In this paper, the authors describe a sensitive method for low-level non-exchangeable OBT determination. This methodology combines suitable sample treatment, a combustion apparatus for large-sized samples and low-background liquid scintillation spectrometry, along with precautions that substantially reduce the risks of sample contamination. Great care must be taken in the measurement of non-exchangeable OBT at environmental levels. Many authors have discussed the opportunities for cross-contamination between samples and contamination by exchange with the laboratory atmosphere. The authors also describe an application of the methodology to a large-scale sampling and measurement campaign, aimed at the determination of the environmental non-exchangeable OBT background level in tree leaves and ferns collected on the site and in the vicinity of a research centre located in the south-west of France, not far from Bordeaux. This study constitutes a 'zero level' for the non-exchangeable OBT activity, as, to our knowledge, there is no tritium source within or in the surroundings of the sampled area capable of producing non-exchangeable OBT activities above the natural levels. Our analyses showed that non-ex changeable OBT activities in the collected samples were very low, ranging from below the detection limit (ca 0.7 Bq kg(-1) of dry material) to ca 2 Bq kg(-1) of dry material. These values are similar to the natural tritium background measured in water samples. No discrepancies can be shown between fern samples and oak tree leaf samples or between samples collected inside and outside the research site. (C) 2003 Elsevier Science Ltd. All rights reserved.
