FI9900129

POSIVA 99-2 3

New data on the Hyrkkola U-Cu mineralization: The behaviour of native copper in a natural environment

Nuria Marcos Helsinki University of Technology Laboratory of Engineering Geology and Geophysics

Lasse Ahonen Geological Survey of Finland

30-4 / May 1 999

POSIVA OY

Mikonkatu 15 A, FIN-OO1OO HELSINKI, FINLAND

Phone (09) 2280 30 (nat.), ( + 358-9-) 2280 30 (int.)

Fax (09) 2280 3719 (nat.), ( + 358-9-) 2280 3719 (int.) ISBN 951 -652-078-2 ISSN 1239-3096

The conclusions and viewpoints presented in the report are those of author(s) and do not necessarily coincide with those of Posiva. - POSiVa Report Raportintunnus- Report code POSIVA 99-23 POSivaOy Julkaisuaika Date Mikonkatu 15 A, FIN-00100 HELSINKI, FINLAND Juikaisuaika- Date Puh. (09) 2280 30 - Int. Tel. +358 9 2280 30 May 1999

Tekija(t) - Author(s) Toimeksiantaja(t) - Commissioned by Nuria Marcos, Helsinki University of Technology D . ~ LasseAhonen, PosivaOy Geological Survey of Finland Nimeke-Title

NEW DATA ON THE HYRKKOLA U-Cu MINERALIZATION: THE BEHAVIOUR OF NATIVE COPPER IN A NATURAL ENVIRONMENT

Tiivistelma - Abstract

The Hyrkkola Cu-U mineralization (SW Finland) is studied as an analogue to the behaviour of copper canister in crystalline bedrock. Uranium-native copper and uranium-copper corrosion products interactions are also addressed in this study. The integration of uranium series disequilibrium (USD) studies gives an estimate of the time-scales of the corrosion processes.

The mineral assemblages native copper - copper sulfide, copper sulfides - copper iron sulfides, and native copper - copper oxide (cuprite) occur in open fractures at several depth intervals within granite pegmatites (GP). The surfaces of these open fractures have accumulations of uranophane crystals and other unidentified uranyl compounds. The secondary uranium minerals are mainly distributed around copper sulfide grains. Microscopic intergrowths of copper sulfides and uranyl compounds also have been observed. The surface of the fracture where native copper and cuprite occur is covered with uranium-rich smectite. The very low 234U/238U activity ratio (0.29 - 0.39) in the main uranium fraction in smectite indicates chemical stable conditions (e.g., oxidising) during at 234 5 least a time period comparable to the half-life of the U isotope (T1/2= 2.44 * 10 a).

Groundwater samples were collected from intervals where copper minerals occur within open fractures. The Eh and pH conditions were measured during long-term pumping (2-4 weeks per sample). Eh was measured both in situ and an the surface using three electrodes (Pt, Au, C). The actual groundwater conditions are oxidising and would not allow the sulfidization of native copper. Sulfidization may be considered as an old phenomenon, older than the precipitation of uranyl phases in the samples. The end of sulfidization may be earlier than the precipitation and/or remobilisation of U(VI) phases in a time span from about 2 * 105 years (precipitation of uranophane) to 2.44 * 105 (remobilisation of U from smectite).

Avainsanat - Keywords native copper, copper canister, corrosion, smectite, uranophane, redox, Hyrkkola ISBN ISSN ISBN 951-652-078-2 ISSN 1239-3096

Sivumaara - Number of pages Kieli - Language 78 English — Posivfl RPDOft Raportin tunnus-Report code POSIVA 99-23

Mikonkatu 15 A, FIN-00100 HELSINKI, FINLAND Julkaisuaika - Date Puh. (09) 2280 30 - Int. Tel. +358 9 2280 30 Toukokuu 1999

Tekijä(t) - Author(s) Toimeksiantaja(t) - Commissioned by Nuria Marcos, Teknillinen korkeakoulu Lasse Ahonen, Posiva Oy Geologian tutkimuskeskus

Nimeke - Title

HYRKKÖLAN U-Cu ANALOGIA: METALLISEN KUPARIN KÄYTTÄYTYMINEN LUONNOLLISESSA YMPÄRISTÖSSÄ

Tiivistelmä -Abstract

Hyrkkölän Cu-U mineralisaatiota on tutkittu analogiana kuparikanisterin käyttäytymiselle kiteisessä kallioperässä. Uraanin ja metallisen kuparin sekä uraanin ja kuparin korroosiotuotteiden vuoro- vaikutuksia on myös käsitelty tässä työssä. Uraanisarjan epätasapainotutkimusten (USD) avulla arvioitiin korroosioprosessien aikaskaalaa.

Mineraaliseurueet luonnonkupari-kuparisulfidi, kuparisulfidi-kuparirautasulfidit ja luonnonkupari- kuparioksidi (kupriitti) esiintyvät avoraoissa useilla syvyyksillä graniittipegmatiiteissa (GP). Avorakojen pinnoilla on uranofaanikiteitä ja muita, tunnistamattomia uranyyliyhdisteitä. Sekundää- risiä uraanimineraaleja on pääasiallisesti kuparisulfidirakeiden ympärillä. Kuparisulfidien ja uranyyliyhdisteiden mikroskooppisia yhteenkasvettumia on myös havaittu. Metallista kuparia ja kuparioksidia sisältävää rakopintaa peittää uraanirikas smektiitti. Smektiitin uraanifraktion isotooppi- suhde 234U/238U on hyvin alhainen (0,29 - 0,39). Tämä viittaa siihen, että kemialliset olosuhteet tutkittujen rakomineraalien ympäristössä ovat pysyneet muuttumattomina (hapettavina); tähän 234 5 vaadittava aika on vähintään U-isotoopin puoliintumisajan suurusluokka (T]/2= 2.44 * 10 a).

Kuparia sisältävistä raoista otettiin vesinäytteet. Veden Eh- ja pH-arvoja mitattiin pitkäaikaisten näyttenottopumppausten aikana (2-4 viikkoa/vesinäyte), redox-potentiaali niitattiin sekä in s itu- tilassa että maan pinnalla kolmea eri elektrodia käyttäen (Pt, Au, C). Nykyiset pohjavesiolosuhteet ovat hapettavat, jollaisissa metallisen kuparin sulfidoituminen ei ole mahdollista. Näytteissä havaittu kuparisulfidin muodostus on tapahtunut aikaisemmin kuin uranyylifaasien saostuminen. Sulfidoi- tuminen on päättynyt ennen U(VI)-faasien saostumista/uudelleenmobiloitumista, joka ajoittuu noin 2 * 105 vuoden (uranofaanin saostuminen) ja 2.44 * 105 vuoden (smektiitin U:n remobilisaatio) aikavälille.

Avainsanat - Keywords metallinen kupari, kuparikanisteri, korroosio, smektiitti, uranofaani, redox, Hyrkkölä ISBN ISSN ISBN 951-652-078-2 ISSN 1239-3096

Sivumäärä - Number of pages Kieli - Language 78 Englanti TABLE OF CONTENTS

Page

Abstract 3

Tiivistelma 5

Preface 9

1 INTRODUCTION AND OBJECTIVES 11

2 GEOLOGY AND DESCRIPTION OF SITE STUDIES 13 2.1 Geological setting 13 2.2 Site studies 16 2.3 Mineralogy and geochemistry 17

3 POROSITY, OCCURRENCES OF COPPER MINERALS AND ASSOCIATED FRACTURE MINERALOGY 19 3.1 Analytical methods and Mineral composition 22 3.1.1 Native copper 22 3.1.2 Copper sulfides associated with native copper 24 3.1.3 Copper sulfides and copper-iron sulfides 26 3.1.4 Occurrences of uranyl compounds 30 3.1.5 Occurrence of copper oxide 35 3.2 Paragenesis 36 3.2.1 Source of sulfur 37

4 HYDROCHEMISTRY 39 4.1 Groundwater sampling, determinations, quality and representativeness 39 4.2 Results 42 4.2.1 Parameters monitored during pumping 42 4.2.2 Chemical characteristics of the groundwater samples 48

5 URANIUM SERIES ISOTOPE STUDIES OF ROCKS AND MINERALS 53 5.1 Rock/mineral samples and methods 54 5.2 Radiochemical data 54

6 SUMMARY AND DISCUSSION 55 7 CONCLUSIONS 59

ACKNOWLEDGEMENTS 60

REFERENCES 61

APPENDICES 65

APPENDIX A: Copper, uranium and trace element analysis of smectite (Sample Hy324/97.85) 67

APPENDIX B: Sequential extraction (SE) methods for smectite (Sample Hy324/97.85) 73

APPENDIX C: Preliminary results of research in the Interface Analysis Centre (University of Bristol, UK): implications for mineral-water interactions and for interpretation of the results obtained by sequential extraction (SE) methods 75 PREFACE

This report presents the results of the Hyrkkola' native copper analogue studies (1997 -1998), a collaborative work between Posiva Oy (Finland) and the Swedish Nuclear Fuel and Waste Management Company (SKB). The contact persons were Margit Snellman at Posiva Oy and Lars Werme at SKB. 11

1 INTRODUCTION AND OBJECTIVES

Studies in laboratory conditions and thermodynamic considerations have been applied to address the chemical stability of metallic copper as a canister material for spent nuclear fuel (e.g., Engman and Hermansson 1994, Ahonen 1995, and references therein). These studies describe well the behaviour of copper in simplified systems, while natural analogues are most suitable in assessing the long-term behaviour of copper in complex natural systems. The durability of copper also has been studied by means of its archaeological analogues, mainly bronzes, which have been preserved hundreds or thousand of years in near-surface environments (e.g. Hallberg et al. 1988, Miller et al. 1994, King 1995).

Among the possible corrosion-resistant canister materials, metallic copper is the only one for which lifetime up to a billion years also can be indicated by means of natural analogues. The basalt-conglomerate-hosted native copper deposits at the Keweenaw Peninsula in Michigan were formed about 1.1 Ga ago, and metallic copper was the stable copper phase during the hydrothermal alteration of the deposit at a temperature of 100° to 200° C. According to Schwartz (1996), 'There is general agreement that the bulk of native copper has remained texturally stable since the Precambrian metamorphic-hydrothermal event that produced the mineralization. Neither sulfidization nor oxidation of native copper is of any importance".

Many authors (e.g., Schwartz 1996, Amcoff 1998) have pointed out that the geological association of native copper deposits corresponds mainly to basalts and to the surpergene weathering of copper sulfide deposits and that the hydrogeochemical environment of those native copper occurrences differs from that expected in the planned deep nuclear waste repository. Amcoff (1998 p. 24) proposed that (op. cit) perhaps the common occurrences of chalcopyrite and other high temperature sulfide phases in glacial boulders constitute more interesting analogues, perhaps, to the behaviour of Cu corrosion products in a granitic environment. Detailed studies on the behaviour of copper and its corrosion products in crystalline bedrock and in contact with groundwater are lacking to date.

The Hyrkkola Cu-U mineralization occurs within the Svecofennian schist belt (crystalline bedrock). Its mineralogical and geological characteristics (Marcos 1996) are similar to those of the sites considered for the actual disposal of nuclear waste in Finland and Sweden. Two new research boreholes were drilled and cored at the site during 1997 in order to get groundwater samples in direct contact with metallic copper or in contact with drill-core samples containing metallic copper (Ahonen & Marcos 1997). A multidisciplinary study of those new samples provides us insights into the long-term behaviour of native copper. Moreover, the association of native copper with copper sulfides and copper oxides in some of the samples as well as the existence of uranium in mineralogical and groundwater samples is a complex natural system, the study of which may assist in assessing the long-term behaviour of the materials/elements involved in the near-field of a nuclear waste repository. 12

The study of the behaviour of native copper in a natural (bedrock-groundwater) system similar to that of potential nuclear waste-disposal sites in Finland and Sweden is our main objective. The occurrence of uranium in the system also allows us to apply U-series isotope studies with aims 1) to determine whether or not native copper and related minerals have been in contact with groundwaters and 2) to elucidate the time-scale of groundwater - rock/mineral interactions. 2 GEOLOGY AND DESCRIPTION OF SITE STUDIES

2.1 Geological setting

The Hyrkkola study site occurs within the Svecokarelian schist belt that extends from SE Finland to central Sweden (Figure 2-1). The bedrock of the area is composed of mica gneisses, quarz feldspar gneisses and amphibolites, which are penetrated by infracrustal rocks, mainly granites and granodiorites. Most of the quartz-feldspar gneisses have been interpreted as arkose sandstones (Simonen 1956). They are often closely associated with micaceous sediments and gradually grade into mica schists and mica gneisses. The amphibolites are mainly mafic-intermediate volcanics, being in many cases pyroclastic in origin. In the stratigraphy, the metasediments are overlain by the metavolcanics. Regional metamorphism took place under low-pressure amphibolite facies conditions (3 - 5 kbar, 550 - 650 °C; Latvalahti 1979, Schreurs & Westra 1986). 14

The lithology and main structures of the Hyrkkola site are shown in Figure 2-2. The geological structure of the study area is characterised by several post-folding faults and shear zones (Makela 1989). The most prominent of them is the Hirsjarvi or Painio shear zone (Makela 1989, Ploegsma 1989), which lies about 2 km north of Hyrkkola. It displays a left-handed movement with both horizontal and vertical components (Makela op.cit.).

The mineralization at Hyrkkola is related to the late stages of the Svecokarelian orogenic events about 1.8 - 1.7 Ga ago (Raisanen, 1986). The uranium-native copper occurrence is hosted by granite pegmatite veins. The mineralization was found in the 1980's by car-borne radiometric surveys, and originally studied by the Geological Survey of Finland (GTK) in connection with a uranium ore exploration campaign. Several reseach pits and trenches were excavated (Ahonen et al. 1997 and references therein), and five cored drillholes were drilled. (Figure 2-3). Native copper and uranium minerals were identified in the core samples (Marcos 1996).

^. + 5717 ( -1-6717 y~ * r 499 / \ / Vhhermanj&rvi Xv /I \ I V \ < \^ X X *<* / "£" J,...... - ...^ •/-•• 76 1 \ ••..._ -V\*

- >=^ • - X

\\

>^ • --Tib ~JJ \ si-j Hein&isienjarvi \ _ \ \ \ ~ -% 3D.. ( r^-Lj^ <

/ 83 ' -' •x "•"'({" •Ov-" --- X / _^ .v X go _— — y^2*s ^ --' V \? - --:( <<: \ / \ \ 1 *• 1 N -^ T 'V<

\ ^^*&>• " I \ \ \ - \ \ ~ X '•

/ x •^J- - - - -f-5/13 X+- 67! 3 sX;< / •a/ / LEGEND: INTERCALATIONS 1 [ HYRKKOLA STUDY SITE AMPHIBOLITE ' i VEINS AND INCLUSIONS LZH QUARTZ-FELDSPAR GNEISS FOLIATION, STRIKE AND DIP 0 1000 M MICA GNEISS -*- VERTICAL FOLIATION • SEOLOGICAL SURVEY OF FINLAND GRANODIORITE 'IT! ^^ FAULT OR FRACTURE /JONE 1997

PEGMATITE PZ PAINIO SHEAR ZONE MODIFIED AFTER YRJOLA (1984)

x OUTCROP Figure 2-2. Bedrock of the Hyrkkold area. 15

HYRKKOLA Lithology

Legend:

[ | Quartz-feldspar gneiss 0 © Garnet

| | Mica gneiss ^cfa Foliation, strike and dip

f [ Amphiboiite/amphibole gneiss ^x^ Vertical foliation •HI Pegmatite e Borehole Q— Research pit or trench

x Outcrop

compiled by: Seppo Paulam&ki Geological Survey of Finland 1997

Figure 2-3. Bedrock of the Hyrkkold study site. Horizontal projections of the old drillholes (R301 - R305) are indicated by arrows. 16

2.2 Site studies

Two new cored boreholes were drilled at the Hyrkkola' study site in 1997. Three dimensional geological modelling was used to optimize the location and direction of the drillholes in such a way that the known copper-bearing pegmatites would be reached and that the initiation point of the drillholes would be suitable for long-term water sampling and other measurements.

The locations and other characteristics of the drillholes are presented in Table 2-1. Drillhole and drillcore diameters are 76 mm and 62 mm, respectively. Hereafter, the drillholes (and cores) are referred as Hy324 and Hy325. After drilling, drill-hole inclination was measured in ten meters interval in both holes. The vertical deviation from the initial drill-hole inclinations given in the Table 2-1 was less than 0.9 degrees in all measurements.

Flushing water for drilling was transported from the near-by fire station. Uranine dye was mixed with the water before use, and uranine concentration of the return water was monitored during drilling. The amounts of injected and returned drilling water were monitored. After drilling, each borehole was pumped continuously for about three days to remove remnants of drilling water. At the end of the pumping, percentage of the flushing water was 1.3 % in Hy324 and 3 % in Hy325. It was estimated that less than 40 percent (20 m3 of 50 m3) of the flushing water remained in the bedrock (Ahonen and Marcos 1977).

Immediately after the drilling and subsequent pumping, both drillholes were video- recorded. The fracture-properties and rock type (with an emphasis on pegmatites) were then analyzed from the video-tape. According to the aperture, fractures were classified into four categories from probably sealed (Rl) to clearly open fractures showing apertures in centimeters (R4). The angle between fracture and drillhole was classified as approximately perpendicular or oblique (Ahonen and Marcos 1997).

The recovered drillcore was preliminary logged immediately in the field, with special reference to fractures in copper-bearing pegmatites. The original orientation of the core sample was traced whenever possible.

Table 2-1. Location, direction and length of the Hyrkkola drillholes

ID-code Collar Z (m above Sea Direction Initial Length coordinates 1), Ground/ Inclination (m) (km) Casing

Y52/2024/97/324 X=6714,874 Gr: 110,29 340° 74,8° 100,25 Y=2495,720 Cs:110,71

Y52/2024/97/325 X=6714,874 Gr: 110,20 310° 74,5° 104,10 Y=2495,720 Cs:110,58 17

2.3 Mineralogy and geochemistry

The amphibolites are dark-gray, homogeneous or banded, fine to medium-grained rocks. The main minerals of the amphibolites are hornblende, plagioclase (A1132-43), and minor diopside. Accessory minerals are titanite (up to 5%), quartz, opaques (magnetite, hematite, pyrite, pyrrhotite, graphite). Epidote, chlorite, and clay minerals are secondary. Hematite, epidote, and carbonates are typical fracture-filling minerals. The chemical analyses (X-ray fluorescence) of these rocks display a high content of magnesium oxide (MgO = 8.2 %), calcium oxide (CaO = 11.8 %) and ferric oxide (Fe2C>3 = 9.2 %). Copper (Cu) and sulfur (S) were not detected by this method, as it is corroborated by the mineralogical absence of copper sulfides in amphibolites. Iron sulfides occur highly disseminated in maximum amounts of about 1 %.

Quartz-feldspar gneisses are light to dark gray, homogeneous, and fine to medium- grained. The main minerals are quartz and plagioclase (An2o-3o); hornblende is a major accessory mineral where gneisses are in contact with amphibolites. Secondary minerals include epidote, chlorite, and hematite. Epidote and chlorite occur as fracture filling minerals.

The granite pegmatites occur as veins (0.05 - 0.5 m thick) parallel to the foliation in amphibolites and quartz-feldspar gneisses. They are heterogeneous, medium to coarse- grained, and red in colour. The granite pegmatites selected in our study occur mainly between amphibolites and amphibole-rich gneisses. The main minerals are quartz, feldspars (mostly microcline), albite, and black tourmaline. Accessory minerals are green apatite, titanite, zircon, graphite, uraninite, and native copper. Native copper occurs as thin flakes in and between feldspar grains and around tourmaline grains. Primary uraninite has been identified as disseminations in granite pegmatites. Epidote, hematite, cuprite, chalcocite, djurleite, calcite, uranophane, and gummite are secondary minerals. Uranophane and gummite are disseminated in the rock matrix, fill microfractures, and coat fracture surfaces.

According to the visual examination, one of the most interesting occurrences of metallic copper was that in Hy325, at depth of 68.3 m, where an approximately 1.5 m thick copper-bearing pegmatite was intersected. Broken rock material in the pegmatite contained native copper in abundance (Figures 2-4 and 2-5). The borehole TV-survey show that there were fissures in the bedrock at that depth. 18

T9SHT-"

Figure 2-4. Overview of drillcore Hy325,from about 64.5 m to 71.5 m.

Figure 2-5. Copper-bearing crushed rock at 68.3 m in drillhole Hy325. 19

3 POROSITY, OCCURRENCES OF COPPER MINERALS AND ASSOCIATED FRACTURE MINERALOGY

In order to evaluate the stability of native copper and its corrosion products in a fractured, water-saturated rock, the main fractures/fracture zones intersecting Cu-bearing pegmatites were thoroughly examined. Rock samples related to open fractures and fractured zones were selected (Table 3-1) for further mineralogical and isotopic studies. Copper minerals include native copper, copper sulfides, copper-iron sulfides, and copper oxides.

Table 3-1. Situation and core length of the selected samples.

Sample Rock Copper minerals Fracture surface minerals Hy324/97.85 m Granite pegmatite Native copper, Smectite cuprite Hy324/53.3 m Granite pegmatite Monoclinic chalcocite, 8-Uranophane, bornite, chalcopyrite, Calcite covellite Hy325/68.3 m Granite pegmatite Native copper, monoclinic Fractured zone, chalcocite,djurleite, crushed material digenite Hy325/68m Granite pegmatite Native copper fills Calcite, microfissures 6-Uranophane Porosity sample Granite pegmatite Native copper - Hy325/78.70

The occurrence of uranyl compounds and smectite associated with native copper and its corrosion products (sulfides, oxides) allows us to study the behaviour of the possible repository near-field materials in a natural environment, which may be similar to that of the spent nuclear fuel repository.

The porosity of a representative sample (Hy325/78.70) where native copper is not within a fracture zone or fracture surfaces in direct contact with groundwater was measured by two methods in order to extrapolate the effect/influence of the groundwater chemistry on the composition of native copper.

method

The method employs impregnation of vacuum dried rock samples by 3H-PMMA, radiation polymerisation, autoradiography and digital image processing techniques. Migration pathways can be identified and visualised and porosity histograms of the matrix (2D images) can be obtained (e.g. Hellmuth et al. 1993, 1994). The porosity value of the surface where native copper occurs within feldspar (Figures 3-la and 3-lb) was about 0.25 %. The porosity value of surfaces where native copper is filling cracks was about 0.63 %. The total porosity of the surface in Figure 3-1 was calculated to be of 0.41 %. 20

Figure 3-la. Sample Hy325/78.70. Section of the drill core. 0 = 6.8 cm. 21

Figure 3-lb. Sample Hy325/78.70. Autoradiography of the same section. 0 = 6.8 cm. 22

Water immersion technique

This method determines the connected pore volume by measuring the increase in weight when sample is saturated with water. The total porosity value calculated by this method was 0.46 ± 0.02%.

3.1 Analytical Methods and Mineral Composition

Fracture surfaces and polished sections were examined using a stereomicroscope and a polarising microscope. The identification of minerals coating fracture surfaces was done by the XRD (X-ray diffraction) method. Quantitative analyses of native copper and related minerals were obtained using an electronprobe microanalyser (EPMA) CAMECA SX 50 at the Geological Survey of Finland. The minimum size of the particles to be analysed is 0 = 5 um. In this study the detection limits of quantitative analyses for copper was 0.18 weight % (wt.%), 0.05 wt.% for sulfur, and 0.19 wt.% for uranium. The detection limits for the other elements analysed may be considered 0.2 - 0.5 weight % depending on the element, current and voltage (Johanson & Kojonen, 1995). The composition of smectite with respect to trace element content was analysed by ICP-MS and ICP-AES (Appendix A). Sulfur isotope determinations (Section 3.2) were performed at the Laboratory of Engineering Geology and Geophysics, Helsinki University of Technology, using the method described by Makela and Tammenmaa (1977).

3.1.1 Native copper

The occurrence of native copper is primarily associated with the crystallization of the feldspars low temperature albite (NaAlSi3Og) and microcline (KAlSi3O8). Copper occurs dispersed within albite and K-feldspar in excess of about 0.5 wt.%; on cooling, copper is exsolved along cleavage planes of K-feldspar and albite (Fig. 3-2). Native copper also fills microfractures and voids. Table 3-2 shows the data of microprobe analyses through the profile marked in Figure 3-3. Analyses of native copper in sample Hy325/68 gave a maximum Cu content of 98.85 weight.%. The elements bismuth and thorium were found as traces in native copper in amounts of 0.2 weight %.

For comparison, the analyses of the sample Hy325/78.70 are presented in Table 3-3. In this sample, thorium occurs as a trace in native copper in amounts of about 0.3 weight %. 23

Figure 3-2. Sample 325/68.30. Width of view = 0.85 mm.

.i 'V-viP '. - • - -i- -' :• - •**''.• " 1 V ': ""

•--,-..-• T

Figure 3-3. Sample 325/68.30. Profile in Table 3 -2. Width of view = 0.85 mm. 24

Table 3-2. Sample 325/68.30 Profile results. Composition in weight %. Figure 3-3

Point no. LSiO2 Cu K,O Na^ S Total #1 64.6 18.2 - 0.75 16.1 0.3 - 99.95 #2 63.8 18.2 0.1 0.55 16.1 0.6 - 99.35 #3 - 0.7 0.7 96.10 - - 0.07 97.57 #4 - 0.3 - 97.50 - - 0.06 97.86 #5 - 0.5 - 95.30 0.2 - 0.11 96.11 #6 64 18.5 0.2 0.62 15.7 0.3 - 99.32 #7 63.7 18.0 - 0.51 15.5 0.3 0.06 98.07

Table 3-3. Porosity sample Hy325/78.70 m. Analyses at random points in the mentioned minerals (Fig. 3-1). Composition in weight %.

Mineral || SiO^ Na^ A12O3 CaO K2O J S | Total K-feldspar 64.2 0.4 18.2 - 0.43 - 15.6 - 98.83 K-feldspar 64.1 0.5 18.4 - - - 15.8 - 98.80 Metallic copper - - 0.5 - 97.40 - - - 97.90 Metallic copper - - 0.3 - 97.10 - - 0.15 97.55 Metallic copper - - - - 98.04 - - - 98.04 Metallic copper - - 0.2 0.2 97.82 - - 0.06 98.28 Albite 67.3 10.9 19.7 0.3 0.6 0.6 - - 99.40 Albite 66.4^ 11.0 19.9 0.7 0.5 0.6 - - 99.10

Alloys of antimony and metallic copper of about 1 um in diameter occur in sample Hy325/78.70. The main accessories are tourmaline in sample Hy325/68 and fluorapatite in sample Hy325/78.70.

3.1.2 Copper sulfides associated with native copper

The copper sulfide phase djurleite (Cu, 93S) occurred within an open fracture surface at about 8.3 m depth (Sample S40; Marcos 1996). Further XRD analyses of the copper sulfide in contact with native copper revealed the existence of monoclinic chalcocite (or low chalcocite; see Table 3-4) closely associated with djurleite and native copper. The average of the Cu/S ratio in new analyses of copper sulfide in sample HY325/68.30 is 1.8 - 1.86 (digenite) for points 1 to 6 and 1.93 (djurleite) for points 7 to 9 (Table 3-5). 25

Table 3-4. Composition, crystal system, and stabilities of Cu-rich copper sulfide minerals.

Mineral Composition System Stability References

Chalcocite (low) Cu, 99_2S monoclinic T < 103 °C Roseboom(1966)

U Chalcocite ^ 1.98-2^ hexagonal 1 1U ) Roseboom (1966) (high) min ~ *- T ~ 435 °C -"•max ^JJ *~" Djurleite ^Ul.93-1.96^ monoclinic T < 93 °C Potter (1977)

Digenite (low) ^-U1.75-1.8^ cubic metastable Morimoto and Koto (1970) Digenite (high) Cu]834S cubic T < 93 °C Potter (1977) Cu,.765S T < 75 °C

Anilite Cu175S orthorhombic T < 72 °C Morimoto et al. (1969)

Posfai and Busek (1994) studied the intergrowths between djurleite and digenite and between djurleite and low chalcocite using high-resolution transmission electronic microscopy. They found that if both djurleite and chalcocite occur intergrown, chalcocite easily converts to djurleite under the electron beam through the rearrangement of Cu atoms. This fact and djurleite-digenite intergrowths may explain the changes in the Cu/S ratio in the analyses of the samples (Table 3-5). Native silver occurs at the contact of native copper and copper sulfide grains (Figure 3-4 b).

Table 3-5. Copper sulfide analyses (wt. %). The values are from random points in the bulk of copper sulfide grains (Sample Hy 325/68.30; Fig. 3-4). Analyses of points 1 to 6 correspond to totally sulfidized grains. Analyses of points 7 to 9 correspond to copper sulfide in close contact with native copper.

Point #1 #2 #3 #4 #5 #6 #7 #8 #9 Cu 77.97 77.39 74.65 73.49 74.40 74.69 74.39 74.40 75.82 U 0.23 - 0.18 - 0.27 - - - - s 20.77 21.33 20.76 20.55 21.18 20.63 19.61 19.51 19.60 Th 0.34 - - - 0.37 - - 0.47 -

Lead (Pb) is present in copper sulfide in amounts of 0.5 weight %. Alloys of native copper and native gold of 10 to 15 um in diameter occur in sample Hy325/68.30. 26

..-'5:, .:,

c?

Figure 3-4. Sample Hy325/68.30. Backscattered electron images, a) The two biggest grains in the center of the image are composed of native copper and copper sulfides. Dark gray is albite and light gray is k-feldspar. b) Magnification of the biggest grain in a). Metallic copper (light gray) and copper sulfide (medium gray). Composition in Table 3-5.

3.1.3 Copper sulfides and copper-iron sulfides

Copper sulfides and copper-iron sulfides occur in sample Hy324/53.30. They are not associated to native copper although they occur in a granite pegmatite sample of similar mineral composition to sample Hy325/78.70, where the main accessory also is green apatite. Two subsamples were examined. In the first one (Hy324/53.30a) chalcocite is associated with chalcopyrite (CuFeS2) and in the second one (Hy324/53.30b) chalcocite is associated with bornite (Cu5FeS4). B-Uranophane crystals coated the surface of the fracture. The study of these samples is important because (1) the existence of intergrowths of copper sulfides and U-silicates/uranyl compounds (Chapter 3.1.4) allows us to study the interaction of copper canister corrosion products and oxidised spent fuel matrix, and (2) the occurrence of copper sulfides may constitute an analogue to the behaviour of Cu corrosion products in a granitic environment.

The results of microprobe analyses of copper sulfides and Cu-Fe sulfides in subsample Hy324/53.30a are presented in Table 3-6. The distance from the fracture surface to the analysed points varies between 0.5 and 5 mm. Sulfides and uranyl compounds occur along interconnected microfractures in apatite and along intergranular contacts (Figures of PLATES 1 to 3). The Cu/S ratio for points 7 to 10 is closed to 1.93 (djurleite, low chalcocite). For point 14 the Cu/S ratio is 1.86. The composition for points 11, 12, and 14 is quite closed to chalcopyrite CuFeS0, or, more exactly, an iron-deficient chalcopyrite. 27

Table 3-6. Analyses of copper sulfides and Cu-Fe sulfides. Points in Figures 1C (PLATE I), 2A and 2B (PLATE 2).

1 Point #7 #8 #9 #10 #11 #12 #14 #15 Cu 78.86 77.25 77.04 78.95 33.54 33.25 77.10 29.43 Fe - - - - 27.65 27.76 - 23.01 S 20.62 20.20 20.19 20.87 35.22 35.36 21.07 30.39

At about 2 cm from the fracture surface, covellite (CuS) replaces low chalcocite, which at the same time replaces chalcopyrite and bornite. Chalcopyrite also replaces bornite. The oxidation state of Cu in sulfides and especially in covellite has been a matter of same debate. 2+ 2 + 2 Covellite used to be thought of as (Cu S2 ')(Cu2 S "), but recent L.edge X-ray Absorption Spectroscopy (XAS) (van der Laan et al., 1992) and earlier X-ray Photoelectron Spectroscopy (XPS) and theoretical studies (Nakai et al., 1978; Tossell, 1978) have shown it to contain only Cu(I). The timing of the reduction of Cu(II) initially present in aqueous solutions when precipitated as a sulfide has been addressed by Pattrick et al. (1997). The reduction from Cu(II) to Cu(I) is fast and completed at least within the first minutes after precipitation of Cu at 5 °C.

The subsample Hy324/53.30b has bornite islands in a big chalcocite grain (0 = 5 mm). Chalcocite also may rim bornite. Metallic bismuth (Bi) rims chalcocite and fills microcracks within it (Fig. 3-5). Due to its affinity for sulfur, metallic Cu is readily sulfidized to univalent Cu+, whereas part of the trivalent Bi3+ component is simultaneously reduced to the metallic state at temperature near to 200 °C (Wang, 1994) in the wet Cu-Bi-S system. The occurrence of altaite (PbTe), a lead telluride (Table 3-7), within chalcocite may indicate a decrease in temperature of the ore fluids (< 300 °C) as the stability field of tellurium species increases at the expenses of all other species with decreasing temperature (Zhang & Spry, 1994).

Table 3-7. Composition of altaite (PbTe) in wt %.

Fe Cu Ni Co Zn Bi Pb Ag Te S Total - 0.5 - - - 0.8 59.5 0.3 38.2 - 99.30 - 0.5 - - - 0.7 59.5 0.3 38.3. - 99.30 28

PLATE 1

Figure 1A Figure IB

Figure 1C Figure ID

All the Figures in Plates 1 and 2 are from Sample 324/53.3a

Figures 1A and IB. Microcracks in apatite filled with copper sulfides, copper iron sulfides and uranyl compounds. Points 1 to 5 in Table 3-8.

Figure 1C. Uranyl compounds (bright gray; Point 6 in Table 3-8) along the contact between apatite and a Cu-Fe sulfide grain (Points 7 and 8 in Table 3-6).

Figure ID. Composed Cu sulfide grain. Points 9 and 10 in Table 3-6. This grain is rimmed by calcite. 29

PLATE 2

Figure 2A Figure 2B

*«

Figwre 2C Figure 2D

Figures 2A and 2B. Copper-iron sulfides (dark gray). Copper sulfide (light gray). Points 11, 12, 14, and 15 in Table 3-6. Point 13 in Table 3-8.

Figure 2C and 2D. Microcracks in albite (black background) and apatite (dark gray) filled with copper sulfides. Figure 2D is a magnification of the fracture filling in apatite in Figure 2C. Points 18, 19, and 20 in Table 3-8. 30

3^tdkA L"td &'•>, s, v!;'. i A

V /%Te.

Figure 3-5. Sample Hy324/53.3b. Bornite islands in a large chalcocite grain. Native bismuth within chalcocite. Width of view = 3 mm.

3.1.4 Occurrences of uranyl compounds

Occurrences of U(VI) compounds are examined in order to get insights in the mobility of copper either coming from native copper or from copper sulfides. Copper is present in all the points analysed corresponding to possible intergrowths of uranophane and uranyl hydroxides (Points 1, 2, 4, 6, 13, 19, 20 in PLATES 1 and 2; Table 3-8). Sulfur is present only in points 1,2,4,5, and 19. Points 3 and 18 correspond to bluish green manganapatite analyses in sample Hy324/53.3a. The fracture surface was coated with B-uranophane (Fig. 3-6) as indicated by XRD analyses. The ideal composition of uranophane is presented in Table 3-8 for comparison (Figures 1A, IB, 1C in Plate 1). ThO2 concentration is below detection limits (also see Point 1 in Table 3-9). Several fractions of B-uranophane were selected for further U-series isotopes analyses (Chapter 5). 31

Table 3-8. Analyses of uranyl compounds and related minerals (weight %). Sample Hy324/53.30a. Points in PLATES 1 and 2. Values marked with * are calculated.

Points Ideal #1 #2 #3 #4 #5 #6 #13 #18 #19 #20 uranophane

SiO2 14.03 16.7 16.2 - 15.6 15.4 14.2 14.3 - 14.2 13.5 MA 1.4 1.7 - 1.8 1.7 - 2.6 - - - UA 61.21 58.22 61.10 - 72.50 71.90 74.80 75.50 - 73.30 72.20 YA 1.1 1.4 - 0.4 0.30 - - - -

Cu 4.80 0.43 - • 0.22 - 1.30 0.45 - 0.55 0.75

BiA 7.8 4.6 - - 0.2 - - - - - CaO 6.54 2.2 3.7 53.1 3.1 3.3 6.0 3.6 53.0 6.1 5.6 S 1.73 0.73 - 0.09 0.08 - - - 0.05 -

P2O5 0.2 0.4 42.6 - 0.2 - - 42.8 - - PbO - 0.3 - - 0.3 0.2 - - 0.2 0.3 MgO - 0.3 - 0.7 0.8 - - - - - MnO - - 1.3 - - - - 1.3 - - FeO - - - - 0.3 - 0.4 - - - F - - 0.4 - - - - 0.4 - -

H2O 18.22 5.85* 9.14* 2.6* 5.59* 5.52* 3.5* 3.15* 2.5* 5.6* 7.65*

The uranium in uranyl compounds occurring along silicate and sulfide grain boundaries and as intergrowths with Cu and Cu-Fe sulfides may come from the leaching of primary uraninite. Figure 3B in PLATE 3 shows an altered grain of uraninite. Several analyses of this sample are shown in Table 3-9. 32

\J .\"- - %M' .Vb-:"

Figure 3-6. fi-uranophane (yellow) coats the fracture surface in sample Hy324/53.3. It is preferentially distributed around copper sulfide grains. Length of copper sulfide (intergrown djurleite & low chalcocite) in the center of the image is 2 mm.

In the analyses of sample Hy325/78.70 (Points 18 to 26; Table 3-9 ) sulfur is below detection limit and does not appear in the Table 3-9. Sulfur is also below detection limit in the analyses of Point 1 in sample Hy324/53.30b (Table 3-9), where Cu-Fe sulfides occur. In the first sample, copper only occurred in its native state (For analyses of native copper in this sample see Table 3-3) and the maximum content of sulfur was 0.14 weight.%. 33

PLATE 3

Figure 3A Figure 3B

»-.*» -— <• - -IF*.-

Figure 3C Figure 3D

Figure 3A. Sample 324/5 3.3 b. Chlorite and hydrated uranyl oxides inter growth in the thickest ribbon-like mineral in the centre of the image. Point 1 in Table 3-9.

Figure 3B. Sample 325/78.80. Altered uraninite grain. Points 18 to 26 in Table 3-9.

Figure 3C. Sample 325/78.80. Points 9 to 12 in Table 3-10.

Figure 3D. Sample 325/78.80. Uranyl compound (Point 15 in Table 3-10) along the contact between apatite (light gray, Point 13 in Table 3-10) and quartz (dark gray, almost black). 34

Table 3-9. Uranyl compounds and related minerals in sample Hy324/53.3b (wt.%) {#1; Figure 3A in PLATE 3) and sample Hy325/78.70 (#18 to #26; Figure 3B in PLATE 3).

Point #1 #18 #19 #20 #21 #23 #24 #25 #26

SiO2 26.9 - 14.8 - 58.0 22.9 19.3 64.1 67.1 A1A 10.1 — 1.4 — 23.9 2.8 1.6 17.9 19.9

U2O3 52.30 — 2.90 — — 7.00 8.18 - -

ThO2 — — — — — 7.9 3.9 — — YA — — 7.5 - - 14.4 14.5 - — Cu 0.45 — 1.03 — — 0.20 — — — BiA — — 1.3 0.2 - 1.0 1.1 - - Nap - — — - 8.1 - — - 11.22 CaO 2.7 54.2 0.9 54.2 0.3 2.0 1.5 — 0.9

K2O 0.5 — - — — — — 15.7 - PbO 0.4 — 11.4 — — 0.9 1.2 — — MgO 2.5 — — — — — — — — MnO — 1.2 20.9 1.0 — 0.2 1.0 — — FeO 0.6 — 2.7 — — 0.3 0.2 — — F — 0.6 — 0.7 — — — — — PA — 43.2 0.9 43.9 — 3.0 2.6 — — Total 96.45 99.2 65.73 100 90.3 62.6 55.08 97.7 99.12 Minera! Chlorite Mn- Hydrated Mn- Hydrated Hydrated Hydrated K- Albite phase & Urano- apatite Y-Pb-Si apatite albite U-Th-Y- U-Th-Y- feldspar phane phase Si phase Si phase

Table 3-10 shows the analyses of intergranular uranyl accumulations in sample Hy325/78.70. Copper values are below detection limit in all the points analysed. 35

Table 3-10. Analyses (weight %) ofuranyl compounds in sample Hy325/78.70 (Figures 3C and 3D in PLATE 3).

Points #9 #10 #11 #12 #13 #14 #15

SiO2 - 14.6 13.6 100 - 14.2 100

U2O3 - 69.60 65.05 - - 75.0 - Cu ------MnO 1.3 - - - 1.3 - - CaO 54.0 7.4 5.00 - 54.0 5.1 - F 0.9 - - - 0.7 - -

P2O5 42.1 2.0 - - 43.5 - - Total 98.3 93.6 83.65 100 99.5 94.3 100 Mineral Mn-F- Uranophane Uranophane Quartz Mn-F- Uranophane Quartz phase apatite + uranyl apatite + uranyl phosphate? hydroxide?

3.1.5 Occurrence of copper oxide

Native copper grains as large as 1 mm in diameter occur within granite pegmatites on a fracture surface about 98.75 m depth in the drillcore Hy324. These native copper grains have rims of cuprite 0.01 to 0.1 mm thick. The smallest grains are totally oxidised. This fracture surface was covered in part by smectite (Fig. 3-7) as indicated by XRD analyses. The copper content for two fractions of smectite varied one order of magnitude from 33.8 ppm to 342 ppm. The first fraction was less coherent than the second one. The uranium content in the same samples varied between 71 and 77 ppm (Appendix A). Several smectite fractions were selected for further U-series isotope analyses (Chapter 5 and Appendix B). 36

Figure 3-7. Fracture surface(Sample Hy434/98.75m); Cuprite partially rims native copper grains. The yellowish material is smectite (XRD analysed). Width of view = 16 mm.

3.2 Paragenesis

Although the paragenetic sequence remains to be documented in detail, the principal stages of mineralization are suggested in Figure 3-8, with the caveat that a large degree of overlap may exist.

Minerals Idealised Mineral Paragenesis Early -

Native copper Uraninite Bornite & High chalcocite Native bismuth Low chalcocite, djurleite, 9 9 & digenite Calcite 7 Covellite ? Smectite 9 9 7 Cuprite 7 7 Uranyl compounds - - 9

Figure 3-8. Idealised paragenesis of native copper and associated fracture filling and fracture coating minerals. Explanation below. 37

Native copper stage (and uraninite). Native copper is a primary ore mineral within the granite pegmatites' rock forming minerals K-feldspar and albite as it is exsolved along cleavage planes and crystallographically oriented fractures in these minerals. Hydrothermal brecciation stage. At this stage, quartz is partially recrystallized and the more brittle tourmaline and apatite grains are brecciated. Native copper also may be partially remobilised at this stage. The majority of sulfide minerals (with or without native copper) are preferentially distributed along microfractures and within recrystallized quartz grains. This suggests that Cu and Cu-Fe sulfide mineralization occurred late in the cycle of "hydrothermal" breccia formation at two stages: a) high temperature stage up to 300 °C at which native bismuth, Cu-Fe sulfides and high chalcocite precipitated and b) low temperature stage (<100 °C) at which low chalcocite, iron depleted chalcopyrite, digenite, and djurleite precipitated. Covellite may belong to this stage although covellite also is a characteristic mineral of the supergene weathering environment (e.g., Blanchard 1968). Calcite, covellite, cuprite and uranyl compounds may correspond to a recent weathering stage. The term recent refers to from 1 million years (Ma) until the present, since the primary mineralization is dated about 1800 -1700 Ma ago (Raisanen 1986).

3.2.1 Source of sulfur

Three sulfides, two of them low chalcocite and one bornite were analysed from the samples 34 Hy324/53.30 and S40 (Dh304; Marcos 1996). The 5 S values fall in the narrow range of -5.0 %o to -5.7 %o, thus suggesting a major component of sedimentary sulfur in these Cu and Cu-Fe sulfides. 39

4 HYDROCHEMISTRY

4.1 Groundwater sampling, determinations, quality and representativeness

Groundwater sampling and associated measurements (e.g., Eh, pH) were carried out using the SKB mobile field laboratory (Almen et al. 1986). The system consist of 1) the down-hole part and its control unit in a separate trailer; 2) laboratory unit containing facilities for chemical analyses, on-line measurement systems, and computers for data storage and operation of the down-hole probe.

The down-hole part of the system contains hydraulically expanded rubber packers, in situ chemical probe, and a hydraulically operated piston pump. All parts coming to contact with groundwater are made of stainless steel. Maximum stroke capacity of the pump is about 120 ml/stroke, corresponding to flow rates of about 100 - 200 ml/min. Pressure-pulse sequences for the pump are regulated by a timer.

The down-hole, in situ, chemical probe consists of special-designed electrodes for redox- potential and pH measurements and a gel-filled silver-chloride reference electrode. The volume of the measuring chamber is about one litre; the measuring chamber is constructed of stainless steel. Inert redox electrodes in the down-hole chemical probe are platinum (Pt), gold (Au), and glassy carbon (C). The probe also contains the electronics required for the signal conversion and data transmission.

The bedrock-fractures to be sampled were isolated, and groundwater from the drillhole section was pumped until the sample was considered to be representative. The amount of drilling water was monitored during pumping by analysing the uranine concentration (original concentration in flushing water was 500 |Xg/l). Uranine was analysed by fluorimetry according to the method describe by Ruotsalainen et al. (1994). The sampling sections, times, flow rates, water volumes pumped, and uranine concentrations at the end of pumping are summarised in Table 4-1.

Table 4-1. Summary of the groundwater sampling survey at Hyrkkold.

DH DH section (m) Sampling time Flow (ml/min) Vtot (1) Uranine (|ng/l) 325 90-104.1 15.7. -4.8. 110 3100 15 325 65-70.6 4.8. - 18.8. 115 ^ 95 2000 6 325 67 - 68.5 20.8. - 17.9 120 -» 60 3400 11 325 43.5 - 45 18.9. - 14.10. 100 3700 3 325 52.4- 53.9 15.10. -3.11 65 1800 5 40

When starting the pumping, the stroke-rate of the pump was adjusted in such a way that the initial flow rate was apparently constant. The flow rate remained approximately constant during the whole pumping time, except in section 65 - 70 m, in which the rate decreased evenly from about 115 ml/min to about 95 ml/min during the two weeks time. This indicates that the transmissivity of the section was lower than, for example, that of the crush zone of the drill-hole bottom (section 90 - 104 meters). This was also observed when sampling the shorter section (67- 68.5 m) and the pumping rate was reduced to 60 ml/min, at which constant flow could be maintained. Due to a low yield, lower pumping rate was also used in the section 52.4 - 53.9 m in the DH 325.

The down-hole part is connected to the surface unit by a umbilical hose, inside of which there are the polyamide tubes for transport of the groundwater sample to the surface, for the operation of the pump, and for pressurizing the packers. The hose also contains the electrical cables for data transmission and power supply for the down-hole probe. A steel wire inside the hose takes up the load of the system. On the surface, the water sample first flows through the remaining part of the umbilical hose (total length of 1000 meters) wind up on a reel, and is then led through a polyamide tube (6/4 mm, i.e. wall thickness of 1 mm) from the control unit to the laboratory wagon.

In the laboratory unit, water passes a magnetic flow-meter and enters the measuring chamber, in which the on line electrodes are mounted. The redox electrodes (Pt, Au, and C), as well as the reference (silver chloride) and the pH-electrode (glass) are commercial products (Yokogawa). In order to avoid air contamination, the measuring chamber has an overpressure of 0.5 - 1 bar during operation. Data from the downhole probe, flow-meter, and surface probes are stored by a computer system.

Before the start of the measurements both down-hole and surface probes are calibrated. For redox-potential calibration, quinhydrone standard is prepared and used as follows: 1) standard pH-buffer solutions are prepared for pH 4, pH 7 and pH 10, one liter of each, 2) 3.6 g of quinhydrone-powder is added to each buffer solution. 3) the solutions are circulated through the measuring systems until stable potential-readings are attained (about 30 min each solution). Calibrations of 'on line' electrodes are made at room temperature (about 20° C), at which the theoretical Eh-values (EhqUjn) of the calibration solutions are: pH 4 = 471 mV, pH 7 = 296 mV, pH 10 = 119 mV. The 'in situ' electrodes are calibrated at about 7-10° C temperature. The redox-correction factor (ECOrr) for each electrode is the difference between the theoretical and observed (Ecai) potential of the calibration: •t'corr~ -t-flquin " -t-cal

The measured redox potentials (Emeas) can then be converted to Eh-scale (i.e., vs. Standard Hydrogene Electrode, SHE) by

till limfifls » ^corr

The pH-calibrations were done using standard pH-buffers (pH 4, pH 7, pH 10) at 20° and 7° C for 'on line' and 'in situ' electrodes, respectively. The measured pH-data (on line) was compensated for temperature variations by extrapolating all values to the temperature of calibration (Figure 4-1). 41

: ; i i i : ' ; ! :

6.80

6.70 1 ; . j 1 i. 6 60 V llfk

6 50 1 ! IV ! I | 1 i | | 6.40 - i i 1 i i 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Temperature (C)

Figure 4-1. Temperature compensation of the pH data: The apparent variation of the 'on line' pH values as a function of temperature was compensated by extrapolating all values to the calibration temperature (20° C), DH325, section 67 - 68.5 m.

After a pumping period of a couple of weeks, when changes in uranine concentration, pH and redox were levelled off, groundwater samples were taken for analysis.

Dissolved sulfide, ferrous iron and total iron were analyzed in the mobile laboratory from filtered samples (0.45 (am) immediately after sampling. Methods used were those described by Ruotsalainen et al. (1994). Other dissolved components were analyzed in the Analytical Laboratory of the Geological Survey of Finland (GTK). Main cations and trace elements were analyzed from filtered (0.45 um), acidified (0.5 ul ultrapure nitric acid, HNO3, in 100 ml) samples using ICP-MS. Anions were analyzed from untreated samples using ion chromatography. The tritium content and the 234U/238U activity ratio of the samples was measured by a-spectrometry (Laboratory of Radiochemistry, University of Helsinki). Carbon isotopes were analyzed by Angstromlaboratory, University of Uppsala, Sweden, sulfur-34 was analyzed by University of Waterloo, Canada.

Alkalinity of water was determined by titration in GTK together with analysis of other dissolved components. During the pumping period before sampling, a measurement apparatus ('Ecolys') carried out automatic titration of alkalinity, measurement of pH, electrical conductivity and chloride concentration (ion-selective electrode). The results (Ahonen and Marcos 1997) indicated about 0.1 - 0.3 mEq lower alkalinities than those determined in laboratory, while the pH values measured by the automatic analyzer were about 0.3 - 0.4 units higher than those measured on line. 42

The quality of chemical analysis was checked by calculating the charge balance from the analytical result, which in four cases of five was within one percent (from the electrical equivalent sum of anions and cations). The chemical analyses of the water samples in GTK were based on an accredited method, except the determination of alkalinity. Cations were analyzed using ICP-MS, anions using IC. In the following, results are given using three digits, if the relative uncertainty of result is 5 percent or less, and using two digits, if the uncertainty is higher than 5 percent or if the method is not accredited. However, more decimals than the determination limit allows were not given.

The main factor possibly affecting the representativeness of the water samples obtained is the disturbance caused by drilling. After drilling, flushing water was removed by pumping about 40 cubic meter of water from each borehole (Ahonen and Marcos 1997). The concentration of uranine tracer indicated that the contribution of flushing water was about 3-5 percent at the end of the pumping. Before the actual water sampling, 2-4 cubic meters of water was pumped during 2 - 4 weeks time from each sampling sections. In all cases, the water amount pumped was more than 50 times the volume of the sampling section. This assures that mixing of water in the drillhole before packer inflation could not affect the results. Only small changes were observed in the parameters monitored during pumping, indicating that there was no further disturbance due to the drilling. The water samples obtained are considered to be representative of the present, undisturbed hydrogeological conditions of the site.

4.2 Results

4.2.1 Parameters monitored during pumping

Redox potential of the water was measured during pumping using six electrodes; platinum (Pt), gold (Au), and carbon (C), both on the surface, on line, and in the bore hole, in situ; recording interval was from 30 min to two hours. Groundwater pH was measured both on the surface and in the borehole. The results are summarized in Figures 4-2 to 4-6. Eh (mV)

14.7.97 00:00 14.7.97 00:00

15.7.97 00:00- 15.7.97 00:00-

: 16.7.97 00:00 - 16.7.97 00:00- I 17.7.97 00:00- 17.7.97 00:00- : 18.7.97 00:00- 18.7.97 00:00-

19.7.97 00:00- 19.7.97 00:00-

20.7.97 00:00 - 20.7.97 00:00 -

8 g 21.7.97 00:00-; 21.7.97 00:00-

22.7.97 00:00-; 22.7.97 00:00 •

: 23.7.97 00:00 - 23.7.97 00:00 -

to 24.7.97 00:00 - 24.7.97 00:00 -

_, 25.7.97 00:00 - CO «s to 25.7.97 00:00 - ft J" tn CO ® 26.7.97 00:00 26.7.97 00:00 - to 9 9 27.7.97 00:00 -; I 27.7.97 00:00 • 2 a s o 28.7.97 00:00 - 28.7.97 00:00

Pii 29.7.97 00:00 - 29.7.97 00:00 •

30.7.97 00:00 - 30.7.97 00:00 - m r 1 rn I I TO i S 31.7.97 00:00- 31.7.97 00:00- c CD 1.8.97 00:00- 1.8.97 00:00-

2.8.97 00:00 - 2.8.97 00:00 - m r 1 m a 3.8.97 00:00 - 3.8.97 00:00 - ;; CD —t c5 s CD •< 4.8.97 00:00 - 4.8.97 00:00 -

5.8.97 00:00 - 5.8.97 00:00 -

6.8.97 00:00 - 6.8.97 00:00 - Eh (mV) TO s. ^' PH

3.8.97 00:00 3.8.97 00:00-- hON 4^ TO i r° 4.8.97 00:00 • 4.8.97 00:00-

* 2 § 5.8.97 00:00 f 5.8.97 00:00-

6.8.97 00:00 - 6.8.97 00:00-

TO "^ t*1 7.8.97 00:00 7.8.97 00:00-

8.8.97 00:00 • 8.8.97 00:00-

§|* 9.8.97 00:00 - 9.8.97 00:00- ||| 10.8.97 00:00 - 10.8.97 00:00-

11.8.97 00:00 11.8.97 00:00- 3

12.8.97 00:00 • 12.8.97 00:00- i 13.8.97 00:00 13.8.97 00:00- | § | 2 HPT B m HPT Y m

Sri § 14.8.97 00:00 • 14.8.97 00:00-

15.8.97 00:00 - m m TO 15.8.97 00:00- l c c S' CD 16.8.97 00:00 - -< 16.8.97 00:00- I pH B pH Y m I I 17.8.97 00:00 - 17.8.97 00:00- £ om S Cd3 03 -< 18.8.97 00:00 - 18.8.97 00:00- O r

SI 19.8.97 00:00 19.8.97 00:00-

20.8.97 00:00 -1 20.8.97 00:00- aTO ao- | § I 19.8.97 00:00 19.8.97 00:00 20.8.97 00:00 - \ 20.8.97 00:00 • 21.8.97 00:00- 21.8.97 00:00- 22.8.97 00:00-! 22.8.97 00:00 •: 23.8.97 00:00 - 23.8.97 00:00 - I 24.8.97 00:00 - \ 24.8.97 00:00 -: 25.8.97 00:00 - 25.8.97 00:00-: 26.8.97 00:00 - 26.8.97 00:00-: 27.8.97 00:00 - 27.8.97 00:00 - \ 28.8.97 00:00 • \ 28.8.97 00:00 •: S' ^ 29.8.97 00:00- 29.8.97 00:00 • TO TO o TO" 30.8.97 00:00 - 30.8.97 00:00 - 31.8.97 00:00-: 31.8.97 00:00-: l 1.9.97 00:00- 1.9.97 00:00-; 2.9.97 00:00 - \ 2.9.97 00:00 -: s u 5 a. 3.9.97 00:00 - 3.9.97 00:00-; to f a 01 4.9.97 00:00-; en 4.9.97 00:00 - -j 5.9.97 00:00-: 5.9.97 00:00 - O) 01 6.9.97 00:00 - \ 00 a a-, -s 6.9.97 00:00-i 01 a a g a, s § 7.9.97 00:00 -; 7.9.97 00:00 - S' ^ a. 8.9.97 00:00 • 8.9.97 00:00 - 9.9.97 00:00 - 9.9.97 00:00 - \ 5? 10.9.97 00:00-: 10.9.97 00:00-; 11.9.97 00:00-: 11.9.97 00:00-; a 12.9.97 00:00 - 12.9.97 00:00- Cr 13.9.97 00:00-: 13.9.97 00:00- 14.9.97 00:00- 14.9.97 00:00-; 15.9.97 00:00- 15.9.97 00:00-; SL D 16.9.97 00:00- 16.9.97 00:00-; 17.9.97 00:00- 17.9.97 00:00-; 18.9.97 00:00- 18.9.97 00:00-; 19.9.97 00:00- 19.9.97 00:00- Eh (mV)

17.9.97 00:00 1 § S 17.9.97 00:00 18.9.97 00:00 18.9.97 00:00- 19.9.97 00:00 19.9.97 00:00- 4^ 20.9.97 00:00 20.9.97 00:00 - 21.9.97 00:00 K 21.9.97 00:00- 22.9.97 00:00 22.9.97 00:00 - CX. o 23.9.97 00:00 til 23.9.97 00:00 - Ob 24.9.97 00:00 24.9.97 00:00 • 25.9.97 00:00 25.9.97 00:00 - 26.9.97 00:00 26.9.97 00:00 - 27.9.97 00:00 27.9.97 00:00 - s 5 a 28.9.97 00:00 28.9.97 00:00 - 29.9.97 00:00 29.9.97 00:00 • 30.9.97 00:00 CO 30.9.97 00:00 - N5 1.10.97 00:00 1.10.97 00:00 TO a 2.10.97 00:00 2.10.97 00:00- OQ 3.10.97 00:00 3.10.97 00:00- a •>! i a" o 4.10.97 00:00 4.10.97 00:00- 1 a m m a a 5.10.97 00:00 5.10.97 00:00- X X a. "• ^ 6.10.97 00:00 6.10.97 00:00- CD ? ^ a 7.10.97 00:00 7.10.97 00:00- K 8.10.97 00:00 8.10.97 00:00- m m O a | a' 9.10.97 00:00 ci 9.10.97 00:00 03 -< 10.10.97 00:00 10.10.97 00:00- i 11.10.97 00:00 11.10.97 00:00- m m I o 12.10.97 00:00 12.10.97 00:00- CsO "< 13.10.97 00:00 13.10.97 00:00- 14.10.97 00:00 14.10.97 00:00- 15.10.97 00:00 15.10.97 00:00- 16.10.97 00:00 16.10.97 00:00- s« re «? pH Eh (mV) re <~> QTQ" Ul 14.10.97 00:00-- | § I 14.10.97 00:00- re o, *< 15.10.97 00:00-: 15.10.97 00:00- 16.10.97 00:00-: 16.10.97 00:00- 17,10.97 00:00- 17.10.97 00:00- 18,10.97 00:00- 18.10.97 00:00- 19.10.97 00:00 •'• 19.10.97 00:00-

20.10.97 00:00- 20.10.97 00:00-

21.10.97 00:00-' 21.10.97 00:00-

: 22.10.97 00:00- 22.10.97 00:00-

: 23.10.97 00:00- 23.10.97 00:00-

24.10.97 00:00- 24.10.97 00:00- CO : to |' 25.10.97 00:00 - .fa 25.10.97 00:00

: 26.10.97 00:00- 26.10.97 00:00- S" 3 2 27.10.97 00:00-: 27.10.97 00:00- l 1 28.10.97 00:00- 28.10.97 00:00- m I I Im "0 : o 29.10.97 00:00- 29.10.97 00:00- CcDz

30.10.97 00:00- 30.10.97 00:00-

31.10.97 00:00-; ——pH B 31.10.97 00:00- m rT m X 1.11.97 00:00- CD Hi 1.11.97 00:00- o ED c : < 2.11.97 00:00- 2.11.97 00:00- a- " 3.11.97 00:00- 3.11.97 00:00 re" -< : to I* 4.11.97 00:00- 4.11.97 00:00- ; 5.11.97 00:00- 5.11.97 00:00 48

All measurements indicate very oxidizing conditions, but different electrodes show systematically different Eh values: platinum « 500 mV, gold « 400 mV, carbon « 300 mV at pH « 6.7. Oxygen concentration was also measured continuously during pumping, but unsystematic fluctuations occurred in many cases. However, all measurements indicated presence of oxygen, concentration varying between about 0.5 - 5 mg/1 (O2). Measured oxygen concentrations also seems to correlate with the measured Eh (compare Figures 4-6 and 4-7).

In general, it is well known that the measured Eh-values in oxic systems seldom or never reach the theoretical O2/H2O boundary (about 1230 mV - 59 pH), but the highest values remain at about 300 mV lower level (Baas-Becking et al. 1960). The 'experimental water-oxygen line' at about 900 mV - 59 pH has been observed to represent the upper redox-boundary also in laboratory experiments with oxygenated distilled water (Natarajan and Iwasaki 1974). Hoare (1968) explained it through the mixed potential theory involving redox pairs Pt-0 / Pt, and O2/H2O.

Studied groundwaters contained also nitrate and, consequently, nitrate/nitrite pair may also contribute the measured redox value. The equilibrium Eh-value of this pair, assuming equal concentrations for both species, is about 550 mV at pH 5, and about 310 mV at pH 9 (e.g., Lide & Frederikse 1995), increasing with increasing NCV/NCV ratio.

The Eh-values at the end of the pumping are summarized in Figure 4-8 as a function of the contemporaneous pH-value. The Eh-values measured by Pt-electrode indicate that the electrode potential is in balance with dissolved oxygen according to the mixed potential concept. The non-platinum electrodes naturally show dissimilar values, because they can not establish the same mixed potential. Electrode kinetics of gold and carbon electrodes are less well known than that of platinum electrode.

Ferric oxides (mainly hematite) are very common in Hyrkkola, but the Eh - pH slope of about 150 - 180 mV / pH unit (proton to electron ratio 3:1), which is typical for waters being buffered by ferric/ferrous system, or for electrodes being in equilibrium with soluble ferric/ferrous pair, was not observed. Because the measured potentials seemed to be dependent on the electrode material, the nitrate/nitrite equilibrium pair was not considered as the most important electron acceptor/donor in the groundwater/measuring electrode system. However, it can also have a contribution to the observed potentials.

A striking feature of the pumping results was that pH values were more slowly stabilized than Eh-values. The pH-values measured automatically by 'Ecolys' were about 0.3 - 0.4 units higher than those measured 'on line'. This is likely to be due to CO2 degassing, during pressure release (high Pco2 in all samples). Because of technical difficulties, in situ pH-measurements show unstable values in some of the experiments.

All waters pumped were fresh. Electrical conductivity (EC) of waters remained constant during pumping, except in pumping from drillhole 324, in which EC increased from about 19 mS/m to about 22 mS/m during the period 16.10 - 29.10, and more slow after that. Also pH showed similar trend (Figure 4-6), but an even more pronounced change was observed in Eh and oxygen concentration (Figure 4-7). 49

o o o o o

o o o o o o o

OOOOO — "^

Figure 4-7. Oxygen concentration during pumping from DH 324, section 52.4 - 53.9 m.

800

700 8PtY »PtB OAuY

600 - OAuB ACY ACS

500 -

400 -

300 -

200 -

100 -

7 PH

Figure 4-8. Summary of Eh-values at the end of pumping. The line shows the 'experimental water-oxygen line' (see text for explanation). Electrodes used in measurement: Platinum (Pt), gold (Au), and carbon (C). Y refers to the on line surface measurement, B refers to the in situ measurements. 50

4.2.2 Chemical characteristics of the groundwater samples

Results of chemical analyses and other characteristics of the Hyrkkola" groundwater samples are summarized in Table 4-2. All samples represent fresh Ca-HCC>3 water, Mg and Na being the other main cations. The amount of total dissolved solid (TDS), and electrical conductivity (E.C.) of these water samples are similar to (or even lower than) those of typical near-surface waters. The lowest salinity is in the deepest sample, SI taken from the crush zone met in the bottom of both bore holes, indicating that there is a rapid infiltration of water into this fracture zone.

Except for sample S2, all waters were slightly acidic, pH between 6.5-7, and all waters contain dissolved oxygen; both factors indicate a short water/rock interaction time. Geochemical modelling calculations with the program PHREEQE show undersaturation with respect to calcite and oversaturation with respect to atmospheric CO2 (i.e., with a gas phase having log Pco2 = -3.44). Equilibrium of the water sample S2 with calcite indicates that the water sample (or a component in it) has had a longer residence time than the other four samples. Sample S3 is a subsection of the sampling section of S2, but the chemical characteristics of S3 indicate more active groundwater flow. The more 'mature' characteristics of S2 may be due to a contribution of slowly moving water from less conductive fractures of the five meter sampling section, while the sampling section of S2 was selected to comprise only the fracture at 68.3 m in DH324. Water samples S3, S4, and S5 are chemically very similar to each others.

The measured 5 C-13 values around -23 indicate an open-system equilibration between soil CO2 and groundwater (Fritz et al 1989). Deuterium and 0-18 isotope compositions of all three analyzed samples are identical, and correspond to the annual mean isotopic composition of the precipitation in this part of Finland .

The groundwaters analyzed also have a clear signal from the agricultural activity, seen for example in the high nitrate concentration. Anomalous concentrations of certain trace elements probably also originate from the fertilizers.

Even though the groundwaters in Hyrkkdla are very oxidizing and 'aggressive', copper concentration in the water samples remains low. Compared to the mean copper concentrations in Finnish wells, copper concentrations analyzed in Hyrkkola are low (Table 4-3). One reason for that is probably the relatively rapid infiltration time and, consequently, the short water/rock interaction time. However, the highest copper concentration observed in Hyrkkola was associated with the most dilute water sample which probably has the shortest contact time with the bedrock, while the lowest concentration was associated with the most 'mature1 water with highest pH. This fact underlines the importance of solubility control for the hydrogeochemical behaviour of copper. In addition to stoichiometric precipitation, the copper-retention mechanism may be sorption and/or coprecipitation.

Uranium concentrations in Hyrkkola groundwaters are anomalously high. The two reasons for that are evident: 1) extremely oxidizing conditions maintain uranium at the soluble hexavalent state, and 2) uranium-rich bedrock with secondary uranium minerals provides an easily mobilized source. 51

Table 4-2. Chemical characteristics of the Hyrkkold groundwater samples : SI = HY325, 90-104 m; S2 = HY325, 65-70 m ; S3 = HY325 67-68.5 m; S4 = HY325, 43.5^5 m; S5 = HY324, 52.4-53.9 m.

SI S2 S3 S4 S5 pH 6.6 8.2 6.6 6.7 6.8 E.C. (mS/m) 16 18 19 20 22 TDS (mg/1) 116 156 151 159 150 O2 (mg/1) 4 0.8 1 2.2 1.8 Ca (mg/1) 14.4 16.9 19.4 19.7 20.3 Mg (mg/1) 4.3 5.3 6.9 7.6 7.2 Na (mg/1) 7.16 15.6 5.94 7.04 7.22 K (mg/1) 2.84 0.62 3.16 1.75 2.23 Cl (mg/1) 3.8 4.4 3.7 3.7 3.4 SO4 (mg/1) 10 13 10 10 11 HCO3(mg/l) 57 85 85 92 85 NO3 (mg/1) 10 5.3 11 11 7.2 Si (mg/1) 5.89 6.2 4.78 5.15 4.98 F (mg/1) 0.2 1.1 0.2 0.2 0.2 Fetot(mg/1) 0.05 <0.01 <0.01 0.02 <0.01 Fe(H) (mg/1) 0.01 <0.01 <0.01 <0.01 <0.01 Sr (ug/1) 93.2 113 108 114 111 Mn (ug/1) 45.5 6.12 104 49.8 102 Cu (ug/1) 13 2.5 4.5 4.1 3.4 U (ug/1) 93 2000 420 630 640 V (|ig/l) 0.05 17 0.07 0.13 0.08 B (ug/1) 43 83 28 31 30 Ba (ug/1) 15.7 0.86 8.12 2.92 5.43 Li (ug/1) 2.3 2.3 3.4 3.9 3.4 AI (iugyi) 26.4 9.4 1.7 2.8 1.0 Rb (ug/1) 4.55 0.51 3.7 1.55 3.23 Co (ug/1) 0.21 0.05 0.43 0.24 0.46 Ni (ug/1) 8.2 6.0 11 6.6 7.4 Mo (ug/1) 61 25 26 14 17 Zn (ug/1) 100 15 130 90 49 Cd (jig/1) 0.22 0.06 0.17 0.10 0.05 As (ug/1) 0.13 3.2 0.1 0.17 0.1 Sb (Mg/1) 1.7 8.8 2.2 2.7 2.4 Se (ug/1) 1.2 15 1.0 1.2 1.1 C-14age(BP) 1285±70 1525±55 1445±60 H-3 (TU) <7.8 <7.8 <7.8 5H-2 (SMOW) -82.8 - -82.5 - -83.9 5O-18 (SMOW) -11.92 - -11.95 - -12.09 5C-13 (POB) -24.96 - -22.09 - -23.53 234U/238U (activity) 1.45 1.39 1.26 1.20 1.20 Uranine (ug/1) 15 6 11 3 5 Charge B. (%) 1.1 0.8 -0.7 -1.1 3.3 *-* Calcite -1.84 -0.06 -1.56 -1.43 -1.34 + 1.61 +0.14 +1.78 + 1.71 + 1.57 52

Table 4-3. Comparison of the copper concentrations in Hyrkkold with the copper concentration in Finnish wells according to Lahermo et al. (1990).

Hg/1 ng/i Copper in Finnish wells: Mean Median Drilled wells (bedrock) 25.3* 9.0 Dug wells (clay) 19.0 6.0 Dug wells (till) 14.5 7.0 Dug wells (sand and gravel) 11.4 6.0 Springs (clay) 19.0 4.5 Spring (till) 13.0 4.0 Springs (sand and gravel) 9.4 3.0 Copper in Hyrkkola water samples: Pg/1 HY 325, 90-104 m 13 HY 325, 67 - 68.5 m 4.5 HY325, HY325, 43.5 - 45 m 4.1 HY324, 52.4-53.9 m 3.4 HY325, 65 - 70 m 2.5 Part of these samples were taken from tap, but water was always allowed to run before sampling until the natural groundwater temperature was reached. 53

5 URANIUM SERIES ISOTOPE STUDIES OF ROCKS AND MINERALS

The study of the relative abundances of radionuclides naturally occurring in the decay series of 238U, 235U and 232Th (Fig. 5.1) attempts to determine whether or not native copper at Hyrkkola has been in contact with groundwaters and to get insights in the time-scale of groundwater-rock interactions. If the host medium has remained undisturbed during the last 1.5 Ma, the uranium and thorium decay series will be in a state of secular equilibrium. If, on the other hand, during some chemical or physical alteration of the geological medium, a daughter nuclide is separated from its parent because of differences in their chemical behaviour, a state of disequilibrium will be created. When parent and daughter radionuclides have been separated from each other, the time required to return to radioactive equilibrium is determined by the daughter half-life. Therefore, the objectives of uranium-series disequilibrium studies in the Hyrkkola U-Cu mineralization are 1) to determine both the response of actinides to the geochemical events affecting the site (e.g., dissolution- precipitation of mineral phases, adsorption and coprecipitation and transport and migration of radionuclides) and 2) to establish the time-scale of these phenomena induced by groundwater movements, which can be responsible for a state of disequilibrium. 23OTh/234U fractionation effects reflect geochemical fractionations which have occurred over the past 104 - 106 years; 226Ra-230Th couple can be used for ages ranging from 1000 to 8000 years.

Uranium-Thorium Decay Series Np U-238 Th-232 U-235

U238 U234 U23S U 0.713 Ga Pa 234 Pa 231 Pa 1.2 m J J 32 500 y Th Th234 Tit 23ft Tit 232 Th22£ Th231 Th227 24.1 d 75 000 y 13,9 Oa 1.9 y 25.6 h 4 18.6 d Ac •i 1 Ac 22 Ac 227 V |eih li. 22 y. ) Ra Ra228 Ra22'X Ra223 6.7 y 3.6 d 11.1 d Fr J 1 i Rn Rn22: Rn22() Rn219 3.82 d 54.5; 3.92 s At Y J ) Po218 Po214 Po210 Po216 Po212 Po215 Po 3.05 itj 0.2 ms Il38d 0.16 i |0.3 us .8 ms Bi214 Bi210 1 Bi212 1 Bi 211 Bi > i 19.7 m i 5 d. V |60.5 d V i 2.16 m Pb214 1 ft>2I0 1 Pb206 Pb212 Pb208 Pb211 =b207 Pb 26.8 m V 2J.4y V (stable 10.6 hi N stable 36.1 m i stable T1210 T1206 m 208 V 207 Tl 1.3 m |4.2m |3.1 m 4.79 m

Figure 5-1. 238 U, 2S5U and 2nTh decay series. 54

5.1 Rock/mineral samples and methods

Three subsamples of Hy325/68.3 (Table 3-1; Hy-a, Hy-b, and Hy-c in Table 5-1), six subsamples of B-uranophane crystals (UF1 to UF6) occurring on the open surface of Hy324/53.3, and three fractions of smectite (Hy324/97.85 in Table 3-1; S-l to S-3 in Table 5-1) were selected as sample material for uranium-series disequilibrium (USD) studies. Radionuclides were separated from the sample material using total dissolution and sequential extraction (Yanase et al. 1991, Suksi et al. 1992) and measured by a-spectrometry. The analysis for uranophane and smectite were performed in two independent laboratories.

5.2 Radiochemical data

Table 5-1 shows the radiochemical data of whole rock and mineral samples. There are deviations from secular radioactive equilibrium in all the samples. The 23OTh/234U activity ratio of all uranophane fractions is below unity (0.59 to 0.91), which indicates recent accumulation of uranium. The age was calculated assuming a thermodynamically closed system after uranophane formation. The ages estimated for groundwater circulations and oxidative precipitation of U(VI) in the form of uranophane range from 88 000 to 210 000 years (Bros et al. 1998). However, only minimum ages can be assigned using these data and the high 234U/238U ratios (up to 1.86 in UF4) may indicate recent accumulations of uranium from groundwater after the initial precipitation of uranophane (UF1). The residual smectite after sequential extraction (*, Appendix B) has a very low 234U/238U activity ratio (0.29 to 0.39) and also a low 23OTh/234U ratio (0.78).

Table 5-1. Radiochemical data of the rock and mineral samples and ages for fi-uranophane fractions.

Total U (ppm) "4U/Z^U /JUTh/"4U Age (ka) Hy-a 278 1.08±0.01 0.93±0.03 no estimations Hy-b 197 0.94±0.01 O.98±O.O3 Li Hy-c 182 1.18±0.01 0.81±0.04 about 300 UF1 5*103 1.34 0.91 210 ± 15 UF2 a 1.46 0.81 155 ± 11 UF3 1.33 0.87 188 ± 12 UF4 it 1.86 0.59 88 ± 10 UF5 1.35 0.90 203 ± 14 UF6 2*105 1.43 0.91 204 ± 15

U-content 234 238 Total U (ppm) after U/ U 230Th/234u sequential extraction [*] S-l 90.5 70.5 0.39 ± 0.02 - S-2 92.5 81.4 0.36 ± 0.01 0.79 ± 0.03 S-3 53 41 0.29 ± 0.02 0.78 ± 0.07 55

6 SUMMARY AND DISCUSSION

The main aims of this study were 1) the characterisation of the groundwater chemistry at Hyrkkola, 2) to check whether or not low temperature corrosion products occur in drill core samples, especially where native copper is in contact with groundwater, and 3) to get the timescales for the possible corrosion processes.

The groundwaters are fresh (Chapter 4), slightly acidic (6.5 < pH < 7) and very oxidising (all waters contain dissolved oxygen between 0.5 and 4 mg/1). The amounts of total dissolved solids (TDS) and the electrical conductivity (E.C.) values are similar -or even lower than- those of typical near-surface waters. . •

Native copper and its corrosion products were found in direct contact with groundwater and also in the vicinity of groundwater conducting fractures. The porosity of a sample (Figs. 3-1 and 3-2) where native copper is not within a fracture zone or fracture surfaces in direct contact with groundwater was measured by two methods in order to extrapolate the effect of groundwater chemistry on the composition of native copper. The maximum porosity value (0.63 %) was for surfaces where native copper is filling cracks and fissures. Groundwater- copper interactions are more expected in these fissures than in places where copper occurs within feldspar grains (porosity value 0.25%). The mineral assemblage native copper-copper sulfides occurs in fissures (sample Hy 325/68.35) and native copper - copper oxide (cuprite) occurs in the surface of an active fracture in direct contact with groundwater (Hy324/97.85).

Although the sulfidization of native copper is not a current process (because waters are oxidising), the temperature at which copper sulfides formed was relatively low, as indicated by the mineralogy, composition, and the 834S values of the Cu and Cu-Fe sulfides. The upper thermal stability of digenite, djurleite and low chalcocite is between 72 ± 2 °C and 103.5 ± 0.5 °C respectively (Potter 1977). The pH value at the time of sulfide-calcite precipitation in Hy324/53.30 should be about 7.8 (25 °C; Mason 1966).

The mineral assemblage native copper-cuprite (Cu(I) oxide) in sample Hy324/97.85 is representative of the present groundwater conditions, as oxidation is likely to be an on-going process, especially where dissolved oxygen is measurable. Native copper grains (0=1 mm) with rims of cuprite are in close contact with smectite. The smallest grains are totally oxidised. The contact between native copper and cuprite is not continuous. Cuprite does not form continuously solid platelets but micrograins disseminated in smectite, so that physically separated native copper grains do not show oxidation. The oxidation process may cause embrittlement in copper, at least where native copper is in contact with smectite.

As uranium is present in the system, uranium series disequilibrium studies (USD) were performed in order to determine the timescale of the processes (end of sulfidization, beginning of oxidation), and as an aid to clarify whether or not native copper and related minerals have been in contact with groundwater. Native copper, copper sulfides and copper oxides have been (and are) in contact with groundwater, since these minerals occur in the 56

same microfractures/fracture surfaces, where uranium is leached from and accumulated in alternately, due to mineral/rock-groundwater interactions. The timescale of the processes is discussed below.

Implications to Performance Assessment

Tectonically the Hyrkkola study site may be considered unsuitable for the stability of metallic copper. Hyrkkola' is situated near the vertex of a triangular bedrock block (area of about 50 km2) bordered by three regional fault zones, the nearest being the Paimio shear zone, about one kilometre to the north of the study site. Based on the interpretation of the topographic and aeromagnetic maps, a NW-SE local fracture was predicted to go through the studied area (Ahonen et al. 1997). The existence of this fracture zone was verified in this study. The drilling of both boreholes ended at about 100 m depth, the last two meters corresponding to a highly fractured zone. The composition of a groundwater sample from this zone (SI in Table 4-2) indicated a rapid recharge of oxic water. The mineral assemblage native copper-cuprite (Hy324/97.85) also corresponds to this fracture zone. The penetration of highly oxidised glacial melt water to repository depth during glaciation could cause a similar copper corrosion process (e.g. Ahonen and Vieno 1994).

Native copper filling microfractures in other samples (Hy325/68.30, Hy325/78.70) does not show oxidation products. Oxygen in groundwater may be consumed by rock-forming minerals before reaching native copper grains. In other words, the bedrock is buffering the system with respect to oxygen.

Uranyl compounds were examined in order to get insights in the mobility of copper either coming from native copper or from copper sulfides. Copper, and sulfur to a minor extent, have co-precipitated with U(VI) minerals (Tables 3-8 and 3-9). Copper and uranium may have been released in acid groundwater (Appendix A) previous to their co-precipitation in uranyl compounds.

The close contact between copper sulfides and uranyl compounds constitutes an analogue to the behaviour of copper corrosion products in a U-rich granitic environment. Although groundwater conditions are extremely oxidising, U(IV) is encountered in direct contact with copper sulfide (sample Hy324/53.30, Appendix C).

The occurrence of uranium in smectite associated with native copper and cuprite constitutes an analogue to the behaviour of the possible repository near-field materials in a natural oxidising environment. The embrittlement of copper after corrosion to copper oxide in contact with smectite (as an analogue to bentonite) has been pointed out above. Most of uranium in smectite (76 - 88 %) was sorbed in irreversible form (Appendices A and B).

Two experiments were performed on sample Hy324/97.85 (Appendix A) because of the interest in characterising the processes which may control migration and retention of rare 57

earth elements (REE), U and Cu in a naturally occurring smectite, and because analogies may be drawn between the behaviour of trivalent REE and the behaviour of trivalent actinides (Choppin 1983; Krauskopf 1986). Copper is released in acid water (pH about 4) more easily than uranium, which is known to be a very mobile element, but in Allard water (pH about 7.8) uranium is the most easily released element after scandium.

The end of sulfidization may be earlier than the precipitation of p-uranophane and other uranyl compounds in samples Hy325/68.30 and Hy324/53.30 on fracture surfaces and around the boundaries of copper sulfide grains, that is about 200 000 years ago. There is no evidence to estimate the beginning of the process.

The very low 234U/238U activity ratio (0.29 - 0.39) in the main U fraction in smectite indicates preferential removal of U-234 from the samples. Based on geometrical considerations a-recoil can not explain 234U/238U activity ratio below 0.5. The easier dissolution of U-234 can be better explained by the different oxidation states of the uranium isotopes: U-234 is in the more soluble +6 valence state whereas most of U-238 is in the less soluble +4 valence state (Appendix B). To attain and maintain such a low 234U/238U activity ratio, oxidising conditions and at least a time period comparable to the half-life of the U-234 5 isotope (TI/2 = 2.44 * 10 a) are required. That means that smectite-groundwater interaction has at least been going on for the same period of time. Therefore, the native copper grains within smectite have also been and still are in contact with oxidising circulating groundwater, and still copper is found in its native state.

I 59

7 CONCLUSIONS

At Hyrkkola native copper has been (and is) in contact with groundwater during long periods of time since it precipitated within granite pegmatite about 1700 Ma ago. This is assessed by uranium-series disequilibrium (USD) studies, as native copper occurs in the same microfractures, where uranium is leached from and accumulated in alternately, due to mineral/rock-groundwater interactions.

Copper sulfides, copper-iron sulfides, and copper oxides also have been (and are) in contact with groundwater. The chemistry of the groundwater at the time of sulfidization can be estimated through thermodynamical considerations (Ahonen 1995, Marcos 1996), as sulfidization is not a current process.

The timing of the processes (e.g., sulfidization, oxidation) can be estimated either considering the system as a whole (all fractures/fracture zones belonging to the same groundwater system) or as two different subsystems (samples Hy325/68.30 and Hy324/53.30 representing one system and sample Hy324/97.85 the other one). Considering the system as a whole, the end of sulfidization and/or remobilisation of sulfide may be earlier than precipitation/leaching of uranium in smectite, that is, earlier than 350 000 years ago and even earlier than one million years ago.

Considering the subsystem represented by samples Hy325/68.30 and Hy324/53.30, the end of sulfidization and/or remobilisation of sulfide may be earlier than the precipitation of B-uranophane and other uranyl (U6+) compounds in these samples (~ 200 000 years ago), on fracture surfaces and around the boundaries of copper sulfide grains.

The subsystem represented by sample Hy324/97.85 (S-l to S-3 in Table 5-1) may be considered as an independent system that is not related to any previous sulfidization process. The very low 234U/238U activity ratio (0.29 - 0.39) in the main uranium fraction, containing 76 - 88 % of the total U, indicates preferential removal of 234U from the samples (Appendix B). Then the oxidation process may have been active since at least 350 000 years ago and even since one million years ago. More exactly, to attain and maintain such a low 234U/238U activity ratio, chemically stable conditions and at least a time period comparable to the half-life 234 5 of the U isotope (T1/2 = 2.44* 10 a) are required.

This work contributes not only with data on the timing of corrosion processes (sulfidization, oxidation) that may happen in the final nuclear waste repository, but also addresses the influence of these corrosion products on the release and transport of uranium. It has been shown that U(VI) may be reduced to U(IV) at the contact surface of copper sulfides (Appendix C). Thus, these secondary copper corrosion products formed within reducing conditions limit the mobilisation of uranium. The interactions of uranium, copper and smectite in samples such as Hy324/97.85 could be worthy of further studies, as new data could be obtained to clarify processes in the near-field in an oxidising environment. 60

ACKNOWLEDGEMENTS

Regis Bros and Juhani Suksi provided data for Chapter 5 (Uranium series isotope studies of rocks and minerals). The EU TMR (Training and Mobility of Researchers) program supported the visit of N.M. to the Interface Analysis Centre (University of Bristol, UK). The authors thank Virginia M. Oversby and John T. Smellie for their helpful comments and suggestions. Virginia M. Oversby also checked the English of the manuscript. 61

REFERENCES

Ahonen L., 1995. Chemical stability of copper canisters in deep repository. Helsinki, Nuclear Waste Commission of Finnish Power Companies, Report YJT-95-19.

Ahonen L. & Vieno T., 1994. Effects of glacial melt water on corrosion of copper canisters. Helsinki, Nuclear Waste Commission of Finnish Power companies, Report YJT-94-13).

Ahonen L., Marcos N., Paananen M., and Paulamaki S., 1997. Research plan for field studies at the Hyrkkoia native copper mineralization, SW Finland. Helsinki. Posiva Working report 97-49e.

Ahonen L. & Marcos N., 1997. Studies of the Hyrkkoia' native copper occurrence during summer 1997. Interim report. Helsinki. Posiva Working report 97-50e.

Almen K.-E., Andersson 0., Fridh B., Johansson B.-E., Sehlstedt M., Gustafsson E., Hansson K., Olsson O., Nilsson G., Axelsen K., and Wikberg P., 1986. Site investigation. Equipment for geological, geophysical, hydrogeological and hydrochemical characterisation. SKB, Technical Report 86-16, 141 p.

Amcoff 6., 1998. Mineral Formation on Metallic Copper in a "Future Repository Site Environment": Textural Considerations Based on Natural Analogues. SKI Report 98:7, January 1998.

Baas-Becking L.G.M., Kaplan I.R., and Moore D., 1960. Limits of the natural environment in terms of pH and oxidation-reduction potentials. The Journal of Geology, 68, 243-284.

Blanchard R., 1968. Interpretation of leached outcrops: Nevada Bur. Mines Bull. 66, 196 p.

Bros R., Roos P., and Holm E., 1998. Mobilisation of U-series radionuclides at the Hyrkkoia U-Cu mineralization, Finland. SKB Utveckling Project Report U-98-13.

Choppin G.R., 1983. Comparison of the solution chemistry of the actinides and lanthanides. Journal of the Less-Common Metals 93, 232-330.

Engman U. and Hermansson H-P., 1994. Corrosion of copper materials for encapsulation of radioactive waste - a literature study (in Swedish). SKI Report 94:6, Stockholm. April 1994.

Fritz P., Fontes J.-Ch., Frape S.K., Louvat D., Michelot J.L., and Balderer, W., 1989. The isotope geochemistry of carbon in groundwater at Stripa. Geochim. et Cosmochim. Acta 53, 1765-1775. 62

Hallberg R.O., Ostlund P., and Wadsten T., 1988. Inferences from a corrosion study of a bronze cannon, applied to high level nuclear waste disposal. Applied Geochemistry, 3, p. 273-280.

Hellmuth K-H., Lukkarinen S. and Siitari-Kauppi M., 1994. Rock Matrix Studies with Carbon-14-Polymethylmethacrylate (PMMA); Method Development and Applications. Isotopenpraxis Environ. Health Stud., 30: 47-60.

Hellmuth K-H., Siitari-Kauppi M. and Lindberg A., 1993. Study of Porosity and Migration Pathways in Crystalline Rock by Impregnation with 14C-Polymethyl-methacrylate, in Journal of Contaminant Hydrology, 13, pp 403-418.

Hoare J.P., 1968. The electrochemistry of oxygen. New York, Wiley Interscience.

Johanson B. and Kojonen K., 1995. Improved electron probe microanalysis services at the Geological Survey of Finland. Geological Survey of Finland, Special Paper 10, 181-184.

King F., 1995. A natural analogue for the long-term corrosion of copper nuclear waste containers-reanalysis of a study of a bronze cannon. Applied Geochemistry, 10, 477-487.

Krauskopf K.B., 1986. Thorium and rare-earth elements as analogues for actinide elements. Chemical Geology 55, 323-335.

Lahermo P., Ilmasti M., Juntunen R., and Taka M., 1990. The Geochemical Atlas of Finland. Part 1. The hydrogeochemical mapping of Finnish Groundwater, Geol. Surv. Finland. 66 p.

Latvalahti U., 1979. Cu-Zn-Pb ores in the Aijala-Orijarvi area, Southwest Finland. Econ. Geol. 74, pp. 1035-1059.

Lide D.R. and Frederikse H.P.R. (Editors), 1995. CRC Handbook of Chemistry and Physics. CRC Press, Inc. Boca Raton.

Marcos N., 1996. The Hyrkkola native copper mineralization as a natural analogue for copper canisters. Helsinki. Posiva Oy. Report Posiva-96-15.

Mason B., 1966. Principles of Geochemistry, John Wiley & Sons, Inc. New York.

Miller W., Alexander R., Chapman N., McKinley I., and Smellie J., 1994. Natural Analogue Studies in the Geological Disposal of Radioactive Wastes. NAGRA, Technical Report 933-03, January 1994.

Morimoto N. & Koto K., 1970. Phase relations of the Cu-S system at low temperatures: Stability of anilite. American Mineralogist, 55, 106-117. 63

Morimoto N., Koto K., and Shimazaki Y., 1969. Anilite, Cu7S4, a new mineral. American Mineralogist, 54, 1256-1268.

Makela U., 1989. Geological and geochemical environments of Precambrian sulphide deposits in southwestern Finland. Ann. Acad. Sci. Fenn., Ser. A, No 151, 102 p.

Makela M. & Tammenmaa J., 1977. A system for precise sulfur isotope analysis by a small mass spectrometer. Report of Investigation No. 20, Geol. Surv. Finland. 22 p.

Nakai I., Sugitani Y., Nagashima K., and Niwa Y., 1978. X-ray photoelectron spectro- scopic study of copper minerals. J. Inorg. Nucl. Chem 40, 789-791.

Natarajan K.A. & Iwasaki I., 1974. Eh measurements in hydrometallurgical systems. Minerals Sci. Engng., 6, 35-44.

Pattrick R.A.D., Mosselmans J.F.W., Charnock J.M., England K.E.R., Helz G.R., Garner CD., and Vaughan D.J., 1997. The structure of amorphous copper sulfide precipitates: An X-ray absorption study. Geochim. et Cosmochim., 61, pp. 2023-1036.

Ploegsma M., 1989. Shear zones in the West Uusimaa area, SW Finland. Doctoral thesis, Vrije Universiteit te Amsterdam, 134 p.

Posfai M. and Busek P.R., 1994. Djurleite, digenite, and chalcocite: Intergrowths and transformations. American Mineralogist, 79, 308-315.

Potter R.W., 1977. An electrochemical investigation of the system copper-sulfur. Economic geology, 72, 1524-1542.

Roseboom E.H., Jr., 1966. An investigation of the system Cu-S and some natural copper sulfides between 25 °C and 700 °C. Economic Geology, 61, 641-672.

Ruotsalainen P., Snellman M., Helenius J., Keinonen M., Vaahtera V., Kuusela H., and Oksa M., 1994. Field manual for the water sampling of TVO (in Finnisk with an English abstract). Teollisuuden Voima Oy, Helsinki, TVO/Site Investigations, Work report PATU- 94-28.

Raisanen E., 1986. Uraniferous granitic veins in the Svecofennian schist belt in Nummi- Pusula, Southern Finland. Technical Committee Meeting on Uranium Deposits In Magmatic and Metamorphic rocks. Report IAEA-TC-521, 37-44.

Schreurs J. and Westra L., 1986. The thermotectonic evolution of a Proterozoic, low pressure, granulite dome, West Uusimaa, SW Finland. Contrib. Mineral. Petrol. 93: 236-250. 64

Schwartz M.O., 1996. Native Copper Deposits and the Disposal of High-Level Waste. International Geology Review, 18, p. 33-44.

Simonen A., 1956. Geological map of Finland 1: 100 000, sheet 2024 Somero. Explanation to the map of rocks (in Finnish with an English summary). Geological Survey of Finland, 30 p.

Suksi J., Ruskeeniemi T., and Rasilainen K., 1992. Matrix diffusion - Evidences from natural Analogue Studies at Palmottu in SW Finland. Radiochimica Acta 58/59, p. 358-393.

Tossell J. A., 1978. Theoretical studies of the electronic structure of copper sulfide in tetrahedral and triangular coordination with sulfur. Phys. Chem. Mineral. 2, 225-236. van der Laan G., Pattrick R.A.D., Henderson C.M.B., and Vaughan D.J., 1992. Oxidation state variations in copper minerals studied with Cu 2p X-ray absorption spectroscopy. J. Phys. Chem. Solids 53, 1185 -1190.

Wang N., 1994. The Cu-Bi-S system: results from low-temperature experiments. Miner- alogical Magazine, 58, 201-204.

Yanase N., Nightingale T., Payne T., and Duerden P., 1991. Uranium distribution in mineral phases of rock by sequential extraction procedure. Radiochimica Acta, 52/53, p. 387-393.

Yrjola M., 1984. Raportti Lansi-Uudenmaan kallioperatutkimuksista 1984. Unpublished report of investigations M19/2023/-84/1/60, 18 p.

Zang X. & Spry P.G., 1994. Calculated stability of aqueous tellurium species, calaverite and hessite at elevated temperatures. Economic Geology, 98, pp. 1152-1166. 65

APPENDICES 67

APPENDIX A (1/6)

COPPER, URANIUM AND TRACE ELEMENT ANALYSIS OF SMECTITE (SAMPLE HY324/97.85)

Smectite occurs on the surface of a water conductive fracture in the drill hole Hy324 at a depth of 98.85 m. It is in direct contact with native copper grains that are, in part, embedded in it. Two experiments were performed on this material because of the interest in characterising the processes which may control migration and retention of rare earth elements (REE), U and Cu in a naturally occurring smectite, and because analogies may be drawn between the behaviour of trivalent REE and the behaviour of trivalent actinides (Choppin 1983; Krauskopf 1986).

Case 1: Smectite in acid water 152.8 mg of smectite was equilibrated in distilled water for four weeks. One drop (1/20 cm3) of 2.74M of HC1 was added to 10 ml of the solution and the pH dropped to 3.8 and equilibrated for three weeks. The solution was separated into solid and liquid phases and analysed in the Chemical Laboratory of the Geological Survey of Finland. The solid phase was dissolved on HC1-HNO3-HF mixture and subsequently analysed by ICP-MS and ICP-AES. The liquid phase was analysed by ICP-MS.

Case 2: Smectite in Allard water 117.7 mg of smectite was equilibrated in Allard water during four weeks. 10 ml of solution was separated into solid and liquid phases and analysed as in Case 1. The composition and pH of Allard water is in Table Al.

Table Al. Composition of Allard water

2 2+ 2+ + + Allard water HCO3 SiO2 so; cr Ca Mg K Na jpH = 7.8 Concentra- 123 12 9.6 55 18 4.3 3 56 tion (mg/1)

Figure Al and Table A2 show the concentration of REE, U, Cu, Ba, and Mn in both phases (mineral and water) in the first case. Copper is released in acid water (pH = 3.8) more easily than uranium, which is known to be a very mobil element. In Allard water (Case 2; pH = 7.8) not even the concentration of scandium (Sc), which hydrolyses readily (Brookins 1990) is greater than that in the solid mineral phase (Fig. A2 and Table A2).

Rare earth element concentrations can be presented in a graphical way, which involves the normalisation of the concentrations in the sample to those of in a chosen reference material (i.e. the concentration of each REE in the sample is divided by the concentration of the same REE in the reference material). The plot is usually given as the logarithm of the normalised abundances versus atomic number. APPENDIX A (2/6)

As smectite is a clay mineral, the REE normalising scheme used is average shale PAAS (Post-Archaean Australian Average Shale) because the REE pattern of average shale is thought to be parallel to the average upper continental crust (e.g. Taylor and McLennan 1988). The PAAS-normalized distribution patterns for smectite and water are shown in Smectite * AcW

1000

Acid water B Smectrte 1 rra

100 -•

10- o c oo

1 ••

0.1 «4 a+ ill La Ce ft (M Sm Eu Gd Tb Dy Ho Er Tm U Yb Lu Y Sc Cu Ba Mn REE C< Figure Al. Concentration of elements (Case 1) in the solid and liquid phases.

Smectite •»- AW

10000

HAIIard water 1000 -• oSwectite 2

ioo ••

10 •• F c o 1 - - u r

0.1 ••

o.oi •-

0.001 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm U Yb Lu Y Sc Cu Ba Mn REE <- Ci Figure A2. Concentration of elements (Case 2) in the solid and liquid phases. 69

APPENDIX A (3/6)

Figure A3. The patterns show a slight enrichment in the heavy rare earth elements {HREE = Dy - Lu) with respect to the light rare earth elements (LREE = La, Pr, Nd, Sm) in both

Shate normalized REE patterns for smectite and water I.OOEt-Ot

1.O0E+O0 ;:

—#— Smectite! 1.00&01 -• •—Smectite2 < -*-Smect1. AcW < -«-Sn»ect2. AW t 1.0O&O2 ••

1.00&05 "•

1.00&06 -

1.OQE-O7 La Ce Pr Nd Sm Eu Gd Tb Dy Ho & Tm Yb tu

Figure A3. Shale normalised REE patterns for smectite and water in both cases. smectite fractions and also in Allard water. A positive Ce anomaly is present in the smectite fraction of Case 2. Eu anomalies are absent in all the patterns. In the fraction Smect2.AW, an enrichment in the intermediate REE (IREE = Sm - Ho) with respect to the LREE and also with respect to Er, Tm, Yb and Lu is seen.

The almost flat pattern for AcW may be explained by the absence of carbonate complexes affecting the dissolution of the REE. On the contrary, in Allard Water (pH = 7.8) IREE and HREE enrichment may be caused by the existence of soluble carbonate complexes of these elements. The formation of carbonate complexes with the HREEs are stronger than those with the LREEs (e.g. Millero 1992, Lee & Byrne 1993). LREE are preferentially retained by smectite in Case 2. 70

APPENDIX A (4/6)

Partition coefficients

1 .OOE+04

1.00E+03 -

T3 '5 D- :.\ ._.„•« C l.OOE+02 - :s:iiii j

1.00E+01 - ;«*"" ::X:::::::::a:^j -S1 (P) Acid water 1.00E+00 " -S2 (P) Allard water

1.00E-01 -I i -i 1 1 H 1 1 H 1 1 1 1 h- La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu U Cu Sc Y

Figure A4. Partition coefficients for REE, Cu, U, Sc, and Y.

Figure A4 shows the partition coefficient (P) between the liquid and solid phases. The solid/liquid REE exchange can be thermodynamically defined as

3+ 3+ = [Ln ]s/ [Ln ]L (1)

Figure A4 also shows that Cu is more soluble than U in acid water, but less soluble than Ho and U in Allard water. 71

APPENDIX A (5/6)

Table A2. Concentrations of REE, U, Cu, Sc, Y, Ba, and Mn in the mineral phase and water in both cases. Concentration is jlgA. in water and jlg/g in the mineral. Casel Case 2

Element Acid water Smectite 1 Allard water Smectite 2 La 19 28.1 0.007 1.2 Ce 28.3 52.5 0.047 15.5 Pr 3.97 7.26 0.004 1 Nd 16.1 31.6 0.04 6.4 Sm 2.55 9.49 0.05 5.2 Eu 0.43 2.08 0.007 1.15 Gd 2.93 15.2 0.029 8.74 Tb 0.38 2.36 0.006 1.53 Dy 1.98 14.5 0.055 9.95 Ho 0.42 3.02 0.021 1.11 Er 1.21 8.32 0.016 5.6 Tm 0.15 1 0.0024 0.7 U 72.6 75.8 16.6 69.6 Yb 0.94 5.63 0.026 4.07 Lu 0.14 0.83 0.0025 0.6 Y 16.5 117 0.271 78.6 Sc 13.1 6.27 3.1 10 Cu 52.1 30.4 3.55 342 Ba 183 783 13.3 1040 Mn 291 300 2 211

REFERENCES:

Brookins D.G., 1990. Aqueous geochemistry of rare-earth elements. In B.R. Lipin & G.A. Me Kay (eds.). Reviews in Min. 21, 201 - 225. Choppin G.R., 1983. Comparison of the solution chemistry of the actinides and lanthanides. Journal of the Less-Common Metals 93, 232-330. 72

APPENDIX A (6/6)

Krauskopf K.B., 1986. Thorium and rare-earth elements as analogues for actinide elements. Chemical Geology 55, 323-335. Lee J.H. and Byrne, R.H., 1993. Complexation of the trivalent rare-earth elements (Ce, Eu, Gd, Tb, Yb) by carbonate ions. Geochim. et Cosmochim. Acta, 57, 295-232. Millero F.J., 1992. Stability constants for the formation of rare-earth inorganic complexes as a function of ionic strength. Geochim. et Cosmochim. Acta, 56, 3123-3132. Taylor S.R. and McLennan S.M., 1988. The significance of the rare-earths in geochemistry and cosmochemistry. In: K.A. Gschneidner Jr. & L. Eyring (Eds.), Handbook on the Physics and chemistry of Rare Earths, Vol. 11, 485-578, Elsevier, Amsterdam. 73

APPENDIX B (1/2)

SEQUENTIAL EXTRACTION (SE) METHODS FOR SMECTITE (SAMPLE HY324/97.85)

Uranium was sequentially extracted from three fractions S-1 to S-3 from the smectite sample. The uranium in samples S-1 and S-2 was extracted by (1) NH4-acetate buffer solution at pH 4 (AA), (2) Tamm's reagent (NH4 oxalate + oxalic acid at pH 3.5) (TAO), and (3) dissolution with HF-HNO3-HC1 mixture (RE) (Table Bl). In the case of sample S-3 a modified procedure was used. This sample was first equilibrated with an artificial groundwater, Allard water (AW), and after that the AA and TAO extractions were performed. Before extraction in the HC1-HNO3-HF mixture, extraction with 1M HN03 (5 h at room temperature) was performed (AN). The meaning of the sequential extraction methods for these samples is discussed below.

Artificial groundwater was used to obtain information of adsorbed U. Ammonium acetate (AA) is used to remove adsorbed uranium and U associated with calcite. To separate U associated with amorphous Fe/Mn oxides and in possible secondary U minerals TAO extraction is performed. In case of sample S-3 1M HNO3 was chosen to complete the TAO extraction.

Table Bl. Radiochemical data for smectite fractions.

Sample Total U (ppm) U-content ^U/^U ijhrh//j4u (ppm) after SE S-1 90.5 - S-l(AA) 5.5 1.51 ±0.01 0.40 ± 0.05 S-1 (TAO) 14.3 0.48 ± 0.01 2.38 ± 0.03 S-1 (RE) 70.5 0.39 ± 0.02 0.78 ± 0.03 S-2 92.5 S-2(AA) 4.5 1.49 ±0.01 - S-2(TAO) 6.1 0.85 ± 0.02 - S-2(RE) 81.4 0.36 ± 0.01 0.79 ± 0.07 S-3 53 S-3(AW) 0.46 1.4 ± 0.7 - S-3(AA) 3.2 1.58 ±0.13 0.45 ± 0.06 S-3(TAO) 2.9 0.86 + 0.10 3.2 ± 0.4 S-3 (AN) 41 0.29 ± 0.02 0.78 ± 0.07 S-3(RE) 6.4 1.12 ±0.06 1.6 ±0.1

In all samples, in spite of variations in U concentration, the same U-234/U-238 activity ratio (A.R.) is observed in the AA fractions. The AA fraction is most accessible to groundwater and is assumed to acquire its U (A.R.) signature from the groundwater. Interestingly, a similar activity ratio is obtained in the groundwater. In samples S-1 and S-3 the same Th-230/U-234 activity ratios (clearly below equilibrium value of unity) was obtained, indicating recent similar/coeval U accumulation process in the fracture coating smectite. The clear decrease in the U-234/U238 and increase in the Th-230/U-234 activity ratios in the TAO fractions is noteworthy and suggest dissolution of U phases with different characteristics. Most of the U still remains fixed after the AA and TAO extraction. In samples S-1 and S-2 this fraction of U was totally released in the third and 74

APPENDIX B (2/2) last extraction. Before total dissolution of the S-3 residue, an extraction with 1M HNO3 was used. Over 70 % of the total U and also smectite dissolve in this extraction step. Only minor quantities of U and traces of smectite were found in S-3 (RE) (XRD analyses of S- 3 (RE) showed that the residue was composed mostly of quartz and K-feldspar and only some traces of smectite). Consequently, the RE-fraction in sample S-3 is different than the RE-fraction in samples S-l and S-2 where most of U still exists.

The slightly higher activity ratios for RE fractions in samples S-l and S-2 (0.36 and 0.39 respectively) than for the fraction AN of S-3 (0.29) is explained by the fact that fractions S-l (RE) and S-2(RE) include minor quantities of quartz and K-feldspars. The 234U/23SU activity ratio for S-3(RE) is 1.12, which would compensate for the lower activity ratio of S-3(AN).

The very low U-234/U-238 activity ratio in the main U fraction, containing 76 - 88 % of the total U, indicates preferential removal of U-234 from the samples. This, in turn, indicates that congruent dissolution of the U fraction does not take place in the system. The easier dissolution of U-234 can be explained by the different oxidation states of the U isotopes: U-234 is in the more soluble +6 valence state whereas most of U-238 is in the less soluble +4 valence state. Possible oxidation from U4+ to U6+ is related to the oxidation potential difference between the displaced site and the original site of the recoiling U atom or to the radioactive decay process itself (e.g. Ordonez Regil et al. 1989, Adloff & Roessler 1991). To attain and maintain such a low U-234/U-238 activity ratio, chemically stable conditions and at least a time period comparable to the half-live of the U- s 234 isotope (T1/2 = 2.44* 10 a) are required.

REFERENCES

Adloff J.P. and Roessler K., 1991. Recoil and Transmutation Effects in the Migration Behaviour of Actinides. Radiochimica Acta 52/53, 269-274. Ordonez Regil E., Schleiffer J.J., and Adloff J. P., 1989. Chemical Effects of a-Decay in Uranium Minerals. Radiochimica Acta 47, 177-185. 75

APPENDIX C (1/4)

PRELIMINARY RESULTS OF A RESEARCH PROJECT AT THE INTERFACE ANALYSIS CENTRE (UNIVERSITY OF BRISTOL, UK): IMPLICATIONS FOR MINERAL-WATER INTERACTIONS AND FOR THE INTERPRETATION OF THE RESULTS OBTAINED BY SEQUENTIAL EXTRACTION (SE) METHODS

Background

It is known that the surface of minerals (sulphides, oxides) may adsorb and/or reduce metals in solution (e.g., Jean and Bancroft 1982, Bancroft and Hyland 1990, White and Peterson 1996, and references therein). Marcos (1996, p. 25) pointed out that uranium is significantly sorbed at the copper sulphide grain boundaries. The sorption mechanism (physical or chemical adsorption) is not known, but it has been suggested that reactions at the sulphide-groundwater interface can significantly contribute to the precipitation of metals on the surface of sulphides (e.g. Jean and Bancroft 1986, Starling et al. 1989, Becker et al. 1997). The characterisation of these sorption/redox reactions has become possible using surface sensitive techniques such as X-ray photoelectron spectroscopy (XPS) and Raman Spectroscopy among others.

Instrumentation

The Interface Analysis Centre at the University of Bristol uses a Kratos XS AM800 X-ray Photoelectron Spectrometer fitted with a dual Al/Mg x-ray source together with a Vacuum Generators imaging XPS instrument (ESCASCOPE).

X-ray photoelectron spectroscopy (XPS) is a technique that is capable of detecting all elements in the periodic table from the top few atom layers in a surface. The technique is made quantitative by the application of sensitivity factors to the measured peak areas and it has a sensitivity for most elements of 0.1%. By detecting the small changes in peak position that occur when elements combine, chemical state information can be obtained. Until recently it has not been possible to provide any significant lateral spatial resolution. Fisons Instruments Surface Systems have developed an XPS instrument that images the surface by detecting electrons emitted at a particular energy in a manner similar to an optical microscope.

Technique: The Production of Photo electrons

The specimen surface is irradiated with x-rays of known energy (most commonly 1486 or 1254 eV), causing electrons to be ejected from the surface as shown in Figure Cl. The kinetic energy of these so called photoelectrons is measured using a hemispherical analyser. Note, the escape depth of these electrons is around 10 atomic layers (2-5 nm) and so the technique is extremely surface specific. Binding energy values are calculated using the following equation: 76

APPENDIX C (2/4)

Electron Binding Energy "BE" = X-ray Energy - Electron Kinetic Energy "KE"

The binding energy is characteristic of both the electronic shell and the oxidation state of the atom from which the electron was ejected. Therefore, XPS can be used for both elemental and chemical analysis of the sample surface. There is no matrix effect and by applying sensitivity factors, a quantitative analysis can be obtained.

Is K

Figure 1C. The production of photoelectrons. This figure was found in this address URL:

More data on the instrumentation and the technique can be search at the following address URL:

Results of an XPS study on fracture coating material in sample H\324/53.30

The fracture coating material consists of a discontinuous thin double layer of calcite and 8- uranophane (Fig. 3.6 in Chapter 3 is drawn schematically in Fig. 2C below). This double layer is on the surface of the rock (granite pegmatite) in general and particularly on the surface of copper sulphide grains. A thin double layer of calcite and 6-uranophane was examined by the XPS method.

1 2 3 i I 4, B-U R c A Groundwater (Sample S5) A N Copper sulfide L O Chalcocite/ C P djurleite Figure 2C. Schematic drawing of the double I H mineral layer in sample Hy324/53.30. Numbers T A show the interfaces involved in the system. 1. E N Rock and mineral-mineral/groundwater interface. E 2. Mineral/groundwater interface 3. Mineral/groundwater interface APPENDIX C (3/4)

Calcite and B-uranophane were analysed by XPS together, as they could not be physically separated. The results showed that 25 % of the uranium content exists as U(IV) and 75% is as U(VI) [U 1 stands for U(IV) and U 2 stands for U(VI) in Appendix C (4/4)]. The uranium in uranophane is thought to be-all as U(VI); thus, U(FV) would correspond to the uranium content in calcite, which is in direct contact with copper sulphide. Further studies of similar samples could be addressed to clarify the mechanism/s of uranium reduction at the copper sulphide-groundwater interface.

The interpretation of the results of 234U/238U activity ratios obtained by sequential extraction (SE) methods should be done taking into account the possible coexistence of U(VI) and U(IV) in the analysed phases (see also Appendix B).

REFERENCES

Bancroft G.M & Jean G., 1982. Gold deposition at low temperature on sulphide minerals. Nature 298, 730-731. Bancroft G.M. & Hyland M.M., 1990. Spectroscopic studies of Adsorption/Reduction of Aqueous Metal Complexes on Sulphide surfaces. In Mineral-Water Interface Geochemistry (Eds. M.F. Hochella Jr. and A.F. White). Amer. Mineral. Soc, Rev. in Miner. 23, p. 511-558. Marcos N., 1996. The Hyrkkola native copper mineralization as a natural analogue for copper canisters. Report Posiva-96-15, Posiva Oy, Helsinki. Jean G. & Bancroft G.M., 1986. Heavy metal adsorption by sulphide mineral surfaces. Geochim. et Cosmochim. Acta, p. 1455-1463. White A.F. & Peterson M.F., 1996. Reduction of aqueous transition metal species on the surfaces of Fe(II) containing oxides. Geochim. et Cosmochim. Acta 60, p. 3799- 3814. 78

APPENDIX C Interface Analysis Centre, Bristol Peak Synthesis V.G.Scientific FIN4B.DA! Region 2 I 2 Level 1 /1 Point 1

Peak Centre FWHM Hght G/L Urea lev) lev)' I I I

0 1 386.0 1.83 28 98 25 U 2 387.3 1.83 84 98

100% Height (Counts) 423 100% Area (kceV/sec) 0.1073 410 405 400 395 390 385 380 375 370 Reduced Chi Squared 1.20 Binding Energy / eV

KHK Ui SOUK; ]50u»/20ui 3m slit; Zoai 2 LIST OF REPORTS 1(5)

POSIVA REPORTS 1999, situation 5/99

POSIVA 99-01 Measurement of thermal conductivity and diffusivity in situ: Literature survey and theoretical modelling of measurements Ilmo Kukkonen, Ilkka Suppala Geological Survey of Finland January 1999 ISBN 951-652-056-1

POSIVA 99-02 An overview of a possible approach to calculate rock movements due to earthquakes at Finnish nuclear waste repository sites Paul R. LaPointe, Trenton T. Cladouhos Golder Associates Inc., Washington, USA February 1999 ISBN951-652-057-X

POSIVA 99-03 Site scale groundwater flow in Olkiluoto Jari Lofman VTT Energy March 1999 ISBN 951-652-058-8

POSIVA 99-04 The psychosocial consequences of spent fuel disposal Jura Paavola, Liisa Erdnen University of Helsinki Department of Social Psychology March 1999 (in Finnish) ISBN 951-652-059-6

POSIVA 99-05 The effects of the final disposal facility for spent nuclear fuel on regional economy Seppo Laakso Seppo Laakso Urban Research March 1999 (in Finnish) ISBN951-652-060-X

POSIVA 99-06 Radwaste management as a social issue Ismo Kantola University of Turku Department of Sociology March 1999 (in Finnish) ISBN 951-652-061-8

POSIVA 99-07 Safety assessment of spent fuel disposal in Hastholmen, Kivetty, Olkiluoto and Romuvaara - TILA-99 Timo Vieno, Henrik Nordman VTT Energy March 1999 ISBN 951-652-062-6 LIST OF REPORTS 2(5)

POSIVA 99-08 Final disposal of spent nuclear fuel in Finnish bedrock - Hastholmen site report Pekka Anttila, Fortum Engineering Oy Henry Ahokas, Fintact Oy Kai Front, VTT Communities and Infrastructure Heikki Hinkkanen, Posiva Oy Erik Johansson, Saanio & Riekkola Oy Seppo Paulamdki, Geological Survey of Finland Reijo Riekkola, Saanio & Riekkola Oy Jouni Saari, Fortum Engineering Oy Pauli Saksa, Fintact Oy Margit Snellman, Posiva Oy Liisa Wikstrom, Posiva Oy Antti Ohberg, Saanio & Riekkola Oy March 1999 (to be published) ISBN 951-652-063-4

POSIVA 99-09 Final disposal of spent nuclear fuel in Finnish bedrock - Kivetty site report Pekka Anttila, Fortum Engineering Oy Henry Ahokas, Fintact Oy Kai Front, VTT Communities and Infrastructure Eero Heikkinen, Fintact Oy Heikki Hinkkanen, Posiva Oy Erik Johansson, Saanio & Riekkola Oy Seppo Paulamdki, Geological Survey of Finland Reijo Riekkola, Saanio & Riekkola Oy Jouni Saari, Fortum Engineering Oy Pauli Saksa, Fintact Oy Margit Snellman, Posiva Oy Liisa Wikstrom, Posiva Oy Antti Ohberg, Saanio & Riekkola Oy March 1999 (to be published) ISBN 951-652-064-2

POSIVA 99-10 Final disposal of spent nuclear fuel in Finnish bedrock - Olkiluoto site report Pekka Anttila, Fortum Engineering Oy Henry Ahokas, Fintact Oy Kai Front, VTT Communities and Infrastructure Heikki Hinkkanen, Posiva Oy Erik Johansson, Saanio & Riekkola Oy Seppo Paulamdki, Geological Survey of Finland Reijo Riekkola, Saanio & Riekkola Oy Jouni Saari, Fortum Engineering Oy Pauli Saksa, Fintact Oy Margit Snellman, Posiva Oy Liisa Wikstrom, Posiva Oy Antti Ohberg, Saanio & Riekkola Oy March 1999 (to be published) ISBN 951-652-065-0 LIST OF REPORTS 3(5)

POSIVA 99-11 Final disposal of spent nuclear fuel in Finnish bedrock - Romuvaara site report Pekka Anttila, Fortum Engineering Oy Henry Ahokas, Fintact Oy Kai Front, VTT Communities and Infrastructure Heikki Hinkkanen, Posiva Oy Erik Johansson, Saanio & Riekkola Oy Seppo Paulamdki, Geological Survey of Finland Reijo Riekkola, Saanio & Riekkola Oy Jouni Saari, Fortum Engineering Oy Pauli Saksa, Fintact Oy Margit Snellman, Posiva Oy Liisa Wikstrom, Posiva Oy Antti Ohberg, Saanio & Riekkola Oy March 1999 (to be published) ISBN 951-652-066-9

POSIVA 99-12 Site scale groundwater flow in Hastholmen Jari Loftnan VTT Energy May 1999 ISBN 951-652-067-7

POSIVA 99-13 Regional-to-site scale groundwater flow in Kivetty Eero Kattilakoski VTT Energy Ferenc Meszdros The Relief Laboratory, Harskiit, Hungary April 1999 ISBN 951-652-068-5

POSIVA 99-14 Regional-to-site scale groundwater flow in Romuvaara Eero Kattilakoski, Lasse Koskinen VTT Energy April 1999 ISBN 951-652-069-3

POSIVA 99-15 Site-to-canister scale flow and transport in Hastholmen, Kivetty, Olkiluoto and Romuvaara Antti Poteri, Mikko Laitinen VTT Energy May 1999 ISBN 951-652-070-7

POSIVA 99-16 Assessment of radiation doses due to normal operation, incidents and accidents of the final disposal facility Jukka Rossi, Heikki Raiko, Vesa Suolanen, Mikko Ilvonen VTT Energy March 1999 (in Finnish) ISBN 951-652-071-5 LIST OF REPORTS 4(5)

POSIVA 99-17 Assessment of health risks brought about by transportation of spent fuel Vesa Suolanen, Risto Lautkaski, Jukka Rossi VTT Energy March 1999 (in Finnish) ISBN 951-652-072-3

POSIVA 99-18 Design report of the disposal canister for twelve fuel assemblies Heikki Raiko VTT Energy Jukka-Pekka Salo Posiva Oy May 1999 ISBN 951-652-073-1

POSIVA 99-19 Estimation of block conductivities from hydrologically calibrated fracture networks - description of methodology and application to Romuvaara investigation area Auli Niemi1'2, Kimmo Kontio2, Auli Kuusela-Lahtinen2, Tiina Vaittinen2 'Royal Institute of Technology, Hydraulic Engineering, Stockholm 2VTT Communities and Infrastructure March 1999 ISBN951-652-074-X

POSIVA 99-20 Porewater chemistry in compacted bentonite Arto Muurinen, Jarmo Lehikoinen VTT Chemical Technology March 1999 ISBN 951-652-075-8

POSIVA 99-21 Ion diffusion in compacted bentonite Jarmo Lehikoinen VTT Chemical Technology March 1999 ISBN 951-652-076-6

POSIVA 99-22 Use of the 14C-PMMA and He-gas methods to characterise excavation disturbance in crystalline rock Jorma Autio, Timo Kirkkomdki Saanio & Riekkola Oy Marja Siitari-Kauppi University of Helsinki Laboratory of Radiochemistry Jussi Timonen, Mika Laajalahti, Tapani Aaitonen, Jani Maaranen University of Jyvaskyla Department of Physics April 1999 ISBN 951-652-077-4 LIST OF REPORTS 5 (5)

POSIVA 99-23 New data on the Hyrkkola U-Cu mineralization: The behaviour of native copper in a natural environment Nuria Marcos Helsinki University of Technology Laboratory of Engineering Geology and Geophysics Lasse Ahonen Geological Survey of Finland May 1999 ISBN 951-652-078-2

POSIVA 99-24 Dissolution of unirradiated UO2 fuel in synthetic groundwater - Final report (1996 - 1998) Kaija Ollila VTT Chemical Technology May 1999 ISBN 951-652-079-0