Improvements to the Accuracy of Atmospheric Oxidized Mercury Measurements

Traceable Calibration of Atmospheric Oxidized Mercury Measurements

Most previous measurements of oxidized mercury were collected using a method now known to be biased low. In this study, a dual-channel system with an oxidized mercury detection limit of 6-12 pg m-3 was deployed alongside a permeation tube-based automated calibrator at a mountain top site in Steamboat Springs Colorado, USA, in 2021 and 2022. Permeation tubes containing elemental mercury and mercury halides were characterized via an International System of Units (SI)-traceable gravimetric method and gas chromatography/mass spectrometry before deployment in the calibrator. The dual-channel system recovered 97 ± 4 and 100 ± 8% (±standard deviation) of injected elemental mercury and HgBr2, respectively. Total Hg permeation rates and Hg speciation from the gravimetric method, the chromatography system, the dual-channel system, and an independent SI-traceable measurement method performed at the Jožef Stefan Institute laboratory were all comparable within the respective uncertainties of each method. These are the first measurements of oxidized mercury at low environmental concentrations that have been verified against an SI-traceable calibration system in field conditions while sampling ambient air, and they show that accurate, routinely calibrated oxidized mercury measurements are achievable.

Application of traceable calibration for gaseous oxidized mercury in air

BACKGROUND

The current speciation methods for mercury (Hg) measurements are fraught with considerable uncertainty, from sample collection to calibration. High reactivity of gaseous oxidized Hg (GOM) species and their ultra-trace level presence makes them difficult to sample and calibrate. Given that improper calibration may lead to measurement biases, reliable and metrologically traceable calibration methods are required for accurately quantifying GOM in air. In the present study, we applied the recently developed calibration method based on non-thermal plasma oxidation of elemental Hg, to a commercially available Hg air speciation system for actual environmental measurements of GOM for the first time.

RESULTS

Hg species such as HgO, HgCl2, and HgBr2 were produced with trace amounts of reactant gases (oxygen and electrolytically produced chlorine and bromine) and the production was driven by plasma-assisted oxidation. The plasma oxidation efficiency of elemental Hg with oxygen was 98.5 ± 7.5 % (k = 2), while that for chlorine and bromine was 96.8 ± 6.9 % (k = 2) and 97.4 ± 9.6 % (k = 2), respectively. The calibration method was tested against the internal permeation (Hg0) source of the Tekran 2537B Hg analyzer on-field by loading HgO to different KCl-coated denuders using the plasma. GOM concentrations were measured using the Tekran speciation system. With internal calibration, concentrations were up to 9.1 % lower than those in plasma calibration, thereby emphasizing the importance of the calibration strategy. Measurement uncertainty (k = 2) further emphasizes this distinction. Internal calibration measurement uncertainty was 36.8 %, while plasma calibration boasted lower uncertainty at 13.8 %.

SIGNIFICANCE

The non-thermal plasma calibration strategy, as a unique and discrete calibration method traceable to the NIST SRM 3133 for ambient air GOM measurements, provide a higher level of confidence in the accuracy of GOM measurements with several advantages over other methods. Calibrations at extreme low concentrations (<100 pg) are possible with this method relevant to ambient air GOM concentrations.

Observations of the chemistry and concentrations of reactive Hg at locations with different ambient air chemistry

The Hg research community needs methods to more accurately measure atmospheric Hg concentrations and chemistry. The Reactive Mercury Active System (RMAS) uses cation exchange, nylon, and PTFE membranes to determine reactive mercury (RM), gaseous oxidized mercury, and particulate-bound mercury (PBM) concentrations and chemistry. New data for Atlanta, Georgia (NRGT) demonstrated that particulate-bound Hg was dominant and the chemistry was primarily N and S HgII compounds. At Great Salt Lake, Utah (GSL), RM was predominately PBM, with NS > organics > halogen > O HgII compounds. At Guadalupe Mountains National Park, Texas (GUMO), halogenated compound concentrations were lowest when air interacting with the site was primarily derived from the Midwest, and highest when the air was sourced from Mexico. At Amsterdam Island, Southern Indian Ocean, compounds were primarily halogenated with some N, S, and organic HgII compounds potentially associated with biological activity. The GEOS-Chem model was applied to see if it predicted measurements at five field sites. Model values were higher than observations at GSL, slightly lower at NRGT, and observations were an order of magnitude higher than modeled values for GUMO and Reno, Nevada. In general, data collected from 13 locations indicated that N, S, and organic RM compounds were associated with city and forest locations, halogenated compounds were sourced from the marine boundary layer, and O compounds were associated with long-range transport. Data being developed currently, and in the past, suggest there are multiple forms of RM that modelers must consider, and PBM is an important component of RM.

Comparison and calibration of methods for ambient reactive mercury quantification

Gaseous oxidized mercury (GOM) is the dominant form of atmospheric mercury (Hg) deposited and sequestered within ecosystems. Thus, accurate, calibrated measurements of GOM are needed. Here, two active membrane-based collection systems (RMAS) were used to determine GOM and particulate-bound Hg (PBM), as well as reactive Hg (RM = GOM + PBM), and compared with two dual-channel systems (DCS) and a Tekran 2537/1130 speciation system. The DCS measured operationally defined GOM by difference, using concentrations of gaseous elemental Hg (GEM) and total gaseous Hg. One DCS was linked to a custom-built, automated calibration system that permeated GEM, HgBr₂, or HgCl₂. The five systems were co-located for one-year to develop a dataset that would allow for understanding limitations of each system, and assessing measurement accuracy and long-term precision of the calibrator. The Tekran system measured ~14.5 % of the GOM measured by the other systems. The USU and UNR DCS and RMAS were significantly correlated, but the DCS was 50 and 30 % higher, respectively, than the RMAS. The calibrator performed consistently in the field and lab, and the DCS fully recovered GOM injected by the calibrator. Since the uncalibrated DCS measured the same concentrations as the calibrated DCS, they are both accurate methods for measuring RM and/or GOM. Some loss occurred from the RMAS membranes. SYNOPSIS: Accurate and calibrated measurements of atmospheric reactive mercury using membranes and two dual-channel systems.

Year-round trace gas measurements in the central Arctic during the MOSAiC expedition

Despite the key role of the Arctic in the global Earth system, year-round in-situ atmospheric composition observations within the Arctic are sparse and mostly rely on measurements at ground-based coastal stations. Measurements of a suite of in-situ trace gases were performed in the central Arctic during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition. These observations give a comprehensive picture of year-round near-surface atmospheric abundances of key greenhouse and trace gases, i.e., carbon dioxide, methane, nitrous oxide, ozone, carbon monoxide, dimethylsulfide, sulfur dioxide, elemental mercury, and selected volatile organic compounds (VOCs). Redundancy in certain measurements supported continuity and permitted cross-evaluation and validation of the data. This paper gives an overview of the trace gas measurements conducted during MOSAiC and highlights the high quality of the monitoring activities. In addition, in the case of redundant measurements, merged datasets are provided and recommended for further use by the scientific community.

Dispersion of airborne mercury species emitted from the cement plant

The cement industry is the second largest source of anthropogenic mercury (Hg) emissions in Europe, accounting for 11% of global anthropogenic Hg emissions. The main objective of this study was to examine the influence of Hg emissions from the Salonit Anhovo cement plant on Hg levels measured in the ambient air at Vodarna, 1 km downwind from the flue gas chimney. The findings reveal that the plant raw mill operational status plays an important role in Hg concentrations in the flue gas emitted from the plant. Emitted total gaseous mercury was, on average, higher (49.4 μg/m³) when raw mills were in the direct mode (both raw mills-off) and lower (23.4 μg/m³) in the combined mode (both raw mills-on). The average Hg concentrations in Vodarna were 3.14 ng/m³ for gaseous elemental mercury, 53.7 pg/m³ for gaseous oxidised mercury, and 41.9 pg/m³ for particulate bound mercury for the whole measurement period. Atmospheric Hg speciation in Vodarna, coupled with plant emissions and wind data, has revealed that the total gaseous mercury emitted from the cement plant is clearly related to all Hg species measured in Vodarna. Wind blowing from the northeastern quadrant (mostly NE, ENE) is responsible for the elevated Hg levels in Vodarna, where gaseous oxidised mercury levels are highly linked to the cement plant emissions. However, elevated levels of Hg species in the absence of northeastern winds indicate potential inputs from other unknown local sources as well as inputs from regional and global transport mechanisms.

Calibration Sources for Gaseous Oxidized Mercury: A Review of Source Design, Performance, and Operational Parameters

Mercury is a neurotoxin that, unlike many localized industrial pollutants, spreads globally through atmospheric transport. Mercury in the atmosphere is operationally partitioned into gaseous elemental mercury (GEM), gaseous oxidized mercury (GOM), and particulate-bound mercury (TPM). Although GOM makes up only a small fraction of Hg in the free troposphere under normal conditions, its role in the dry and wet deposition of mercury makes GOM a significant species for understanding the transport and fate of mercury in the atmosphere. Although instruments for atmospheric mercury speciation are commercially available, significant uncertainty is associated with the current speciation methods, from sample collection to calibration, for GOM measurements. This paper examines the custom-made calibration sources that have been developed for GOM measuring instruments, evaluates the factors influencing the source performance, and synthesizes recommendations for the design and operation of GOM calibration sources in the future.

Evidence against Rapid Mercury Oxidation in Photochemical Smog

Mercury pollution is primarily emitted to the atmosphere, and atmospheric transport and chemical processes determine its fate in the environment, but scientific understanding of atmospheric mercury chemistry is clouded in uncertainty. Mercury oxidation by atomic bromine in the Arctic and the upper atmosphere is well established, but less is understood about oxidation pathways in conditions of anthropogenic photochemical smog. Many have observed rapid increases in oxidized mercury under polluted conditions, but it has not been clearly demonstrated that these increases are the result of local mercury oxidation. We measured elemental and oxidized mercury in an area that experienced abundant photochemical activity (ozone >100 ppb) during winter inversion (i.e., cold air pools) conditions that restricted entrainment of air from the oxidized mercury-rich upper atmosphere. Under these conditions, oxidized mercury concentrations decreased day-upon-day, even as ozone and other pollutants increased dramatically. A box model that incorporated rapid kinetics for reactions of elemental mercury with ozone and OH radical overestimated observed oxidized mercury, while incorporation of slower, more widely accepted reaction rates did not. Our results show that rapid gas-phase mercury oxidation by ozone and OH in photochemical smog is unlikely.

Elevated Gaseous Oxidized Mercury Revealed by a Newly Developed Speciated Atmospheric Mercury Monitoring System

Gaseous oxidized mercury (Hg²⁺) monitoring is one of the largest challenges in the mercury research field, where existing methods cannot simultaneously satisfy the measurement requirements of both accuracy and time precision, especially in high-particulate environments. Here, we verified that dual-stage cation exchange membrane (CEM) sampler is incapable of gaseous elemental mercury (Hg⁰) uptake even if particulate matter is trapped on CEM, whereas the Hg²⁺ capture efficiency of the sampler is more than 90%. We then developed a Cation Exchange Membrane-Coupled Speciated Atmospheric Mercury Monitoring System (CSAMS) by coupling the dual-stage CEM sampler with the commercial Tekran 2537/1130/1135 system and configuring a new sampling and analysis procedure, so as to improve the monitoring accuracy of Hg²⁺ and ensure the simultaneous measurement of Hg⁰, Hg²⁺, and Hg_p in 2 h time resolution. We deployed the CSAMS in urban Beijing in September 2021 and observed an unprecedented elevated Hg²⁺ during the daytime with an average amplitude of 510 pg m^-3. Using a zero-dimensional box model, the elevated Hg²⁺ production rate was attributed to high atmospheric oxidant concentrations, Hg⁰ heterogeneous and interfacial oxidation processes on the surface of atmospheric particles, or potential unknown oxidants.

Calibration Approach for Gaseous Oxidized Mercury Based on Nonthermal Plasma Oxidation of Elemental Mercury

Atmospheric mercury measurements carried out in the recent decades have been a subject of bias largely due to insufficient consideration of metrological traceability and associated measurement uncertainty, which are ultimately needed for the demonstration of comparability of the measurement results. This is particularly challenging for gaseous Hg^II species, which are reactive and their ambient concentrations are very low, causing difficulties in proper sampling and calibration. Calibration for atmospheric Hg^II exists, but barriers to reliable calibration are most evident at ambient Hg^II concentration levels. We present a calibration of Hg^II species based on nonthermal plasma oxidation of Hg⁰ to Hg^II. Hg⁰ was produced by quantitative reduction of Hg^II in aqueous solution by SnCl₂ and aeration. The generated Hg⁰ in a stream of He and traces of reaction gas (O₂, Cl₂, or Br₂) was then oxidized to different Hg^II species by nonthermal plasma. A highly sensitive ¹⁹⁷Hg radiotracer was used to evaluate the oxidation efficiency. Nonthermal plasma oxidation efficiencies with 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 HgO, HgCl₂, and HgBr₂ was confirmed by temperature-programmed desorption quadrupole mass spectrometry (TPD-QMS). This work demonstrates the potential for nonthermal plasma oxidation to generate reliable and repeatable amounts of Hg^II compounds for routine calibration of ambient air measurement instrumentation.

Validating an Evaporative Calibrator for Gaseous Oxidized Mercury

Understanding atmospheric mercury chemistry is the key for explaining the biogeochemical cycle of mercury and for improving the predictive capability of computational models. Increased efforts are being made to ensure comparable Hg speciation measurements in the air through establishing metrological traceability. While traceability for elemental mercury has been recently set, this is by no means the case for gaseous oxidized mercury (GOM). Since a calibration unit suitable for traceable GOM calibrations based on evaporation of HgCl₂ solution was recently developed, the purpose of our work was to extensively evaluate its performance. A highly specific and sensitive ¹⁹⁷Hg radiotracer was used for validation over a wide range of concentrations. By comparing experimental and calculated values, we obtained recoveries for the calibration unit. The average recoveries ranged from 88.5% for 1178 ng m⁻³ HgCl₂ gas concentration to 39.4% for 5.90 ng m⁻³ HgCl₂ gas concentration. The losses were due to the adsorption of oxidized Hg on the inner walls of the calibrator and tubing. An adsorption isotherm was applied to estimate adsorption enthalpy (ΔH_ads); a ΔH_ads value of −12.33 kJ mol⁻¹ was obtained, suggesting exothermal adsorption. The results of the calibrator performance evaluation suggest that a newly developed calibration unit is only suitable for concentrations of HgCl₂ higher than 1 µg m⁻³. The concentration dependence of recoveries prevents the system from being used for calibration of instruments for ambient GOM measurements. Moreover, the previously assessed uncertainty of this unit at µg m⁻³ level (2.0%, k = 2) was re-evaluated by including uncertainty related to recovery and was found to be 4.1%, k = 2. Calibrator performance was also evaluated for HgBr₂ gas calibration; the recoveries were much lower for HgBr₂ gas than for HgCl₂ gas even at a high HgBr₂ gas concentration (>1 µg m⁻³). As HgBr₂ is often used as a proxy for various atmospheric HgBr species, the suitability of the unit for such calibration must be further developed.

Use of Membranes and Detailed HYSPLIT Analyses to Understand Atmospheric Particulate, Gaseous Oxidized, and Reactive Mercury Chemistry

The atmosphere is the primary pathway by which mercury enters ecosystems. Despite the importance of atmospheric deposition, concentrations and chemistry of gaseous oxidized (GOM) and particulate-bound (PBM) mercury are poorly characterized. Here, three membranes (cation exchange (CEM), nylon, and poly(tetrafluoroethylene) (PTFE) membranes) were used as a means for quantification of concentrations and identification of the chemistry of GOM and PBM. Detailed HYSPLIT analyses were used to determine sources of oxidants forming reactive mercury (RM = PBM + GOM). Despite the coarse sampling resolution (1-2 weeks), a gradient in chemistry was observed, with halogenated compounds dominating over the Pacific Ocean, and continued influence from the marine boundary layer in Nevada and Utah with a periodic occurrence in Maryland. Oxide-based RM compounds arrived at continental locations via long-range transport. Nitrogen, sulfur, and organic RM compounds correlated with regional and local air masses. RM concentrations were highest over the ocean and decreased moving from west to east across the United States. Comparison of membrane concentrations demonstrated that the CEM provided a quantitative measure of RM concentrations and PTFE membranes were useful for collecting PBM. Nylon membranes do not retain all compounds with equal efficiency in ambient air, and an alternate desorption surface is needed.

Development of an Understanding of Reactive Mercury in Ambient Air: A Review

This review focuses on providing the history of measurement efforts to quantify and characterize the compounds of reactive mercury (RM), and the current status of measurement methods and knowledge. RM collectively represents gaseous oxidized mercury (GOM) and that bound to particles. The presence of RM was first recognized through measurement of coal-fired power plant emissions. Once discovered, researchers focused on developing methods for measuring RM in ambient air. First, tubular KCl-coated denuders were used for stack gas measurements, followed by mist chambers and annular denuders for ambient air measurements. For ~15 years, thermal desorption of an annular KCl denuder in the Tekran® speciation system was thought to be the gold standard for ambient GOM measurements. Research over the past ~10 years has shown that the KCl denuder does not collect GOM compounds with equal efficiency, and there are interferences with collection. Using a membrane-based system and an automated system—the Detector for Oxidized mercury System (DOHGS)—concentrations measured with the KCl denuder in the Tekran speciation system underestimate GOM concentrations by 1.3 to 13 times. Using nylon membranes it has been demonstrated that GOM/RM chemistry varies across space and time, and that this depends on the oxidant chemistry of the air. Future work should focus on development of better surfaces for collecting GOM/RM compounds, analytical methods to characterize GOM/RM chemistry, and high-resolution, calibrated measurement systems.

Use of Multiple Lines of Evidence to Understand Reactive Mercury Concentrations and Chemistry in Hawai'i, Nevada, Maryland, and Utah, USA

To advance our understanding of the mercury (Hg) biogeochemical cycle, concentrations and chemistry of gaseous oxidized Hg (GOM), particulate-bound Hg (PBM), and reactive Hg (RM = GOM + PBM) need to be known. The UNR-RMAS 2.0 provides a solution that will advance knowledge. From 11/2017 to 02/2019, the RMAS 2.0 was deployed in Hawai'i, Nevada, Maryland, and Utah to test system performance and develop an understanding of RM at locations impacted by different atmospheric oxidants. Mauna Loa Observatory, Hawai'i, impacted by the free troposphere and the marine boundary layer, had primarily -Br/Cl RM compounds. The Nevada location, directly adjacent to a major interstate highway and experiences inputs from the free troposphere, exhibited -Br/Cl, -N, -S, and organic compounds. In Maryland, compounds observed were -N, -S, and organic-Hg. This site is downwind of coal-fired power plants and located in a forested area. The location in Utah is in a basin impacted by oil and natural gas extraction, multiday wintertime inversion episodes, and inputs from the free troposphere. Compounds were -Br/Cl or -O, -N, and -Br/Cl. The chemical forms of RM identified were consistent with the air source areas, predominant ion chemistry, criterion air pollutants, and meteorology.