|Abstract Title:||Evidence for Deciphering the Underestimation in Gaseous Oxidized Mercury Measurements by the KCl Denuder-Based Method|
|Presenter Name:||Lei Zhang|
|Session:||Special Session - New developments in understanding reactive mercury concentrations and chemistry|
|Day and Session:||Thursday 28th July - Session Four|
|Start Time:||15:00 UTC|
|Co-Authors:||Lei Zhang,Weikang Zhang,Guichen Zhang,Peisheng Zhou,Hangtian Zhou,Yu Zhao|
Abstract Information :
Accurate measurements of speciated atmospheric mercury (Hg), including gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate-bound mercury (PBM), are crucial to atmospheric Hg behaviors, especially the redox chemistry and the deposition process. The Tekran 2537/1130/1135 monitoring system has been the most widely used instrument for speciated atmospheric Hg observation worldwide. However, the KCl denuder-based design in the Tekran system for GOM could be hampered by humidity and O3 interferences which cause Hg loss and the resulting underestimation. To investigate the performance of the KCl denuder-based method in a polluted region, comparison between the Tekran system and the Reactive Mercury Active System (RMAS 2.0) developed by the University of Nevada, Reno (UNR) was conducted at a suburban site in eastern China as part of the Speciation and Transformation of Atmospheric Mercury in a Polluted region (STAMP) campaign.
In the RMAS configuration, two cation exchange membranes (CEMs) were utilized to measure GOM with a polytetrafluoroethylene (PTFE) membrane upstream capturing PBM, and two nylon membranes were used to sample reactive mercury (RM=GOM+PBM) with programmed thermal desorption for profiles of chemical compounds. Results from the method comparison show that the RMAS (CEM-P) GOM was 2.2 times that of the Tekran GOM (R2=0.54, with the correlation intercept fixed to 0). Assuming CEM has 100% GOM capture rate, the recovery rate of the KCl denuder-based Tekran system in this study was poorly correlated with humidity and O3, which was quite different from existing studies. Instead, the recovery rate was positively correlated with the ratio of PBM/GOM (measured by RMAS, R2=0.33, p=0.004). Moreover, when the proportion of [-Br/Cl] in RM based on deconvolution of the RMAS thermal desorption profiles was further introduced into the linear regression model as an independent variable, the R2 value rose to 0.51.
Here we propose a theory on the mechanism for the underestimation of GOM in the denuder-based method. The Hg loss on the inner wall of the denuder is probably due to the Hg reduction in the KCl matrix by the aqueous HO2 radicals which was a key reaction in chemical transport models. High O3 level indicates more HO2, and high humidity accommodates more HO2 to the aqueous phase, resulting in the observed low bias of GOM by Tekran. The positive effect of the PBM/GOM ratio on the GOM recovery rate in this study was likely related to the particles acting as a sink of HO2. The contribution of the Hg halides to the GOM recovery rate could be benefited from the enhancement of denuder capture capacity (to form [HgCl3]? and [HgCl4]2?).