*Revised Manuscript

STUDY OF ARSENOPYRITE WEATHERING PRODUCTS IN MINE WASTES 1 2 FROM ABANDONED TUNGSTEN AND TIN EXPLOITATIONS 3 4 Murciego A.a, Álvarez-Ayuso E.b,*, Pellitero E.a, Rodríguez Mª.A.c, García-Sánchez 5 b d d d e 6 A. , Tamayo A. , Rubio J. , Rubio F. and Rubin J. 7 a 8 Department of Geology. Plza. de los Caídos s/n. Salamanca University, 37008 9 10 Salamanca (Spain). 11 b 12 Department of Environmental Geochemistry. IRNASA (CSIC). Apdo. 257, 37071 13 14 Salamanca (Spain). 15 c 16 Faculty of Sciences, Crystallography and Mineralogy Area, Avd. Elvas s/n. 17 18 Extremadura University, 06071 Badajoz (Spain). 19 d 20 Ceramic and Glass Institute (CSIC). c/ Kelsen, 5, 28049 Cantoblanco, Madrid (Spain). 21 e 22 Material Science Institute of Aragón, CSIC-Zaragoza University. c/ María de Luna 3, 23 24 50009 Zaragoza (Spain). 25 26 *Corresponding author. Fax number: +34 923219609. E-mail: [email protected] 27 28 29 30 ABSTRACT 31 32 Arsenopyrite-rich wastes from abandoned tungsten and tin exploitations were studied to 33 34 determine the composition and characteristics of the secondary phases formed under 35 natural weathering conditions so as to assess their potential environmental risk. 36 37 Representative weathered arsenopyrite-bearing rock wastes collected from the mine 38 39 dumps were analysed using the following techniques: X-ray powder diffraction (XRD) 40 41 analysis, polarizing microscopy analysis, electron microprobe analysis (EMPA) and 42 43 microRaman and Mössbauer spectroscopies. Scorodite, pharmacosiderite and 44 45 amorphous ferric arsenates (AFA) with Fe/As molar ratios in the range 1.2-2.5 were 46 identified as secondary arsenic products. The former showed to be the most abundant 47 48 and present inAccepted the different studied mining areas Manuscript. Its chemical composition showed to 49 50 vary in function of the original surrounding rock mineralogy in such a way that 51 52 phosphoscorodite was found as the mineral variety present in apatite-containing 53 54 geoenvirons. Other ever-present weathering phases were goethite and hydrous ferric 55 56 oxides (HFO), displaying, respectively, As retained amounts about 1 and 20% 57 (expressed as As O ). The low solubility of scorodite, the relatively low content of AFA 58 2 5 59 and the formation of compounds of variable charge, mostly of amorphous nature, with 60 61 62 63 1 64 Page 1 of 30 65 high capacity to adsorb As attenuate importantly the dispersion of this element into the 1 2 environment from these arsenopyrite-bearing wastes. 3 4 5 6 Keywords: Arsenopyrite, arsenic, mine wastes, secondary or weathering products, 7 8 environmental risk assessment. 9 10 11 12 1. INTRODUCTION 13 14 Arsenic is currently regarded as one of the most toxic inorganic pollutants, being 15 16 responsible for severe environmental and health impacts. Contamination with As is a 17 18 global concern due to its worldwide distribution. In nature As occurs as a constituent of 19 different minerals, but its release has been promoted by anthropogenic activities such as 20 21 mining. More than 300 As minerals are known, with approximately 60% being 22 23 arsenates, 20% sulphides and sulphosalts and the remaining 20% including arsenides, 24 25 arsenites, oxides and elemental As [1,2]. Sulphide deposits are the main mineral source 26 27 of As, in which this element can be present in high concentrations. Particularly, 28 29 arsenopyrite (FeAsS) is the most common As-bearing mineral, being found in a range 30 of ore deposits. Thus, arsenopyrite is the ubiquitous As-bearing mineral in a variety of 31 32 hydrothermal environments, from high-temperature magmatic-hydrothermal porphyry- 33 34 style Sn-W and Cu(±Au) deposits to mesothermal polymetallic Cu-Pb-Zn-Ag and gold 35 36 deposits [3]. These deposits have been extensively exploited for the economic ore 37 38 minerals, inducing a wide legacy of As in mine spoils. 39 40 Arsenopyrite is stable under reducing conditions, and mine tailings containing 41 42 arsenopyrite should be chemically stable during long-term storage provided they are 43 kept water-saturated and moderately reduced [4]. Conversely, under oxidising 44 45 conditions the mobilisation and migration of As into the environment can take place. 46 47 All over the world, there are many historic derelict mines that have generated a large 48 Accepted Manuscript 49 amount of arsenopyrite-bearing waste rocks and tailings. These wastes have been poorly 50 51 managed, allowing oxidation processes to occur. Arsenopyrite can be easily oxidised by 52 3+ 53 both O2 and Fe , this process being promoted by micro-organisms, especially by 54 acidophilic Fe- and S-oxidising bacteria [5]. 55 56 Secondary As compounds such as scorodite (FeAsO ·2H O), pharmacosiderite 57 4 2 58 (KFe4(AsO4)3(OH)4·6-7H2O), arsenolite (As2O3), arseniosiderite 59 60 (Ca2Fe3(AsO4)3O2·3H2O) and amorphous ferric arsenate-(sulphate) have been reported 61 62 63 2 64 Page 2 of 30 65 as the most common products of arsenopyrite weathering in natural environments, 1 2 together with other secondary compounds such as goethite (FeO(OH)), ferrihydrite 3 (Fe O ·0.5H O) or amorphous hydrous ferric oxides (HFO) [6-11]. The genesis of these 4 2 3 2 5 secondary products is dependent on the original ore mineralogy and on the processes 6 7 that have occurred during their formation (mainly redox and pH changes) [2]. The 8 9 release of As from mine wastes is controlled by precipitation-dissolution and 10 11 adsorption-desorption reactions involving the secondary compounds generated during 12 13 the weathering of arsenopyrite [12]. Therefore, their identification and characterisation 14 are crucial for the risk assessment and environmental control of the historic 15 16 arsenopyrite-bearing mine spoils. Whereas many works studying wastes from gold 17 18 mining areas have been performed [4,7,8,11,13-17], only some efforts have been 19 20 devoted to study wastes from tungsten- and tin-mining districts [18-20]. 21 22 The aim of this work is to study the mine wastes from abandoned tungsten and tin 23 24 exploitations in order to identify and characterise the secondary products generated 25 26 from arsenopyrite weathering and to assess the potential environmental risk of such 27 wastes. 28 29 30 31 32 2. MATERIALS AND METHODS 33 34 2.1. Study areas 35 36 Two abandoned mine exploitations were considered in this study, the Barruecopardo 37 38 and Terrubias mines, located, respectively, in the north-west and centre of the 39 Salamanca province, Spain (Figure 1). The Barruecopardo mine corresponds to the most 40 41 important tungsten deposit in Spain, exploited extensively from 1912 to 1983. This is 42 43 constituted by subvertical quartz veins disposed in parallel intragranitic bands. The 44 45 main minerals present are scheelite (CaWO4) and wolframite ((Fe,Mn)WO4), which 46 47 constitute the ore, and arsenopyrite, pyrite (FeS2), chalcopyrite (CuFeS2) and ilmenite 48 (FeTiO3) as primaryAccepted minerals. These minerals appearManuscript in the veins and on the borders of 49 50 granite in which apatite (Ca (PO ) (OH,F,Cl,Br) ) appears as an accessory mineral 51 10 4 6 2 52 [21]. The Terrubias mine corresponds to a tungsten- and tin-rich deposit worked 53 54 intermittently during the last century, and whose exploitation ceased more than three 55 56 decades ago. Two Sn-W mineralisation types are present in this mining area: stratiform, 57 58 in calcosilicate bands, and in subvertical veins of NNE-NE and N100-130E directions, 59 60 cutting the scheelite-rich calcosilicate bands. The paragenesis of these veins is 61 62 63 3 64 Page 3 of 30 65 dominated by quartz, mica, tourmaline, wolframite, scheelite, cassiterite (SnO2) and 1 2 arsenopyrite [22]. 3 4 2.2. Characterisation of weathered arsenopyrite-bearing wastes 5 6 About 25 samples of weathered arsenopyrite were collected from the Barruecopardo 7 8 and Terrubias mining areas. All these samples were surface samples. In the studied 9 mining areas wastes were dispersed on surface or accumulated in very small dumps 10 11 where the existence of weathering profiles was not observed. The weathering products 12 13 show yellow-greenish colours, graduating towards orange-reddish tonalities in the 14 15 sample borders, where these display numerous holes (Figs. 2a and 3a). Such weathered 16 17 arsenopyrite samples were studied by X-ray powder diffraction (XRD) analysis, 18 19 polarizing microscopy analysis, electron microprobe analysis (EMPA) and microRaman 20 and Mössbauer spectroscopies. X-ray diffraction analyses were performed on a D8 21 22 Advance Bruker diffractometer equipped with a diffracted-beam graphite 23 24 monochromator, using the CuK radiation (λKα =1.54Å). Solids were scanned as 25 26 unoriented powder samples from 4 to 60º 2 with a 0.04 2 step and a 1 s per step 27 28 counting time. The analyses by polarizing microscopy were carried out on polished thin 29 30 sections using a Leica DMLP microscope in both transmitted and reflected light. 31 32 Electron microprobe analyses for chemical analyses and for electron backscatter images 33 were conducted on carbon-coated polished thin sections in wavelength-dispersive mode 34 35 using a JEOL Superprobe JXA-8900 M electron probe microanalyser. The analysed 36 37 elements were Al, Mn, P, Ba, As, Si, Fe, S, K, Cu, Pb, Ca and Zn. Operating conditions 38 39 were an accelerating voltage of 15 kV and a beam current of 2 nA with a spot size of 1- 40 41 5 µm and spot analysis and background collection times of 10 and 5 s, respectively. 42 43 Calibration was carried out using the following standards: sillimanite for Al, almandine 44 for Fe, pyrolusite for Mn, apatite for P, witherite for Ba, gallium arsenide for As, albite 45 46 for Si, galena for S and Pb, microcline for K, chalcopyrite for Cu, kaersutite for Ca and 47 48 gahnite for Zn.Accepted The microRaman spectra of polishedManuscript thin sections of samples were 49 50 recorded with a Via Renishaw spectrometer equipped with a Charge Coupled Device 51 52 (CCD) detector. Samples were excited with a 784 nm laser operating at 25 mW. A 50x 53 54 objective lens was employed to focus the laser beam. The overall spectral resolution 55 was 2 cm-1. The Mössbauer spectra of powder samples were recorded at room 56 57 temperature in transmission mode on a constant acceleration Mössbauer spectrometer 58 59 equipped with a 57Co-Rh matrix source. -Fe foil was used for velocity calibration. 60 61 62 63 4 64 Page 4 of 30 65 Spectra were fitted using the Mössbauer fitting program NORMOS written by R.A. 1 2 Brand. 3 4 2.3. Environmental risk assessment of the arsenopyrite-bearing mine wastes 5 6 In order to more accurately assess the environmental risks that may entail the mine 7 8 wastes present in the Barruecopardo and Terrubias mining areas, these wastes were 9 subjected to the subsequent determinations: pH, net neutralisation potential (NNP), 10 11 toxicity and bioavailable content of As and other toxic elements (Cu, Pb, Zn and Cd). 12 13 pH was analysed potentiometrically in a paste saturated with water. NNP was evaluated 14 15 following the method of Sobek et al. (1978) [23] as a measure of the difference between 16 17 the neutralising potential (NP) and the acid potential (AP). Accordingly, AP was 18 19 derived by calculating the theoretical quantity of acid that could be produced if the total 20 S content of wastes is converted to sulphuric acid, and expressing this quantity in kg of 21 22 CaCO3 required to neutralise the acidity produced from 1 t of waste material (AP (kg 23 -1 24 CaCO3 t ) = 31.25 wt.% S), and NP was determined by neutralising finely ground 25 26 waste samples with 0.1 M HCl at 90 °C, and expressing the corresponding results in kg 27 -1 28 CaCO3 t . Toxicity was established according to the TCLP test method. Thus, waste 29 30 samples underwent an agitation period of 18 h with buffered acetic acid (pH 4.93±0.05) 31 on a vertical rotary shaker (30 rpm), using a liquid/solid (L/S) ratio of 20 L kg-1. The 32 33 bioavailable content of As was extracted according to the method of Wenzel et al. 34 35 (2001) [24] by subjecting waste samples to an agitation period of 16 h with 0.05 M 36 -1 37 (NH4)2SO4, using a L/S ratio of 25 L kg . The bioavailable content of the other toxic 38 39 elements (occurring as cationic species) was determined following an EDTA extraction 40 41 procedure [25] in which waste samples underwent an agitation period of 1 h with 1 M 42 CH COONH + 0.02 M EDTA, employing a L/S ratio of 10 L kg-1. The extracts 43 3 4 44 resulting from these three extraction processes were analysed for As and/or Cu, Pb, Zn 45 46 and Cd by inductively coupled plasma-atomic emission spectrometry (ICP-AES) using 47 48 a Varian 720-AcceptedES unit. Manuscript 49 50 Additionally, in order to obtain more information on the environmental impact of these 51 52 mine wastes the water quality of the nearby ponds was also evaluated, determining their 53 54 pH, Eh and concentrations of sulphate, As and other toxic elements (Cu, Pb, Zn and 55 Cd). The concentrations of such species were analysed by ICP-AES. 56 57 58 59 60 3. RESULTS AND DISCUSSION 61 62 63 5 64 Page 5 of 30 65 3.1. Characterisation of weathered arsenopyrite-bearing wastes from the Barruecopardo 1 2 mine 3 4 3.1.1. X-ray diffraction analyses 5 6 XRD analyses were performed on the bulk samples collected from the Barruecopardo 7 8 mine waste rock dumps and on the yellow-greenish and orange-reddish phases 9 distinguished on them. Bulk samples showed scorodite, pharmacosiderite and goethite 10 11 as secondary minerals, and the presence of primary minerals such as quartz, muscovite, 12 13 K-feldspar and arsenopyrite, which are known constituents of the ore gangue and 14 15 surrounding rocks. The XRD analyses carried out specifically on the yellow-greenish 16 17 phases revealed that these are mainly composed of scorodite; the presence of the 18 19 primary minerals muscovite, K-feldspar and arsenopyrite was also detected (Fig. 4a). 20 Pharmacosiderite and goethite were found as the main secondary minerals composing 21 22 the orange-reddish phases, with scorodite only present in minor amounts; the presence 23 24 of the primary mineral K-feldespar was also detected (Fig. 4b). 25 26 3.1.2. Polarizing microscopy analysis 27 28 Microscopically different modes of occurrence of arsenopyrite can be distinguished. 29 30 Thus, arsenopyrite appears as unweathered idiomorphic crystals, in slightly weathered 31 32 crystals and in highly to completely weathered crystals, mostly keeping their original 33 34 morphology. Two kinds of arsenopyrite weathering products are observed: scorodite 35 36 and red-blackish phases. Scorodite is the most abundant oxidation product. This occurs 37 on the arsenopyrite crystal borders, along small cracks, inside of arsenopyrite (Fig. 2b) 38 39 and following the crystallographic planes. On occasions scorodite completely replaces 40 41 arsenopyrite crystals, pseudomorphing them. More usual is the presence of corroded 42 43 relicts of arsenopyrite inside scorodite. This mineral is present in different textures: 44 45 microcrystalline massive, spherulitic and colloform, and in prismatic crystals of varying 46 47 sizes (from µm to tens of µm) (Fig. 2c). The red-blackish phases occur bordering the 48 scorodite. ThesAcceptede could correspond to HFO andManuscript goethite (Fig. 2d), but no clear 49 50 identification of them could be established due to the product mixture and to the low 51 52 crystallinity and relatively low content of goethite, as reflected by the corresponding 53 54 XRD pattern. 55 56 3.1.3. Electron microprobe analysis 57 58 A total of 95 chemical analyses were performed by EMPA on the observed arsenopyrite 59 60 weathering products (scorodite (80) and red-blackish phases (15)). The specific number 61 62 63 6 64 Page 6 of 30 65 of analyses performed on each weathering phase is representative of the phase 1 2 abundance. These analyses are indicated in Table 1 (only analyses given element 3 contents for single phases were reflected, whereas those indicating phase mixture or 4 5 presence of arsenopyrite were omitted). According to these analyses, the main general 6 7 differential characteristics of scorodite are: relatively high P contents (from 0.26 to 8 9 9.01% P2O5), moderately low contents of S (up to 1.49 % SO3) and minor or trace 10 11 concentrations of other elements (Al, Si and Cu). This scorodite displays Fe/As molar 12 13 ratios in the range 1.04-1.47, differing from the scorodite theoretical value (Fe/As = 1). 14 The frequency distribution is as follows: 13% (Fe/As = 1-1.1), 41% (Fe/As = > 1.1-1.2), 15 16 31% (Fe/As = > 1.2-1.3) and 15 % (Fe/As = > 1.3). Therefore, an important proportion 17 18 of the analyses performed on scorodite (> 85%) shows Fe/As molar ratios increased in 19 20 more than 10% compared to the ideal value. This As deficiency is explained by the 21 22 corresponding P enrichment shown by this mineral, giving Fe/(As+P) molar ratios in the 23 24 range 0.92-1.26. The frequency distribution is as follows: 80% (Fe/(As+P) = > 0.9-1.1), 25 16% (Fe/(As+P) = > 1.1-1.2) and 4% (Fe/(As+P) = > 1.2). Hence, most of the analyses 26 27 performed on scorodite agrees well with the proposed elemental constitution. This 28 29 compositional characteristic denotes that this mineral is an intermediate term between 30 31 scorodite and strengite (FePO4·2H2O), corresponding to P-rich scorodite or 32 33 phosphoscorodite (Fe(As,P)O4·2H2O). P exists in scorodite by isomorphous substitution 34 3- 3- 35 of PO4 for AsO4 or by simultaneous incorporation into the crystal lattice, and is 36 mineralogically expressed as a solid solution of scorodite and strengite, though they 37 38 form only a partial solid-solution series. No continuous solid solution is produced due to 39 3- 3- 40 the size difference between AsO4 and PO4 [26]. The electron backscatter images also 41 42 suggest such elemental composition. Thus, concentric zonation is observed in scorodite 43 44 crystals (Fig. 2e), showing P enrichment from the borders to the core (with P2O5 values 45 from 1.42 to 9.01%). Accordingly, grey tonalities are intensified towards the crystal 46 47 cores as elements with low atomic number (P) appear darker than those with high 48 49 atomic numberAccepted (As). Zonation is also observed inManuscript the colloform scorodite (Fig. 2f), with 50 51 P2O5 contents varying from 0.75 to 6.36%. The P enrichment shown by the scorodite 52 53 developed in this mining area should be related to the occurrence of apatite in the 54 55 surrounding rocks of the exploited ore deposit. 56 57 The analyses performed on the red-blackish phases show Fe2O3 and As2O5 contents 58 59 comprised, respectively, in the ranges 55.8-65.6% and 5.66-11.9%. Such values suggest 60 that these phases are not composed of single goethite as their Fe O content should be 61 2 3 62 63 7 64 Page 7 of 30 65 higher (theoretically close to 80%) and their As2O5 concentration should display much 1 2 lower values. Different studies performed on the As(V) adsorption capacity of goethite 3 [27,28] have revealed that the maximum As(V) adsorption values attained by this 4 5 mineral are about 1% As2O5. Conversely, HFO display much elevated As(V) adsorption 6 7 capacities due to their higher number of reactive hydroxyl groups [27,29]. The relatively 8 9 high As2O5 contents shown by these phases together with their apparent hydrous 10 11 character (as derived from the low total elemental composition) suggest the presence of 12 13 HFO as important constituents of them. It is also worth mentioning that the P2O5 14 content of these weathering phases ranges from 0.30 to 0.78%. It is well known the 15 16 ability of compounds of variable charge, as those present in such phases, to adsorb 17 18 phosphate. This ability and the presence of apatite in the studied area explain such P2O5 19 20 contents. Phosphate and arsenate are chemical analogues with comparable dissociation 21 22 constants for their acids and similar chemical properties, competing directly for binding 23 24 sites. Moreover, at similar concentrations phosphate outcompetes arsenate for 25 adsorption sites because of its smaller size and higher charge density [30]. 26 27 3.1.4. MicroRaman spectroscopy 28 29 30 The microRaman spectra of the weathering phase identified as scorodite agree well with 31 that of scorodite reported in literature [31-33], confirming the important occurrence of 32 33 this mineral as a secondary product of arsenopyrite. The representative microRaman 34 35 spectrum corresponding to the most crystalline scorodite (Fig. 5a) displays in the 36 -1 37 arsenate stretching region two high intensity bands at 812 cm ( 1, symmetric stretching 38 -1 39 mode) and at 892 cm ( 3, asymmetric stretching mode), characteristic of scorodite. 40 -1 41 The low wavelength region is characterised by several bands at 338 and 388 cm ( 2, 42 -1 43 symmetric bending mode) and at 419, 458 and 485 cm (asymmetric bending mode), 44 45 also in accordance to the low wavelength region of scorodite. The microRaman spectra 46 derived from zoned crystals of scorodite with different P concentrations do not reflect 47 48 such compositionalAccepted variations. Also similar vibrationalManuscript positions are shown by the 49 50 microRaman spectra corresponding to scorodite in microcrystalline forms, although 51 52 exhibiting poor definition. 53 54 The microRaman spectra corresponding to the red-blackish phases distinguished by 55 56 polarizing microscopy show goethite as the only crystalline secondary mineral 57 58 occurring in them. Its representative microRaman spectrum is displayed in Figure 5b, 59 where the following bands are distinguishable: 249, 303, 390, 485 and 556 cm-1. These 60 61 62 63 8 64 Page 8 of 30 65 bands are in good agreement with the characteristic pure goethite Raman modes 1 2 reported by de Faria et al. (1997) [34]. Several microanalyses performed on these red- 3 blackish phases give spectra with no apparent Raman modes, indicating the amorphous 4 5 nature of some of their constituents, which could be compatible with the presence of 6 7 HFO. 8 9 3.1.5 Mössbauer spectroscopy 10 11 The Mössbauer analyses were performed on the yellow-greenish and orange-reddish 12 13 phases distinguished on the waste rocks collected from the mine dumps. The 14 15 representative Mössbauer spectra obtained from the analyses of these phases are shown 16 17 in Figures 6a and 6b, respectively. According to Takano et al. (1971) [35], scorodite has 18 -1 -1 19 values of = 0.39 mm s and QS = 0.42 mm s , whereas pharmacosiderite displays 20 -1 -1 21 values of = 0.36 mm s and QS = 0.68 mm s . The spectrum corresponding to the 22 yellow-greenish phases is solved assuming a single doublet fit. Its Mössbauer 23 24 parameters are = 0.39 mm s-1 and QS = 0.44 mm s-1, which are in agreement with 25 26 those of scorodite. The spectrum corresponding to the orange-reddish phases is 27 28 principally a doublet accompanied by a weaker sextet. The Mössbauer parameters of the 29 30 main component are = 0.37 mm s-1 and QS = 0.57 mm s-1. The doublet has values 31 32 intermediate between those of scorodite and pharmacosiderite, suggesting the presence 33 34 of both minerals in these phases. The sextet is related to goethite with BHF = 34T. This 35 36 value is significantly reduced when compared with the typical BHF value of goethite, 37 which is 38T [36]. Structural defects, like vacancies and/or isomorphic substitution, and 38 39 also small particle size are the main contributors to reduced BHF values [37]. 40 41 3.2. Characterisation of weathered arsenopyrite-bearing wastes from the Terrubias mine 42 43 3.2.1. X-ray diffraction analyses 44 45 XRD analyses performed on the bulk samples collected from the Terrubias mine waste 46 47 rock dumps showed scorodite and goethite as secondary minerals, and the presence of 48 49 the primary mineralsAccepted quartz, dravite (a tourmaline Manuscript variety) and arsenopyrite. Similar 50 51 mineralogical composition, except for the presence of goethite, was displayed by the 52 53 XRD analyses performed on the yellow-greenish phases occurring on the weathered 54 55 waste rock samples (Fig. 4c), whereas goethite was the only crystalline secondary 56 mineral detected in the orange-reddish phases together with quartz as primary mineral 57 58 (Fig. 4d). 59 60 3.2.2. Polarizing microscopy analysis 61 62 63 9 64 Page 9 of 30 65 Microscopically arsenopyrite appears as unweathered euhedral crystals embedded in 1 2 quartz, as partial or totally weathered crystals that keep their original shape and in 3 relicts with corroded borders. Several arsenopyrite weathering products are 4 5 distinguished: scorodite, goethite, red-blackish phases and orange-reddish phases. 6 7 Scorodite is the most abundant oxidation product. This mineral is present in different 8 9 textures: microcrystalline massive, spherulitic and colloform. The latter is very profuse, 10 11 whereas the spherulitic texture is mainly limited to zones of high porosity. The massive 12 13 texture is also widespread, appearing around the arsenopyrite, filling small cracks in 14 quartz and locally aggregating the scorodite spherulites (Fig. 3b). Goethite and the other 15 16 red-blackish phases occur bordering the scorodite (Fig. 3c). These latter phases are 17 18 isotropes, with numerous internal reflections. Such secondary products seem to 19 20 correspond to HFO. The orange-reddish phases are very rare and generally occur 21 22 embedded in scorodite masses and, more rarely, in direct contact with corroded relicts 23 24 of arsenopyrite. These show a cracked appearance (Fig. 3d), being consistent with 25 amorphous ferric arsenates (AFA). 26 27 3.2.3. Electron microprobe analysis 28 29 30 A total of 115 chemical analyses were performed by EMPA on the observed 31 arsenopyrite weathering products (scorodite (78), goethite (9), red-blackish phases (14) 32 33 and orange-reddish phases (14)). The specific number of analyses performed on each 34 35 weathering phase is representative of the phase abundance. These analyses are indicated 36 37 in Table 2 (only analyses given element contents for single phases were reflected, 38 39 whereas those indicating phase mixture or presence of arsenopyrite were omitted). The 40 41 scorodite chemical composition varies within short ranges, namely Fe2O3 = 34.0-37.8% 42 and As O = 46.4-51.2%, displaying really minor or trace concentrations of other 43 2 5 44 elements (S, Ba, Cu and Al). This scorodite shows Fe/As molar ratios in the range 0.99- 45 46 1.09, values very close to the theoretical value, with all the variations accounting for 47 48 less than 10%Accepted as compared to the ideal value. TheManuscript corresponding electron backscatter 49 50 images reveal the important occurrence of colloform scorodite (Fig. 3e). The chemical 51 composition of the orange-reddish phases is very variable, without clear stoichiometry, 52 53 showing Fe2O3 and As2O5 contents in the ranges 30.7-49.1% and 25.3-42.1%, 54 55 respectively. Their SO3 content is relatively low, ranging from 0.04 to 3.84%. 56 57 Occasionally some impurities of Cu, K, Ba and, in a lesser extent, Zn and Ca are 58 59 present. The described chemical composition of these amorphous gels with the typical 60 61 62 63 10 64 Page 10 of 30 65 mudcrack-type texture, also observed in the electron backscatter images (Fig. 3f), 1 2 confirms that these are constituted by AFA. The presence of amorphous ferric 3 sulphoarsenates (AISA) can be discarded in view of their low SO content. The Fe/As 4 3 5 molar ratios of these compounds vary from 1.21 to 2.51. 6 7 The analyses carried out on goethite show Fe O and As O contents comprised, 8 2 3 2 5 9 respectively, in the ranges 59.0-79.1% and 0.32-1.38%. The indicated As2O5 values 10 11 agree well with the As(V) adsorption capacities reported in literature for this mineral 12 13 [27,28]. Significant SiO2 contents are also displayed by this mineral, from 0.56 to 14 15 9.43%, with the highest values in correspondence to the lowest Fe2O3 contents. The 16 17 chemical composition of the other red-blackish phases displays Fe2O3 and As2O5 18 contents comprised, respectively, in the ranges 50.1-63.3% and 16.4-21.7%. Such high 19 20 As2O5 values together with the apparent hydrous character of these phases (as derived 21 22 from the low total elemental composition) suggest that these are constituted by HFO on 23 24 which As(V) is adsorbed. 25 26 3.2.4. MicroRaman spectroscopy 27 28 Of the different weathering phases distinguished by polarizing microscopy (scorodite, 29 30 goethite, red-blackish phases and orange-reddish phases) the corresponding 31 32 microRaman spectra confirm the presence of scorodite and goethite. The representative 33 microRaman spectrum of the analysed scorodite (Fig. 5c) displays the typical bands in 34 35 the arsenate stretching (809 and 891 cm-1) and low wavelength (337, 385, 420, 451 and 36 37 486 cm-1) regions of scorodite. The Raman modes reflected by the representative 38 -1 39 spectrum of the analysed goethite (Fig. 5d) (240, 294, 383, 479 and 552 cm ) also 40 41 correspond to the characteristic bands of pure goethite. The microanalyses performed on 42 43 the red-blackish and orange-reddish phases give spectra with no apparent Raman 44 modes, confirming their low-crystalline or amorphous nature, as is the case of HFO and 45 46 AFA, with which are, respectively, compatible. 47 48 3.2.5. MössbauerAccepted spectroscopy Manuscript 49 50 The Mössbauer spectra of the yellow-greenish and orange-reddish phases distinguished 51 52 on the waste rocks collected from the mine dumps are indicated in Figures 6c and 6d, 53 54 respectively. The representative spectrum of the yellow-greenish phases is a 55 56 superposition of two doublets, one of them really weak and with a very low 57 -1 -1 58 contribution. The main doublet has values of = 0.40 mm s and QS = 0.43 mm s , 59 60 61 62 63 11 64 Page 11 of 30 65 compatible with those of scorodite. The representative spectrum of the orange-reddish 1 2 phases is fitted by a sextet with BHF = 37T, corresponding to that of goethite. 3 4 3.3. Environmental risk assessment of the arsenopyrite-bearing mine wastes 5 6 The environmental characterisation performed on the Barruecopardo and Terrubias 7 8 mine wastes is reflected in Table 3. Mine wastes exhibit negative NNP. Theoretically, 9 wastes with positive NNP values have no potential for acidification whereas wastes 10 11 -1 with negative NNP values do, especially at levels < - 30 kg CaCO3 t . Significant lower 12 13 values are displayed by concerned mine wastes, indicating their potential of acid 14 15 generation. Accordingly, the corresponding pH values are low. These pH values are also 16 17 consistent with the occurrence and persistence of scorodite as main arsenopyrite 18 19 weathering product. According to the TCLP test, the Terrubias mine wastes are 20 characterised as toxic as the regulatory limits are greatly exceeded for As. The As 21 22 leached from the Barruecopardo mine wastes also displays a concentration close to the 23 24 limit for this characterisation. The bioavailable fraction of As as determined by the 25 26 (NH4)2SO4 extraction procedure shows important contents, reflecting higher values for 27 28 wastes from the Barruecopardo mine. These concentrations could imply an As 29 30 transference to different environmental compartments. In this regard, soil total 31 concentrations of As about 1200 and 1350 mg kg-1 have been reported in the environs of 32 33 the Barruecopardo and Terrubias mine dumps, respectively [27]. The bioavailable 34 35 fraction of the other toxic elements as determined by the EDTA extraction procedure 36 37 displays in general relatively low contents, especially in wastes from the Barruecopardo 38 39 mine, with Pb showing in both cases the highest values. 40 41 The parameters analysed to establish the water quality of the nearby ponds are indicated 42 43 in Table 4. Lower pH and higher Eh values as well as lower concentrations of As and 44 other toxic elements were shown by the pond present in the Barruecopardo mining area. 45 46 The unexpected water characteristics found in the Terrubias mining area pond could be 47 48 explained takingAccepted into account that carbonate and Manuscriptcalcosilicate rocks outcrop in the basin 49 50 of this pond. This feature minimises importantly the impact of mine wastes on water, 51 52 raising water pH and decreasing the sulphate content as well as those of toxic elements 53 54 existing as cationic species due to the formation of insoluble compounds. 55 56 P-rich scorodite, pharmacosiderite, goethite and HFO were the secondary products 57 58 identified in the weathered arsenopyrite-rich wastes from the Barruecopardo mine, 59 whereas scorodite, AFA, goethite and HFO were the weathering phases detected in 60 61 62 63 12 64 Page 12 of 30 65 wastes from the Terrubias mine. Such specific secondary phases are responsible for the 1 2 As control in the studied mining areas. 3 4 In both of the studied mine wastes scorodite was found the most abundant oxidation 5 6 product. Its formation takes place according to the following reaction [38]: 7 3+ 2+ 2- + 8 FeAsS + 14 Fe + 10 H2O → 14 Fe + SO4 + FeAsO4·2H2O + 16 H (1) 9 10 Many studies have been performed studying the solubility and stability of scorodite [38- 11 12 41] and, although some discrepancy has been found between the reported solubility 13 products (K : 10-21.7-10-25.4), scorodite is considered to have low solubility and relatively 14 s 15 high stability under oxidising conditions. The solubility of scorodite is minimum (< 0.5 16 17 mg As L-1) at pH slightly less than 3, from which its incongruent dissolution takes place 18 19 with the concomitant precipitation of HFO and As release according to the following 20 21 reactions [42]: 22 - + 23 FeAsO4·2H2O + H2O → H2AsO4 + Fe(OH)3 + H (2) 24 2- + 25 FeAsO4·2H2O + H2O → HAsO4 + Fe(OH)3 + 2 H (3) 26 27 The released As is adsorbed on the precipitated HFO. These processes keep soluble As 28 29 at really low levels up to pH values about 6 [39,40]. The scorodite solubility has been 30 31 reported not to be influenced by the mineral particle size [43]. Moreover, its dissolution 32 kinetic has been proven to be slow, with dissolution rates on the order of 10-10 to 10-11 33 34 mol m-2 s-1 [43,44], which explains why scorodite persists at mine waste sites after 35 36 many years of weathering. Therefore, the prevalence of scorodite in the studied mine 37 38 wastes should limit greatly the release of As into the surrounding environmental 39 40 compartments, unless drastic condition changes (mainly redox and pH changes) take 41 42 place. Anyway, it must be taken into account that phosphoscorodite is the mineral 43 variety present in the Barruecopardo mine wastes, which could entail a differential 44 45 behaviour. Thus, although considerable lower solubility values have been reported for 46 47 strengite [45], it has been found that the presence of phosphate leads to the formation of 48 Accepted Manuscript 49 P-rich scorodite of somewhat reduced stability [46]. This aspect together with the 50 51 simultaneous incorporation of phosphate during the scorodite crystal lattice formation 52 53 or the latter phosphate substitution for arsenate implies a reduced scavenger action of 54 this mineral in the Barruecopardo mining area. 55 56 Low solubility values have been also reported for HFO and goethite, with K values of 57 s 58 10-37.1 and 10-44.2, respectively [47]. In spite of the low solubility of HFO, their 59 60 conversion to goethite can occur spontaneously in acidic media (also in basic media) 61 62 63 13 64 Page 13 of 30 65 through reconstructive transformation. However, high loads of As on HFO inhibit the 1 2 transformation of these to the crystalline goethite [48]. This has important 3 environmental implications since such decomposition proceeds with a release of As. 4 5 Moreover, when goethite dominates the stability field of scorodite is significantly 6 7 smaller than when HFO are the dominant species [44]. In this regard, the As loads 8 9 shown by the HFO present in the Barruecopardo mine wastes are lower than those of 10 11 the Terrubias mine, which may suppose a reduced inhibition of the HFO transformation 12 13 to goethite, and so higher environmental risks. It is worth mentioning that, although 14 goethite was found in mine wastes from both areas, in the case of Barruecopardo mine 15 16 wastes HFO and goethite are present as a continuous mixture, whereas in the Terrubias 17 18 mine wastes HFO and goethite are mainly present as differentiated phases, with goethite 19 20 only appearing very locally. 21 -23.0 22 Concerning AFA, these have been proven to display higher solubilities (Ks: 10 ) and 23 24 dissolution rates than scorodite, about 100 times higher [2,40,43,49]. AFA show low 25 26 stability, dissolving incongruently from a lower pH value than scorodite [43]. These 27 compounds are considered the precursors to scorodite formation, being a dissolution 28 29 process suggested as the mechanism involved in this transformation [50]. AFA were 30 31 only detected in the Terrubias mining area, showing a really restricted presence, which 32 33 is consistent with the mentioned characteristics and transformation. Such conversion 34 35 should limit the As dispersion into the environment. 36 37 To our knowledge, no solubility and thermodynamic data have been determined for 38 39 pharmacosiderite, so its role in the As control in the Barruecopardo mine wastes in 40 41 which this mineral appears cannot be assessed. 42 43 In general, it seems that the secondary products present in the studied mining areas 44 should restrict importantly the As release from mine wastes provided that environmental 45 46 conditions keep more or less stable. This aspect is especially important for scorodite 47 48 (the most abundantAccepted arsenopyrite weathering product). Manuscript The low pH displayed by mine 49 50 wastes is responsible for the stability and low solubility of this mineral, and so for the 51 52 derived As immobilisation. This low pH should be also determinant for the presence 53 54 and stability of AFA. Anyway, in the case of Barruecopardo the P-rich environment and 55 the relatively reduced As load shown by the HFO there present could enhance the As 56 57 mobility and migration. 58 59 60 61 62 63 14 64 Page 14 of 30 65 4. CONCLUSIONS 1 2 The long-term weathering of arsenopyrite-rich wastes from tungsten and tin mines, 3 4 particularly from the abandoned exploitations of Barruecopardo and Terrubias (Spain), 5 6 has generated diverse secondary As products, namely scorodite, pharmacosiderite and 7 AFA with Fe/As molar ratios ranging from 1.2 to 2.5. Scorodite was found to prevail in 8 9 both studied mining areas. This mineral occurs in variable particles sizes and textures 10 11 and with different chemical composition depending on the original surrounding rock 12 13 mineralogy. Thus, phosphoscorodite was the mineral variety detected in apatite- 14 15 containing geoenvirons, with P contents up to 9% P2O5. Goethite and HFO were also 16 17 identified as weathering phases, showing, respectively, As loads up to 1.4 and 22% 18 As2O5. The characteristics and contents of all these secondary phases, i.e. low solubility 19 20 of scorodite, low content of AFA and high capacity of HFO to adsorb As, should limit 21 22 importantly the As release from these arsenopyrite-bearing wastes. Although in the case 23 24 of wastes generated from P-rich geoenvirons and where the generated HFO display 25 26 relatively reduced As loads (as happens in the Barruecopardo mining area) the As 27 mobility and migration could be enhanced. Anyway, the environmental characterisation 28 29 of these wastes reveals their toxic or hazardous character in terms of the TCLP test, the 30 31 bioavailable fraction of As and the acid generation capacity. Therefore, proper planning 32 33 should be undertaken to manage these wastes. 34 35 36 37 ACKNOWLEDGMENTS 38 39 This study was carried out under the projects SA091A08 funded by the “Junta de 40 41 Castilla y León” and CGL2008-06357 funded by the Ministry of Science and 42 43 Innovation of Spain. 44 45 46 47 REFERENCES 48 49 [1] B.J. Alloway,Accepted Heavy metals in soils, Blackie Manuscript Academic & Professional, Glasgow, 50 51 1995. 52 53 [2] P. Drahota, M. Filippi, Secondary arsenic minerals in the environment: A review, 54 55 Environ. Int. 35 (2009) 1243-1255. 56 57 [3] G.S. Pokrovski, S. Kara, J. Roux, Stability and solubility of arsenopyrite, FeAsS, in 58 59 crustal fluids, Geochim. Cosmochim. Ac. 66 (2002) 2361-2378. 60 61 62 63 15 64 Page 15 of 30 65 [4] D. Craw, D. Falconer, J.H. Youngson, Environmental arsenopyrite stability and 1 2 dissolution: theory, experiment, and field observations, Chem. Geol. 199 (2003) 3 71-82. 4 5 6 [5] C.L. Corkhill, D.J. Vaughan, Arsenopyrite oxidation-A review, Appl. Geochem. 24 7 (2009) 2342-2361. 8 9 [6] R.J. King, Minerals explained 35: Arsenopyrite, Geol. Today 18 (2002) 72-75. 10 11 12 [7] M. Filippi, V. Goliáš, Z. Pertold, Arsenic in contaminated soils and anthropogenic 13 deposits at the Mokrsko, Roudný, and Kašperské Hory gold deposits, Bohemian 14 15 Massif (CZ), Environ. Geol. 45 (2004) 716-730. 16 17 [8] D. Paktunc, A. Foster, S. Heald, G. Laflamme, Speciation and characterization of 18 19 arsenic in gold ores and cyanidation tailings using X-ray absorption spectroscopy, 20 21 Geochim. Cosmochim. Ac. 68 (2004) 969-983. 22 23 [9] R.L. Flemming, K.A. Salzsauler, B.L. Sherriff, N.V. Sidenko, Identification of 24 25 scorodite in fine-grained, high-sulfide, arsenopyrite mine-waste using micro x-ray 26 27 diffraction (μXRD), Can. Mineral. 43 (2005) 1243-1254. 28 29 [10] D. Mains, D. Craw, Composition and mineralogy of historic gold processing 30 31 residues, east Otago, New Zealand, New Zeal. J. Geol. Geop. 48 (2005) 641-647. 32 33 [11] K.A. Salzsauler, N.V. Sidenko, B.L. Sherriff, Arsenic mobility in alteration 34 35 products of sulfide-rich, arsenopyrite-bearing mine wastes, Snow Lake, Manitoba, 36 Canada, Appl. Geochem. 20 (2005) 2303-2314. 37 38 [12] M.F. Lengke, C. Sanpawanitchakit, R.N. Tempel, The oxidation and dissolution of 39 40 arsenic-bearing sulfides, Can. Mineral. 47 (2009) 593-613. 41 42 [13] H.W. Nesbitt, I.J. Muir, Oxidation states and speciation of secondary products on 43 44 pyrite and arsenopyrite reacted with mine waste and waters and air, Miner. Petrol. 45 46 62 (1998) 123-144. 47 48 [14] R. Gieré,Accepted N.V. Sidenko, E.V. Lazareva, Manuscript The role of secondary minerals in 49 50 controlling the migration of arsenic and metals from high-sulfide wastes (Berikul 51 52 gold mine, Siberia), Appl. Geochem. 18 (2003) 1347-1359. 53 54 [15] S.R. Walker, H.E. Jamieson, A. Lanzirotti, C.F. Andrade, G.E.M. Hall, The 55 56 speciation of arsenic in iron oxides in mine wastes from the giant gold mine, 57 58 N.W.T.: Application of synchrotron micro-XRD and micro-xanes at the grain 59 scale, Can. Mineral. 43 (2005) 1205-1224. 60 61 62 63 16 64 Page 16 of 30 65 [16] S.R. Walker, M.B. Parsons, H.E. Jamieson, A. Lanzirotti, Arsenic mineralogy of 1 2 near-surface tailings and soils: Influences on arsenic mobility and bioaccessibility 3 in the nova scotia gold mining districts, Can. Mineral. 47 (2009) 533-556. 4 5 6 [17] L. Haffert, D. Craw, Mineralogical controls on environmental mobility of arsenic 7 from historic mine processing residues, New Zealand, Appl. Geochem. 23 (2008) 8 9 1467-1483 10 11 [18] P.-K. Lee, M.-J. Kang, S.-H. Choi, J.-C. Touray, Sulfide oxidation and the natural 12 13 attenuation of arsenic and trace metals in the waste rocks of the abandoned Seobo 14 15 tungsten mine, Korea, Appl. Geochem. 20 (2005) 1687-1703. 16 17 [19] M.E.P. Gomes, P.J.C. Favas, Mineralogical controls on mine drainage of the 18 19 abandoned Ervedosa tin mine in north-eastern Portugal, Appl. Geochem. 21 20 21 (2006) 1322-1334. 22 23 [20] B.M. Petrunic, T.A. Al, L. Weaver, A transmission electron microscopy analysis of 24 25 secondary minerals formed in tungsten-mine tailings with an emphasis on 26 arsenopyrite oxidation, Appl. Geochem. 21 (2006) 1259-1273. 27 28 29 [21] E. Pellitero, Geologic and genetic factors in the tungsten-deposits of the north of 30 the Salamanca province. Cuadernos do Laboratorio Xeolóxico de Laxe 2 (1981) 31 32 245-257; (in Spanish). 33 34 [22] E. Pellitero, The tungsten central-eastern region of Salamanca. Cuadernos do 35 36 Laboratorio Xeolóxico de Laxe 2 (1981) 227-244; (in Spanish). 37 38 [23] A.A. Sobek, W.A. Schuller, J.R. Freeman, R.M. Smith, Field and laboratory 39 40 methods applicable to overburden and minesoils, U.S EPA (EPA 600/2-78-054), 41 42 Cincinnati, Ohio, 1978. 43 44 [24] W.W. Wenzel, N. Kirchbaumer, T. Prohaska, G. Stingeder, E. Lombi, D.C. 45 46 Adriano, Arsenic fractionation in soils using an improved sequential extraction 47 48 procedureAccepted, Anal. Chim. Acta 436 (2001) 309 Manuscript-323. 49 50 [25] A. Kontopoulos, K. Komnitsas, A. Xenidis, N. Papassiopi, Environmental 51 52 characterisation of the sulphidic tailings in Lavrion, Miner. Eng. 8 (1995) 1209- 53 1219. 54 55 [26] S. Endo, Y. Terada, Y. Kato, I. Nakai, Chemical speciation of arsenic- 56 57 accumulating mineral in a sedimentary iron deposit by synchrotron radiation 58 59 multiple X-ray analytical techniques, Environ. Sci. Technol. 42 (2008) 7152-7158. 60 61 62 63 17 64 Page 17 of 30 65 [27] A. -Sanchez, E. Alvarez-Ayuso, F. Rodriguez-Martin, Sorption of As(V) by 1 2 some oxyhydroxides and clay minerals. Application to its immobilization in two 3 polluted mining soils, Clay Miner. 37 (2002) 187-194. 4 5 6 [28] P. Lakshmipathiraj, B.R.V. Narasimhan, S. Prabhakar, G. Bhaskar Raju, 7 Adsorption of arsenate on synthetic goethite from aqueous solutions, J. Hazard. 8 9 Mat. 136 (2006) 281-287. 10 11 [29] M.L. Pierce, C.B. Moore, Adsorption of arsenite and arsenate on amorphous iron 12 13 hydroxide, Water Res. 16 (1982) 1247-1253. 14 15 [30] B.A. Manning, S. Goldberg, Modeling competitive adsorption of arsenate with 16 17 phosphate and molybdate on oxide minerals, Soil Sci. Soc. Am. J. 60 (1996) 121- 18 19 131. 20 21 [31] E.E. Coleyshaw, W.P. Griffith, R.J. Bowell, Fourier-transform raman spectroscopy 22 23 of minerals, Spectrochimica Acta Part A: Molecular Spectroscopy, 50 (1994) 24 25 1909-1918. 26 27 [32] M. Filippi, V. Machovič, P. Drahota, V. Böhmová, Raman microspectroscopy as a 28 29 valuable additional method to x-ray diffraction and electron 30 microscope/microprobe analysis in the study of iron arsenates in environmental 31 32 samples, Appl. Spectrosc. 63 (2009) 621-626. 33 34 [33] M.A. Gomez, H. Assaaoudi, L. Becze, J.N. Cutler, G.P. Demopoulos, Vibrational 35 36 spectroscopy study of hydrothermally produced scorodite (FeAsO4·2H2O), ferric 37 38 arsenate sub-hydrate (FAsH; FeAsO4·0.75H2O) and basic ferric arsenate sulfate 39 40 (BFAS; Fe[(AsO4)1-x(SO4)x(OH)x]·wH2O), J. Raman Spectrosc. 41 (2010) 212- 41 42 221. 43 44 [34] D.L.A. de Faria, S. Venâncio Silva, M.T. de Oliveira, Raman microspectroscopy of 45 46 some iron oxides and oxyhydroxides, J. Raman Spectrosc. 28 (1997) 873-878. 47 48 [35] M. Takano,Accepted T. Takada, T. Wada, K. Okada, Manuscript Magnetic properties of FeAsO4·2H2O 49 and KFe (OH) (AsO ) ·8H O, J. Phys. Soc. Jpn 31 (1971) 298. 50 4 4 4 3 2 51 52 [36] R.E. Vandenberghe, E. De Grave, C. Landuydt, L.H. Bowen, Some aspects 53 concerning the characterization of iron oxides and hydroxides in soils and clays, 54 55 Hyp. Interact. 53 (1990) 175-196. 56 57 58 59 60 61 62 63 18 64 Page 18 of 30 65 [37] T.S. Berquó, R.A.L. Imbernon, A. Blot, D.R. Franco, M.C.M. Toledo, C.S.M. 1 2 Partiti, Low temperature magnetism and Mössbauer spectroscopy study from 3 natural goethite, Phys. Chem. Miner. 34 (2007) 287-294. 4 5 6 [38] P.M. Dove, J.A. Rimstidt, The solubility and stability of scorodite, FeAsO4·2H2O, 7 Am. Mineral. 70 (1985) 838-844. 8 9 [39] E. Krause, V.A. Ettel, Solubilities and stabilities of ferric arsenate compounds, 10 11 Hydrometallurgy 22 (1989) 311-337. 12 13 [40] D. Langmuir, J. Mahoney, J. Rowson, Solubility products of amorphous ferric 14 15 arsenate and crystalline scorodite (FeAsO4·2H2O) and their application to arsenic 16 17 behavior in buried mine tailings, Geochim. Cosmochim. Ac. 70 (2006) 2942- 18 19 2956. 20 21 [41] M.C. Bluteau, G.P. Demopoulos, The incongruent dissolution of scorodite - 22 23 Solubility, kinetics and mechanism, Hydrometallurgy 87 (2007) 163-177. 24 25 [42] D.K. Nordstrom, D.G. Archer, Arsenic thermodynamic data and environmental 26 27 geochemistry. In: Arsenic in ground water (A.H. Welch, K.G. Stollenwerk, 28 29 editors.), Kluwer Academic Publishers, Boston, Massachusetts, USA, 2003. 30 31 [43] D. Paktunc, K. Bruggeman, Solubility of nanocrystalline scorodite and amorphous 32 ferric arsenate: Implications for stabilization of arsenic in mine wastes, Appl. 33 34 Geochem. 25 (2010) 674-683. 35 36 [44] M.C. Harvey, M.E. Schreiber, J.D. Rimstidt, M.M. Griffith, Scorodite dissolution 37 38 kinetics: Implications for arsenic release, Environ. Sci. Technol. 40 (2006) 6709- 39 40 6714. 41 42 [45] M. Iuliano, L. Ciavatta, G. De Tommaso, On the solubility constant of strengite, 43 44 Soil Sci. Soc. Am.J. 71 (2007) 1137-1140. 45 46 [46] S. Singhania, Q. Wang, D. Filippou, G.P. Demopoulos, Acidity, valency and third- 47 48 ion effectsAccepted on the precipitation of scorodite Manuscript from mixed sulfate solutions under 49 50 atmospheric-pressure conditions, Metallurgical and Materials Transactions B: 51 52 Process Metallurgy and Materials Processing Science 37 (2006) 189-197. 53 54 [47] D.O. Whittemore, D. Langmuir, Ferric oxyhydroxide microparticles in water, 55 Environ. Health Persp. 9 (1974) 173-176. 56 57 58 [48] R.G. Ford, Rates of hydrous ferric oxide crystallization and the influence on 59 coprecipitated arsenate, Environ. Sci. Tecnol. 36 (2002) 2459-2463. 60 61 62 63 19 64 Page 19 of 30 65 [49] N.J. Welham, K.A. Malatt, S. Vukcevic, The stability of iron phases presently used 1 2 for disposal from metallurgical systems-a review, Miner. Eng. 13 (2000) 911-931. 3 4 [50] D. Paktunc, J. Dutrizac, V. Gertsman, Synthesis and phase transformations 5 6 involving scorodite, ferric arsenate and arsenical ferrihydrite: Implications for 7 arsenic mobility, Geochim. Cosmochim. Ac. 72 (2008) 2649-2672. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Accepted Manuscript 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 20 64 Page 20 of 30 65 Table 1. Chemical analyses of the arsenopyrite weathering products in the Barruecopardo mine wastes (Scms, Scsf, Sccl and Sccry: scorodite microcrystalline 1 massive, spherulitic, colloform and in crystals, and R-b ph: Red-blackish phases). 2 Phases Fe2O3 (%) As2O5 (%) P2O5 (%) SO3 (%) Al2O5 (%) SiO2 (%) CuO (%) 3 Sc 35.0 44.9 4.16 n.d. 2.21 n.d. n.d. 4 ms Sc 34.9 45.0 4.34 n.d. 2.53 n.d. n.d. 5 ms Sc 35.3 43.4 4.24 n.d. 2.14 n.d. n.d. 6 ms Sc 35.6 48.1 0.26 0.76 0.17 n.d. n.d. 7 ms Sc 35.5 48.2 0.27 0.61 0.53 n.d. n.d. 8 ms Sc 36.1 49.5 0.36 0.42 0.84 n.d. n.d. 9 ms Sc 35.0 48.3 0.33 0.32 0.94 n.d. n.d. 10 ms Sc 36.6 47.0 1.33 1.15 n.d. n.d. n.d. 11 ms Sc 34.8 43.2 2.20 0.12 0.27 0.73 n.d. 12 ms 35.3 45.9 1.59 0.48 0.10 0.23 n.d. 13 Scms 14 Scms 37.4 43.3 5.35 n.d. n.d. n.d. n.d. 15 Scms 37.3 42.8 3.53 0.87 n.d. n.d. n.d. 16 Scms 37.3 42.9 4.25 0.38 n.d. n.d. n.d. 17 Scms 35.9 48.0 1.37 0.75 n.d. n.d. n.d. 18 Scms 36.8 40.5 6.32 0.45 n.d. n.d. n.d. 19 Scms 37.4 43.4 5.28 n.d. n.d. n.d. n.d. 20 Scms 37.9 41.4 6.73 n.d. n.d. n.d. n.d. 21 Scms 33.1 39.2 8.02 n.d. 0.40 n.d. n.d. 22 Scms 35.0 47.0 0.88 1.25 n.d. n.d. n.d. 23 Scms 35.5 46.8 0.90 1.02 n.d. n.d. n.d. 24 Scms 36.6 42.2 5.33 n.d. n.d. n.d. n.d. 25 Scms 37.4 42.8 5.64 n.d. n.d. n.d. n.d. 26 Scsf 37.2 46.5 1.26 0.86 n.d. n.d. n.d. 27 Scsf 37.1 46.1 1.13 0.91 n.d. n.d. n.d. 28 Scsf 36.3 45.0 0.96 1.05 0.18 n.d. n.d. 29 Scsf 36.5 45.6 1.82 0.99 0.45 0.50 n.d. 30 Scsf 35.2 45.7 0.85 0.79 1.17 1.17 n.d. 31 Sccl 37.0 46.8 2.08 n.d. 0.16 n.d. n.d. 32 Sccl 37.1 45.7 1.44 0.90 0.33 n.d. n.d. 33 Sccl 37.5 45.7 2.02 0.51 0.20 n.d. n.d. 34 Sccl 36.9 47.8 0.93 n.d. 0.37 n.d. n.d. 35 Sccl 37.0 46.4 2.07 0.26 0.27 n.d. n.d. 36 Sccl 37.7 43.9 4.74 0.01 0.24 n.d. n.d. 37 Sccl 37.2 44.9 1,60 1.09 0.18 n.d. n.d. 38 Sccl 36.5 46.9 0.75 0.96 n.d. n.d. n.d. 39 Sccl 36.7 45.2 1.87 1.29 0.16 n.d. 0.56 40 Sccl 38.2 40.7 6.36 1.49 0.44 n.d. n.d. 41 Sccl 39.2 42.8 1.17 0.74 0.34 n.d. n.d. 42 Sccl 37.8 44.4 1.52 1.03 0.21 n.d. n.d. 43 Sccry 38.8 46.8 1.42 n.d. 0.14 n.d. n.d. 44 Sccry 39.0 47.3 2.29 n.d. 0.20 n.d. n.d. 45 Sccry 39.0 40.9 8.84 n.d. 0.13 n.d. n.d. 46 Sccry 37.8 42.5 7.92 n.d. 2.22 n.d. n.d. 47 Sccry 39.2 45.1 5.54 n.d. 0.91 n.d. n.d. 48 Sc 37.4 43.0 7.02 n.d. 3.08 1.10 n.d. cry Accepted Manuscript 49 Sccry 39.2 45.0 7.06 n.d. 0.66 n.d. n.d. 50 Sccry 41.7 40.7 9.01 n.d. 0.42 0.26 n.d. 51 Sccry 39.8 46.6 2.87 n.d. 0.69 n.d. n.d. 52 Sccry 37.4 47.4 4.59 n.d. 2.15 n.d. n.d. 53 Sccry 36.1 41.8 7.77 n.d. 2.57 0.12 n.d. 54 R-b ph 57.3 5.66 0.30 n.d. 1.86 n.d. n.d. 55 R-b ph 55.8 5.93 0.34 n.d. 1.89 n.d. n.d. 56 R-b ph 61.3 6.10 0.37 n.d. 2.26 n.d. n.d. 57 R-b ph 57.1 11.9 0.78 0.24 n.d. n.d. n.d. 58 R-b ph 62.3 6.66 0.63 0.68 n.d. n.d. n.d. 59 R-b ph 65.6 6.89 0.71 0.78 n.d. n.d. n.d. 60 61 62 63 21 64 Page 21 of 30 65 Table 2. Chemical analyses of the arsenopyrite weathering products in the Terrubias mine wastes (Sc: scorodite and Gth: goethite). 1 Fe O As O SO Al O SiO CuO K O BaO ZnO CaO Phases 2 3 2 5 3 2 3 2 2 2 (%) (%) (%) (%) (%) (%) (%) (%) (%) (%) 3 Sc 37.0 49.2 0.20 n.d. n.d. 0.16 n.d. n.d. n.d. n.d. 4 Sc 34.0 49.4 0.04 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 5 Sc 35.5 51.2 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 6 Sc 37.5 49.5 0.12 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 7 Sc 36.1 49.5 0.31 n.d. n.d. 0.12 n.d. 0.04 n.d. n.d. 8 Sc 35.8 49.2 0.12 n.d. n.d. 0.28 n.d. 0.09 n.d. n.d. 9 Sc 35.9 49.6 0.12 n.d. n.d. 0.11 n.d. n.d. n.d. n.d. 10 Sc 36.2 50.0 0.26 n.d. n.d. 0.23 n.d. n.d. n.d. n.d. 11 Sc 37.8 49.1 0.08 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 12 Sc 36.8 49.7 0.10 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 13 Sc 35.5 49.7 0.20 0.12 n.d. n.d. n.d. n.d. n.d. n.d. 14 Sc 35.1 47.1 0.34 0.20 n.d. n.d. n.d. n.d. n.d. n.d. 15 Sc 36.0 49.4 0.14 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 16 Sc 37.2 49.8 0.08 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 17 Sc 36.9 50.0 0.19 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 18 Sc 37.2 49.6 0.34 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 19 Sc 36.8 49.7 0.28 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 20 Sc 36.5 49.3 0.24 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 21 Sc 36.3 49.2 0.24 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 22 Sc 34.9 48.8 0.25 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 23 Sc 36.9 50.4 0.23 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 24 Sc 36.9 49.9 0.22 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 25 37.5 49.5 0.29 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 26 Sc 36.6 46.4 0.63 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 27 Sc 28 Sc 36.9 50.9 0.11 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 29 Sc 37.0 50.0 0.16 0.12 n.d. n.d. n.d. n.d. n.d. n.d. 30 Sc 37.2 49.8 0.20 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 31 AFA 40.4 39.0 2.14 n.d. n.d. 0.98 0.23 0.19 n.d. n.d. 32 AFA 37.1 25.3 3.07 n.d. n.d. 0.33 n.d. n.d. n.d. n.d. 33 AFA 50.2 28.7 0.36 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 34 AFA 33.2 28.1 0.18 n.d. n.d. n.d. n.d. n.d. 0.22 n.d. 35 AFA 49.1 31.3 0.42 n.d. n.d. n.d. n.d. n.d. n.d. 0.10 36 AFA 49.0 31.2 0.40 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 37 AFA 37.3 39.1 0.04 n.d. n.d. n.d. 2.42 0.20 n.d. n.d. 38 AFA 36.3 37.3 2.28 n.d. n.d. n.d. 0.67 4.10 n.d. n.d. 39 AFA 43.5 42.1 0.33 n.d. n.d. 0.37 n.d. n.d. n.d. n.d. 40 AFA 43.2 41.2 0.47 n.d. n.d. 0.11 n.d. n.d. n.d. 0.10 41 AFA 30.7 33.7 1.47 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 42 AFA 31.1 37.0 1.27 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 43 AFA 36.4 34.2 3.84 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 44 AFA 39.9 35.4 3.73 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 45 Gth 59.0 0.38 n.d. n.d. 9.43 n.d. n.d. n.d. n.d. n.d. 46 Gth 65.6 0.32 n.d. n.d. 6.08 n.d. n.d. n.d. n.d. n.d. 47 Gth 66.2 0.73 n.d. n.d. 5.86 n.d. n.d. n.d. n.d. n.d. 48 Gth 75.1 Accepted0.51 n.d. n.d. 1.65 Manuscriptn.d. n.d. n.d. n.d. n.d. 49 Gth 68.3 0.70 n.d. n.d. 2.21 n.d. n.d. n.d. n.d. n.d. 50 Gth 77.3 0.86 n.d. n.d. 0.99 n.d. n.d. n.d. n.d. n.d. 51 Gth 79.1 1.38 n.d. n.d. 0.69 n.d. n.d. n.d. n.d. n.d. 52 Gth 74.3 0.91 n.d. n.d. 0.77 n.d. n.d. n.d. n.d. n.d. 53 Gth 73.2 0.92 n.d. n.d. 0.56 n.d. n.d. n.d. n.d. n.d. 54 HFO 62.2 18.1 1.41 n.d. n.d. 0.29 n.d. 0.13 n.d. n.d. 55 HFO 51.7 18.7 1.80 n.d. n.d. 0.10 n.d. n.d. n.d. 0.20 56 HFO 50.1 16.7 1.08 n.d. n.d. 0.11 n.d. n.d. n.d. 0.10 57 HFO 61.3 21.7 0.59 n.d n.d 0.36 n.d n.d n.d 0.10 58 HFO 63.3 16.4 0.45 n.d. 1.56 0.45 n.d. n.d. n.d. 0.10 59 60 61 62 63 22 64 Page 22 of 30 65

Table 3. Environmental characterisation of the Barruecopardo and Terrubias mine wastes. 1 2 Barruecopardo Terrubias 3 -1 NNP (kg CaCO3 t ) - 60 - 350 4 5 pH 1.82 1.82 6 7 As 3.67 12.9 8 9 Cu 0.008 0.319 10 TCLP concentrations (mg L-1) Pb 0.056 0.671 11 12 Zn 0.013 0.743 13 14 Cd 0.011 0.049 15 Bioavailable fraction (mg kg-1) 16 As 76.4 52.1 17 (NH4)2SO4 procedure 18 Cu 0.608 12.9 19 20 Bioavailable fraction (mg kg-1) Pb 33.0 65.9 21 EDTA procedure 22 Zn 0.328 12.6 23 Cd < 0.05 0.288 24 25 TCLP limits for toxic characterisation: As (5 mg L-1), Cd (1 mg L-1) and Pb (5 mg L-1). 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Accepted Manuscript 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 23 64 Page 23 of 30 65

Table 4. Water quality parameters of ponds present in the Barruecopardo and Terrubias 1 mining areas. 2 3 Barruecopardo Terrubias 4 pH 3.45 7.81 5 6 Eh (mV) 518 142 7 Sulphate (mg L-1) 600 185 8 -1 9 As (mg L ) 0.218 0.133 10 Cu (mg L-1) 0.444 < 0.005 11 -1 12 Pb (mg L ) < 0.05 < 0.05 13 Zn (mg L-1) 1.863 < 0.005 14 -1 15 Cd (mg L ) 0.009 < 0.005 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Accepted Manuscript 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 24 64 Page 24 of 30 65 1 2 3 4 5 6 7 8 9 10 11 12 Barruecopardo Salamanca 13 Terrubias 14 15 16 17 18 19 20 21 22 23 24 25 26 1 Km 27 28 Shales Granites xx Granites Granites Mineralised veins 29 ++ 30 31 32 33 34 35 36 37 1 Km 38 Accepted Manuscript 39 Granites Tertiary Porfiroids Quartzites 40 Infraordovicic Mineralised veins 41 42 Figure 1. Localisation of the Barruecopardo and Terrubias mines and geologic outcrops in these areas. 43 44 45 46 47 25 48 Page 25 of 30 49

1 a b Sc Q 2 Apy 3 4 5 W.P. 6 Apy 7 8 9 10 Sc 11 12 1 cm 1 mm 13 14 15 16 17 c d Sc 18 Apy 19 Apy 20 21 22 Q 23 Sc Gth/HFO 24 25 26 27 28 200 µm 29 500 µm 30 31 32 33 e f 34 35 S 36 c 37 38 39 40 41 42 Sc 43 44 50 µm 50 µm 45 46 47 48 Figure 2. ImagesAccepted of the arsenopyrite weathering Manuscript products from Barruecopardo mine 49 wastes: a) Hand sample of arsenopyrite, b) Microscopic image in polarizing transmitted 50 51 light of scorodite on the borders, in veins and inside of arsenopyrite, c) Microscopic 52 image in polarizing transmitted light of scorodite crystals, d) Microscopic image in 53 polarizing transmitted light of red-blackish phases occurring on the borders of scorodite, 54 e) Electron backscatter image of zoned crystals of scorodite, showing intensified grey 55 tonalities towards the crystal cores (detail of c) and f) Electron backscatter image of 56 57 colloform scorodite with zonation, showing intensified grey tonalities towards the cores 58 (WP: weathering products, Sc: scorodite, Apy: arsenopyrite, Gth: goethite and Q: 59 quartz). 60 61 62 63 26 64 Page 26 of 30 65

1 a b 2 3 4 5 6 Sc1 7 8 Apy 9 W.P. Sc2 10 11 12 100 µm 13 1 cm Q 14 15 16 17 c d 18 Gth/HFO 19 20 21 Sc 22 Gth/HFO Apy 23 24 AFA 25 Sc 26 27 28 Apy 29 500 µm 1 mm 30 31 32 33 e f 34 Sc AFA 35 36 37 38 39 40 41 42 43 Apy 44 50 µm 50 µm 45 Apy 46 47 48 Figure 3. ImagesAccepted of the arsenopyrite weathering Manuscript products from Terrubias mine wastes: 49 a) Hand sample of arsenopyrite, b) Microscopic image in polarizing transmitted light of 50 sferulitic scorodite (Sc1) and microcrystalline massive scorodite (Sc2), c) Microscopic 51 image in polarizing transmitted light of weathering products of arsenopyrite (scorodite, 52 53 goethite and HFO), d) Microscopic image in polarizing transmitted light of weathering 54 products of arsenopyrite, showing orange-reddish phases corresponding to AFA, e) 55 Electron backscatter image of colloform scorodite and f) Electron backscatter image of 56 AFA (detail of d) (WP: weathering products, Sc: scorodite, Apy: arsenopyrite, Gth: 57 goethite and Q: quartz). 58 59 60 61 62 63 27 64 Page 27 of 30 65

1 Sc + M 2 400 Sc Fd 3 a b 4 Sc + Fd 150 5 300 6 7 100 Ph 8 200 M

9 (cps) Intensity Sc + Apy Sc (cps) Intensity Fd Gth 10 Apy M 50 Gth 100 Gth

Sc 11 Sc Sc 12 13 14 0 0 15 20 40 60 20 40 60 2 (º) 2 (º) 16 400 17 Sc Sc 18 c Q Intensity (Counts) Intensity 500 19 Q d 20 300 Sc 21 400 22

23 200 300 24 Sc Sc

Intensity (cps) Intensity Gth

25 Dr (cps) Intensity 200 26 100 Q 27 Gth 100 Q 28 Dr Gth Q Gth 29 0 0 30 20 40 60 20 40 60 31 2 (º) 2 (º) 32 33 2 (°) 34 Figure 4. XRD patterns of the yellow-greenish and orange-reddish phases occurring on the 35 weathered waste rocks from the Barruecopardo and Terrubias mines (a: yellow-greenish phases 36 37 (Barruecopardo), b: orange-reddish phases (Barruecopardo), c: yellow-greenish phases 38 (Terrubias) and d: orange-reddish phases (Terrubias); M: muscovite, Sc: scorodite, Fd: K- 39 feldespar, Apy: arsenopyrite, Ph: pharmacosiderite, Gth: goethite, Dr: dravite and Q: quartz). 40 41 42 43 44 45 46 47 48 Accepted Manuscript 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 28 64 Page 28 of 30 65

800 1 1000 a b 2 700 3 900 4 600 800 5 700 6 500 600 7 400 8 500 9 300 400 10 300 11 200 200 12 100 13 100 14 0 0 15 500 1000 1500 2000 2500 3000 16 500 1000 1500 2000 2500 3000 17 18 19 700 700 20 c d 21 600 600 22 (Counts) Intensity Raman 500 23 500 24 400 25 400 26 300 300 27 28 200 200 29 100 30 100 31 0 0 32 33 500 1000 1500 2000500 25001000 30001500 2000 2500 3000 34 35 Wavenumbers (cm-1) 36 37 Figure 5. Representative microRaman spectra of the arsenopyrite weathering products 38 in the Barruecopardo and Terrubias mine wastes (a: scorodite (Barruecopardo), b: 39 goethite (Barruecopardo), c: scorodite (Terrubias) and d: goethite (Terrubias)). 40 41 42 43 44 45 46 47 48 Accepted Manuscript 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 29 64 Page 29 of 30 65

1 2 3 4 5 6 7 8 9 a 10

11 b 12 -3 13 2x10 14 0 15 Error -2 16 17 18 1.0 19 Intensity Relative 20 21 22 0.9 23 RelativeCounts c 24 25 0.8 -3 -2 -1 0 1 2 3 26 d 27 Velocity (mm/s) 28 29 30 31 32 Velocity (mm s-1) 33 34 Figure 6. Representative Mössbauer spectra of the yellow-greenish (a and c) and 35 orange-reddish phases (b and d) occurring on the weathered waste rocks from the 36 Barruecopardo and Terrubias mines (a: scorodite (Barruecopardo), b: scorodite + 37 38 pharmacosiderite and goethite (Barruecopardo), c: scorodite (Terrubias) and d: goethite 39 (Terrubias)). 40 41 42 43 44 45 46 47 48 Accepted Manuscript 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 30 64 Page 30 of 30 65