REPOSITORY > RESULTS

Abstract

Atmospheric mercury (Hg) plays an important role in the biogeochemical Hg cycle. Hg species in the atmosphere are present in very low concentrations. While gaseous elemental mercury (GEM) is relatively inert, gaseous oxidized (GOM) and particulate-bound (PBM) mercury are highly reactive. These characteristics of atmospheric Hg species make the speciation measurements challenging, and instruments measuring atmospheric Hg species have been shown to be subject to bias and uncertainty. Most of the challenges originate from the sampling and calibration of the measurement instrumentation. Different sampling and calibration methods exist for atmospheric Hg, but their validation is often lacking or non-existent at ambient concentration levels. Additionally, the results obtained by different methods are commonly not comparable due to the lack of measurement traceability and uncertainty evaluation.
Traceability of GEM calibration has previously been established, though researchers still often use non-traceable GEM calibrations. Our first objective was to compare three different GEM calibration approaches: primary gas standard, calibration via reference material, and bell-jar calibration. The first two calibrations were traceable to the International System of Units (SI) and gave comparable results, while the latter calibration was not traceable to SI and gave statistically different results. Bell-jar calibration was shown to give 8% underestimated results compared to the SI traceable primary gas standard.
Sampling and calibration are the most challenging for reactive Hg species: GOM and PBM. Our second objective was to validate GOM sampling with KCl sorbent traps and KCl impinging solutions and an evaporative calibrator for GOM; validation was focused on ambient GOM concentration levels. Validation experiments at ambient concentration levels were mostly performed with the 197Hg radiotracer. The results showed that KCl sorbent traps are feasible for ambient GOM sampling, showing low HgII losses (under 5%) and high specificity (negligible retention of Hg0) under simulated sampling conditions. On the other hand, KCl impinging solutions were found to be unsuitable for ambient GOM sampling due to low specificity originating from Hg0 solubility and oxidation in the solution. The evaporative calibrator was not accurate and precise; its HgII output was concentration- and time-dependent; near-ambient HgII concentrations were the most problematic due to HgII adsorption. At the lowest HgCl2 concentration tested (5.90 ng m−3), the calibrator recovery (accuracy measure) was as low as 39.4%.
The results indicated that new GOM calibration methods are needed for ambient GOM concentrations. In our final work, we developed a calibration approach based on nonthermal plasma (NTP) oxidation of Hg0 to HgII species in the presence of a reaction gas. Validation work was done using the 197Hg radiotracer. The obtained oxidation efficiencies with the corresponding expanded standard uncertainty values were 100.5 ± 4.7% (k = 2) for 100 pg of HgO, 96.8 ± 7.3% (k = 2) for 250 pg of HgCl2, and 77.3 ± 9.4% (k = 2) for 250 pg of HgBr2. The presence of each species was confirmed by temperature-programmed desorption quadrupole mass spectrometry (TPD-QMS). Since mercury analyzers detect mercury in elemental form, we thermally reduced the produced HgII species to Hg0. The quantitative thermal reduction was achieved with the Al2O3 catalyst.
The 197Hg radiotracer was successfully applied for our work and was used for the first time for studies of atmospheric Hg. Radiotracer has been shown to be a more suitable validation tool than stable isotopes and isotope dilution methods due to its unique characteristics that allow validation for reactive Hg species at ambient concentration levels.