Intramolecular N2O isotopic composition using laser spectrometers: Correction functions, uncertainty budget, freeze-thaw events and source process identification

N₂O isotopic composition, i.e., δ¹⁵N-N₂O, δ¹⁸O-N₂O and especially site preference (SP; difference of substitution frequencies at terminal or central position in N-N-O molecule) has been shown to provide information on N₂O source processes, and allows for source partitioning of N₂O emissions to nitrification and denitrification. The advent of laser spectrometers more than a decade ago has spawned first datasets of N₂O isotopic composition in daily resolution, but they have remained scarce. This is because until recently, the precision of commercially available spectrometers did not allow direct determination of N₂O isotopic composition without technically challenging liquid nitrogen free cryogenic preconcentration of N₂O. The specifications of the latest commercially available spectrometers promised preconcentration free in-situ determination of N₂O isotopic composition, but a recent instrument intercomparison showed that for most of the analyzers, specific correction functions are still necessary. While some available instruments were thoroughly characterized with regard to short term precision, repeatability, drift, amount effects, matrix effects and spectral interferences, instrument performance during field deployment and on the time scale of long measurement campaigns has not been analysed so far.

Here we present a setup and results of an automated chamber system in conjunction with a laser spectrometer that was installed in the field and in use for a period of approx. two years. Initially, amount dependence was in the range of 4 to 2 ‰ ppm N₂O^-1 for the various isotopic species, but instrument optimizations reduced this dependence to less than 1 ‰ ppm N₂O^-1. CH₄ dependence was constant through the whole period and in the range of 1 to 2 ‰ ppm CH₄^-1, with affecting only δ¹⁵N^α and δ¹⁸O. In contrast, CO₂ dependence was variable and in the same range as N₂O amount dependence. The uncertainty budget was dominated by instrument noise, calibration and N₂O amount dependence, indicating that improvements of instrument precision and availability of more suitable reference materials have a high potential to further decrease uncertainty of measurements. Analysis of the effect of uncertainty on the error of determined soil air N₂O isotopic composition based on Keeling plots resulted in an error of 2 ‰ and 1 ‰ at N₂O concentration increases of 70 and 140 ppb, respectively. Consequently, source partitioning based on SP will be associated with an error of 17 and less than 12% at the moment. Compared to growing-season emissions, SP and δ¹⁸O-N₂O during freeze-thaw cycles were distinctly different. SP was ~0, indicating that N₂O reduction to N₂ was negligible during freeze-thaw events.