Speciation, Associations, and Geochemical Transformations of Antimony and Arsenic in a Mine-Contaminated Freshwater System

Abstract

<p>Antimony (Sb) and arsenic (As) are toxic Group 15 metalloids. The two elements often cooccur in hypogene environments which has led to widespread contamination events following mining-related disturbance. The speciation and solid-phase associations of the metalloids are the primary determinants of toxicity, mobility, and exposure risk in these contaminated systems. While the drivers of Sb and As speciation and mobility have been studied in controlled laboratory trials, field studies examining metalloid geochemistry in co-contaminated environments remain rare, largely due to the difficulties maintaining sample and species integrity. Field studies, nevertheless, are crucial to verify that the species and biogeochemical mechanisms and processes observed under laboratory conditions occur in the complex, dynamic conditions of environmental systems. This thesis provides a comprehensive examination of the speciation and sediment associations of Sb and As in the co-contaminated Bakers Creek – Macleay River waterway of New South Wales, Australia. Historic mining in the Hillgrove Mineral Field, which hosts significant deposits of stibnite (Sb₂S₃) and arsenopyrite (FeAsS), has led to the downstream dispersion of Sb and As through 320km of the waterway. The work aimed to identify Sb and As speciation and associations across the varied geomorphological environments of the system, and understand the fundamental drivers of metalloid mobility in the co-contamination event. A suite of complementary advanced analytical techniques was applied to examine metalloid mineralogy, associations, and speciation in field waters and sediments. This was supplemented by laboratory experiments using field-contaminated media to investigate metalloid oxidation and dissolution rates.</p> <p>Antimony(V) predominated (maximum 1042 µg L^-1 ) in the waters of the primary contamination zone of Bakers Creek, as identified by high-performance liquid chromatography coupled to atomic fluorescence spectroscopy (HPLC-AFS). Antimony(III), at low concentration (3.4 µg L^-1 ), was observed in only one water sample during the sampling program, collected in a period of extreme drought. This was notable, as the persistent release of Sb(III) was expected from the abundant stibnite identified in the creek bed and the relatively slow rate of Sb(III) oxidation (2.9 x 10^-3 µM hour^-1 ) observed using filtered Bakers Creek water. Mineralogical X-ray diffraction analysis of Bakers Creek stibnite samples identified oxidation rims of roméite-minerals on exposed surfaces potentially limited the release of the more toxic Sb(III) from stibnite in this system. Solid-state Sb speciation in Bakers Creek – Macleay River sediments via X-ray absorption near-edge structure (XANES) analysis, aligned with the limited Sb(III) detection in the waters of the primary contamination zone, and confirmed Sb(V) as dominant in the system sediments overall (69 – 93 % total Sb). Sb(III) was identified only in tidally-flooded sediments of the lower catchment where reducing conditions intermittently prevailed. </p> <p>In contrast, dissolved As(III) was detected in almost all of the Bakers Creek water samples (up to 6.6 µg L^-1 ), although As(V) was the dominant species (maximum 270 µg L^-1 ). As(V) was also the dominant solid-state species (49 – 92%) identified in XANES spectra of sediments, although orpiment (As₂S₃, maximum 88%) was locally important in Bakers Creek, and As(III) (maximum 26%) was an important fraction of total sediment As in both Bakers Creek and tidally-flooded sites, and likely contributing to dissolved As(III). A significant increase in dissolved As(V) observed during the drought period during sampling, was considered a result of reductive dissolution of oxyhydroxide host phases. No comparable increase in Sb mobility was observed, possibly due to Sb sequestration into crystalline phases and sulfur moieties of organic material. This field data clearly demonstrated that mobility and exposure risk of the two metalloids diverge significantly during droughts and occasions of O₂stress. In this system, As is of greater concern which has important consequences for risk characterisation in contaminated systems with increasing extremes of climate predicted into the future. </p> <p>The percentage of Sb (10 – 59%) and As (16 – 43 %) associated with amorphous sediment components, as determined in sequential extractions, increased with downstream distance from the primary contamination, with a corresponding decrease in residual phases. Bioavailable Sb phases (1.3 – 4.8 %) significantly increased with downstream distance, and bioavailable As (2.2 – 7.7 %) was greater than Sb in most sediments. The decreased risk with downstream attenuation of metalloid concentrations was, therefore, somewhat offset by the increased lability of sediment metalloid associations downstream in the trunk waterway. The importance of amorphous-phases for metalloid retention in the lower catchment may represent increased risk particularly during periods of reduced redox potential. Nevertheless, for Sb this seemed to be partially mitigated by sequestration into crystalline phases during redox cycling, due to an increase in residual phases at certain tidally-inundated sites.</p> <p>The particle size distribution of Sb and As was also examined in both waterway sediment and floodplain soils, as this can similarly affect subsequent mobility and uptake. Antimony concentrations were greatest in mid-size particles (63 – 250 µm) in the upper-to-mid catchment due to products of stibnite breakdown that were heterogeneously distributed in sediment for up to 100 km downstream of the primary contamination. In the lower catchment Sb concentrations were larger in the < 63 µm fraction as the development of sorption complexes with fine particles became the primary source of sediment Sb. As such, the balance between remnant minerals and sorbed species controlled Sb size distribution in the system. This was not the case for As. Arsenic concentrations were greatest in the < 63 µm fraction over the full span of the contaminated waterway due to the fine-grained nature of the primary mineralisation. The assessment of particle size distribution indicated that the mass of metalloids in the system and possible associated risks may have been significantly underestimated in prior work because of the particle sizes analysed. This also clearly demonstrated the importance of both contaminant source and proximity in design of contamination assessments.</p> <p>A dissolution trial using different fractions of contaminated Bakers Creek sediment showed that metalloid mobility varied considerably between geomorphological zones based on the size and composition of the dominant sediment substrate. Areas dominated by fine-grained alluvium showed the most metalloid release (20 % total Sb, 30 % total As) and dissolved Sb(III) (up to 53 µg L^-1, 5 % of total dissolved Sb).. Arsenic was more labile than Sb in all sediments, reflecting the greater affinity of As to partition to solid-phases via the formation of sorption complexes. The results highlighted the need to consider the potential impacts of metalloid contamination on highly localised, microenvironment scales, where the risk of exposure can vary significantly. </p> <p>Overall, this work significantly adds to the field data available on Sb and As speciation and associations in co-contaminated environments, validating metalloid geochemical processes previously only reported under laboratory conditions, and fills many gaps in our understanding of the biogeochemical cycle of the metalloids in co-contaminated environments. Numerous metalloid mitigation or sequestration processes that influence mobility in co-contaminated environments in surficial freshwater systems are reported for the first time. The observed increased partitioning of As to dissolved phases during a period of drought provides an early warning of possible increased risk in these systems with a changing climate. Similarly, the importance of site-specific assessments for Sb is demonstrated due to the highly variable associations observed throughout the waterway. The work underpins accurate risk characterisation in these co-contaminated environments and identifies areas for further field research such as metalloid-microbial interactions.</p>