Expanding the scope and applicability of laser-based spectroscopy to studies of ecohydrology by removing organic contaminants in natural water

Abstract:

The ability to measure the stable isotopic composition of plant and soil waters, surface waters and ambient atmospheric vapor is essential to understanding an ecosystem’s water budget, including how water cycles between the air, plants and the subsurface. With the advent of laser-based spectroscopy, e.g., Cavity Ring-Down Spectroscopy (CRDS), the isotopic analysis of waters has become increasingly cost-effective and prevalent, with comparable precision to conventional isotope ratio mass spectrometry methods. However, early work [1,2] demonstrated that the accuracy of isotopic analysis by laser-absorption techniques could be compromised by the spectral interference from organic compounds, in particular methanol and ethanol [1], which are prevalent in ecologically-derived waters.

Here we present results from the Picarro Micro-Combustion Module (MCM), which acts to destructively remove these interfering organic species from the analyzed water vapor stream by oxidizing them to CO₂ and H₂O. Analyzed samples include simulated plant water solutions, waters doped with varying concentrations of potentially problematic organic compounds, and actual plant water extracts. We find that the median offset between IRMS and a Picarro L2130-i outfitted with a MCM is less than 0.5 ‰ for δ¹⁸O and less than 1 ‰ for δD. In parallel to the destruction of organic contaminants, a software tool can also be used to assess the probability of spectral interference. This software performs a statistical analysis of spectral fit parameters, e.g., the shift in the spectral baseline, and compares unknown samples to clean standard waters.

In general, the most common primary metabolites present in plant materials include the light organic acids, e.g., benzoic and formic acid. At low concentrations (0.1 and 1%) formic acid does not appear to interfere with the resolved absorption spectra for H₂O, HDO and H₂¹⁸O. Similar tests will be conducted for benzoic acid. Conversely, although methanol and ethanol are only present in trace amounts in plants, these alcohols can cause large interferences even at the low concentrations (1% and 0.025% for ethanol and methanol, respectively). Using these results, we will propose when CRDS for ecologically-derived waters functions best.

[1] Brand et al. (2009), RCM, 23

[2] West et al. (2010), RCM, 24