Measuring C-enriched CO in air with a cavity ring-down spectroscopy gas analyser: Evaluation and calibration.

Rationale: Cavity ring-down spectroscopy (CRDS) is becoming increasingly popular for δ C-CO analysis of air. However, little is known about the effect of high C abundances on the performance of CRDS. Overlap between CO and CO spectral lines may adversely affect isotopic-CO CRDS measurements of C-enriched samples. Resolving this issue is important so that CRDS analysers can be used in CO flux studies involving C-labelled tracers.

Methods: We tested a Picarro G2131-i CRDS isotopic-CO gas analyser with specialty gravimetric standards of widely varying C abundance (from natural to 20.1 atom%) and CO mole fraction (xCO : <0.1 to 2116 ppm) in synthetic air. The presence of spectroscopic interference between CO and CO bands was assessed by analysing errors in measurements of the standards. A multi-component calibration strategy was adopted, incorporating isotope ratio and mole fraction data to ensure accuracy and consistency in corrected values of δ C-CO , x CO , and x CO .

Results: CRDS measurements of x CO were found to be accurate throughout the tested range (<0.005 to 100 ppm). On the other hand, spectral cross-talk in x CO measurements of standards containing elevated levels of CO led to inaccuracy in x CO , total-xCO (x CO  + x CO ), and δ C-CO data. An empirical relationship for x CO measurements that incorporated the C/ C isotope ratio (i.e. CO / CO , RCO2) as a secondary (non-linear) variable was found to compensate for the perturbations, and enabled accurate instrument calibration for all CO compositions covered by our standard gases.

Conclusions: C-enrichement in CO leads to minor errors in CRDS measurements of x CO . We propose an empirical correction for measurements of C-enriched CO in air by CRDS instruments such as the Picarro G2131-i.