Recapturing sublimated ice core gas samples & implementing δ15N – N2 mass spectrometry measurements

Ice cores represent invaluable tools for palaeoclimatic reconstructions. Of particular interest is the gas trapped in air bubbles and clathrates within the ice. It serves as a direct record of the changes to atmospheric composition over the last several 100kyrs. Recently, the University of Bern established joint high precision greenhouse gas (GHG) analyses (CO₂, CH₄ & N₂O concentration, as well as δ¹³C – CO₂) on small ice core samples (15g) thanks to sublimation extraction and multi-beam quantum cascade laser spectrometry.

Although gas bubbles form a direct record of past atmosphere, they are not an unbiased record of past atmosphere. Several processes take place prior and during bubble formation which fractionate gas concentration and gas isotopes, requiring several corrections to be applied (especially for δ¹³C – CO₂ data). These corrections can be achieved by studying the isotope ratios of noble gases and N₂. However, this requires a mass spectrometry analysis which would normally call for the use of additional samples. This is problematic for two reasons: first, the corrections applied are most accurate when comparing the various isotope ratios of the same gas sample. Even ice in close proximity is subject to small scale disruptions within the ice and can skew the results. Second, near the bottom of ice cores, glacial thinning – compression of the ice caused by the overlying material – causes thousands of years of ice to be compacted in only one metre, making two different samples, even if adjacent, incomparable.

Hence, the objective of this study is to reuse gas samples which have already undergone laser- spectrometric GHG analysis to implement δ¹⁵N – N₂ & δ⁴⁰Ar analysis. We outline the development and building of an apparatus which reuses the previously extracted and analysed air and captures it for mass spectrometry. Sample recapture is achieved using a helium cryostat to cool dip tubes down to approximately 10K. When connected, the gas samples are drawn from the laser spectrometer into said tubes, where they are cryogenically trapped. To avoid interferences from other gases, the samples go through a liquid nitrogen trap to remove the already measured CO₂ and N₂O. The sample tubes are then disconnected, warmed up to room temperature and brought to an isotope-ratio mass spectrometer for major gas isotope analysis.

Despite significant fractionation of the isotopic composition of major gas compounds (N₂, O₂, Ar) during the recapture process, the developed method achieves its initial objectives, with a >99% recapture efficiency. It features a δ¹⁵N – N₂ reproducibility of standard gas measurements of ∼ 10 permeg and an offset of 0.1‰ and ∼ 30 permeg and 0.2‰ for δ⁴⁰Ar. With this reproducibility, sufficiently precise corrections of the gravitational enrichment of isotopes in the firn column are possible, however temperature reconstructions using thermodiffusion thermometry are not yet possible. Further improvements are thought to be possible to reduce the signal to noise ratio, as well as reducing the offset.