• No results found

Hydrothermal alteration and lithogeochemical marker units at the Svärdsjö Zn-Pb-Cu deposit, Bergslagen, Sweden, and their implications for exploration

N/A
N/A
Protected

Academic year: 2022

Share "Hydrothermal alteration and lithogeochemical marker units at the Svärdsjö Zn-Pb-Cu deposit, Bergslagen, Sweden, and their implications for exploration"

Copied!
29
0
0

Loading.... (view fulltext now)

Full text

(1)

Hydrothermal alteration and

lithogeochemical marker units at the Svärdsjö Zn-Pb-Cu deposit, Bergslagen,

Sweden, and their implications for exploration

Anton Fahlvik

Geosciences, master's level (120 credits) 2018

Luleå University of Technology

Department of Civil, Environmental and Natural Resources Engineering

(2)

Abstract

In exploration, a lithogeochemical approach can be used to aid the characterisation of rocks surrounding metamorphosed and hydrothermally altered deposits. Accurate description of the geological setting of deposits is crucial for understanding the ore forming processes and identifying targets for exploration. The Svärdsjö Zn-Pb-Cu deposit is located in the heavily mineralised and metamorphosed Bergslagen ore province of south-central Sweden. The deposit and surrounding minor occurrences were actively mined for over 500 years, producing more than 1 Mt of Zn-Pb-Cu-Ag massive sulphide ore. The combination of strongly metamorphosed and hydrothermally altered rocks in Svärdsjö makes geological interpretation challenging. Therefore, an approach combining lithogeochemical and petrographic methods is used in this study. The characterisation of the rocks and hydrothermal alteration surrounding the deposit allowed for an interpretation of ore formation and its implications for further exploration in the Svärdsjö area. The results verified that the Svärdsjö mineralisations are hosted by 2–15 m thick dolomitic marble units, commonly altered to skarn. Surrounding the deposit are subvolcanic intrusions and volcanoclastic rocks of mainly dacitic composition. The combined approach also helped identifying a strong to intense hydrothermal chlorite-sericite alteration enveloping the mineralised marble units and resulted in large mass gains of Fe and Mg whereas Na was depleted. Multiple episodes of alteration and metamorphism are evident from cross- cutting relationships with less altered dykes and overprint by metamorphic minerals such as cordierite and anthophyllite. An ore formation model involving sub-seafloor volcanic-associated replacement is suggested for the Svärdsjö deposit based on (i) the presence of a zoned hydrothermal alteration system within a volcanoclastic rock sequence and (ii) the irregular stratabound sulphide lenses hosted by thin marble units in the centre of the alteration system. Additionally, it is inferred that the stratabound nature of the deposit is caused by the neutralisation of a hot acidic fluid, resulting in precipitation of the sulphides within the marble.

Finally, two geochemically distinct lithological units have been identified adjacent to the mineralised zones, providing new, larger exploration targets in the area. Mass change calculations reveal that Fe and Mg enrichment and Na depletion are useful vectors towards mineralisation, with detectable changes extending for up to 100 m from the mineralised lenses. These findings showcase the usefulness of the incorporation and careful interpretation of lithogeochemical data when exploring for metamorphosed hydrothermal ore deposits in mineralised provinces of the Fennoscandian Shield or elsewhere in the world.

Front page: Photo of the Norra Kompanigruvan pit at the Svärdsjö deposit.

1 Introduction

The Svärdsjö Zn-Pb-Cu sulphide deposit is situated in the Palaeoproterozoic ore province of Bergslagen, south-central Sweden. Extensive mining over the last centuries has made the province a major metal producer with more than 6000 historical mine sites and three currently active mines (Allen et al., 1996; Stephens et al., 2009). Bergslagen is located in the south-western part of the Svecokarelian orogen in the Fennoscandian Shield. Felsic plutonic rocks of 1.8–1.9 Ga age are dominant (Allen et al., 1996), with slivers of volcanic and subvolcanic rocks of similar age occurring as inliers in the plutonic

Test

rocks (Stephens et al., 2009). Greenschist- to amphibolite facies metamorphism and deformation associated with the Svecokarelian orogeny is present throughout the province (Stephens et al., 2009). Mining in the Svärdsjö area started during the 15

th

century and the Svärdsjö deposit itself has been mined since the mid 1700’s. Mining operations closed down in 1989, but there is potential for extensions of the known mineralisations at depth and for new mineralisations to be discovered in the area.

Although there is ongoing exploration at the

(3)

2 Svärdsjö deposit, very little research has been conducted since the mid 1980’s.

Exploration drilling has revealed a continuation of the Svärdsjö deposit towards the south-west. Two mineralised units, herein referred to as the Fäbodgruvan mineralisation and the Vilnäset mineralisation (figure 2) are the focus of this study. The extent and relationship between these units and the Svärdsjö deposit is not yet fully understood. The mineralisations are semi-massive to massive sulphides with sphalerite, galena and chalcopyrite as the main ore minerals.

In this study, a geochemical approach is used to characterise the rocks and alteration surrounding the deposit. The specific research objectives are to; (i) determine the nature and extent of hydrothermal alteration and how it relates to mineralisation, (ii) suggest a genetic model of ore formation, (iii) evaluate the potential of geochemistry and alteration as vectors towards mineralisation in the Svärdsjö area.

Increasing the understanding of this deposit and identifying potential vectors to mineralisation will aid further exploration in the Svärdsjö area and potentially also at similar deposits in Bergslagen and in other metamorphic ore provinces around the world where geological interpretations may otherwise be difficult.

1.1 Geological setting and background The geological units at the Svärdsjö deposit are part of the Bergslagen lithotectonic unit (BLU; figure 1). The BLU is made up of rocks belonging to the Svecokarelian orogen, which represents a major part of the Fennoscandian shield, stretching through most of eastern Sweden (figure 1) and western Finland. The Svecokarelian orogen formed by intense igneous activity, followed by ductile deformation and metamorphism in the interval from 2.0 to 1.8 Ga (Stephens et al. 2009; Stephens and Andersson, 2015). The rocks of the BLU were interpreted to be formed in the continental back-arc of an active subduction zone (Allen et al., 1996) where

granitic intrusions and felsic caldera volcanism have produced the majority of the rocks.

Metasedimentary turbidites and quartzites constitute the lowermost stratigraphic unit of the BLU and mark the beginning of the Svecokarelian orogeny (Stephens et al., 2009).

Pyroclastic flows from volcanic centres deposited thick (1–8 km) sequences of volcanoclastic breccia and sandstones on the metasediments during the stage of intense felsic caldera volcanism (Allen et al., 1996). Proximal to volcanic centres, subvolcanic intrusions occurred, emplacing rhyolitic porphyry rocks into the volcanoclastic sequence (Allen et al., 1996). A shallow marine setting of the sequence has been inferred by the interpretation of stromatolitic limestones and current affected sediments within the metamorphosed volcanoclastic rocks.

Subsidence during the waning stages of volcanism facilitated sedimentation of finely bedded volcanoclastic silt- and sandstones with abundant stromatolitic limestone interbeds in moderately shallow, below storm wave base waters distal to active volcanic centres (Allen et al., 1996). This more fine-grained upper sequence of the volcanic rocks hosts the vast majority of Bergslagen’s mineral deposits (Stephens et al., 2009).

Quartz-rich plutons of granitic to dioritic to gabbroic (GDG) composition intruded the volcanic rock suite shortly after its emplacement.

In central and western Bergslagen these intrusions are the dominating rock type with only minor areas of volcanic rocks preserved as inliers and belts in between (Stephens et al., 2009; Beunk and Kuipers, 2012). A second set of intrusions into the volcanic and GDG suite of rocks occurs mainly along the southern and western borders of the BLU. The composition of these rocks is different from the GDG intrusions and varies from granite and syenite to diorite and gabbro (GSDG; Stephens et al., 2009). Granitic intrusions and pegmatite veins (GP) commonly cross-cut the older rock suites and are found scattered around most of Bergslagen (figure 1;

Sundblad et al., 1993; Stephens et al., 2009).

Regional metamorphism at amphibolite facies

has overprinted most rock textures, although,

(4)

3 minor areas of greenschist facies overprint including well-preserved primary rock features do occur (Stephens et al., 2009). Synvolcanic hydrothermal alteration has affected mineralised parts of the volcanic rock suite. Regional sodic- potassic alteration is also common around mineralised areas, but less widespread in barren rock, whereas skarn alteration is present in calcareous parts of the metavolcanic rock suite (Trägårdh, 1991). The present skarn mineral assemblage is thought to have formed by both

hydrothermal alteration and subsequent regional metamorphism and is often associated with Fe- oxide and base metal sulphide mineralisation (Jansson and Allen, 2015). Two phases of ductile deformation and folding of extensional to transtensional character have been identified in the Bergslagen lithotectonic unit, although the pervasivity of the resulting foliation and folding varies throughout the BLU (Stephens and Andersson, 2015).

Figure 1. Regional geology of Bergslagen, the location of Svärdsjö and major deposits are marked by large

dots and text. Smaller deposits are marked by small red dots. Modified after Stephens et al. (2009).

(5)

4 Svärdsjö has previously been classified as a Falu-type deposit (Levi et al., 1980), one of two types into which deposits in Bergslagen have been subdivided (e.g. Magnusson, 1953;

Sundblad, 1994). The Falu-type also corresponds largely to the stratabound volcanic-associated limestone-skarn-hosted (SVALS) type in Allen et al. (1996), where a division of deposits based on volcanological facies was conducted. A recent geological study at the Falun deposit, located ca.

20 km from Svärdsjö, suggests a syn-volcanic hydrothermal system with subsequent metamorphism and deformation as a model of ore formation (Kampmann et al., 2016a; Kampmann et al., 2016b). The mineralised fluid was precipitated in a sub-seafloor setting replacing mainly volcanic rocks and minor limestone (Kampmann et al., 2016b) during a period of high volcanic activity at 1.89 Ga (Kampmann et al., 2016c).

1.2 Local geology

The Svärdsjö deposit occurs in a 2–3 km wide suite of felsic metavolcanic and metasedimentary rocks bordered by granitic intrusions (Levi et al., 1980; Bergman and Sundblad, 1981).

Mineralisations in the area are situated along two parallel northeast-southwest belts, the western one of which hosts the Svärdsjö deposit (figure 2).

The eastern part contains a number of smaller deposits in a mica schist-dominated area (Bergman and Sundblad, 1981). Multiple deformation zones with a northeast-southwest strike are present throughout the area and aligned with the main orientation of the known mineralised bodies. Regional metamorphism at amphibolite facies and subsequent retrograde metamorphism at greenschist facies overprinted the rocks, including the hydrothermally altered zones which formed during and immediately after the emplacement of the volcanic rocks (Levi et al., 1980; Stephens et al., 2009).

Mining at Svärdsjö has been performed during several periods over the last 600 years, particularly during the 19

th

and 20

th

centuries (Sundberg et al., 2015). The mine produced approximately 1 Mt of ore from 1951 until 1989, with average grades of 5.96 wt % Zn, 2.68 wt %

Pb, 0.62 wt % Cu and 111.6 g/t Ag. The Svärdsjö deposit consist of a series of mineralised lenses occurring along a skarn-bearing marble unit striking NE–SW (Billström, 1980; Levi et al., 1980). The main ore minerals mined were sphalerite, galena and chalcopyrite, although mineralogical variations can be observed along strike. At the north-eastern end the ore is rich in pyrite and pyrrhotite, in the central part sphalerite, galena and chalcopyrite dominate and at the south-western end the main ore minerals are sphalerite and galena, along with an increased abundance of magnetite (Billström, 1980).

Levi et al. (1980) described the Svärdsjö deposit as a typical Falu-type ore, with lenses of massive Cu-Pb-Zn-Fe sulphides in a skarn- bearing dolomite host rock. Several generations of hydrothermal alteration, metamorphism and tectonic movement have been suggested to have affected the deposit and the surrounding rocks. A previous study at the nearby Boviksgruvan (figure 2) identified two separate types of mineralisation;

Zn-Pb-Cu rich sulphides similar to those of the

Svärdsjö deposit, as well as Au-Bi-(Te)

mineralisation (Bergman and Sundblad, 1991). It

has been argued that the Au-Bi-(Te)

mineralisation is epigenetic, but formed shortly

after the emplacement of the volcanic rock

sequence and the stratiform sulphides. However,

the lead isotope composition of the two ore types

is similar, so that a formation by remobilisation of

the sulphide ore cannot be excluded (Bergman

and Sundblad, 1991).

(6)

5

2 Method

2.1 Core logging and sampling

Eight exploration drill holes located 100–500 m south-southeast of the Svärdsjö deposit were selected for investigation in this study (figure 2).

Five of these form a section through the Fäbodgruvan- and Vilnäset mineralisations, two drill holes are located to the northeast of the section and one to the southwest, along the general strike of the Svärdsjö deposit. Six of the eight drill cores have been mapped in this study, geological information from the remaining two cores is based on previous logs generated by geologists at the mining and exploration company Boliden Mineral AB. 85 samples of the major rock units were acquired for geological characterization, as well as thin section and lithogeochemical analysis. Mineralised samples were generally avoided, even though this was usually not possible within the skarn-bearing marble units. All samples used in this study were taken from diamond drill cores, a minimum of 25 cm whole core of BQ size (36.5 mm diameter) or 20 cm of NQ size (47.6 mm diameter) was taken for lithogeochemical analysis. In addition, a

reference sample was taken adjacent to each sample.

2.2 Lithogeochemical analysis

Core samples were dried, crushed to 70% less than 2 mm, split to 250 g sample and pulverized to 85 % less than 75 microns according to international standards at ALS Minerals, Piteå, Sweden before being shipped to Bureau Veritas, Vancouver, Canada, for lithogeochemical analysis. Whole rock major and trace elements were analysed using a combination of methods.

For major oxides, lithium borate fusion with inductively coupled plasma emission spectrometry (ICP-ES) was used, whereas trace elements were analysed using inductively coupled plasma mass spectrometry (ICP-MS).

Base metals and Au-associated trace elements were digested in an aqua regia solution (HNO

3

- HCl-H

2

O) and analysed using a combination of ICP-ES and ICP-MS. Carbon and S were analysed using combustion furnace infrared detection (LECO).

For quality control, two Boliden and five Bureau Veritas certified reference materials were interspersed randomly with the samples.

Additionally, duplicate analysis of ten samples

Figure 2. Local geology at the Svärdsjö deposit. Approximate location of Fäbodgruvan- and Vilnäset mineralisations

are indicated. Modified after Sukotjo et al., (2005).

(7)

6 and one reference material was conducted.

Results showed a sample deviation of less than 5% for all the major oxides and relevant trace elements. Reference material analyses deviated less than 2% from the expected value for major oxides and less than 6% for trace elements except for Ga, U and Lu, which deviated less than 10%.

Lithogeochemical data was analysed and graphs produced using the softwares ioGAS and Excel. 3D visualisation of the data was done using Leapfrog Geo. Immobile element techniques as described by MacLean and Barrett (1993) and Barrett and MacLean (1994) were used to identify precursors to the altered rocks, their magmatic affinity and to calculate mass changes during hydrothermal alteration. All calculations were done using the volatile-free weight percent of major oxides and ppm of trace elements. Mass change calculations followed a multiple precursor system approach (Barrett and MacLean, 1994), in which least altered samples were used to determine magmatic fractionation trends of the co-magmatic rocks. After establishing trends as a function of Zr for all major oxides, mass changes of these elements were calculated for each sample. To verify the immobility of Zr in the Svärdsjö rocks mass change calculations of another immobile element, Ti (measured as TiO

2

), were also conducted. The results showed very minor mass changes of TiO

2

, suggesting that both these elements have remained immobile in the sampled rocks.

2.3 Thin sections

15 polished thin sections were produced from the reference samples by Vancouver Petrographics Ltd., Vancouver, Canada. These were investigated using a petrographic microscope at Luleå University of Technology (LTU), Luleå, Sweden, including examination of mineral assemblages, abundances and overprinting relationships, as well as rock textures. In addition, thin sections were used to examine some of the samples classified as least altered (further discussed under geochemical results).

3 Results

3.1 Geological description

The eight diamond drill cores logged in this study are located to the south-southeast of the known Svärdsjö deposit and are drilled perpendicular to the dominating northwest- southeast strike of the nearly vertically dipping rock sequence (Billström, 1980). The drillholes vary in length from 116 m to 741 m, dip 50–55°

and reach a maximum depth of approximately 530 m below surface. Based on the core logging and work by previous researchers (Levi et al., 1980; Bergman and Sundblad, 1981), rocks surrounding the Svärdsjö deposit are assumed to be predominantly volcanic to subvolcanic, with a greenschist to amphibolite facies metamorphic overprint.

Throughout the studied area rock types and mineral assemblages vary considerably. At the south-eastern end, sericite-chlorite schists form a distinct unit (figure 3A). These rocks are fine- to medium grained and faintly quartz-phyric, with the quartz showing signs of recrystallization such as undulating extinction, as well as sericite inclusions. Foliation is typically pervasive in sericite-rich, and moderate in quartz-rich rock domains. Thin section examination revealed a quartz-rich (>70 vol. %) matrix with minor feldspar and elongated sheets of sericite defining the foliation (figure 4A). The sericite bends around clusters of finer quartz and feldspar, resembling eye-like shapes, which may explain the faint porphyritic appearance in hand sample.

A similar but more distinctly quartz-phyric rock is encountered, in smaller sections, on either side of the Fäbodgruvan mineralisation. In this unit, the sericite is more fine-grained, mainly occurring as groundmass in between larger quartz porphyroblasts and only partially forming elongated sheets together with biotite. In thin section, this rock is more sericite-rich but lacks feldspar and chlorite.

Surrounding the mineralised zones are a set of

volcanic rocks with a mineral assemblage

dominated by quartz, chlorite, sericite, biotite and

(8)

7 cordierite (figure 3B). The relative abundance of these minerals varies considerably, but no distinct discrimination of units can be made within this set of rocks. The rocks are fine-grained to very fine- grained and generally lack phenocrysts. However, in some places a faint quartz-phyric texture is visible and the rocks commonly show strong schistosity. Garnet, serpentine and andalusite occur in minor amounts, typically in proximity to intruding dykes and skarn-bearing marble (described in the following paragraphs, figure 3E). Thin section analysis suggests at least two different rock types within the unit; a quartz- sericite-chlorite rich rock (figure 4B) and a biotite-sericite-cordierite-wollastonite rock (figure 4E) with only minor quartz, occurring closer to the mineralised skarn-bearing marble zones.

Mafic dykes cross-cut the rock sequence in multiple places but are particularly abundant on

the south-eastern side of the Fäbodgruvan mineralisation (figure 8). The dykes are commonly fine-grained to very fine-grained close to the contacts and medium-grained at the centre with a dark green-black colour. The width of these dykes is typically <2 m. Biotite and acicular black amphibole are the most abundant minerals.

Minor chlorite ± epidote is typically present, along with disseminated magnetite and pyrite (<

5 vol. %). Two mafic units, with similar composition and appearance as the dykes, have a width of up to 20 m. These units occur on the south-eastern side of the Fäbodgruvan mineralisation (figure 8).

Subordinate bodies of very fine-grained biotite-rich rocks occur as 5–20 m thick sections locally between the Fäbodgruvan- and Vilnäset mineralisations (figure 8). The rocks are considerably more fine-grained, biotite-rich and homogenous than any of the other rock types. In

Figure 3. Examples of rock types and mineral assemblages in diamond drill cores from Svärdsjö. A. Sericite-chlorite schist (SVASJ288-95 m). B. Chlorite-sericite-cordierite rock closer to the Fäbodgruvan mineralisation. C. Tremolite- sericite skarn proximal to mineralised marble unit. D. Chlorite-garnet ± anthophyllite rock (SVASJ279- 269 m). E.

Cordierite-sericite-serpentine ± andalusite rock proximal to Fäbodgruvan mineralisation. F. Feldspar-phyric rock with

minor sericite-biotite (SVASJ279).

(9)

8 thin section quartz and biotite are identified as the dominating minerals (55 and 30 vol. %, respectively). Some chlorite (<10 vol. %), sericite and garnet (<3 vol. %) can also be observed.

Biotite and chlorite define a strong foliation, which bends around the garnet.

The skarn-bearing marble at the Fäbodgruvan mineralisation (figure 8) is up to 15 m thick and occurs along strike of the Svärdsjö deposit. The unit can be followed at depth and along strike through multiple drill core intersections (figure 8–

9). The thickest part of the unit is found close to the surface and progressively narrows with increasing depth. At shallow depth, the Fäbodgruvan mineralisation comprises both fairly well-preserved marble, ophicalcite and tremolite skarn, whereas at greater depth the unit is almost exclusively comprised of tremolite skarn (figure 3C). Sphalerite, pyrite, pyrrhotite, galena and magnetite occur as disseminated to massive mineralisations in the marble. The Vilnäset mineralisation is hosted by a second, smaller skarn-bearing unit encountered in two drillholes approximately 150 m south-east of the main unit. Abundant, close to massive, chlorite and pyrite occur within and proximal to this mineralisation. In places, durchbewegung textures can be seen, where rounded fragments of wall rock have been incorporated in the massive sulphides, a texture typically associated with ductile deformation in sulphides (Vokes, 1969;

Geijer, 1971; Marshall and Gilligan, 1989). Shear zones (< 2 m wide) with both brittle and ductile deformation, as well as abundant chlorite and talc, are common at the contact between the marble and the surrounding rock. Microscopic investigation of the skarn-bearing parts of the marble confirms that tremolite is the most abundant mineral, whereas only trace calcite or dolomite could be found (figure 4C). Other commonly occurring minerals in the tremolite- altered rock are sericite, hornblende and talc.

A dark green-grey quartz-chlorite-garnet- anthophyllite rock was encountered adjacent to the skarn-bearing marble at Fäbodgruvan. The rock is fine-grained and lacks any primary

textures but can be distinguished by the presence of red-brown garnet porphyroblasts and acicular anthophyllite (figure 3D). Other minerals observed in this rock include chlorite, sericite and cordierite. A thin section of this rock (figure 4D), taken proximal to the mineralised marble, also how the presence of tremolite and strongly pinitized cordierite. Chlorite and recrystallized quartz are, however, identified as the main constituents of this rock based on thin section observations. No precursor rock could be determined in core or thin section, due to the lack of primary textures and minerals.

At the north-western end (figure 8), a sequence of feldspar-phyric rocks is observed. These rocks are fine-grained with 5–30 vol. % feldspar and quartz phenocrysts, typically 2–4 mm in size (figure 3F). The overall grain size varies slightly and phenocrysts are unevenly distributed in the rock. In a few places, the foliation kinks sharply by 10–15°. Viewed in thin section, quartz and feldspar are identified as the dominant minerals in the groundmass, occurring in approximately equal proportions (figure 4F). Phyllosilicates in the matrix consist of biotite (5–20 vol. %) with minor sericite and chlorite, defining the foliation.

Sericite growth has caused a poikilitic texture and

coating on feldspar crystals. The foliation bends

around phenocrysts, resulting in an eye-shaped

appearance.

(10)

9

Figure 4. Photomicrographs of rock types and mineral assemblages found in Svärdsjö. A. Sericite schist from the south-eastern area (SVASJ288-93.55 m), crossed polarisation. B. Quartz-rich sericite-chlorite rock closer to Fäbodgruvan mineralisation (SVASJ269-48.30 m), split plane- and crossed polarisation. C. Tremolite-rich rock from skarn-bearing marble zone at Fäbodgruvan mineralisation (SVASJ290-432.05 m), split plane- and crossed polarisation. D. Quartz-chlorite-garnet- anthophyllite rock adjacent to skarn-bearing marble at Fäbodgruvan (SVASJ279-269.65 m), crossed polarisation. Inset shows anthophyllite-rich part of thin section, split plane- and crossed polarisation. E. Sericite-biotite-cordierite ± wollastonite- hornblende rock proximal to Vilnäset mineralisation (SVASJ290-240.20 m), crossed polarisation, inset shows wollastonite- biotite rich part of thin section. F. Quartz-feldspar phyric rock with minor sericite-chlorite at the north-western side of the mineralised zone (SVASJ265-316.20 m), split plane- and crossed polarisation. Qtz = quartz, Ser = sericite, Bt = biotite, Chl

= chlorite, Trem = tremolite, Hbl = hornblende, Grt = garnet, Antho = anthophyllite, Wo = wollastonite, Cord = cordierite,

Fsp = feldspar.

(11)

10 3.2 Geochemical results

Results from the lithogeochemical analyses are presented as median values of each logged rock type in table 1. Median absolute deviations (MAD) were calculated to provide a measure of the variation found in each element and rock type.

Values below the limit of detection were re- calculated to half the detection limit before calculation of median and MAD values (cf.

Kampmann et al., 2016b). The major oxides are presented in wt. % whereas trace elements are presented in ppm. Fe is measured as FeO+Fe

2

O

3

, from here on denoted as wt. % Fe

2

O

3

T. A few elements, such as Cr

2

O

3

, were at or below the detection limit in most samples, resulting in the same median value for all rock types. Some remarks can be given regarding the chemical composition of the rock types. The chl-ser-bt- cord rock (table 1) is the most abundant rock type in the area but also show large variations, particularly in SiO

2

and Fe

2

O

3

T content. The generally high values of Fe

2

O

3

T and MgO fits with the observations of abundant chlorite and biotite. In contrast, the feldspar phyric rocks show small variations and are comparatively rich in SiO

2

and Na

2

O but low in Fe

2

O

3

T and MgO. The skarn-bearing marble samples were separated into two groups in the geochemical results; one with predominantly calc-silicate minerals such as tremolite and one with predominantly carbonate minerals. The marble samples are rich in CaO and MgO, indicating a dolomitic composition. The calc-silicates have considerably higher SiO

2

values but lower CaO and a high variability in both these elements as well as in Fe

2

O

3

T and MgO.

Interpretation

4.1 Rock types

In altered and metamorphosed volcanic provinces, the identification of rock types and precursor rocks can be challenging (e.g Barrett et al., 2005). This can lead to difficulties in interpreting the stratigraphy of an area and in correlating lithological units between drill holes

(Barrett and MacLean, 1994). At Svärdsjö the rocks have been subject to intense hydrothermal alteration, obscuring precursor rock textures and composition. However, using geochemical data and elements such as TiO

2

, Al

2

O

3

and Zr, which typically remain immobile during hydrothermal alteration and metamorphism, the precursor volcanic rock types can be identified (MacLean and Barrett, 1993; Jansson and Allen, 2015).

Immobile elements have varying degrees of compatibility in igneous processes, which can be utilized to identify fractionation trends and magmatic affinity of volcanic rocks. A binary plot of immobile and incompatible elements, such as Th-Yb (figure 5) provides an indication of the magmatic affinity of altered volcanic rocks (MacLean and Barrett, 1993; Barrett et al., 2005).

The rocks in Svärdsjö, with the exception of a single tholeiitic mafic dyke sample, exhibit a strong calc-alkaline affinity.

The absolute mass of immobile elements remains constant during hydrothermal alteration and the ratio between two immobile elements should therefore be the same in the precursor rock and the altered rock, regardless of mass gain or loss of mobile elements (MacLean and Barrett, 1993). It has been shown that ratios of Zr/TiO

2

and Nb/Y provide a good discrimination of

different rock types and magma series when

plotted in a binary diagram, and the immobile

nature of these elements allows for the

identification of precursors to altered volcanic

rocks (Winchester and Floyd, 1977). The rocks in

Svärdsjö plot mainly as a large group overlapping

the dacitic and rhyolitic border, with a smaller

group of more mafic dacite and slightly scattered

samples of andesitic-basaltic compositions

(figure 6). To identify groups of rocks belonging

to a single precursor, binary plots of immobile

elements and immobile element ratios can be

utilised (MacLean and Barrett, 1993). In binary

plots involving two immobile elements, rock

samples with a common precursor will form lines

passing through the origin. If immobile element

ratios are used, rocks with a single precursor will

tend to form clusters (Barrett and MacLean,

1994). The Svärdsjö volcanic rocks can be

(12)

11

Rock type Bt-rich rock Feldspar phyric rock

Chl-ser-bt-

cord rock Mafic dyke Qtz-chl-grt- antho rock

Ser-chl schist

Calc-silicate

rocks Marble

n 3 13 25 8 6 16 11 3

Med. MAD Med. MAD Med. MAD Med. MAD Med. MAD Med. MAD Med. MAD Med. MAD

SiO

2

66.98 3.84 75.03 0.60 68.59 8.97 57.13 5.07 69.31 5.29 73.69 3.65 53.68 9.26 9.08 3.76

Al

2

O

3

14.07 1.57 13.31 0.26 13.18 1.40 16.26 0.75 9.44 1.44 13.23 0.79 4.33 2.54 2.41 1.02

Fe

2

O

3

T 6.91 1.87 3.07 0.35 7.80 3.73 12.12 1.14 13.20 2.66 3.32 0.95 8.72 6.47 2.04 0.24

MgO 4.87 0.86 0.87 0.41 5.48 2.58 4.44 1.55 5.43 1.38 4.27 2.02 17.38 8.08 33.13 0.65

CaO 0.58 0.32 1.55 0.41 0.30 0.23 3.78 1.92 0.15 0.04 0.31 0.15 5.53 5.36 50.17 4.35

Na

2

O 0.98 0.89 3.44 0.65 0.15 0.06 2.06 0.83 0.09 0.03 0.88 0.69 0.12 0.12 0.01 0.00

K

2

O 2.49 0.14 2.23 0.77 1.78 0.78 1.44 0.67 0.34 0.11 2.29 0.59 0.02 0.02 0.01 0.00

TiO

2

0.53 0.02 0.22 0.02 0.20 0.07 0.90 0.26 0.06 0.01 0.25 0.08 0.05 0.03 0.07 0.03

P

2

O

5

0.13 0.00 0.04 0.01 0.03 0.02 0.21 0.12 0.01 0.00 0.05 0.02 0.02 0.02 0.02 0.01

MnO 0.05 0.01 0.07 0.02 0.08 0.03 0.19 0.05 0.08 0.01 0.05 0.02 0.35 0.25 2.19 0.81

Cr

2

O

3

0.001 0.000 0.001 0.000 0.001 0.000 0.001 0.000 0.001 0.000 0.001 0.000 0.001 0.000 0.001 0.000

LOI (wt %) 2.50 0.10 1.20 0.30 3.50 0.90 2.65 0.70 3.00 1.25 2.60 0.60 3.00 1.00 41.50 0.10

T. Ox (wt%) 99.83 0.01 99.93 0.01 99.86 0.03 99.85 0.03 99.88 0.02 99.87 0.03 99.76 0.19 99.12 0.51

Ba 1037 563 877 197 342 200 436 124 52.63 10.80 573 149 6.13 5.63 3.42 0.22

Sc 15.38 0.02 10.04 1.93 9.29 4.11 33.15 7.60 3.57 0.57 9.25 4.73 4.41 1.33 3.21 0.21

Be 0.50 0.00 1.02 0.52 0.50 0.00 0.50 0.00 0.50 0.00 1.53 0.78 0.50 0.00 0.50 0.00

Co 11.99 3.62 2.13 0.51 4.22 3.18 26.69 6.85 8.68 3.56 2.32 0.67 5.93 3.51 1.88 1.37

Cs 1.66 0.03 0.81 0.20 1.44 0.60 1.09 0.42 0.77 0.09 1.04 0.42 0.20 0.15 0.05 0.00

Ga 16.21 0.80 12.86 0.99 14.69 1.86 16.74 1.85 10.84 1.55 14.13 2.00 7.16 4.03 1.54 0.71

Hf 4.10 0.71 4.34 0.18 3.89 0.75 3.35 1.09 2.75 0.40 3.70 0.41 1.51 1.14 1.03 0.10

Nb 8.41 2.69 8.32 0.51 8.32 1.45 7.48 2.50 5.81 0.96 8.02 1.14 1.05 0.83 2.22 0.66

Rb 59.00 9.85 42.54 14.43 44.96 13.58 36.26 10.80 9.68 2.03 48.61 13.02 0.60 0.55 0.05 0.00

Sn 3.07 0.00 2.02 0.01 3.07 1.02 2.06 1.31 1.54 0.52 2.08 0.98 5.25 3.14 0.50 0.00

Sr 44.72 39.10 128 19.47 8.54 3.65 128 45.69 2.99 0.94 24.19 15.88 1.86 1.09 60.10 0.24

Ta 0.62 0.10 0.71 0.10 0.72 0.20 0.52 0.21 0.56 0.15 0.72 0.09 0.05 0.00 0.17 0.12

Th 9.13 1.53 12.14 1.18 12.13 1.86 5.87 2.08 11.20 2.11 11.74 0.79 3.78 3.05 2.74 0.95

U 3.49 0.57 4.72 0.67 4.79 1.25 2.52 0.46 4.80 0.37 4.44 0.48 2.86 1.81 3.76 0.57

V 146 4.93 17.16 7.11 4.00 0.00 229 60.36 4.00 0.00 8.70 4.70 4.00 0.00 13.68 0.75

W 0.82 0.09 0.80 0.29 1.16 0.91 1.13 0.56 0.57 0.32 1.80 1.30 1.69 1.44 14.38 6.53

Zr 149 7.94 155 7.12 138 22.35 117 34.58 89.57 5.93 135 16.55 51.22 42.70 37.78 3.57

Y 30.26 3.96 23.10 2.41 25.32 5.71 23.82 3.84 19.24 2.62 20.60 3.02 9.75 5.56 8.39 1.21

La 31.59 7.76 31.00 0.99 29.48 4.95 23.98 6.51 26.26 4.42 31.89 3.17 10.61 7.82 8.38 1.87

Ce 63.59 16.24 58.26 1.42 59.57 11.77 49.58 10.67 50.77 8.84 58.01 3.87 21.03 15.38 18.46 4.94

Pr 6.92 1.70 6.35 0.27 6.34 0.97 5.78 1.37 5.45 0.69 6.22 0.45 2.44 1.61 2.07 0.48

Nd 24.92 5.36 21.94 1.29 22.67 2.34 21.62 5.76 18.81 1.66 21.97 1.97 9.03 5.37 8.03 2.06

Sm 5.42 1.29 4.19 0.39 4.31 0.56 4.31 1.23 3.58 0.23 3.82 0.50 1.68 1.27 1.44 0.68

Eu 1.31 0.07 0.73 0.07 0.80 0.29 1.12 0.24 0.23 0.09 0.72 0.14 0.48 0.31 0.58 0.10

Gd 4.95 1.15 3.83 0.41 4.34 0.54 4.12 0.95 3.33 0.25 3.57 0.46 1.85 1.42 1.25 0.20

Tb 0.86 0.10 0.60 0.06 0.68 0.09 0.65 0.11 0.54 0.03 0.53 0.07 0.33 0.25 0.19 0.08

Dy 5.30 0.27 3.90 0.40 4.36 0.67 4.04 0.69 3.46 0.30 3.34 0.36 1.88 1.45 1.21 0.39

Ho 1.14 0.06 0.83 0.08 0.91 0.18 0.88 0.11 0.73 0.10 0.72 0.11 0.37 0.28 0.29 0.06

Er 3.29 0.25 2.57 0.24 2.82 0.54 2.70 0.29 2.34 0.30 2.28 0.35 1.18 0.82 0.77 0.24

Tm 0.45 0.07 0.39 0.04 0.41 0.09 0.41 0.08 0.34 0.05 0.33 0.05 0.17 0.11 0.12 0.02

Yb 3.21 0.43 2.69 0.24 2.83 0.60 2.66 0.38 2.33 0.47 2.34 0.37 1.11 0.87 0.77 0.26

Lu 0.49 0.11 0.42 0.04 0.44 0.10 0.41 0.07 0.37 0.07 0.37 0.05 0.17 0.13 0.12 0.01

Table 1. Lithogeochemical data for the lithological units in Svärdsjö, presented as median and median average deviations

(MAD). Major oxides are given in weight %, trace elements in ppm. n = number of samples, T. Ox = Total oxides.

(13)

12 grouped into five precursors using plots such as Al

2

O

3

/TiO

2

vs. Zr/TiO

2

and Zr/Al

2

O

3

vs. Al

2

O

3

/ TiO

2

(figure 7A–B). Combining this with the geological data from core logging, a section of the precursor rock type stratigraphy was constructed (figure 8). The colours of the symbols used in diagrams and figures throughout this study reflect the classification into precursor rocks, as shown in the legend of figure 5. A distinct group of samples plot in the upper right corner of figure 7A and are classified as rhyolitic using the classification diagram (figure 6; Winchester and Floyd, 1977). These samples are all adjacent to the Fäbodgruvan mineralisation (figure 8) and mainly represent quartz-chlorite-garnet- anthophyllite rock. Two groups near the centre of the diagram make up the bulk of the samples (figure 7A–B). The groups are classified as a dacite (Dacite A) and transitional between dacite and rhyolite (Dacite B), respectively, and comprise both feldspar-phyric rocks, sericite- chlorite schists and fine-grained to faintly quartz- phyric rocks with a chlorite-sericite-biotite- cordierite mineral assemblage. Mafic dykes form a separate group in the left lowermost corner of the immobile-element diagrams and in the andesitic to basaltic fields of the classification diagram (figure 6). A small group of samples plots between the andesite-basalt and the dacitic to rhyolitic units. These rocks plot in the dacitic field in the rock classification diagram (Dacite C;

figure 6), but are substantially more mafic in composition than the other dacites. The rocks are mainly quartz-phyric sericite-chlorite schists but other rock types are also included, indicating that

this group could not be distinguished from the more felsic dacites in drill core. The group occurs in a limited area close to the Vilnäset mineralisation, as well as in the area between the mineralised zones (figure 8).

An efficient method for quantifying hydrothermal alteration and reveal alteration trends around ore deposits is to construct a box plot using the chlorite-carbonate-pyrite index (CCPI) and the Ishikawa alteration index (figure 9; Large et al., 2001). Although the box plot was originally designed for use on pre-metamorphic it has also been successfully applied at medium- to high-grade metamorphosed deposits (Theart et al., 2010). Plotting the Svärdsjö samples in an alteration box plot confirms the altered nature of the rocks. Samples with AI >70 approximately correspond to rocks identified in hand sample as strongly to intensely altered and are denoted by a triangular symbol. Apparently less altered rocks are indicated by squares. The majority of the rock types cluster in the upper right part of the diagram, indicating that chlorite-pyrite-sericite hydrothermal alteration affected them after their formation. A typical trend in volcanic hydrothermal systems is one of weak sericite alteration at the margins of the system and progressively moving towards intense sericite- chlorite ± pyrite alteration in the proximal footwall zone (Large et al., 2001). This trend is evident at Svärdsjö, as indicated by the vertical trail of samples along the right side of the plot.

Some of the least altered dacitic and rhyolitic rocks show signs of diagenetic albite alteration,

Figure 5. Th/Yb plot of magmatic affinity in the Svärdsjö rocks, the symbology is further discussed in the following paragraphs.

(14)

13 plotting closer to the lower left corner of the box plot, while most mafic samples plot within or close to the least altered box. The feldspar-phyric rocks at the north-western end of the section (figure 8, 10) plot mainly in the least altered dacite and rhyolite boxes (figure 9).

4.2 Hydrothermal alteration

By combining geological data from core logging with lithogeochemical trends in the alteration box plot, it is possible to establish a zonation model of hydrothermal alteration around the studied Svärdsjö mineralisations. Using data from the five drillholes drilled in this profile, a schematic section showing the alteration zones has been generated (figure 10). Differentiation by

mineral assemblage and alteration intensity allowed for a subdivision into five zones (figure 10). The zones and their respective alteration types are summarised in table 2 and further described in the following section.

4.2.1 Calc-silicate zone

Within and proximal to the mineralisations a spatially limited zone dominated by a tremolite- hornblende-talc ± quartz mineral assemblage has been identified, herein referred to as the ‘calc- silicate zone’ (table 2). This zone is pronounced and continuous surrounding the Fäbodgruvan mineralisation, but discontinuous at the Vilnäset mineralisation and is therefore only indicated around the former (figure 10). An abundance of

Figure 7A–B. Precursor discrimination of the Svärdsjö rocks using two sets of

immobile element ratios.

(15)

14 calc-silicate minerals such as tremolite, wollastonite and minor actinolite are characteristic for this zone. Quartz is present in variable amounts.

4.2.2 Strong chlorite zone

Surrounding the Vilnäset mineralisation an alteration zone characterised by abundant chlorite-sericite-cordierite mineralogy occurs, herein referred to as the ‘Strong chlorite zone’

(table 2). This zone is characterisd by sericite- chlorite schists and faintly quartz-phyric, moderately sericite altered rocks with < 5 vol % poikilitic feldspar. Proximal to the Vilnäset mineralisation (figure 10) the rocks exhibit intense chlorite-pyrite ± anthophyllite alteration.

Mafic dykes cross-cutting this zone are weakly to moderately chlorite altered.

Table 2. Summary of alteration zones and associated mineral assemblages surrounding the Svärdsjö deposit.

Alteration zone Mineral assemblage Alteration intensity

Calc-silicate zone Tremolite-hornblende-talc-quartz Intense Strong chlorite zone Chlorite-sericite-cordierite Strong

Sericite zone Sericite-chlorite±biotite Weak

Intense chlorite zone Chlorite-sericite-cordierite±garnet- anthophyllite

Intense

Least altered zone Sericite-biotite Subtle

Figure 8. Interpreted stratigraphy based on core logging and lithogeochemical rock types.

(16)

15 4.2.3 Sericite zone

Between the Vilnäset zone and the Fäbodgruvan zone the alteration intensity decreases considerably and the mineral assemblage is dominated by sericite-chlorite with minor biotite. This zone is denoted as the ‘Sericite zone’ (table 2). Subtly chlorite altered mafic dykes and weakly sericite-chlorite altered quartz

± feldspar-phyric rocks are interfingered in this zone. In the alteration box plot (figure 9) samples from the sericite zone plot within, or close to, the least altered boxes, indicating minor alteration.

4.2.4 Intense chlorite zone

A further alteration zone is seen surrounding the Fäbodgruvan mineralisation and is referred to as the ‘intense chlorite zone’ (table 2, figure 10).

A pervasive chlorite-sericite-cordierite assemblage overprinting all primary textures is characteristic for this zone. Downhole (NW) of the mineralisation abundant garnet and anthophyllite are encountered for approximately 30 m, before a thin mineralisation and a sharp

decrease in alteration occurs approximately 50 m downhole.

4.2.5 Least altered zone

The intense chlorite zone is followed by a zone of subtle sericite-biotite alteration in well- preserved quartz-feldspar phyric rocks, here referred to as the ‘least altered zone’ (table 2, figure 10). These rocks show no signs of hydrothermal alteration in hand sample and plot as least altered dacites and rhyolites in the alteration box plot (figure 9).

Figure 9. Alteration box plot (after Large et al., 2001) showing trends of hydrothermal alteration in Svärdsjö.

CCPI = Chlorite-carbonate-pyrite index:

100 𝑀𝑔𝑂+𝐹𝑒𝑂

𝑀𝑔𝑂+𝐹𝑒𝑂+𝑁𝑎2𝑂+𝐾2𝑂

, AI = Ishikawa alteration index:

100 𝐾2𝑂+𝑀𝑔𝑂

𝐾2𝑂+𝑀𝑔𝑂+𝑁𝑎2𝑂+𝐶𝑎𝑂

.

(17)

16 4.3 Mass changes

The effects of hydrothermal alteration around ore deposits can be quantified by calculating the mass change of the elements involved (Barrett and MacLean, 1994). The first step in this process is to select a set of least altered samples, from which magmatic fractionation trends applicable to all co-magmatic samples can be obtained. The conditions used for selecting least altered samples in this study largely follow the criteria set by Kampmann et al. (2016b), applied at the Falun deposit in Bergslagen (figure 1). The criteria for determining a least altered rock are: (i) the mineralogy and chemical rock classification of the sample should be in agreement, (ii) the alteration box plot (figure 9) should indicate a least altered rock, (iii) minerals found in the strongly altered rocks at Svärdsjö, such as garnet, tremolite, cordierite and anthophyllite or textures indicative of hydrothermal alteration should not be present, (iv) the samples should have <0.2 wt

% S and <3 wt % LOI.

Using these criteria, 15 volcanic to sub- volcanic samples were identified as least altered.

The selection comprises ten quartz-feldspar phyric rocks, three fine-grained felsic rocks and two mafic intrusive rocks. The fractionation trends of element pairs in the least altered samples are shown in figure 11A–E.

The immobile elements Al

2

O

3

and Zr form a linear trend closely resembling the magmatic fractionation trends observed at metamorphosed hydrothermal deposits in the Fennoscandian Shield (e.g. Barrett et al., 2005; Kampmann et al., 2016b). Using the Al

2

O

3

/Zr trend (figure 11A) the precursor Zr content of each altered sample could be determined, and subsequently be used to calculate the precursor content of major oxides using the trends shown in figure 11B–E. The major oxides mainly follow a polynomial, concave-upwards trend with oxide content

Figure 10. Schematic view of the hydrothermal alteration zones surrounding the Svärdsjö deposit. Section oriented 50°

NE along the main strike of the lithology. The location and alteration intensity of samples from the drill cores are

indicated by symbols.

(18)

17 decreasing as Zr increase (figure 11B–E), which is the expected trend in rocks of basaltic-andesitic to rhyolite composition (Barrett and MacLean, 1994). The calculated fractionation trends in figure 11 show a good fit, with a coefficient of determination (R

2

) higher than 0.7. Trends and precursor values of Na

2

O (R

2

= 0.27), K

2

O (R

2

= 0.11) and MnO (R

2

= 0.71) vs. Zr were also calculated, where the two former oxides show a weaker correlation, possibly due to some alteration in the least altered samples.

The mass change calculations show strong variations in elements lost and gained, both between alteration zones and between lithological units (table 3). A general trend of Fe

2

O

3

T and

MgO enrichment, as well as CaO and Na

2

O depletion can be observed (table 3, figure 12).

Plotting mass changes in binary diagrams can give information on the alteration processes associated with the mass changes (MacLean and Barrett, 1993). Mass loss of CaO and Na

2

O indicate the breakdown of feldspar, whereas mass gain of Fe

2

O

3

and MgO indicates chloritisation (MacLean and Barrett, 1993; Gifkins et al., 2005). Both these alteration processes can be detected in the Svärdsjö rocks (figure 12).

Furthermore, all rocks classified as strongly altered have lost significant CaO and Na

2

O (table 3), indicating extensive feldspar breakdown.

Subdividing into the alteration zones previously identified (figure 10, table 2), some observations

Table 3. Mass change calculations for the Svärdsjö rocks, presented as median, MAD (Median Average Deviation), and maximum gains and losses. All elements are reported as wt % mass change.

Rock type Bt-rich siltstone

Feldspar- phyric rock

Chl-ser-bt-cord

felsic rock Mafic dyke Qtz-grt-antho

felsic rock Ser-chl schist

# of samples 3 13 25 8 6 16

∆SiO

2

Median 0.73 -0.75 -4.01 -7.12 27.60 2.39

MAD 20.47 2.03 9.12 5.81 5.22 6.82

Max. gain 0.86 4.41 67.69 4.15 71.66 33.09

Max. loss -19.75 -4.26 -31.96 -28.56 -2.13 -16.24

∆Fe

2

O

3

Median 1.74 -0.10 2.71 3.09 13.71 -0.37

MAD 0.62 0.73 2.12 2.18 3.09 1.15

Max. gain 2.25 1.81 21.25 5.80 18.78 2.75

Max. loss 1.12 -1.74 -2.97 -1.24 10.43 -2.47

∆MgO Median 3.34 0.26 4.20 1.58 6.70 2.60

MAD 1.72 0.42 1.80 1.53 1.88 1.63

Max. gain 4.38 1.69 12.78 3.76 49.82 6.01

Max. loss 1.62 -0.88 0.49 -1.38 3.21 -0.25

∆CaO Median -0.88 -0.08 -1.63 -0.06 -1.88 -1.50

MAD 1.34 0.32 0.40 0.72 0.20 0.32

Max. gain 0.46 1.49 0.32 1.94 -1.20 0.88

Max. loss -2.86 -0.79 -4.25 -2.74 -2.31 -2.62

∆Na

2

O Median -2.39 -0.05 -2.99 -0.81 -2.94 -2.34

MAD 1.66 0.30 0.63 0.68 0.10 0.69

Max. gain -0.73 1.52 -0.14 2.33 -2.67 1.90

Max. loss -2.61 -0.61 -4.13 -2.30 -4.12 -3.58

∆K

2

O Median 0.11 -0.25 -0.47 -0.16 -1.64 0.37

MAD 0.00 1.11 0.58 0.91 0.21 0.56

Max. gain 0.77 3.88 1.63 0.90 -0.83 2.01

Max. loss -0.42 -1.50 -2.31 -1.41 -2.19 -2.41

∆MnO Median -0.02 -0.01 0.00 0.00 0.02 -0.05

MAD 0.04 0.02 0.03 0.05 0.03 0.01

Max. gain 0.01 0.06 1.02 0.08 0.20 0.06

Max. loss -0.09 -0.03 -0.10 -0.06 -0.02 -0.07

(19)

18 can be made: Rocks within the intense chlorite zone (figure 10) show the largest gains of Fe

2

O

3

T and MgO, indicating a strong chloritisation (figure 12). These rocks have also lost almost all their initial Na

2

O and CaO. SiO

2

varies significantly within the alteration zone, both gains and losses of > 30 wt % are recorded, particularly from samples close to the mineralised marble.

The calc-silicate zone associated with the mineralised marble units had very low Zr, TiO

2

and Al contents, indicating only a minor detrital volcanic rock component, which could not be characterised using mass calculations. The sericite zone shows moderate gains of Fe

2

O

3

T and MgO, in agreement with the weak sericite- chlorite alteration observed. Mass changes in the strong chlorite zone are characterised by

moderate gains of MgO and strong depletion of Na

2

O. Fe

2

O

3

T was variably gained or lost but the mass change is limited to < 5 wt %. In the least altered zone mass changes are close to zero. A slight mass gain of Na

2

O in some samples can be attributed to albitisation and a minor K

2

O mass gain can occur by subtle sericite alteration (Gifkins et al., 2005).

A downhole plot of mass changes along SVASJ279 illustrates the mass change and alteration zone relations described above. The strong enrichment of Fe

2

O

3

T and MgO as well as the depletion of Na

2

O in rocks of the intensely chlorite altered zone above and below the Fäbodgruvan mineralisation is clearly visible (figure 13). The sericite zone can be identified in

Figure 11. Magmatic fractionation trends from least altered samples at the Svärdsjö deposit. Blue

points = mafic samples, cyan points = intermediate samples, yellow points = felsic samples.

(20)

19 the upper part of the hole where close to zero mass changes of Na

2

O are observed. The five uppermost samples show the variation of mass changes in the strong chlorite zone, including a typical slight gain of MgO and depletion of Na

2

O.

A sharp contact to the subtle alteration zone where all mass changes are close to zero can be observed downhole of the intense chlorite zone.

A single sample taken within the intensely altered zone from a considerably less altered 1.5 m thick quartz-feldspar phyric dyke can clearly be identified by the sudden lack of mass change compared to surrounding samples (figure 13).

SiO

2

mass changes vary, either showing no significant change or strong depletion (figure 13), possibly indicating effects of both silicification and chloritisation.

The spatial extent of mass changes has been evaluated using 3D interpolation and visualisation, cross-sections and plan views. The most significant trends were found in the mass changes of Fe

2

O

3

T, MgO and Na

2

O. The interpolated models (figure 14A–C) show a similar pattern of element enrichment and depletion surrounding the mineralised parts as seen in the downhole plot of SVASJ279 (figure 13). Enrichment of MgO and Fe

2

O

3

T form a sub-

vertical sheet-like zone around the Fäbodgruvan mineralisation and a smaller, slightly scattered zone around the Vilnäset mineralisation (figure 14A–B). Mass gains increase towards the more strongly mineralised parts of the zones, indicating stronger chloritisation (Barrett et al., 2005). A lower threshold of +3 wt % mass gain was used for both elements, which provided a large zone around the mineralisations, without indicating areas of altered barren rock. An upper threshold of 6.5 wt % and 7 wt %, respectively, provides a smaller, more focused zone around the mineralisations. An area of less Fe

2

O

3

T mass change can be identified in the centre of the high mass gain zone in figure 14A. This is likely an effect of the presence of a less altered dyke with no significant mass change regarding Fe

2

O

3

T (figure 13).

The enrichments of MgO and Fe

2

O

3

T generally follow similar patterns, increasing strongly as the mineralisations are approached (figure 14A–B). However, a strong enrichment of Fe

2

O

3

T with moderate MgO gains is present away from the mineralisations on the NW side of Fäbodgruvan (figure 8). This discrepancy can be explained by the presence of a thin Fe-bearing sulphide layer (mainly pyrite and pyrrhotite)

Figure 12. After MacLean and Barrett (1993). Mass changes indicative of alteration observed in

the Svärdsjö rocks. For comparison, the symbology is the same as used in the precursor and

alteration graphs presented previously.

(21)

20 occurring at the contact between the intense chlorite- and the least altered zone (figure 8–9).

A strong to almost complete depletion of Na

2

O, presumably associated with feldspar destruction (MacLean and Barrett, 1993; Gifkins et al., 2005) occurs surrounding the mineralised zones when interpolated into 3D (figure 14A–C).

A threshold value of < -3 wt % ∆Na

2

O indicate near-complete depletion relative to the initial igneous composition, and is only visible close to the Fäbodgruvan mineralised zone where hydrothermal alteration is most intense. The slightly higher threshold zone of -2 wt % also shows the feldspar destruction associated to hydrothermal alteration around the Vilnäset mineralisation (figure 14A–C) and provides a larger zone than the Fe

2

O

3

T and MgO enrichment.

4.4 Precursor rocks and chemostratigraphy

The results presented in this study indicate that strong hydrothermal alteration and subsequent metamorphic overprint has occurred at and around the Svärdsjö deposit. Consequently, it is

challenging to accurately distinguish precursor rock types and pre-metamorphic alteration in hand sample and correlate these between different drill holes. To address this problem, five precursor rock types have been identified using geochemical proxies (figures 7A–B). These differ somewhat from the rock types identified through geological core logging. The mafic precursor group of andesitic to basaltic composition, have been identified in accordance with the geochemical results to a large degree. For the other rock types, the core logging deviated from the geochemical results. Specifically, it was possible to identify three distinct rock compositions within a rock suite previously including a mix of rhyolites, volcanoclastic rocks and mafic dykes. A dacite and a transitional dacite to rhyolite make up the bulk of these volcanic rocks and can be separated mainly by the higher Zr/Al

2

O

3

ratio in the transitional rock type (figure 7, Dacite A and B). The third rock type, a rhyolite, forms a distinctly separate unit, geochemically and stratigraphically.

Figure 13. Downhole log of SVASJ279 showing mass changes in some of the major oxides. The intense chlorite

zone with strong Fe, Mg enrichment and Na, Ca depletion is marked in yellow.

(22)

21

Figure 14. 3D interpolations of mass changes. The oblique views are oriented towards 65° (NE) with a 20° inclination

downward. A. MgO mass changes. B. Fe

2

O

3

T mass changes. C. Na

2

O mass changes. Drillhole traces and sample locations

are marked in black.

(23)

22 Dacite A is the most abundant rock type within the Svärdsjö area and occurs throughout the entire rock sequence. It occurs as both strongly altered rocks and subtly altered feldspar-phyric rocks.

The close chemical similarity between these two indicates a common, feldspar-phyric precursor.

This is further supported by faint phyric textures and large mass losses regarding Na

2

O and CaO (indicative of feldspar destruction) in the strongly altered rocks (figure 12). Additionally, a volcanoclastic origin of the dacitic rocks is inferred by textural characteristics in the subtly altered rocks such as poor sorting of phenocrysts, variable grain sizes, possible graded beddings and slight shifts in foliation, interpreted as cross- bedding. The rhyolite and dacite B units have a limited extent, occurring on either side of the Fäbodgruvan mineralisation (figure 8). The rhyolite is found on the north-west side of the mineralisation and corresponds to the quartz- chlorite-garnet-anthophyllite felsic rock identified during core logging. Intense hydrothermal alteration overprinted any primary textures in this rock and thus no determination of a coherent or clastic precursor could be made.

Dacite B occurs on the south-east side of the Fäbodgruvan mineralisation (figure 8) and has mainly been logged as a chlorite-sericite-biotite- cordierite felsic rock, though no distinction could be made between this unit and strongly altered dacite A in hand sample. The appearance of these rocks in distinct stratigraphic units could imply that they are coherent rocks, emplaced as subvolcanic intrusions prior to hydrothermal alteration. The position of the more rhyolitic rocks, surrounding a marble unit, could also suggest a decrease in volcanic activity caused the formation of these rocks, as limestones (now marbles) in Bergslagen are mainly stromatolitic in origin and could only form during periods of low volcanic activity (Allen et al., 1996).

5 Discussion

5.1 Character and extent of hydrothermal alteration

The alteration zonation identified at the Svärdsjö deposit (figure 10) is indicative of a VHMS associated hydrothermal system (e.g.

Trägårdh, 1991; Large et al., 2001; Jansson and Allen, 2015). The Fäbodgruvan and Vilnäset zones, surrounding the mineralised areas, are characterised by strong to intense chlorite-sericite alteration with high mass gains of Fe and Mg, particularly in the Fäbodgruvan zone. These zones are interpreted to represent the medial part of the hydrothermal system, where extensive chlorite- and sericite alteration is characteristic (Large et al., 2001; Gifkins et al., 2005). Proximal to the mineralisation, and particularly within the rhyolitic rocks, Si mass gains are high, indicating strong silicification, which typically occurs at the core of hydrothermal systems (Galley, 1993;

Gifkins et al., 2005). Rocks surrounding the Vilnäset mineralisation are comparably less altered and show smaller mass gains, possibly indicating a more distal position in the system.

Weakly altered dacite and basalt-andesite dykes crosscut the Vilnäset and Fäbodgruvan zones, indicating emplacement after the hydrothermal alteration.

The mineralised bodies at the Svärdsjö deposit are hosted by stratabound marble units, around which intense alteration generated an alteration assemblage dominated by calc-silicate minerals such as tremolite and wollastonite (figure 4C, 4E). This alteration is restricted to the marble itself and extends a few meters into the surrounding volcanic rocks. The high Mg content (median 33 wt%, table 1) of the marble is compatible with extensive dolomitisation of calcitic limestone. Levi et al. (1980) suggested that this occurred by leaching of volcanic rocks in association with the hydrothermal alteration.

However, mass calculations only show mass

gains of MgO within the investigated area, which

suggests a more distal source of MgO. Therefore,

it seems more likely that diagenetic alteration by

(24)

23 Mg-rich groundwater caused dolomitisation of the limestone (cf. Barrett et al., 2005; Gifkins et al., 2005) prior to hydrothermal alteration.

Weakly altered, partially feldspar-phyric rocks in the sericite zone (figure 10, table 2) are interpreted as dacitic dykes or subvolcanic intrusions, emplaced after, or during the waning stage of hydrothermal activity. Minor mass changes (figure 13) and a low intensity of alteration (figure 9) suggest these rocks are mainly affected by regional metamorphism and only limited metasomatism. Several fault zones are encountered within this rock sequence and tectonic movement can therefore not be excluded but the displacement required to move the less altered rocks into their current position would be substantial and is, therefore, less likely.

North-west of the Fäbodgruvan mineralisation a sharp, non-faulted boundary separates the Fäbodgruvan zone from the least altered zone (figure 10). The abrupt change between intensely- and weakly altered rocks implies that the latter were either emplaced after the hydrothermal activity had seized or were displaced to their current position by tectonic deformation. Since no faults or shear zones have been identified in the proximity of this sharp alteration boundary, the former explanation is favoured.

If the weakly altered rocks were deposited directly onto the intensely altered rocks the sharp boundary to the NW of the Fäbodgruvan mineralisation (figure 10) could represent the seafloor at the time when hydrothermal activity seized. Alternatively, the boundary could result from post-hydrothermal erosion of the strongly altered zone and subsequent deposition of the weakly altered rocks. In this case, the boundary would instead record the position of the seafloor at the change from erosion to deposition of sediments. These interpretations gives a stratigraphic up direction of the rock sequence, trending north-west, and suggests a sub-seafloor setting of the mineralisations. The presence of presumably stromatolitic limestone (now marble) gives further support to this, as these formed in shallow marine settings (Allen et al., 1996).

5.2 Metamorphic overprint and metasomatism

Rocks of the BLU underwent multiple episodes of deformation and regional metamorphism under low-P, high-T conditions (Stephens et al., 2009). At Svärdsjö, cordierite, anthophyllite and minor andalusite are present within the strongly to intensely altered Fäbodgruvan and Vilnäset zones. These minerals can form by metamorphism of chlorite-rich felsic volcanic rocks and show that hydrothermal alteration was overprinted by regional metamorphism (Barrett et al., 2005). Ophicalcite occurs in multiple sections of the dolomitic marble and is commonly formed by redistribution of Mg and Si between the dolomitic- and the volcanic rock during metamorphism.

Furthermore, small sections of biotite-rich siltstone encountered in the hydrothermally altered zones may represent metamorphosed chlorite-sericite altered rocks (c.f. Trägårdh, 1991). The presence of these minerals indicates that regional metamorphism has partially overprinted the rocks at the Svärdsjö deposit.

However, the presence of well-preserved feldspar-phyric rocks shows that regional metamorphism was, at least partially, of sufficiently low grade to preserve some primary rock features.

The mineralogy and mass changes identified in Svärdsjö suggests at least three post- emplacement processes have likely affected the rocks; diagenesis during compaction and burial, hydrothermal alteration related to the formation of the volcanic-associated deposit, and regional metamorphism at greenschist- to amphibolite facies. The alteration modified the mineralogy, textures and composition of the rocks (Gifkins et al., 2005). Although metamorphism is by definition isochemical, some degree of metasomatism can occur during metamorphism of submarine volcanic rocks due to the common presence of fluids (Gifkins et al., 2005).

The highly variable mass changes of Si

observed in Svärdsjö could be the result of

References

Related documents

It leads to unique contact surfaces (floor and roof) regardless the structural setting. Besides, in the database there are two sets of 'no data' : on the one hand a lack

The Bellviksberg deposit in the Dorotea area is part of the sandstone-hosted lead-zinc deposits occurring all along the eastern front of the Scandinavian Caledonides in Precambrian

The seven dolomite marble sample points have been divided into two groups for analysis; the three points on the right hand side of the study area, relatively proximal to

Sulfides occur disseminated or in veins with dark green actinolite or with chlorite (Figure 29). The grain size varies from fine- to coarse-grained for both sulfides

GARPN3740 are all adjacent to the marble breccia and is no more than a few meters thick. The green color is pale and not very strong. The microscopy and SEM analysis show

Albite is a common alteration mineral (Figure 30A and 30B) and occur as replacement of primary plagioclase grains and is recognized with the parallel twinning. The albitization

Red-feldspar veins (Fig. 13, B, E and G) with hematite staining are less important, but a distinct increase of vein intensity in the tuff unit is observed towards the

46.. shares of Silverstone. Cash flow from opera- tions during 2007 totalled $425.7 million and financing activities generated $34.6 million net, largely on a $37.2 million