A Rapid and Portable DIC Analysis for Aquatic Systems: Rise of the Blanks.

Abstract:

Dissolved inorganic carbon analysis, in both fresh water and marine systems, can help determine carbon sources, sinks and flows through an ecosystem (Apongwa et al. 2013). Methods to measure DIC with small quantities in situ are becoming more important as the need for DIC estimation increases across aquatic ecosystems. Recently a number of papers have measured both δ¹³C_VPDB and concentrations by injecting a relatively small amount of sample, e.g. 1 to 2 mL, into prepared sample tubes and then analyzing the headspace gas (Spötl 2004, Torres et al. 2005, Assayag et al. 2006). The initial sampling can be done in the field and samples are stable for weeks to months (Capasso et al. 2005, Taipale and Sonninen 2009) prior to analysis. However, CO₂ gas samples can have a contaminant interfering with measurements when concentrations are low, e.g. 0.04% CO₂ (Knohl et al. 2004), though the effect can be negligible at higher concentrations, e.g. 1% (Spötl 2005).

We investigate fitting a blank correction to a suite of standards to quantify the contamination and more accurately measure the DIC concentration and isotopic values. We examined 6 and 18 month time points, using the suite of standards with δ¹³C_VPDB of 18.69, -2.69 and -16.86 ‰, 2 to 3 mM concentration range. Fitting the blank correction allows us to detect the blank in the 6 month time point. The blank only has a minor effect on the -2.69 and -16.86 ‰ standards, <0.1 ‰, but a significant one for the 18.69 ‰ standard, ~0.4 ‰. Samples run ca. 2 mM in the range of 0 ‰ only see a blank effect in the range of 0.15 ‰, but samples run at < 1 mM show potential shifts of up to 0.8 ‰. The 18 month test showed the tubes are under vacuum—apparently the He diffuses thru the chlorobutyl rubber septa. Those samples can still be analyzed, either by charging them with He so they will not be under vacuum, or running as is and letting the vacuum in the tubes draw in air during the analysis. After correcting for the blank, we see a loss of CO₂ from the vials, ca. 15%, and a shift in the per mil values. We show the DIC measurements can be robust over time, if a full suite of standards prepared with the sampling tubes are used for proper corrections.