Neutron Filtering System for Neutron Capture Cross Section Measurement at the ANNRI beamline of MLF/J-PARC

. A neutron ﬁltering system has been designed and implemented in the Accurate Neutron-Nucleus Reaction Measurement Instrument (ANNRI) beamline in the Materials and Life Science (MLF) facility of the Japan Proton Accelerator Research Complex (J-PARC) to bypass the e ﬀ ect of the double-bunch mode of J-PARC by molding the incident neutron ﬂux into quasi-monoenergetic neutron beams. Filter assemblies using Fe, Si and Cr as ﬁlter materials were analyzed by means of experimental analysis, together with Monte Carlo simulations. The characteristics of the ﬁltered neutron beam are presented and discussed alongside its viability in future applications for neutron cross-section measurements in the keV neutron region.


Introduction
The Accurate Neutron-Nucleus Reaction Measurement Instrument (ANNRI) beamline at the Material and Life Science (MLF) experimental facility of the Japan Proton Accelerator Research Complex (J-PARC) features one of the most intense neutron beams available in the world. This allows for neutron-induced reactions, such as the neutron capture reaction, to be measured using tiny amounts of sample, even in the keV region where the neutron capture cross section is rather small. Nonetheless, the present operation pattern of the J-PARC accelerator hampers the energy determination for fast neutrons, making cross section measurements in the keV region unattainable. J-PARC is currently operated in double-bunch mode, where two proton bunches with a time difference of 0.6 µs are injected into the Hg spallation target to generate pulsed neutrons. This mode is aimed at increasing the thermal neutron fluence while reducing the thermal stress induced in the spallation target. Since most of the beamlines at the MLF use slow neutrons, the 0.6 µs time difference is negligible in comparison to the time-of-flight (TOF) for thermal neutrons. However, the double-bunch mode introduces serious ambiguities when determining the neutron energy by means of timing measurements for neutrons higher than 10 keV (TOF < 20 µs), since it is impossible to ascertain the originating proton pulse of each neutron. A neutron filtering system has been designed and implemented in the ANNRI beamline to bypass the aforementioned effects of the double-bunch mode as part of the project "Study on accuracy improvement of fast-neutron capture reaction data of long-lived Minor Actinides (MAs) * e-mail: gerard.rovira@jaea.go.jp for development of nuclear transmutation system" [1]. Materials with the characteristic of sharp minima in the neutron total cross section were employed as neutron filters in order to produce quasi-monoenergetic neutron beams.
In this paper, a brief overview of the neutron filtering system at ANNRI is provided. The experimental setup for the neutron filtering system is presented in Sec. 2 followed by a description of the characteristics of the filteredneutron beams derived from both experiments and simulations in Sec. 3. Finally, results of 197 Au neutron capture cross section measurements to validate the present system are shown and discussed in Sec. 4 with the conclusions following in Sec. 5.

Setup of the neutron filter
In order to bypass the effects of the double-bunch mode in the keV region, quasi-monoenergetic beams were molded in the ANNRI beamline by using materials as neutron filters with the characteristic of sharp minima in the neutron total cross section. The neutron filter apparatus was introduced in the rotary collimator of the ANNRI beamline. This apparatus consisted of several 5-cm-thick cylindrical metallic slabs of filter material that were inserted to obtain the desired thickness, as shown in Fig 1. The position of the neutron filter within the ANNRI beamline can be seen in the vertical section view of AN-NRI in Fig. 2. Fe, Si and Cr were considered in the present work in order to create quasi-monoenergetic neutron beams with expected averaged neutron energy around 24 keV (Fe), 54 and 144 keV (Si), and 46 and 136 keV (Cr). Configurations with thicknesses of 20 cm for Fe and Si, and 15 cm for Cr were experimentally tested in separate measurements to accurately determine the characteristics of the incident filtered-neutron flux using the NaI(Tl) spectrometer of the ANNRI beamline. The time distribution of the incident filtered-neutron flux was deduced using the 10 B(n,αγ) 7 Li reaction in a measurement of a 90%-enriched 10 B sample with a diameter of 10 mm, a thickness of 0.5 mm and a mass of 86.3 mg. This technique is commonly used in the ANNRI beamline since this reaction emits a sole γ-ray with the energy of 478-keV and, thus, events from this reaction can easily be isolated. The measurement of a carbon sample, having also a diameter of 10 mm and thickness of 0.5 mm, was employed, together with a no-sample (blank) measurement, to determine the sample-dependent and sample-independent backgrounds, respectively. Finally, the neutron flux was obtained by removing the energy-dependent influence of the 10 B(n,αγ) 7 Li reaction from the measured reaction yield using the simulation results from PHITS [2]. This correction also removes the influence from other sample-related effects such as self-shielding and multiple scattering. Further information about the neutron capture experimental setup can be found elsewhere [3].

Filtered-neutron beams
The neutron time distributions of the filtered neutron flux using the filter configurations of 20 cm for Fe and Si, and 15 cm foe Cr were determined in measurements using the NaI(Tl) spectrometer of the ANNRI beamline. The expected peaks with neutron energies around 24 keV (Fe), 54 and 144 keV (Si), and 46 and 136 keV (Cr) were clearly observed in the experimental results. However, due to the double-bunch effect, the neutron energy distribution within the peaks cannot be accurately inferred from the experimental time distribution. Thus, simulations were required to complement the experimental results in order to accurately determine the characteristics of the filtered neutron beams. The neutron transmission ratios for each filter configuration were derived from simulations with PHITS and were broadened using the resolution function of AN-NRI, determined in the work of Kino et al. [4], to obtain two-dimensional histograms with the neutron energy and time. Figure 3 shows a comparison between the experimental and simulated neutron time distributions for each filter configuration  The very good agreement between the experimental and simulated results validates the present simulation results for both the neutron time and energy, as they were simultaneously calculated. This is very important since the simulations were used to determine the neutron en-

Neutron capture cross section measurements of 197 Au
The final validation of the neutron filtering system implemented in the ANNRI beamline was performed by measuring the neutron capture cross section of 197 Au using the characteristics of the filtered neutron beams determined in sec 3. For the present experiments, a 197 Au sample with a diameter of 10.0 mm and a thickness of 0.1 mm. The sample mass amounted to 153.9 mg with an area density of 5.99 × 10 −4 at/b. The sample was thick enough to completely saturate the first resonance and was used to normalize the incident filtered neutron flux with each filter configuration. The neutron capture cross section was determined using the TOF gates described in Sec. 3 for each filtered peak as: where < σ Au (E g ) > is the averaged neutron capture crosssection for each of the five TOF gates (E g ), Y Au (E g ) and C(E g ) are the neutron capture yield and the correction coefficient for self-shielding and multiple-scattering for each TOF gate (E g ) obtained from PHITS simulations, respectively. φ n (E g ) stands for the normalized incident neutron flux at the TOF gate (E g ) and n Au is the sample area density in at/b. The 197 Au neutron capture cross-section results for each filter configuration are shown in Fig. 4 in comparison with the evaluated data from JENDL-5 [5] and the IAEA neutron data standards [6], and the recent reported data from Lederer et al. [7] and Massimi et al. [8]. The present cross-section results are consistent with the experimental data by Lederer et al. and Massimi et al., and the evaluated nuclear data values from JENDL-5 and IAEA neutron data standards. Consequently, the present agreement validates the neutron filtering system implemented at the ANNRI beamline in order to accurately measure neutroninduced reaction in the keV region bypassing the effects of the double-bunch mode.

Conclusions
A neutron filtering system has been designed and implemented at the ANNRI beamline to accurately measure keV-neutron-induced reaction bypassing the effects of the double-bunch mode. The present system consists of a series of neutron filter arrays made of Fe, Si and Cr that al- low for the incident neutron flux to be molded into quasimonoenergetic neutron flux. The characteristics of the neutron filter arrays of 20 cm for Fe and Si, and 15 cm for Cr were determined by means of experimental analysis together with simulations using the PHITS code. The present filter configurations are able to tailor beams with averaged neutron energies of 23.5 keV (Fe), 51.5 and 127.7 keV (Si), and 45.0 and 133.4 keV (Cr). The present system was validated in 197 Au neutron capture cross section measurements by the good agreement of the present results with both the latest experimental data and the evaluated nuclear data from JENDL-5 and the IAEA neutron data standards. Meaning that, the present neutron filtering system is a suitable solution to accurately measure keV-neutron-induced reactions avoiding the effect of the double-bunch mode in the incident neutron flux.