two small glass containers. The spectrogram of lutetium oxide very nicely marks the channel numbers on which the two photon energies 202 KeV and 307 KeV fall.

Fig. 9b: Pulse height spectrum from a Lu2O3-Preparation
measured with NaI (Tl) detector

A Cs-137 preparation provides the label at 662 KeV photon energy.

Fig. 9c: Pulse height spectrum from a Cs137 preparation
measured with NaI (Tl) detector
The measured curve shapes are compared with the references (here the red spectral lines in Fig. 9a - Fig. 9c) and after some experience can be brought to congruence with these. Corresponding calibration points are now as pairs of values
(Channel number; photon energy) can be read off.

Picture 9a to picture 9c serve as an example, which I produced for exercise purposes based on measurement data from Mr. Andreas Öller. The following four pairs of values ​​can be read:

x y
Channel E [KeV]
 75             202
113            307
238            662  
508          1460

If these value pairs (channel number; photon energy) = (x; y) are transferred to a Cartesian coordinate system, a largely linear relationship between the channel number and the photon energy can be seen (see Fig. 9e).

An approximation function can now be calculated, which maps the channel number to the photon energy.

Fig. 9e: Examples for mapping the channel number to the photon energy using approximation functions

Effectiveness calibration /Intensity calibration :
The variable of interest is mostly the activity of the nuclide in the measured sample, sometimes also the gamma flux or the flux density.
If one of these variables is to be determined absolutely, the counting yield or response probability of the detector must be calibrated as a function of the gamma energy. The number of pulses counted for a specific energy differs from the number of gamma quanta emitted by a gamma source for this specific energy. By means of the effectiveness calibration, the number of counted pulses and the number of gamma quanta actually emitted by the source should be set in a constant ratio over the widest possible energy range.

Scintillators have the peculiarity of delivering an increased number of pulses counted at the energy in question when the energy of the gamma quanta is low. To calibrate the effectiveness, the number of pulses counted for low energy values ​​must be divided by a numerically larger quotient. In the case of higher energies, it is divided by a lower quotient.

As a suitable course for these quotients falling towards higher energies, I like to use the spectrogram from the backgram in simplified form as a yardstick. heran .. This has the advantage that expensive and difficult to obtain radioactive test and calibration preparations can be largely dispensed with during calibration.

If the spectrogram of the gamma background (backgr.) Is plotted on a double logarithmic scale, then according to my observation the values ​​are arranged in a wide range on a straight line. I would like to implement this observation in order to normalize a spectrogram, so to speak, to the backgram spectrogram, as follows. For this purpose, the Backgr. Spectrogram is plotted on a double-logarithmic scale and a straight line of best fit is now drawn through the value pairs of the double-logarithmic Backgr.

What is rectified:
Y = ln (Zb) and X = ln (N): Y = a + b * X

N is the channel number and Zb is the counter reading for the respective channel number in the backgr. Pulse height spectrogram. Once the coefficients a and b of the regression line have been determined, the pulse height spectrogram of the sample is then related to the regression function Zb '= e ^ (a + b * X).

Effectiveness calibrations below 100 keV are not trivial.

I did not consider support points with photon energy below 100keV for my effectiveness calibration.

Note: Graphics can be displayed larger in the Firerfox browser if required by clicking on the right mouse button and "Show graphic"

<--- page back

next page ->

 I ask for information and comments by e-mail to: [email protected]

[1] Identify, program for nuclide identification from gamma spectra
[2]Gamma spectroscopy with NaI and Ge detectors
[3] Basic features of practical radiation protection By Hans-Gerrit Vogt, Heinrich Schultz

(last text change on this page: 04/12/2012)