ABSTRACT

ICP-AES analyses of REE in two GSJ carbonate reference rocks of JDo-1 and JLs-1 have been made, and then compared with reported INAA results. Both of our ICP-AES and reported INAA results for JDo-1 are in reasonably good agreement, whereas most of the INAA analyses of Ce, Nd, Sm and Lu for JLs-1 are higher than our ICP-AES values by factors of 4 to 20. The contrasting results between JDo-1 and JLs-1 are related to their greatly different U/REE ratios: JLs-1 has 20–100 times higher U/REE ratios than JDo-1. The apparent disagreement in JLs-1 reveals the significance of fission and spectral interferences due to coexisting ²³⁵U and ²³⁸U in INAA determination of REE. We have estimated the fission interferences for ¹⁴⁰La, ¹⁴¹Ce, ¹⁴⁷Nd, and ¹⁵³Sm from theoretical calculations. Our estimates agree reasonably with the reported experimental data for the fission interferences. The fission interference in ¹⁴¹Ce can explain the difference between INAA and ICP-AES data of Ce for JLs-1. The extent of fission interference in ¹⁴⁷Nd is estimated. to be almost the same as that in ¹⁴¹Ce. Even the fission interference in ¹⁴⁰La becomes significant, when samples with U/La ratio greater than unity are irradiated for 8–12 hours and then cooled for more than several days. This must be a reason for that most of INAA values for La in JLs-1 are about 50% higher than our ICP-AES value. Most of the INAA analyses of Sm and Lu for JLs-1 are biased to about ten times higher values by the γ-ray interferences of ²³⁹Np produced from ²³⁸U. Because of the great potential of fission and spectral interferences in INAA for geochemical samples with high U/REE ratios, we should be careful for their Ce anomalies and REE patterns based on such INAA data.