Optimisation of a batch thermal combustion method using a tube furnace oxidation system (pyrolyser) and LSC for carbon-14 determination in environmental matrices.

Accelerator mass spectrometry and benzene synthesis coupled with liquid scintillation spectrometry are often used for accurate measurements of C activity in the environmental matrices. Thermal oxidation is one of the methods employed for C determination in environmental matrices. In this method, the sample is oxidised at high temperature (600-900 °C) to convert carbon species to CO and trapped in an amine-based absorber for determining the activity in a liquid scintillation counting (LSC) system. In this study, the performance of a commercially available tube furnace system (pyrolyser), for batch combustion of samples, was evaluated for the determination of C specific activity in terrestrial biota samples. Significant improvements over the manufacturer specified method, which is primarily designed for analysis of samples with activity well above the environmental background level, was implemented to achieve accurate determination of C specific activity at ambient background level. In the improved method, the CO produced from the combustion of the sample was isolated from the combustion products through cryogenic trapping and then absorbed in the absorber (Carbo-Sorb E) through a simple off-line transfer process. This allowed (i) optimisation of CO absorption by the absorber (2.2477 g of CO/10 mL), (ii) achieving good accuracy and precision in the measurements, and a minimum detectable activity value of 13 Bq kgC for a counting time of 300 min (7 Bq kgC for 1000 min), (iii) avoiding uncertainty associated with the determination of recovery of C in the combustion and trapping process, and (iv) elimination of the need for an independent determination of carbon content (%) for expressing the results in terms of C specific activity. The method is capable of yielding accurate results with a deviation of <2.4% from the target value for IAEA C3 quality assurance reference material (with a relative standard deviation of 1.40%, and relative error of 0.34%). The combined uncertainty (1σ) associated with the measurements was computed to be 3.4%. Upon optimisation, the suitability of the method for the determination of C specific activity in typical terrestrial biota samples of clean air region (region not affected by local anthropogenic sources) and for the quantification of a small increase in the C activity above ambient levels in the vicinity of a nuclear power plant is demonstrated.