RU2120646C1 - Process of identification of radionuclides in sample with use of liquid scintillation counter - Google Patents

Abstract

FIELD: environmental control. SUBSTANCE: process includes taking samples of environment and of technological samples, preparation of samples for measurement on liquid scintillation counter, measurement and recording of sample spectrum and measurement parameters, convolution of sample spectrum into groups, formation of model spectrum of sample on basis of library of basic spectra, minimization of deviation of model spectrum from sample spectrum and identification of content of radionuclides in sample. In this case sample spectrum is convoluted into groups which boundary values Ni are quasi- arithmetic progression of type of N_i+1= N_i+[(i+1)/2] where I=1, 2,... n and minimization of deviation of model spectrum from sample spectrum is carried out by composition of model spectrum of definite type. Difference between model spectrum and sample spectrum is minimized to determine relative contributions of spectra of radionuclides into model spectrum of sample. Minimization is carried out in several stages selecting first of all bigger value δ and zero values c_j and determining minimum of expression. After this δ is decreased and minimization process is repeated . In this case starting values of contributions c_j for each next step of minimization are results of preceding step. Calculation is stopped when process of minimization of expression turns to be unstable and identification of radionuclides in examined sample is completed by recalculation of obtained values c_j in terms of absolute activities of radionuclides in sample. Technical result of invention consists in possibility of isolation of radionuclides with low levels of activity against background of radionuclides having higher activities by selection of proper scale of convolution of spectra into groups and in determination of criterion of minimization of sample spectrum from model spectrum. EFFECT: improved efficiency of process. 4 dwg, 4 tbl

Description

 The present invention is used when conducting work in the field of radioecological monitoring to measure the content of radionuclides in various components of the environment, when processing the results of measurements in a complex of hardware and software, allowing to operate with large amounts of radioecological information. In addition, the proposed method is used in the accumulation, storage, updating and transmission of radioecological data, followed by mathematical processing using a computer display of the results on an electronic cartographic basis. In particular, the method is used to determine the activity of beta-emitting radionuclides, such as H-3, C-14, Ni-63, Sr-90, Cs-137, Co-60, Fe-55, in prepared liquid samples using hardware processing spectra in the range of activities from 0.5 Bq / sample using a liquid scintillation counter.

 Known methods for measuring the activity of samples containing several radionuclides that use a liquid scintillation counter. For the estimated N radionuclides, at least N + 1 windows of the pulse height analyzer are selected and the count in each window is measured. Next, for the estimated level of extinguishing, the previously obtained counting efficiencies of individual radionuclides for each window are selected and N + 1 equations are compiled, from which, using the least-squares method, the desired activity values of individual radionuclides are obtained. Then a different level of quenching is chosen, and the activity values are recalculated for it. This cycle is repeated until a minimum deviation of the sum of these activities from the measured activity is achieved [1].

 In addition, a method is known in which the spectrum of the test sample is measured using a multichannel amplitude analyzer connected to an analog-to-digital converter, its quenching level is determined, the normalized spectra of individual radionuclides for a given quenching level are calculated, and, using the least squares method, they are determined factors, it is necessary to multiply the spectra of single samples in order to obtain the total spectrum closest to the studied one. These factors are proportional to the desired radionuclide content in the sample [2].

где
i - номер канала анализатора;
j - индекс радионуклида [3].The closest to the proposed method according to the technical essence and the achieved effect is a method for identifying radionuclides in a liquid scintillation sample, in which the spectrum of the test sample is measured, after which, for the appropriate level of extinction from the library of basic spectra of individual radionuclides for various levels of extinction, normalized model spectra of individual radionuclides. Next, the difference between the spectrum of sample P _i and the sum of model spectra of individual radionuclides M _ij multiplied by coefficients c _j determining the activity of individual radionuclides is minimized by the least square method. The minimized expression in this case is as follows:

Where
i is the channel number of the analyzer;
j is the radionuclide index [3].

However, the known methods do not allow the identification of radionuclides having low activity in the presence of radionuclides with high activity in the sample. This is due to the following: due to the statistical nature of the quantities P _i squared deviation of each term in the sum (1) has a statistical variation values of order ≈ P _i. Accordingly, by minimizing the expression to be minimized F will affect only terms with a large value of P _i analyzer account and terms with a small value of P _i will be "absorbed" due to the above dispersion. This leads to the fact that the identification of radionuclides is carried out almost exclusively on the "tops" of the spectra, not taking into account the additional information that weakly active areas can give.

 The aim of the proposed method is to increase the sensitivity of liquid scintillation analysis of radionuclides having low activity, in the presence of radionuclides in the samples having high activity.

где
M_ij - модельные спектры радионуклидов;
c_j - относительные вклады спектров радионуклидов в модельный спектр пробы;
i - номер группы;
j - индекс радионуклида.To solve this problem, a method for identifying radionuclides in samples using a liquid scintillation counter is proposed, including sampling the environment and process samples, preparing samples for measurement on a liquid scintillation counter, measuring and recording the spectrum of the sample and measurement parameters, folding the hardware spectrum of the sample into groups, creating a model spectrum of a sample based on a library of basic spectra, minimizing the deviation of the model spectrum from the spectrum of the sample, and identifying the content Ania radionuclides in the sample. In this case, the hardware spectrum of the sample is collapsed into groups whose boundary values N _i in which are a quasi-arithmetic progression of the form N _i + 1 = N _i + [(i + 1) / 2], where i = 1, 2, ... n (1 ) The deviation of the model spectrum from the spectrum of the sample is minimized by compiling a model spectrum

Where
M _ij - model spectra of radionuclides;
c _j are the relative contributions of the radionuclide spectra to the model spectrum of the sample;
i is the group number;
j is the radionuclide index.

где
P_i - спектр пробы;
M_i - модельный спектр;
δ - коэффициент устойчивости процесса минимизации.To determine the relative contributions of the radionuclide spectra to the model spectrum of the sample, the difference between the model spectrum and the spectrum of the sample is minimized by the expression

Where
P _i is the spectrum of the sample;
M _i - model spectrum;
δ is the stability coefficient of the minimization process.

Minimization is carried out in several stages, first choosing a large value of δ and zero values of c _j , determine the minimum of expression (3), then reduce δ by repeating the minimization process, while the initial values of the contributions c _j for each next minimization step are the results of the previous step. The calculation is stopped when the process of minimizing expression (3) becomes unstable. The identification of radionuclides in the test sample is completed by recalculating the obtained values of c _j into the values of the absolute activities of the radionuclides in the sample.

 In FIG. 1 is a flow chart of a method for identifying radionuclides in samples using a liquid scintillation counter; in FIG. 2 shows beta spectra of radionuclides at various scales; in FIG. 3 - a fragment of the base spectra library; in FIG. 4 shows a comparison of the results of processing a test sample by the method of the prototype (A) and the proposed method (B).

 The method is as follows. The sequence of operations is shown in FIG. one.

 I. Sampling and preparation of samples for measurement.

 Spectrometric analysis based on liquid scintillation counting can be performed for any samples (environment and technological) after their appropriate preparation [4]. As a rule, these are water samples, soil samples, vegetation and air samples.

 I.1. Sample selection.

 Environmental water samples, as well as technological samples, are taken in clean containers of 1 liter in an amount of at least 3 pieces per controlled object, acidified to pH 3-4, the containers are sealed, indicate the name of the sample, time and place of sampling and delivered to the lab.

 Soil samples are taken from a plot of 10 • 10 m in size. The sampling area is marked with an envelope, samples are taken from envelope nodes (5 points) in a monolithic piece of 10 • 10 cm, 5 cm thick, from which grass cover is previously removed. The sample is cleaned of the roots and packaged in a plastic bag with the name of the sample, place of sampling, date of sampling, mass of the sample, dose rate at the place of sampling.

Vegetation samples for grass samples are taken from elementary sites with an area of 0.25 to 1 m ² depending on the density of the grass cover. A mixed sample is composed of 4-5 individual samples, and aerial parts of plants are taken.

 Samples of woody vegetation are made by combining several individual samples from various tree species in proportion to their share in the stand. As a sample, the shoot of the last annual growth (conifers) or foliage from one branch from the lower part of the crown is taken.

 The selected vegetation samples are packed in a plastic bag with the name of the sample, the place of sampling, the date of sampling, the mass of the sample and delivered to the laboratory.

Air samples for determining the content of radioactive aerosols are taken using suction units with a capacity of at least 1000 m ³ / h on a fine fiber Petryanov filter. After pumping the required volume of air, the filter is removed and packaged in clean, thick paper, the name of the sample, the place and date of sampling, the volume of pumped air are indicated and delivered to the laboratory.

 I.2. Sample preparation for measurement.

 I.2.1. Preparation of water samples.

 I.2.1.1. Technological tests.

If the estimated volumetric activity of the radionuclide in the sample is more than 10 ² Bq / l, then the sample is not concentrated, but analyzed immediately. To do this, 5 ml of each of the three parallel samples are taken, the pH of the solution is checked and, if necessary, adjusted to neutral (with concentrated hydrochloric acid or ammonia), then neutral solutions are added to the bottles for liquid scintillation counting, pre-filled with 10 ml of scintillation cocktail. Vigorously mix the contents of each vial to obtain a homogeneous solution, keep the vials in the dark for at least 12 hours (to attenuate possible chemo and photoluminescence processes), and then install them in a liquid scintillation analyzer for further measurements. Simultaneously with the analyzed sample, a background is prepared by analogy, into which the same volume of deionized water is added as the prepared sample, but not more than 10 ml.

 I.2.1.2. Low active samples.

Samples with low volumetric activity of radionuclides (<10 ² Bq / L) must be concentrated before analysis. For this, three parallel samples of 1 liter each are taken, if necessary, acidified with concentrated hydrochloric acid to pH 3-4 and evaporated on an electric stove to a volume of 50 - 70 ml, transferred to a heat-resistant glass with a volume of 0.1 liter and evaporated to a volume of 5 - 8 ml Next, the solution is cooled, neutralized with a concentrated solution of ammonia to pH 5-7 (the volume of the prepared sample should not exceed 10 ml) and put into vials for liquid scintillation counting, pre-filled with 10 ml of scintillation cocktail. The contents of the vials are thoroughly mixed and kept in the dark for at least 12 hours before measurement. Simultaneously with the analyzed sample, a background is prepared by analogy, into which the same volume of deionized water is added as the prepared sample.

 When concentrating the samples by evaporation, some radionuclides, for example, carbon-14, sulfur-35, which are in water in the form of dissolved gases or volatile organic compounds, can be lost, therefore their presence and quantity must be estimated by measuring 10 ml of the non-concentrated sample on a liquid scintillation analyzer.

 I.2.2. Preparation of soil samples.

 I.2.2.1. Preparation of soil samples using acid decomposition (leaching).

The selected soil sample weighing 1 kg is dried for 6 to 8 hours at a temperature of 60 ^o C in an oven. Depending on the type of sample (sand, clay) and its hardness, it is either pre-crushed in a porcelain mortar, or immediately dispersed on a vibrating table with a set of sieves of various diameters. The smallest fraction (0.5 mm or less) is divided by quartering and a 20 g sample is taken for analysis. A sample is transferred to a heat-resistant glass beaker with a volume of 150-200 ml, 40 ml of concentrated nitric acid are added, 2-4 ml of 30% hydrogen peroxide the mixture is stirred, covered with a watch glass and heated on an electric stove for 1 hour. Then the mixture is filtered on a funnel with a blue ribbon filter, the residue with the filter is transferred to the same glass, the same volumes of nitric acid and hydrogen peroxide are added and the process is repeated .

 The filter cake is washed with 20 ml of 6N nitric acid, all three acid extracts are combined. The extract is evaporated to wet salts, 20 ml of hot distilled water are added and evaporated to wet salts. The minimum volume of a solution of hot 1N hydrochloric acid is added to the residue until the precipitate is completely dissolved. As a rule, depending on the composition of the sample and the content of iron in it, the color of the sample changes from light yellow to dark brown and is the reason for the different color haze of the sample. If the volume of the solution does not exceed 5 ml, then you can take half of the sample or an aliquot, add 10 ml of scintillation cocktail into a glass bottle for liquid scintillation counting, and preliminarily estimate the degree of quenching of the sample by measurement on a liquid scintillation analyzer for 1 - 2 min. If the quenching parameter is at least 100 in magnitude, then the prepared sample can be measured, and the remainder of the sample can be added to the same bottle. If the quenching is less than 100, which is most likely due to the color quenching of the sample due to the presence of ferric chloride, then the sample must be clarified by adding an equal volume of a saturated solution of ferrous tin chloride to restore ferric iron to ferrous, the compounds of which have a less intense color. Next, the sample is analyzed as described above.

 I.2.2.2. Preparation of soil samples using microwave technology.

 The finely dispersed fraction of the quartered sample is crushed in a planetary micromill to a size of 0.1 mm or less, a sample of 5-10 g is taken and introduced into the fluoroplastic glasses of the microwave rotor. 10 ml of a mixture of concentrated nitric acid and 30% hydrogen peroxide in a volume ratio of 9: 1 are also added to each of the glasses. Glasses are closed with special covers with bypass valves, placed in the rotor and installed in the microwave. The sample decomposition process is carried out in the following modes: 5 min at 250 W of pulsed microwave radiation, 10 min - 400 W and 5 min - 500 W. Due to the uniform microwave heating of the sample in a closed volume, high temperature and pressure are created in the reaction medium, which make it possible to reduce the opening time of the sample by an order of magnitude and, in most cases, facilitate its complete decomposition.

 After the process is completed and the sample is cooled, it is filtered (in case of incomplete decomposition), the filter residue is washed 2 times with 5 ml of 5 M nitric acid, the acid fractions are combined and prepared for liquid scintillation counting as described above.

 I.2.3. Preparation of vegetation samples.

Vegetation samples are dried for 6 to 8 hours at a temperature of 60 ^o C in an oven. The dried samples are crushed using a mill, weighed, placed in porcelain crucibles with lids and ashed in a muffle furnace for 6 - 7 hours at 400 - 450 ^o C. After completion of ashing and cooling of the crucibles, they fill 10 - 20 ml of concentrated nitric acid ( depending on the amount of ash obtained) and carry out acid leaching and preparation for liquid scintillation counting as described for soil samples.

 I.2.4. Air sample preparation.

Filters consisting of Petryanov’s fabric are crushed with scissors, placed in porcelain crucibles with lids and ashed in a muffle furnace for several hours at 400-450 ^° C. After completion of the ashing and cooling of the crucibles, they fill them with 10-20 ml of concentrated nitric acid (in depending on the amount of ash obtained) and carry out acid leaching and preparation for liquid scintillation counting as described for soil samples.

 I.2.5. Preparation of background samples.

 When analyzing water, soil, vegetation, and air samples, samples prepared by mixing in vials for liquid scintillation counting (volume 20 ml) of 10 ml scintillation cocktail with a volume of 1 N hydrochloric acid (or its mixture with a solution of tin chloride) equal to the volume are used as background samples. prepared and entered into the measuring vial of the sample.

 II. Measurement and recording of the spectrum of the sample.

 When measuring the spectrum of the sample, the following operations are performed.

 A background sample and prepared samples are placed in a special cassette. The cartridge is installed in the analyzer. On the left, insert a clip with a code corresponding to the sample measurement program in the mode of counting recorded pulses in min. The measurement program sets the measurement time in the energy range from 0 to 2000 KeV. After that, the analyzer is transferred to the counting mode, in which all the sample bottles placed on the cassette are automatically calculated. All parameters specified in the program are recorded, namely: sample index, measurement time, count, error, quenching parameter, and the resulting spectrum of the sample is saved in the form of a file with the distribution of the number of recorded pulses over the energy channels of the multichannel analyzer. After that, according to a specially developed program, the resulting energy spectrum of the sample is analyzed and the qualitative composition of the sample and the volumetric (specific) activity of its components are determined.

 III. Collapsing the hardware spectrum of the sample into groups.

 Due to the irrationality of using the scintillation beta spectrometer in the initial hardware spectrum in calculations, it is advisable to collapse this spectrum into groups, while losing a minimum of information.

 Using a uniform linear partition is irrational, because in this case, the most informative initial part of the hardware spectrum is collapsed into a small number of groups and a significant amount of information is lost for most spectra. In the upper graph of FIG. 2 shows graphs of beta spectra of some radionuclides on a linear scale. It can be seen from the graph that the “soft” spectra (in this case, the tritium spectrum — H-3) practically merge at the origin and, when minimized, fall into groups, which, of course, does not allow us to judge their shape.

 Using a logarithmic partition (see the average graph) is also impossible, since the low-energy part of the spectra is stretched over most of the range, compressing the high-energy part of the spectra in the last groups, which will also lead to loss of information.

In this method, folding of the hardware spectrum of the sample into groups, the boundary values of which are presented in table. 1 and are a quasi-arithmetic progression of the form
Ni + 1 = Ni + [(i + 1) / 2],
Where
i = 1, 2, ... n
(the square brackets here indicate the integer part)
In the table. Figure 1 shows the folding of the hardware spectrum into 90 groups for 2115 analyzer channels. The number of groups is selected depending on the detail of the hardware spectrum.

 Such a partition (see the graph on the scale of the groups in Fig. 2) allows you to detail, without losing information, to distinguish low-energy radionuclide spectra. In this case, when moving to high-energy regions, the difference in the width of the groups is leveled, so that the detail of the representation of the high-energy region in the proposed group scale becomes the same as with the linear scale.

 IV. Creating a library of basic spectra.

 To create a library of basic spectra do the following. For each of the control sources (tritium, carbon-14, nickel-63, strontium-90, etc.), a series of 10 bottles with a known equal amount of radioactive label is prepared. Starting from the second one, an increasing amount of a chemical quencher is introduced into each bottle, after which the entire series is filtered through a liquid scintillation analyzer with recording of all spectra in a nuclide library, with each value of the quenching parameter corresponding to a certain radionuclide emission detection efficiency. In FIG. Figure 3 shows the energy spectra of some radionuclides with various chemical quenching. With increasing quenching in the sample (the quenching value decreases in this case), the spectral curve recorded by the analyzer is located in the region of lower energies, and the recording efficiency, as a rule, is greatly reduced.

 Into the library are introduced recalculated into the group spectra of individual radionuclides for 10 different quenches.

 The library also includes registration efficiency values for each quenching.

 All library spectra are normalized to a single activity.

 The resulting library of spectra of the most common nuclides and quenching curves allows us to analyze real samples in a wide range of quenching for this brand of scintillation cocktail.

 V. Recalculation of the basic spectra for the quenching of the sample.

 The library provides basic spectra for the final quenching selection. Quenching of the test sample may not coincide with the library. Therefore, when processing samples with a certain quenching, the basic spectra lead to this quenching by interpolation between library spectra. For this quenching, the registration efficiency for each radionuclide is also interpolated. The resulting spectra are normalized and used further to create a model spectrum.

 VI. Minimization of deviation of the model spectrum from the spectrum of the sample.

где
M_ij - модельные спектры отдельных радионуклидов;
c_j - относительные вклады в модельный спектр этих радионуклидов;
i - номер группы;
j - индекс радионуклида.The next stage is the model range

Where
M _ij - model spectra of individual radionuclides;
c _j are the relative contributions to the model spectrum of these radionuclides;
i is the group number;
j is the radionuclide index.

To find the contributions c _j, the difference between the model spectrum and the spectrum of the sample P _{i is} minimized.

 When solving minimization problems, the most important point is the correct choice of minimization criterion.

 Using minimization of the absolute deviation of the spectrum of the sample from the spectrum of the model (see formula (1)), as is done in the prototype, does not allow you to "feel" small activity against the background of large ones.

где
P_i спектр пробы;
M_i - модельный спектр;
δ - - коэффициент устойчивости процесса минимизации.In order for the method to allow radionuclides with low activity to be isolated in samples against the background of radionuclides with high activity, it was proposed to minimize the following expression:

Where
P _{i the} spectrum of the sample;
M _i - model spectrum;
δ - is the coefficient of stability of the minimization process.

Here, dividing by min (P _i , M _i ) (transition to relative deviation), and not just by P _i , is used to increase the sensitivity to the "low-active" regions of the spectrum, since this makes it possible to further increase the role of terms with small values of M _i .

The stability coefficient δ is a small additive necessary to eliminate division by zero at min (P _i , M _i ) = 0 and to eliminate instability of solution (4) due to terms in which the term min (P _i , M _i ) is less than the statistical spread values of P _i .

The choice of the coefficient δ simultaneously affects both the stability of the solution (the more δ, the better the stability), and the sensitivity of the method to inactive elements (the more δ, the worse the sensitivity). In the extreme case, as δ _ → ∞, the minimized expression reduces to the form (1) used in the prototype. In the calculation, first choose a large value of the coefficient δ, find the minimum of expression (4), after which δ reduce and repeat the minimization process. In this case, the initial values of the contributions c _j for each next step are the results of the previous step. The calculation is stopped when the process of minimizing expression (4) becomes unstable.

 VII. Calculation of the activity of radionuclides.

The obtained values of c _j determine the relative contribution of radionuclides to the activity of the sample. The transition to absolute activity is carried out according to the formula
A _j = c _j • A / E _j , (5)
Where
A is the integral count of the sample (the sum of the hardware spectrum).

E _j is the registration efficiency.

 Thus, as a result of measuring and processing the spectrum of the sample, the radionuclide composition was identified and the activities of each radionuclide included in the sample were obtained.

 The technical efficiency of the proposed method compared to the prototype is that this method allows to isolate in the samples against the background of radionuclides with high activity, radionuclides with low levels of activity by choosing the appropriate scale for folding the hardware spectra into groups and determining the criterion for minimizing the deviation of the spectrum of the sample from the model spectrum.

 To confirm technical efficiency, we analyze the measurement data of 2 test samples. In the table. 2 and 3 show the real compositions of these samples and the results of determining the radionuclide composition by the method of the prototype and by the method presented.

 From the table. 2 shows that in the first sample, using the proposed method, the entire radionuclide composition is determined quite accurately, while in a known manner it is possible to determine only highly active Cs-137 and one element out of 3 weakly active. From the table. 3 it can be seen that in the second sample, using the proposed method, the entire radionuclide composition was also quite accurately determined, while the known method only allowed accurate determination of tritium, the activity of which is significantly higher than the activity of related elements. The results of determining the radionuclide composition of sample No. 2 are well illustrated by the graphs in FIG. 4. They show the spectrum of the sample, the model spectrum, the contributions to the model spectrum of the spectra of individual radionuclides and the deviation of the model spectrum from the spectrum of the sample. The upper part shows the full spectra, and the middle and lower ones show the lower part of the upper graph, enlarged by 35 and 1750 times, respectively. On the left graphs (Fig. 4A) - the results of the known method, on the right (Fig. 4B) - according to the proposed method, which shows that the model spectrum obtained by the known method only in the upper part coincides with the spectrum of the sample, while time as the model spectrum obtained by the proposed method is close to the spectrum of the sample along the entire axis.

 Example.

Water technological samples are taken from one of the radioactive waste disposal facilities (RWDFs) in Russia. This sample has an activity of about 10 ⁶ Bq / L, due to the presence of a large amount of tritium, but it is necessary to determine other weakly active radionuclides in it, which in their radiation hazard exceed tritium by several orders of magnitude.

 This sample is prepared as follows: three parallel samples are taken, the pH of the solution is adjusted to neutral, the resulting solutions are added to liquid scintillation counting vials pre-filled with 10 ml of scintillation cocktail. The resulting homogeneous solution is kept in the dark for 12 hours to attenuate possible processes of chemo and photoluminescence. Then the bottles are installed in a liquid scintillation analyzer for measurements. Measure the spectrum of the sample and record the measurement parameters, the measurement time is 60 minutes, the quenching is 270, the sample is 5 ml.

 The following operations are carried out to determine these weakly active radionuclides against the background of highly active tritium.

 Since the measured hardware spectrum ends in channel 1280, it is collapsed into 70 groups. This folding allows you to detail, without loss of information, to highlight the spectra of radionuclides with low energies. In this case, when moving to high-energy regions, the difference in the width of the groups is leveled, so that the detail of the representation of the high-energy region in the proposed group scale becomes the same as with the linear scale.

для чего выбирают набор радиоизотопов, характерных для данного пункта захоронения. Здесь c_j - относительные вклады в модельный спектр радионуклидов, i - номер группы, j - индекс радионуклида. Далее, установив δ = 1 и c_j = 0 для всех j, минимизируют следующее выражение

Далее уменьшают δ в 10 раз и заново проводят минимизацию, используя в качестве начальных значений вкладов c_j полученные на предыдущем шаге. Повторяют процедуру до тех пор, пока процесс минимизации не становится неустойчивым (изменение вкладов c_j приводит к уменьшению минимизируемого выражения).A model spectrum is formed from the library of basic spectra for the measured level of quenching

why choose a set of radioisotopes characteristic of a given burial site. Here c _j are the relative contributions to the model spectrum of radionuclides, i is the group number, j is the radionuclide index. Next, setting δ = 1 and c _j = 0 for all j, minimize the following expression

Then, δ is reduced by a factor of 10 and minimization is carried out anew using the initial values of the contributions c _j obtained in the previous step. Repeat the procedure until the minimization process becomes unstable (changing the contributions c _j leads to a decrease in the minimized expression).

The coefficients c _j obtained in this way determine the relative contribution of radionuclides to the activity of the sample. The transition to absolute activity is carried out according to the formula
A _j = c $\genfrac{}{}{0ex}{}{\text{*}}{\text{j}}$ A / E _j ,
Where
A is the integral count of the sample (the sum of the hardware spectrum);
E _j is the registration efficiency.

 As a result of the selection, sample preparation, measurement and processing of the spectra of the sample, the radionuclide composition was identified and the activities of each radionuclide included in the sample were obtained.

 The results of sample processing are presented in table. 4.

 Thus, the proposed method allows to detect small additives of accompanying isotopes in water containing a large amount of tritium. This result is confirmed by reprocessing after evaporation of the sample, in which the tritium content is minimized.

Information Sources Used
1. US patent N 4918310 IPC G 01 T 1/204.

 2. US Patent N 5134294 IPC G 01 T 1/204.

 3. PCT N 91/10922 IPC G 01 T 1/204 - prototype.

 4. "Method for measuring the volumetric activity of beta-emitting radionuclides in aqueous samples using a beta spectrometer based on a liquid scintillation analyzer of the brand" TRI-CARB 2550 TR / AB '' '', MRK-MO-1597, Moscow, 1997.

Claims (1)

A method for identifying radionuclides in samples using a liquid scintillation counter, including environmental and process samples, preparing samples for measurement on a liquid scintillation counter, measuring and recording the spectrum of the sample and measurement parameters, folding the hardware spectrum of the sample into groups, creating a model spectrum of the sample on the basis of the library of basic spectra, minimizing the deviation of the model spectrum from the spectrum of the sample and identification of the content of radionuclides in the sample, different t We note that the hardware spectrum of the sample is collapsed into groups whose boundary values N _i in which are a quasi-arithmetic progression of the form N _{i + 1} = N _i + [(i + 1) / 2], where i = 1, 2, ... n, minimize the deviation of the model spectrum from the spectrum of the sample by compiling a model spectrum

where M _ij - model spectra of radionuclides;
c _j are the relative contributions of the radionuclide spectra to the model spectrum of the sample;
i is the group number;
j is the radionuclide index,
and to determine the relative contributions of the radionuclide spectra to the model spectrum of the sample, the difference between the model spectrum and the spectrum of the sample is minimized by the expression

where P _i is the spectrum of the sample;
M _i - model spectrum;
δ is the stability coefficient of the minimization process,
minimization is carried out in several stages, first choosing a large value of δ and zero values of c _j , determine the minimum, then δ is reduced by repeating the minimization process, while the initial values of the contributions c _j for each next minimization step are the results of the previous step, the calculation is stopped when the minimization process becomes unstable, and the identification of radionuclides in the test sample is completed by recalculating the obtained values of c _j into the values of the absolute activities of the radionuclides in the sample.

Images