OPTIMIZATION AND CONFIGURATION OF A HYDROCARBON GAS REMOVAL SYSTEM FOR THE DETERMINATION OF δ2H AND δ18O ON PICARRO CRDS

Methane and other gaseous hydrocarbons cause interference in the spectroscopic measurement of stable water isotopes. A study has been undertaken to assess and quantify the use of various pre-treatment options for stripping C1-C6 gases in samples and include: 1) catalytic oxidation using CuO and high temperature via a tube oven; 2) coupling a Nafion gas drying system to remove/separate water vapour from hydrocarbon gas; 3) activated carbon system to absorb C1-C6 hydrocarbons. Initial testing was undertaken using a tube oven operated at three temperatures (850, 925, 1000° C) with samples containing 5 concentrations of methane (0 to 4.4%) as well as 5 concentrations of C1-C6 mixture (0 to 5%-CH₄) which were vapour equilibrated with two isotopic reference standards (SK-T and BIR). Water isotope measurements (δ₂H and δ₁₈O) were recorded on a Picarro L2120i. Preliminary results show a 99-100% reduction of methane and all C1-C6 hydrocarbons at 925° C and 1000° C. The used of the CuO catalyst does not provide reliable δ18O values due to additional oxygen entering into the vapour stream. C1-C6 mixtures appear to fluctuate more than just CH₄ in vapour in the less depleted waters. CH₄ only vapour appears to plot closer to the standard water values for δ₂H, for SK-T specifically, isotope measurements are approximately 16‰ more depleted, and for BIR standard waters, about 10 ‰ more enriched. This worsens with the addition of C1-C6 gasses, with SK-T measuring 20-60 ‰ more depleted and BIR measuring 20 ‰ more depleted after concentrations exceed 2.5% (CH₄). Further refinement of this in in progress. Pretreatment of vapour was also performed with a PermaPure Nafion gas drying tube at four temperatures (70, 80, 90 & 100° C). The Nafion tube removed 98-100% of the methane but only provided a peak concentration of water vapour of approximately 800 ppmv, far below the requirement for a stable isotope measurement on Picarro. Measured isotope values plotted inconsistently with no correlation to the actual standard water value and consequently this method was considered unacceptable. The final pre-treatment option consisted of an activated carbon inline filter. The activated carbon was successful in absorbing methane, but also absorbed all water vapour as well, causing a null measurement on the Picarro and due to an unacceptably low H₂O-ppmv of 300.