Accuracy and precision for measurements of the mass ratio 30/28 in dinitrogen from air samples and its application to the investigation of N losses from soil by denitrification.

Abstract In the 1950s Hauck introduced a special version of the (15)N dilution technique ((15)N flux method) for the determination of N losses from the soil by denitrification. Although this method is very useful and reliable its application has been rather infrequent up to now. This is mainly due to the need to measure the m/z 30 in addition to the usually measured m/z 28 and 29 for dinitrogen, because the (15)N in the enriched air sample taken from an enclosure (cover box) at the soil surface is nonrandom. The signal from the m/z 30 is very low and difficult to measure with sufficient precision because other species (e.g. NO) also having the m/z 30 often interfere with its measurement. In this study the accuracy and precision of an easy to use CF-IRMS with sample batch operation to measure the ratio 30/28 was investigated. The relative standard deviation (RSD = precision) from natural abundance up to 2 at.% was always <1%. After correction of the mass ratio 30/28 (R30), by means of a formula obtained by linear regression of theoretical R30 against measured R30, the accuracy of the abundance calculated from this corrected R30 was very high. From the achieved precision and assuming a cover box height of 10 cm (headspace volume of 7 1), and a collection time of 2 h, a limit of detection for N(2) losses by denitrification equivalent to 16 g N/ha*d or 6 kg N/ha*a can be estimated. The performance of the (15)N dilution method using the equipment and procedure described is demonstrated by means of results from an incubation experiment with [(15)N]nitrate-amended soils.