A simple setup for online laser spectroscopy gas isotope analysers in online chamber systems

Laser spectroscopy-based gas isotope analysers (LSIA) are cheaper in acquisition and maintenance but still lack the accuracy and precision available through isotope ratio mass spectrometry (IRMS). One of the applications where LSIA are particularly advantageous are online measurements that follow gas release over time. This is even more the case in labelling experiments, where requirements regarding isotope ratio precision are lower. Yet, experiments that implement such setups remain relatively rare.

Here, we present a simple, low-cost setup that conducts automated isotope ratio measurements in gases released from various materials. The setup consists of a LSIA instrument (Picarro G2201-i or G5131-i) which is connected to up to 16 measurement chambers using a VICI selector valve actuated by a Raspberry Pi which also records the measurement data and valve position. An auxiliary pump equipped with a needle valve is placed in parallel to the LSIA to regulate the total flow rate to 500 mL min^-1. Each chamber is connected to the analyser for 10 minutes, before switching to the next chamber. We therefore allow the target gas (CO₂, CH₄, or N₂O) to accumulate over 150 minutes between measurements in each chamber. During the measurement, chamber air is pulled to the analyser and replaced by ambient air. The analyte concentration therefore decreases during the measurement time, which allows us to calculate the source isotope value through the Keeling plot method. At the end of the measurement, the analyte concentration and isotope ratio is near ambient air, such that the chamber is reset for the next cycle.

We present and compare three different experiments that used this approach. First, we studied phloem transport rates in Beech trees. For this, we pulse-labelled branches with ¹³CO₂ and followed the release of ¹³CO₂ from stem respiration at different heights over time. Second, we studied the conversion of ¹³C-acetate label injected into intact peat cores into CO₂ and CH₄. We quantified the fractions of label recovered as CO₂ and CH₄ as well as the timing of label-derived gas emissions as a function of injection depths. Finally, we adjusted this setup for measurements of natural abundance isotope ratios in soil N₂O emissions to study N₂O source processes from permafrost soils. Our presentation will compare these implementations and report on the experience gained during their setup.