Development of a novel passive sampler for gaseous mercury monitoring

Abstract

Mercury as a pollutant has a longstanding history of environmental and health impacts. The ubiquitous global presence of mercury in the environment in combination with its toxicity has given rise to global efforts to control releases and limit exposure and risks associated with the use of mercury. Despite the global transport and deposition of mercury in the environment, there are some areas which lack monitoring data, mainly the southern hemisphere. South Africa, specifically, requires more robust and consistent monitoring, however, implementation of commercially available methods is costly and alternative means of Hg quantitation are needed. Consequently, passive air sampling provides a promising approach to obtain additional data regarding the spatial distribution of Hg across South Africa. Herein a set of novel sulfur-doped 3-D graphene foams were synthesized, characterized, and then employed in a radial passive sampling setup to capture total gaseous mercury (TGM). A chemical vapour deposition (CVD) method was optimized to generate a pristine graphene foam as a cylindrical cartridge. The pristine material was subsequently doped by means of various methods and sulfur sources, namely dimethyl sulfoxide (DMSO), sodium sulfide (Na2S.9H2O), thiourea (SC(NH2)2) and elemental sulfur (S8). Prior to their use in real-world sampling of atmospheric Hg, these derivatized foams were chemically and physically characterized alongside the pristine precursor (where possible) to provide insights into doping efficiencies, surface area differences between foam variants, mass losses upon analysis, surface functionality introduced during doping, and the influence of the polymethyl methacrylate (PMMA) support layer. These characterizations included scanning electron microscopy (SEM), transmission electron microscopy (TEM), energy dispersive X-ray spectroscopy (EDX), Raman spectroscopy, thermogravimetric analysis (TGA), X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared spectroscopy (FTIR). It was found that the multiple-layered nature of the material, as well as the inherent thinness of the graphene, contributed significantly to the variance seen in the characterization data sets. In addition to this, the influence of the PMMA support layer was identified and the efficient removal thereof was also catered for as it was found to impact characterization and doping efficiencies. Following characterization, selected sorbents, which included the pristine precursor as a candidate foam and point of comparison, were utilized in passive air samplers (PASs) that were deployed in a total of four different deployments, encompassing two at a chosen impacted site (Witbank) and a two at a less impacted site (University of Pretoria). Despite inherently low atmospheric Hg concentrations (ng/m3), the radial passive sampler employed in this study was expected to facilitate high uptake rates of Hg and be viable for ambient atmospheric Hg capture, especially at the impacted site where there are surrounding coal-fired power station point sources. The PASs were also further evaluated by comparison to a globally calibrated commercial passive sampler, the MerPAS®, which was deployed concurrently in the final deployment set at each site. iii Studies were done to determine the linear uptake of Hg by the DMSO-doped graphene foam sorbent specifically, but although linear Hg uptake was not observed, successive deployments offered better Hg capture from the pool of other derivatized foams, including the pristine graphene foam precursor. For all sorbents there was a difference in the mean Hg concentration between the two chosen sites, confirming Witbank as an impacted site. The two best performing foams were the thiourea-doped variant, followed by the sodium sulfide variant. Direct comparison to the Hg captured by the PASs deployed to the responses for the MerPAS® showed a promising similarity in effective Hg capture when the mass differences between the MerPAS® activated carbon sorbent and the graphene foams were considered. The methods which offered the highest doping efficiencies were the thiourea and sodium sulfide-based methods, respectively, which were expected to directly improve Hg captured on the surface of the resulting foams. The results of the deployment, however, inferred that a direct increase in sulfur content alone did not lead to a corresponding increase in captured Hg. Instead, an interplay between the surface area, sulfur content and Hg sorbed was identified which requires further investigation. Quantification of Hg on the deployed samplers was achieved by thermal desorption from the sorbent, amalgamation with gold and then atomic absorption spectroscopy (AAS) by means of a direct mercury analyser. The method was validated by successful participation in a global interlaboratory proficiency testing scheme, for which a Z score of -1.042 was obtained. The calibration procedure for analysis of deployed foams was also optimized as a final matrix-matched calibration, for which the influence of the graphene foam matrix on calibration accuracy and precision was also assessed. The use of intermittent cleaning steps and their effect on the accuracy of the calibration was found to contribute to achieving better fits and more reproducible responses. With respect to a NIST 2962c bituminous coal reference material (0.5% sulfur mass fraction), a reproducibility of 7 % RSD (N=11) and error of 1 % (N=11) were achieved for the final deployments and a limit of detection (LoD) of 0.083 ng Hg was achieved for the final optimized calibration. Quality control limits were also established with respect to the calibration, wherein threshold limits of ±20% around the expected value for the NIST standard reference material (SRM) and a 30 μg/kg aqueous quality control standard were maintained for up to ~six months. Overall, the sorbents utilized in this study require more extensive characterization to offer further insight into the observed trends in their variable Hg capture, surface areas and sulfur contents. The PAS also requires further optimization prior to widespread applications and use in areas of low ambient airborne concentrations. The PASs based on thiourea and sodium sulfide-doped graphene foams that were developed for the first time in this study offered the best Hg capture and should be further optimized. The novel S doped graphene foam PAS shows good potential to bridge the monitoring gaps prevalent in atmospheric Hg monitoring data in South Africa.