Probing the limits of sampling gaseous elemental mercury passively in the remote atmosphere

Reliably recording very low ambient concentrations of gaseous elemental mercury (GEM) in remote regions is often required, for example in the context of evaluating how effective the Minamata Convention is in reducing global Hg emissions. However, sampling over extended periods of time at sites that are difficult to access can be very challenging. In order to establish what role inexpensive and easy-to-use passive air samplers may play in this regard, we deployed a sampler using a Radiello diffusive barrier and activated carbon sorbent for periods of up to three years and with sampling periods ranging from one to three months in some of the most extreme, remote and challenging global environments: at Concordia station on the Antarctic plateau, on Amsterdam Island in the remote Southern Indian Ocean and at several sites on the tropical island of La Reunion. The ability to reliably record the GEM concentrations at these sites was strongly influenced by the size of the sequestered amount of mercury relative to the extent and variability of the contamination of field blank samples. In some cases, acceptably low and consistent field blank contamination could only be achieved by storing samplers in sealed glass jars during transport and storage. The size of the sequestered amount is easily increased by extending deployment times, and the experience of the current study suggests that deployment periods in excess of two months are advisable. Sampling in Antarctica was compromised by the extreme low temperatures, which caused unknown sampling rates, hoar frost accumulation, material failure and potential failure of storage seals. While good agreement with GEM concentrations measured with an active sampler was noted on Amsterdam Island, the passive sampler derived levels at the Maido Observatory in La Reunion were notably higher than concentrations measured simultaneously with a Tekran vapour analyzer, which was possibly related to sampling rates being temporarily elevated at the very beginning of deployment at low GEM concentrations.