Measurements of keV neutron capture cross sections with a 4π barium fluoride detector: Examples ofNb93,Rh103, andTa181

A new experimental method has been implemented for precise measurements of neutron capture cross sections in the energy range from 3 to 200 keV. Neutrons are produced via the ${}_{}{}^7{}_{}{}^{}\mathrm{Li}$(p,n${)}^7$Be reaction using a pulsed 3-MV Van de Graaff accelerator. The neutron energy is determined by the time-of-flight (TOF) technique using flight paths of less than 1 m. Capture events are detected with a 4π barium fluoride detector. This detector is characterized by a resolution in gamma-ray energy of 14% at 662 keV and 7% at 2.5 MeV, a time resolution of 500 ps, and a peak efficiency of 90% at 1 MeV. Capture events are registered with ∼95% probability above a gamma-ray threshold of 2.5 MeV. The combined effect of the relatively short primary flight path, the 10-cm inner radius of the detector sphere, and of the low capture cross section of ${\mathrm{BaF}}_2$ shifts the background due to capture of sample scattered neutrons in the scintillator to later TOF and therefore leaves the high-energy portion of the TOF spectrum undisturbed. The high efficiency and good energy resolution for capture gamma-rays yields a further reduction of this background by using only the relevant energy channels for data evaluation. In the first measurements with the new detector, the neutron capture cross sections of ${}_{}{}^{93}{}_{}{}^{}\mathrm{Nb}$, ${}_{}{}^{103}{}_{}{}^{}\mathrm{Rh}$, and ${}_{}{}^{181}{}_{}{}^{}\mathrm{Ta}$ were determined in the energy range from 3 to 200 keV relative to gold as a standard. The cross-section ratios could be determined with overall systematic uncertainties of 0.7 to 0.8%; statistical uncertainties were less than 1% in the energy range from 20 to 100 keV, if the data are combined in 20-keV wide bins. This represents an improvement of factors 5–10 compared to existing experimental methods. The necessary sample masses were of the order of one gram. Maxwellian averaged capture cross sections were calculated in the temperature range relevant for s-process studies. Severe discrepancies were found compared to the data reported in literature.