The paper presents the results of material testing for the purpose of obtaining radiator material – acomposite coating with neutron conversion material – for ionization chambers (IC) which contain the ⁶Li isotope and convert neutron radiation to a flux of high-energy charged particles through the ⁶Li(n, α)³H nuclear reaction. The proposed method for forming lithium-containing radiator material allows ensuring a high temperature resistance of up to 600 °C and a mechanical strength at the expense of adhesion to the IC electrode material (grade 321 steel). The advantages of a lithium-containing radiator, compared to a boron radiator, are explained by the smaller cross-section of the ⁶Li-neutron interaction: the smaller efficiency of the "neutron → charged particle" conversion is made up for by a high power density and a prolonged free path of reaction products in the radiator material, which makes it possible to increase the surface density of ⁶Li atoms, while reducing the extent of "burnup" in neutron fields. The IC electrode radiator material consists of a two-layer composite coating comprising an adhesive silicate layer and a functional neutron-sensitive lithium fluoride layer. Measurements at an alpha spectrometric facility have shown that the coating has a high energy output (~ 2.8·10^–3 MeV/neutron), which remains stable after four thermal cycles of up to 600 °C. The coating is resistant to vibration when exposed to frequencies of 35 to 200 Hz. The paper presents the results of testing the IC mockup with a lithium-containing radiator material. When irradiated with a neutron flux of 6·10³ cm^–2·s^–1, the IC mockup sensitivity value was about 10^–15A·s·cm²/neutron, which agrees with the calculated value.

neutron detectors, ionization chamber, radiator material, functional neutron-sensitive lithium fluoride layer

Ionization fission chambers with radiator coatings based on isotopes,such as ²³⁵U, ²³⁸U and ²³²Th, and boron ionization chambers (with the¹⁰B isotope used as the radiator), the technology of which was developed in the past century, are currently employed for neutron diagnostics at nuclear power plants.

The need for developing new neutron detectors is dictated by the change in operating conditions of radiation devices. Temperatures of up to 600 °С, neutron fluxes of up to 1·10¹⁷ cm^{- 2}·s^-1, and vibration loads of up to 200 Hz are reached in the present-day reactor cores (Baskov et al. 2022). Expanding the temperature range of ionization fission chambers, based on the electrical principle of operation, by organizing heterostructured nano-level uranium-oxide radiator coatings with a surface density of 1.2 mg/cm² is proposed in (Sakharov et al. 2012a, 2012b; Egorov et al. 2015; Glazyuk et al. 2017; Baskov et al. 2018) to support neutron diagnostics of current reactor plants.

There is no information in open sources on materials based on the ⁶Li isotope used as the IC radiator coatings. The key types of neutron detecting lithium-containing materials are lithium-silicate glass (Voytovetsky and Tolmacheva 1959, 1961; Shulgin et al. 2005) and crystal scintillators with the ZnS(Ag)/⁶LiF composition (Osovizky et al. 2018; Mosset et al. 2014). The use of the ⁶Li isotope and its compounds as the IC radiator materials is limited by their high reactivity and volatility (Allen 1962; Subbotin et al. 1999).

Using the ⁶Li isotope is promising in terms of forming radiator coatings which will feature a smaller "burnup" as compared with existing boron (¹⁰B isotope) radiator coatings. The balance of the number of active radiator nuclei is determined by equation (Dmitriev, Malyshev 1975):

$\frac{dN}{dt}=-\Phi ·\sigma ·N$ (1)

where Φ is the neutron flux;σ is the cross-section of interaction with the neutron; N is the number of active radiator nuclei; and t is time.

It follows from equation (1) that the larger is the cross-section, the larger is the radiator material "burnup". The ⁶Li (n,α)³H nuclear reaction cross-section (940 barns) is four times as small as for the ¹⁰B(n,α)⁷Li reaction (3840 barns). At the same time, the power density of the ⁶Li(n, α)³H reaction (4.80 MeV) is 1.7 times as high as in the event of ¹⁰B(n,α)⁷Li reaction (2.79 MeV), and the path length of the⁶Li(n,α)³H reaction products is larger than that for the ¹⁰B(n,α)⁷Li. This makes it possible to reduce the "burnup" of the lithium-containing layer and increase its efficiency. A higher energy yield creates conditions for using ICs based on the ⁶Li isotope both in the current and pulsed neutron recording modes. The development of the electronic component base and computer equipment allows nowadays to record individual fission events, which is expected to replace the classic actinide-based fission chambers.

The method proposed to obtain the IC lithium-containing radiator coatings is formation of a composite two-layer coating. The adhesion of the functional neutron-sensitive lithium-containing layer to the surface of the grade 321 steel (IC electrode material) will be achieved using an "adhesive" intermediate silicate layer.

Lithium fluoride, the most chemically and thermally stable compound (melting point of 848 °C), has been chosen as the functional base for the lithium-containing radiator coating.

Fluoride-ion, as a ligand, has a relatively small atomic weight (19 g/mole), which will allow as many target atoms of the ⁶Li isotope as possible to be formed on the surface. It needs to be taken into account that the thickness of the formed functional lithium-containing layer of the composite coating requires to be proportional to the path lengths of the alpha particle and tritium in the lithium fluoride environment, which is determined by the Bragg equation (Bolozdynya 2017)

$\frac{R_1}{R_0}=\frac{{\rho }_0\sqrt{A_1}}{{\rho }_1\sqrt{A_0}}$ (2)

where R₀ is the path length of the alpha particle (tritium) in air; R₁ is the path length of the alpha particle (tritium) in the environment; ρ₀ is the density of air; ρ₁ is the density of the environment; A₀ is the molar mass of air; and A₁ is the molar mass of the environment.

The path lengths of the alpha particle (R₀ = 4.5 μm) and tritium (R₀ = 26.1 μm) were determined from the nomograms of the path length and energy ratio (Experimental nuclear physics. Volume 1. 1955).

The material of the intermediate silicate layer in a plastic state is mechanically applied to the electrode surface, the surface roughness parameters of which are Ra = 5.3 μm and Rz = 29.2 μm. The material of the intermediate silicate layer penetrates into the microscopic dimpleson the electrode surface and is retained in these mechanically. Thixotropically, the solidifying silicate base also has a developed surface with parameters Ra = 12.3 μm and Rz = 23.9 μm. This promotes the adhesion of the deposited finely-dispersed lithium fluoride powder to the surface of the intermediate silicate layer while the latter is in a plastic state. In the process of drying, a composite two-layer coating with a lithium fluoride layer surface density of 2.5 mg/cm² is finally formed.

The microstructure of the composite radiator coating sample is shown in Fig. 1. A lithium fluoride layer with a thickness of 10 to 20 µm is applied to the surface of the intermediate silicate layer with a thickness of about 150 to 170 µm.

Figure 1. 

Microstructure (cross-sectional view) of a composite two-layer coating: 1 – intermediatesilicate layer; 2 – functional lithium fluoride layer.

It was found by the X-ray phase analysis (XRD) (Papko, Kravchuk 2013) that the silicate base of the intermediate coating layer consists predominantly of alphaquartz. Additional infrared spectroscopy investigations (Fig. 2) show that the silicate base includes a clay mineral (kaolinite). A broad band with 1098 cm^-1 is caused by the valence vibrations of the silicon-oxygen backbone tetrahedrons, and the narrow absorption bands with 798, 779 and 695 cm^-1 belong to alpha quartz. The bands in the region with 515 and 463 cm^-1 belong to the deformation oscillations of the alpha quartz silicon-oxygen groups (Plyusnina 1977). The narrow absorption lines with maximums at 3704 and 3620 cm^-1 coincide with the analytical absorption lines of the valence vibrations of the kaolinite hydroxyl groups (Nakamoto 1966). Following the heat treatmentat 600 °C, the disappearance of these absorption bands in the sample’s infrared spectrum is observed, which is associated with the kaolinite dehydration. The texture of the radiator coating intermediate layer (see Fig. 1) consists of alphaquartz grains (up to 30 μm), which are bound by the kaolinite veins.

Figure 2. 

Infrared absorption spectra of the composite coating intermediate layer silicate base: 1 – prior to heat treatment; 2 – afterheat treatment at 600 °С.

It was shown by investigating the thermal action effect on the radiator coating material structure using the XRD method that, following a heat treatment at 600 °C, the materials of the intermediate silicate layer (alphaquartz, kaolinite) and the functional (lithium fluoride) layers of the radiator coating did not chemically interact with each other, since the diffraction patterns for the mixture of these substances is a superposition of the alphaquartz and lithium fluoride diffraction patterns (Fig. 3).

Figure 3. 

Diffraction patterns of the radiator coating components: 1 – lithium fluoride after heat treatment at 600°C, 2 – silicate base after heat treatment at 600°C, 3 – mixture of lithium fluoride and silicate base (ratio 1:2) after heat treatment at 600 °C.

Lithium-containing radiator coatings were tested for resistance to mechanical overloads in horizontal and vertical planes under the action of vibrations at the VEDS-1500 vibration test bench. The samples were sequentially exposed to high frequency (200 Hz) vibrations with different acceleration amplitudes (1 to 8 g), the exposure time for each of the modes having been 30 min. It was found that the surface density of the composite radiator coating functional layer of 2.5 mg/cm² does not vary and, therefore, the coating remains intact. Composite coatings were also tested under the action of low-frequency resonant vibrations in accordance with GOST 29075-91 (GOST 29075-91, 2004). Test conditions: frequency intervals 6–8, 8–10, 10–13, 13–16, 16–20, 20–26, 26–35 Hz, acceleration amplitude 1.5 g, vibration exposure time for each frequency interval 15 min. Vibrations were changed from a lower frequency to an upper frequency with exposure at the highest upper frequency of each interval. The surface density of the functional layer prior to and after the tests remained unchanged and was 2.5 mg/cm². The results of the vibration resistance tests evidence that the lithium-containing radiator coating has high adhesive properties.

Radiometric testing of electrode samples with a lithium-containing radiator coating were undertaken at an experimental test bench under the action of the neutron flux of 104 cm^-2·s^-1 from a Pu-Be source. The energy yield from the radiator coating surface is determined from the spectrum obtained as a result of measurements, which is converted to the specific sensitivity of the coating using the following formula

$\eta =\frac{\Sigma E}{{\Phi }_0·S·\tau }·\frac{e}{\epsilon }=K·\frac{e}{\epsilon }$ (3)

where ΣE is the total energy yield of the radiator coating, eV; K is the specific energy yield of the radiator coating, MeV/neutron; ε is the ionization energy of the IC working gas (argon), 30 eV; e is the electron charge, C; Φ₀ is the neutron flux, cm^-2·s^-1; S is the electrode area, cm²; τ is the spectrum recording time, s; and η is the specific sensitivity of the radiator coating, C/neutron.

Fig. 4 presents the output pulse spectra for the lithium-containing functional layer of the composite radiator coating and the industrial boron radiator coating for the KNK-53 chamber. Spectrum 1 (see Fig. 4) includes a distinctly observable peak in the 2.75 MeV region corresponding to tritium, and a broad peak with an energy below 2.05 MeV related to the alpha particle. The peaks have a broad shape due to partial energy losses in the radiator material. The energy yield of the coating when irradiated with thermal neutrons was 3.7 10^-3 MeV/neutron. The specific neutron sensitivity of the lithium-containing composite radiator coating was η = 2.0·10^-17 C/neutron.

Figure 4. 

Spectrum of output pulses from charged particles for electrode samples: 1 – with a lithium-containing radiator coating; 2 – with a boron radiator coating for the KNK-53 chamber.

Spectrum 2 (see Fig. 4) includes visible broad bands with energies of less than 0.83 MeV (due to the presence of ⁷Li), as well as with energies of 1.47 and 1.7 MeV (owing to the alpha particle). The energy yield of the coating when irradiated with thermal neutrons was 0.9·10^-3 MeV/neutron. The specific neutron sensitivity value of the boron coating was η = 5.0·10^-18 C/neutron.

The sensitivity of the lithium-containing radiator coating is 4.0 times higher, and the energy yield is 4.1 times higher. The values of the ratios are generally comparable, and the slight difference is explained by different geometric characteristics of the electrodes.

A sample of the lithium-containing radiator coating was tested thermocyclically. Annealing conditions for a single cycle: inert medium – argon (absolute pressure 10⁵ Pa); annealing temperature – 600 °С; exposure time – 4h; heating rate – 5°/min; inertial cooling. The results of determining the specific neutron sensitivity of the coating in the thermocyclic test process are shown in Table 1.

It follows from the above data that the initial annealing is followed by a decrease in sensitivity due to the thermal emission of a portion of the fluoride-lithium coating. The specific neutron sensitivity, energy yield and radiation coating surface density values are constant with a confidence probability of p = 0.95. Following the decrease in the value in the course of thethermocyclic loads, the specific neutron sensitivity of the functional layer is 1.5∙10^-17 C/neutron on the average. The specific neutron sensitivity value of the boron radiator coating for the KNK-53 chamber in conditions of similar thermocyclic tests decreases from 5.0·10^-18 to 1.0∙10^-18 C/neutron, which is lower than the lithium-containing coating characteristics.

An IC model was developed and manufactured (Fig. 5) based on the data obtained. The chamber consists of two parallel plates (electrode profiles), each of which had four earlier tested electrode samples with a lithium-containing composite radiator coating applied to it. The electrode profiles are contained in a cylindrical case with a diameter of 50 mm, and the gap between the electrodes is 3 mm. Since the radiator material (lithium fluoride) is a dielectric, a metal mesh of a wire with a diameter of about 180 μm is attached to the surface of the electrode plates for the current collection (Fig. 6). The chamber is filled with theworking gas (argon) with an absolute pressure of 1.2·10⁵ Pa.

Figure 5. 

IC model: 1 – body; 2 – flange; 3 – guide disc; 4 – O-ring; 5 – flange with electrode leads; 6 – electrode profile with a mesh; 7 – electrodes with lithium-containing radiator coating; 8 – support insulators; 9 –electrode contact leads; 10 – electrical insulators.

Figure 6. 

Electrode profile with a mesh.

The IC model was tested on a dedicated test bench being irradiated with a neutron flux of 6·10³ cm^-2·s^-1 from a Pu-Be source. Measurements were undertaken in the current mode using a KVVS controller from the Avtotest APK suite. The measured current value is about 5·10^-12 A.

The volt-ampere characteristic (VAC) of the model is shown in Fig. 7. The VAC slope at a voltage of U ≥ 300V is s = 0.02%/V, which corresponds to the non-linearity of the loading characteristic, θ, of not more than 3.8%. The deviation from the linearity of the loading characteristic θ is determined by equation (Dmitriev, Malyshev 1975):

Figure 7. 

Volt-ampere characteristic of an IC model with a lithium-containing radiator coating.

$\theta =\frac{U·s}{2+{10}^{-2}·U·s}$ (4)

With a total radiator coating area of 56.5 cm², the measured sensitivity of the IC model was 0.8·10^-15 A·s·cm²/neutron, which agrees with the design value of 1.0·10^-15 A·s·cm²/neutron.

The composite lithium-containing coating has shown to be serviceable in the IC model design, which indicates that there is a high potential for using the proposed radiator material in neutron flux detection devices.

The paper presents the results of investigating a radiator lithium-containing coating with a composite two-layer microscale structure that consists of an intermediate silicate layer and a fluoride-lithium functional neutron-sensitive layer of a given thickness (10 to 20 µm). An examination of the microstructure and phase composition of the coating components has shown that the basis for the intermediate adhesive layer is formed by alphaquartz and clay mineral (kaolinite). The adhesion of the functional lithium fluoride layer is achieved due to the forces caused by the developed surface of the intermediate silicate layer. It has been found that the coating remains intact after vibration tests at a low-frequency resonant impact of up to 35 Hz and at a high-frequency impact of up to 200 Hz.

Radiometric characteristics of the lithium-containing radiator coating were studied. The specific neutron sensitivity was η = 2.0·10^-17 C/neutron. This value is four times higher than that for the boron radiator coating of the industrial KNK-53 IC. It has been found that thermocyclic tests at 600 °C after the initial annealing cause the specific sensitivity to decrease by a factor of about 1.5 due to the partial heat emission of the lithium-containing layer. The value of the specific neutron sensitivity stabilizes in the course of subsequent thermal cycles and remains constant (about 1.5·10^-17 C/neutron).

It is for the first time that possibilities have been demonstrated for using the ⁶Li isotope as a component of the IC radiator coating for the neutron flux detection sensors. Results are presented from neutronic studies of the lithium-containing radiator coating as part of the developed IC model. The measured value of the IC model current was 5·10^-12 A, the neutron flux from the Pu-Be source being F = 6·10³ cm^-2·s^-1. The sensitivity of the chamber model was about 10^-15 A·s·cm²/neutron. The volt-ampere characteristic slope for the chamber model with a lithium-containing radiator coating was 0.02%/V at a voltage of over 300 V.

Further studies will be aimed at improving the IC model’s design features and determining its pulse characteristics, and conducting durability tests to increase the sensitivity and stability of the lithium-containing radiator coating’s characteristics in the course of thermal cyclic testing.