Improving Particle-Phase Nitrate Measurement in PM2.5 Filter Sampling: Evaluation of a Denuder–Nylon Filter Modification in the SPARTAN Network

Accurate measurement of PM2.5 composition is of global importance. Evaporation-induced mass loss introduces potential bias in filter-based PM2.5 measurements, with ammonium nitrate volatilization from Teflon filters being a major contributor to negative mass artifacts. Previous studies have demonstrated that using an acid gas denuder in combination with nylon filters can help recover the nitrate mass. This study evaluates design modifications to the AirPhoton sampling station setup within the globally distributed Surface Particulate Matter Network (SPARTAN) incorporating an additional acid gas denuder upstream and a nylon filter downstream of the existing Teflon filter. Two co-located AirPhoton Sampling stations were installed on the roof of an academic building at Washington University in St. Louis. One of the stations used the design modifications and one did not. The design modifications were also installed in SPARTAN stations worldwide. The post sampling analysis for the Teflon filters remained unchanged. Ion Chromatography (IC) was used for laboratory measurement of nitrate on nylon filters. Adding the design modification increased nitrate and ammonium collection in St. Louis by an average of 60% and 4% during the spring 2024. The design modifications showed no significant impact on the Teflon filter PM2.5 mass and the major ionic species composition. Global deployment of the denuder–nylon modification revealed substantial variability in nitrate partitioning, with higher nitrate fractional recovery on nylon filters more common in warmer climates and during summer months. Temperature emerged as the dominant factor influencing nitrate fraction on nylon filters, while correlations with relative humidity, daily fluctuations, and aerosol chemical characteristics—including cation-to-anion balance and dust composition—were weak. Post-sampling nitrate stability in St. Louis was assessed using paired filters with and without immediate refrigeration. Most measurements fell within detection bounds, indicating limited evaporative loss under the observed conditions. Teflon-associated nitrate showed the most evidence of loss, while PM2.5 mass remained largely stable with only minor deviations. The findings suggest that under moderate climate and nitrate conditions, post-sampling volatilization is a minor contributor to overall measurement uncertainty, though losses may be different in warmer or more polluted environments.