Oxygen interference with membrane inlet mass spectrometry may overestimate denitrification rates calculated with the isotope pairing technique

Membrane inlet mass spectrometry (MIMS) is used to measure gas concentrations in aqueous samples, and is often employed to measure N 2 in denitrification studies. While most denitrification studies using MIMS have employed the N 2 :Ar method, MIMS is increasingly used with the isotope pairing technique (IPT), an alternative method that calculates denitrification based on 29 N 2 and 30 N 2 concentrations, measured as mass to charge ratios ( m / z ) 29 and 30. Dissolved O 2 is known to interfere with m / z 28 and 40 ( 28 N 2 and Ar, respectively) in the MIMS, and in this study, we show that the effect on m / z 29 and 30 is also significant. This O 2 effect may result in artificially high denitrification rates with IPT if O 2 varies between samples. In lab‐based experiments, m / z 29 and 30 were impacted immediately as O 2 was depleted. The resulting effect on simulated denitrification rates was minor if final O 2 concentrations remained above 50% initial O 2 , but increased by more than 100 μ mol m −2 h −1 as O 2 fell below 50%. Similarly, denitrification rates from sediment incubations were an order of magnitude higher when analyzed in the presence of O 2 , compared with analysis without O 2 using an inline furnace with reduction column. We further show that this effect is variable over the life of the MIMS, and/or with different ion sources. Given these uncertainties, we recommend using an inline furnace to remove O 2 for all IPT experiments. Alternatively, the magnitude of the O 2 effect should be assessed often to determine the range of minimal O 2 interference.