Hydrogeochemical controls on rare earth element behaviour in an acidic metal mine-impacted stream

Abstract

Due to the high value of rare earth elements (REEs) and their significant pathogenic potential, understanding the sources, fate and transport of REE in aqueous systems is of increasing interest. The behaviour of REE in mine water polluted streams has been well investigated in America, European countries and China. Very few studies of REE transport and fate in watercourses receiving mine water in the UK have been undertaken. The Woodend low level mine water, from the abandoned Threlkeld mine, discharges to the Gate Gill with a pH of ~3.5, and contains high concentration of Zn, Pb, Al, Mn. The Gate Gill was selected for the detailed investigation of REE source, transport and fate, with the aim of understanding (1) the behaviour of REE in watercourses downstream of a mine discharge under varied hydrological conditions; and (2) the factors that influence REE transformation and fractionation in freshwater. ICP-MS was adopted for measuring REE concentration in waters due to its high sensitivity and very low detection limits. However, Ba can cause significant oxide interference on Eu during sample measurement, due to the generally low abundance of REE in nature compared with Ba. In addition, the lanthanide has the strongest metal oxide bond among all the metals, and the lighter rare earth elements oxides formed can cause interferences on the heavier rare earth elements (the mass of which is 16amu more than the interfering lighter REE). To ensure the accuracy of measurement, a mathematical correction method was developed to remove the polyatomic interferences on measured REE isotope. This was based on the signal of the interfering element in the sample and the correction factor of each interference species’ intensity to the interfering element intensity. The interference of BaO on Eu was quite serious on specific samples and the highest error induced was more than 50%. The error was reduced to less 6% after adopting the developed BaO correction equations. The interference of lighter REE on heavier REE was determined to be negligible. This was due to the relatively small concentration difference between the interfering REE and interfered REE analytes. Threlkeld mine is the main source of REE content in the Gate Gill, but the contribution of REE from Threlkeld mine to the River Glenderamackin, into which Gate Gill flows, is rather limited. ii pH is a master control on REE transformation from truly dissolved (<0.005µm) to (suspended) solid phase and there is an inverse relationship between pH and the degree of REE transformation degree. The transformation of REE from truly dissolved (<0.005µm) to suspended solid phase is strongly inhibited when pH is ≤ ≈ 4.5. A higher degree of REE transformation from truly dissolved (< 0.005µm) to suspended solid is likely to occur when pH in water reaches ≈ 5. Fe and Al flocs/ oxyhydroxides are likely to be the main materials scavenging REE from the truly dissolved phase. At sampling sites on the Gate Gill downstream of the mine water the main REE species in water are free REE ions and REE sulphate complexes. Solution chemistry-related processes (especially surface complexation) begin to fractionate REE during the transformation process when pH in the water reaches around pH 5. HREE have greater affinity for the Fe, Al flocs and/ or secondary Fe, Al precipitates which are remobilised from the stream bed relative to MREE and LREE during the REE adsorption process. In the River Glenderamackin, which has a much higher pH (mean pH of 6.05) relative to Gate Gill, the main REE species are free REE ions and REE carbonates. Solution chemistry-related processes (especially solution complexation) results in LREE having a greater affinity for the Fe, Al flocs and/ or secondary Fe, Al precipitates during the REE adsorption process. But source-related processes still have a dominant control on the REE distribution pattern at all downstream sampling sites of mine discharge.