RU2478934C2 - Method for elemental analysis of media and apparatus realising said method - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: fast neutron source is placed in a slowing unit; a γ-spectrometer employing a non-overload linear detection technique is used to detect instantaneous γ-quanta resulting from radiation capture of neutrons by nuclei of elements; calibration responses of separate elements making up the sample being identified are determined; apparatus spectra of γ-quanta are used to determine concentration of elements in the sample through weight coefficients of responses of elements, wherein chemical composition of the media is determined by: a priori determining the chemical composition - known chemical compounds - of the medium; determining calibration responses of compounds through the sum of calibration responses of elements making up the compounds; determining concentration of compounds and elements in the medium being identified; determining concentration of elements making up the compounds; establishing conformity of the obtained total concentration of elements of the medium, based on elements making up compounds of the medium being identified, concentration of elements of the medium obtained by identification decryption of only the elemental composition of the medium; in case of mismatch with a given error of the total concentration of elements of the medium, obtained by identification decryption based on elements making up the identified compounds of the medium, with concentration of elements of the medium, obtained by identification decryption of only the elemental composition of the medium, the procedure, starting from the a priori determining of the structure of compounds, is repeated.

EFFECT: faster elemental analysis, high resolution of identification of elements, high sensitivity of determining impurities in media.

3 cl, 13 dwg

Description

The invention relates to the field of elemental analysis - qualitative detection and quantification of the content of elements and elemental composition of substances, materials and various objects. It can be liquids, solid materials, gases. Elemental analysis allows you to answer the question - what atoms (elements) the analyte consists of.

Elemental analysis is one of the most important tasks in any production, the purpose of which is to control the raw materials used, production, as well as finished products. Ferrous and non-ferrous metallurgy, oil production and refining, agriculture, geology, mining and much more are almost impossible without the availability of analytical laboratories that carry out elemental analysis. Elemental analysis is important in environmental-analytical and sanitary-epidemiological control, analysis of food and feed, metals and alloys, inorganic materials, highly pure substances, polymeric materials, semiconductors, oil products, detection of explosives, etc., in medical and scientific research.

The effectiveness of elemental analysis largely depends on instrumental methods of analysis. Among instrumental analysis methods, X-ray fluorescence, atomic emission, atomic absorption spectrometry, spectrophotometry and luminescent analysis, electrochemical methods (polarography, potentiometry, etc.), mass spectrometry (spark, laser, etc.), various options for activation analysis are widespread. When choosing a method and analysis technique, the structure of the analyzed materials, the requirements for the accuracy of determination, the limit of detection of elements, the sensitivity of determination, selectivity and specificity, as well as the cost of analysis, staff qualifications, speed of analysis, the level of necessary sample preparation and the availability of necessary equipment are taken into account.

Of particular note is the radioactive analysis. Radioactive analysis is a method of analyzing a substance by the nature of the radiation of radioactive isotopes formed during the bombardment of a test substance with high-energy nuclear particles. Radioactive analysis is highly sensitive and is used to determine impurities in metals, alloys, semiconductor materials and other substances. Activation analysis refers to the basic nuclear physics methods for detecting and determining the content of elements in various natural and technogenic materials and environmental objects. The method is based on fundamental concepts and data on the structure of atomic nuclei, cross sections of nuclear reactions, schemes and probabilities of decay of radionuclides, radiation energies, as well as on modern methods of separation and preliminary concentration of trace elements. Activation analysis was widely used due to such advantages over other methods as low detection limits for elements (10 ^-12 -10 ^-13 g · cm ^-3 ), expressivity and reproducibility of the analysis, the possibility of non-destructive simultaneous determination of 20 or more elements in a sample. Common to all methods of activation analysis is the activation of matter by neutrons, gamma rays, or charged particles and the subsequent recording of the spectral composition of the radiation of excited nuclei or the resulting radioactive isotopes. The first two methods are most common. Activation analysis on charged particles, due to their low range in the substance, is mainly used for the analysis of thin layers and in the study of surface effects.

The main disadvantage of these methods are high requirements for sample preparation and a large or relatively long analysis time, and in some cases, such as, for example, in the case of γ-activation analysis, the cumbersome and low-tech tool equipment.

For a wide range of problems of elemental analysis with acceptable accuracy, within 10%, sensitivity and resolution of about 10 ^-3 ÷ 10 ^-4 %, the speed of elemental analysis and the high penetration ability of tools for detecting and controlling elemental composition are important. This takes place when identifying hidden bookmarks that are in places inaccessible for visual viewing, for example, in containers of bulky cargo transportation, bookmarks from explosives, smuggling at customs control posts, etc. The most optimal combination of capabilities for identifying the elemental composition of media under the conditions listed above is possessed by the elemental analysis method (s) based on spectral analysis of instantaneous γ-quanta generated during radiation capture of neutrons by element nuclei.

This method is most fully developed and used by Rapiscan Systems (www.rapiscansystems.com), in which the quantitative composition of the medium (identification of elemental composition) is controlled by neutron irradiation of materials through spectral analysis of γ-quanta arising from the interaction of fast neutrons with materials ( substances). Elemental analysis is used to control bulky luggages (containers).

In the method proposed in this application, the elemental analysis also uses spectral analysis of γ-quanta arising from the radiation capture of neutrons by element nuclei. However, in contrast to the techniques used by Rapiscan Systems, the proposed method analyzes the spectra of gamma quanta that are generated when only neutrons are captured that do not enter the nuclei of identifiable elements in the inelastic scattering reaction [1]. In addition, the proposed method identifies not only the elemental composition, but also the chemical (molecular) composition. Using the modified method of discrimination of time intervals significantly improves the characteristics of a non-overloading γ-ray spectrometer, increases statistical accuracy and reduces systematic errors. The measurement error is no more than 5–15% for exposures (measurement intervals) of no more than several tens of seconds and sensitivity up to 10 ^{–5–10 –6} g, for example, in samples of biological tissue with a volume of V ~ 1 cm ³ with a specific gravity of 1 g Cm ^-3 . In general, the method allows estimating the concentrations of elements and chemical compounds in condensed and gaseous media up to ^10-3 ÷ 10 ^-4 % with an error of less than 10%.

The authors of this application seems very promising to use the proposals put forward in medicine and especially in the treatment of cancer. A new method for the treatment of cancer, which experts have high hopes for, is neutron capture therapy (NRT). The essence of NRT is as follows: a preparation containing boron-10 or gadolinium-157 is injected into the tumor and irradiated with a beam of high-intensity thermal neutrons. The drug has the ability to accumulate in the tumor, therefore, the energy released from the particles that destroy the tumor cells formed as a result of the nuclear reaction occurs only in it and healthy tissues are practically not damaged. For a complete cure, one session of NRT is enough. Developed technology is the technology of choice for patients who cannot be helped by other methods of treatment.

BNZT - Bor-Neutron-Capture-Therapy - one of the central areas of modern non-surgical oncology has not yet left the category of experimental treatment only due to a number of technical, financial and other factors. For example: a slight deviation from the corresponding accuracy criteria in determining the concentration of the drug with boron, established by oncologists, and, as a result, the incorrectness of planning a treatment session, can lead to the death of significant areas of healthy tissue or to an insufficient dose of radiation to the tumor tissue, requiring repeated procedures, and therefore, to the danger of an increased risk of treatment sessions. Therefore, the development of non-invasive methods for controlling boron content in the area of tumor growth directly in the process of NRT is a fundamentally important factor in the success of therapy. The proposed method and equipment for its implementation to identify the elemental and chemical composition of the media allows us to determine the concentration of boron-10 in biological tissue in the range of 10-40 μg / g with a specific gravity of tissue ρ ~ 1 g · cm ^-3 with an error of not more than 10% per a time of several tens of seconds and even units of seconds.

A well-known method for identifying elemental composition by spectral analysis of instantaneous γ-quanta generated during the radiation capture of fast neutrons is used in identification systems of Rapiscan Systems (www.rapiscansystems.com). The company effectively uses this method to control the elemental composition of bulky luggage in closed freight containers. The disadvantage of this method is the ambiguity in determining the chemical composition of the media due to the imperfection of the method in compiling the calibration reference data identifying calibration measurements.

The known method [2, 3], adopted as a prototype, the identification of elemental composition by spectral analysis of instantaneous γ-quanta generated in the reactions of radiation capture of neutrons by element nuclei. In the considered method [2, 3], the possibility of identifying the chemical (molecular) composition of the media is mentioned. However, the authors nowhere describe how such identification is carried out. The disadvantage of this method is the large statistical error of the measurements and, in connection with this, prolonged exposure during the measurement. In the proposed application, the noted deficiencies are absent.

The essence of the proposed method for determining the elemental composition is as follows: a source of fast neutrons is placed in a slowing unit, a γ-spectrometer using the non-overloading linear detection method, instantaneous γ-quanta generated during radiation capture of neutrons by element nuclei are recorded, calibration responses of individual elements included in the composition of the identified sample, according to the hardware spectra of γ-quanta, determine the concentration of elements in the sample through weight response coefficients elements, characterized in that the chemical composition of the media is determined according to the scheme: a priori chemical composition — known chemical compounds — media are defined, calibration responses of compounds are determined through the sum of the calibration responses of the elements included in the compounds, concentrations of compounds and elements in the identified medium are determined, and the concentration of elements that make up the compounds, establish the correspondence of the obtained total concentration of elements of the environment, taking into account the elements that make up the identifiers of identifiable medium compounds, the concentration of medium elements obtained by identifying the decoding of only the elemental composition of the medium, if the given concentration does not coincide for a given error, the total concentration of medium elements obtained by identifying decryption, taking into account the elements that make up the identifiable medium compounds, with the concentration of medium elements obtained by identifying decoding only elemental composition of the medium, the procedure starting with a priori setting the structure of the compounds is repeated. When performing identification measurement procedures, a γ-spectrometer is used that uses the non-overloading linear detection method with a modified method of discrimination of time intervals or, more precisely, with the method of double discrimination of time intervals (see below, p. 18).

The known device for the identification of elemental composition is implemented by Rapiscan Systems [4]. The disadvantages of the device used by the company are: the use of isotopic sources of fast neutrons; low resolution; low level of elemental analysis speed, accuracy and sensitivity; lack of reliable criteria for the chemical (molecular) composition of controlled environments. The disadvantages of the implemented installation should also include the following: the identifying installation is equipped with various auxiliary equipment, which makes it cumbersome and not always convenient. The weight of the installation can reach 20 ÷ 25 tons. The unit is transported unassembled and disassembled by sea or rail.

In the installation, implemented by the proposed method, these disadvantages are absent.

The prototype of the presented invention is a known device for the identification of elemental composition, proposed in [2, 3]. The method of operational identification and control, implemented by the installation, is based on recording and decoding the spectral composition and output of instantaneous γ-quanta generated when neutrons are irradiated with element nuclei. As neutron sources, pulsed neutron generators are used. The disadvantage of the device of the prototype is the low statistical accuracy and, as a result, long measuring exposures (several tens of minutes or more). The device proposed in this application is devoid of this drawback.

The claimed method is implemented using a device including a neutron source, a neutron moderation unit, a γ-spectrometer using the method of non-overloading linear detection, a program unit including a library of calibration characteristics (responses of the γ-spectrometer in the form of arrays of instrumental spectra of instantaneous γ-quanta generated by radiation neutron capture by nuclei of various elements), a measuring complex for recording the instrumental spectra of γ-quanta, a block of a software complex for converting total contributions (weighting coefficients) of the element-wise calibration characteristics to the values of the concentration of the element-wise composition, the interface for converting the measurement information to the values of the element-wise concentration on-line, characterized in that a γ-spectrometer is installed in a device that implements the method of elemental and chemical analysis of media non-overloading linear detection method with a modified method of discrimination of time intervals or, more precisely, with a method of double discrimination of time int intervals, which allows reducing the measurement procedure by several decimal orders of magnitude for identifying the elemental and chemical composition of the media, a program unit for converting the measurement information into neutron energy spectra is installed, adapted to calculate the concentrations of chemical compounds and media elements based on the measurement information and using modified identifying calibration characteristics ( spectrometer responses), allowing to evaluate the concentration of elements and chemical their compounds and condensed gaseous media up to 10 ^{-4 -3} ÷ 10% with an error less than 10%, the device is performed in the stationary and mobile (portable) embodiments.

The implementation of the method and the operation of the installation are illustrated by the following figures.

Figure 1. Block diagram of a stationary version of an identifying conveyor-type installation.

1 - pulsed neutron source; 2 - neutron moderation unit; 3 - radiation detectors (γ-spectrometer); 4 - Pentium I with a software unit for processing experimental data on-line; 5 - measuring and recording complex; 6 - briquette TBPO; 7 - conveyor.

Figure 2. Temporal dependence of the energy of instantaneous γ-rays of radiation capture absorbed in a detector placed in a neutron spectrometer according to the deceleration time.

▲ - H, v - C, --- - N, - ·· - - О, ···· - S, Δ - Cl, □ - Fe, - - Ni, - · - - Gd, + - ₈ U .

Figure 3. Hardware spectra of gamma rays.

1 - ▲▲▲ - mixture of radionuclides (experiment); 2 - °°° - background; 3 - +++ - ¹³³ Ba; 4 - *** - ¹⁵² Eu; 5 - ♦♦♦ - ¹³⁷ Cs; 6 - ▲▲▲ ⁶⁰ Co.

Figure 4. The distribution of time intervals N (T).

Δ - calculation, <n> = 1549 s ^-1 ; ° - experiment, <n> = 1549 s ^-1 , detector - NaI (Tl), radiation - γ-quanta; • - experiment, <n> = 1549 s ^-1 , detector - NaI (Tl), radiation - γ-quanta, level of discrimination of time intervals T _d = 150 μs.

Figure 5. Detector operating in linear detection mode.

9 - n, γ; 10 - scintillating single crystal; 11 - PMT; 12 - output of the photomultiplier, R _o > 1000 MΩ; 13 - storage capacitor, C _n ; 14 - preamplifier, R _in ~ 1000 MΩ.

6. The initial instrument spectrum φ (V) of the mixed flux of γ-quanta from ¹³⁷ Cs and ⁶⁰ Co. radionuclides Experiment. Detector: single crystal - NaI (Tl).

7. Convolutional instrumental spectrum of gamma quanta Φ (V) for <n> = 0.01. The initial hardware gamma-ray spectrum is shown in Fig.6.

Fig. 8. Hardware spectra of gamma rays. Radionuclides: ⁹⁴ Nb; ¹³⁷ Cs; ⁶⁰ Co. Detector: NaI (Tl) single crystal with a height of h = 18 cm and a diameter of d = 20 cm. Using discrimination of time intervals. ° - ⁹⁴ Nb; × - ¹³⁷ Cs + ⁶⁰ Co.

Fig.9. Hardware spectrum of gamma rays. Radionuclide ⁹⁴ Nb. Detector: NaI (Tl) single crystal with a height of h = 18 cm and a diameter of d = 20 cm. The load is 540 kHz. Without discrimination of time intervals.

Figure 10. The distribution of time intervals N (T), <N> = 10 MHz.

.1 - T ₁ ; 2 - T ₂ = T _D ;

.

11. Block diagram of a spectrometric setup.

9 - γ-quanta; 10 - NaI (Tl) - scintillating single crystal; 11 is an electronic photomultiplier optically coupled to a NaI (Tl) single crystal; 12 - dynode; 16 - electronic unit (charge-energy); 19 is a block with a controlled linear pass or linear gate; 22 - G5-78 - rectangular pulse generator; 25 - amplifier and driver of impulses coming from a linear transmitter; 26 - single-board analyzer SBS-60; 4 - Pentium I, in the slot of which a single-board pulse analyzer SBS-60 is inserted; Note: 15, 17, 18, 20, 21, 23, 24 - coaxial connectors through which switching is carried out with various electronic units of the spectrometric complex.

Fig. 12. Block diagram of the charge block.

27 - timer IMS1; 28 - the front panel of the electronic unit; 11 - PMT; 12 - the output of the dynode of the photomultiplier PMT (11); 15, 17, 18, 20, 23 - front panel connector; 16 - electronic unit (charge-energy); 19 - line skip or linear gate; 22 - generator G5-78; 29 - repeater UDP15; 30 - input ("leg" 2) of the timer IMS1 (1); 31 - repeater UDP18; 32 - capacitor C; 33 - input ("leg" 4) of the timer IMS1 (27); 34 - timer IC5; 35 - repeater UDP 42; 36 - shaper IMS7; 37 - repeater UDP 43; 38 - repeater UDP28; 39 - repeater UDP29; 40 - input ("leg" 2) of the timer ICM5 (34); 41 - storage capacitor C ₂ , forming the output pulses of the timer IC 2 (61) of duration T = T ₁ , similar to the storage capacitor C ₁ (42); 42 - storage capacitor C ₁ participating in the formation of pulses at the output of the timer IMS5 (34) of duration T = T ₀ ; 43 - output ("leg" 6) of the timer (IMS5) (34); 44 - output ("leg" 7) of the timer (IMS5) (34); 45 - output ("leg" 3) of the timer (IMS5) (34); 46 - repeater UDP 25; 47 - shaper IMC6; 48 - repeater UDP 27; 49 - switch P; 50 - repeater UDP20; 51 - output ("leg" 7) of the timer IMS1 (27); 52 - output ("leg" 3) of the timer IMS1 (27); 53 - repeater UDP 9; 54-repeater UDP10; 55 - repeater UDP11; 56 - UDP2 repeater "; 57, 58 - front panel connectors; 59 - UDP14 repeater; 60 - UDP16 repeater; 61 - IMS2 timer; 62 - output (" leg "3) of the IMS2 timer (61); 63 - UDP2 repeater; 64 - repeater UDP3; 65 - driver IMC3; 66 - output (output) (leg 1) of the driver IMC3 (65); 67 - output (output) ("leg" 6) of the driver IMC3 (65); 68 - amplifier UD1; 69 - repeater UDP; 70 - amplifier UD2; 71 - repeater UDP6.

Fig.13. Grouping of events within the interval ΔT = T ₂ -T ₁ = 70 ns.

Let us dwell in more detail on the method and settings of the prototype due to the fact that the prototype includes the basics of using the principle of regularization of A.N. Tikhonov when conducting both elemental and chemical analyzes of the composition of the media. The plants are designed in two versions:

- stationary option;

- "portable" mobile option.

The stationary version of the installation, with dimensions of approximately 2 × 2 × 1.8 m, can in forced mode under the conditions of the conveyor flow of identifiable material provide a capacity of up to 1000 t / day. When controlling a briquette weighing 10 ÷ 20 kg, the elements are identified at their concentration (by weight) of 0.5 ÷ 1% of the composition of the entire briquette mixture.

The radiation background when the installation is operating at a distance of 5 ÷ 10 m from the installation does not exceed the maximum permissible standards adopted for the public during round-the-clock irradiation during the year. There is no radiation background on an idle installation, because in the off state, the neutron generator does not emit neutrons.

The unit is mounted from unified modules for 2.5 ÷ 3 hours. Understands the same amount of time. It can be easily transported and installed practically in any place where appropriate radiation safety can be ensured.

The setup allows you to identify the elemental composition of the materials of the media in the presence of their own (in the material of the media) background of γ-quanta. The cost (estimated) of the installation is not more than 150 thousand USD.

An analogue of the stationary variant is the devices used by Rapiscan Systems (www.rapiscansystems.com [4]). The advantage of the installation developed in [2, 3] is the best (in comparison with the installations developed by Rapiscan Systems) identifying resolution (10–20 times) with 2–3 times less weight and cost of the installation itself. Moreover, the estimated scanning speed of large bookmarks (containers) is 2–3 times higher.

[5]. Контролируемый материал транспортируют вблизи среды, замедляющей нейтроны. Скорость радиационного захвата замедляющихся нейтронов регистрируют по выходу мгновенных γ-квантов, образующихся при радиационном захвате нейтронов материалом среды. Скорости радиационного захвата нейтронов от времени замедления, а следовательно, и от энергии нейтронов f[Ф_γ(E_н(t))], где t - время замедления нейтронов, определяемые через регистрацию потока эмиссии мгновенных γ-квантов, существенно различаются для разных элементов. Рассмотрим функциональные зависимости f[Ф_γ(E_н(t))] в качестве функций чувствительности замедляющего блока к выходу γ-квантов в зависимости от атомного веса элементов А. ТогдаConsidered in the prototype method is based on the registration and identification of the spectral composition and output of instantaneous γ-quanta that are generated during the capture of neutrons by nuclei of elements. A pulsed neutron generator with a neutron energy of E _n = 14 MeV and a neutron pulse duration of τ _and = 10 ^-6 s is placed in a heavy low-absorbing moderator, which is a spatial medium that forms neutron fluxes with different strictly fixed average energies at a certain point in time

[5]. Controlled material is transported near a neutron-slowing medium. The rate of radiation capture of decelerating neutrons is recorded by the output of instantaneous γ-quanta generated during radiation capture of neutrons by the material of the medium. The rates of radiation capture of neutrons from the deceleration time, and therefore from the neutron energy f [Ф _γ (E _н (t))], where t is the neutron deceleration time, determined through registration of the emission flux of instantaneous γ-quanta, differ significantly for different elements . Let us consider the functional dependences f [Ф _γ (E _н (t))] as functions of the sensitivity of the slowing down unit to the output of γ-quanta depending on the atomic weight of elements A. Then

where K (t, A) = f (Φ _γ (t), A) is the kernel of the integral equation (1).

We define a slowdown unit, coupled with a transport system for supplying materials and a system for recording instantaneous γ-quanta, as an identifying installation. Then the kernel K (t, A) of equation (1) is the hardware response function of the identifying setup and is completely identified by the sensitivity function f (Ф _γ (t)). Solving equation (1), it is possible to estimate the initial concentrations of the elements φ (A) in the mixture of materials from the temporal instrumental behavior of the γ-ray output from the nuclei of materials u (t).

The solution of integral equation (1) (Fredholm equations of the first kind) is performed in accordance with the regularization theory developed by A.N. Tikhonov [6] to solve essentially ill-posed problems such as the Fredholm equation of the first kind. The solution of equation (1) is obtained by minimizing the smoothing functional:

- производная 1-го порядка искомого спектра концентраций элементов φ(А);

- производная 1-го порядка априорного спектра концентрации φ₀(А); c₁ - заранее выбранные весовые коэффициенты; p(t) и q₁(A) - заданные весовые коэффициенты.where α is the regularization parameter;

- the first-order derivative of the desired spectrum of element concentrations φ (A);

- the first-order derivative of the a priori concentration spectrum φ ₀ (A); c ₁ - pre-selected weights; p (t) and q ₁ (A) are the given weights.

The real set of parameters included in equation (1) is discrete. Therefore, equation (1) is written in matrix form and the system of equations is solved:

In relation to the operator equation (3), the smoothing functional (2) can be written in the form:

where | P | and | Q _l | are weight matrices, | R _l | - 1st order differentiation operator.

The considered approach for decoding measurement information is widely used in neutron spectrometry for spectrometric systems of a non-classical type. A typical example of spectrometric systems of a non-classical type is a multisphere spectrometer [7], [8]. A multisphere spectrometer has a set of spheres made of a material that effectively slows down neutrons, for example, polyethylene. Spheres - of different diameters. The sensitivity function or response in a multisphere spectrometer is the counting function of a neutron detector located inside a slowing sphere, depending on the energy of neutrons incident on the sphere. For slowing down spheres having different diameters, the calculated functional dependencies will be different. In relation to a multisphere spectrometer, equation (1) is written in the form:

where φ (E) is the spectrum of neutrons incident on the decelerating sphere; E is the neutron energy.

The neutron spectrum φ (E) is an analogue of the concentration spectrum φ (A) in equation (1). Each sphere of a multisphere spectrometer, which differs from others in its moderating and scattering properties, is similar to a chronosection at a certain value of time t with a finite time interval Δt in the problem of the elemental composition of materials. However, unlike a multisphere spectrometer, in which the real number of spheres is limited (as a rule, not> 10 spheres), the number of chronosections in the problem of elemental composition is practically unlimited and the upper limit of their number is determined by the attainable rank of the inverse matrix | U | ^-1 when solving the system of equations (3) and observing the criteria of the principle of regularization of solutions to ill-posed problems. For practical purposes, this means that the simultaneous analysis of media consisting of 50-100 elements is not an overwhelming task in determining the elemental composition of materials. The method for decoding measurement information, considered using the multispheric neutron spectrometer as an example, has been tested for several decades and is characterized by the stable reliability of the obtained experimental results. Therefore, this method is taken as the basis for the formation of an algorithm for determining the elemental and chemical composition of identifiable materials, adapted in accordance with relations (1) ÷ (5).

According to the results of studies presented in [2, 3], using the functional dependences f (Ф _γ (t)) ≡K (t, A) obtained by Monte Carlo calculations [9, Figure 2], based on The regularization method has developed an algorithm and a software package for calculating the concentrations of chemical elements. The testing of the algorithm and the software package for decoding the composition of the mixture of medium prepared from predetermined concentrations of known elements was performed using the example of the installation shown in FIG. 1. From the pulsed source 1, the neutrons enter the heavy moderator 2. The neutron flux leaving the moderator 2 falls on the briquette 6 (TBPO) located on the conveyor belt 7. As a result of the interaction of neutrons with the nuclei of the material medium of the briquette 6, γ-rays of radiation capture are generated, which are recorded by detectors of a gamma-ray spectrometer 3. The measurement information enters the program unit 4 of the measuring and recording complex 5, at the output of which on-line output information is generated on the elemental and chemical th part of the preform 6.

The sensitivity functions were calculated by the Monte Carlo method [9] in geometry simulating a real setup and real physical processes accompanying the interaction of instantaneous γ-rays of radiation capture with the detector medium. The detector is a scintillation detector made in the form of a plate based on a scintillating Bi ₃ GeO ₁₂ single crystal. Scintillator dimensions: 20 × 40 × 40 cm. The parameter Ф _γ (t), which reflects the behavior of the sensitivity functions K (t, A), characterizes the total absorbed energy E _{γ of the} instantaneous γ-rays of radiation capture during their interaction with the scintillator depending on time neutron moderation in a lead moderator.

The identified sample is a briquette in the form of a cube with a side h = 40 cm. The briquette is located above the horizontal plane of the heavy moderator. The distance between the plane of the lower face of the sample and the plane of the moderator is 40 cm.The plane of the moderator practically acts as a planar neutron source with a diameter of about 100 cm.The decrease in the neutron flux of a plane source at a distance of 100 cm from the plane of the moderator is no more than 2-3 times. Such neutron flux characteristics above the plane of the heavy moderator are provided by a special configuration of channels and voids inside the volume of the heavy moderator. The working range of time intervals within which the "measuring" calculation procedure is carried out is from 10 to 1000 μs, sometimes up to 2000 ÷ 3000 μs. This range of operating time intervals covers a range of neutron energies from about 1 ÷ 10 eV to 100 keV. The energy resolution at any time slice within the specified energy (and time) interval, as a rule, is not worse than 50–70%.

The sensitivity functions f (Ф _γ (t)) ≡K (t, A) were calculated by the Monte Carlo method using the MCNP software package [8] for the elements: Cl, Fe, Gd, Н, N, О, S. Briquette sizes - 40 × 40 × 40 cm. From the above elements, a homogeneous mixture was prepared having densities: ρ ₁ = 0.1 g · cm ^-3 ; ρ ₂ = 0.2 g · cm ^-3 ; ρ ₃ = 0.4 g · cm ^-3 . The function was calculated (see (1)) φ (A) = Σ _i b _i · f (E _γ , t, A _i ) characterizing the yield of instantaneous γ-quanta of the mixture with a predetermined spectrum of element concentrations determined by the "weight" coefficients b _i . Table 1 shows the results of the decoding procedure described above.

Table 1 * ⁾ No. p / p The elemental composition of the mixture The concentration of elements in the mixture (rel. Units), (input information) Concentration of elements in the mixture (rel. Units), (output information - decoding) Relative error of the output information,% the density of the mixture ρ = 0.1 g · cm ^-3 ; the geometry of the mixture is a cube with a side l = 40 cm; the mixture is homogeneous one H 0.05 0,0515 18.1 2 C 0.15 0,1517 16.7 3 N 0.10 0,0996 14.6 four O 0.20 0.1968 12.0 5 S 0.10 0.0971 8.95 6 Fe 0.20 0.1884 6.2 7 Cl 0.20 0.1804 5.1 the density of the mixture ρ = 0.2 g · cm ^-3 ; the geometry of the mixture is a cube with a side l = 40 cm; the mixture is homogeneous one H 0.05 0,0491 10.3 2 C 0.15 0.1465 9.58 3 N 0.10 0.0975 8.58 four O 0.20 0.1950 7.38 5 S 0.10 0.0975 6.1 6 Fe 0.20 0.1945 5.1 7 Cl 0.20 0.1925 4.8 the density of the mixture ρ = 0.4 g · cm ^-3 ; the geometry of the mixture is a cube with a side l = 40 cm; the mixture is homogeneous one H 0.05 0,0514 29.5 2 C 0.15 0,1506 27.16 3 N 0.10 0,0999 23,4 four O 0.20 0,2002 18.75 5 S 0.10 0,1010 13,2 6 Fe 0.20 0,1990 8.26 7 Cl 0.20 0.1910 5.77 * ^{) The} data shown in table 1 were obtained using monitoring the attenuation of the neutron flux over the thickness of the identified samples.

In the above-described embodiment of the interpretation of the measurement information, sensitivity functions having an integral character are used as initial data. In each time interval of the neutron moderation range, an integral characteristic is determined: the total absorbed energy released in the medium during its interaction with instantaneous γ-rays of radiation capture. This approach in determining the sensitivity functions provides high statistical accuracy with a duration of exposure of the measuring procedure within 1 ÷ 3 s. However, the relatively smooth nature of the behavior of the sensitivity functions and their, in connection with this, “soft” difference between themselves leads to noticeable systematic errors. Therefore, errors in the decoding output results, as a rule, amount to, on average, 5–10%.

To significantly reduce systematic errors when using the above method of decoding the elemental composition, a mobile (or "portable") version of the identifying installation was developed and experimentally tested. The information basis of the mobile version is the instrumental spectra of recoil electrons, which are formed in the scintillator-detector during the interaction of the scintillator medium with instantaneous γ-quanta arising from the radiative capture of neutrons by an identifying medium.

In the mobile version, a heavy moderator is not used. A moderator is used whose maximum weight is not more than 10 kg. When using a pulsed neutron generator with a neutron energy E _n = 14 MeV, the neutron deceleration time in such a moderator is no more than 3–5 μs.

Hardware spectra or amplitude distributions of recoil electrons at the output of scintillators N (E _γ , E _e -, V) ≡N (V) are characterized by more pronounced irregularities, which provides a significantly higher conditionality of the corresponding matrices when solving equations of type (1), (2) . Therefore, in the mobile version of the identifying installation, the amplitude hardware distributions N (V) are used as responses or sensitivity functions. More fully: N (V) ≡ N (E _γ , E _e -, V, A), where E _γ is the energy of the instantaneous γ-rays of radiation capture, V are the amplitude distributions of the recoil electrons having the energy E _e -, A≡A _i - i-th element - the source of the birth of instantaneous γ-rays of radiation capture.

In the mobile version mode, experiments were performed on a laboratory installation simulating a mobile version of an identifying installation. In the created experimental setup there was no neutron generator. Simulators appeared as sources of γ-rays of radiation capture. Sources-simulators of instantaneous γ-rays of radiation capture selected radionuclides that are part of the set of OSGI (exemplary spectrometric γ-sources). For experimental studies, γ sources were selected: ¹³⁷ Cs; ⁶⁰ Co; ¹⁵² Eu; ¹³³ Ba. The fifth source is the premises, the activity of which is from 10% to 50% of the activity of the sources of the OSGI set. The abbreviation of the fifth source is background (fon). Figure 3 illustrates the instrumental spectra of γ-quanta of radionuclides: a mixture of radionuclides - 1; background - 2; ¹³³ Ba - 3; ¹⁵² Eu - 4; ¹³⁷ Cs - 5; ⁶⁰ Co - 6.

Hardware spectra (Figure 3) were measured using a NaI (Tl) single crystal, having dimensions: height h = 18 cm; diameter d = 20 cm. Sources are placed on the installation line and the hardware spectrum is recorded from the mixed γ-ray flux of the entire ensemble of the listed sources.

In one version of the geometry of the experiment, the fraction of the intensity of each source in the mixed flux of gamma rays incident on the scintillator from ¹³⁷ Cs radionuclides; ⁶⁰ Co; ¹⁵² En; ¹³³ Ba; background: 0.160; 0.157; 0.466; 0.139; 0.078 respectively. The results of the decoding of the measurement information appearing in the form of an instrumental spectrum of gamma quanta of the mixture were performed by the method of statistical regularization and are presented in table 2.

table 2 No. p / p Radionuclide Relative intensity of gamma radiation of a radionuclide in a mixture of radionuclides (input information) The relative intensity of the gamma radiation of the radionuclide in the mixture
radionuclides (output - transcript) Relative error of the output information,% one Cs 0.160 0.1597 0.50 2 Co 0.157 0.1585 <0.6 3 Eu 0.466 0.4674 0.63 four Ba 0.139 0.1393 1.20 5 fon 0,078 0,0756 ~ 3.4

Table 2 shows the almost complete coincidence of the input and output data. Input data: a priori selected and experimentally verified fraction of the γ-radiation intensity of each radionuclide in the mixture of all used γ-sources. Output data: the fraction of the radiation intensity of each radionuclide in the mixture of all used γ sources, determined by calculation by the method of statistical regularization using experimental data on the instrumental spectra of the γ quanta of the mixture and experimental data on the shape of the instrumental spectra of the initial (reference) sources of γ quanta. The errors in the results of the calculated decryption data are mainly in the range 0.5–1.5%. But the implementation of an insignificant level of errors requires large exposures to perform measurement procedures in comparison with a stationary installation for the identification of elemental composition. This is due to an increase in the "weight" of statistical errors. To ensure statistical errors of less than 3%, it is necessary to increase the exposure of measurement procedures by 10–20 times. Under conditions similar to those given above, for a stationary installation, the exposure time will be not 1 ÷ 3 s, but ~ 20 ÷ 60 s. It should be noted: the results of decoding the elemental composition obtained for the conditions of a stationary installation, if necessary, can always be supplemented with data obtained under conditions characteristic of mobile identifying installations. Moreover, using the same measuring equipment and without any special additional adjustment operations.

A feature of the regularization method used in decoding measurement information is that it allows the successful identification of not only the elemental, but also the chemical composition of materials and media. It is this circumstance that is used in the method for determining the chemical composition of media and a device implementing this method proposed in this application.

Table 3 shows examples in which the results of the analysis (identification) of the elemental and chemical (molecular) composition of 2 different media (mixtures) formed using experimental data on the responses (instrumental spectra of γ-quanta) of composite elements and molecules are presented. As in the results given in table 2, the corresponding radionuclides are imitators of the elements of the environment. Similarly, the simulators of the molecules of the medium are combinations of radionuclides, combined in a certain ratio. In this case, the response of the synthesized molecule is not taken as a set of responses of its constituent "elements" (radionuclides), but the combined (superpositioned) total response of its constituent ("molecule") "elements" (radionuclides). For example, consider a “molecule” of the type Cs2Ba1Co2≡Cs ₂ BaCo ₂ . The response of the molecule is an instrumental spectrum of gamma rays, consisting of a mixture of gamma fields of the "elements" of Cs, Ba, Co, the weight fractions of which must take into account the chemical composition of the "molecule" of Cs ₂ BaCo ₂ , taking into account normalization to the molecular weight of the Cs ₂ BaCo molecule ₂ in the mixture. The table shows a very good agreement between the input and output data of the analysis (identification) of the elemental and chemical composition of the media (the difference is significantly less than the relative errors of the identifying analysis).

Table 4 presents the results of the analysis (identification) of the elemental and chemical composition of media with an elemental composition similar to the composition of the media shown in table 3 using the same responses for the elemental and molecular composition as for the media presented in table 3. But in table 4, the molecular composition is intentionally "erroneously" changed. The results of the identifying analysis show a complete discrepancy (discrepancy) in the input and output data. Moreover, these discrepancies are far beyond the relative errors of the identifying analysis.

Table 3 Wednesday No. p / p Radionuclides that mimic the elemental and molecular composition of a medium Relative intensity of γ-radiation of radionuclides simulating the elemental and molecular composition in a mixture of radionuclides (input information) Relative intensity of gamma radiation of radionuclides simulating the elemental and molecular composition in a mixture of radionuclides (output) Relative error of the output information,% one one fon 65000E-01 53545E-01 11539E-01 2 Cs2BalCo6 45000E + 00 45475E-00 94792E-02 3 Eu 48500E + 01 49317E + 00 10590E-01 2 one fon 65000E-01 55435E-01 99891E-02 2 Cs 30000E + 00 29324E + 00 72800E-02 3 Ba 13500E + 00 13694E + 00 79878E-02 four Co2eu3 50000E + 00 51344E + 00 81363E-02

Table 4 Wednesday No. p / p Radionuclides that mimic the elemental and molecular composition of a medium Relative intensity of γ-radiation of radionuclides simulating the elemental and molecular composition in a mixture of radionuclides (input information) Relative intensity of gamma radiation of radionuclides simulating the elemental and molecular composition in a mixture of radionuclides (output) Relative error of the output information,% one one fon 82000E-01 65004E-01 11539E-01 2 Cs2BalCo2 31200E + 00 45000E + 00 94784E-02 3 Eu 60600E + 01 48500E + 00 10590E-01 2 one fon 92000E-01 65000E-01 99891E-02 2 Cs 42900E + 00 30000E + 00 72800E-02 3 Ba I9300E + 00 13500E + 00 79878E-02 four Coleul 28600E + 00 50000E + 00 81367E-02

Our proposed method for identifying elemental and chemical composition is implemented using a device developed on the basis of the prototype described in [2, 3]. The device is based on a measuring complex with a scintillation detector capable of providing undistorted instrumental spectra of gamma rays when the detector photomultiplier up to 1000 MHz is loaded onto the photocathode. However, the known prototype device has a significant drawback: despite the ability to provide high loads on the photomultiplier photomultiplier, in some cases, lengthy measurement procedures are necessary, since the control channel of the sampling (discrimination) of time intervals operates in a non-linear mode. The problem is solved using the modified method of discrimination of time intervals or the method of double discrimination of time intervals.

To understand the essence of the modified method of discrimination of time intervals proposed in this application, we consider in more detail the features of the unmodified method of discrimination of time intervals used in the prototype. Let us briefly consider the possibilities of a scintillation spectrometer using the method of discrimination of time intervals and the method of non-overlapping linear detection used in the prototype [10, 11, 12].

The essence of the method of discrimination of time intervals used in the prototype is that an additional channel is allocated in the spectrometer in which a spectrum of time intervals is formed (Figure 4). The pulses responsible for the formation of the spectrum of time intervals arrive at the input of the integrated discriminator, the discrimination threshold of which is above the level of T _d . The spectrum of time intervals below the T _d level is distorted due to dead time effects and overlays. Impulses above the threshold (level T _d ) are normalized by the integral discriminator and from the output of the integrated discriminator go to control the linear gate. The linear input of the linear gate receives pulses from the output of the spectrometer channel of the spectrometer. The pulses of the spectrometric channel are passed to the output of the linear gate only if they correspond to the region of time intervals (in the spectrum of time intervals) above the level of T _d .

The method of non-overlapping linear detection (prototype) is reduced to the rejection of the use of nonlinear forming circuits when fixing analog information at the output of the radiation detector. Figure 5 illustrates a linear detection method using an example of a scintillation detector of nuclear radiation. Nuclear radiation 9 (neutrons ″ n ″ and ″ γ ″ quanta) is detected by a scintillation single crystal 10. Scintillation flashes at the output of a single crystal are incident on the photocathode of photomultiplier 11 (PMT). The resulting current pulses i (t) from the output 12 of the photomultiplier 11, having an output resistance R _o > 1000 MΩ, are fed to the storage capacitor 13 (C _n ). The time-varying charge of the capacitor Q (t) after the pre-amplifier 14, having an input resistance R _xx ~ 1000 MΩ, is recorded by the subsequent recording equipment. The resulting information is presented in the form of instrumental spectra of voltage pulses U (t) = Q (t) / C _n . In implementing the described approach should be satisfied: R _{n O} · C · _Rin ≥R C _{and N} >> τ where τ _and - the duration of the current pulses at the output of the photomultiplier 11, and hence the duration of scintillation flashes produced in the single crystal 10 during registration nuclear radiation 9.

The developed spectrometric system in the study of Poisson streams of events provides a number of unique capabilities.

1. The processes accompanying the formation of distributions at the output of the storage capacitor after the corresponding transmitting device lend themselves to a rigorous analytical description.

2. By changing the duration of the accumulation interval at the storage capacitor T, any operating mode of the recording system is ensured - from spectrometric to integral.

3. In the spectrometric mode, it is allowed to load a photomultiplier up to 10–100 MHz onto the photocathode.

4. In integrated mode, i.e. in the recording mode of the total absorbed energy in the time interval T, it is allowed to load a photomultiplier up to 100-1000 MHz onto the photocathode, which is 4-5 more decimal orders than existing spectrometers.

Studies have established that for a Poisson stream of events, the instrumental distributions (instrumental spectra) recorded at the output of the storage capacitor C _n for the time interval T are described by the expression

where [φ _i (V) * φ _j (V) * ... * φ _k (V)] is the convolution of the “k” th multiplicity; φ (V) = φ _i (V) = φ _j (V) = ... = φ _k (V) - initial “unit” hardware distributions or hardware spectra of γ-quanta (or neutrons); P ₀ , P ₁ , P _k - coefficients whose values are determined by the Poisson formula

where m = 0, 1, k; [φ _i (V) * φ _j (V) * ... * φ _k (V)] is the convolution of the hardware distributions φ (V) of multiplicity "k".

It has been established by calculation and experimental studies that the distributions at the output of the storage capacitor C _n for 2 extreme values of the parameter <n>, namely for <n><< 1 and <n>>> 1, have very important properties. For <n><< 1, the convolutional instrumental spectra Ф (V) almost completely coincide with the differential instrumental spectra φ (V) ≡N (V), where N (V) is the count in the analyzer channel of the recording setup, recorded by classical methods. For example, already at <n> = 0.01, the spectrum of Φ (V) almost completely coincides with the spectrum of φ (V) (see Fig.6, 7). For <n>>> 1, more precisely for <n>> 10–15, the distributions of Φ (V) are described by the Poisson distribution with the parameter <n> _q , which is related to the parameter <n> to within a fraction of a percent as

On Fig, 9 presents the hardware spectra of γ-quanta, measured using the method of discrimination of time intervals and without the use of discrimination of time intervals. The instrumental spectra of radionuclides were measured: ⁹⁴ Nb, E _{γ, 1} = 0.702 MeV, E _{γ, 2} = 0.871 MeV; ¹³⁷ Cs, E _γ = 0.661 MeV; ⁶⁰ Co, E _{γ, 1} = 1.173 MeV, E _{γ, 2} = 1.3232 MeV (Fig. 8). The detector is loaded with a flux of γ-quanta of the radionuclide ⁹⁴ Nb - 540 kHz; the detector is loaded with a flux of γ-quanta of radionuclides ( ¹³⁷ Cs + ⁶⁰ Co) - 1.5 kHz.

However, the use of the described method of discrimination of time intervals does not allow high statistics in the hardware spectra of gamma quanta at high loads of detectors and short-term exposures of measurement procedures due to the restrictions imposed by the forming nodes of the control channel.

In the installation proposed in this application, this drawback is eliminated, which allows for high statistical accuracy in combination with minimal systematic errors. With a sensitivity of 10 ^-4 ÷ 10 ^{-6, it is} possible to provide a total error of the measurement results at a level no worse than 10% when the exposure of the measurement procedures is only a few tens of seconds.

As noted above, the problem is solved using the modified method of discrimination of time intervals or the method of double discrimination of time intervals.

The essence of the modified method of discrimination of time intervals or the method of double discrimination of time intervals, which is the subject of novelty, proposed in this application, is illustrated by an example: the photocathode of a photomultiplier tube (PMT) receives a stream of optical photons generated in the scintillator as a result of registration of γ-quanta. Let the γ-ray registration intensity be 10 MHz. As a scintillator, we use a NaI (Tl) single crystal. The limiting values of the load (registration intensity) on the photocathode photocathode in modern γ-ray spectrometers made on NaI (Tl) single crystals are 1–5 kHz [10]. This is due to the fact that NaI (Tl) emission pulses are characterized, at best, by two components: fast τ _b ~ 200 ns and slow τ _m ~ 1 μs [1, 2]. In accordance with expressions (5), (6) [10], the registration of γ-ray radiation using only the fast emission component τ _b ~ 200 ns determines the maximum load on the photomultiplier of the PMT around N = 1–5 kHz. Figure 10 shows the distribution of time intervals for the case of loading onto the photocathode of the PMT <N> = 10 MHz.

Set the level of discrimination of time intervals equal to T ₁ = 0.4 · 10 ^-6 s = 400 ns (1, Figure 10). The discrimination procedure is performed using a charging device (charging unit), a block diagram of which is presented below in Fig.12. We establish the second level of discrimination of time intervals equal to T ₂ = T _d = 0.470 · 10 ^-6 s (2, Figure 10), i.e. the sampling time interval in the apparatus spectrum of gamma quanta on the time interval scale T of the time interval distribution N (T) will be ΔT _d = T ₂ -T ₁ = 0.7 · 10 ^-7 s = 70 ns. The fast emission component of NaI (Tl) is about τ _b ~ 200 ns. Thus, spectra recorded at the output of the spectrometer are recorded on the basis of voltage pulses formed by a portion (~ 30%) of the charge of the pulses of the emission currents.

. При Т_д=Т₂ скорость регистрации составит

. Среднее число импульсов, попадающих в интервал ΔТ_д=70 нс, составит ΔN=(N₁-N₂)·ΔТ_д=1,5·10⁵·0,7·10^-7~0,01=10^-2 отсчетов. В момент Т_д ¹≈0,5·10^-6 с (3, Фиг.10) зарядовый блок переходит в исходное состояние и режим функционирования зарядового блока повторяется.Thus, the time interval between discrimination T _d = 70 ns. When T _d = T ₁ the registration speed will be

. When T _d = T ₂ the registration speed will be

. The average number of pulses falling in the interval ΔT _d = 70 ns will be ΔN = (N ₁ -N ₂ ) · ΔT _d = 1.5 · 10 ⁵ · 0.7 · 10 ^-7 ~ 0.01 = 10 ^-2 samples . At the moment T _d ¹ ≈0.5 · 10 ^-6 s (3, Fig. 10), the charge block returns to its initial state and the mode of functioning of the charge block is repeated.

The spectral pulse count rate is

With the number of analyzer channels N _A = 256, the number of samples in each analyzer channel will be:

During the time Δt≈100 s, the number of samples in each channel will be ΔN _A ~ 10 ⁴ s ^-1 .

As you can see, the statistical error in each channel will be δ ~ 1%. The error due to the general dispersion of the entire spectrum will be δ _g << 1%.

It was established by experiments and calculations [11, 12] that the distortion of the instrumental spectra of gamma quanta during the registration of emissions by spectrometers operating in the linear detection mode (Fig. 7) is negligible (the error in the entire spectral range is << 1%) if the condition : the number of current pulses arriving at the storage capacitor C _n during the time interval ΔT ₁ , averaged over a large number of time intervals n _T >> 1, must satisfy the condition: <n _q ><< 1, more precisely <n _q > ≤0, 01, where <n _q> - the number of current pulses, post Payuschie on storage capacitor C _n. Observing this condition and using the algorithm of the method of double discrimination of time intervals described above, when using scintillators with faster emission times when excited by nuclear radiation, it is possible to ensure loading of a photomultiplier <N> ≥1000 MHz on the photocathode. For example, for plastic scintillators, characterized by only a fast emission component when excited by nuclear radiation, δ _b ≈ 10 ^-9 s, when a photomultiplier <N> ~ 10 ³ MHz is loaded onto the photocathode, the speed of the measurement procedures and the quality of the measurement information will be similar, in the case <N> ~ 10 MHz, analyzed above using the NaI (Tl) single crystal as an example.

In the device proposed in this application, the indicated drawback of the “conventional” method of time interval discrimination is eliminated with the help of the developed electronic circuit tested in the electronic unit 6, which is an integral part of the γ-spectrometer block diagram (Fig. 11), used as the leading unit of the identifying installation elemental and chemical composition of media. The spectrometer is manufactured and tested in a prototype breadboard version.

в течение времени T₀ до поступления на вход 33 таймера 27 (ИМС1) импульса блокировки отрицательной полярности. Импульс блокировки отрицательной полярности, поступающий на вход 33 ("ножка" 4) таймера 27 (ИМС1), переводит триггер таймера в состояние, при котором таймер переходит в открытое состояние, при котором конденсатор 32 (С) разряжается и накопления заряда на нем не происходит. Импульс блокировки на входе 33 ("ножка" 4) таймера 27 (ИМС1) вырабатывается таймером 34 (ИМС5), который в исходном состоянии также открыт. Запуск таймера 34 (ИМС5) происходит, практически, одновременно с запуском таймера 27 (ИМС1) через повторитель 35 (УДП42) и формирователь 36 (ИМС7). Импульс отрицательной полярности и амплитудой 2,5÷3,0 B после повторителей 37 (УДП43), 38 (УДП28), 39 (УДП29) поступают на управление таймером 34 (ИМС5) (вход 40 ("ножка" 2)) и закрывает таймер 34 (ИМС5). С помощью генератора тока, выполненного на переменном сопротивлении, происходит накопление заряда на накопительном конденсаторе 42 (C₁) до напряжения, равного опорному, задаваемому на компараторе таймера 34 (ИМС5), с которым соединен вывод 43 ("ножка" 6) таймера 34 (ИМС5). Вывод 43 ("ножка" 6) таймера 34 (ИМС5) соединен с выводом 44 ("ножка" 7), который, в свою очередь, соединен с накопительным конденсатором 42 (C₁). При достижении на накопительном конденсаторе 42 (С₁) напряжения, равного уровню опорного напряжения на компараторе таймера 34 (ИМС5), триггер таймера 34 (ИМС5) опрокидывается (переключается в другое состояние) и открывает ключ, соединенный с выводом 44 ("ножка" 7) таймера. Таймер открывается, а на выводе 45 ("ножка" 3) таймера выделяется прямоугольный импульс напряжения положительной полярности и длительностью Т₀, которая задается сопротивлением генератора тока таймера 34 (ИМС5). На выходе вывода 45 ("ножка" 3) таймера - четыре ветвления с выходами на соответствующие повторители, выполненные на операционных усилителях. По одному из этих ветвлений через повторитель 46 (УДП25) импульсы поступают на формирователь 47 (ИМС6) нормализованных сигналов длительностью - 0,5÷1,0 мкс, передний фронт которых по времени соответствует заднему фронту прямоугольных импульсов на выходе 45 таймера 34 (ИМС5). С выхода формирователя 47 (ИМС6) импульсы напряжения отрицательной полярности и с выхода повторителя 48 (УДП27) через переключатель 49 (П), повторитель 50 (УДП20) поступают на вход блокировки таймера 27 (ИМС1) (вывод 33 таймера 27 (ИМС1)). Накопление заряда на конденсаторе 32 (С) прекращается, т.к. таймер 27 (ИМС1) переходит в исходное открытое состояние, и на выходах таймера 27 (ИМС1) (выводы 51, 52) выделяются импульсы напряжения длительностью

(3, Фиг.10). Пилообразный импульс напряжения положительной полярности ("пила") длительностью Т₀, выделяемый на выводе 51 таймера 27 (ИМС1), через повторители 53 (УДП9), 54 (УДП10), 55 (УДП11), 56 (УДП2'') поступает на выходные разъемы 17, 18, 57, 58 блока 16. Сформированная таким образом "пила" является генеральной.Let us consider briefly the interaction of electronic nodes in the developed electronic unit, with the help of which the ideology of the method of double discrimination of time intervals described above is implemented (Fig. 12). From the output of the 10th (or 12th or other) dynode 12 of the photomultiplier photomultiplier 11, current pulses of positive polarity are fed to the input of connector 15 of the front panel 28 of unit 16. From the internal output of connector 15 (inside unit 16), current pulses are fed to input 51 ( the findings of chip 27 (IMS1)) normally open timer KR1006 VI1 (IMS1). From the output of the generator 22 (G5-78), the pulses through the connector 23 of the block 16 and the repeater 29 (UDP15) are fed to the input 30 of the timer 27 (IMS1) to control the timer. Upon receipt of the output of the generator 22 (G5-78) pulses of negative polarity (chain: connector 23 of the front panel 28 of block 16, repeater 29 (UDP15), repeater 31 (UDP18)) to the input 30 of timer 27 (IMS1), the timer trigger caps over and the timer closes. On the capacitor 32 (C), the accumulation of charge begins

during the time T ₀ before the input 33 of the timer 27 (IMS1) blocking pulse of negative polarity. A negative polarity blocking pulse, input to the 33 (“leg” 4) of the timer 27 (IMS1), puts the timer trigger in a state in which the timer goes into the open state, in which the capacitor 32 (C) is discharged and there is no charge accumulation on it . The blocking pulse at the input 33 ("leg" 4) of the timer 27 (IMS1) is generated by the timer 34 (IMS5), which is also open in the initial state. The start of the timer 34 (IMS5) occurs, almost simultaneously with the start of the timer 27 (IMS1) through the repeater 35 (UDP 42) and the shaper 36 (IMS7). The pulse of negative polarity and an amplitude of 2.5 ÷ 3.0 V after the repeaters 37 (UDP 43), 38 (UDP 28), 39 (UDP 29) are sent to the timer 34 (IMS5) (input 40 ("leg" 2)) and closes the timer 34 (IMS5). Using a current generator made on a variable resistance, the charge is accumulated on the storage capacitor 42 (C ₁ ) to a voltage equal to the reference voltage set on the comparator of the timer 34 (IMS5), to which the output 43 ("leg" 6) of the timer 34 ( IMS5). The terminal 43 (“leg” 6) of the timer 34 (IMS5) is connected to the terminal 44 (“leg” 7), which, in turn, is connected to the storage capacitor 42 (C ₁ ). When the accumulator capacitor 42 (C ₁ ) reaches a voltage equal to the level of the reference voltage on the comparator of the timer 34 (IMS5), the trigger of the timer 34 (IMS5) overturns (switches to another state) and opens the key connected to terminal 44 ("leg" 7 ) timer. The timer opens, and at the terminal 45 ("leg" 3) of the timer, a rectangular voltage pulse of positive polarity and a duration of T ₀ , which is set by the resistance of the current generator of the timer 34 (IMS5), is allocated. At the output of terminal 45 (“leg” 3) of the timer, there are four branches with outputs to the respective repeaters made on operational amplifiers. In one of these branches, through the repeater 46 (UDP25), the pulses are fed to the former 47 (IMS6) of normalized signals with a duration of 0.5 ÷ 1.0 μs, the leading edge of which in time corresponds to the trailing edge of the rectangular pulses at the output 45 of timer 34 (IMS5) . From the output of the shaper 47 (IMS6) voltage pulses of negative polarity and from the output of the repeater 48 (UDP27) through the switch 49 (P), the repeater 50 (UDP20) are fed to the input of the blocking timer 27 (IMS1) (terminal 33 of the timer 27 (IMS1)). The accumulation of charge on the capacitor 32 (C) stops, because the timer 27 (IMS1) goes into the initial open state, and voltage pulses of duration duration are allocated at the outputs of the timer 27 (IMS1) (pins 51, 52)

(3, Fig. 10). A sawtooth voltage pulse of positive polarity ("saw") of duration T ₀ , allocated at terminal 51 of timer 27 (IMS1), through repeaters 53 (UDP9), 54 (UDP10), 55 (UDP11), 56 (UDP2 '') connectors 17, 18, 57, 58 of block 16. The “saw” formed in this way is general.

In addition to starting the timer 34 (IMS5), from the output of the shaper 36 (IMS7) through the repeaters 59 (UDP14), 60 (UDP16), the timer 61 (IMS2) is launched. Rectangular pulses of positive polarity and a duration of T ₁ <T ₀ from terminal 62 of timer 61 (IMS2) are fed to a forming device made at 63 (UDP2), 64 (UDP3), 65 (IMS3), then through an amplifier 70 (UD2), a repeater 71 (UDP6) and then through the switch 49 (P) and a repeater 50 (UDP20) to the input (terminal 33) of the timer 27 blocking (IMS1). At the output 62 of the timer 61 (IMS2) and at the outputs of the timer 27 (IMS1) (pins 51 and 52), voltage pulses of duration T ₁ <T ₀ are allocated. At the output (output) 66 of the shaper 65 (IMC3), the generated pulses along the leading edge correspond to the trailing edge of the pulses T ₁ . Pulses of positive polarity and a duration of ΔТ ₁ = ΔТ _Д from the output (output) 67 of the shaper 65 (IMC3) after amplifiers 68 (UD1) and 69 (UDP5) are fed to the connector 20 of the front panel 28 of block 16 and then to control the line pass (line pass or linear gate 19).

Figure 11 shows a block diagram of a gamma-ray spectrometric complex operating in the mode of a non-overloading gamma-ray spectrometer using the method of double discrimination of time intervals, gamma-quanta 9 are recorded by a scintillating single crystal 10 (NaI (Tl)) optically coupled to the photocathode of the photomultiplier 11 (PMT-93). From the output of the PMT dynode 12, the current pulses arrive at the connector 15 of the block 16. From the output of the 17 block 16, the voltage pulses, whose amplitudes are directly proportional to the energy of the detected radiation and which are recorded in the time interval ΔT = T _d (Figure 10), are input 18 block 19 linear gate (or linear skip). From the output of the connector 20 of the block 16, the normalized voltage pulses of duration ΔT≤T _d -T ₁ (see the description of the charging block, Fig. 12) are fed to the input 21 of the linear gate control channel (or linear gate). Control pulses of duration ΔT≤T _d -T _{1 are} synchronized with the pulses from the output of the generator 22 (G5-78) to control the charging unit 16 (connector 23). From output 24 of the linear gate block 19, voltage pulses are fed to an amplifier-driver 25, information on the spectral composition of which is recorded by a single-board analyzer 26 SBS-60 and is stored in 4 PC Pentium I.

There is a very important circumstance that should be taken into account when recording the spectra of pulse amplitudes with the spectrometric complex shown in FIG. 11: possible distortions in the spectral composition of pulse amplitudes due to the finite duration of the time interval ΔT≤T _d -T ₁ , i.e. the so-called boundary effects. For the case under consideration (parameters: <n> = 0.01, <N> = 10 MHz, ΔT = 70 ns), we calculated the probability of grouping the recorded events within the ΔT interval depending on time. On Fig illustrates the results of calculations performed on the basis of the convolution of three distributions: the Poisson stream of events, the binomial distribution and the distribution of time intervals within the double discrimination interval ΔT≤T _d -T ₁ taking into account the above parameters. The result is an almost 100% grouping of events far from the interval. An experimental check for the case <n> = 0.01 fully confirms this fact (Fig. 7 [11, 12]).

Using a γ-spectrometer, the block diagram of which is shown in Fig. 11, experiments using radionuclides were performed using the double-discrimination method for time intervals when a photomultiplier <N>> 540 kHz was loaded on the photocathode. The results obtained are similar to those shown in Fig. 8, but with a duration of the measurement exposure of one hundred times less than that used in the experiment, the results of which are presented in Fig. 8 (exposures of 40 s and 4000 s, respectively).

LIST OF SOURCES OF INFORMATION

1. Vlasov N.A., Neutrons. M .: Nauka, 1971.

2. Trykov O. A., Leonova O. O., Soloviev N. A., Semenov V. P., Khachaturova N. G. Installation for the operational determination of the elemental composition of organic and inorganic substances: Preprint FEI - 3095. Obninsk, 2007.

3. Trykov O. A., Leonova O. O., Semenov V. P., Soloviev N. A., Khachaturova N. G. Method and installation for the operational determination of the elemental composition of organic and inorganic substances. SSC RF IPPE. Questions of atomic science and technology. Series: Nuclear Constants, 2007, p. 81.

4. www.rapiscansystem.com. Rapiscan Systems web pages. Copyright © 2010.

5. Bekurts K., Wirtz K. Neutron physics. M .: Atomizdat, 1968.

6. Tikhonov A.N. On the solution of incorrectly posed problems and on the regularization method // DAN SSSR, 1963. V.151, N3. S.501-504.

Tikhonov A.N. On the regularization of incorrectly posed problems // DAN SSSR, 1963. V.153, N1. S.49-52.

7. Bramblett R.I., Ewing R.I., Bonner T.W. A new type of neutron spectrometer // Nuclear Instruments and Methods. Vol. 9, No. 1. P.1-12.

8. Semenov V.P., Trykov L.A., Tyufyakov N.D. Some methods for reconstructing neutron spectra from measurements with a multispheric spectrometer: Preprint FEI-507. Obninsk, 1974.

9. Briesmeister J.F. "MSNP-A General Monte Carlo N-Particle Code System, Version 4A." Los Alamos National Laboratory report LA-12625-M, 1993.

10. Trykov O.A. The method of discrimination of time intervals in the spectrometry of nuclear radiation: Preprint FEI - 2796, Obninsk, 1999.

11. Leonova O.O., Semenov V.P., Soloviev N.A., Trykov O.A., Khachaturova N.G. Non-reloading scintillation spectrometer of fast neutrons and gamma rays. SSC RF IPPE. Questions of atomic science and technology. Series: Nuclear Constants, 2008, VYP.1-2, S.87.

12. Leonova O.O., Semenov V.P., Soloviev N.A., Trykov O.A., Khachaturova N.G. Non-reloading scintillation spectrometer of fast neutrons and gamma rays. SSC RF IPPE. Nuclear Instrumentation. Metrology. Collection of reports SIC SNIIP, 2007.

13. Medvedev M.N. Scintillation detectors. M .: Atomizdat, 1977.

14. Trykov O.A., Mokhov A.S. Spectrometric method for measuring nuclear radiation and its spectrometric system. Patent for invention No. 2269798 dated February 10, 2006

Claims (3)

1. The method of elemental analysis of media, which consists in the fact that the source of fast neutrons is placed in the slowing down unit, with a γ-spectrometer using the non-overloading linear detection method, instantaneous γ-quanta generated during the radiation capture of neutrons by element nuclei are recorded, the calibration responses of individual elements are determined, included in the composition of the identified sample, from the hardware spectra of γ-quanta, determine the concentration of elements in the sample through the weight coefficients of the responses of the elements, which differs I mean that the chemical composition of the media is determined according to the scheme: a priori chemical composition — known chemical compounds — of the medium is defined, the calibration responses of the compounds are determined through the sum of the calibration responses of the elements included in the compounds, the concentrations of the compounds and elements in the identified medium are determined, the concentrations of the elements are determined, included in the composition of the compounds, establish the correspondence of the obtained total concentration of elements of the environment, taking into account the elements that are part of the identified compounds of medium the concentration of the elements of the medium obtained by identifying the decoding of only the elemental composition of the medium, if the given concentration does not match the total concentration of the elements of the medium obtained by the identifying decryption, taking into account the elements that make up the identifiable compounds of the medium, with the concentration of the elements of the medium obtained by identifying the decoding of only the elemental composition environment, the procedure starting with a priori setting the structure of the compounds is repeated. 2. The method according to claim 1, characterized in that when performing identifying measurement procedures, a γ-spectrometer is used that uses the method of non-overloading linear detection with a modified method of discrimination of time intervals or, more precisely, with the method of double discrimination of time intervals. 3. The device that implements the method according to claim 1, comprising a neutron source, a neutron moderation unit, a γ-spectrometer using the non-overloading linear detection method, a software package unit including a library of calibration characteristics (responses of the γ-spectrometer in the form of arrays of instrumental spectra of instantaneous γ- quanta generated during the radiation capture of neutrons by nuclei of various elements), a measuring complex for recording the instrumental spectra of gamma quanta, a block of a software complex for transforming weight contributions (weight coefficients) of element-wise calibration characteristics to element-wise concentration values, an interface for converting measurement information to element-wise concentration values in on-line mode, characterized in that a γ-spectrometer is installed in the device using the non-overloading linear detection method with a modified method of time discrimination intervals or, more precisely, with the method of double discrimination of time intervals, which allows reducing by several decimal pores During the measurement procedure for identifying the elemental and chemical composition of the media, a software package was installed to convert the measurement information into neutron energy spectra, adapted to calculate the concentrations of chemical compounds and media elements on the basis of the measurement information and using identifying calibration characteristics (spectrometer responses) to evaluate concentrations of elements and chemical compounds in condensed and gaseous media up to 10 ^-3 -10 ^-4 % s with an error of less than 10%, a device that implements the method of elemental and chemical analysis of media is performed in stationary and mobile (portable) versions.

Images