An examination of k-factor in TEM-EDS analysis for grain boundary segregation in neutron-irradiated stainless steels

Grain boundary (GB) segregation in stainless steels due to neutron irradiations is considered to be one of the factors causing irradiation assisted stress corrosion cracking (IASCC); therefore, accurate quantitative evaluations are required to understand the IASCC behavior. GB segregation is typically evaluated by using transmission electron microscope (TEM) and energy dispersive X-ray spectroscope (EDS). To quantify the concentration of each element, the Cliff-Lorimer method is generally used, which assumes that the intensity ratio of the characteristic X-rays is proportional to the concentration ratio of the elements. There are some methods for determining the proportionality constant for each element, the "k-factor", however, there are only a few studies about the effect of the difference in k-factor determination methods on quantified values of GB segregations. In the present study, the extent of the effect was investigated by comparing the elemental concentrations quantified with k-factors determined by three methods with respect to TEM-EDS data obtained for a GB of a baffle former bolt that had irradiated to a dose of 18.8 dpa (displacements per atom). For Cr, Mn, and Mo, there was not much difference for the GB composition and the amount of GB segregation even if the k-factor determination methods were different, while for Ni and Si, the effect of the difference in k-factor determination methods was relatively large. The change in the amount of GB segregation due to the difference in the k-factor determination methods was smaller than the reported variation in the amount of GB segregation of Cr and Ni; therefore, the effect of the difference in the k-factor determination methods on the quantified values of GB segregation was confirmed to be small. It was also found that the method to correct each k-factor using the chemical composition of the material being analyzed is the most convenient and the most accurate for quantifying GB segregation among the three k-factor determination methods. (author)