Creation of a high-performance and fast-responding scintillators is one of the most essential concept of the ma- terial authority of late. Combination of the high conversion efficiency and the short lifetime which gadolinium oxysulfide (Gd₂O₂S) obtain at the alloying with trivalent praseodymium (Pr³⁺) in combination with impurity of Ce³⁺ and F⁻, have defined a choice of the phosphor as a base at the development of the ceramic scintillators for a computed tomography [1] [2] .

The Pr³⁺ is a scintillating ion. The ion emits the radiation quantum at 513 nm corresponding to the transition ³P₀ → ³H_J, ³F_J, with a decay constant of about 3 μs. The Ce³⁺ secures the necessary level of the afterglow; F⁻ ion was introduced in the host lattice for increasing the luminescence intensity and reducing the afterglow. In the literature, hot pressing of the material with traces of Li₂GeF₆ as a sintering aid is usually used in order that the ceramic scintillator has high translucent. The density of the scintillator is 7.34 g∙cm⁻³. The Gd₂O₂S:Pr,Ce as well as Gd₂O₂S:Tb,Ce and Gd₂O₂S:Eu scintillation optical ceramics were investigated and were developed at Com- pany “NITIOM Vavilov State Optical Institute”, St. Petersburg, Russia since 2000. The uniaxial hot pressing method was used toward this end. Various domestic and foreign firms produce powder phosphors which were used for ceramics manufacture. The pressing process of the high-melting compound Gd₂O₂S is carried out together with the easily melted additive of the LiF to produce the ceramics with the high density and the transparency. The scintillation and the optical characteristics of the ceramics depend on the quality of the base phosphor powder and on its synthesis feature which determines a distribution of the lattice and the impurity defects in the crystal lattice. High transparency, uniformity of the ceramics and the required scintillation parameters are defined with the quality of the initial powder with other things being equal (Table 1).

The optical properties of the ceramics worsen, the specific light yield decreases and the afterglow increases, if the initial powder contains uncontrollable impurities. The purpose of the paper is to reveal the causes of deteri- oration of optical and scintillation characteristics of the Gd₂O₂S:Pr,Ce samples produced from the raw materials containing uncontrollable impurities.

The research of the trace, minor and major elements has been carried out with instrumental neutron activation analysis in Gd₂O₂S:Pr,Ce in view of the above mentioned.

The impurity content determination of Gd₂O₂S:Pr samples with neutron activation analysis is very intricate problem. First nuclei ³²P, ³³P and ³⁵S are accumulated in the sample when Gd₂O₂S:Pr is subjected to neutron ir- radiation. The ³²P, ³³P and ³⁵S isotopes are obtained with nuclear reactions described with the equations ³²S(n,p)³²P (T_1/2 = 14.262 d), ³³S(n,p)³³P (T_1/2 = 25.34 d), ³⁴S(n,p)³⁵S (T_1/2 = 87.51 d). These isotopes are power- ful sources of beta radiation with E_β = 1710.66 keV (³²P), E_β = 248.5 keV (³³P) and E_β = 167.14 keV (³⁵S). The decay schemes of the isotopes are given in the Figure 1 [3] . Beta radiation generates the low energy bremsstrah- lung in the gamma spectrum of the sample.

The next problem to be solved is the presence twelve isotopes ^{151,153,159,161,162}Gd, ^155,157,158Eu, ^153,155,157Sm and ¹⁶¹Tb which natural Gd produces under a reactor irradiation. The radioactive isotopes have a lot of gamma-rays of the large intensity at the energy of E_γ < 250 keV. The calculations and the experiment have shown that specific activities of the daughter Gd and Pr isotopes vary from 10⁵ to 10¹¹ Bq if irradiations were performed in the reactor in a thermal flux of 6.3 × 10¹³ s⁻¹∙cm⁻² and in a epithermal neutron flux of 3.1 × 10¹² s⁻¹∙cm⁻² after irradiation of 1 hour (Table 2). As the multichannel analyzer used had dead time about 95% - 100%, the gamma-ray spectrum of the sample cannot be registered with HPGe detector with a desired statistical precision because ¹⁵³Gd, ¹⁵¹Gd and ¹⁶⁰Tb are long-lived isotopes. The main nuclear characteristics of Gd and Pr isotopes are given in the Table 2.

The continuous background of the gamma-ray spectra is one of the limiting factors for the sensitivity of the analysis. It may be reduced by absorption of the interfering radiation, mainly beta radiation or low energy gam- ma radiation, by more or less thick foils placed between the sample and the detector. This technique is useful if

weak high-energy photo-peaks have to be measured in the presence of intensive low-energy radiation. In the fa- vorable case, photons of the certain energies can be selectively absorbed by materials whose K X-ray absorption edge corresponds to that energy.

The concentrations of the trace elements were measured by the absorption method in Gd₂O₂S:Pr. The Cu (0.1 mm thick), Cd (0.5 mm), W (0.1 mm) and Pb (0.1 mm) plates have been used as absorbing filters because these elements have the K X-ray absorption edges at the energies of 8.982 keV, 26.712 keV, 69.51 keV and 87.95 keV, respectively. The Cu and Cd plates have been used for the absorption of beta radiations and bremsstrahlung of ³²P, ³³P, ³⁵S. The W and Pb plates have been used for the absorption of low energy gamma and X-ray radiation of ¹⁵¹Gd (X-ray: 40.902 keV, 41.542 keV, 46.905 keV, 47.038 keV, 48.249 keV, gamma radiation: 21.532 keV, 63.91 keV, 64 keV, 93.17 keV), ¹⁵³Gd (X-ray: 5.85 keV, 40.902 keV, 41.542 keV, 46.905 keV, 47.038 keV, 48.249 keV, gamma radiation: 14.06383 keV, 19.81295 keV, 69.673 keV, 75.42212 keV, 83.36717 keV, 89.48594 keV), ¹⁶¹Tb (X-ray: 6.5 keV, 45.208 keV, 45.988 keV, 51.947 keV, 52.113 keV, 53.476 keV, gamma radiation: 25.65134 keV, 48.91532 keV, 57.19169 keV, 74.56668 keV).

The gamma-ray spectra of the multielement reference standard of IAEA 433 are given in the Figure 2. The reference standard of IAEA 433 was irradiated and cooled for 2 hours and 3 days respectively. One of the spectra has been measured by using the Cu-Cd-W-Pb filter and another spectrum has been measured without the filter. The gamma-ray spectra were measured with the 15% coaxial HPGe detector (FWHM 1.7 keV at 1173.238 keV, Canberra, USA) coupled with “Lynx” multichannel analyzer (Canberra, USA).

The counting rate of the complete absorption peaks at the energy of E_γ < 250 keV is much less in case of use of the Cu-Cd-W-Pb absorbing filters compared to the case without the filter. This distinction strongly decreases, if the gamma energy increases. The ratio S₀/S_F is plotted against the gamma energy in the Figure 3, where S_F and S₀ are the peak areas in the spectra of Gd₂O₂S:Pr sample measured with and without the Cu-Cd-W-Pb gamma rays and X-rays absorbing filter. The S₀/S_F value tends to 1 at the energy of E_γ > 250 keV (Figure 3) when Cu-Cd-W-Pb filter is placed between sample and detector. Thus, on the one hand we suppress gamma radiation at the energy of E_γ < 250 keV and on the other hand optimum conditions are achieved when we measured gamma spectrum at the gamma energy of E_γ > 250 keV.

The spectra of the sample at the energy of E_γ < 200 keV measured under the same conditions are presented in the Figure 4.

In our work the NAA technique of Gd₂O₂S:Pr,Ce samples has been developed. The technique provides the high

sensitivity and the results reproducibility for the large number of elements in one run. The analyses of the sample were performed using two methods. The first method is the k₀-method (Gd, Pr) of NAA and the second method is relative NAA (As, Ce, Co, Cr, Cs, Eu, Fe La, Sc, Tb, Zn, Zr, Na).

The k₀-method was used to determine concentrations of the Gd and Pr in Gd₂O₂S:Pr,Ce samples. Masses of the sample and the monitors were about 30 - 50 mg. The Al(99.9%)-Co(0.1%) and Al(99.9%)-Au(0.1%) alloys were used as the thermal and the epithermal flux monitors. The Ni foil (0.1 mm thick) was used as the fast neutron flux monitor. The Fe foil (0.1 mm thick) was used as the k₀-monitor. The samples, the k₀-monitor and the flux monitors were packed up into quartz glass ampoules. The ampoule’s glass was HQS high-purity quartz glass (Heraeus Quarzglas, Germany). The samples and the monitors were irradiated into the water channel of the WWR-M reactor with thermal neutrons (f_th), epithermal neutrons (f_epi), and fission-spectrum fast neutrons (f_f). Typical values of these fluxes at a sample irradiation positions in the 17 MW reactor are thermal f_th = 6.3 × 10¹³ s⁻¹∙cm⁻², epithermal f_epi = 3.1 × 10¹² s⁻¹∙cm⁻²and fast f_f = 4.1 × 10¹¹ s⁻¹∙cm⁻². For irradiation the time-period of two hours has been taken. The ampoules were washed out with 0.5 M HNO₃ aqueous solution and distilled water after the irradiation. The sample, the k₀-monitor and the flux monitor activity were measured with the 15% coaxial HPGe detector (Canberra, FWHM 1.7 keV at E_γ=1332.5 keV). The Cu-Cd-W-Pb gammas and X-rays absorbing filter was placed between the sample and the detector when we measured spectra. The gamma-ray spectra were measured after 7 and 15 days of cooling time.

The mass of Gd and Pr was calculated following the equation

The k₀ factor includes only natural constants

The asterisk marks the values belonging to the monitor. The η is counting efficiency of HPGe detector, the M is the atomic weight of the element, the a is the abundance of the isotope yielding the radionuclide to be measured. The σ is the thermal cross section. The I(α) value is determined with the equation

where I₀ is the resonance integral of the isotope and is effective resonance energy [6] [7] . The P is counting rate in photo-peak, the h is the gamma ray abundance. The Φ_th and Φ_epi are thermal and epithermal neutron fluxes, respectively. The is build up term. The is the decay and the measuring terms, the is the decay constant. The t_b, t_d and t_m are the irradiation, the decay and the measurement times, respectively. The T_1/2 is half-life of the analytical isotope. The α value is the epithermal flux deviation from the ideal (1/E) distribution.

The thermal and the epithermal neutron fluxes were measured with the double monitor method [6] . The Al(99.9%)-Co(0.1%) and Al(99.9%)-Au(0.1%) alloys were used as the monitors. These monitors were irradiated together with the samples without disturbing the neutron flux. The flux ratio was calculated following the equation

where . The deviation α of the epithermal flux from the ideal (1/E) distribution is determined by the “Cd ratio” method (Au, Co monitors) [6] . The result of measurement was α = 0.024 with the uncertainty of about 10%.

Relative technique was used to measure the concentrations of 13 trace elements in the samples.

The Gd₂O₂S:Pr,Ce samples and comparison standard (IAEA 433 reference material) were packed in high-pu- rity quartz glass ampoules. The sample and the comparison standard mass were about 30 - 50 mg. The samples were irradiated in a water channel of the WWR-M reactor for two hours at thermal f_th = 6.3 × 10¹³ s⁻¹∙cm⁻² and epithermal f_epi = 3.1 × 10¹² s⁻¹∙cm⁻² neutron fluxes. After irradiation the ampoules were washed out with the 0.5 M HNO₃ aqueous solution and the distilled water. The total the sample and the comparison standard activity was measured with the 15% coaxial HPGe detector (Canberra, FWHM 1.7 keV at E_γ = 1332.5 keV). In the time of measurements the Cu-Cd-W-Pb gamma and X-ray absorbing filters were placed between the sample and the detector. The gamma ray spectra were measured after 6, 20 and 40 days of cooling time. The gamma and X-ray absorbing filter was used in the measurements of the gamma spectra with Gd₂O₂S:Pr and the reference material of IAEA433. Parts of Gd₂O₂S:Pr,Ce spectra are presented in the Figure 5 and Figure 6. The gamma ray spectra were measured after a decay period of 6 and 20 days, respectively.

The Figure 5 shows the complete absorption peaks with respect to the well-known 1368.633 keV, 1460.8 keV and 1596.21 keV lines from 14.959 hours half-life ²⁴Na, 1.265 × 10⁹ years half-life ⁴⁰K and 1.6781 days half-life ¹⁴⁰La respectively.

The complete absorption peaks of the ⁵⁹Fe (1099.245 keV), ¹⁵²Eu (1112.069 keV), ⁶⁵Zn (1115.546 keV), ⁴⁶Sc (1120.545 keV), ⁶⁰Co (1173.2 keV), ¹⁶⁰Tb (1177.954 keV) long-lived isotopes (T_1/2 > 20 days) are presented in Figure 6. The gamma line of ⁶⁵Zn isotope at the energy of E_γ = 1115.546 keV coincides with a gamma line of ¹⁶⁰Tb isotope at the energy of 1115.12 keV (Figure 6). So, when calculating Zn concentration, the contribution of the 1115.12 keV gamma line of ¹⁶⁰Tb into the intensity of 1115.546 keV gamma line of ⁶⁵Zn was taken into account. The complete absorption peak area of ⁶⁵Zn at the 1115.546 keV was counted according to the equation

where S₁₁₁₅ is the peak area in the spectrum of the sample, S₂ is the peak area of ¹⁶⁰Tb at the E_γ = 1177.954 keV in a spectrum of the sample, η₁ and η₂ are the registration efficiency of the HPGe-detector at the 1115.546 keV and 1177.954 keV, respectively. The V₁ = 50.6% and V₂ = 14.8694% are the relative intensities of the 1115.546 keV gamma ray of ⁶⁵Zn and of the 1177.954 keV of ¹⁶⁰Tb.

Mass concentration of Gd and Pr was measured with the k₀-method and other elements were measured with the relative activation analysis technique in the analyzable samples. The analysis results of the sample and the nuclear characteristics [5] of the nuclides used to determine the concentrations are shown in the Table 3.

Specific activity of the radioactive isotope is calculated following equation:

where is time factor. The minimum activity required to enable measurement

with the desired statistical precision to be taken can be estimated for example, following the equations given by Gerhard Erdtmann [6] . He calculated the limit quantitative determination as

where A_q is the activity, which can be measured with a requisite precision. The A_q is calculated with

, where k_q is the reciprocal value of the requisite relative standard deviation and σ₀ is the

standard deviation of the background measurement. The determination limit of the trace and minor elements

were calculated with, where m_s is sample mass.

Spectrum of the tested scintillation ceramics is shown in the Figure 7. The Gd₂O₂S:Pr,Ce ceramics spectrum has the characteristic spectral band at the wavelength of λ_max = 513 nm (transition ³P₀ → ³H₄). The spectral band is Pr³⁺ emission in the matrix Gd₂O₂S:Pr,Ce.

The basic optical properties of the tested Gd₂O₂S:Pr,Ce ceramics are given in the Table 4.

Given in the Table 5 are the determination limits, reactions and the nuclear characteristics of the nuclides used to determine the concentrations of the 15 major, minor and trace elements, the conditions of irradiation, the cooling time and measurement time.

Neutron activation analysis technique of the Gd₂O₂S:Pr,Ce scintillation ceramics was developed. The concentrations of As, Ce, Co, Cr, Cs, Eu, Fe, La, Sc, Tb, Zn, Zr, Pr, Gd, Na have been measured with the instrumental neutron activation analysis. Relative light yield (λ_max = 513 nm) reduces when Ce content increases in the sam-

Following designations are accepted in the table: The Е_γ is gamma energy; p is isotopic abundance; σ_th is neutron radiative capture cross-section measured at 2200 m/sec; ¹Neutron radiative capture cross section measured in a Maxwellian flux; I is resonance integral; T_1/2 is half-life time; t_cool is cooling time; t_meas is measurement time; γ-abn is the average number of gamma photons of energy Е_γ per 100 decay events.

ple. The Gd₂O₂S:Pr,Ce has maximum value of the relative light yield (40%) at the Ce content about 10⁻⁵% in sample. Doping of Gd₂O₂S:Pr,Ce with Tb (~10⁻⁶%) leads to reducing the afterglow (0.045% - 0.055% after 5 ms and 0.001% afte 500 ms).