Successive and automated stable isotope analysis of CO2, CH4 and N2O paving the way for unmanned aerial vehicle‐based sampling

Rationale Measurement of greenhouse gas (GHG) concentrations and isotopic compositions in the atmosphere is a valuable tool for predicting their sources and sinks, and ultimately how they affect Earth's climate. Easy access to unmanned aerial vehicles (UAVs) has opened up new opportunities for remote gas sampling and provides logistical and economic opportunities to improve GHG measurements. Methods This study presents synchronized gas chromatography/isotope ratio mass spectrometry (GC/IRMS) methods for the analysis of atmospheric gas samples (20‐mL  glass vessels) to determine the stable isotope ratios and concentrations of CO2, CH4 and N2O. To our knowledge there is no comprehensive GC/IRMS setup for successive measurement of CO2, CH4 and N2O analysis meshed with a UAV‐based sampling system. The systems were built using off‐the‐shelf instruments augmented with minor modifications. Results The precision of working gas standards achieved for δ13C and δ18O values of CO2 was 0.2‰ and 0.3‰, respectively. The mid‐term precision for δ13C and δ15N values of CH4 and N2O working gas standards was 0.4‰ and 0.3‰, respectively. Injection quantities of working gas standards indicated a relative standard deviation of 1%, 5% and 5% for CO2, CH4 and N2O, respectively. Measurements of atmospheric air samples demonstrated a standard deviation of 0.3‰ and 0.4‰ for the δ13C and δ18O values, respectively, of CO2, 0.5‰ for the δ13C value of CH4 and 0.3‰ for the δ15N value of N2O. Conclusions Results from internal calibration and field sample analysis, as well as comparisons with similar measurement techniques, suggest that the method is applicable for the stable isotope analysis of these three important GHGs. In contrast to previously reported findings, the presented method enables successive analysis of all three GHGs from a single ambient atmospheric gas sample.


| INTRODUCTION
There is an increased awareness of the anthropogenic impact on climate change. Identifying the sources and sinks of greenhouse gases (GHGs) and monitoring their atmospheric abundance 1 are essential in managing the required global GHG emission reductions to achieve the 1.5 C target 2 or to establish pathways to zero emissions. 3 GHG monitoring is a valuable input to facilitate technology improvement, leading to more efficient resource utilization. The most important GHGs are carbon dioxide (CO 2 ), methane (CH 4 ) and nitrous oxide (N 2 O). For 2018 the global mean concentrations of CO 2 , CH 4 and N 2 O were 407.8 ± 0.1 μmol mol −1 , 1869 ± 2 nmol mol −1 and 331.1 ± 0.1 nmol mol −1 , respectively. 4 In light of their low absolute concentrations, instrumentation has to be used that will ensure accurate and precise measurements. In addition to the compound concentration, the stable isotopic composition enables identification of GHG sources and sinks. [5][6][7][8][9] However, the inclusion of isotopic analysis requires different considerations and an appropriate analyzer, such as an isotope ratio mass spectrometer, or optical devices based on tunable diode laser adsorption spectroscopy or Fourier transform infrared spectroscopy. 10 While optical devices are customized to analyze specific compounds only (e.g. CO 2 , H 2 O or N 2 O), isotope ratio mass spectrometry (IRMS) can be used to measure a multitude of compounds in addition to GHGs. The key process in efficient and accurate IRMS is sample preparation, which encompasses the various steps from specifically designed sample collection and manual sample preparation, through to introduction of the gases into the device and treatment options for specific gas separation.
Most IRMS systems use gas chromatography (GC) to separate and isolate gas compounds of similar physical and chemical behavior (e.g. CO 2 and N 2 O). Furthermore, pre-concentration steps connected in series with GC are often needed, if injection volumes (e.g. 20 mL to 2.5 L) 11,12 yield compound amounts (e.g. of CH 4 , N 2 O) below IRMS detection limits, which are in the range of hundreds of picomoles to nanomoles. A common pre-concentration approach is the use of cryogenic traps, filled with adsorbent material to trap CH 4 or N 2 O while other residual compounds are vented away. In this setup the entire gas contents of the vessel, whatever the volume, are purged out though the trapping system with a carrier gas such as helium. Pre-concentrated CH 4 held on the trap can then be released and oxidized to CO 2 , while trapped pre-concentrated N 2 O requires no chemical transformation for detection. Both gases then undergo similar preparation to that for atmospheric CO 2 measurements. That is, compounds are focused and separated from any residuals using GC and transferred to the isotope ratio mass spectrometer, which measures the intensity of m/z 44, 45 and 46 to calculate the stable isotope ratios of carbon, oxygen or nitrogen of sampled CO 2 , CH 4 or N 2 O.
There are already numerous methods available to measure atmospheric samples for the concentration and isotopic composition of CO 2 , CH 4 and N 2 O. [11][12][13][14][15][16][17][18] While the majority of published methods focus on one of the three GHGs only, the aim of the study reported here was to enable the measurement of all three gases from identical sample vessels in a single-push measurement approach. Specifically, taking into consideration the compatibility of the sampling approach with unmanned aerial vehicle (UAV)-based sampling systems, the key for success was the identification of appropriate sample vessels fitting the requirements of both the sampling system and the measurement setup. To our knowledge there is no comprehensive GC/IRMS setup for the successive measurement of CO 2 , CH 4 and N 2 O meshed with a UAV-based sampling system. This paper presents a method for automated GC/IRMS based on simultaneous and/or successive measurement of atmospheric CO 2 , CH 4 and N 2 O provided by a single air sample. Therefore, sample vessels have been identified fitting the requirements of UAV-based sampling systems and GC/IRMS instrumentation. Such a UAV-based sampling system was also designed and tested, but will be presented elsewhere. In accordance with Schauer et al, 17 the aim was to employ off-the-shelf instruments needing minor modification only, so that the presented methods can be considered by the scientific community as an alternative to specially designed instruments. It should also be kept in mind that, after adjusting the current measurement system, it is still ready to use for its ordinary purpose of analyzing a large variety of other gaseous and volatile compounds.
The presented methods were customized to the current abundance status of CO 2 , CH 4 and N 2 O in the atmosphere. The detection range for atmospheric CO 2 and CH 4 was established at 372 to 944 μmol mol −1 and 1.7 to 5.0 μmol mol −1 , respectively. 19 Appropriate limits of determination for tracing atmospheric gases are recommended at 100 μmol mol −1 for CO 2 and 500 nmol mol −1 for CH 4 20 in Schuyler and Guzman. 21 For N 2 O, a limit of determination of 300 nmol mol −1 was the target to guarantee the measurement of current atmospheric global mean abundance.

| MATERIALS AND METHODS
In the process of setting up the methods described below, there have been many process iterations with several discarded options, in terms of sample vial specifications, sample transfer, GC columns, adsorbents and cryogenic trapping. These are summarized in the supporting information.

| Sample vial preparation
Air was sampled in 20-mL headspace vials (La-pha-pack GmbH, Langerwehe, Germany) sealed with grey butyl-PTFE-lined septa (DWK Life Sciences GmbH, Mainz, Germany) and aluminum crimp caps. 22,23 The vials were pre-conditioned by flushing with helium or synthetic air for 1 min at an inlet pressure of 2 bar before evacuation for 1 min with  were calculated as:

| Sampling of CH 4 and N 2 O via purge and trap GC/C-HTC/IRMS
where R is the ratio of the abundance of 13 C to 12 C, 18 O to 16 Figure S1 (supporting information) for individual data points of measured working gas standards and presented in Table 1, giving a summary of grouped data points.

| CH 4 and N 2 O
Internal calibration of CH 4  T A B L E 2 Results of CH 4 and N 2 O working gas standard measurements: n tot represents the total number of measurements, and n is the number of measurements inside a 95% confidence interval. Nominal concentrations are indicated by c nom in μmol mol −1 and measured concentrations (c mean ) are calculated as arithmetic means of μmol mol −1 with their standard deviation (SD). Delta values are given as arithmetic means in ‰ with their SD. For N 2 O, the column S vol indicates the total sample volume, which can represent either a single vial volume (20.5 mL) or the volume of three vials (61.5 mL)

CH 4 N 2 O
c nom n tot n c mean ± SD RSD (%) δ 13 C ± SD c nom S vol n tot n c mean ± SD RSD (%) δ 15 N ± SD  Figure S2 (supporting information

| CH 4
Sampling of the atmosphere for CH 4    concentration and isotopic ratios. It was therefore shown that successive quantification and stable isotope analysis of all three GHGs in a single ambient atmospheric gas sample can be accomplished by modifying a purge and trap autosampler connected to a GC/C-HTC/IRMS system. Given that, the envisaged UAV-based air sampling system can be used to sample the atmosphere for GHGs. Such a system facilitates sampling campaigns at hard-to-access areas and enables automated sampling by remote control.