Accurate nuclear data for neutron-induced reactions are essential for the design of nuclear transmutation system. However uncertainties of nuclear data such as Minor actinides (MA) do not fulfill requirement for the design of transmutation facilities. Measurement of the neutron capture cross section is ongoing at the Accurate Neutron Nucleus Reaction Measurement Instrument (ANNRI) in the Materials and Life Science Experimental Facility of the Japan Proton Accelerator Research Complex (J-PARC). The determination of the absolute values of cross sections is one of the main causes that affect the final uncertainties of cross section results.

In the present work, we suggest two new methods to reduce systematic uncertainties of capture cross section measurement. Both methods employ change of the self-shielding effect with sample rotation angle. The first method is for thickness determination of a B-10 sample which is used for measurement of the incident neutron spectrum. In capture cross section measurements in ANNRI, we place B-10 sample and determine the incident neutron spectrum by counting the 478 keV γ-rays from the B-10(n,γ)Li-7 reaction. The uncertainty of the B-10 sample thickness, more precisely the area density that is usually calculated from the mass and the area, introduces the uncertainty of the shape of the incident neutron spectrum. In this method, the B-10 sample is tilted with respect to the neutron beam direction, thereby changing the effective thickness. The neutron self-shielding effect increases with the effective thickness. This results in change of the shapes of time-of-flight (TOF) spectrum of 478 keV γ-ray counts from the B-10(n,γ)Li-7 reaction with the tilted angle. Comparing the difference of the TOF spectra at different angles and assuming the 1/v energy dependence of cross section of the B-10(n,γ)Li-7 reaction, the area density of the B-10 sample can be determined without using sample mass and area.

The second method suggested in this work is for absolute determination of the capture cross section. For absolute determination of capture cross section. a typical and often-used technique is the saturated resonance method that is based on the fact that the number of neutron capture events becomes equal to the number of the incident neutrons at a large resonance when the sample is very thick and the resonance is fully saturated. The principle of our method is similar to the saturated resonance method but this method does not require a fully saturated resonance. In this method, the neutron self-shielding factor is determined from change of resonance peak height with sample tilted angle. The self-shielding factor at the resonance peak is used for normalization to give the absolute capture cross section.

Theoretical and experimental studies on the methods are ongoing. Calculations using Monte Carlo simulation code PHITS were carried out to study the feasibility of the present methods. Test experiments using a sample rotation system at ANNRI were also performed. We used Np-237 sample in the measurements. Preliminary results will be given in this contribution.