δ2H and δ18O analyses of solid, liquid and gaseous samples
High-Temperature Conversion – Isotope Ratio Mass Spectrometry (HTC-IRMS)
Isotopic ratios of oxygen and hydrogen are analysed on Ianosch (Isotope Analyser for Oxygen and Hydrogen), a Delta+XL that is coupled to either a TC/EA furnace from Thermo Fisher, or to a similar furnace produced by Hekatech. The former is equipped with a liquid autosampler, while the latter is set up for solid samples via a Costech Zero Blank autosampler. The quantitative catalytic conversion of water or organic compounds is achieved in a glassy carbon reactor at temperatures above 1350 °C (Gehre et al., 2004). The products of this reaction are H2 and CO gas that are consequently analysed. The reactor contains a tube in tube design consisting of a ceramic outer, and a glassy carbon inner tube.
This design ensures a stable carrier gas flow and leads to a lower CO background (Gehre et al., 2004). The conversion of an analyte to H2 gas is not quantitative if the analyte contains halogens, nitrogen or sulphur (Gehre et al., 2017; Gehre et al., 2015; Nair et al., 2015). When analysing hetero-atom containing analytes, elemental chromium needs to be used as reactor filling instead of glassy carbon. Gaseous samples can be analysed by connecting a 10-Port valco valve equipped with two 1 ml sample loops to a needle which is pushed into the injector septum. That way gases (e.g. CH4 and N2O can be injected manually into the HTC.)
The BGC-IsoLab employs three internal standards (working, scaling and QC) in a daily δ2H/δ18O water run consisting of 48 samples and 69 standards. These standards are calibrated against Vienna Mean Ocean Water (VSMOW) and Standard Light Antarctic Precipitation (SLAP) so that all measurement results are given on the VSMOW/SLAP scale. Solid samples are analysed in the same fashion. Here the internal standards are standardised using SMOW and SLAP water in silver capsules that is available at the Reston Stable Isotope Laboratory (USGS). The raw data undergoes a standardised evaluation to bring all data on the VSMOW/SLAP scale. This evaluation is described in Gehre et al. (2004).