Fractionation of Fe and Cu isotopes in acid mine tailings: Modification and application of a sequential extraction method

Highlights

• Sequential extraction method for Cu and Fe isotope analysis

• No isotope fractionation during the procedure

• The method was applied to mine tailings samples from the Chañaral Bay, Chile.

• Isotope analysis of the fraction revealed several metal mobilization processes.

Abstract

Sulfidic mine tailings have a high potential for contamination of the environment by triggering acid mine drainage. Hence, it is crucial to understand metal mobilization processes and to develop monitoring tools. Metal isotope fingerprinting as a potential monitoring tool for metal sulfide oxidation processes was in the focus of this study by using stable isotope signatures of Cu and Fe. For Fe, a six-step sequential extraction method was applied, in order to separate potential Fe-bearing minerals (water-soluble, exchangeable fraction, Fe(III)(oxyhydr)oxides, Fe-oxides, sulfides and organic compounds and residual/silicates). For Cu, this method was modified into a four step extraction method (water-soluble, exchangeable fraction, oxalate fraction/bound to Fe-oxides and sulfides/residual). To verify accuracy and precision of the sequential extraction method for metal isotope analysis, isotope fractionation during the extraction procedure was investigated employing minerals for which the mineral composition and the isotopic composition was known. The developed procedure is suitable to separate target minerals with only a small loss in the elemental budget. No significant isotope fractionation was observed during the extraction procedure.

Application of this method on two sites of porphyry copper mine tailings in the Atacama Desert in Chile (Chañaral bay) revealed several implications about the mobilization of Fe and Cu in an environmental setting. Iron contents and Fe isotope compositions are homogeneous with depth (0–61 cm; δ⁵⁶Fe ~0.2–0.3‰) for the bulk and the Fe(hyr)oxide fraction and only the deepest samples at ~60 cm exhibited lower δ⁵⁶Fe values (~0‰), which are likely related to the occurrence of an alluvium at this depth. The Fe silicate fraction shows higher δ⁵⁶Fe values (0.6–0.9‰), most likely because of preferential leaching of the light Fe isotopes. This consequently indicates a more pronounced Fe isotope fractionation with depth, as is expected from longer weathering. The Fe sulfide fraction is isotopically lighter compared to the Fe(hydr)oxide fraction, because during sulfide oxidation the heavy Fe isotopes prefer the oxidized forms and oxidative precipitation results in an enriched Fe isotopic signature for Fe(hydr)oxides. The Cu isotope compositions of all bulk samples and individual fractions (except the Cu sulfides) of one site (Ch1) exhibited a decrease of the δ⁶⁵Cu values from the depth towards the surface, which is potentially related to Cu mobilization during capillary water rise in the arid climate. A correlation of δ⁶⁵Cu with pH indicates preferential adsorption of ⁶⁵Cu on Fe(oxy)hydroxides at site Ch1, which is evident by a change of δ⁶⁵Cu from 0.5‰ to −0.7‰ in the water-soluble fraction. At another site (Ch12), where pH at depths was potentially not high enough for the formation of Fe-minerals that could adsorb Cu, only minor Cu isotope fractionation was observed in the water-soluble fraction. The Cu sulfide fraction at site Ch1 exhibits higher δ⁶⁵Cu values with an increase from the bottom (0.42‰) to the surface (0.92‰), which might be related to preferential leaching of the light isotopes, e.g. by microorganisms.

Introduction

Mine tailings are generally stored in piles or in impoundments close to mining sites and consist of heterogeneous materials that are mostly fine grained (Hudson-Edwards et al., 2011; Kossoff et al., 2014). In the past and sometimes also nowadays tailings are deposited directly or nearby of marine environments, rivers, and lakes, and contaminate the surroundings through erosion, the dispersal of dust, or the release of acid mine drainage (AMD) (Gault et al., 2005; Hayes et al., 2009; Hudson-Edwards et al., 2011; Kelm et al., 2009; Lottermoser, 2010; Wang et al., 2014). Large environmental damage can be caused by AMD (e.g. Akcil and Koldas, 2006; Kefeni et al., 2017; Simate and Ndlovu, 2014) which is a result of sulfide minerals reacting with oxygen and water at supergene conditions, hence triggering an increase of acidity. Microorganisms may further accelerate the oxidation of sulfide minerals. Acidophilic iron(II)- and sulfur-oxidizing bacteria are able to produce sulfuric acid during mineral degradation, which in turn further enhances the dissolution of minerals, and thus may mobilize toxic elements like As, Pb, and Cd (Dold, 2017; Hedrich and Schippers, 2017; Schippers et al., 2010). Consequently, large areas may be contaminated, resulting in a degraded water quality and a high risk for life (Akcil and Koldas, 2006; Dold, 2010; Korehi et al., 2013).

Moreover, mine tailings may function as sinks for many elements. Trace metals are accumulated in the material via e.g. adsorption or precipitation, and these may be released due to changes of pH or redox potential (Hindersmann and Mansfeldt, 2014; Kossoff et al., 2014; Ramirez et al., 2005). Even more significant is metal association with different mineralogical fractions, here the metal availability strongly depends on chemical binding partners and speciation (Mansfeldt and Overesch, 2013; Ramirez et al., 2005; Sparks, 2005).

Consequently, the knowledge about metal concentration is not sufficient to delineate mobilization, bioavailability and environmental consequences of mine tailings. Sequential extractions enable investigation of metal distributions in different sediment or soil fractions and can be furthermore applied for exploration purposes or to gain knowledge about biogeochemical element cycling in natural and in mine waste environments (Dold, 2003; Gleyzes et al., 2002; Silveira et al., 2006). They provide a powerful tool to identify the potential mobility and consequently the toxicity and bioavailability of certain elements like As, Hg, and Se (Bacon and Davidson, 2008). The general principle is based on the treatment of sample material with different selective reagents to dissolve discrete mineral fractions without attacking other phases (Gleyzes et al., 2002).

Sequential extractions have already been combined with stable isotope analysis for several elements, e.g. for Fe (Guelke et al., 2010; Mansfeldt et al., 2012; Wiederhold et al., 2007a), Se (Schilling et al., 2014), Zn (Thapalia et al., 2010), Mo (Siebert et al., 2015), and Hg (Wiederhold et al., 2013, Wiederhold et al., 2015). In previous studies, the combination of a sequential extraction method with Cu isotopes (Kusonwiriyawong et al., 2016) was used to determine the response of Cu release in a soil during a flooding event. For Fe, isotope fractionation during pedogenic Fe transformation and translocation processes were examined under oxic conditions by Wiederhold et al. (2007b) and the same was done to analyze the relationships between pedogenic Fe transformation and redistribution processes (Wiederhold et al., 2007a).

Numerous studies determined the variation of the Fe isotope composition in low temperature environments which are caused by diverse fractionation mechanisms (see e.g. reviews by Johnson et al., 2008; Wiederhold, 2015), hence demonstrating that Fe isotope fractionation is a powerful tool to trace reactions and understand mechanisms that are mobilizing or immobilizing Fe (e.g. Akerman et al., 2014; Fekiacova et al., 2013; Mansfeldt et al., 2012; Opfergelt et al., 2017; Schuth et al., 2015; Schuth and Mansfeldt, 2016; Yesavage et al., 2012). Oxidation and reduction of Fe by various microorganisms (aerobic and anaerobic) results in Fe isotopic fractionation, the heavier Fe isotopes are preferentially enriched in the oxidized form (e.g. Balci et al., 2006; Croal et al., 2004; Crosby et al., 2007; Johnson et al., 2004; Kappler et al., 2010). Furthermore, Fe isotope fractionation has also been observed during diverse abiotic processes: Fe(II) oxidation produces for example an isotopically heavier isotope composition of ferrihydrite compared to Fe(II)_aq (Bullen et al., 2001). Iron isotope fractionation was moreover investigated during Fe(III) precipitation, with isotopically lighter remaining solutions and heavier precipitates (e.g. Balci et al., 2006; Beard and Johnson, 2004; Bullen et al., 2001). Igneous rocks in contrast display only limited variations in their Fe isotope composition (e.g. Johnson et al., 2004; Weyer et al., 2005).

In addition to the above mentioned need to identify chemical forms and the speciation of the contaminants, sequential extractions may reveal more details about microbial processes in tailings. Herbert and Schippers (2008) already examined Fe isotope fractionation in mine tailings at the interface between the oxidized and non-oxidized zone. Iron isotope fractionation in these environments is mostly controlled by redox reactions, precipitation, and adsorption processes yielding isotopic variations within the different Fe pools. At the redox interface in mine tailings microorganisms catalyze sulfide oxidation. Ferric iron (FeIII) oxidizes metal sulfides like pyrite, thereby forming sulfuric acid, and Fe(III) is regenerated from ferrous Fe (Fe(II)) by iron-oxidizing microorganisms (Herbert and Schippers, 2008; Nordstrom, 2011):

$$\mathit{Fe}{S}_{2\left(s\right)}+14F{e}^{3+}+8H_{2}O\to 15F{e}^{2+}+2S{O}_4^{2-}+16H^{+}$$

$$4F{e}^{2+}+O_2+4H^{+}\to 4F{e}^{3+}+2H_{2}O$$

Consequently, the resulting AMD is rich in dissolved metals, which may have a severe effect on the environment. Herbert and Schippers (2008) ascertain that pyrite oxidation (Eq. (1)) was mediated by microorganisms and Fe was precipitated abiotically (Eq. (2)), which both affected the Fe isotope composition of the pore water and the solid phase. The Fe isotope fractionation between the pore water and secondary ferric oxyhydroxides can produce δ⁵⁶Fe values of −1.4 to −2.4‰ with a light Fe isotope composition in the dissolved Fe pool and heavier δ⁵⁶Fe values for the solid phase (Herbert and Schippers, 2008). Similar results were also obtained experimentally by Schuth et al. (2015) for variable redox conditions and reductive Fe mobilization.

Copper isotope fractionation is mostly based on redox chemistry and on elemental exchange between dissolved and solid phases. The Cu isotope composition in most primary magmatic Cu-sulfides exhibit only minor isotope variations (e.g. chalcopyrite average δ⁶⁵Cu: 0.12‰, 1σ = 0.33‰; Markl et al., 2006; primary ores δ⁶⁵Cu: 0 ± 0.5‰, 2.s.d.; Mathur et al., 2009) but secondary alteration and low temperature redox processes are the main reason for isotope variations (e.g. Bigalke et al., 2013; Moynier et al., 2017; Wiederhold, 2015).

In several earlier studies, experiments indicated that redox reactions between Cu(I) and Cu(II) are the most important processes that fractionate Cu in natural settings. For instance, reduction of Cu(II)_aq to Cu(I) results in the preferential reduction of the lighter Cu isotope (Ehrlich et al., 2004; Zhu et al., 2002). Mathur et al. (2005) explored biotic and abiotic Cu oxidation with contrasting results: Abiotic oxidation entails an enrichment of the heavy Cu isotope in solution whereas the contribution of microorganisms leads to a Cu pool depleted in ⁶⁵Cu because the heavy Cu isotope is incorporated in metal oxides on the external membranes of the cells of Fe-oxidizing bacteria. This was further demonstrated by Balistrieri et al. (2008) and Pokrovsky et al. (2008). Adsorption onto amorphous ferric oxyhydroxides during the mixture of metal-rich acid rock drainage and river water resulted in lower ⁶⁵Cu values of the water (Balistrieri et al., 2008). Similarly, during adsorption experiments on Fe-oxy(hydr)oxide surfaces and ascorbic acid, preferential adsorption of ⁶⁵Cu was observed (Pokrovsky et al., 2008; Bigalke et al., 2011), while on biological cell surfaces ⁶⁵Cu was depleted (Kimball et al., 2009; Navarrete et al., 2011; Pokrovsky et al., 2008). Song et al. (2016) used Cu isotope fractionation in streams near a porphyry copper mine in China (Dexing Mine) to identify the source of contaminants. As a result, they identified mine tailings as source for the contamination, because of a heavier Cu isotope composition compared to average Cu isotope values from rivers and oceans. Kimball et al. (2009) examined Cu isotope fractionation in acid mine drainage (San Juan Mountains, Colorado, USA) and they determined higher δ⁶⁵Cu values for the stream waters compared to mineral samples (1.38‰ ≤ δ⁶⁵Cu ≤ 1.69‰, 2 s.d. 0.10‰). Borrok et al. (2007) measured various stream waters affected by acid mine drainage and they got heterogeneous results varying between −0.7‰ to +1.4‰ (± 0.3‰, 2 s.d.). Those samples did not reflect a single averaged isotopic composition within the continental crust.

Metal isotope fingerprinting as a potential monitoring tool for metal sulfide oxidation processes was in the focus of our study. It comprises an experimental part with the aim to examine and modify a sequential extraction method (after Dold, 2003, and Mehra and Jackson, 1958) for Fe and Cu separation from sediments and mine tailings. The developed procedures were tested with regard to their ability to quantitatively separate the Cu and Fe fractions that are incorporated in different minerals without generating an unwanted artificial isotope fractionation. To verify the precision of the sequential extraction procedure employed for Cu and Fe isotope analysis, potential isotope fractionation during the extraction was investigated by using minerals with a known chemical and Fe-Cu isotopic composition. In the second part of the study, the developed extraction methods were applied to samples from Cu mine tailings in Chile for which the mineralogy, geochemistry and microbiology was already described (Dold, 2006; Korehi et al., 2013).