Evaluating O₂ : Ar, N₂ : Ar, and ^29,30N₂ using membrane inlet mass spectrometry configured to minimize oxygen interference

Abstract Membrane inlet mass spectrometry (MIMS) provides detailed measures of dissolved 28,29,30 N 2 , O 2 , and argon (Ar) for estimating important gas fluxes and concentrations in aquatic ecosystems. Previous studies demonstrated a large O 2 scavenging effect while using a MIMS, where varying concentrations of O 2 can affect measured N 2 : Ar because O 2 interacts with N 2 in the ion source to produce NO + ( m / z = 30), potentially decreasing the detected current for 28,29 N 2 and increasing the detected current for 30 N 2 . A common solution is to use a muffle furnace heated to 600°C with a copper reduction column to reduce the concentration of O 2 to minimal levels and accurately measure 28,29,30 N 2 . However, this solution eliminates the detection of O 2 in environmental samples, which is a major benefit of using a MIMS. We questioned whether the MIMS was sensitive enough to provide accurate O 2 estimates when using the furnace and whether the O 2 scavenging effect was real and consistent among MIMS. We conducted four separate experiments on three different MIMS to test the O 2 scavenging effect and the potential detection of O 2 when using a MIMS with furnace. The furnace removed ~ 99% of O 2 , and O 2 scavenging had little to no detectable effect on N 2 : Ar and an unclear effect on 29 N 2 : 28 N 2 , but increased 30 N 2 : 28 N 2 . In most cases, accurate O 2 data could be retrieved despite using the furnace. The need for O 2 reduction may be limited to measuring accurate 30 N 2 : 28 N 2 in isotope pairing studies, but without substantial loss of MIMS measurements used to describe O 2 dynamics.