Neutron capture cross section measurements of 120Sn, 122Sn and 124Sn with the array of Ge spectrometer at the J-PARC/MLF/ANNRI

Preliminary neutron capture cross section of 120 Sn, 122 Sn and 124 Sn were obtained in the energy range from 20 meV to 4 keV with the array of germanium detectors in ANNRI at MLF,J-PARC. The results of 120 Sn, 122 Sn and 124 Sn were obtained by normalizing the relative cross sections to the data in JENDL-4.0 at the largest 426.7-, 107.0- and 62.05-eV resonances, respectively. The 67.32- and 150-eV resonances for 120 Sn and the 579- and 950-eV resonances for 124 Sn which are listed in JENDL-4.0 and/or ENDF/B VII.1 were not observed.


Introduction
Accurate neutron capture cross section data for long-lived fission products (LLFPs) are required in the study of transmutation of radioactive waste [1]. One of the most important LLFPs is 126 Sn, which is included in spent fuels of light water reactors with relatively large yields and long half-life. However, only one experimental data set is available at the thermal energy [2]. Accurate cross section measurements of 126 Sn are strongly required.
It is expected that a 126 Sn sample for a cross section measurement is contaminated with a large amount of tin stable isotopes, 117−120,122,124 Sn, because these stable isotopes also have fission yields and the sample is normally prepared only through a chemical process from spent fuels. These isotopes have large influence on neutron capture cross section measurements of 126 Sn. Therefore, to obtain accurate cross section data for 126 Sn, the measurements of all tin stable isotopes had been started with Accurate Neutron-Nucleus Reaction measurement Instrument (ANNRI) of Materials and Life science experimental Facility (MLF) in Japan Proton Accelerator Research Complex (J-PARC). The results for 112 Sn and 118 Sn have been reported in ND2013 [3]. In this paper, results of the neutron capture cross section measurements of 120 Sn, 122 Sn and 124 Sn are reported in the neutron energy region from 20 meV to 4 keV.

Experimental procedure
Capture cross section measurements with neutron Timeof-Flight (TOF) method were performed with the array of Ge spectrometer in ANNRI. The array of Ge detectors is a e-mail: kimura.atsushi04@jaea.go.jp installed at the flight length of 21.5 m and is composed of two cluster-type Ge detectors, eight coaxial-type Ge detectors and anti-coincidence shields around each Ge detector described in Refs. [4] and [5]. The neutron intensity at the 21.5-m sample position is described in Ref. [6]. J-PARC is normally operated with "double-bunch mode", in which each proton pulse consists of two bunches (each with a width of 100 ns) at intervals of 600 ns [7]. The simulated resolution function at the 21.5-m sample position is described in Ref. [8].
In the measurements, two cluster-type Ge detectors were used, but the coaxial-type Ge detectors were not used because they suffered from severe electrical noise. The pulsed neutron beam was collimated to a 7 mm at the sample position. J-PARC was operated with a proton beam power of 270 kW and at a repetition rate of 25 Hz in the "double-bunch mode".
Samples were isotopically enriched metallic tin with a diameter of 5 mm. The weight of the 120 Sn, 122 Sn and 124 Sn samples was 68.7 mg, 99.7 mg and 88.2 mg, respectively. Isotopic distribution and chemical impurities of each sample are listed in Table 1. The samples were put in fluorinated ethylene propylene (FEP) film bag and attached to a polytetrafluoroethylene (PTFE) sample holder. The total measuring times for the 120 Sn, 122 Sn and 124 Sn samples were about 63, 30 and 32 hours, respectively. To deduce the background, measurements for a 208 Pb sample with a diameter of 5 mm, a weight of 159.7 mg, and an isotopic enrichment of 99.60 mole% and a sample holder with an empty FEP film bag (Blank) were also carried out during 16 and 22 hours. For dead-time correction, pulses from a random-timing pulse generator were fed to the pre-amplifier of every Ge crystal [9]. The data acquisition system in ANNRI has a typical dead time of 6 µs.

Data analysis
The analysis procedure was almost the same manner as that described in Ref. [3]. The frame-overlap backgrounds were estimated and subtracted in almost the same manner as that described in Ref. [10]. Because the beam duct and cases of the detectors are made of aluminum, main source of the TOF dependent were γ rays from 27 Al(n, γ ) reactions due to scattered neutrons. The TOF dependent backgrounds were estimated using the capture γ -ray yields for the empty FEP bag sample and the 208 Pb sample. The backgrounds were normalized using the areas of the photo peaks due to the 7724-keV γ rays from 27 Al(n, γ ) reactions.
Correction factors for neutron self-shielding and multiple scattering were calculated with the Monte Carlo simulation code MCNP [11]. In the calculation, the sample size, shape, mass, isotope abundances, and intensity distribution of neutrons were taken into consideration.
The neutron spectrum was measured by detecting the 478-keV γ rays emitted from the 10 B(n,αγ ) 7 Li reaction. In the measurement, clear photo-peak of the 478-keV γ rays were observed. The photo-peak efficiency for 478-keV γ rays was deduced using 60 Co, 152 Eu, 133 Ba, and 137 Cs standard isotopes. The cross section data for 10 B(n,αγ ) Li reactions were taken from JENDL-4.0 [12]. Using these data and results, the absolute neutron flux was obtained.
The energy dependences of the relative cross section for 120 Sn, 122 Sn and 124 Sn samples were deduced by using the TOF spectra with the dead time correction, the deduced backgrounds, the self-shielding and multiple scattering correction factors and the neutron flux. Results

Result
The results of neutron capture cross section for 120 Sn, 122 Sn and 124 Sn samples were obtained in the energy range from 20 meV to 4 keV. Because of "double-bunch mode", the structure appeared on the obtained cross section in the neutron energy range above 100 eV. Figures 1, 2  and 3 show the results for 120 Sn, 122 Sn and 124 Sn samples together with uncertainties due to statistical uncertainty and normalization uncertainty, values of JENDL-4.0 for T = 300 K (broadened with the resolution function [8]) and those with the impurities.
In Fig. 1, the 67.32-and 150-eV resonance were not observed. These resonances were reported by G.V. Muradyan [13] and are listed in ENDF/B VII.1 [14]. This result agreed with the result by P.E. Koehler [15] and evaluation in JENDL-4.0. The 579-and 950-eV resonances for 124 Sn were not observed. These resonances were reported by Yu.V. Adamchuk [16] and Fuketa [17], and are listed in both JENDL-4.0 and ENDF/B VII.1.
Pulse-height spectra gated at all resonances of the samples were obtained by subtracting off-resonance : The number of the events was not enough to observe photo peaks. ×: The resonance was not observed. b Resonances were overlapped. spectra from on-resonance spectra [4]. Many prompt γ -rays from 120 Sn, 122 Sn and 124 Sn are observed. The 1114-, 1747-and 2006-keV γ rays observed in the 122 Sn (n, γ ) reactions were previously unknown γ -rays. The other γ -rays were already reported by R.F. Carlton [18,19] and/or A.I. Egorov [20]. The origin of the resonances were decided using the gated spectra. For example, a photopeak of 273-keV γ rays was clearly observed at the 1.457-eV resonance of the 122 Sn sample. The 273-keV γ rays : The number of the events was not enough to observe photo peaks. ×: The resonance was not observed. b The prompt γ rays were observed clearly, but some resonances were overlapped.
was not prompt γ -rays from 122 Sn (n, γ ) reactions but those from 115 In (n, γ ) reactions. It was validated that the 1.457-eV resonance is one of the 115 In resonances. In the same manner, the other observed resonances were confirmed. The results and the observed resonance energies are listed in Tables 2 and 3, and along with the evaluated resonance energies in JENDL-4.0 and ENDF/B VII.1.
In Fig. 1, the 120 Sn sample was contaminated by In of 770 ppm, Sb of 450 ppm, Te of 5 ppm and Ag of 8 ppm. The 122 Sn sample was contaminated by In of 7.5 ppm as shown in Fig. 2 However, except for Ag in the 120 Sn sample, these values do not agree with the certificated values in Table 1. Unexpected contaminations by In, Sb and Te were found in the 120 Sn sample and that by In was found in the 122 Sn sample.

Summary
The preliminary neutron capture cross section of 120 Sn, 122 Sn and 124 Sn were obtained in the energy range from 20 meV to 4 keV with the array of germanium detectors in ANNRI at MLF/J-PARC. The results were obtained by normalizing the relative cross sections to the data in JENDL-4.0 at the largest resonances, respectively. The 67.32-and 150-eV resonances for 120 Sn and the 579-and 950-eV resonances for 124 Sn which are listed in JENDL-4.0 and/or ENDF/B VII.1 were not observed. Three new prompt γ -ray emissions were observed in the 122 Sn (n, γ ) reactions.