Measurements from the AirSentry II were compared to simultaneous measurements of gas phase ammonia made with the traditional annular denuder method in order to assess the applicability of the AirSentry II for use at ambient conditions. To prevent particles from entering the IMS cell, a 2 micron Teflon filter was installed at the sample inlet. Additionally, the inlet tubing was heated downstream of the filter pack to prevent water from condensing and damaging the IMS cell. The denuder measurements were made with three channel glass annular denuders (URG Corp.) coated with a 2% solution of phosphorous acid and installed downstream of a PM2.5 cyclone. The 0.1 Hz AirSentry II data were averaged over the duration of each denuder sample period for comparison.
Seventy seven denuder samples were obtained, with ammonia concentrations ranging from 1.5 to 15 ppb. The average concentrations reported by the AirSentry II during these periods were well correlated (r = 0.93) with the denuder measurements. However, the AirSentry II values were generally higher than those reported by the denuders. The magnitude of disagreement increased with increasing ammonia concentration, reaching nearly 50% at the highest concentrations. System checks indicated that the calibration of the AirSentry II was valid. Time series data suggested that ammonium nitrate volatilization from the inlet filter may be responsible for the observed discrepancy. A new virtual impactor inlet that removes particles from the sample airstream is currently in development.