Reproducibility and accuracy of noble gas measurements on water samples in the microlitre range

RATIONALE In order to derive meaningful noble gas temperatures (NGTs) from speleothem fluid inclusions, precise and accurate measurements of noble gas concentrations on very small water samples are necessary. To optimise these measurements and their reliability, an investigation of the reproducibility and accuracy of the analytical procedure is essential. METHODS Water equilibrated with air under controlled conditions was filled into copper capillaries to produce small (about 1 μL) air-equilibrated water samples (so-called μAEWs). As speleothem samples, the μAEWs were opened in a crusher and the released noble gases analysed with an electron ionisation sector field mass spectrometer run in static mode. For better comparability with speleothem samples, a defined amount of air was added to the noble gases derived from one group of μAEWs. RESULTS The reproducibility of the experimental procedure was found to be better than 2.2% for all relevant noble gases. Within these uncertainties, the measured noble gas concentrations of μAEWs agree with expectations. The corresponding NGTs reproduce within 0.5°C and deviate by less than 1°C from the equilibration temperature. In the case of air addition, the air to water ratio was determined accurately and, excluding one outlier, the NGTs were determined with a reproducibility and accuracy well below 1°C. CONCLUSIONS The results show that the used measurement procedure is generally suitable to reveal major climatic temperature changes, for which an overall NGT error of less than 1°C is desirable, from very small water samples as obtained from speleothem fluid inclusions. Copyright © 2013 John Wiley & Sons, Ltd.