Stability and selectivity of pre-concentration methods for gaseous oxidized mercury in the air

Abstract. The determination of gaseous oxidized mercury (GOM, Hg^II) in the atmosphere at ultra-trace levels necessitates a preconcentration step. KCl-coated denuders and cation exchange membranes (CEMs) are widely used preconcentration methods for sampling Hg^II. Sampling losses during preconcentration could result in measurement biases. This study evaluated the performance of denuders and CEMs in retaining Hg^II, to accurately estimate biases through a precise mass balance approach using ¹⁹⁷Hg radiotracer and generation of specific Hg^II species by non-thermal plasma oxidation of Hg⁰ in the presence of reactant gases. The results showed that recovery of Hg^II by thermal decomposition from freshly prepared denuders were approximately 100 %. The retention of Hg^II on denuders during sampling period was much poorer compared to CEMs. Freshly prepared denuders lost up to 61.5 % of Hg^II, while reused denuders lost up to 79.7 % after 2 hours of ambient air exposure. CEMs exhibited much lower losses over 72 hours: on average, 16.7 % for HgO, 2.7 % for HgCl₂, and 3.8 % for HgBr₂. The standard digestion procedure for CEMs by BrCl digestion was found to be incomplete, with a few residual amounts detected on CEMs post-digestion (5.3 % on average). Additionally, the overall mass distribution within the CEM filter pack cartridges revealed that some Hg^II was retained on their inner Teflon parts. These findings underscore the critical importance of addressing biases in methods adopted for Hg^II measurements, to improve the accuracy and reliability of atmospheric Hg data, an integral component in evaluating Hg model outputs for Minamata convention effectiveness evaluation.