Piatak 2004

Applied Geochemistry 19 (2004) 10391064 www.elsevier.
com/locate/apgeochem
Mineralogical and geochemical controls on the release of trace elements from slag produced by base- and precious-metal smelting at abandoned mine sites
Nadine M. Piatak*, Robert R. Seal II, Jane M. Hammarstrom
US Geological Survey, 954 National Center, Reston, VA 20192, USA
Abstract Slag collected from smelter sites associated with historic base-metal mines contains elevated concentrations of trace elements such as Cu, Zn and Pb. Weathering of slag piles, many of which were deposited along stream banks, potentially may release these trace elements into the environment. Slags were sampled from the Ely and Elizabeth mines in the Vermont copper belt, from the copper Basin mining district at Ducktown, Tennessee and from the Clayton silver mine in the Bayhorse mining district, Idaho, in the USA. Primary phases in the slags include: olivine-group minerals, glass, spinels, sulde minerals and native metals for Vermont samples; glass, sulde minerals and native metals for the Ducktown sample; and olivine-group minerals, clinopyroxenes, spinels, sulde minerals, native metals and other unidentied metallic compounds for Clayton slag. Olivine-group minerals and pyroxenes are dominantly fayalitic and hedenbergitic in composition, respectively and contain up to 1.25 wt.% ZnO. Spinel minerals range between magnetite and hercynite in composition and contain Zn (up to 2.07 wt.% ZnO), Ti (up to 4.25 wt.% TiO2) and Cr (up to 1.39 wt.% Cr2O3). Cobalt, Ni, Cu, As, Ag, Sb and Pb occur in the glass phase, suldes, metallic phases and unidentied metallic compounds. Bulk slag trace-element chemistry shows that the metals of the Vermont and Tennessee slags are dominated by Cu (190013,500 mg/kg) and Zn (231010,200 mg/kg), whereas the Clayton slag is dominated by Pb (63,000 mg/kg), Zn (19,700 mg/kg), Cu (7550 mg/kg), As (555 mg/kg), Sn (363 mg/kg) and Ag (200 mg/kg). Laboratory-based leach tests indicate metals can be released under simulated natural conditions. Leachates from most slags were found to contain elevated concentrations of Cu and Zn (up to 1800 and 470 mg/l, respectively), well in excess of the acute toxicity guidelines for aquatic life. For the Idaho slag, the concentration of Pb in the leachate (11,000 mg/l) is also in excess of the acute toxicity guideline. Geochemical modeling of the leachate chemistry suggests that leachates from the Vermont, Tennessee and Clayton slags are saturated with amorphous silica and Al hydroxide. Therefore, the dissolution of silicate and oxide phases, the oxidation of sulde phases, as well as the precipitation of secondary phases may control the composition of leachate from slags. The presence of secondary minerals on slag deposits in the eld is evidence that these materials are reactive. The petrographic data and results of leaching tests from this study indicate slag may be a source of potentially toxic metals at abandoned mine sites. Published by Elsevier Ltd.
1. Introduction Metal mining produces a variety of waste materials including waste rock, otation tailings and slag. Slag is either dumped on site or may be used as road ballast or
* Corresponding author. E-mail address: [email protected] (N.M. Piatak). 0883-2927/$ - see front matter Published by Elsevier Ltd. doi:10.1016/j.apgeochem.2004.01.005
backll material, as sandblasting agents, as cement additives, or as material for roong shingles (Collins and Miller, 1979; Robbins et al., 1983; Wilson, 1994). Slag produced by base-metal smelting typically contains high concentrations of potentially toxic metals. Slag is generally considered chemically inert because metals are associated with glass and silicate and oxide minerals that have low solubilities in water (Scales, 1986; Wilson, 1994). However, recent studies have shown that smelter
1040
N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064
slag is not as inert as originally believed (Kucha et al., 1996; Manz and Castro, 1997; Parsons et al., 2001; Lottermoser, 2002). Trace metals leached from slag may contaminate soils and streams and have deleterious eects on the environment. The mineralogy of slags has been well characterized by a number of studies that emphasized the geochemistry and weathering behavior of historical base- and precious-metal slags (e.g. Maskall et al., 1996; Gee et al., 1997; Maskall and Thorton, 1998; Parsons et al., 2001; Lottermoser, 2002). These studies demonstrated that metals are released from slag waste by natural weathering processes and thus may represent an important source of metals in waters and soils at historical smelting sites. Kucha et al. (1996) and Ettler et al. (2001b) conducted leaching tests on the surfaces of polished sections. Laboratory-based batch-extraction tests have been carried out in several studies using a variety of protocols including the Extraction Procedure Toxicity Test (EP Tox, USEPA Method 1310; Johnson et al., 1982; Robbins et al., 1983) and a modication of it (Woodley and Walters, 1986); the Toxicity Characteristic Leaching Procedure (TCLP, USEPA Method 1311; Lagos and Luraschi, 1997; Parsons et al., 2001) and a nonstandardized leaching procedure (Manz and Castro, 1997). Description of the US Environmental Protection Agency protocols can be found in USEPA (1994). The EP Tox and TCLP were designed to simulate leaching in a sanitary landll and use a leachate solution either of, or containing acetic acid, which preferentially binds Pb. Thus, the leaching studies do not approximate the ambient eld conditions for mine wastes. The Synthetic Precipitation Leaching Procedure (SPLP, Method 1312; USEPA, 1994) was developed to assess the impact of contaminated soils and wastes on groundwater; thus, it probably more accurately simulates eld conditions at mine sites when compared with the EP Tox and the TCLP. Hageman and Briggs (2000) developed a eld-leach test for characterization of mine waste, which is a modication of SPLP. The concentration of some elements in the eld-leach test leachate are lower than in leachate produced by the SPLP, although overall the results from these two procedures are similar. For the present study, a modied version of the eld-leach test by Hageman and Briggs (2000) was used and the results were compared with USEPA waterquality guidelines for aquatic biota to determine the potential for the release of metals. This paper reports the mineralogy, mineral chemistry, bulk major and trace-element chemistry and leachate chemistry of slags collected from a variety of historical base-metal smelting sites to understand the hosts of trace metals in the slags and the relative leachability of metals from these hosts. The suite of samples included pot slag (air-cooled) from the Elizabeth and Ely mines in the Vermont copper belt (Cu), granulated slag (quenched in
water) from the Ducktown mining district (Cu), Tennessee and slag (air-cooled) from the Clayton smelter (Pb Ag) in the Bayhorse mining district, Idaho to highlight the eects of ore composition and smelting techniques.
2. Geology and metallurgical operations 2.1. Vermont copper belt The Elizabeth and Ely mines worked Besshi-type massive sulde deposits located in the Vermont copper belt in Orange County, Vermont. Bedrock includes metasedimentary and minor mac metaigneous rock of Silurian and Early Devonian age. The mineral deposits are stratabound and generally stratiform and massive; they are composed of pyrrhotite, chalcopyrite and minor sphalerite and pyrite (Oeld and Slack, 1993; Slack et al., 2001). The following metallurgical history of the Vermont copper belt is summarized from Kierstead (2001) unless otherwise noted. The Elizabeth and Ely mines were in operation from 1809 to 1958 and from the 1850s to 1905, respectively. The Elizabeth mine had several smelting campaigns throughout its history. Samples were collected from 3 smelter locations: Furnace Flat on the north bank of the Ompompanoosuc River, active from 1854 to 1867 and in 1884; Sargent Brook, active from 1880 to 1890; and the Heckscher smelter, active from 1905 to 1909. Furnace Flat smelters included charcoal-fueled blast furnaces lined with acid-silica rebrick. The coke-fueled smelter erected at Sargent Brook was a steam-powered vertical blast furnace with a water-jacket cooled hearth. The Sargent Brook smelter was the site of the rst known American experiments using basic chromite refractory materials to prolong furnace life. Chromite sand was used to line the hearth bottom to counteract the corrosive acidsilica slag. The slag from this operation was deposited along Sargent Brook at the foot of the west slope of Copperas Hill. The matte produced by these two smelting operations was usually shipped out to be rened. The Heckscher smelter was a pyritic furnace designed to use the sulde sulfur in the ore as fuel. A tailings pile now buries the smelter site and slagheaps protrude from the north toe of the tailings. The ores from the Elizabeth mine were self-uxing; ore high in pyrrhotite, which was roasted separately from that high in chalcopyrite, was added to the charge when needed (Packard, 1908). The Ely smelting plant was built in 1867 and eventually became a massive operation consisting of a smelter shed 210 m in length. Initially, the facility consisted of 4 Continental, or German cupola-type furnaces. Besides processing ore from the Ely mine, the Ely smelter processed ore from the Union mine, located at Pike Hill several kilometers to the north between 1878 and 1882. By 1879, the plant consisted of a total of 24 furnaces of
1041
several types for processing ore in various stages of renement. In 1889, a water-cooled smelting furnace was installed and by 1900 a new ore separator, a Colorado-style furnace and Bessemer converters were built. Around this time, experiments were carried out on smelting low-grade ore without roasting. Basic magnesite refractory furnace linings were used instead of acid-silica rebricks. The pot-slag deposit extends for more than 90 m along the bank of Schoolhouse Brook and is locally up to several meters high. 2.2. Ducktown mining district The Ducktown mining district is located in the southeastern corner of Tennessee in Polk County. These deposits, like those in the Vermont copper belt, are Besshi-type massive sulde deposits. The host rock is a highly deformed and metamorphosed sequence of Late Proterozoic clastic metasedimentary rocks and local amphibolites. Nine separate sulde deposits have been mined in the district. Primary ore minerals include pyrrhotite with subordinate pyrite and minor chalcopyrite, sphalerite, magnetite and sparse galena (Magee, 1968). The deposits are tabular and generally conformable to the surrounding rocks (Slater et al., 1985). The Ducktown mining district was in operation from 1847 to 1987. Both pot slag and granulated slag were produced as smelter waste products. The sample from Ducktown is a granulated slag, in contrast to the pot slags collected from Vermont. At Ducktown, the process of granulating slag began around 1920. Only granulated slag was produced since shortly after World War II. The processing of the ore is as follows: after the ore was crushed and concentrates of Fe, Zn and Cu suldes separated in the otation plant, the Cu concentrate was roasted. The Cu calcine was then mixed with silicasand ux and melted in a reverberatory furnace with coal as the fuel. The matte was transferred to a Pierce-Smith converter, which is a cylindrical, bricklined vessel where air was blown through the molten matte to oxidize the sulde to SO2 and the Fe impurities to Fe oxide. The Fe impurities combined to form slag, which, after allowing any remaining Cu to settle, was granulated in a stream of owing water, which caused it to fracture into sand-sized (0.11 cm) particles (Anonymous; Blair, 1966). The SO2 was collected in a ue system to manufacture sulfuric acid. The granulated slag for this study was collected from a waste pile on the northwestern bank of Davis Mill Creek. Additional discussion on the characterization of mine waste at this site is discussed in Moyer et al. (2000, 2002). 2.3. Clayton smelter site The Clayton smelter processed PbAg ore from mines in the Bayhorse mining district in Custer County, Idaho.
The deposits in this district include polymetallic vein and replacement deposits, most of which are hosted by the Ordovician Ramshorn Slate, the Bayhorse Dolomite and the Kinnikinic Quartzite. Granitic rocks are also associated with many of the deposits. The major sulde minerals are galena, sphalerite, pyrite, tetrahedrite and chalcopyrite; both the galena and the tetrahedrite were argentiferous (Ross, 1937). The Clayton smelter operated intermittently from the 1880s to 1902, reopened for a time in 1913 and then again in 1935. The original 30-ton smelter was enlarged to a 60-t smelter in 1888. This smelter processed ore from several mines within the mining districts of Greyhound Ridge, Seafoam, Germania Basin and Bayhorse. The variety of ore supplied to the smelter was mixed to optimize chemical reactions for eective smelting. Imported coke, as well as local charcoal, was used as fuel (Wells, 1983). The smelter is located on the north bank of the Salmon River and slag was dumped directly into the river.
3. Methodology 3.1. Sampling and analysis of smelter slags Samples of smelter slag for this study were collected from 4 mine sites (Appendix). Four pot slag samples were collected from the Elizabeth mine, from 3 dierent smelting operations. For two of the samples (01JH28 and 01JH37), the rind (rst few centimeters from the chilled margin) and core were analyzed separately. Also, a third texturally distinctive split of sample 01JH28 was examined. One sample was collected from the Ely mine, one from the Ducktown mining district and one from the Clayton smelter site. Transmitted- and reected-light petrographic studies were conducted on polished thin sections. Mineralogical identication was aided by powder X-ray diraction (XRD), scanning electron microscopy (SEM) and electron probe microanalysis (EPMA) in USGS laboratories in Reston, VA. XRD analyses were performed using a Scintag X1 automated powder diractometer equipped with a Peltier detector with CuKa radiation; XRD patterns were analyzed using Material Data Inc.s JADE software and standard reference patterns (International Centre for Diraction Data, 1997). Polished thin sections and secondary minerals on weathered surfaces were examined using a Jeol JSM-840 scanning electron microscope equipped with a back-scattered electron (BSE) detector, a secondary electron (SE) detector and a PGT X-ray energy-dispersive system (EDS). SEM spectra were collected using an accelerating voltage of 1520 kV and a specimen current of 110 nA. Electron microprobe analyses were performed on a Jeol JXA 8900R microprobe equipped with 5
1042
wavelength-dispersive spectrometers. The microprobe was operated at an accelerating voltage of 1520 kV, a beam current of 2030 nA and a beam diameter of up to 5 mm. The analyses were corrected for electron beam/ matrix and instrumental drift and dead time using either a ZAF or a Phi-Rho-Z-CITZAF scheme as supplied with the Jeol instrument software. Standard reference materials used to calibrate the instrument for quantitative analysis included natural and synthetic silicates, oxides, sulde minerals and glasses. In general, precision for microprobe analyses is 1% relative concentration for major and minor elements and equal to counting statistics for trace elements (510%). Instrument precision for 10 analyses of a chalcopyrite standard was 0.4% or less. The bulk chemical composition of the slag was determined using inductively coupled plasmaatomic emission spectrometry (ICPAES) and inductively coupled plasma-mass spectrometry (ICPMS) following aciddigestion by a mixture of HClHNO3HClO4HF in USGS laboratories in Denver, CO (Crock et al., 1999). Total S was measured using a LECO S analyzer and the wt.% SiO2 was measured using wavelength dispersive X-ray uorescence spectroscopy (XRF) following fusion by a mixture of LiBO2 and LiB4O7. LECO S and XRF analyses were performed by XRAL Laboratories, Ontario, Canada. 3.2. Leaching tests A modied version of a Field Leach Test developed by Hageman and Briggs (2000) was used on splits of all
Table 1 Major, minor and trace phases identied in samples of slag Group Silicates Fayalitefosterite Fayalitekirschsteinite Hedenbergite Quartz Oxides Spinel series Hematite Glass suldes Galena Pyrrhotite Sphalerite/wurtzite CuFe suldes CuFeZn suldes CuPb suldes Metals Intermetallic compounds Cu, Pb, Sb, As, Ag, Fe Pure metals Pb, Cu, Ni, Sb PbS Fe1XS (Zn, Fe)S Fe2SiO4Mg2SiO4 Fe2SiO4CaFeSiO4 CaFeSi2O6 SiO2 Name Ideal/range composition
samples in USGS laboratories in Reston, VA. The leach protocol involves crushing a sample and sieving it to a diameter less than 2 mm. Fifty grams of sample are combined with 1 l deionized water (DI) or 1 l of a solution that approximates eastern United States precipitation (ESP). A solution approximating synthetic eastern precipitation was made by combining a mixture of H2SO4 and HNO3 with deionized water to adjust the pH to 4.2 0.1 (USEPA, 1994). The sample and solution mixtures are vigorously shaken for 1 min and allowed to settle for 24 h. Filtered (0.45-mm nitrocellulose lter) splits were analyzed for cations by ICPMS and ICPAES and for anions (SO2 and Cl) by ion 4 chromatography (IC) in USGS laboratories in Denver, CO and Ocala, FL, respectively.
4. Results 4.1. Petrographic characteristics of the slag deposits Slags are composed of various silicate, oxide, glass, sulde and metallic phases as well as several secondary phases formed by weathering (Table 1 and Fig. 1). Silicates include olivine-group minerals, pyroxene and rare exotic quartz. Oxide phases include spinels and hematite. Suldes, present as either mono or polymineralic inclusions, occur with metallic phases. Slag textures record varying rates of cooling. Those slags quenched in water are devoid of crystalline phases other than sulde minerals (and local hematite) (Fig. 1e), whereas air-cooled slags, including pot slags, contain
Elizabeth mine Ely mine Ducktown Clayton smelter *** *** *** *** *a *** *** * * *** *** * * * * * * * * * * *** *** *** *
(Zn, Mg, Fe2+)(Fe3+, Al, Cr)2O4 Fe2O3
*** major (595%), **minor (15%), *trace (< 1%) indicates modal abundance of phase based on visual estimates. a Exotic phase, interpreted as refractory brick from furnace assimilated into slag.
1043
Fig. 1. Backscattered scanning electron photomicrographs of polished thin-sections (ad, fh) and granulated slag mounted in epoxy resin (e): (a) Randomly oriented elongate skeletal laths of olivine (dark gray), euhedral spinel (gray) and ne- to coarse-sized sulde blebs (white) set in a glass matrix (Elizabeth mine); (b) Parallel growths of elongate olivine skeletons (dark gray), euhedral spinels (gray) and ne-grained disseminated suldes (white) set in a glass matrix (Ely mine); (c) quartz inclusion (dark gray) within a chilled margin of glass (gray) and suldes (white), also note ning of elongate olivine crystals (light gray) (Elizabeth mine); (d) coarse-grained crystals of hedenbergite (dark gray), elongate olivine (gray), ne-grained euhedral spinels (gray) and polymineralic suldes (white) set in a glass matrix containing ne-grained inclusions of suldes and spinels (Clayton smelter slag); (e) sulde blebs (light gray) set in glass (dark gray) mounted in epoxy resin (black) (Ducktown); (f) sulde bleb displaying exsolution lamellae and myrmekitic texture within a matrix of ne-grained olivine laths and glass (Elizabeth mine); (g) large polymineralic particles displaying myrmekite texture enclosed by silicate phases (Clayton smelter site); (h) close up of polymineralic particle in g (Clayton smelter site).
1044
crystalline silicate and oxide phases (Fig. 1ad, fh). Air-cooled slags also display chilled, glassy outer margins and an increase in crystallinity away from these margins towards the center of the mass. In general, the texture, grain-size and mineralogy, as well as color and chemical composition of the slag deposits are dierent from sample site to sample site. Additional micrographs are presented in Piatak et al. (2003). 4.1.1. Vermont copper belt Slags from the Elizabeth and Ely mines display some similarities due to the analogous composition of the ore and smelting technology. In general, the Vermont slags are heterogeneous, ranging from colors of gray to black, glassy to coarsely crystalline, with a luster from metallic, commonly iridescent, to matte (Appendix). Flow textures are evident in slag deposited in the molten state. Individual slag-pot casts, which are approximately 0.75 m in diameter, display an increase in grain size from glassy rinds to megascopically crystalline interiors. Vesicles, ranging from sub-micrometer up to a centimeter in diameter, are common in some samples. Phases present in the Vermont slags include: olivine, glass, spinel, sulde, native metals and quartz, as well as secondary phases (Table 1). Olivine-group minerals are abundant in all slag samples and commonly display spinifex texture within a glassy or nely-crystalline matrix. Olivines predominantly occur as ne-grained to several centimeter long skeletal laths, either growing radially or parallel indicating a rapidly cooled environment (Fig. 1ac). The glass phase comprises both interstitial and bulk glass. Interstitial material is illustrated in Fig. 1ac, whereas bulk glass refers to either surface glass in chilled margins of slag casts or to areas that consist of only glass, such as those exhibiting ow textures. Spinel is present as euhedral to subhedral crystals throughout the samples, ranging in grain size from 1 mm to approximately a 100 mm in diameter (Fig. 1a and b). Suldes of predominantly Fe and Cu occur disseminated throughout the samples, typically within the glass phase, as rounded blebs ranging in size from <1 mm to 500 mm in diameter (Fig. 1a, b, c, f). Also, one slag sample from the Elizabeth mine contains quartz inclusions up to 1 mm in diameter (Fig. 1c). Small pieces of the silica-brick used to line the smelting furnace may have been incorporated into this sample. 4.1.2. Ducktown mining district The sample from the Ducktown mining district is a granulated slag consisting of conchoidally fractured material <1 cm in diameter. The sample is black, has a glassy texture and is characterized by relatively homogeneous glass and FeCu suldes or metallic phases with rare hematite. No crystalline silicate phases occur in this sample. Hematite is euhedral and 1020 mm in diameter. The suldes occur as rounded blebs up to 300
mm in diameter, some displaying myrmekitic textures on 1 mm or less scale. Sulde minerals locally contain mm-size inclusions and veinlets of native metals. 4.1.3. Clayton smelter The Clayton smelter samples are gray, glassy to coarsely crystalline and commonly display a metallic luster. These samples, like the Vermont copper belt samples, display ow textures, have glassy chilled margins containing vesicles and coarse-grained crystalline interiors. The silicates present in this sample include olivine group minerals and pyroxene. Olivine is present as radial laths up to several cm long, whereas the pyroxene occurs as subhedral crystals scattered throughout the sample (Fig. 1d). The sample examined for petrographic analysis contained only interstitial glass. The euhedral to subhedral spinel is up to 50 mm in diameter. The suldes (which are mainly of Pb, or CuFe) occur as discrete grains disseminated throughout the silicate matrix, or as admixtures with native metals and alloys of various metals in large (up to several 100 mm in diameter) blebs (Fig. 1g and h). Some sulde grains display myrmekitic textures and mm-scale exsolution. 4.2. Chemical composition of minerals and glasses in slag 4.2.1. Olivine The composition of olivine-group minerals in the Elizabeth and Ely mine slag samples lie along the fosterite-fayalite join, with the olivine from the Elizabeth mine being more Mg-rich than that from the Ely mine (Table 2 and Fig. 2). In contrast, the olivine in slag from the Clayton smelter site contains signicant Ca (up to 19 wt.% CaO) and thus falls between fayalite and kirschsteinite in composition (Fig. 2). The substitution of Ca for Fe in natural fayalitic olivine is limited, but olivines with compositions similar to those found in the Clayton smelter slag sample have been reported in slag from the Pr bram mining district in the Czech Republic (Ettler et al., 2001a) and from Roman slags (Manasse et al., 2001) (Fig. 2). Also, quenched experimental studies have demonstrated complete solid solution between fayalite and kirschsteinite at temperatures above 1040 C (Mukhopadhyay and Lindsley, 1983). Other cations found in signicant abundances in the olivines include Zn and Mn. Olivines from Vermont contain up to 1.17 wt.% ZnO and those from the Clayton smelter reach 1.25 wt.%. In contrast, fayalitic olivines in slags from the Penn mine (California) average 6.83 wt.% ZnO and willemite (Zn2SiO4) also is an important phase (Parsons et al., 2001). Ettler et al. (2001a) reported up to 6.28 wt.% ZnO in olivines from slags from Pr bram, Czech Republic. Olivine in the sample from the Clayton smelter site also contains minor Mn (up to 3.31 wt.% MnO); and some laths are
1045
Table 2 Average electron microprobe analyses of olivine, pyroxene, glass, spinel and hematite in slag samplesa (Concentrations reported in wt.%)
Al2O3 MnO MgO Cr2O3 K2O ZnO SiO2 Olivine Elizabeth FeOb Fe2O3b CaO TiO2 Na2O CuO PbO SO3c P2O5 CoO NiO nd
maximum minimum average 1 maximum minimum average 1 maximum minimum average 1
2.12 0.05 0.25 0.43 3.14 0.03 0.44 0.80 0.05 0.01 0.03 0.02
0.31 0.07 0.15 0.07 0.36 0.11 0.24 0.08 3.31 2.64 2.99 0.25
12.63 2.33 7.22 2.54 3.98 1.34 2.57 0.89 2.72 1.08 2.09 0.61
0.05 0.00 0.02 0.01 0.04 0.00 0.01 0.01 0.02 0.00 0.01 0.01
0.64 0.00 0.03 0.12 0.13 0.00 0.01 0.04 0.00 0.00 0.00 0.00
1.17 0.25 0.76 0.25 0.82 0.29 0.58 0.15 1.25 0.84 1.05 0.14
34.63 66.15 29.47 48.95 31.54 60.73 0.84 3.15 31.92 67.50 27.27 64.27 30.58 65.80 1.09 1.03 31.78 53.32 31.01 43.60 31.45 47.90 0.25 3.25
0.52 0.12 0.25 0.11 0.76 0.13 0.26 0.17 19.14 8.22 14.54 3.67
1.05 0.00 0.07 0.17 0.20 0.00 0.06 0.05 0.03 0.00 0.01 0.01
0.04 0.00 0.02 37 0.02 0.00 0.00 0.00 14
Ely
Clayton
Pyroxene Clayton
maximum minimum average 1
7.57 5.88 6.64 0.61
0.95 0.71 0.83 0.09
0.83 0.07 0.43 0.29 4.53 1.25 2.68 0.83 1.56 1.35 1.44 0.11 2.16 0.90 1.54 0.36
0.04 0.00 0.01 0.02
0.01 0.00 0.00 0.00 2.15 0.26 1.27 0.52 1.27 1.18 1.24 0.05 0.60 0.27 0.43 0.10
0.90 0.60 0.70 0.12 1.00 0.35 0.74 0.24 0.62 0.49 0.55 0.07 2.17 1.53 1.85 0.18
42.62 27.44 41.28 26.15 41.81 26.93 0.51 0.57 41.92 63.57 28.04 34.22 36.60 45.20 3.93 8.96 42.75 44.29 41.95 42.14 42.38 43.30 0.40 1.09 38.36 57.93 31.75 45.53 34.83 52.05 1.80 3.42
22.11 21.61 21.80 0.22 2.91 0.81 2.27 0.74 2.11 1.87 1.99 0.12 5.93 2.30 4.05 0.98
0.32 0.23 0.27 0.04 0.51 0.17 0.40 0.11 0.43 0.37 0.40 0.03 0.19 0.06 0.13 0.04 4.29 0.24 1.73 0.98 0.95 0.86 0.89 0.05 0.64 0.26 0.44 0.11 0.58 0.03 0.29 0.22 0.26 0.12 0.19 0.07 1.43 0.21 0.53 0.28 0.10 0.00 0.03 0.04 0.05 0.01 0.03 0.02 0.14 0.00 0.03 0.04 1.43 0.20 0.81 0.33 0.85 0.70 0.76 0.08 5.60 2.53 3.98 0.87 0.00 0.00 0.00 0.00 0.19 0.06 0.11 0.05 0.13 0.10 0.11 0.02
Bulk glass Elizabeth
maximum 12.02 minimum 2.38 average 8.47 1 2.88 maximum minimum average 1 maximum minimum average 1 7.94 7.29 7.55 0.34 4.11 2.00 3.07 0.62
13
Ely
Ducktown
19
Interstitial glass Elizabeth maximum 21.22 minimum 10.41 average 18.73 1 2.17 Ely maximum 20.24 minimum 12.60 average 16.89 1 2.67 maximum 12.94 minimum 8.74 average 10.73 1 1.39
4.22 0.00 0.30 0.88 0.58 0.00 0.14 0.21 0.09 0.00 0.03 0.03
4.91 1.11 3.17 0.88 5.24 2.32 3.83 1.13 3.87 1.47 3.06 0.86
1.14 0.18 0.58 0.19 0.75 0.40 0.51 0.11 4.16 2.12 3.19 0.79
55.23 33.77 42.77 8.58 50.66 14.74 3.33 5.30 56.16 26.08 52.28 12.27 53.47 16.29 1.16 5.31 45.46 22.43 40.82 14.31 42.99 17.20 1.52 2.96
8.78 2.77 5.69 1.44 6.39 2.57 4.51 1.34 17.13 13.82 15.66 1.07
2.88 0.16 0.45 0.39 0.75 0.13 0.29 0.21 0.49 0.18 0.34 0.10
6.06 1.94 4.19 0.77 3.37 1.53 2.62 0.69 1.44 0.82 1.18 0.22
1.24 0.00 0.18 0.22 0.24 0.03 0.12 0.07 0.03 0.01 0.02 0.01
0.08 0.00 0.01 0.02 0.06 0.00 0.02 0.02 5.85 3.00 4.26 1.09
4.11 0.06 0.51 0.66 0.67 0.06 0.26 0.24 0.71 0.39 0.54 0.12
0.12 0.00 0.04 0.06
0.08 0.00 0.03 0.02 0.06 0.01 0.03 0.02
49
Clayton
1.37 0.00 0.46 0.48
Spinel Elizabeth
maximum 23.89 minimum 8.65 average 13.44 1 5.25 maximum minimum average 1 9.87 9.45 9.66 0.23
0.06 0.00 0.03 0.02 0.05 0.03 0.04 0.01
1.06 0.24 0.71 0.25 0.40 0.33 0.38 0.04
1.39 0.05 0.39 0.36 1.09 0.00 0.28 0.54
0.40 0.00 0.03 0.09 0.00 0.00 0.00 0.00
2.07 0.73 1.08 0.48 0.56 0.52 0.54 0.02
6.95 33.61 57.45 0.42 30.54 35.05 1.17 31.62 49.78 1.44 0.86 7.17 0.65 31.69 54.88 0.53 31.49 54.18 0.61 31.59 54.51 0.05 0.08 0.29
0.34 0.00 0.05 0.09 0.03 0.00 0.01 0.01
4.25 1.67 2.69 0.79 2.21 1.79 2.08 0.19
21
Ely
(continued on next page)
1046
Table 2 (continued)
Al2O3 MnO MgO Cr2O3 K2O ZnO SiO2 Clayton maximum minimum average 1 3.27 3.03 3.15 0.17 0.64 0.60 0.62 0.03 0.18 0.12 0.15 0.04 0.06 0.05 0.05 0.01 0.00 0.00 0.00 0.00 1.41 1.21 1.31 0.14
FeOb Fe2O3b CaO 0.39 0.35 0.37 0.03
TiO2 Na2O CuO PbO SO3c P2O5 CoO NiO nd 1.13 0.92 1.02 0.15
0.99 29.50 63.33 0.98 28.60 62.23 0.99 29.05 62.78 0.01 0.63 0.78
Hematite Ducktown
maximum minimum average 1
0.07 0.06 0.07 0.01
0.89 0.85 0.87 0.02
0.70 0.68 0.69 0.01
99.12 98.09 98.49 0.52
0.05 0.00 0.02 0.02
0.02 0.00 0.01 0.01
0.02 0.02 0.02 0.00
a b c d
Complete tabulation of microprobe data is provided in Piatak et al. (2003). Total Fe calculated as FeO except for spinel and hematite, for which Fe2+/Fe3+ ratio was estimated using site occupancy constraints. Total S reported as SO3. Number of analyses.
zoned, with the interior being slightly enriched in Ca and Mg. A complete tabulation of microprobe analyses is presented in Piatak et al. (2003). 4.2.2. Pyroxene Pyroxene is found only in slag from the Clayton smelter site and is a major phase, constituting up to 40 vol.% (Fig. 1d). Pyroxene is hedenbergitic in composition and aluminous (up to 7.57 wt.% Al2O3) (Table 2). Also, the pyroxene contains up to 0.95 wt.% MnO, 0.90 wt.% ZnO, 0.83 wt.% MgO and 0.32 wt.% TiO2. Ettler et al. (2001a) also found aluminous hedenbergitic clinopyroxenes in slags from Pr bram, Czech Republic. They reported MnO concentrations up to 1.80 wt.% and ZnO concentrations to 3.12 wt.%. Saez et al. (2003) reported clinopyroxenes with compositions ranging from hedenbergite to diopside with up to 7.60 wt.% Al2O3 and up to 11.28 wt.% MnO. Individual microprobe analyses for pyroxene are given in Piatak et al. (2003).
4.2.3. Glass The glass, like other silicate phases, is Fe-, Al- and Ca-rich (Table 2). Bulk glass compositions can be described in terms of the components FeO, Al2O3, SiO2 and CaO and are depicted in FeOAl2O3SiO2 and CaOFeOSiO2 ternary diagrams (Fig. 3a and b). As shown, the composition of the interstitial glass is distinctly dierent than the composition of the bulk glass. For example, in Vermont samples, the bulk glass contains Fe5Si >Al, whereas the interstitial glass contains Si > Al> Fe. Also, interstitial glass is enriched in Na, K and Ca and depleted in Mg, S and Zn relative to bulk glass for samples from Vermont. The composition of glass, all of which is interstitial (Fig. 1d), in the Clayton smelter slag is similar to the composition of interstitial glass in the Vermont copper belt samples, although it contains higher concentrations of Ca and lower concentrations of Al (Fig. 3). The composition of the glass in the granulated slag from Ducktown is similar in composition to the bulk glass from the Vermont copper
Fig. 2. Olivine compositions in the larniteforsterite(fayalite+tephroite) ternary diagram. Olivine compositions for PbZn slag from Pr bram, Czech Republic, are from Ettler et al. (2001a,b). Olivine from Cu slags from Campiglia Marittima, Italy, are from Manasse et al. (2001).
1047
Fig. 3. Glass compositions and bulk chemistry compositions on FeOAl2O3SiO2 (a) and CaOFeOSiO2 (b) ternary diagrams. Bulk glass analyses are shown with small lled symbols, interstitial glass analyses with small open symbols and whole rock analyses with large open symbols. Glass analyses for PbZn slag from Pr bram from Ettler et al. (2001a,b). Phase boundaries and isotherms ( C) from Levin et al. (1964).
1048
belt. The similarity probably reects similar ore and smelting techniques. The higher concentrations of Ca and lower concentrations of Al and K in the glass in the Ducktown sample relative to bulk glass in the Vermont copper belt samples may indicate variations in original ore chemistry or possibly dierences in uxes added during the smelting process (Fig. 3). The bulk glass analyses for the Vermont and Ducktown samples typically lie in the elds for the initial crystallization of either tridymite/cristobalite or olivine on both the CaOFeOSiO2 and FeOAl2O3SiO2 ternary diagrams (Fig. 3). Thus, if these analyses represent the quenched molten material, the phases likely to form are tridymite, olivine and, for a few Elizabeth and Ducktown samples, hercynite. Primary quartz has not been identied in this study nor in other slag studies, such as those by Ettler et al. (2001a,b), Parsons et al. (2001) and Lottermoser (2002). Experimental studies of the crystallization of silicate melts document kinetic barriers to the nucleation of framework silicates (De Jong et al., 1984). Therefore, the absence of silica minerals, such as tridymite, may reect the inability to nucleate these phases during quenching. Also, the four components, CaO, Al2O3, FeO and SiO2, represented by the ternary diagrams in Fig. 3 are, on average, 94 and 90% of the total weight percent for bulk and interstitial glass analyses for this study, respectively. Thus, the lack of quartz observed in the slag may indicate that the chemistry of the glass is not completely represented by these 4 components. The compositions of the interstitial
glass lie in the two-liquids, tridymite/cristobalite, or mullite elds. The position of the bulk glass analyses, which may represent a quenched layer, suggests that the molten temperature of the slag was above 1100 C for the Vermont and Tennessee slags (Fig. 3). This temperature is consistent with temperatures in excess of 1000 C for Cu smelting furnaces reported by Peters (1911). The interstitial glass is enriched in Al, K and Na relative to associated olivine, pyroxene, or spinel. Zinc concentrations in glass reach 1.14 wt.% ZnO from Vermont, 4.16 wt.% ZnO from Clayton and 2.17 wt.% ZnO from Ducktown. Glasses from Vermont and Ducktown contain traces of Cu and variable S, with the Ducktown sample containing the highest concentration of S (maximum of 5.60 wt.% SO3). Analytical totals from microprobe analyses of the glass in the Ducktown sample are slightly high (up to 104.8 wt.% and average 102.9 wt.%), which may be due to the presence of very ne-grained suldes or dissolved S in the glass due to rapid quenching of the molten slag. The S in the microprobe analyses was calculated as an oxide, instead of S as sulde, which would reduce the totals. The presence of metal-sulde grains indicates S may be in the reduced state and supports this alternative. The interstitial glass in the Clayton smelter sample contains signicant Pb (up to 5.85 wt.% PbO). Parsons et al. (2001) found that interstitial glass from the Penn mine averaged 5.00 wt.% ZnO and 33.55 wt.% PbO, but bulk glass averaged 5.58 wt.% ZnO and only 0.06 wt.% PbO. Ettler et al. (2001a) reported matrix glass from Pr bram,
Fig. 4. Composition of spinel within the hercynitemagnetitegahnitefranklinite quadrilateral. Analyses for Cu slag from the Penn mine, CA, are from Parsons et al. (2001) for the average microprobe analyses of zincian spinel and magnetite. Representative compositions of spinel in PbZn slag from Pr bram, Czech Republic, are from Ettler et al. (2001a,b). Appropriate valence distributions are calculated from electron microprobe analyses assuming ideal stoichiometry.
1049
Czech Republic that averaged 4.74 wt.% ZnO and 2.35 wt.% PbO; also, surface glass averaged 5.20 wt.% ZnO and 0.57 wt.% PbO. A complete tabulation of microprobe analyses for this study is presented in Piatak et al. (2003). 4.2.4. Spinel Spinel compositions plot between magnetite and hercynite (Table 2 and Fig. 4). Spinel in Vermont slag samples contains up to 1.39 wt.% Cr2O3; spinel from the Vermont and Clayton smelter samples contain signicant Zn, up to 2.07 and 1.41 wt.% ZnO, respectively and Ti to 4.25 and 1.13 wt.% TiO2, respectively. Some spinels in slag from the Vermont Cu belt are zoned. In one sample, the core of the spinel contains slightly higher concentrations of Cr, whereas the rim is enriched in Al and Fe; in another sample, the core of the spinel contains higher concentrations of Fe and Ti and the rim contains higher concentrations of Al. Kucha et al. (1996) reported zoned spinels in slag from a ZnPb mine in Belgium with the outermost zone rich in Fe and Ti, an intermediate zone rich in Al and then an inner core strongly enriched in Cr and V. Parsons et al. (2001) analyzed magnetites from the Penn mine, which averaged 8.60 wt.% ZnO and zincian spinels (gahnite franklinite) with an average of 32.45 wt.% ZnO (Fig. 4). Ettler et al. (2001a) reported ZnO concentrations up to 19.91 wt.% in spinels within the hercynitemagnetite gahnitefranklinite quadrilateral from Pr bram, Czech Republic (Fig. 4). Individual microprobe analyses for spinel are presented in Piatak et al. (2003).
4.2.5. Suldes and metallic phases and unknown compounds Suldes and metallic phases are volumetrically minor components of the slag samples, but they are environmentally signicant because of their metal content and their potentially more reactive state. The compositions of most sulde minerals in slag from the Vermont copper belt correspond to the digenitebornite, intermediate, or pyrrhotite solid solutions (Fig. 5). Electron microprobe analyses for representative suldes within these solid solutions are listed in Table 3. Microprobe analyses for all suldes examined are presented in Piatak et al. (2003). Small scale exsolution may account for the compositions of sulde between these solid solution elds (Fig. 1f). The suldes in the Vermont samples also contain minor Zn (up to 2.07 wt.%), Co (to 1.31 wt.%), Ni (to 0.13 wt.%), Se (to 0.07 wt.%) and Ag (to 0.05 wt.%). Metallic Cu and Ni phases occur as small inclusions within the FeCu suldes. Also, semi-quantitative analyses indicate a possible sulde phase containing Fe, Cu and Zn. The ratios of these elements vary widely, with the weight percent of Zn in one phase examined making up more than one-third of the total. The composition of sulde blebs in granulated slag from Ducktown is similar to those from the Vermont Cu belt, although the concentration of Cu is lower generally (Fig. 5). The blebs (Fig. 1e), commonly contain minor Zn, traces of Ni, Se and Ag and veinlets of metallic Cu (Fig. 6). Analytical totals from electron microprobe analyses of sulde blebs in the granulated
Fig. 5. Atomic proportions of Cu, Fe and S in suldes from slag. Phase relations in the central portion of the CuFeS system at 600 C after Cabri (1973). Squares represent stoichiometeric compositions for the minerals chalcocite (cc), bornite (bn), chalcopyrite (cp), pyrite (py) and troilite (tr). Solid solution elds are represented by pyrrhotite (po), intermediate solid (iss) and bornitedigenite (bndg).
1050
Ducktown slag are low, ranging from 78 to 94 wt.% (representative analyses are given in Table 3). The low totals are partially due to sub-microscopic inclusions of glass (or possibly quartz) disseminated throughout many of the sulde blebs (Fig. 6). Random spot analyses of the SiO2 in the suldes are signicant and range up to a maximum of 6.84 wt.%, supporting this hypothesis, but the SiO2 contents are not suciently high to explain the low totals, even though, if glass, not all glass-derived elements were included in the sulde analyses. The low totals for sulde analyses may relate to the presence of oxides, such as magnetite. Admixtures of suldes of galena, bornite, sphalerite/ wurtzite ((Zn, Fe) S) and metallic phases with varying proportions of Fe, Cu, As, Sb and Pb occur in the slag sample from the Clayton smelter site (Table 4). Metallic Pb and sphalerite/wurtzite were determined qualitatively using a SEM, whereas all other phases were identied quantitatively using the electron microprobe. The phases occur as complex polymineralic sulde-metal blebs that display myrmekitic textures (Fig. 1g and h). Thus,
analyses undoubtedly represent mixtures of phases because of the ne-grained exsolved nature of the material. According to electron microprobe analyses in Table 4, phases such as furutobeite [(Cu,Ag)6 PbS4], cuprostibite [Cu2(Sb,Tl)], domeykite (Cu3As) and dyscrasite (Ag3Sb) may be present. All suldes, metallic phases and other compounds contain variable traceelement concentrations (Table 4). 4.2.6. Secondary minerals The slags have been subjected to weathering and oxidation since dumping, some since the mid nineteenth century. Thus a series of secondary minerals have been observed on the surface of the slag piles. The minerals occur as coatings on surfaces and along fractures and as precipitates at seepage points and in protected overhangs, all of which reect the reactivity of the material. Not all precipitates are crystalline and some solids forming on slags from Vermont are amorphous Fe and Al oxyhydroxides. Crystalline phases on samples from Vermont include chalcanthite, siderotil,
Fig. 6. Wavelength spectrometry electron-microprobe X-ray maps of the slag from Ducktown, showing the distribution of Fe, Cu, Si and S. Note the metallic Cu vein and micrometer-size inclusions rich in Si, likely present as glass. Lighter colors represent higher concentrations.
Table 3 Representative electron-microprobe analyses of FeCu suldes in Vermont and Ducktown slaga,b Ely 00JH05 25.12 0.00 0.02 0.00 30.84 43.75 0.01 0.01 0.02 0.18 0.00 0.00 0.00 99.95 16.80 0.00 0.00 0.00 27.24 55.34 0.01 0.01 0.05 0.08 0.00 0.00 0.00 99.53 9.67 0.00 0.00 0.00 24.18 66.42 0.03 0.03 0.05 0.06 n.a. 0.00 n.a. 100.44 4.97 0.00 0.01 0.00 23.10 73.21 0.01 0.00 0.03 0.06 0.00 n.a. n.a. 101.37 62.58 0.02 0.02 0.00 35.68 1.41 0.31 0.00 0.00 0.00 n.a. 0.07 n.a. 100.10 28.36 0.01 0.01 0.00 32.40 36.67 0.05 0.06 0.03 1.51 0.00 n.a. n.a. 99.09 19.57 0.01 0.00 0.00 28.58 51.82 0.04 0.06 0.00 0.06 0.00 n.a. n.a. 100.14 14.76 0.00 0.00 0.00 24.09 60.84 0.03 0.00 0.00 0.09 n.a. 0.00 n.a. 99.81 52.48 0.18 0.01 n.a. 28.01 10.80 0.06 0.02 0.00 1.38 n.a. 0.00 0.28 93.21 50.82 0.17 0.00 n.a. 27.86 11.89 0.05 0.00 0.00 1.45 n.a. 0.01 0.30 92.56 41.33 0.13 0.00 0.03 22.77 28.38 0.05 0.03 0.00 1.02 n.a. 0.00 0.27 94.09 28.28 0.09 0.00 n.a. 13.74 44.54 0.02 0.00 0.00 0.67 n.a. 0.01 0.18 87.53 00JH05 01JH28 01JH37 01JH34 01JH34 01JH34 01JH34 01CB25 01CB25 01CB25 01CB25 01CB25 8.70 0.03 0.00 0.00 16.33 60.56 0.03 0.04 0.00 0.16 0.01 0.00 0.00 85.84 Ducktown
Elizabeth
01JH27
00JH05
01JH28
wt.% Fe Mn As Pb S Cu Co Se Ag Zn Sb Ni Si total
73.25 0.00 0.00 0.00 24.41 0.74 1.31 0.01 0.00 0.00 0.00 n.a.c n.a. 99.73
61.23 0.02 0.00 0.00 35.34 3.13 0.23 0.02 0.00 0.00 0.00 0.11 0.00 100.09
29.61 0.02 0.00 0.00 32.09 35.39 0.17 0.00 0.00 2.07 n.a. 0.00 n.a. 99.36
62.25 0.00 0.00 0.00 36.13 0.56 1.06 0.01 0.00 0.00 0.00 0.00 100.0 100.0 100.0 0.06 100.0 100.0
24.98 0.02 0.00 0.00 47.14 26.24 0.14 0.00 0.00 1.49
8.77 0.00 0.00 0.00 38.19 52.93 0.03 0.02 0.02 0.05
4.54 0.00 0.00 0.00 36.70 58.70 0.01 0.00 0.01 0.04 0.00
49.53 0.01 0.01 0.00 49.18 0.98 0.24 0.00 0.00 0.00
23.95 0.01 0.00 0.00 47.65 27.21 0.04 0.04 0.01 1.09 0.00
17.01 0.01 0.00 0.00 43.27 39.60 0.03 0.04 0.00 0.04 0.00
13.38 0.00 0.00 0.00 38.04 48.48 0.03 0.00 0.00 0.07 0.00 100.0
46.55 0.16 0.01 43.27 8.42 0.05 0.01 0.00 1.05
45.43 0.16 0.00 43.37 9.34 0.05 0.00 0.00 1.11
38.43 0.12 0.00 0.01 36.87 23.19 0.04 0.02 0.00 0.81
30.61 0.09 0.00 25.90 42.36 0.02 0.00 0.00 0.62
at.% Fe Mn As Pb S Cu Co Se Ag Zn Sb Ni Si total 21.38 0.00 0.01 0.00 45.72 32.73 0.01 0.01 0.01 0.13 0.00 0.00 0.00 100.0 14.87 0.00 0.00 0.00 41.99 43.04 0.01 0.01 0.02 0.06 0.00 0.00 0.00 100.0
0.00
100.0
48.64 0.01 0.00 0.00 48.89 2.19 0.17 0.01 0.00 0.00 0.00 0.08 0.00 100.0
100.0
0.00 0.49 100.0
0.01 0.54 100.0
0.00 0.51 100.0
0.01 0.39 100.0
9.60 0.03 0.00 0.00 31.39 58.76 0.03 0.03 0.00 0.15 0.00 0.00 0.00 100.0
Complete tabulation of microprobe data is presented in Piatak et al. (2003). Sn was sought but not found. Not analyzed.
1051
1052
Table 4 Average electron-microprobe analyses of sulphides, metallic phases, alloys and unknown compounds in slag from the Clayton smelter, Idahoa furutobeite? (Cu,Ag)6PbS4 (n=1) FeAs (n=5) AsFeCu (n=2) AsFe (Pb) (n=2) 33.95 0.00 54.34 5.58 0.20 0.95 0.15 0.06 0.25 0.24 2.43 0.00 1.16 0.00 99.35 0.18 0.00 1.53 0.43 0.04 51.31 0.00 0.05 0.93 0.06 46.72 0.16 0.05 0.00 101.43 0.03 0.01 5.11 4.32 0.02 15.75 0.00 0.11 0.02 0.01 75.00 0.69 0.00 0.00 101.07 0.02 0.01 24.61 0.38 0.04 70.71 0.00 0.07 0.06 0.08 4.06 0.01 0.01 0.00 100.05 0.00 0.01 5.48 4.84 0.00 0.05 0.02 0.12 0.00 0.05 88.22 0.00 0.00 0.00 98.78 0.00 0.01 0.03 0.11 0.01 0.22 0.00 0.03 72.46 0.00 26.10 0.11 0.00 0.00 99.09 15.16 0.17 3.36 0.45 29.46 14.91 0.02 0.00 0.01 32.43 0.10 0.00 0.04 0.01 96.14 10.57 0.02 0.16 2.99 24.54 61.62 0.01 0.02 0.07 0.10 0.06 0.01 0.01 0.00 100.15 1.96 0.00 0.06 27.47 18.59 54.42 0.01 0.00 0.00 0.09 0.00 0.05 0.00 0.00 102.64 42.62 0.00 52.17 0.57 0.22 1.63 0.12 0.07 0.03 0.05 0.75 0.00 0.57 0.00 98.87 38.06 0.00 51.57 0.17 0.14 7.00 0.23 0.04 0.15 0.01 4.80 0.00 0.64 0.00 102.86 cuprostibite? Cu2(Sb,Tl) (n=7) CuAsSb (n=1) domeykite? Cu3As (n=3) Sb (AsPb) (n=1) dyscrasite? Ag3Sb (n=1) ZnCuFeS (n=2) bornites CuFeS (n=8)
metallic Sb (nb=1)
metallic Cu (n=1)
galena (PbS) (n=8)
wt.% Fe Mn As Pb S Cu Co Se Ag Zn Sb Sn Ni Si Total
0.00 0.00 3.73 18.58 0.67 0.56 0.01 0.10 0.00 0.00 74.10 0.78 0.00 0.00 98.52
0.49 0.00 0.01 0.06 0.76 100.53 0.00 0.00 0.06 0.08 0.03 n.a.c n.a. n.a. 102.02
0.38 0.01 0.01 81.43 13.05 4.41 0.00 0.01 0.03 0.07 0.12 0.05 0.01 0.00 99.55
at.% Fe Mn As Pb S Cu Co Se Ag Zn Sb Sn Ni Si Total 2.18 0.00 0.05 8.25 36.09 53.31 0.01 0.00 0.00 0.08 0.00 0.03 0.00 0.00 100.0 50.39 0.00 45.98 0.18 0.45 1.69 0.13 0.06 0.02 0.05 0.41 0.00 0.64 0.00 100.0 44.22 0.00 44.66 0.05 0.28 7.14 0.25 0.03 0.09 0.01 2.56 0.00 0.71 0.00 100.0 42.55 0.00 50.67 1.89 0.43 1.04 0.18 0.05 0.16 0.26 1.39 0.00 1.37 0.00 100.0 0.26 0.00 1.65 0.17 0.10 65.64 0.01 0.05 0.70 0.07 31.20 0.11 0.07 0.00 100.0 0.06 0.01 7.09 2.17 0.06 25.77 0.00 0.14 0.02 0.02 64.06 0.61 0.00 0.00 100.0
0.00 0.00 6.34 11.41 2.64 1.13 0.02 0.15 0.00 0.00 77.47 0.83 0.00 0.00 100.0
0.54 0.00 0.00 0.02 1.46 97.85 0.00 0.00 0.03 0.08 0.02
100.0
0.76 0.03 0.02 44.85 46.33 7.71 0.00 0.01 0.03 0.11 0.11 0.05 0.01 0.00 100.0
0.02 0.01 22.18 0.13 0.09 75.13 0.00 0.06 0.04 0.08 2.25 0.01 0.01 0.00 100.0
0.00 0.02 8.87 2.83 0.00 0.09 0.03 0.18 0.00 0.09 87.87 0.00 0.00 0.00 100.0
0.00 0.03 0.04 0.06 0.03 0.38 0.00 0.04 75.28 0.00 24.03 0.11 0.00 0.00 100.0
13.78 0.16 2.32 0.11 46.53 11.81 0.02 0.00 0.00 25.17 0.04 0.00 0.04 0.01 100.0
9.69 0.02 0.11 0.80 39.33 49.90 0.01 0.01 0.03 0.08 0.02 0.00 0.01 0.00 100.0
Complete tabulation of microprobe data is presented in Piatak et al. (2003). Number of analyses. Not analysed.
1053
jarosite, brochantite and gypsum, thus demonstrating the mobility of Fe, Cu, S, Ca and alkali elements. 4.3. Bulk slag chemistry Bulk chemical analyses of slag from the 3 mining districts show signicant chemical heterogeneities between areas (Table 5). Also, 4 samples from the Elizabeth mine (two of which were split into several subsamples) show little variation in major element concentrations (except for Ca, which ranges from 0.82 to 4.1 wt.%), but high variability in trace-element concentrations (Table 5). The bulk chemistry of the slag is plotted on the CaO FeOSiO2 and FeOAl2O3SiO2 ternary diagrams in Fig. 3. Iron and silica are major constituents in all samples (up to 37 wt.% Fe and up to 20.28 wt.% Si). All of the slag material studied from Vermont contains higher concentrations of Al (over 3.2 wt.%), K (over 0.85 wt.%) and Na (over 0.62 wt.%), compared with the Ducktown granulated slag and the Clayton smelter slag. In contrast, the slag from the Clayton smelter contains higher concentrations of Ca (3.1 wt.%) and Mn (2.2 wt.%) than the average concentration of Ca and Mn in slag from Vermont and Ducktown. Presumably, this reects the carbonate hosts for many of the ore deposits in the Bayhorse mining district.
Slags also display variations in trace-element concentrations (Table 5). The sample from Ducktown is low in Cu (2000 mg/kg) compared with Elizabeth (maximum of 13,500 and averages 7390 mg/kg), Ely (6390 mg/kg) and Clayton (7550 mg/kg). The higher concentrations of Cu in slags from the Vermont copper belt relative to Ducktown probably reect improvements in Cu smelting technology between the nineteenth and twentieth centuries. The sample from the Clayton smelter site contains the highest concentrations of Pb (63,000 mg/kg), Zn (19,700 mg/kg), Sb (1710 mg/kg), As (555 mg/kg) and Sn (363 mg/kg), which is consistent with the enrichment of these elements in the ores feeding the Clayton smelter, relative to the Cu ores in Tennessee and Vermont. 4.4. Leachate 4.4.1. Chemical composition Leaching tests conducted on crushed slag (< 2 mm in diameter) with deionized water (DI) or eastern synthetic precipitation (ESP) indicate that signicant concentrations of metals can be released from the slag. In general, leachate produced by eastern synthetic precipitation contained equal or higher concentrations of trace elements compared with leachate produced by deionized water, as shown for Zn, Ni, Cu and Co in Fig. 7. This
Fig. 7. Concentration of Zn, Ni, Cu and Co in leachate from leach tests using deionized water (DI) and eastern synthetic precipitation (ESP).
1054
trend undoubtedly reects the lower pH of the eastern synthetic precipitation (pH=4.1) relative to deionized water (pH=5.0). For example, the concentration of Cu in leachate using ESP is more than double that in leachate using DI for the Ducktown sample and for several samples from the Elizabeth mine (Table 6). For most major elements, the concentrations in the DI leachate and ESP leachate are similar, except for Al, which for most samples is leached more readily by ESP than by DI. The specic conductivity for all samples is higher in leachate produced by ESP than in leachate produced by
Table 5 Average chemical composition of slag samplesa Elizabeth mine
DI, with a single exception. Also, the leachates of an Elizabeth mine slag have the highest specic conductivity (88.5 mS/cm for DI and 73 mS/cm for ESP). For all samples, the pH of the leachate is higher than the original pH of the leaching solution, with the exception of two samples from Vermont leached with DI. The greatest increase in pH was for leachate produced from the Clayton smelter slag, which rose to pH 5.7 for both DI and ESP. The concentrations of major, minor and trace elements in leachates vary from sample to sample (Table 6).
Ely mine Ducktown Clayton smelter 01CB25 (n=1) 1 1.8 21 0.23 0.62 0.31 0.25 0.04 2.01 16.56 0.063 53JH00 (n=1) 1.2 3.1 32 0.39 0.63 2.2 0.1 0.2 0.86 0.061
Sample 00JH05 01JH27 01JH28 01JH28rind 01JH28core 01JH37rind 01JH37core Average 01JH34 Major elements (wt.%)c Al 3.6 4.3 Ca 0.83 4.1 Fe 37 26 K 0.85 1.1 Mg 1.1 1.1 Mn 0.054 0.052 Na 1 1 P 0.08 0.04 S 1.41 0.85 Si 14.50 19.30 Ti 0.31 0.363 4.9 1.9 24 1.2 2 0.15 1.9 0.1 0.82 20.28 0.426 4 1.8 32 0.94 1.2 0.094 1.1 0.1 0.41 16.63 0.312 4.1 1.8 33 0.93 1.2 0.097 1.1 0.1 0.46 16.41 0.338 4.1 0.92 33 1 1.4 0.062 1.2 0.09 0.85 16.91 0.302 4 0.89 33 0.97 1.4 0.06 1.2 0.1 1.12 16.20 0.291 (nb=7) 4.1 1.7 31 1.0 1.3 0.081 1.2 0.09 0.85 17.18 0.335 (n=1) 3.2 1.1 36 0.89 0.7 0.13 0.62 0.1 1 16.23 0.287
Minor or trace elements in ppm (mg/kg)d Ag 1.2 0.21 1.2 0.61 As 0.8 <0.5 1 1.5 Au <0.05 <0.05 <0.05 <0.05 Ba 41 112 99 76 Be 0.5 1.1 0.9 0.5 Bi <0.05 <0.05 0.09 0.2 Cd 2 0.43 1.7 2.2 Ce 5.2 13 12 9.2 Co 600 247 260 472 Cr 114 276 164 102 Cs 3.6 3.2 6.5 3.7 Cu 8420 1900 7040 5900 Mo 51 25 24 50 Ni 19 8.5 14 24 Pb 14 8.1 26 33 Sb 0.1 <0.1 <0.1 <0.1 Se 6.1 2.9 2.4 2 Sn 0.8 <0.5 <0.5 0.9 Sr 34 113 64 56 W 1.2 1.5 2.2 2.6 Y 11 14 18 14 Zn 7650 2310 6160 7390
a b c d
0.84 1.5 <0.05 77 0.7 0.2 2.4 9.6 491 150 3.8 7320 52 26 35 <0.1 2.5 1 58 2.6 15 7640
1 1.5 <0.05 41 0.6 0.08 3.1 5.9 448 127 4.5 7640 33 18 24 <0.1 2.9 0.7 36 2 12 10,200
2.4 2 <0.05 39 0.6 0.2 3.4 5.6 445 122 4.4 13,500 32 27 37 <0.1 5.1 0.7 35 1.9 12 9870
1.1 0.98 1.2 < 0.5 <0.05 < 0.05 69 162 0.7 1.6 0.1 0.2 2.2 1.9 8.6 23 423 716 151 56 4.2 2.5 7390 6390 38 47 20 18 25 11 <0.1 0.1 3.4 5.8 0.6 1.4 57 71 2.0 1.4 14 17 7320 6070
0.61 1.2 <0.05 70 0.3 <0.05 0.51 16 66 40 0.8 2000 3 2.8 47 0.3 1.8 8.2 82 0.4 7.6 9020
200 555 <0.05 170 1.2 14 0.3 22 6.1 19 1.4 7550 17 5.6 63,000 1710 2.6 363 285 175 9.4 19,700
Complete tabulation of bulk chemistry data is presented in Piatak et al. (2003). Number of analyses. Major elements analyzed by ICPAES (except Mn and Ti which were analyzed by ICPMS and Si by XRF). Minor elements analyzed by ICPMS.
Table 6 Results of leachate tests on samples of slaga,b
Parameter
Method
Units
Leach Elizabeth Ely Ducktown Clayton solutionc 00JH05 01JH27 01JH27Dup 01JH28 01JH28rind 01JH28core 01JH37rind 01JH37core Average 01JH34 01CB25 53JH00 (n=7)d
pH
Specic cond.
mS/cm
Alkalinity
mg/l CaCO3
DI ESP DI ESP DI
4.83 4.56 30.6 35.6 0
4.6 4.16 88.5 73 0.19
4.71 4.3 78.8 83 0
5.16 4.95 21.7 27.2 2.72
5.5 5.27 7.4 12.1 4.05
5.23 5.45 6.6 7.9 2.82
5.5 5.41 10.7 12.1 2.33
5.5 5.19 8.2 12.5 1.93
5.2 5.00 24.8 25.8 2.01
5.07 4.73 18.6 23.7 1.22
5.5 5.16 35.3 41.7 3.14
5.65 5.73 20.4 25.6 3.19
Al
ICPMS
mg/l
As
ICPMS
mg/l
ICPAES mg/l
Ba
ICPMS
mg/l
Ca
ICPAES mg/l
Cd
ICPMS
mg/l
Cl
IC
mg/l
Co
ICPMS
mg/l
Cs
ICPMS
mg/l
Cu
ICPMS
mg/l
Fe
ICPAES mg/l
ICPAES mg/l
Mg
ICPAES mg/l
Mn
ICPMS
mg/l
Mo
ICPMS
mg/l
ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI
0 50 88 0.2 <0.2 72 56 2 2 0.64 0.59 2.5 2.6 0.2 0.1 55 55 0.08 0.09 1100 1800 1.6 1.8 0.31 0.3 0.16 0.15 20 19 0.11
0.3 53 99 0.3 0.3 60 55 0.8 4 3.6 2.8 0.2 0.2 0.2 0.2 20 18 0.1 0.1 19 55 11 8.9 0.31 0.26 0.88 0.72 37 30 2.3
0 44 95 0.3 <0.2 56 54 0.8 1 2.8 3.5 0.2 0.2 0.1 0.2 17 22 0.09 0.1 19 54 10 9.7 0.27 0.3 0.78 0.78 32 34 2.3
2.22 6.9 27 0.3 0.2 58 43 0.8 1 0.88 0.84 0.9 1 0.1 0.1 28 29 0.1 0.2 470 870 0.28 0.5 0.23 0.21 0.86 0.87 68 73 0.25
3.13 3.2 9.5 <0.2 <0.2 73 59 0.9 2 0.48 0.62 0.5 0.81 0.1 0.1 3.9 5.7 0.02 0.03 240 520 0.077 0.2 0.23 0.22 <0.1 <0.1 8.6 13 0.84
1.87 4.4 3.1 <0.2 <0.2 66 61 1 2 0.45 0.48 0.71 0.71 0.1 0.1 4.8 4.8 0.02 0.02 200 210 0.077 0.062 0.16 0.19 <0.1 <0.1 9.7 11 0.74
1.21 3.2 2.6 <0.2 <0.2 52 42 0.8 2 0.32 0.34 1 1 0.1 0.1 17 17 0.03 0.04 720 860 0.1 0.13 0.23 0.21 <0.1 <0.1 6.7 6.8 0.33
1.08 2.5 5.8 0.2 <0.2 53 43 1 2 0.21 0.24 1.6 1.8 0.1 0.1 16 18 0.03 0.04 580 920 0.099 0.18 0.18 0.16 <0.1 <0.1 4.2 5.4 0.44
1.40 18 34 0.2 <0.2 62 51 1 2 0.94 0.84 1.1 1.2 0.1 0.1 21 21 0.05 0.07 480 750 1.9 1.7 0.24 0.22 0.33 0.31 22 23 0.72
0.28 2.48 9.7 1.6 29 2.4 <0.2 0.2 <0.2 <0.2 44 41 32 27 4 4 4 4 0.45 4.9 0.43 5.2 0.86 0.07 0.82 0.1 0.2 0.2 0.1 0.2 35 2 33 2.3 0.02 0.02 0.03 0.02 1100 99 1700 260 1.1 0.037 1.3 <0.02 0.26 0.56 0.26 0.58 0.12 0.15 0.12 0.15 35 67 33 75 0.21 0.14
1.73 7.1 9.2 52 52 50 40 5 5.8 0.82 1 0.3 0.4 0.2 0.1 0.2 0.2 0.03 0.04 650 840 <0.02 <0.02 0.34 0.36 <0.1 <0.1 120 140 0.39 (continued on next page)
1055
1056
Table 6 (continued)
Parameter
Method
Units
Elizabeth Ely Ducktown Clayton Leach solutionc 53JH00 00JH05 01JH27 01JH27Dup 01JH28 01JH28rind 01JH28core 01JH37rind 01JH37core Average 01JH34 01CB25 (n=7)d
Na
ICPAES mg/l
Ni
ICPMS
mg/l
Pb
ICPMS
mg/l
Sb
ICPMS
mg/l
Si
ICPAES mg/l
SO4
IC
mg/l
Sr
ICPMS
mg/l
ICPMS
mg/l
ICPMS
mg/l
Zn
ICPMS
mg/l
ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP DI ESP
0.12 0.81 0.7 16 16 0.2 0.3 0.75 0.37 4.6 4.1 10 11 3.1 2.9 <0.3 <0.3 2.7 3.2 420 420
7.5 0.59 0.5 21 30 <0.05 <0.05 0.1 0.15 19 15 35 27 4 3.4 <0.3 <0.3 3.4 4 70 76
0.84 0.47 0.56 23 34 <0.05 <0.05 0.15 0.1 6.2 15 30 32 3.4 4.1 <0.3 <0.3 2.9 4.3 64 78
0.16 0.39 0.34 8.8 10 0.62 0.2 8.2 2.8 23 3.7 4.8 8.2 1 1 <0.3 <0.3 0.48 1.3 270 300
0.33 0.39 0.41 1.9 3.1 <0.05 <0.05 <0.03 <0.03 24 4.1 0.4 2 1.4 2 0.4 <0.3 0.03 0.11 74 110
0.69 0.26 0.29 2 2 <0.05 <0.05 0.04 0.03 22 15 0.7 1.5 1.3 1.4 0.6 0.4 0.042 0.02 76 79
0.28 0.57 0.56 16 17 <0.05 <0.05 <0.03 <0.03 6.9 5.6 1.9 3.1 1.1 1.2 <0.3 <0.3 0.089 0.1 180 180
0.22 0.4 0.42 15 16 <0.05 <0.05 <0.03 <0.03 23 5.6 1.4 2.9 0.9 1.1 0.4 <0.3 0.04 0.17 170 200
1.3 0.49 0.46 12 13 0.15 0.11 1.3 0.49 18 7.6 7.7 8.0 1.8 1.9 0.4 0.3 0.97 1.3 180 200
0.14 0.25 0.23 9.9 9.3 <0.05 0.2 <0.03 2.3 7.5 4 5.5 7.5 1.9 1.8 <0.3 <0.3 0.66 1 150 150
0.092 <0.1 <0.1 0.2 0.3 0.08 <0.05 0.45 0.24 19 3.9 11 13 23 25 <0.3 <0.3 0.02 0.03 290 470
0.25 <0.1 0.11 3 3.8 8400 11000 230 290 20 5.5 4.7 7.5 17 23 0.9 0.4 0.38 0.58 310 370
Complete tabulation of leachate results is given in Piatak et al. (2003). Concentrations of Ag (0.01 mg/l), Be (0.05 mg/l), Cr (1 mg/l), Se (0.2 mg/l) and Ti (50 mg/l) were at or below the Detection limit. Detection limit given in parentheses. Samples were leached with deionized water (DI) and eastern synthetic precipitation (ESP). Calculation of average does not include the duplicate sample 01JH27Dup.
1057
The dominant anion in the leachate is SO2. The highest 4 (35 mg/l) and the lowest (0.4 mg/l) concentrations of SO2 are found in leachates produced from Elizabeth 4 mine slag. Cations are dominated by Si, Ca, Na and K. Iron, Cu and Mg are major constituents in a few samples. The highest concentration of Ca is found in leachate from the Ducktown granulated slag. The bulk chemical analysis of this slag sample indicates an intermediate concentration of Ca relative to samples from other locations; thus, the anomalous dissolved Ca concentration in the leachate may be due to the dominance of glass in this sample ( > 95 vol.%), which is expected to be more reactive than the olivines and spinels that dominate the other samples. The total dissolved basemetals (Cd+Co+Cu+Ni+Pb+Zn) range from 120 to 1590 mg/l for DI and from 180 to 2290 mg/l for ESP, for the Vermont and Tennessee leachates. Leachates from samples collected at the Elizabeth mine contain both the highest (1800 mg/l, sample 00JH05) and the lowest (19 mg/l, sample 01JH27) concentrations of Cu when compared to other leachates. The low concentration of Cu in leachate from sample 01JH27 reects the low Cu content of the bulk slag (Table 5). In contrast, the total dissolved base-metal concentrations in leachate from the Clayton slag are 9360 mg/l for DI and 12,210 mg/l for ESP. These leachates contain the highest concentrations of elements such as As (52 mg/l), Ba (5.8 mg/l), Mn (140 mg/l), Pb (11,000 mg/l) and Sb (290 mg/l), which reect the higher concentrations of these elements in the bulk slag sample (Table 5). In contrast, the concentration of Zn in leachate from the Clayton slag is intermediate in comparison with the other samples even though the concentration of Zn in the bulk slag sample from Clayton is anomalously high. This suggests that not only bulk chemistry but also mineralogy inuences the leachability of trace metals in these samples. 4.4.2. Speciation and chemical controls on leachate waters Geochemical analyses of leachates were used to model the speciation of dissolved constituents and to evaluate the saturation state of the solutions with various solid phases to investigate potential controls on water compositions. PHREEQC (version 2.6) was used for the calculations (Parkhurst and Appelo, 1999) with the WATEQ4F database (Ball and Nordstrom, 1991). The saturation state of the solutions was investigated through the use of saturation indices. The saturation index (SI) for an aqueous solution with respect to a mineral indicates the thermodynamic tendency to precipitate or to dissolve certain phases. The saturation index is dened as: SI log Q=K where Q is the ion activity product and K is the equilibrium solubility product. If the saturation index is zero,
the solution is saturated at equilibrium. A negative saturation index indicates undersaturation, whereas a positive value indicates supersaturation with respect to a certain phase. Model calculations indicate that leachates are at or near equilibrium with respect to several phases. Saturation indices calculated for chalcedony in several of the leachates from Elizabeth, Clayton and Ducktown range from 0.0 to 0.2 (Fig. 8a), suggesting that microcrystalline silica may be limiting the upper concentration of dissolved SiO2. The remainder of the samples range between 0.7 and 0.3. Likewise, dissolved Al concentrations may be limited by saturation with respect to gibbsite or amorphous Al(OH)3; saturation indices for gibbsite range between 1.5 and 1.1. The presence of thin coatings of amorphous Al hydroxides on slag piles at the abandoned Ely mine site is consistent with this conclusion. The variation of the activity of Al3+ with pH indicates that only the Elizabeth leachate with the lowest observed pH (4.2) is signicantly undersaturated with respect to gibbsite and amorphous Al(OH)3 (Fig. 8b). In contrast, saturation indices for ferrihydrite (Fe(OH)3) are negative (6.2 to 3.2). The variation of the activity of Fe3+ with pH indicates that all leachates are signicantly undersaturated with respect to amorphous Fe(OH)3 (Fig. 8c). Hematite and goethite both have positive saturation indices in all samples; however, they are not expected to exert controls on leachate chemistry due to kinetic barriers to their precipitation (Nordstrom and Alpers, 1999). All other geologically reasonable phases were signicantly undersaturated. Inverse modeling of the major element chemistry (pH, Ca, Fe, Al, Si, O2, SO2 and alkalinity) of the starting 4 waters (DI or ESP) and leachates from Elizabeth, Ely and Clayton slags using PHREEQC indicates that the massbalance relationships of the primary phases identied in the slag alone cannot account for the leachate chemistry strictly through the dissolution of phases. Instead, precipitation of secondary phases is required to fully account for the observed leachate chemistry. For the Elizabeth and Ely slags, the leachates can be reasonably described by the dissolution of fayalite, glass, spinel and pyrrhotite with the addition of atmospheric O2. Some models also require the dissolution of silica. However, the precipitation of Al(OH)3 and ferrihydrite or other Al- and Fe-bearing phases is also required. The saturation indices indicate that Al(OH)3 is near saturation in most of the leachates, but ferrihydrite is far from being saturated. Other Fe phases not in the database are not likely to be important. A possible explanation for this discrepancy is that average glass compositions from Table 2 were used in the calculations, but in the leaching experiments, less Fe-rich portions of glass could have been dissolving, thus removing the need for a secondary sink for Fe. The decreased Fe content would need to be oset by increased contents of Si, Ca, or Al. The
1058
requirement of some of the models to dissolve silica is consistent with this hypothesis. For the Clayton slag, the leachates can be reasonably described by the dissolution of fayalitekirschsteinite, hedenbergite, spinel,
glass and galena with the addition of atmospheric O2. As with the Elizabeth and Ely slags, some models also require the dissolution of silica. The precipitation of Al(OH)3 and ferrihydrite is also required. The satura-
Fig. 8. (a) Saturation indices (SI) for chalcedony plotted as a function of pH. (b) Logarithm of the activity of the dissolved Al ion plotted against pH with the equilibrium solubility lines for gibbsite and amorphous Al(OH)3 shown as dashed lines. (c) Logarithm of the activity of the Fe(III) plotted against pH with the equilibrium solubility for amorphous Fe(OH)3 show as a dashed line.
1059
tion indices indicate that Al(OH)3 is near saturation in most of the leachates, but ferrihydrite is far from being saturated. As with the Elizabeth and Ely slags, the average glass compositions from Table 2 used in the calculations may not accurately describe the most reactive glass in the slag. The requirement of the dissolution of silica and the precipitation of ferrihydrite produced by some of the models and the absence of evidence of these processes in the experiments suggest that glass higher in silica and lower in Fe than the average glass composition used in the modeling may be important in the weathering of the Clayton slag. The inverse modeling lacks the sensitivity to rigorously address trace metals such as Cu and Zn. For the slags, chalcopyrite and related sulde minerals are the only signicant sources of Cu in these systems. Therefore, it is reasonable to assume that the factors controlling the dissolution of these phases will control the mobility of Cu. Zinc is distributed among several solid phases, including olivine (up to 1.25 wt.% ZnO), spinel (up to 2.07 wt.% ZnO) and glass (up to 4.16 wt.% ZnO). Clinopyroxene from the Clayton slag contains up to 0.9 wt.% ZnO. Notably, sphalerite is absent in all samples, except Clayton slag. Laboratory and modeling studies of the reactivity of phases in slags with aqueous solutions suggest that olivines and glasses typically are 13 orders of magnitude more reactive than spinels (Kucha et al., 1996; Ettler et al., 2001a,b; Parsons et al., 2001). Thus, for the slags in this study, spinelgroup minerals are less likely to be important sources of Zn in leachates than olivines and glasses. Similar conclusions probably are applicable to Fe and Al.
5. Potential environmental impacts The results of this study provide insights into the potential environmental impacts from slag deposits. Mineralogical and geochemical characterization of the slags has demonstrated that various metals are present in signicant quantities. Furthermore, leaching studies demonstrate that many of the elements are readily liberated from slags by deionized water and solutions approximating precipitation in the eastern United States. The concentrations of several elements exceed water-quality guidelines for the protection of aquatic life. The comparison of the leachate chemistry with hardness-based water-quality guidelines was made assuming a hardness of 100 mg/l CaCO3 equivalent (USEPA, 1993)a value that is characteristic of the unimpacted watersheds in the Vermont copper belt (Seal et al., 2001). In general, the exceedances of water-quality guidelines correlate well with the geochemical nature of the ore being smelted. The concentrations of Cu in the leachate for all samples are higher than both the acute and
chronic toxicity guidelines (Fig. 9a). Copper is an important metal in all of the mining districts, especially in Vermont. Likewise, Zn is important in all of the districts and its concentrations in most leachates are above both acute and chronic guidelines, except for a few samples from Elizabeth (Fig. 9b). In contrast, Pb only exceeds acute water-quality guidelines in leachates from the Clayton slag, which was the most enriched in this element (Fig. 9c). Cadmium exceeds the chronic waterquality guidelines for some samples of Elizabeth slag. The results of this study indicate that the compositions of leachates do not reach the extreme concentrations of the most contaminated drainage in these mining districts nor do they approach the concentrations of the least contaminated waters in these districts (Fig. 9). However, laboratory reaction times may be signicantly less than those in nature. Thus, although slag deposits may not be the most signicant sources of metals in watersheds, the potential magnitude of their impact cannot be ignored; instead, their role in determining the overall water quality in a watershed should be evaluated. Comparison of leachate chemistry with surface waters from the mining districts highlights some of the limitations on using leaching studies to describe natural systems. Clearly, the metals of concern in a given district are strongly related to the geochemical nature of the ores being processed. As discussed, leachates from the slags associated with the Cu mines in Vermont and Tennessee and the polymetallic deposits around Clayton are high in Cu and Zn. Nevertheless, the local geologic and hydrologic setting can serve to either mitigate or amplify the environmental impact of slag deposits on water quality. For example, the variations of the dissolved concentrations of Cu, Zn and Pb with pH for leachates from Vermont and Tennessee fall within the elds dened by the local impacted watersheds, which suggest that the leaching experiments may provide a reasonable representation of the natural system. In contrast, the Clayton leachate falls well outside of the eld dened by drainage from the vicinity of the slag piles. The reason for this dierence is that carbonate rocks are an important part of the geology underlying the watershed. The carbonate rocks serve to elevate pH, alkalinity and hardness and provide conditions signicantly dierent from those for the leaching experiments. Thus, the application of the results of the leaching studies to this watershed must be done with caution. In addition, the proximity of mine waste to streams inuences the potential for watershed impact. Several of the slag deposits examined are along the banks of rivers, thereby increasing the potential for metals to be released into the watersheds. The pot slag deposit from the Ely smelter was dumped along the north bank of Schoolhouse Brook. The granulated slag produced in the Ducktown mining district was depos-
1060
ited along Davis Mill Creek, also, slag from the Clayton smelter in Idaho was placed along the bank of the Salmon River. For these sites, watersheds are more likely to be contaminated by metals released from slag piles.
Overall, the toxicity of leachate produced in a laboratory can be combined with information regarding regional geology and proximity to streams and rivers to better understand the possible extent of environmental
Fig. 9. Concentration of Cu (a), Zn (b), Pb (c) in leachates. Closed symbols are for leach tests using deionized water (DI) and open symbols are for leach tests using eastern synthetic precipitation (ESP). Water-quality guidelines for protection of aquatic life (USEPA, 1993) based on a hardness of 100 mg/l CaCO3 are as labeled. The composition of contaminated drainage for mine sites is plotted as open elds for Vermont and Ducktown mining districts and the Clayton mining district. Drainage water data for Vermont and Ducktown are from Seal et al. (2001) and D. Williams (unpublished data, 2002), respectively. Drainage water data for the Clayton mine are from Hammarstrom et al. (2002).
1061
contamination through leaching of potentially toxic elements.
6. Conclusions This study examined slag deposits from 3 historic mining districts. Deposits at the Elizabeth and Ely mines in the Vermont copper belt and at the Clayton smelter site in Idaho are pot slag casts or solidied slag ows, in contrast to the granulated slag deposit in the Ducktown mining district. Slag samples from the 3 mining districts are mineralogically distinctive, but overall deposits within each district are similar. Primary slag phases include: olivine-group minerals, glass, spinel-group minerals, sulde minerals and native metals for Vermont samples; glass, sulde minerals and native metals for the Ducktown sample; olivine-group minerals, clinopyroxenes, spinels, sulde minerals, native metals and unidentied metallic compounds for Clayton smelter slag. The olivine, pyroxene and spinel phases contain up to several weight percent ZnO and thus represent important sources of Zn. In contrast, Co, Ni, Cu, As, Ag, Sb and Pb occur in the glass phase and in suldes, metallic phases, and unknown metallic compounds. Mineralogical characterization to determine which phases sequester trace elements, in conjunction with leaching studies, indicate that slag contains metals that can be released under ambient conditions. The results of laboratory leachate experiments show that signicant amounts of Cu and Zn are released from slag in all 3 mining districts. The concentration of Cu in the leachate from all samples exceeds the USEPA chronic toxicity guideline for aquatic habitats and the concentration of
Zn in most leachates produced from the Vermont samples exceed the acute Zn toxicity standard. Also, the leachate from the Clayton smelter slag contains high concentrations of Pb, Sb and As. The concentration of Pb in the Clayton leachate exceeds the chronic toxicity guideline by several orders of magnitude. Thus, slag releases a variety of elements in the laboratory and the reactivity of the slag deposits in nature is evident in the formation of secondary minerals. Several of the waste piles occur proximal to streams and the combination of their reactivity and common proximity to streams indicates that slag may be a source of potentially toxic elements at abandoned mine sites. Further studies are needed to determine on a site-by-site basis if the waste piles are to be considered in remediation plans.
Acknowledgements The authors would like to thank Ed Hathaway and Bill Lovely of the US Environmental Protection Agency and Scott Acone of the US Army Corps of Engineers for their assistance and cooperation. The manuscript beneted from interesting and helpful discussions with Matthew Kierstead, PAL, Inc. and Carter Hearn, US Geological Survey. Harvey Belkin provided guidance on SEM and electron-microprobe operations. We also thank Tom Moyer, Science Applications International Corporation and David Williams, US Environmental Protection Agency. Harvey Belkin and Bruce Lipin of US Geological Survey provided thoughtful reviews of the manuscript. The use of trade, product, industry, or rm names in this report is for descriptive purposes only and does not constitute endorsement of the US Government.
1062
Appendix Sample description

Sample number 00JH05 Mining district Mine Smelter/slag pile location Sargent Brook/ west slope of Copperas Hill Smelting date 18801890 Slag type Air-cooled Sample description Gray to black; ow banding with some areas botryoidal; iridescent and gray glassy surface Gray to black with yellow coatings Gray to black; ow banding and rough blocky surface; yellow coatings Textural description Vesicular, spinifex
Vermont Elizabeth copper belt
01JH27a
Vermont Elizabeth copper belt Vermont Elizabeth copper belt
Heckscher smelter/ 19051909 base of tailings pile Heckscher smelter/ 19051909 base of tailings pile
Pot slag
Vesicular, spinifex
01JH27b
Air-cooled
Fine-grained to glassy
01JH28
Vermont Elizabeth copper belt
Furnace Flat/ north bank of the Ompompanoosuc River Furnace Flat/ north bank of the Ompompanoosuc River Furnace Flat/ north bank of the Ompompanoosuc River Not known
185467, 1884 Air-cooled
Black and dark gray Vesicular, spinifex layers, ow banding; brown, red and iridescent coatings Dark gray; red coatings and glassy surfaces Core, spinifex and blocky
01JH28core Vermont Elizabeth copper belt
185467, 1884 Pot slag
01JH28rind Vermont Elizabeth copper belt
185467, 1884 Pot slag
Dark gray; red coatings and glassy surfaces
Rind, vesicular, spinifex
01JH37core Vermont Elizabeth copper belt
Not known
Pot slag
Gray to black; red, brown and iridescent coatings Gray to black; red, brown and iridescent coatings Gray; iridescent, brown and green coatings Gray to black; ow banding; iridescent, brown and red coatings
Core, spinifex
01JH37rind Vermont Elizabeth copper belt
Not known
Not known
Pot slag
Rind, spinifex
01JH34a
Vermont Ely copper belt
Ely/north bank of 178671905 Schoolhouse Brook
Pot slag
Vesicular, spinifex
01JH34b
Vermont Ely copper belt
Ely/north bank of 18671905 Schoolhouse Brook
Pot slag
Vesicular, spinifex
01CB25
Ducktown mining district Bayhorse mining district
Ducktown along Davis Mill Creek
19200 s1987
Granulated Black; conchoidally fractured sandy bits
Glassy
53JH00
various mines
Clayton smelter/ north bank of Salmon River
18800 s1902
Pot slag
Gray; metallic luster; iridescent, white and green coatings
Vesicular, spinifex
1063
References
Anonymous. Sulfuric Acid Production: From Ore to Acid. Polk County Publishing, Benton, TN. Ball, J.W., Nordstrom, D.K., 1991. Users Manual for WATEQ4F, with revised thermodynamic data base and test cases for calculation speciation of major, trace and redox elements in natural waters. US Geol. Surv. Open-File Rep. 91-183. Blair, J.C., 1966. Fluosolids roasting of copper concentrates at Copperhill increases smelter capacity and improves the recovery of sulfur. J. Metals 18 (3), 324327. Cabri, L.J., 1973. New data on phase relations in the CuFeS system. Econ. Geol. 68, 443454. Collins, R.J., Miller, R.H., 1979. Utilization of mining and mineral processing wastes in the United States. Miner. Environ. 1 (1), 819. Crock, J.G., Arbogast, B.F., Lamothe, P.J., 1999. Laboratory methods for the analysis of environmental samples. In: Plumlee, G.S., Logsdon, M.J. (Eds.), The Environmental Geochemistry of Mineral Deposits, Part A: Processes, Techniques and Health Issues. Reviews in Economic Geology, 6A, pp. 265287 (Chapter 13). De Jong, B., Schramm, C.M., Parziale, V.E., 1984. Polymerization of silicate and aluminate tetrahedra in glasses, melts and aqueous solutionsV. The polymeric structure of silica in albite and anorthite composition glass and the devitrication of amorphous anorthite. Geochim. Cosmochim. Acta. 48, 26192629. Ettler, V., Legendre, O., Bodenan, F., Touray, J., 2001a. Primary phases and natural weathering of old lead-zinc pyrometallurgical slag from Pr bram, Czech Republic. The Can. Mineral 39, 873888. Ettler, V., Mihaljevic, M., Touray, J., 2001b. Metallurgical slag behavior in extreme conditions: surface leaching and metal mobility. ENSMP Mem. Sc. De la Terre 40, 8183. Gee, C., Ramsey, M.H., Maskall, J., Thornton, I., 1997. Mineralogy and weathering processes in historical smelting slags and their eect on the mobilization of lead. J. Geochem. Explor. 58, 249257. Hageman, P.L., Briggs, P.H., 2000. A simple eld leach test for rapid screening and qualitative characterization of mine waste dump material on abandoned mine lands. ICARD 2000, Proc. 5th Internat. Conf. Acid Rock Drainage II 14631475. Hammarstrom, J.M., Eppinger, R.G., Van Gosen, B.S., Briggs, P.H., Meier, A.L., 2002. Case study of the environmental signature of a recently abandoned, carbonate-hosted replacement deposit: The Clayton Mine, Idaho. US Geol. Surv. Open-File Rep. 02-10. Johnson, E.A., Oden, L.L., Sanker, P.E., 1982. Results of E.P.A. extraction procedure toxicity test applied to copper reverberatory slags. US Bur. Mines Rep. Invest. 8648. International Centre for Diraction Data, 1997. Powder Diraction File Sets 147. Kierstead, M.A., 2001. History and historical resources of the Vermont copper belt. Society of Economic Geologists Guidebook Series 35, 165191. Kucha, H., Martens, A., Ottenburgs, R., De Vos, W., Viaene, W., 1996. Primary minerals of ZnPb mining and metallur gical dumps and their environmental behavior at Plombieres, Belgium. Envir. Geol. 27, 115.
Lagos, G.E., Luraschi, A., 1997. Toxicity characteristic leaching procedure (TCLP) applied to Chilean primary copper slags. Trans. Inst. Min. Metall. Sec. C 106, C95C97. Levin, E.M., Robbins, C.R., McMurdie, H.F., 1964. Phase Diagrams for Ceramists. The American Ceramic Society, Columbus, Ohio. Lottermoser, B.G., 2002. Mobilization of heavy metals from historical smelting slag dumps, north Queensland, Australia. Mineral. Mag. 66, 475490. Magee, M., 1968. Geology and ore deposits of the Ducktown district, Tennessee. In: Ridge, J.D. (Ed.), Ore Deposits of the United States, 19331967, 1. American Institute of Mining, Metallurgical and Petroleum Engineers, Inc, New York, NY, pp. 207241. Manasse, A., Mellini, M., Viti, C., 2001. The copper slags of the Capattoli Valley, Campiglia Marittima, Italy. Eur. J. Mineral. 13, 949960. Manz, M., Castro, L.J., 1997. The environmental hazard caused by smelter slags from the Sta. Maria de la Paz mining district in Mexico. Environ. Pollut. 98, 713. Maskall, J., Whitehead, K., Gee, C., Thornton, I., 1996. Longterm migration of metals at historical smelting sites. Appl. Geochem. 11, 4351. Maskall, J.E., Thorton, I., 1998. Chemical partitioning of heavy metals in soils, clays and rocks at historical lead smelting sites. Water, Air and Soil Pollut. 108, 391409. Moyer, T.C., Jourdan, R., Taylor, M., Mozingo Jr., J.T., 2000. Challenges of restoring an environment impacted by a century of copper mining: the Copper basin mining district, Tennessee. Geol. Soc. Am. Abs. With Programs 32, 476477. Moyer, T.C., Dube, T.E., Johnsen, M.G., 2002. The impacts of hardrock mining in eastern Tennessee: integrated studies of Davis Mill Creek and the Copper Basin mine site. Geol. Soc. Am. Abs. With Programs 34, 143144. Mukhophadhyay, D.K., Lindsley, D.H., 1983. Phase relation in the join kirschsteinite (CaFeSiO4)- fayalite (Fe2SiO4). Am. Mineral. 68, 10891094. Nordstrom, D.K., Alpers, C.N., 1999. Geochemistry of acid mine waters. Reviews in Economic Geology 6A, 1331160. Oeld, T.W., Slack, J.F., 1993. Structure and origin of the Ely copper deposit, east-central Vermont. U.S. Geol. Surv. Bull. 2039, 5968. Packard, G.A., 1908. Mining in the United States during 1907, reviews of progress: Vermont. Engin. Min. J. 85, 4748. Parkhurst, D.L., Appelo, C.A.J., 1999. Users guide to PHREEQC (Version 2)- A computer program for speciation, batch-reaction, one-dimensional transport and inverse geochemical calculations. US Geol. Surv. Water-Resour. Invest. Rep. 99-4259. Parsons, M.B., Bird, D.K., Einaudi, M.T., Alpers, C.N., 2001. Geochemical and mineralogical controls of trace element release from the Penn Mine base-metal slag dump, California. Appl. Geochem. 16, 15671593. Peters, E.D., 1911. The Practice of Copper Smelting. McGrawHill Book Company, NY. Piatak, N.M., Seal, R.R., II, Hammarstrom, J.M., Meier, A.L., Briggs, P.H., 2003. Geochemical characterization of slags, other mine waste and their leachate from the Elizabeth and Ely mines (Vermont), the Ducktown mining district (Tennessee) and the Clayton smelter site (Idaho). US Geol. Surv. Open-File Rep. 03-260.
1064
N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064 Slater, W.R., Misra, K.C., Acker, C.P., II. 1985. Field trip #5: Massive sulde deposits of the Ducktown district, Tennessee. In: Woodward, N.B. (Ed.), Field Trips in the Southern Appalachians, Department of Geological Science, Univ. Tennessee, Studies of Geology, no. 9, pp. 171190. US Environmental Protection Agency, 1993. EPA Region VIII Clean Water Act Section 304(a) criteria chart, 1 July 1993. US Environmental Protection Agency. US Environmental Protection Agency, 1994. Test Methods for Evaluating Solid Waste, Physical/chemical Methods (SW846), third ed., update 2B. Environmental Protection Agency National Center for Environmental Publications, Cincinnati, Ohio. Wells, M.W., 1983. Gold camps and silver cities-nineteenth century mining in central and southern Idaho. Idaho Bur. Mines Geol. Bull., second ed., 22. Wilson, L.J., 1994. Literature review on slag leaching. Canada Centre for Mineral and Energy Technology, Mineral Sciences Laboratories Division Report 94-3 (CR), Ottawa. Woodley, N.K., Walters, J.V., 1986. Hazardous waste characterization extraction procedures for the analysis of blastfurnace slag from secondary lead smelters. Envir. Prog 5, 217223.
Robbins, D.A., Bundy, S.D., Stanga, G.R., 1983. Availability of toxic metals from non-ferrous metallurgical slags using various procedures. In: Sohn, H.Y., George, D.B., Zunkel, A.D. (Eds.), Advances in Sulde Smelting, vol. 2. The Metallurgical Society of AIME, pp. 923934. Ross, C.P., 1937. Geology and ore deposits of the Bayhorse region, Custer County, Idaho. US Geol. Surv. Bull. 877. Saez, R., Nocete, F., Nieto, J.M., Capitan, M.A., Rovira, S., 2003. The extractive metallurgy of copper from Cabezo Jure, Huelva, Spain: chemical and mineralogical study of slags dated to the third millennium B.C. Can. Mineral. 41, 627 638. Scales, M., 1986. Smelter modernization. Can. Mining J. 107, 4450. Seal II., R.R., Kornfeld, J.M., Meier, A.L., Hammarstrom, J.M., 2001. Geochemical setting of mine drainage in the Vermont copper belt. Soc. Econ. Geol. Guidebook Series 35, 255276. Slack, J.F., Oeld, T.W., Woodru, L.G., Shanks III., W.C., 2001. Geology and geochemistry of Besshi-type massive sulde deposits of the Vermont copper belt. Soc. Econ. Geol. Guidebook Series 35, 193211.

Piatak 2004

Uploaded by

Copyright:

Available Formats

Piatak 2004

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Piatak 2004

Uploaded by

Copyright:

Available Formats

Applied Geochemistry 19 (2004) 10391064 www.elsevier.

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

(Zn, Mg, Fe2+)(Fe3+, Al, Cr)2O4 Fe2O3

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

maximum minimum average 1 maximum minimum average 1 maximum minimum average 1

0.04 0.00 0.02 37 0.02 0.00 0.00 0.00 14

maximum minimum average 1

7.57 5.88 6.64 0.61

0.95 0.71 0.83 0.09

0.04 0.00 0.01 0.02

Bulk glass Elizabeth

0.12 0.00 0.04 0.06

0.08 0.00 0.03 0.02 0.06 0.01 0.03 0.02

1.37 0.00 0.46 0.48

0.06 0.00 0.03 0.02 0.05 0.03 0.04 0.01

1.06 0.24 0.71 0.25 0.40 0.33 0.38 0.04

1.39 0.05 0.39 0.36 1.09 0.00 0.28 0.54

0.40 0.00 0.03 0.09 0.00 0.00 0.00 0.00

2.07 0.73 1.08 0.48 0.56 0.52 0.54 0.02

0.34 0.00 0.05 0.09 0.03 0.00 0.01 0.01

4.25 1.67 2.69 0.79 2.21 1.79 2.08 0.19

(continued on next page)

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

FeOb Fe2O3b CaO 0.39 0.35 0.37 0.03

maximum minimum average 1

0.07 0.06 0.07 0.01

0.89 0.85 0.87 0.02

0.70 0.68 0.69 0.01

99.12 98.09 98.49 0.52

0.05 0.00 0.02 0.02

0.02 0.00 0.01 0.01

0.02 0.02 0.02 0.00

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

46.55 0.16 0.01 43.27 8.42 0.05 0.01 0.00 1.05

45.43 0.16 0.00 43.37 9.34 0.05 0.00 0.00 1.11

30.61 0.09 0.00 25.90 42.36 0.02 0.00 0.00 0.62

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

0.00 0.49 100.0

0.01 0.54 100.0

0.00 0.51 100.0

0.01 0.39 100.0

galena (PbS) (n=8)

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

N.M. Piatak et al. / Applied Geochemistry 19 (2004) 10391064

Table 6 Results of leachate tests on samples of slaga,b

4.83 4.56 30.6 35.6 0

4.6 4.16 88.5 73 0.19

4.71 4.3 78.8 83 0

5.16 4.95 21.7 27.2 2.72

5.5 5.27 7.4 12.1 4.05

5.23 5.45 6.6 7.9 2.82

maximum minimum average 1 maximum minimum average 1 maximum minimum average 1

maximum minimum average 1

maximum minimum average 1