Abstract
This paper describes the source of variation in the SIMS determination of impurities in an insulating powder sample mixed together with graphite powders for electrical conductivity. The predominant factors were the local charge buildup on the sample surface and the local distribution of impurities originally present in the graphite powders. The local charge buildup due to the inhomogeneity of the tablet sample resulted in obvious changes of the secondary ion energy distributions. The local distribution of impurities in the graphite powders caused the occasionally anomalous high values in the determination.