Release of elements from unprocessed- and processed shale from Kvarntorp : Effect from different pH

(1)

Release of metals from unprocessed- and processed

shale from Kvarntorp

As a function of solution pH

Lovisa E. Karlsson 2013-01-04

Örebro University, School of Science and Technology Environmental science, advanced level, 15 ECTS credits

(2)

Abstract

The alum shale of Sweden is a sulphidic black shale which was mined for its contents of alum during the 17th-19th centuries. During the thirst half of the 20th century alum shale from Kvarntorp was pyrolysed for refinement of its organic matter.

The refuse from the pyrolysis process were deposited in a deposit area which is today known as Kvarntorpshögen. This deposit is still hot due to on-going oxidation of the organic matter, pyrite oxidation and the neutralisation of acid by limestone. This heat generation prevents percolation of water to any greater extent. Once it cools the deposit will be susceptible to extraction.

Three materials from the Kvarntorp deposit and a block of alum shale were sampled from the Kvarntorp area. These four materials were crushed, sieved and extracted. The extraction experiment was performed at controlled pH which corresponded to dissolution equilibrium of pyrite, carbon dioxide/oxygen, calcite and slaked lime. In the fifth extraction series the material was allowed to set the pH of the water phase. The extraction procedure lasted for 26 days.

Metal content of the four shale materials were analysed by XRF (including contents of sulphur, phosphorous and chloride) and by digestion followed by ICP-MS analysis. Mineralogy of the four materials and the extraction residues were studied by XRD analysis. Metal concentrations of the aqueous samples were analysed by ICP-MS and concentrations of dissolved organic- and inorganic carbon by TOC. Electrical conductivity and concentrations of sulphate was also analysed in

aqueous phases with equilibrated pH.

Sulphate and calcium concentrations were highest in aqueous sample from weathered fines indicating the presence of gypsum. Presence of gypsum was also confirmed by XRD. Gypsum indicates weathering since it is a neutralisation product of calcite and the sulphuric acid generated by pyrite oxidation.

Concentrations of dissolved organic carbon in samples with pH 12.5 were found to be highest from red processed shale and lowest from samples of shale. This indicates different hydrolysis properties or origin of the organic matter in the materials.

The majority of analysed metals had highest solid solution distribution under acidic, i.e. pH 3.0, conditions. Under these conditions it was samples from the relatively non-weathered shale that had the highest distribution of metals, with a few exceptions.

At pH 12.5 high distribution of V, As and Mo was extracted from almost all four materials. The red processed shale released the highest percentages of its total contents of V, As and Mo under these conditions; 28 %, 55 % and 61 % respectively. The high release of V, As and Mo from red processed shale was most likely a result from the pyrolysis process. The roasting induced the formation of oxides which then easily forms vanadate, arsenate and molybdate when exposed to water.

Water percolating the Kvarntorp deposit in the future will initially most likely have a near neutral or slightly acidic pH. Lowest concentrations of metals were found in aqueous samples with pH between 5.5-8.5. There is a great risk that the pH of the drainage from the Kvarntorp deposit will in time become acidic due to depletion of buffering capacity. Thus the environmental threat will increase.

(3)

1 Introduction

1.1 Sulphidic black shale

Shale is a fissile mud rock consisting of clay and silt, i.e. particles with a diameter less than 0.062 mm (Nesse, 2009; Blatt et.al., 2006). Mudrocks consists of approximately 60 % clay minerals, 30 % quartz and 10 % various other minerals such as feldspar, carbonates and oxides (Blatt et.al., 2006). Shale is formed in aquatic environments with very little or no turbidity since silt and clay particles are easily kept in suspension (Blatt et.al., 2006).

Sulphidic black shale is a shale type rich in free organic material and pyrite, FeS2, caused by the

oxygen-deficient environments (i.e. reducing conditions) in which the shale is deposited (Blatt et.al., 2006). The black colour of this shale indicates a content of minimum 1 % free organic material (Blatt et.al., 2006).

Since shale consists of small silt and clay particles they tend to accumulate metals during settling due to formation of surface complexes. Sulphidic black shale often contains a higher degree of rare elements such as Cu, As, Sb, Mo, Ni, Ag, U and rare earth elements, REE, than average shale (Blatt et.al., 2006; Bolonin and Gradovsky, 2012; d’Hugues et.al., 2008; d’Hugues and Spolaore, 2008; Grawunder et.al.,2009). The reducing condition caused by anaerobic bacteria and a surplus of oxidizable organic material reduces sulphate to sulphide (Blatt et.al., 2006; Falk et.al., 2006). Under these conditions transition metals precipitates as sulphide minerals and are incorporated in the shale.

Due to the sulphidic black shales high content of sulphide and heavy metals it poses an environmental risk when exposed to oxidising conditions. Atmospheric oxygen or oxygenated groundwater oxidises sulphide minerals and produces sulphuric acid (Nesse, 2009; Lavergren et.al., 2009b; Sohlenius and Öborn, 2004). The acid contributes to a pH decrease which mobilise cations enabling dissolution of heavy metals in the leachates (Drever, 1997; Sartz, 2010). The environmental problems associated with oxidation of sulphidic black shale are similar to those posed by sulphidic mine waste and acid mine drainage (Lavergren et.al., 2009b; Yu et.al., 2012).

The pyrite oxidation is a series of reactions which involves both aerobic- and anaerobic steps. The reactions 1 and 2 are the aerobic reactions (Maia et.al., 2012; Jönsson et.al., 2006; Puura, 1998; Puura and Neretnieks, 2000):

FeS2(s) + 3.5 O2 + H2O → Fe2+ + 2 SO42- + 2 H+ (1)

Fe2+ + 0.25 O2(aq) + H+ → Fe3+ + 0.5 H2O (2)

Reaction 3 is the anaerobic oxidation of pyrite (Maia et.al., 2012; Puura, 1998; Puura and Neretnieks, 2000):

FeS2(s) + 14 Fe3+ 8 H2O → 15 Fe2+ +2 SO42- + 16 H+ (3)

Dissolution of iron in water induces hydrolysis of the ions which results in further pH decrease according to reactions 4 and 5 (modified from Cotton et.al., 1995):

(6)

[Fe(H2O)6]3+ + H2O = [Fe(H2O)5(OH)]2+ + H3O+ (4)

[Fe(H2O)5(OH)]2+ + H2O = [Fe(H2O)4(OH)2]+ + H3O+ (5)

1.2 Kvarntorp

The sulphidic black shale in Sweden is often referred to as alum shale, from its contents of alum, KAl(SO4)2x12H2O, (Lavergren et.al., 2009b; Falk et.al., 2006) which was used in the paper and

dyeing industries from the 17th to19th century (Eklund et.al., 1995).

Due to the alum shales contents of organic matter it has been pyrolysed for the production of fuels at different sites in Sweden (Dyni, 2006). This study focuses on shale materials from the Kvarntorp site in the region of Närke, some 200 km W of Stockholm. Due to the movement of the ice sheet during the last ice age the overlaying limestone was physically weathered and thus the alum shale layer in Kvarntorp was exposed (Anderberg and Johansson, 1981).

Shale was pyrolysed in Kvarntorp from the beginning of the 1940 decade until 1966 (SWECO, 2005a). The refuse from the pyrolysis process was deposited and formed what is now known as the Kvarntorp deposit (swe. Kvarntorpshögen).

There are three chemical reactions going on inside the Kvarntorp deposit which generates heat. Oxidation of organic compounds may generate temperatures of 500 °C, or even up to 700 °C (SWECO, 2005a). Pyrite oxidation generates temperatures around 100 °C (SWECO, 2005a). Formation of carbon dioxide from neutralisation of acid by calcite also generates heat to some extent (SWECO, 2005a).

1.3 This study

Due to the heat generation, water is unable to infiltrate into the deposit to any greater depth. This prevents extraction of the solid phase. Once the deposit cools, extraction of toxic elements will be possible. It is therefore important to study the mobilisation of metals from the shale materials in contact with water. However this study was not primary made for an environmental risk assessment.

This study focuses on the chemical and mineralogical processes involved when unprocessed- and processed shale materials are exposed to solutions with different natural pH. The different pH-values chosen for this study correspond to dissolution equilibriums of gases and minerals present in the area. Pyrite is a very abundant mineral in the area due to its presence in the sulphidic black hale. For this study its dissolution equilibrium was presumed to be at pH 3.0. The dissolution equilibrium is sensitive to redox and under high oxidative conditions the dissolution equilibrium of pyrite may be < 3.0. Equilibrium between water and atmospheric oxygen and carbon dioxide is at pH 5.5. Calcite or limestone is also very abundant at the Kvarntorp site due to the orthoceratite limestone in the SE of Närke (Anderberg and Johansson, 1981). The dissolution equilibrium of calcite is at pH 8.5 (Mõtlep et.al., 2010). Due to the pyrolysis processes at Kvarntorp some amount of brunt lime is

(7)

CaCO3 + heat → CaO + CO2(g) (6)

CaO + H2O → Ca(OH)2 (7)

Ca(OH)2 → Ca2+ + 2OH- (8)

A fifth leaching series was also made were the pH was not controlled, i.e. the materials were allowed to equilibrate to the aqueous phase. Even though this leaching series is hereafter called “equilibrated” the reader should have in mind that it is not actually at equilibrium, rather striving towards it.

Table 1. The five pH conditions of water phases in this study

pH Explanation

Equilibrated The pH set by the material

3.0 Dissolution equilibrium of pyrite

5.5 Dissolution equilibrium of oxygen and carbon dioxide

8.5 Dissolution equilibrium of calcite, CaCO3

12.5 Dissolution equilibrium of slaked lime, Ca(OH)2

2 Materials and methods

2.1 Analytical procedures

2.1.1 pH

Measurements of pH were done with a pH electrode from Metrohm, 6.0228.000, Pt 1000/B/2/3 M KCl. The pH-electrode was calibrated daily using buffer solutions with pH 4, 7 or 10 (LabService AB) depending on the pH-range of the samples.

2.1.2 EC

Electrical conductivity (EC) was only determined in aqueous samples with equilibrated pH using a conductivity electrode from Radiometer Analytical S.A., model CDC866T.

2.1.3 Sulphate in aqueous samples

Sulphate was analysed by ion chromatography with conductometric detection equipped with a supressor. The ion chromatograph was a system from Metrohm equipped with two columns from Dionex; AG12 guard column and AS12A analytical column. The mobile phase was a solution of

(8)

2.7 mM sodium carbonate and 0.3 mM sodium bicarbonate which was pumped at 1 ml/min. The suppressor was supplied with deionised water and a solution of 0.073 M H2SO4 at 0.5 ml/min.

Calibration was done using standard solutions with the sulphate concentrations 1 mg/L, 5 mg/L, 15 mg/L, 30 mg/L and 60 mg/L. Recalibration was made every 40th sample.

2.1.4 Dissolved organic carbon (DOC) and dissolved inorganic carbon (DIC)

Analysis of DOC and DIC was performed with a TOC-Ucph from Schimadzu. Calibration solutions

for total carbon (TC) were solutions of potassium hydrogen phthalate with varying concentrations. Calibration standards for inorganic carbon (IC) were solutions of sodium carbonate and sodium bicarbonate with varying concentrations. A solution of 25 % phosphoric acid (diluted from reagent grade, 85 %) was used for the IC reactor.

Standards containing 10 mg/L TC and 10 mg/L IC was analysed every 10th sample to ensure that the analytical performance of the instrument did not drift.

All samples were filtered through polypropylene filters (0.20 µm) before analysis. Concentration of organic carbon was calculated by subtracting IC from TC.

2.1.5 Spectrophotometry

Some samples were selected for spectrophotometric analysis. Measurements of absorbance were made with a Hewlett Packard 8453 spectrophotometer using a PMMA cuvette. Absorbance was determined in the wavelength range of 190-1100 nm.

2.1.6 XRF

Analysis by XRF was performed with a Spectro Xepos which was a polarised ED XRF with three beam targets and a radiation source made of copper. Calibration of the XRF was performed according to the method Turboquant which is a fundamental parameter method with matrix correlation. The matrix correlation was based on the relationship between Rayleigh- and Compton scattering and the density of the sample which was predicted from the samples weight.

2.1.7 Digestions

Digestions were performed with a MARS5 CEM microwave oven in HP500 super (CEM) Teflon bombs. To each bomb 50 mg material (particle size <0.25 mm) and 5 ml conc. HNO3 (sub boil

distilled) were added.

The power used was set to 600 W at 50 % of the time. Maximum temperature was 180°C and highest allowed pressure was 250 psi. Ramp time was set to 60 minutes i.e. for the whole run time. This caused the samples to be exposed to microwave radiation for the whole run time.

(9)

An internal standard (Rh, 10µg/L) was used to verify that the instrument did not drift during analysis and to correct for the intensity’s dependence of the samples matrices. The original rhodium solution was delivered from MERCK as Rhodium standard solution, 1000 mg/l for ICP, rhodium(III)-chloride in hydrochloric acid 8 %. This solution had then been diluted to a concentration of 1 mg/L with 1 % HNO3.

Analysis was performed by an ICP-MS, Agilent 7500 cx positioned in a clean room. Plasma effect during analysis was 1500 W. The element concentrations of the samples were based on analysis of the following isotopes; 24Mg, 27Al, 43Ca, 51V, 55Mn, 57Fe, 59Co, 60Ni, 63Cu, 66Zn, 75As,

88

Sr, 95Mo, 111Cd, 137Ba, 208Pb and 238U.

Standards containing 1 µg/L of each analysed metal and 10 µg/L of each analysed metal was analysed every 25th sample to ensure that the analytical performance of the instrument did not drift.

2.1.9 XRD

X’pert PRO Powder diffractometer from PANalytical was used for the analysis. The radiation source was constructed of a copper electrode and operated at 45 kV and 40 mA. The detector had a continuous motion between the angles 5°-70° during measurements, its speed was set to 0.1°/sec for the 5 min. scans and 0.02°/sec for the 22 min. scans. The set speed did not correspond to the total time of the scans, what caused this error in the software is unknown.

Scans of 22 min. each were performed on non-leached materials, including a sample of red processed shale sampled in 2011 (Karlsson, 2011), particle size <0.25 mm. Scans of 5 min. each were performed on all samples. Evaluation of diffractograms was made in X’pert Highscore with the database ICDD 2008 and X´pert Dataviewer.

2.2 Sampling of the shale materials

The study included four different shale materials, two unprocessed- and two processed materials. The first was black and relatively pristine alum shale. This was sampled as a block (circa ½ m across and 1 dm thick) from the Östersätter quarry. The second was a material called weathered fines (swe. stybb), which was a crushed fine fraction (<1 cm Ø) of unprocessed shale sampled at the SW base of the Kvarntorp deposit (fig. 1). Its crushed state has exposed the weathered fines to more extensive weathering than the non-crushed black shale.

The third material was red processed shale (swe. rödfyr) which had been pyrolysed and with visibly none of its original organic matter remaining. The fourth material was black processed shale (swe. aska), a partially processed shale which still had some black colour indicating organic matter. The weathered fines, red- and black processed shale were sampled at the SW base, the NE base and the top of the Kvarntorp deposit respectively (fig. 1).

(10)

Fig. 1. The Kvarntorp deposit and the sampling locations for weathered fines (1), red processed shale (2) and black processed shale (3). Picture modified from Google Earth.

2.3 Sample preparation

2.3.1 Extraction of the solid materials and sample preparation of aqueous samples

All four materials were dried at 35°C for two days. The block of black shale was initially crushed into manageable particle sizes with a sledge hammer. The four materials were then crushed using a jaw crusher. The crushed materials were sieved and two particle fractions were collected with particle diameters of 2.0-0.50 mm and <0.50 mm.

The extraction was initially performed at a liquid to solid ratio (L/S) of 10 in 2.5 L plastic cans. The L/S ratio decreased with each sampling. To each can 230 g solid material (2.0-0.50 mm) and 2.3 L deionised water was added. Five cans for each type of shale material were prepared. The pH of the water phase was controlled to those given in table 1 with either 1 M nitric acid (reagent grade, Scharlau) or 1 M sodium hydroxide (reagent grade, Merck). One can with each material was allowed to strive for equilibrium. Nitric acid was chosen over H2SO4 and HCl since nitrate usually

forms weaker complexes than sulphate and chloride.

Adjustments of pH in the cans were made at the beginning, after 6 hours, 2 days, 5 days, 6 days, 7 days, 11 days, 15 days and 22 days. The cans were agitated manually for 10 seconds three times every workday.

Sampling was done after 1 h, 5 d and 26 d from the beginning (first pH adjustment) of the extraction period. Sampling of aqueous samples was performed twice after 1 h and 26 d. Sampling

(11)

tubes. Three mL of each sample was added to separate tubes for metal analysis and was acidified with concentrated HNO3 (sub boil distilled) to a final concentration of 1 %.

Measurements of pH in non-acidified samples were performed within an hour after sampling on sampling occasion 2 (day 5) and 3 (day 26). Measurements of pH in non-acidified samples sampled after 1 hour of leaching were performed the following day. These samples had been stored in a freezer (-20°C) overnight.

Non acidified samples were stored in a freezer until analysis of sulphate, dissolved organic- and inorganic carbon.

2.3.2 Sample preparation of solid samples

Once the extraction process had been concluded extraction residues were rinsed with deionised water, dried at 35°C, crushed and sieved to a particle size < 0.25 mm. Non-extracted sample materials (2.0-0.50 mm) were also crushed and sieved to a particle size < 0.25 mm.

Elemental contents of non-extracted materials were measured by XRF on particle fractions 2.0-0.50 mm and <2.0-0.50 mm. Metal contents of non-extracted materials (particle fraction <0.25mm) was determined by acid digestion and ICP-MS (2.1.7 and 2.1.8). The particle fraction <0.25mm of non-extracted materials and leaching residues were used for mineralogical analysis by XRD.

3 Results and discussion

3.1 XRD

3.1.1 Comparison of extraction residues

Visual inspection of diffractograms was done in X´pert Dataviewer. Diffractograms from non-extracted samples and extraction residues of the same material were compared to each other. All diffractograms from the same type of material were visibly identical. This indicated that the extraction processes had not altered the mineralogy to any extent that was detectable under the used analytical settings.

3.1.2 Identified mineral phases in non-extracted materials

The tables 2, 3, 4 and 6 show the results from two different evaluation methods. The left columns present the mineral phases identified using the program X’pert Highscore while the right columns presents the mineral phases identified by comparing the diffractograms with diffractograms of shale ash from Estonia (Mõtlep et.al.,2010) and of black schist from Finland (Bhatti et.al., 2012).

Quartz had the highest correspondence score to the intensities of the diffractograms when evaluated in X’pert Highscore. This indicates that quartz was very abundant in the materials and was most likely the bulk mineral (tables 2-6). This corresponds to previous mineralogical data reported by SWECO (2005a). Quartz is also reported to be the bulk material for shale from Degerhamn, SE Sweden (Lavergren et.al., 2009a). The presence of pyrite (table 2) was also expected in the shale (Lavergren et.al., 2009b; Yu et.al., 2012). Alunite is a mineral form of alum which is the namesake for black shale of Sweden. Muscovite was also identified in shale from

(12)

South China black shale horizon (Yu et.al., 2012). The minerals muscovite, orthoclase and microcline was identified in shale and weathered shale from Degerhamn (Lavergren et.al., 2009a).

The presence of pyrite indicates that the sampled shale was relatively non-weathered since pyrite had not been completely oxidised.

Table 2. Identified mineral phases in shale by XRD scan

Shale Identified mineral phases by

X’pert Highscore

Identified mineral phases by comparing diffractograms with Mõtlep et.al. (2010) and

Bhatti et.al. (2012)

Quartz – SiO2 Quartz – SiO2

Pyrite – FeS2 Pyrite – FeS2

Muscovite – H2KAl3(SiO4)3 Mica (Biotite/Phlogopite) –

K(Mg,Fe)3AlSi3O10(OH,F)2 / KMg3AlSi3O10(OH,F)2

Alunite – K(Al3(SO4)2(OH)6) Chlorite –

(Mg,Fe)3(Si,Al)4O10(OH)2x(Mg,Fe)3(OH)6

Orthoclase – (K0.931Na0.055)(Al0.97Si3.03O8) Feldspar/Anorthite – CaAl2Si2O8

Wenkite – Ca5Ba4Al9Si11S3O53(OH)4 Pyrrhotite – FeS

Ettringite – Ca6Al2(SO4)3(OH)12x26(H2O)

Anhydrite – CaSO4

Feldspar/Microcline – KAlSi3O8

Calcite – CaCO3

Chalcopyrite – CuFeS2

Sodium chloride – NaCl Lime – CaO

(13)

Fig. 2. Diffractogram from shale in the detector angle range 5-30° 2Theta

(14)

Fig. 4. Diffractogram from shale in the detector angle range 50-70° 2Theta

The presence of gypsum is an indication of weathering (table 3). Limestone is abundant in Kvarntorp and due to the acid drainage from shale it will dissolve. Calcium ions will then bind with the excessive sulphate ions from pyrite oxidation and form gypsum (SWECO, 2005a; Bhatti et.al., 2012; Puura, 1998).

Oxidation of pyrite will also enable extensive formation of iron sulphate minerals such as jarosite. Jarosite forms in iron- and sulphide rich environments with low pH (Sartz, 2010; Carlsson and Büchel, 2005; SWECO, 2005a).

(15)

Table 3. Identified mineral phases in weathered fines by XRD scan

Weathered fines Identified mineral phases by

X’pert Highscore

Gypsum – Ca(SO4)(H2O)2 Gypsum – Ca(SO4)(H2O)2

Szomolnokite – Fe(SO4)(H2O) Jarosite –KFe3(SO4)2(OH)6

Muscovite – H2KAl3(SiO4)3 Ettringite – Ca6Al2(SO4)3(OH)12x26(H2O)

Sodium Silicate – Na2Si3O7 Chlorite –

Feldspar/Anorthite – CaAl2Si2O8

Anhydrite – CaSO4

Mica (Biotite/Phlogopite) –

Calcite – CaCO3

Vermiculite –

(Mg,Fe,Al)3(Al,Si)4O10(OH)2x4(H2O)

Pyrite – FeS2

Pyrrhotite – FeS

Sodium chloride – NaCl Lime – CaO

(16)

(17)

Fig. 7. Diffractogram from weathered fines in the detector angle range 50-70° 2Theta

Quartz and hematite were the only mineral phases that could be identified with any certainty in samples of red processed shale by X´pert Highscore (table 4). Presence of hematite (table 4 and 5) corresponds to data of burnt shale from a shale ash plateau in NE Estonia (Mõtlep et.al., 2010). The mineral phases microcline and hematite (table 4) was also identified in burnt shale from Degerhamn (Lavergren et.al., 2009a). According to Puura (1998), ettringite and the mineral sanidine, which shares the same chemical composition as microcline (KAlSi3O8), was also identified in a shale ash

(18)

Table 4. Identified mineral phases in red processed shale by XRD scan

Red processed shale (sampled 2012) Identified mineral phases by

X’pert Highscore

Hematite – Fe2O3 Mica (Biotite/Phlogopite) –

Chlorite –

Feldspar/Anorthite – CaAl2Si2O8

Anhydrite – CaSO4

Feldpar/Microcline – KAlSi3O8

Calcite – CaCO3

Lime – CaO

Sodium chloride – NaCl

(19)

Fig. 9. Diffractogram from red processed shale in the detector angle range 30-50° 2Theta

(20)

The red processed shale that had been sampled 2011 was used in a previous leaching study made by Karlsson (2011). The electrical conductivity and concentrations of calcium and sulphate in the aqueous phase after contact with this red processed shale indicated the presence of a gypsum phase (Karlsson, 2011). This XRD analysis confirms this assumption (table 5).

The red processed shale used by Karlsson (2011) was sampled at the NW base of the Kvarntorp deposit and might have been pyrolysed under different conditions than the red processed shale sample used in this study. Weathering processes might also have been different at the two different sampling locations.

Table 5. Identified mineral phases in red processed shale sampled in 2011 by XRD scan

Red processed shale (sampled 2011) Quartz – SiO2

Hematite – Fe2O3

Gypsum – CaSO4x2(H2O)

Portlandite – Ca(OH)2

The mineralogy of black processed shale has similarities to red processed shale (table 6). Both of them were expected to contain oxides as a result of the pyrolysis process. See the discussion about pH under heading 3.2 for further information.

Table 6. Identified mineral phases in black processed shale by XRD scan

Black processed shale Identified mineral phases by

X’pert Highscore

Hematite – Fe2O3 Mica (Biotite/Phlogopite) –

Portlandite – Ca(OH)2 Chlorite –

Gypsum – CaSO4x2(H2O) Feldspar/Anorthite – CaAl2Si2O8

Anhydrite – CaSO4

Calcite – CaCO3

Lime – CaO

(21)

Fig. 11. Diffractogram from black processed shale in the detector angle range 5-30° 2Theta

(22)

Fig. 13. Diffractogram from black processed shale in the detector angle range 50-70° 2Theta

3.2 Metal content

Metal contents determined by XRF or digestion followed by ICP-MS analysis gave quite similar results for most elements (fig. 14-17). Some exceptions were all four materials contents of Cu, Zn and Ba, which were reported higher by the digestion and ICP-MS than by XRF. The detected concentrations of uranium in red processed shale were considerably higher by XRF than results from digestion and ICP-MS.

The metal contents determined by digestion and ICP-MS will hereafter be referred to as the total contents.

Trace elements often show a heterogeneous quantitative abundance in solid materials particle fractions, i.e. smaller particles tends to accumulate more metals than larger particles (Arocena et.al., 1995; Rutherford et.al., 1996). Analysis with XRF of the fractions 2.0-0.5 mm and <0.5 mm indicate heterogeneous quantitative abundance of some elements in the different materials (fig. 18).

(23)

Fig. 14. The content of Mg, Al, Ca, V, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Ba, Pb and U in shale quantified by XRF (n=1), and those reported from digestion/ICP-MS (n=2).

Fig. 15. The content of Mg, Al, Ca, V, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Ba, Pb and U in weathered fines quantified by XRF (n=1), and those reported from digestion/ICP-MS (n=2).

(24)

Fig. 16. The content of Mg, Al, Ca, V, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Ba, Pb and U in red processed shale quantified by XRF (n=1), and those reported from digestion/ICP-MS (n=2).

Fig. 17. The content of Mg, Al, Ca, V, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Ba, Pb and U in black processed shale quantified by XRF (n=1), and those reported from digestion/ICP-MS (n=2).

(25)

Fig. 18. Distribution ratios of elements between the two particle fractions: 2.0-0.5 mm / < 0.5 mm

3.3 Contents of P, S and Cl in solid samples quantified by XRF

The total content of sulphur in red- and black processed shale was noticeable lower than in shale and weathered fines (table 7). This was due to the emission of sulphur dioxide during pyrolysis of shale (Eklund et.al., 1995; Falk et.al., 2006; Puura, 1998) and thus was some acidity removed from the materials (Puura, 1998). The sulphur content of shale and red processed shale from Kvarntorp (table 7) seem to have a lower sulphur content compared to shale materials from Degerhamn where the sulphur content of non-weathered- and processed shale was found to be 5.9 and 2.4 % respectively (Lavergren 2008; Lavergren et.al., 2009a; Lavergren et.al., 2009b). The differences between the quantified sulphur contents could be a result of different methods. The sulphur content of the shale materials from Degerhamn were analysed by a sequential leaching process modified from Hall et.al., 1996a, b, (Lavergren, 2008; Lavergren et.al., 2009a) while the sulphur content of Kvarntorp shale was analysed by XRF. Black shale from Ronneburg, Germany, was reported to contain 6-7 % pyrite (Bolonin and Gradovsky, 2012) which equals 3.3-3.8 % sulphur. The sulphur content of a black shale sample from China contained 0.7 % sulphur (Li et.al., 2009) which was lower than the sulphur content of unprocessed shale from Kvarntorp.

Mudrocks may contain up to 0.17 % P2O5 (Blatt et.al., 2006). Phosphate also forms rather

insoluble minerals with rare earth elements, REEs (Blatt et.al., 2006). Since alum shale is presumably rich in REEs, the presence of phosphate is expected.

The phosphorous contents of shale and weathered fines exceed the content of 0.17 % P2O5 (table

7) stated by Blatt et.al. (2006). The reported content of P2O5 in black shale from Ronneburg ranged

from 0.2 to 0.5 % (Carlsson and Büchel, 2005) and the corresponding content of shale from China was 0.19 % (Li et.al. 2009; Li et.al. 2010b).

(26)

The contents of chloride were relatively low compared to sulphur (table 7) which was expected due to previous extraction data reported by Karlsson (2011). Lower contents of chloride compared to sulphur was also found in soil from Ronneburg where the median contents were determined to be 31 mg/kg chloride compared to 4000 mg/kg sulphur (Carlsson and Büchel, 2005).

Table 7. Contents of P, S and Cl analysed by XRF (n=1)

Material and particle size P (mg/kg) S (mg/kg) Cl (mg/kg)

shale (2.0-0.5 mm) ₂₂₀₀ ₃₂₀₀₀ ₃₁₀

weathered fines (2.0-0.5 mm) 1800 45000 440

red processed shale (2.0-0.5 mm) 1300 5300 360

black processed shale (2.0-0.5 mm) 950 5600 360

3.4 pH and EC in equilibrium aqueous samples

As was mentioned under heading 1.1 and shown by XRD analysis under heading 3.1.2 (table 2); the shale was rich in pyrite. Pyrite has dissolution equilibrium around pH 3.0 and was most probable to be the main reason for the low pH in aqueous samples from shale (fig. 19, see oxidation reactions 1, 2 and 3 under heading 1.1). It was most likely that pyrite was not as common in the mineral structure of the weathered fines since the fines have been exposed to more extensive oxidation.

Many metals share irons ability to hydrolyse and thereby lower pH. The low pH in the water phase in contact with weathered fines (fig. 19) may be due to the dissolution of mineral phases containing various pH-lowering metals. The weathering that the fines been exposed to may have modified or weakened their mineral structure and therefore increased the solid solution distribution compared to the shale. This was also indicated by the higher EC in the water phase from weathered fines than in the water phase from shale presented in fig. 20.

Red- and black processed shale contributed to a higher pH in their water phases (fig. 19) than shale and weathered fines. This pH increase was caused by the presence of oxides formed during pyrolysis of the materials. One example is the formation of lime from calcite:

CaCO3 + heat → CaO + CO2(g) (9)

Lime increases pH when slaked with water:

CaO + H2O → Ca(OH)2 (10)

Ca(OH)2 → Ca2+ + 2OH- (11)

Higher pH in water phase in contact with red processed shale than in water phase from black processed shale was expected since the black processed shale was not completely processed and therefore should contain oxides to a less extent. The lower pH in water phase in contact with black

(27)

Fig. 19. The pH of equilibrium samples. n=2 for day 1 and 26, n=1 for day 5.

The electrical conductivity in aqueous sample from red processed shale was very low, not exceeding 50 µS/cm on day 26 (fig. 20). This indicates that the red processed shale either was very resistant to extracion or that there were a high degree of precipitation and coprecipitation of elements at near neutral pH. In the previous extraction study made by Karlsson (2011); aqueous sample of red processed shale had an EC of approximately 2 mS/cm, mainly due to the presence of gypsum mentioned under heading 3.1.2 (table 5).

Fig. 20. Electrical conductivity of equilibrium samples. n=2 for day 1 and 26, n=1 for day 5.

3.4.1 Buffering capabilities

The exact volumes of added HNO3 and NaOH were not noted since they were at the moment

(28)

volumes made by the author is presented in table 8. The presented volumes are the estimated total volumes added during the time period for the extraction process.

Table 8. Estimated sum of added volumes HNO3 or NaOH for pH adjustments during the extraction period, 26 days

pH 3.0 pH 5.5 pH 8.5 pH 12.5

Shale 2 ml HNO3 8 ml NaOH 16 ml NaOH 80 ml NaOH

Weathered fines 0 ml 12 ml NaOH 24 ml NaOH 100 ml NaOH

Red processed shale 10 ml HNO3 2 ml HNO3 6 ml NaOH 70 ml NaOH

Black processed shale 8 ml HNO3 4 ml NaOH 10 ml NaOH 75 ml NaOH

Although the total volumes presented in table 8 are unsure the relationship between the added volumes are more accurate. This provides a rough estimation of the four materials relative buffering capabilities.

Alkalinity is mainly controlled by the carbonate system (Drever, 1997) with the following two reactions:

HCO3- + H+ → H2CO3 pK1 = 6.4 (12)

CO32- + H+ → HCO3- pK2 = 10.3 (13)

The data in table 8 show that the alkalinity was slightly higher in samples from red processed shale compared to samples from the black processed shale. The different amounts of added HNO3 might

be a result of the different pH of the samples water phases. The water phases of the shale and weathered fines samples equilibrated to a low pH, thus inducing the formation of H2CO3 to a high

extent. Carbonic acid in turn dissociates into gaseous carbon dioxide which leaves the system (Drever, 1997):

H2CO3 → H2O + CO2(g) (14)

Due to the high contents of aluminium in the materials detected by XRF (fig. 14-17), it is possible that aluminosilicates contributes to the alkalinity. Aluminosilicates, such as illite, dissolves in acid generating other mineral phases and base cations as in the reaction presented by Puura (1998):

5/3 Illite + H+ → 5/3 Smectite + K+ (15)

Acidity is controlled by cations which are capable of coordinating hydroxide ions. Some examples of such ions are iron and aluminium. Reactions 16-18 show the example of aluminium:

Al3+ + OH- → Al(OH)2+ (16)

(29)

The data in table 8 indicate that the unprocessed shale samples have a higher acidity than the processed materials. The lower acidity in the processed materials might be a result of the higher contents of oxides in these materials as was mentioned under heading 3.4. The water phase from the weathered fines had a higher acidity than that from shale. This could be caused by the higher EC in the water phase from weathered fines, indicating higher concentration of ions.

3.5 Sulphate in aqueous samples

Sulphate is the dominating extractable anion in shale materials from Kvarntorp (Karlsson, 2011). As is indicated by the EC in fig. 20; the weathered fines released most sulphate of the four materials (fig. 21). This was probable because the oxidation of pyrite has induced the formation of various sulphate bearing minerals such as gypsum (3.1.2, table 3).

The higher concentration of sulphate in aqueous samples from black processed shale compared to aqueous samples from red processed shale is most likely caused by the presence of gypsum in the black processed shale observed by XRD (3.1.2, table 6).

Fig. 21. Concentration of sulphate in equilibrium samples. Note logarithmic scale. Filtered (0.20 µm) samples. n=2 for day 1 and 26, n=1 for day 5.

Very low amounts of the materials total contents of sulphur were extracted (table 9). The 0.48 % sulphur released from shale corresponds quite well to extraction data reported by Lavergren (2008) and Lavergren et.al. (2009a) where unprocessed black shale released 0.74 % of its total sulphur content after an extraction procedure with water at L/S 10 according to EN 12457-3. Processed shale from Degerhamn released 9.6 % of its total sulphur content (Lavergren, 2008; Lavergren et.al., 2009a) which is significantly higher than the release of sulphur from red- and black processed shale from Kvarntorp (table 9).

(30)

Table 9. Sulphur extracted in % of the materials total sulphur content analysed by XRF (3.3)

Shale Weathered fines Red processed shale

Black processed shale Sulphur extracted

(%), day 26 0.48 1.2 0.10 2.0

3.6 Dissolved organic- and inorganic carbon

3.6.1 TOC analysis

Increases of DOC concentration in aqueous samples with pH 12.5 indicates a hydrolysis of the organic matter bound to the material (fig. 22-25). Why the aqueous samples of weathered fines (pH 12.5) had a higher concentration of DOC than aqueous samples from shale (pH 12.5) is unsure (fig. 22 and 23). The shale should contain more organic material than the weathered fines due to their different exposures to weathering. Possible explanations are that the weathering of the fines has altered the structure and properties of the organic material making it more susceptible to hydrolysis or it could be that biological matter such as leaves and roots have accumulated in the weathered fines deposit. The latter scenario is more probable.

The increased concentration of IC in aqueous samples (pH 12.5) was most likely a consequence of the addition of NaOH (fig. 22-25). Solutions of NaOH are known to accumulate carbonate from the dissolution of atmospheric carbon dioxide. The NaOH solution used in this study had been stored for a couple of months in a plastic bottle. The NaOH solution was not analysed in regard to its IC concentration.

Shale

Fig. 22. Concentration of DOC (left) and DIC (right) in aqueous samples from shale. Filtered (0.20 µm) samples. n=2 for day 1 and 26, n=1 for day 5.

(31)

Weathered fines

Fig. 23. Concentration of DOC (left) and DIC (right) in aqueous samples from weathered fines. Filtered (0.20 µm) samples. n=2 for day 1 and 26, n=1 for day 5.

The aqueous sample from red processed shale (pH 12.5) had the highest concentration of approximately 36 mg/L organic carbon on extraction day 26 compared to 22 mg/L in weathered fines (pH 12.5) and 26 mg/L in black processed shale (pH 12.5) (fig. 22-25). This was unexpected since the red processed shale has been pyrolysed and no remaining black hydrocarbons can be seen on the material.

The particle size used for the extraction tests was 2.0-0.5 mm. Observations during the sample preparation confirmed that particles of red processed shale had great adhesive capabilities. This resulted in that particles with the size 2.0-0.5 mm were coated with finer particles of red processed shale. This might have made the extraction properties of red processed shale different compared to the other three materials.

The pyrolysis process might not have completely removed all organic carbon from the red processed shale. Surfaces with organic carbon may have been exposed by the crushing during the preparation process. But the most probable explanation for the high concentration of organic carbon in aqueous samples from red processed shale was the presence of biological matter.

Red processed shale

Fig. 24. Concentration of DOC (left) and DIC (right) in aqueous samples from red processed shale. Filtered (0.20 µm) samples. n=2 for day 1 and 26, n=1 for day 5.

(32)

Black processed shale

Fig. 25. Concentration of DOC (left) and DIC (right) in aqueous samples from black processed shale. Filtered (0.20 µm) samples. n=2 for day 1 and 26, n=1 for day 5.

3.6.2 Spectrophotometry

Determination of DOC may be done by ultraviolet (UV) absorbance (Brandstetter et.al., 1996; Simonsson et.al., 2005; Wieshaar et.al., 2003). Absorbance for the aqueous samples from weathered fines, red- and black processed shale (pH 12.5, day 26) was measured since these had a yellow colour and relatively high concentrations of DOC; 22 mg/L, 36 mg/L and 26 mg/L respectively (fig. 22-25). No absorbance peaks were detected at wavelengths longer than 400 nm. The only absorbance peaks observed were at 254 nm and 340 nm, the latter only in samples from red- and black processed shale (fig. 26-28). Absorbance at 254 nm is a general absorbance for all organic carbon compounds (Brandstetter et.al., 1996; Wieshaar et.al., 2003).

Aromaticity of DOC may be estimated by the specific UV absorbance at 254 nm, SUVA254

(Wieshaar et.al., 2003). The calculated SUVA254 for weathered fines, red- and black processed

shale were 0.038, 0.059 and 0.049 L mg-1 m-1 respectively.

Common inorganic interferences in the UV region is iron and nitrate (Wieshaar et.al., 2003). The observed yellow colour of the samples may also have been caused by the presence of hydrous ferric oxide colloids such as FeOOH. The samples contents of iron will be further discussed under heading 3.7.1.

The interference in wavelength region 190-240 nm observed in figures 26-28 was most likely caused by the PMMA cuvette. The same interference was observed in blanks of deionised water and in solutions of 1 % HNO3 and 0.05 M NaOH. A quartz cuvette should have been used instead

(33)

Fig. 26. Absorbance spectrum of wavelengths 190-400 nm for an aqueous sample of weathered fines adjusted to pH 12.5, day 26 (n=1).

Fig. 27. Absorbance spectrum of wavelengths 190-400 nm for an aqueous sample of red processed shale adjusted to pH 12.5, day 26 (n=1).

(34)

Fig. 28. Absorbance spectrum of wavelengths 190-400 nm for an aqueous sample of black processed shale adjusted to pH 12.5, day 26 (n=1).

3.7 Metals in aqueous samples and their implications

Outputs from the Visual MINTEQ (v. 3.0, compiled by Gustafsson J. P., KTH, Dept. of Land and Water Resources Engineering, Stockholm, Sweden) modelling of aqueous samples with pH 12.5 indicate a minimum impact of the DIC concentrations on the metals. Formation of anionic carbonate complexes and precipitation of carbonates were therefore negligible and will not be considered further.

3.7.1 The hydroxide and oxyhydroxide forming metals: Al, Mn and Fe

Ions of Al, Mn and Fe hydrolyse in water according to reactions 16-18 (3.4.1) forming hydroxide complexes (Cotton et.al., 1995) which may precipitate as hydroxide or oxyhydroxide phases (Jönsson et.al., 2006; España et.al., 2005; Arranz González et.al., 2011; Maia et.al., 2012; Drever, 1997). Iron hydroxide is formed according to reaction 19 (Jönsson et.al., 2006; Maia et.al., 2012; Puura and Neretnieks, 2000) and aluminium according to a similar reaction 20 (Drever, 1997):

Fe3+ + 3 H2O → Fe(OH)3(s) + 3 H+ (19)

Al3+ + 3 H2O → Al(OH)3(s) + 3 H+ (20)

In acidic conditions with high sulphate concentration Fe and Al precipitates as hydroxysulphates or oxyhydroxysulphates such as schwertmannite Fe8O8(OH)6SO4, K-jarosite KFe3(SO4)2(OH)6,

jurbanite Al(SO4)(OH), and alunite K(Al3(SO4)2(OH)6) (Jönsson et.al., 2006; Arranz González

et.al., 2011; Maia et.al., 2012; Puura, 1998, Drever, 1997). Jarosite precipitates at pH 1.5-3.0 (Puura, 1998), schwertmannite precipitates at pH 2-4 and alunite at pH 3.5-4 (referred to by Sartz,

(35)

Arranz González et.al., 2011; Maia et.al., 2012; España et.al., 2005). The phases adsorption capabilities for cations are however low in acidic conditions due to the net positive charge of their surfaces (Drever, 1997; Stumm and Morgan, 1996; Maia et.al., 2012). Most iron hydroxides and oxyhydroxides have a pHpzc around 8 (Stumm and Morgan, 1996; Stumm, 1992; Maia et.al., 2012).

Aluminium had highest solid solution distribution in samples with pH 3.0 and equilibrated samples of shale (pH 3.0-3.8) and weathered fines (pH 2.8-3.1), (table 10 and fig. 29). The aluminium solid solution distribution decreased when pH was adjusted to 5.5 and 8.5, indicating precipitation of Al(OH)3 or AlOOH. As pH was increased further to 12.5 dissolved aluminium

increased due to the formation of Al(OH)4- (Cotton et.al., 1995).

The content of manganese in the four materials ranged between approximately 10-120 mg/kg according to XRF and digestion data (3.2, fig. 14-17). Sequential leaching data of shale materials from Degerhamn indicate concentrations ranging between 67-293 mg/kg (Lavergren et.al., 2009a). Manganese contents in a world average shale is 850 mg/kg (referred to by Lavergren et.al., 2009a), i.e. the manganese contents of these shale materials from Kvarntorp were low compared to other shales.

The aqueous samples with low pH contained higher concentrations of manganese than those with high pH (fig. 30). High pH most likely induced the precipitation of manganese hydroxide which was confirmed by Visual MINTEQ modelling.

Highest solid solution distribution of iron was found under acidic conditions from shale, weathered fines and black processed shale. The formation of iron hydroxides and oxyhydroxides was most probable the reason for the lower concentrations of iron in samples with pH >5.5 (table 12 and fig. 31).

An exception was the sample from red processed shale which had been adjusted to pH 12.5. Since the samples were not filtered before analysis by ICP-MS, the higher concentrations of iron in these samples may be caused by the presence of suspended colloids of hydrous ferric oxides. This sampled had, which was mentioned under heading 3.6.2, a yellow colour which may indicate the presence of colloidal FeOOH.

The relatively low distributions of Al, Mn and Fe in the pH range 5.5-8.5 indicate precipitation of hydroxides and oxyhydroxides with these elements. Due to their scavenging capabilities it should be expected that concentrations of other elements tend to be lower in this pH range. This will be further mentioned under the following headings of 3.7.

(36)

Fig. 29. Concentration of Al in aqueous samples from weathered fines (left) and black processed shale (right). Note logarithmic scale. (n=2 for day 1 and 26, n=1 for day 5).

(37)

Fig. 31. Concentration of Fe in aqueous samples from weathered fines (upper left), red processed shale (upper right) and black processed shale (bottom). (n=2 for day 1 and 26, n=1 for day 5).

Table 10. Percentages extracted of the materials total contents of aluminium (heading 3.2)

% Al extracted

pH Shale Weathered fines Red processed

shale Black processed shale nat. 0.61 0.42 0.14 0.01 3.0 0.65 0.42 3.7 0.08 5.5 0.14 0.02 0.08 0.01 8.5 0.01 0.01 0.08 0.01 12.5 0.64 0.02 2.0 0.14

(38)

Table 11. Percentages extracted of the materials total contents of manganese (heading 3.2)

% Mn extracted

shale Black processed shale nat. 3.3 0.95 0.12 5.0 3.0 3.2 0.95 5.1 15 5.5 2.7 0.62 0.65 3.8 8.5 0.92 0.01 0.13 0.76 12.5 0.04 0.02 0.22 0.13

Table 12. Percentages extracted of the materials total contents of iron (heading 3.2)

% Fe extracted

shale Black processed shale nat. 0.35 0.24 0.03 <0.01 3.0 1.1 0.23 0.02 0.02 5.5 0.04 0.01 0.03 <0.01 8.5 0.01 0.01 0.04 <0.01 12.5 0.02 0.03 0.21 0.01

3.7.2 The sulphate forming metals: Ca, Sr, Ba and Pb

Calcium, Sr and Ba forms precipitates with sulphate and their solubility decreases by the increasing atomic number (Cotton et.al., 1995), Ksp for CaSO4 is 4.93x10-5, 3.44x10-7 for SrSO4 and 1.08x10 -10

for BaSO4 (Generalic et.al., 2012). Lead sulphate has a Ksp 2.53x10-8 (Generalic et. al., 2012) and

according to decreasing lead concentrations presented in fig. 35 lead was either precipitating or was adsorbed over time. Since the distribution of sulphate were increasing (fig. 21) over time it is a high probability that lead precipitated as PbSO4.

Low distribution of lead were also achieved from non-weathered black shale, 2 µg/kg Pb, and processed shale, 3.5 µg/kg Pb, from Degerhamn with water at L/S 10 (EN 12457-3) (Lavergren et.al., 2009a). It could be that these low concentrations of lead were caused by the relatively high concentrations of dissolved sulphur (no speciation presented) found by Lavergren et.al. (2009a) after water extraction at L/S 10 (EN 12457-3) of non-weathered black shale (204 mg/kg Pb) and processed shale (1360 mg/kg Pb). Their sequential extraction data (modified from Hall et.al., 1996a, b and Land et.al., 1999) indicated that a large portion of lead was adsorbed to the shale materials, or incorporated into amorphous iron oxides. In the case of processed shale lead also seem

(39)

indicated by the increased calcium distribution in samples with pH <5.5 (fig. 32). The high release from weathered fines is probably a result of the presence of gypsum identified by XRD (3.1.2). The weathered fines higher contents of calcium, compared to the other three materials, were also determined by XRF and digestion followed by ICP-MS analysis (3.2, fig. 15).

High concentrations of strontium (fig. 33) and sulphate (fig. 21) might indicate the presence of SrSO4 in samples of the weathered fines. An unknown decrease in strontium concentration can be

observed in the samples adjusted to 8.5. Output from modelling in Visual MINTEQ did not give any explanation to this phenomenon.

The solid solution distribution from shale, red- and black processed shale all indicate that strontium minerals are more soluble at low pH, indicating precipitation or adsorption of strontium species at high pH.

The leachability of barium seem to be unaffected by pH (fig. 34).

Fig. 32. Concentration of Ca in aqueous samples from shale (upper left), weathered fines (upper right), red processed shale (bottom left) and black processed shale (bottom right). (n=2 for day 1 and 26, n=1 for day 5).

(40)

Fig. 33. Concentration of Sr in aqueous samples from shale (upper left), weathered fines (upper right), red processed shale (bottom left) and black processed shale (bottom right). (n=2 for day 1 and 26, n=1 for day 5).

(41)

Fig. 35. Concentration of Pb in aqueous samples from shale (upper left), weathered fines (upper right), red processed shale (bottom left) and black processed shale (bottom right). (n=2 for day 1 and 26, n=1 for day 5).

Relatively high percentages of calcium were extractable in the extraction experiments (pH <5.5) compared to the materials total contents (table 13). Unprocessed and non-weathered shale from Degerhamn released 1.2 % of its total calcium content while the processed shale released 37 % (Lavergren, 2008; Lavergren et.al., 2009a) with water at L/S 10 (EN 12457-3).

The highest amounts of strontium were extracted from shale and weathered fines, especially at low pH, indicating an extraction pattern similar to that of calcium (table 14).

Relatively small amounts of barium were extracted under these extraction conditions. The percentages were almost the same for all four shale materials (table 15) indicating that barium most likely was present in the same mineral form in all four materials. The mineral form was most probably BaSO4 due to its low solubility.

The distribution of lead on extraction day 26 equalled very small released fractions of the shale materials total lead contents (table 16). Due to the distribution behaviour of lead (fig. 35) one can assume that the concentrations of lead on day 26 do not represent the total lead that have been released from the materials. Larger fractions of lead have actually been extracted from the material but been reincorporated to the material by either adsorption or precipitation.

(42)

Table 13. Percentages extracted of the materials total contents of calcium (heading 3.2)

% Ca extracted

shale Black processed shale nat. 23 20 2.7 15 3.0 23 21 10 32 5.5 17 20 3.9 12 8.5 10 19 0.96 5.4 12.5 0.11 16 0.37 0.16

Table 14. Percentages extracted of the materials total contents of strontium (heading 3.2)

% Sr extracted

Table 15. Percentages extracted of the materials total contents of barium (heading 3.2)

% Ba extracted

(43)

Table 16. Percentages extracted of the materials total contents of lead (heading 3.2)

% Pb extracted

shale Black processed shale nat. 0.24 0.02 0.06 <0.01 3.0 0.26 0.01 0.02 <0.01 5.5 0.07 <0.01 0.05 <0.01 8.5 0.02 0.03 0.07 <0.01 12.5 0.02 0.02 0.27 0.02

3.7.3 Metals with the highest solid solution distribution at low pH: Mg, Co, Ni, Cu, Zn and Cd Characteristics for the elements Mg, Co, Ni, Cu, Zn and Cd are that these for the most part form cationic species in acidic conditions and precipitates mainly as hydroxides or form anionic hydroxide complexes in alkaline conditions. Therefore adsorption of these elements was expected to be low in acidic conditions and the concentrations of dissolved species low at high pH. Copper, Zn and Cd may also form complexes with DOC (Drever, 1997).

Magnesium is capable of forming both ionic- and non-ionic compounds and binds often to carbon compounds (Cotton et.al., 1995). This could explain the higher concentrations of magnesium detected by XRF in shale, weathered fines and black processed shale compared to red processed shale (3.2, fig. 14-17).

Low pH increases the distribution of magnesium from all four materials (fig. 36). The concentrations of magnesium were slightly lower in aqueous samples with pH 8.5 and drastically lower at pH 12.5. The high pH most likely induced the precipitation of Mg(OH)2(s) (Cotton et.al.,

1995).

Cobalt was only detectable in aqueous samples from shale, weathered fines and black processed shale with pH <5.5 (fig. 37). Concentrations of cobalt decreases in aqueous samples with pH 8.5 and 12.5. Cobalt precipitates most likely as Co(OH)2 in the presence of OH- (Cotton et.al., 1995).

The distribution of nickel and zinc increases at low pH and decreases as pH is increased. The nickel distribution in acidic water phases corresponds to findings by Lavergren et.al. (2009b) and SWECO (2005a). At high pH nickel and zinc precipitates as Ni(OH)2 and Zn(OH)2 respectively

(Rötting et.al., 2006; Navarro et.al., 2006).

The lower solid solution distribution of nickel from red processed shale compared to from shale and black processed shale might be due to the grade of pyrolysis. If the pyrolysis was done in an excess of air nickel might have formed NiO(s) (Cotton et.al., 1995). The most probable forms of nickel in non-pyrolysed materials should be as nickel sulphides due to the reducing conditions during the lithification of the shale. Nickel sulphides are soluble under oxidising conditions, while its oxide is soluble in acids, but insoluble in neutral solutions (Lancashire and Swaby, 2010). This could explain the higher distribution at pH 3.0 from red- and black processed shale (fig. 38).

It was a large difference in nickel distribution from shale (pH 3.0-5.5) and from weathered fines (pH 3.0-5.5), fig. 38. This was either a result of heterogeneity or that the weathered fines already have released a large portion of its nickel contents during its 50-60 years in the deposit.

(44)

Higher distribution of copper was found under acidic conditions from shale and weathered fines than under alkaline conditions (fig. 39). According to Drever (1997) and Stumm and Morgan (1996) copper should not be adsorbed to any hydrous ferric oxide surfaces at pH < 4.

The highest copper distribution from red- and black processed shale was found to be under alkaline conditions (fig. 39). According to the output from Visual MINTEQ modelling; the dominating dissolved copper species in aqueous samples from all four materials with pH 12.5 were Cu(OH)2(aq), Cu(OH)3- and Cu(OH)42-. Data from aqueous samples from shale and weathered fines

(pH 12.5) indicate however no formation of negatively charged copper species (fig. 39). The modelled systems at pH 12.5 indicate an undersaturation for Cu(OH)2(s) but since modelling

softwares are not completely accurate it could be possible that copper has formed Cu(OH)2(s) in the

aqueous phases from shale and weathered fines (pH 12.5).

Another reason for the low distribution of copper from shale and weathered fines (pH 12.5) could be that copper was bound to organic matter. Due to possible differences of the organic compounds present in the four materials, metals may have different tendencies to bind to them. According to DOC data (3.6.1), the lowest concentrations of DOC in samples with pH 12.5 were found in samples of shale and weathered fines (6.7mg/L and 22 mg/L respectively). The lower rate of hydrolysis of the organic matter in the shale sample indicates higher amounts of organic compounds still adsorbed to particles. If copper have bound to the organic matter, it might not have dissolved as pH was increased.

The only detectable concentration of cadmium was in the aqueous sample from black processed shale with pH 3.0 (fig. 41). The contents of cadmium detected by XRF were minimal (See appendix 1).

The decreased concentrations of Mg, Co, Ni, Cu, Zn and Cd at pH 5.5 compared to pH 3.0 might be a result of coprecipitation with hydrous oxide phases of iron, aluminium and manganese (fig. 36-41).

(45)

Fig. 36. Concentration of Mg in aqueous samples from shale (upper left), weathered fines (upper right), red processed shale (bottom left) and black processed shale (bottom right). (n=2 for day 1 and 26, n=1 for day 5).

Fig. 37. Concentration of Co in aqueous samples from shale (upper left), weathered fines (upper right) and black processed shale (bottom). (n=2 for day 1 and 26, n=1 for day 5).

(46)

Fig. 38. Concentration of Ni in aqueous samples from shale (upper left), weathered fines (upper right), red processed shale (bottom left) and black processed shale (bottom right). (n=2 for day 1 and 26, n=1 for day 5).

(47)

Fig. 40. Concentration of Zn in aqueous samples from shale (upper left), weathered fines (upper right), red processed shale (bottom left) and black processed shale (bottom right) (n=2 for day 1 and 26, n=1 for day 5).

Fig. 41 Concentration of Cd in aqueous samples from black processed shale (n=2 for day 1 and 26, n=1 for day 5).

Relatively large percentages of cobalt are extracted from the shales total content of cobalt. The lower concentrations and percentages extracted from weathered fines may be a result of previous release or heterogeneity of the materials (table 18).

The shale releases considerable amounts of nickel under acidic conditions, 14 % of its total content (table 19). The non-weathered black shale from Degerhamn released only 0.05 % of its total nickel content while the weathered shale from the same region released 14 % (Lavergren, 2008; Lavergren et.al., 2009a).

Only a very small amount of the materials copper contents were accessible by the extraction processes (table 20). Copper in shale materials from Degerhamn was also very resistant to extraction (Lavergren, 2008; Lavergren et.al., 2009a).

(48)

Highest percentage of zinc release, 1.0 %, was that of shale with pH around 3. Zinc was rather unaffected by the extraction processes (table 21). The release of zinc from shale materials from Degerhamn was also low except from weathered black shale where 10 % was released (Lavergren, 2008; Lavergren et.al., 2009a).

Table 17. Percentages extracted of the materials total contents of magnesium (heading 3.2)

% Mg extracted

(49)

Table 18. Percentages extracted of the materials total contents of cobalt (heading 3.2)

% Co extracted

pH Shale Weathered fines Black processed

shale

nat. 17 0.76 0.60

3.0 18 0.72 4.4

5.5 13 0.44 0.38

Table 19. Percentages extracted of the materials total contents of nickel (heading 3.2)

% Ni extracted

shale Black processed shale nat. 14 0.49 0.03 1.0 3.0 14 0.50 0.13 2.8 5.5 9.6 0.37 0.03 0.45 8.5 1.3 0.03 0.06 0.05 12.5 0.12 0.06 0.04 0.08

Table 20. Percentages extracted of the materials total contents of copper (heading 3.2)

% Cu extracted

(50)

Table 21. Percentages extracted of the materials total contents of zinc (heading 3.2)

% Zn extracted

3.7.4 The Mg/Ca- and Na/K ratios

The aqueous sample from the weathered fines (equilibrated pH) had the highest Na/K ratio of 4.56 (table 22). When the clay mineral illite, K0.6Mg0.25Al2.3Si3.5O10(OH)2, is exposed to acidic

conditions it will dissolve into smectite while releasing potassium ions according to following reaction (Puura, 1998):

5/3 Illite + H+ → 5/3 Smectite + K+ (21)

As was mentioned under heading 3.7.1; precipitation of K-jarosite may occur in solutions with a pH between 1.5-3.0 and with high concentrations of iron and sulphate (Jönsson et.al., 2006; Arranz González et.al., 2011; Maia et.al., 2012; Puura, 1998, Drever, 1997). Since the aqueous sample from weathered fines with equilibrated pH contained 8800 µg/L iron, 1600 mg/L sulphate and had pH 2.8 on extraction day 26, K-jarosite most likely precipitated. Thus the concentration of dissolved potassium decreases and the Na/K increases.

The ratio between Mg and Ca (table 23) gives an indication of the distribution of magnesium in the limestone that may be incorporated in the materials (Puura, 1998).

Release of elements from unprocessed- and processed shale from Kvarntorp : Effect from different pH

Release of metals from unprocessed- and processed

shale from Kvarntorp

As a function of solution pH

Abstract

Table of contents

1 Introduction

2 Materials and methods

3 Results and discussion