Skip to main content

Validation of ammonia diffusive and pumped samplers in a controlled atmosphere test facility using traceable Primary Standard Gas Mixtures

Literature Reference
Peer Reviewed Literature

Nicholas A. Martina, Valerio Ferraccia, Nathan Cassidy, Josh Hook, Ross M. Battersby,
Elena Amico di Meane, Yuk S. Tang, Amy C.M. Stephens, Sarah R. Leeson, Matthew R. Jones,
Christine F. Braban, Linda Gates, Markus Hangartner, Jean-Marc Stoll, Paolo Sacco,
Diego Pagani, John A. Hoffnagle, Eva Seitlerg


We report the determination of ammonia (NH3) diffusive sampling rates for six different designs of commercial
diffusive samplers (CEH ALPHA sampler, Gradko diffusion tube, Gradko DIFRAM-400, Passam ammonia sampler,
and ICS Maugeri Radiello radial sampler (blue and white turbulence barriers)), together with the validation
test results for a pumped sampler (CEH DELTA denuder). The devices were all exposed in the UK's National
Physical Laboratory's (NPL) controlled atmosphere test facility (CATFAC). For each of the seven diffusive
sampler exposure tests there were traceable concentrations of ammonia (in the range 3–25 μgm−3) generated
under well-defined conditions of temperature, relative humidity and wind speed, which are applicable to a
variety of ambient monitoring environments. The sampler exposure time at each concentration was 28 days,
except for the radial devices, which were exposed for 14 days. The work relied on the dilution of newly developed
stable Primary Standard Gas Mixtures (PSMs) prepared by gravimetry in passivated gas cylinders as a
method of improving the metrological traceability of ammonia measurements. The exposed diffusive samplers
were sent blind to the participants for analysis and the reported NH3 concentrations were then compared against
the known reference concentration. From the results for each sampler type a diffusive sampling rate was calculated
and compared against the rate used routinely by the participants. Some measurement results were in
good agreement with the known traceable reference concentration (particularly for one diffusive sampler design
(ALPHA)), while other devices exhibited over-reading and under-reading (each with a clear bias). The new
diffusive sampling rates determined in the laboratory study were then applied to measurements in a field
comparison campaign, and this was found to deliver an improvement in agreement between the different devices