An oxygen standard for the determination of oxygen in steel by 14-MeV neutron activation analysis

Abstract

A relative method for the determination of oxygen in steel via the 16O(n,p) 16N reaction by means of 14-MeV neutrons is described. A standard is irradiated immediately behind the sample and the induced activities are counted simultaneously with two separate but identical detector systems. The standard mixture ( ca. 5 g of graphite plus iron oxide containing 80 mg of oxygen per g) is compressed into a steel capsule of the same external dimensions as the samples (26 mm diameter, 9 mm thick). Dimensional tolerances, choice and purity control of the oxygen compound and preparation of the standard mixture are discussed. Fast neutron shielding, absorption of fast neutrons, self-absorption of the 16N /gg-rays and the average neutron flux in sample and standards are considered and a total correction factor is established. Flux inhomogeneities and differences in counting geometry and discriminator setting can be determined by irradiation and counting of two identical standards. The accuracy of this method was checked by comparison of the results with those of the reducing fusion method; satisfactory agreement was observed, although the activation results tended to be slightly higher. The mean long-term standard deviation for analysis of a given sample over a period of 6 months was found to be ±3%.