Systematic low bias of passive samplers in characterizing nitrogen isotopic composition of atmospheric ammonia
Yuepeng Pan,Mengna Gu,Linlin Song,Shili Tian,Dianming Wu,Wendell W. Walters,Xingna Yu,Xuemei Lü,Xue Ni,Yanjun Wang,Jing Cao,Xuejun Liu,Yunting Fang,Yuesi Wang
DOI: https://doi.org/10.1016/j.atmosres.2020.105018
IF: 5.965
2020-10-01
Atmospheric Research
Abstract:<p>Identifying the sources of increased ammonia (NH<sub>3</sub>) in the atmosphere has important implications for air quality, ecosystems, and climate. While analysis of the nitrogen isotopic composition (δ<sup>15</sup>N) can aid in the apportionment of NH<sub>3</sub> emission sources, the widely-used passive collection methods have not been verified for their ability to accurately characterize δ<sup>15</sup>N-NH<sub>3</sub>. In this study, three types of passive samplers (ALPHA, Analyst, and Radiello) were employed and collocated with a denuder based active sampler (DELTA, serves as a reference method) to compare and contrast their ability in collecting atmospheric NH<sub>3</sub> for subsequent concentration and δ<sup>15</sup>N-NH<sub>3</sub> analysis in urban Beijing during the summer of 2018. The results showed that NH<sub>3</sub> concentrations determined using passive collection samplers that included ALPHA (<em>n</em> = 11), Analyst (<em>n</em> = 12), and Radiello (<em>n</em> = 7) samplers were 13.4%, 14.4%, and 27.5% lower than that of the active sampler (<em>n</em> = 12), respectively. The low NH<sub>3</sub> concentration bias in the passive collection samplers was suggested to be the result of inaccurate effective sampling rates due to incorrect mass transfer correction factors for the environmental conditions of this study. The δ<sup>15</sup>N-NH<sub>3</sub> values did not show significant differences between passive samplers, with weekly means of −31.0 ± 1.0‰ (<em>n</em> = 4), −28.4 ± 5.5‰ (<em>n</em> = 4), and − 29.1 ± 0.9‰ (<em>n</em> = 4) for the ALPHA, Analyst, and Radiello, respectively. Notably, the overall mean δ<sup>15</sup>N-NH<sub>3</sub> values collected by these passive samplers (−29.5 ± 3.2‰, <em>n</em> = 12) was significantly lower than that of the active sampler (−14.1 ± 1.6‰, <em>n</em> = 4), with a difference of 15.4 ± 3.5‰. The large underestimation of δ<sup>15</sup>N-NH<sub>3</sub> values by passive samplers can be explained by the diffusive isotope fractionation of NH<sub>3</sub> isotopologues, which is driven by relative mass differences of <sup>14</sup>NH<sub>3</sub> and <sup>15</sup>NH<sub>3</sub>. Overall, these findings highlight a potential low bias in NH<sub>3</sub> concentrations and a substantial uncertainty in previous source interpretation using δ<sup>15</sup>N-NH<sub>3</sub> as previous studies have almost exclusively characterized δ<sup>15</sup>N-NH<sub>3</sub> using passive samplers. We suggest that the denuder-based active sampler should be used in the future for characterizing δ<sup>15</sup>N-NH<sub>3</sub> whether at the source or in an atmospheric environment with mixed sources. This work also has important implications for characterizing δ<sup>15</sup>N of other gaseous nitrogen compounds.</p>
meteorology & atmospheric sciences