부산대학교 도서관

In environmental samples, the major analytical limitation on the use of uranium {sup 238}U/{sup 235}U determinations as an indicator of uranium enrichment is mass dependent bias occurring during the measurement. The double-spike technique can be used to correct the data for this effect. The purpose of this experiment was to evaluate the variation of mass bias among several laboratories and to determine the extent to which the double-spike could be used to reduce analytical uncertainty. Four laboratories performed replicate analyses on each of three samples. Generally mass bias was determined to be small compared to the random scatter of the measurements, but in at least one case, the bias was > 1%. In 8 of 12 cases, intra-laboratory variance was reduced when the double-spike correction was applied. For all three samples, the inter-laboratory variance was decreased, though the decrease was small. Based on a reasonable assumption about the true isotopic compositions of the samples, the accuracy of 11 of the twelve analyses was improved by applying the double spike correction. When the double spike is used to correct for mass bias, the {sup 238}U/{sup 235}U accuracy is better than 1% even for samples as small as 1 ng. For 50 ng of uranium, 0.1% accuracy was achieved.