**4. Radiation detection**

The following overview of radioactivity laboratory equipment for food control is only a brief summary of the required equipment. For information about the theory of radiation detection and measurements, technical details, or special applications, we refer to the standard literature [13–16].

### **4.1 Dose rate monitors**

Measurement devices to count the dose rate of radioactive sources are not sensitive enough for the precise contamination measurements of food. However, they can be adequate for

<sup>1</sup> Treaty banning of nuclear weapon tests in the atmosphere, in outer space, and under water. 5 August 1963.

screening analysis. In 2013 and 2014, the State Laboratory Ticino and the Federal Office of Public Health used dose rate monitors to screen wild boars shot in the southern parts of Switzerland. They only sampled animals over a certain dose rate for precise γ-spectrometric analysis for radiocaesium [17]. This kind of measurement equipment is not sensitive enough to detect contamination in the case of food imports from Ukraine, Japan, and other contami‐ nated countries. Therefore, one cannot save the counting on a γ-spectrometer.

### **4.2. γ-Spectrometry**

Be),

Uranium (235U, 238U) and thorium (232Th) belong to them. Secondary radioactive elements are built by the decay of these primordial radionuclides and belong to the inventory of our soils. Some dose-relevant radionuclides of these decay chains are radium (224Ra, 226Ra, and 228Ra), radon (220Rn and 222Rn), lead (210Pb), polonium (210Po), and thorium (228Th, 230Th, and 232Th). Another ubiquitous radionuclide is potassium-40 (40K), Which is a large part to the annual received dose. Cosmogenically produced radionuclides are, for example, beryllium-7 (7

Artificial radionuclides have been produced and released to the environment since the 1940s, when nuclear weapons development and tests began in the United States. These tests of nuclear fission in the United States, USSR, China, GB, and France led to fallout with a great number of radioactive fission products and activated radionuclides. Over 600 atmospheric bomb tests from 1945 to 1963 led to a contamination of the Northern Hemisphere with artificial radionu‐

(239Pu). [10]. It is because of scientists, such as Ernest Sternglass, who investigated and proved the negative effects of the bomb fallout on childhood mortality. Their warnings helped to enact

clearly decreased. One can clearly see this in sediment profiles of lakes (e.g., lake sediment

NPPs are also emitters of artificial radionuclides. Several accidents caused the release of large quantities of radioactive fallout to the environment. Accidents were the burn of one pile at the Windscale reactor in Sellafield in 1957, the partial core melting of the Three Mile Island reactor in Harrisburg in 1979, the nuclear catastrophe of the Chernobyl NPP in 1986, and the core meltings of the NPPs of Fukushima-Daiji in Japan 2011. The catastrophe of Chernobyl affected

The use of radionuclides in diagnoses and therapies against cancer leads to the release of shortlived radionuclides, such as iodine (131I), technetium (99mTc), indium (111In), lutetium (177Lu), yttrium (90Y), and others. They do not enter the food chain because of their short half-lives and

The following overview of radioactivity laboratory equipment for food control is only a brief summary of the required equipment. For information about the theory of radiation detection and measurements, technical details, or special applications, we refer to the standard literature

Measurement devices to count the dose rate of radioactive sources are not sensitive enough for the precise contamination measurements of food. However, they can be adequate for

Treaty banning of nuclear weapon tests in the atmosphere, in outer space, and under water. 5 August 1963.

many European countries, many thousands of miles away from the NPP ground.

H, radiocaesium (134Cs and 137Cs), radiostrontium (90Sr), and plutonium

which most nuclear powers ratified [11]. After 1965, the fallout

tritium (3

134 Radiation Effects in Materials

clides such as 14C, 3

the partial test ban treaty,1

investigations in Switzerland) [12].

are therefore of minor concern.

**4. Radiation detection**

**4.1 Dose rate monitors**

[13–16].

1

H), or radiocarbon (14C) [8, 9].

It is of primary importance to be equipped with a γ-ray detector. γ-Spectrometry can be operated with inorganic scintillators, such as pure crystals of sodium iodide (NaI) and caesium iodide (CsI) or doped crystals, such as NaI(Tl) or CsI(Tl). NaI(Tl) is the most used scintillator and has excellent light yield and a good linear response, but energy resolution is quite limited. Today, the best choice is semiconductor detectors. Crystals of silicon, germanium (Ge), cadmium-telluride, and others are the detector materials. Ge detectors have been widespread since 1980, when the production of high-purity Ge monocrystals became possible. They have to be kept under vacuum and cooled with liquid nitrogen. The detectors have an excellent energy resolution. Therefore, even complex spectra can be analysed without prior chemical separation steps. Two criteria are of importance: resolution and sensitivity. Energy resolution of Ge detectors is excellent: 1.5 to 2.5 keV at 1.33 MeV, below 100 keV, less than 1 keV FWHM (full width at half maximum). The efficiency of the detector (relative efficiency compared to a NaI detector at 1.33 MeV) and the relation of the peak to Compton background are the most important factors for the sensitivity. Today, Ge detectors with efficiencies of 25% and higher are available. The peak/Compton quotient is more than 46. In our laboratory, we use Ge detectors with 50% efficiency. Important factors for quantitative γ-spectrometry are the shielding of the detector and the efficient suppression of the electronic noise of the amplifier system. γ-Spectrometry has the advantage that γ-radiation can be measured without the elimination of the matrix. Therefore, sample preparation takes only a short time. γ-Spectrom‐ eters need an exact calibration over the whole energy range (e.g., 50–2000 keV). Calibrations with certified radioactive sources are necessary for every counting geometry used (volume and shape of the sample, and distance from the detector) Calibration solutions consist of a mix of short-lived radionuclide with emission lines over the whole energy range (e.g., 109Cd, 57Co, 113Sn, 137Cs, 88Y, and 60Co). After 1 year, the short-lived radionuclides are partly disintegrated; therefore, the calibration mix cannot be used anymore. This can be overcome using mixtures of 152Eu (half-life of 13.5 years) in combination with a low-energy γ-nuclide, such as 210Pb or 241Am. When using 152Eu, summation effects have to be corrected properly.

There are other important factors to consider besides the counting geometry. The sample matrix itself absorbs γ-rays before they arrive at the detector. These self-absorption effects are a function of the sample geometry and the elemental composition and the density of the sample matrix. Normally, calibrations and efficiency curves are produced with calibration standards in water or gels of density 1. Density corrections can be calculated for every material and geometry with means of a software based on Monte Carlo simulations. The background radiation has to be considered. Some background radiation remains, even with good shielding with lead and copper. This background consists of radionuclides of the natural decay series, such as 214Pb and 214Bi and others. For every counting geometry, the background has to be measured with water-filled containers of the needed counting geometries. The γ-spectra have to be subtracted by the specific background spectrum. Radionuclides with cascade emissions show coincidence summing effects (e.g., 134Cs and 152Eu). Spectra have to be corrected or the measurement must be repeated with a distance between sample and detector. For short-lived radionuclides, their partial decay has to be corrected to the reference date (e.g., the date of the sampling). Further advice and descriptions over quantitative γ-spectrometry are given in the literature [18].

### **4.3. β-Spectrometry**

For the counting of β-rays, the sample matrix has to be eliminated. An exception is water samples (e.g., the measurement of 3 H needs not much sample preparation, only the mix of the water sample with a scintillator cocktail). Some important β-nuclides, such as 3 H, 14C, 89Sr, 90Sr, and 90Y, can be analysed with scintillation counting. Commercially available scintillation counters can detect α- and β-decays. This widens the spectrum of radionuclides (e.g., 222Rn can be analysed in water samples or in charcoal air samples). When samples are analysed directly, the sensitivity is given by the small sample amount of typically some millilitres.

Radiostrontium, 89Sr and 90Sr, are important fission products. One possibility is to extract the 90Sr with the use of specific crown ethers from the sample. In our laboratory, we have developed a fast analysis scheme for water samples [19]. Another possibility is to clean up extracts over a column filled with crown ethers. These methods are suitable for activity concentrations higher than 1 Bq/kg. For sensitive analyses, the β-spectrometers of choice are gas flow proportional counters. We use this technique for the analyses of 90Sr traces in food, human, and environmental samples. The method is based on the counting of the daughter nuclide, 90Y. Before the counting starts, a rigorous elimination of the matrix and disturbing β-nuclides, such as 40K, is necessary. With an oxalate precipitation step, most of the 40K is eliminated. Then, 90Y is separated from 90Sr by precipitation as hydroxide. The Y(OH)3 is dissolved and precipi‐ tated as Y2(oxalate)3. These β-sources are pure enough for counting. Counting is performed in 10 consecutive runs, as 90Y decays during the counting (half-life is 64 h). A good-quality criterion for the purity is the measured decay of the source. Decay should be near 64 h. When decay is slower, impurities are present. The conserved 90Sr solution may be prepared and analysed again after 20 days (the built-up 90Y will then arrive equilibrium with 90Sr). Quite sensitive analyses may be performed down to 10 mBq/kg. Counting time is 3 days. Therefore, several detectors should be available. Our Canberra α/β-counter LB 4100 can take up to four drawers with four sample holders each [20].

#### **4.4. α-Spectrometry**

Like β-spectrometry, α-spectrometry requires an elimination of the sample matrix. Only water samples need a minimal preparation. Two counting techniques are common today: scintilla‐ tion counting and passivated implanted planar silicon (PIPS) detectors. We use liquid scintil‐ lation counters in our laboratory for the analyses of uranium, thorium, radium, and polonium. According to the methods published by W. Jack McDowell, the water sample is extracted once with some millilitres of a nuclide selective extractant, which contains the scintillator cocktail for the α-analysis. Very low activity concentrations (5–10 mBq/L) can be achieved [21]. The disadvantage is the poor energy resolution. Also, α/β-discrimination has to be set carefully. Analysis time is 24 hours for water samples containing 100 mBq.

Analyses with PIPS detectors show good energy resolution and efficiencies from 20% to 30%. Counting has to be performed under vacuum and very thin layer α-sources are needed. The sample matrix has to be destroyed (by ashing, etc.) following a clean-up (specific nuclide extraction, scavenging, etc.) and a preparation of a thin-layer source by means of electroplat‐ ing or coprecipitation. Many actinides, such as uranium, thorium, and plutonium, may be analysed (well described in [22]). Some elements such as 210Po and radium nuclides may be auto-deposited onto special surfaces. 210Po is auto-deposited under reductive conditions onto silver or copper disks after a microwave digestion of the sample [23, 24]. Radium nuclides ( 224Ra and 226Ra) are auto-deposited onto MnO2 surfaces at pH 8 as has been shown by Surbeck [25, 26]. These methods are suitable for drinking water and mineral water analyses. Some new developments were done for uranium and thorium analyses in honey and spices [27, 28] (**Figure 1**).

**Figure 1. Equipment of a radioactivity laboratory:** (A) γ-spectrometry; (B) α-liquid scintillation (PERALS); (C) α-Spec‐ trometry (PIPS detectors); and (D) β-spectrometry (gas proportional counter), drawer with 90Y oxalate sources.

### **4.5. Neutron activation analysis (NAA)**

with lead and copper. This background consists of radionuclides of the natural decay series, such as 214Pb and 214Bi and others. For every counting geometry, the background has to be measured with water-filled containers of the needed counting geometries. The γ-spectra have to be subtracted by the specific background spectrum. Radionuclides with cascade emissions show coincidence summing effects (e.g., 134Cs and 152Eu). Spectra have to be corrected or the measurement must be repeated with a distance between sample and detector. For short-lived radionuclides, their partial decay has to be corrected to the reference date (e.g., the date of the sampling). Further advice and descriptions over quantitative γ-spectrometry are given in the

For the counting of β-rays, the sample matrix has to be eliminated. An exception is water

and 90Y, can be analysed with scintillation counting. Commercially available scintillation counters can detect α- and β-decays. This widens the spectrum of radionuclides (e.g., 222Rn can be analysed in water samples or in charcoal air samples). When samples are analysed directly,

Radiostrontium, 89Sr and 90Sr, are important fission products. One possibility is to extract the 90Sr with the use of specific crown ethers from the sample. In our laboratory, we have developed a fast analysis scheme for water samples [19]. Another possibility is to clean up extracts over a column filled with crown ethers. These methods are suitable for activity concentrations higher than 1 Bq/kg. For sensitive analyses, the β-spectrometers of choice are gas flow proportional counters. We use this technique for the analyses of 90Sr traces in food, human, and environmental samples. The method is based on the counting of the daughter nuclide, 90Y. Before the counting starts, a rigorous elimination of the matrix and disturbing β-nuclides, such as 40K, is necessary. With an oxalate precipitation step, most of the 40K is eliminated. Then, 90Y is separated from 90Sr by precipitation as hydroxide. The Y(OH)3 is dissolved and precipi‐ tated as Y2(oxalate)3. These β-sources are pure enough for counting. Counting is performed in 10 consecutive runs, as 90Y decays during the counting (half-life is 64 h). A good-quality criterion for the purity is the measured decay of the source. Decay should be near 64 h. When decay is slower, impurities are present. The conserved 90Sr solution may be prepared and analysed again after 20 days (the built-up 90Y will then arrive equilibrium with 90Sr). Quite sensitive analyses may be performed down to 10 mBq/kg. Counting time is 3 days. Therefore, several detectors should be available. Our Canberra α/β-counter LB 4100 can take up to four

Like β-spectrometry, α-spectrometry requires an elimination of the sample matrix. Only water samples need a minimal preparation. Two counting techniques are common today: scintilla‐ tion counting and passivated implanted planar silicon (PIPS) detectors. We use liquid scintil‐ lation counters in our laboratory for the analyses of uranium, thorium, radium, and polonium.

water sample with a scintillator cocktail). Some important β-nuclides, such as 3

the sensitivity is given by the small sample amount of typically some millilitres.

H needs not much sample preparation, only the mix of the

H, 14C, 89Sr, 90Sr,

literature [18].

136 Radiation Effects in Materials

**4.3. β-Spectrometry**

samples (e.g., the measurement of 3

drawers with four sample holders each [20].

**4.4. α-Spectrometry**

INAA (instrumental neutron activation analysis) is based on the production of radionuclides by nuclear reactions. Thermal neutrons can activate many elements. The efficiency of this irradiation process depends on the flux density of the neutrons and the cross-section of the nuclear reaction. The thermal neutrons are generated in a nuclear reactor, for example, at the University of Basel (AGN-211-P, a light water moderated swimming pool reactor).

The samples (1–5 g material) are inserted into the core over a cannula through the so-called glory hole and irradiated for 30 min with a power rate of 2 kW. After a cooling time of some hours, the samples are counted on an HPGe detector.

We used INAA for the analysis of total bromine content. Bromide is built by the decay of the fumigant methyl bromide. We used this technique for many years to determine the total bromine in spices, tea, and dried mushrooms. Another application is the determination of the total bromine content as a screening analysis for flame-retardants in plastic materials [29]. 238U and 232Th can be determined by INAA in suspended matter and sediments [30]. In addition, total iodine content in iodine rich food, such as algae, can be determined over the activation of 127I to 128I, which decays to 128Xe [31].

We mentioned the use of INAA as a completion of the possibilities of γ-spectrometry. These applications will not be discussed further, because the analytes are not radionuclides.
