1 Introduction

Both the radionuclides ³⁶Cl and ²³⁶U have a very long half-life (T_1/2) up to 3.01×10⁵ a, and 2.342 ×10⁷ a^[1], respectively. Because of the long half-life, these isotopes persist longer periods than human lifetimes in the environment. But this feature can be turned to advantage by using ³⁶Cl and ²³⁶U as isotopic tracer in many fields, such as the environmental, geological, and the nuclear safeguard research^{[2,3,4,5,6,7]}.

In nature, the natural ²³⁶U is very low^[2], and the ratio of ²³⁶U/²³⁸U in different minerals are expected in the level range of 10⁻¹⁰–10⁻¹⁴. The ²³⁶U is mainly produced by the ²³⁵U(n,γ)²³⁶U reaction, where neutrons could be produced by spontaneous fission of ²³⁸U or the (α,n) reaction on light elements such as Na, Mg, and Al. In 1996, Purser et al.^[3] proposed that the ²³⁶U can be used as a neutron flux integrator, which was extended further by Valenta et al.^[8] a few years later. The ²³⁶U is potentially applied to monitor the environmental impact of uranium mining by measuring the levels in drainage water of the mine and uranium exploration by searching the ²³⁶U in ground water^[9], respectively. Also, it is applicable to environmental and geological research. Like ²³⁶U, the natural ³⁶Cl under the deep subsurface is produced by the neutron capture reaction of ³⁵Cl(n,γ)³⁶Cl.

The produced ²³⁶U and ³⁶Cl depend on the neutron fluxes in the thermal and epithermal energy ranges. The combination of the ³⁶Cl measurements with ²³⁶U offers the possibility of determining subsurface neutron fluxes in the thermal and epithermal energy ranges^[9]. Because the natural ²³⁶U and ³⁶Cl are ultratrace, their measurement is difficult or even impossible without accelerator mass spectrometry (AMS)^{[10,11,12,13,14]}. Wilcken et al.^[9,15] have conducted some investigations on the nucleogenic of ³⁶Cl, ²³⁶U, and ²³⁹Pu in uranium minerals from several region to understand the essential features of the subsurface production of nucleogenic isotopes in uranium rich rocks. In order to promote practical application of ²³⁶U in uranium exploration and monitor the environmental impact of uranium mining in geological research, the nucleogenic of ³⁶Cl and ²³⁶U in uranium minerals should be conducted extensively to obtain the data from different region.

In this paper, the ³⁶Cl and ²³⁶U in the same uranium mineral are measured by AMS, which is developed in China Institute of Atomic Energy (CIAE). Two samples are from Guangxi and Shanxi province, China, respectively. The ratios of ³⁶Cl/Cl and ²³⁶U/²³⁸U are measured, and calculated by MCNP to understand the neutrons production model in uranium minerals.

2 Sample preparation

Sample 1 was collected from Guangxi province, China; and Sample 2 from Shanxi province, China (Fig.1). Two samples were under the deep subsurface of more than 1 m, avoiding the ³⁶Cl produced by the cosmic- ray spallation on Ar, K, and Ca.

Fig.1Full-screen View

Pictures of uranium samples. (a) Sample 1 from Guangxi province, China; and (b) Sample 2 from Shanxi province, China.

To precisely measure the ratios of ³⁶Cl/Cl and ²³⁶U/²³⁸U in a uranium mineral, a method of extraction and purification sample for AMS measurement was developed. In order to ensure products of an identical neutron flux, the ³⁶Cl and ²³⁶U were separated from the same uranium mineral and sample preparation procedure (Fig.2). Hard to dissolve, the big block uranium mineral was grounded into 300-μm powder.

2.1 ³⁶Cl sample

2.1.1 Dissolution

After the uranium mineral sample (5.0 g) was put into a Teflon vessel, 1 M HNO₃ (30 mL) was added. Teflon vessel was closed tightly for about 36 h, shaken by hand at every 5 h, and vibrated by ultrasonic device for 20 min. Most of Cl in the rock was extracted into the solution. A few residual rocks could be dissolved by adding the new HNO₃ solution. The Cl concentration was 1% less in the second dissolution solution than in the first. Because the AMS only needs to measure the ³⁶Cl/Cl, it is not necessary to take into account the second dissolution. The few residual rocks, which contained some insoluble uranium-bearing minerals, were transferred to another Teflon vessel to prepare the ²³⁶U sample.

2.1.2 Separation and AMS sample preparation

(1) the Cl sample solution was added to slight excess AgNO₃ (0.25 M) in the dark room to form AgCl within 24 h. (2) After discarding the supernatant, the AgCl was collected by centrifugal machine for 10 min. Besides, the residual solution was collected to prepare the ²³⁶U sample. (3) 10 mL high-purity water and ammonia was used to dissolve the AgCl at pH of more than 10. (4) Saturation Ba(NO₃)₂ was added to remove the sulfur for about 24 h at 30ºC to enough form BaSO₄ precipitate. (5) The Cl-containing solution was collected and filtered. (6) The excess high-purity HNO₃ was added to form AgCl, and collected with centrifugal machine for 10 min. In order to remove the sulfur, the steps of (3)–(6) were three times repeated.

Fig.2Full-screen View

The sample preparation procedure.

The AgCl was dried at 100ºC, put into a vial, and wrapped by black paper. In the AMS measurement, each sample was mixed with high purity (99.99%) silver powder according to their similar volume to improve thermal and electric conductivity, and pressed into a Cu sample holder (40-Sample NEC MC-SNICS ion source).

To minimize any additional contamination by sulfur, all the samples were prepared in an air-purified room without the reserved HCl and H₂SO₄. All reagents were 99.99% purity.

3 AMS measurement

The ³⁶Cl and ²³⁶U were measured by using CIAE AMS based on HI-13 accelerator. The CIAE AMS can be used to measure ²⁶Al, ³⁶Cl, ⁴¹Ca, ⁵⁵Fe, ⁶⁴Gu, ⁷⁹Se, ⁹⁹Tc, ¹²⁶Sn, ¹²⁹I, ¹⁵¹Sm, ¹⁸²Hf and ²³⁶U in the fields of biomedicine, nuclear physics, nuclear astrophysics, geosciences, nuclear environment engineering, and environmental science^{[16,17,18,19,20,21,22]}. In the recent years, a new beam line was granted to the upgraded CIAE AMS system^[25], as shown in Fig.3. The ³⁶Cl and ²³⁶U was measured as following.

The ²³⁶U was measured on AMS beam line 1, as described by Wang and Jiang et al.^[22]. Negative UO^– ions were extracted from the ion source by accelerating voltage of about 7.240 MV, and ²³⁶U¹¹⁺ ions were selected after acceleration. Because the isobar ²³⁶U is not expected to measure, it is the key how to identify the interest of ²³⁶U ions from isotope ²³⁸U and ²³⁵U. The energy difference between the ²³⁸U and ²³⁵U background ions and the ²³⁶U interested ion is 0.8% and 0.4%, respectively. On happening to pass the ion optical filters, the time of flight (TOF) detector has sufficient resolution to distinguish ²³⁶U from the isotopes ²³⁸U and ²³⁵U. A sensitivity is 10⁻¹¹ lower for ²³⁶U/²³⁸U than for ²³⁶U due to the good time resolution of ~500 ps^[24], high terminal voltage of HI-13 accelerator and the relatively small energy straggling of the flying ions.

Fig.3Full-screen View

Schematic representation of the accelerator mass spectrometer based on the HI-13 accelerator at the CIAE.

3.1 ³⁶Cl

It is difficult to heighten sensitivity of the medium- heavy mass nuclides because of the limitation of beam time and measurement conditions, thus resulting in insufficient difference of energy losses and ineluctable energy straggling of medium-heavy radioisotopes. A new beam line equipped with a ∆E-Q3D detection system has been installed on the HI-13 tandem (AMS beam line 2)^[23] (Fig.3). The new beam line is equipped with some ion-optical elements, a large Q3D magnetic spectrometer, SBD and multi-anode ionization chamber. The ³⁶Cl is measured by AMS beam line 2.

The Cl⁻ ions were selected to inject into the HI-13 Tandem Accelerator. Its typical output beam current at the Lower Energy Faraday Cup (LEFC) was about 3 μA. The Cl⁻ ions were accelerated by the tandem terminal voltage which was typically set at 11.0 MV. The carbon foil (3 μg/cm²) was employed to break up the Cl⁻ ions and produce atomic ions with high charge states. The resulting Cl⁺ ions were further accelerated by the same terminal voltage. A 90° double focusing High Energy Analyzing Magnet (HEAM) with a maximal mass energy product of 200 MeV·amu was used to select ³⁶Cl⁸⁺ (and ³⁶S⁸⁺) with 99.11 MeV.

None of processes can separate ³⁶Cl from its stable isobar ³⁶S. After switching magnet, the particles were transported to the ΔE-Q3D system to separate ³⁶Cl from ³⁶S and detect ³⁶Cl. A homogeneous Si₃N₄ membrane of 3.0-μm thickness was mounted at the entrance of Q3D as an absorber to produce different energy losses of ³⁶Cl with ³⁶S. After passing through the Si₃N₄ membrane, these ³⁶Cl⁸⁺ and ³⁶S⁸⁺ ions have 81.48 MeV and 82.64 MeV, and the charge state of 14⁺ were analyzed by the Q3D magnetic spectrometer. The ³⁶Cl and ³⁶S were separated by their residual energies. The peak distance between ³⁶Cl and ³⁶S on the focal plane was about 81 mm with a separation factor of 2.4 (Fig.4), which is defined as the ratio of the peak difference to the full width of half maximum (FWHM). In order to increase the detection efficiency and further separation of isobaric interferences, a multiple-anode ionization chamber (MAIC, four anodes in this work) with an entrance window of 100 mm×40 mm Mylar foil was accurately mounted at the Q3D focal plane, thus recording the interested nuclide ³⁶Cl. The ionization chamber was filled with propane of 4.2 kPa. By choosing the suitable magnetic field, a suppression factor of ³⁶S ions about 10⁵ was achieved, while most of the ³⁶Cl ions are recorded by the MAIC.

Fig.4Full-screen View

Position spectra of ³⁶Cl and ³⁶S on the Q3D focal plane.

4 Results

Figure 5 shows the TOF spectra and their two- dimensional spectra vs. energy for uranium mineral samples, distinguishing the interested nuclide ²³⁶U from the ²³⁸U and ²³⁵U has sufficient resolution.

Fig.5Full-screen View

TOF spectra and TOF vs. the energy two-dimensional spectra for ²³⁶U. (1) For a standard sample with ²³⁶U/²³⁸U =5.00×10⁻⁹ in 100 s counting time, (b) for sample with ³⁶Cl/Cl=2.03×10⁻¹⁰ in 300 s counting time.

Figure 6 shows two-dimensional spectra of E_r vs. E_t for ³⁶Cl standard and one uranium sample (E_r and E_t are the energy loss signals from anodes 4 and the total energy, respectively), indicating that the Q3D method can identify ³⁶Cl from its isobar ³⁶S clearly. These results suggest that the sample preparation method for the ³⁶Cl and ²³⁶U in the uranium minerals were satisfied for the AMS measurement.

Fig.6Full-screen View

Two-dimensional spectra of E_r vs. E_t for ³⁶Cl standard and one sample (E_r and E_t are the energy loss signals from anodes 4 and the total energy, respectively). (a) For a standard sample with ³⁶Cl/Cl=1.60×10⁻¹¹ in 300 s counting time, (b) for sample with ³⁶Cl/Cl=6.8×10⁻¹¹ in 300 s counting time.

The measurements of natural ³⁶Cl and ²³⁶U from the same uranium mineral sample as Fig.6 are listed in Table 1. The uncertainties for isotope ratios include (1σ) counting statistics, reproducibility of the measurement, and a systematic contribution from the measurement relative to reference materials.

5 Discussion

The ³⁶Cl yield in samples depends on these reactions including ³⁵C1(n,γ)³⁶CI, ⁴⁰Ca(n,2n3p)³⁶C1, ³⁹K(n,2n 2p)³⁶Cl, ⁴⁰Ca(μ⁻, γ)³⁶Cl, ³⁹K(n,α)³⁶Cl and ³⁹K(μ⁻,2np) ³⁶Cl. Relative to the σ(³⁵C1(n,γ)³⁶CI) ~ 43b, the cross section for ⁴⁰Ca(μ^–, α) ³⁶Cl(σ~4.3 mb) and ³⁹K(μ^–, 2np) ³⁶Cl (σ ~5 mb) is small, thus the contribution from the μ⁻ induced reaction is generally negligible. The ²³⁶U yield in samples depends on the ²³⁵U(n,γ)²³⁶U reaction with the large thermal neutron of σ~8.3b. Therefore, the natural yield of two isotopes depends on the neutron yield in the uranium mineral and the fractions of the absorbed reaction elements.

The neutron was produced by (α,n) reactions on light elements such as B, Na, Mg and Al, and its yield depends on the major element abundances in the sample. The elemental compositions of the samples are listed in Table 2.

Also, the neutron yield depends on the ²³⁸U spontaneous fission besides the uranium is the principal source of α-particles. Two points for the neutron absorption should be taken into account. (1) The neutron energy thermalization process in the rock is essential to produce the ³⁶Cl and ²³⁶U because their reaction cross section excitation functions vary with energy. (2) The concentrations of trace neutron absorbers such as B, Gd and Sm. Also, the ratios of ³⁶Cl/Cl and ²³⁶U/²³⁸U calculated with the MCNP is similar to the works of Fabrka-Martin et al.^[25] and Wilcken et al.^[8]. The MCNP calculations assume an infinite, homogeneous mineral body with an elemental composition is equal to each individual mineral. Individual neutron with starting 1–2 MeV are tracked by the matrix using Monte-Carlo methods; and the probability of scattering or absorption by various elements at each step.

In Table 1, the MCNP calculation results were basically consistent with that of AMS measurement. More ³⁶Cl and ²³⁶U experimental data conduce to understand the natural nuclides produced in the uranium mineral and the nucleogenic ²³⁶U as a reference data for the background. This is important for the practical application of ²³⁶U in environmental and geological research.

6 Conclusions

In this paper, the ³⁶Cl and ²³⁶U in the same uranium mineral are measured by AMS at CIAE. The ratios ³⁶Cl/Cl and ²³⁶U/²³⁸U were studied. The measurement method of ³⁶Cl and ²³⁶U in the same uranium provides for extending the potential application of ²³⁶U in uranium mining, environmental and geological research.