Tracing the SW border of the Svecofennian Domain in the Baltic Sea region: evidence from petrology and geochronology from a granodioritic migmatite

Abstract

Geological investigations of a part of the crystalline basement in the Baltic Sea have been performed on a drill core collected from the depth of 1092–1093 m beneath the Phanerozoic sedimentary cover offshore the Latvian/Lithuanian border. The sample was analyzed for geochemistry and dated with the SIMS U–Pb zircon method. Inherited zircon cores from this migmatized granodioritic orthogneiss have an age of 1854 ± 15 Ma. Its chemical composition and age are correlated with the oldest generation of granitoids of the Transscandinavian Igneous Belt (TIB), which occur along the southwestern margin of the Svecofennian Domain in the Fennoscandian Shield and beneath the Phanerozoic sedimentary cover on southern Gotland and in northwestern Lithuania. It is suggested that the southwestern border of the Svecofennian Domain is located at a short distance to the SW of the investigated drill site. The majority of the zircon population shows that migmatization occurred at 1812 ± 5 Ma, with possible evidence of disturbance during the Sveconorwegian orogeny.

Introduction

The Precambrian basement in the Baltic Sea region is covered by km-thick successions of Phanerozoic sedimentary rocks, Quaternary sediments and overlying sea water. The relations between individual crustal segments in this region and the well-exposed adjacent Fennoscandian Shield have therefore historically been restricted to geophysical (gravity, magnetic and seismic) data, which have recorded a significant decrease in the crustal thickness towards the south in the central parts of the Baltic Sea (BABEL Working Group 1993; Lund et al. 2001).

In a recent contribution, Salin et al. (2019) demonstrated that this geophysical feature is related to the border between two major crustal complexes in the Fennoscandian Shield; the Svecofennian Domain in the northeast and the Transscandinavian Igneous Belt in the southwest.

In this paper, the southwestern border of the Svecofennian Domain is further traced in the eastern parts of the Baltic Sea. Using petrographic, geochemical and geochronological evidence from migmatite drilled from beneath the east-central Baltic Sea, correlations are made with known magmatic and metamorphic events recorded in southeastern Sweden.

Geological background

Fennoscandian framework

The Fennoscandian Shield consists of a 2.9–2.6 Ga and older Archaean craton in the northeast, the ∼1.92–1.87 Ga Svecofennian Domain intruded by the ∼1.86–1.67 Ga Transscandinavian Igneous Belt (TIB) in the center and the 1.1–0.9 Ga Sveconorwegian orogen in the southwest (Fig. 1 inset; Gaál and Gorbatschev 1987; Gorbatschev and Bogdanova 1993; Bingen et al. 2005; Stephens and Weihed 2020). The geological evolution in the earliest Proterozoic (2.5–2.0 Ga) was manifested by protracted extension and rifting of the craton and formation of oceanic basins along its western and southern margins. The Svecofennian crust was mainly formed from 1.92 to 1.87 Ga, which represents continued marine conditions to the west of the craton and associated formation of juvenile crust associated with several magmatic arcs, which eventually amalgamated with the Archaean continent (Korja et al. 2006; Stephens 2020). Supracrustal rocks of the domain include metavolcanic and intercalated metasedimentary successions formed between 1.96 and 1.84 Ga (e.g. Skiöld 1988; Welin 1992).

Fig. 1

Simplified geological map of the southern part of the Fennoscandian Shield including Precambrian units beneath the Phanerozoic cover in the Baltic Sea region. The map is modified after Koistinen et al. (2001) by adding information from Sundblad et al. (2003), Motuza and Motuza (2011) and Salin et al. (2019). BLU Bergslagen lithotectonic unit, LLDZ Loftahammar-Linköping deformation zone, LSGM Late Svecofennian granite-migmatite zone, OJB Oskarshamn-Jönköping belt, R Riga batholith, SLU Småland lithotectonic unit, TIB Transscandinavian Igneous Belt, V Västervik quartzites, WLG West Lithuanian Granulite domain. The TIB 0 granites: F Finspång, G Graversfors, H Hälla, AT Askersund-Tiveden area, K Kuršiai. Late Svecofennian granites: Fe Fellingsbro, L Lisjö, Va Vallentuna. The location of E6-1 and other drill sites of interest for this study are marked with red dots. Drill sites: Fr Frigsarve, N När

Svecofennian magmatic rocks have traditionally been divided into early and late orogenic (Gaál and Gorbatschev 1987). The intrusion of extensive early Svecofennian granitoids commenced almost simultaneously with calc-alkaline volcanism (Allen et al. 1996; Stephens et al. 2009; Kampmann et al. 2016). The oldest granitoid plutons consist of a differentiated suite of calcic and calc-alkaline, I-type rocks that range in composition from gabbroic to granodioritic and granitic. They form large plutonic complexes in central Finland and Sweden. Although the early Svecofennian plutonism is mostly restricted to 1.92–1.87 Ga, intrusive activity lasted locally until 1.85 Ga (Vaasjoki and Sakko 1988; Väisänen et al. 2002). A first stage of metamorphism affecting the Svecofennian crust was recorded at 1.87 Ga by Andersson et al. (2006) in the Bergslagen area, leading to the creation of a Fennoscandian proto-continent. Subsequent crustal reworking, resulting in the final cratonization of the Svecofennian units, caused the formation of migmatites, anatectic partial melts and the emplacement of late Svecofennian I- and S-type granites at 1.83–1.80 Ga (e.g. Vaasjoki and Sakko 1988; Lindh 2014).

The emplacement ages for the late Svecofennian granites span from 1.85 to 1.79 Ga in southern Finland, Late Svecofennian granite-migmatite zone (LSGM, Fig. 1 inset; e.g. Kurhila et al. 2011; Skyttä and Mänttäri 2008). In Bergslagen, the late Svecofennian granites have ages ranging from 1.84 to 1.75 Ga (Öhlander and Zuber 1988; Stephens and Andersson 2015; Johansson and Stephens 2017). The granites occur as two principal types: large homogeneous massifs of coarse-grained porphyritic K-feldspar granites (e.g. Fellingsbro, Lisjö) or smaller intrusions of heterogeneous granites associated with migmatites (e.g. Vallentuna granite complex).

Continued extension along the western border of the Svecofennian Domain resulted in the emplacement of the 1.85–1.65 Ga Transscandinavian Igneous Belt (TIB; Gaál and Gorbatschev 1987). The southern part of the Svecofennian Domain was intruded by anorogenic rapakivi granites at 1.64–1.47 Ga (e.g. Ahl et al. 1996; Rämö and Haapala 2005). The westernmost component of the Fennoscandian Shield was reworked and polyphaser-deformed during the Sveconorwegian orogeny at 1.1–0.9 Ga (Bingen et al. 2005).

Bergslagen lithotectonic unit

The southwestern part of the Svecofennian Domain is composed of two major lithotectonic units: Bergslagen and Småland (Fig. 1; Stephens 2020). The Småland unit is dominated by post-Svecofennian units, mainly TIB (Jarl and Johansson 1988; Mansfeld 1991; Kleinhanns et al. 2015), distinguished based on the mode of deformation and metamorphic conditions (Stephens et al. 1997; Korja and Heikkinen 2005). The southern border of the Bergslagen lithotectonic unit runs along the Loftahammar-Linköping deformation zone (LLDZ), which separates it from the Småland lithotectonic unit (Stephens et al. 1997; Beunk and Page 2001; Wik et al. 2005). The LLDZ is a shear belt of several kilometers width which is composed of strongly banded orthogneisses. The age dating indicates that the rocks have undergone ductile strain under amphibolite facies conditions in the time interval 1.86–1.85 Ga (Stephens and Andersson 2015), but later was reactivated in colder, brittle regimes (Stephens and Jansson 2020). The LLDZ extends from southeastern Sweden through Gotland towards the Latvian coast (Korja and Heikkinen 2005).

The Bergslagen lithotectonic unit is dominated by metamorphosed 1.91–1.86 Ga early Svecofennian granitoids, diorite and gabbro showing a calc-alkaline affinity (Stephens et al. 2009). The southernmost position for such a rock unit is 17 km south of Valdemarsvik (Fig. 1), where Stephens and Andersson (2015) documented a 1.88 Ga age for an early Svecofennian granodioritic orthogneiss. Felsic metavolcanic rocks, with subordinate carbonate and siliciclastic metasedimentary rocks of similar or older ages, are also prominent in most parts of the Bergslagen region. However, amphibolite is the dominating supracrustal rock in the southernmost segment (Valdemarsvik area). Subsequent igneous activity related to mafic underplating and anatexis took place at 1.87–1.85 Ga and 1.84–1.81 Ga and partly overlaps in space and time with the emplacement of the TIB felsic intrusive rocks (Stephens and Andersson 2015; Johansson and Stephens 2017). Late Svecofennian migmatites and pegmatites are mostly observed in the southeastern and northern parts of the Bergslagen area (Andersson 1997; Hermansson et al. 2007; Stephens et al. 2009).

The southern parts of the Bergslagen lithotectonic unit were affected by unevenly distributed ductile foliation (D1) developed prior to static recrystallization of the rocks. Later folding (D2) of the boundaries between the rock units and the tectonic foliation occurred at different scales (Hermansson et al. 2007; Stephens et al. 2009). Metamorphic temperatures correspond to amphibolite or granulite facies conditions (Stephens et al. 2009; Stephens and Andersson 2015). The southern parts of the Bergslagen lithotectonic unit were also subject to high-grade metamorphism involving partial melting with development of migmatites. Mineral parageneses with andalusite (or sillimanite), cordierite and garnet in aluminous rocks, as well as geobarometric and geothermometric data indicate metamorphism under low-P conditions at 4–6 kbar (Stephens et al. 2009; Johansson and Stephens 2017). Stephens and Andersson (2015) and Andersson et al. (2006) showed that two metamorphic events, associated with anatexis under low-pressure conditions in the central and southern parts of the Bergslagen lithotectonic unit, took place at ≤ 1.87 Ga (M1) and 1.84–1.81 Ga (M2) intimately associated with the D1 and D2 deformation events (accordingly). The age of 1854 Ga was recorded for the Askersund granites in the Tiveden area by Wikström (1996), whereas Andersson (1997) recorded an 1818 Ma age of contact metamorphism in the same area.

Småland lithotectonic unit

The Småland lithotectonic unit (Fig. 1) mainly consists of three generations of felsic intrusive rocks belonging to the Transscandinavian Igneous Belt: TIB 0 (1855–1845 Ma), TIB 1a (1808–1794 Ma) and TIB 1b (1793–1769 Ma); (Larson and Berglund 1992; Ahl et al. 2001; Salin et al. 2019). The oldest generation, TIB 0, occurs mainly in the Bergslagen lithotectonic unit (Askersund-Tiveden areas, as well as the Finspång, Graversfors and Hälla intrusions), and compositionally has been assigned by Stephens et al. (2009) to the 1.87–1.84 Ga intrusive suite of the Bergslagen lithotectonic unit. Nevertheless, the TIB 0 granitoids are also observed in the northeastern part of the Småland lithotectonic unit (Loftahammar area), which is suggested to be reminiscent of the Bergslagen lithotectonic unit inside the Småland lithotectonic unit representing a tectonic link between Bergslagen and Småland units (Wahlgren and Stephens 2020).

The TIB 0 granitoids are quartz monzonites and quartz syenites with alkali-calcic metaluminous to peraluminous geochemical signatures (Andersson 1997; Ahl et al. 2001). All other parts of the Småland lithotectonic unit are dominated by I-type quartz monzonites to granites constituting homogeneous TIB 1a and heterogeneous TIB 1b granitoid generations. In the central part of the unit, the E-W trending Oskarshamn-Jönköping Belt (OJB) was formed between 1.83 and 1.82 Ga (Mansfeld 1996; Åhäll et al. 2002) and separates the TIB 1a from the TIB 1b granitoids. The OJB is dominated by a continuous belt of calc-alkaline tonalites to granodiorites emplaced between 1.83 (Mansfeld 1996; Åhäll et al. 2002) and 1.82 Ga (Wik et al. 2003). The supracrustal rocks of the OJB include felsic and mafic volcanic rocks of the Fröderyd Group (Sundblad et al. 1997), volcanic and volcano-sedimentary formations in the Nömmen area and the Vetlanda metasedimentary formation (Röshoff 1975). The 1.85 Ga age was recorded by Salin et al. (2020) for a metamorphosed rhyolite in the Fröderyd Group indicating the oldest crustal component southwest of the Svecofennian Domain.

Southwestern limit of the Svecofennian Domain

The Västervik metasedimentary succession has a key position in the understanding of the evolution along the southwestern margin of the Svecofennian Domain. It is a km-thick continental margin sequence of fluvial, tidal and turbiditic deposits with 3.64–1.87 Ga detrital components (Sultan et al. 2005; Sultan and Plink-Björklund 2006). The presence of metabasaltic intercalations or dykes (Kresten 1972) indicates that the sedimentation was associated with some kind of rifting. The U–Pb ages of the two youngest detrital grains (1872 ± 24 Ga and 1870 ± 12 Ma) demonstrate that deposition in this basin cannot have started until shortly after the first metamorphic event (1.87 Ga) of the Svecofennian crust recorded by Andersson et al. (2006) in the adjacent Bergslagen area. The sedimentation in the Västervik basin must, however, have been completed already before 1859 ± 9 Ma, when the TIB 0 granitoids of Loftahammar intruded into the Västervik succession (Bergström et al. 2002) and sealed it to the Svecofennian granitoids and amphibolites in the Valdemarsvik area and thus the Fennoscandian proto-continent. For this reason, it is considered relevant to include the Västervik succession in the concept “Svecofennian Domain”.

Baltic Sea region

The Precambrian crystalline bedrock continues beneath the Baltic Sea where it is covered by up to 2300 m thick sequences of Phanerozoic sedimentary rocks of the East European Platform (Fig. 1; Bogdanova et al. 2015; Salin et al. 2019). The Precambrian bedrock beneath Gotland and adjacent offshore region has been studied through drillings (approximately 35 percussion drillings and drill cores; Sundblad et al. 2003; Salin et al. 2019) and geophysical investigations (EUROBRIDGE seismic working group 1999). Amphibolite is the dominant Svecofennian supracrustal rock unit beneath the sedimentary cover on southern Gotland. On the southernmost tip of Gotland, it is intruded by the TIB 1a unit (Sundblad et al. 2003). The amphibolite is also intruded by 1.88 Ga orthogneisses at När (Sundblad et al. 2003) and TIB 0 granitoids at Frigsarve (Salin et al. 2019), which correlates well with the Hälla-Valdemarsvik and Loftahammar regions.

Furthermore, early Svecofennian Pb-Pb signatures in igneous K-feldspar grains were recorded by Salin (2014) in two offshore drill cores, B-12 and B-21, east of southern Gotland (Fig. 1). The southwestern border of the Svecofennian Domain has been drawn in Fig. 1 from the southwestern side of the Västervik quartzites in the Fennoscandian Shield to the southwestern limit of the amphibolites on southernmost Gotland.

Western Latvia and Lithuania

The Precambrian rocks in western Latvia and Lithuania are also concealed by a thick cover of Phanerozoic sedimentary rocks (Fig. 1 inset), but have been studied through numerous drill cores (e.g. Skridlaite and Motuza 2001) and geophysical investigations (EUROBRIDGE seismic working group 1999; Bogdanova et al. 2006).

The Mid-Lithuanian Domain (MLD) occupies the northwestern and northern parts of Lithuania and continues farther north in the Precambrian basement of western Latvia. The MLD is a composite domain characterized by charnockitic rocks (sensu lato) and mafic granulites in the north and northwest (Kuršiai pluton) and migmatites and granites in the northeast. Supracrustal rocks of the MLD include metamorphosed felsic volcanic and sedimentary sequences. The tectonic setting was transitional from a volcanic island arc in the south to active continental margin in the north (Motuza et al. 2008; Bogdanova et al. 2015).

Based on U–Pb and Sm–Nd isotopic characteristics, Mansfeld (2001) implied that the Precambrian basement beneath the Phanerozoic cover in Latvia and Lithuania was formed between 1.9 and 1.82 Ga, which together with lithological similarities suggests a continuation of the Svecofennian Precambrian rocks in this area. Formation of charnockites and garnet-cordierite-bearing granites (Kuršiai pluton) occurred during the time interval of 1850–1815 Ma (Claesson et al. 2001; Motuza et al. 2008). The final stage of magmatism coincides with the granulite peak metamorphism recorded in the metasedimentary rocks at 1.81–1.79 Ga (Claesson et al. 2001; Skridlaite et al. 2014). This metamorphic age is not documented in the MLD charnockites, except for some alteration in the magmatic zircon at 1.79–1.74 Ga (Skridlaite et al. 2014).

The northernmost extent of the MLD is defined by the intrusion of the ~ 1.58 Ga Riga pluton, a rapakivi-anorthosite suite (Kirs et al. 2004). The western border of the MLD is not well defined, however, based on geochemical and geochronological data, Motuza and Motuza (2011) correlated the 1.85 Ga Kuršiai plutonic suite with the TIB 0 granitoids in southeastern Sweden.

Materials and methods

Materials

The Precambrian basement in the study area is covered by a km-thick sequence of Phanerozoic sedimentary rocks and ~ 100 m of seawater, and all access to study material is restricted to drill core material. The sample for this study comes from the drill core E6-1 (Figs. 1, 2), courtesy of the Geological Survey of Latvia. This drilling was conducted as a part of a program with the main purpose to document stratigraphy of the Palaeozoic sedimentary sequences. All available geophysical, geological and borehole data starting from 1980 had been used to produce geological maps of the Baltic Sea and compiled in a map description by Grigelis (1991). According to Grigelis (1991), the Phanerozoic cover in the E6-1 drill core is composed of Cambrian-Devonian sedimentary rocks (e.g. dolomites, sandstones and claystone); the Precambrian basement is encountered at the depth of 1045 m (Fig. 2a). The crystalline basement is composed of Mesoproterozoic sedimentary and volcanic rocks (1045–1056 m) and underlying Palaeoproterozoic migmatites (1056–1095 m). The migmatite is cut by a 3 cm wide lamprophyre dyke at 1095 m. Our sample from the E6-1 drill core was collected from the depth interval of 1092.0–1093.0 m and is composed of a migmatitic orthogneiss containing dark grey melanocratic and bright pink leucocratic domains (Fig. 2b, c).

Fig. 2

a Generalized geological section of the E6-1 drill hole (after Grigelis 1991). The sampling depth (1092.0–1093.0 m) is marked with a thick line; b The sample from the E6-1 drill hole investigated in this study. Migmatitic orthogneiss with grey granodioritic palaeosome and red granitic leucosome domains displaying gneissose texture; c Fracture filled with adularia that is reddish-brown colored due to micrograins of hematite. Some calcite is also present in the fractures

Analytical methods

Melanocratic and leucocratic layers are rather thin and the contact between them is gradational (Fig. 2). Separate rock pieces of each layer were cut from the core sample using a rock saw, discarding diffuse contact areas and carbonate-quartz veins. These were analyzed for geochemistry at either CAF, Stellenbosch University, South Africa, or Actlabs, Ontario, Canada (Table 1). Methods at Stellenbosch employ XRF and LA-ICP-MS from fusion glass, while Actlabs uses fusion inductively coupled plasma spectrometry (FUS-ICP) and fusion mass spectrometry (FUS-MS) methods. The F content was determined by fusion ion-selective electrode (FUS-ISE). No significant deviation in results is observed between the two laboratories.

Table 1 Chemical composition of the E6-1 migmatitic orthogneiss

For the zircon study, a composite sample of both melanocratic and leucocratic components was used. As a result of the thin layered structure and diffuse contact between the components, it was assumed that the similarity in thermochemical history between the two domains was not sufficiently distinct to warrant two separations. Furthermore, combining the components ensured a significant population of zircons from the limited material available. Zircon separates from the sample were obtained using a conventional heavy liquid method. Magnetic minerals were removed using Frantz magnetic separator. The zircon grains were handpicked from a heavy fraction under a binocular microscope and cast together with the 1065 Ma Geostandard reference zircon 91,500 in an epoxy mount (Wiedenbeck et al. 2004) for SIMS analyses. After hardening, mounts were polished to approximately half of zircon thickness to reach the core of a grain. Back-scattered electron (BSE) images were prepared to study growth features and to target the spot analysis sites using JEOL TM JSM-7100F Field Emission Scanning Electron Microscope at the Geological Survey of Finland in Espoo.

The U-(Th)-Pb analysis of the E6-1 sample was conducted using a large geometry Cameca IMS1280 Secondary Ion Mass Spectrometer (SIMS) at the Swedish Museum of Natural History in Stockholm (Nordsim facility). The SIMS instrument setup parameters generally followed those described in Whitehouse et al. (1999) and Whitehouse and Kamber (2005). A 23 kV incident energy (− 13 kV primary, + 10 kV secondary) O²⁻ primary beam was used in aperture illumination (Köhler) mode, resulting in a ~ 15 μm spot. Age calculations were performed using version 4.15 of the Isoplot software (Ludwig 2003).

Results

Petrography and nomenclature remarks

In hand specimen, the migmatite is dominated by a medium-grained melanocratic domains with gneissic structure. This irregularly alternates with a coarse-grained homogeneous granitic domains creating a gneissose banding (Fig. 2b). Melanocratic and leucocratic domains have both diffuse and sharp contacts. The rock is cut by narrow veins filled with carbonate and quartz as well as fractures filled with hematite and adularia (Fig. 2c).

Quartz, plagioclase, and biotite form up to 90% of the melanocratic domains (Fig. 3a). Rare K-feldspar and titanite grains are also observed. The titanite grains replace biotite and the plagioclase grains exhibit myrmekitic and antiperthitic textures (Fig. 3a). The plagioclase grains are sometimes replaced by clinozoisite, sericite and epidote. Small cracks in quartz and altered plagioclase crystals are filled with sericite/muscovite (Fig. 3a). At least two quartz generations are present: larger (up to 0.01 mm) crystals as well as recrystallized finer-grained (0.001–0.003 mm) metasomatic quartz replacing plagioclase and biotite. Some clusters of quartz grains are characterized by triple junctions (Fig. 3b), indicating grain boundary migration. The biotite and plagioclase grains are bent and kinked (Fig. 3c), which together with the observation of deformation twins of plagioclase (Fig. 3c) and formation of small quartz crystals, indicate deformation in a solid state. Accessory minerals are apatite, epidote, zircon, ilmenite and magnetite. Only one garnet inclusion (26 μm in diameter) was observed in zircon by EDS, Energy Dispersive X-Ray Spectroscopy.

Fig. 3

The sample from the E6-1 drill hole. a Plagioclase grains showing antiperthite unmixing and sericitic alteration in the palaeosome; b Clusters of quartz grains characterized by triple junctions in the palaeosome; c Bent plagioclase grains and deformation twins in plagioclase in the palaeosome; d Perthitic K-feldspar grain slightly altered by sericite in the leucosome; e Myrmekites in altered plagioclase grain in the leucosome; f Altered euhedral plagioclase grain surrounded by quartz in the leucosome

The leucocratic domains are dominated by medium- to coarse-grained quartz and feldspars (Fig. 3b). In hand specimen, the rock is homogeneous and massive with feldspar grains up to 1 cm in diameter. Plagioclase and K-feldspar are in equal proportions. Minor biotite is present. Alteration types in the leucocratic domains are the same as in the melanocratic domains. Plagioclase grains are partly sericitized and contain myrmekites (Fig. 3e), while K-feldspar (oligoclase) is mostly devoid of alteration and exhibits perthitic texture (Fig. 3d). Some plagioclase grains that occur within quartz are euhedral, although affected by sericitic alteration (Fig. 3f). Accessory minerals include rutile partly replaced by titanite, apatite, zircon, ilmenite and magnetite.

In terms of Sawyer (2008), neosome is a part of a migmatite newly formed by partial melting, whereas palaeosome is a non-neosome part that was not affected by partial melting. A leucosome is defined as a light-colored part of the neosome, dominantly composed of feldspar and quartz (ibid). A melanosome is a darker-colored part of the neosome, dominantly composed of biotite, garnet, cordierite, orthopyroxene, clinopyroxene and hornblende. If an anatectic melt has migrated from the place where it was formed, then it is an in-source leucosome, while an anatectic melt remained at site where the melt was formed is an in situ leucosome (Sawyer 2008). Although it is difficult to define what has been melted and what has not, a palaeosome coexisting with the neosome cannot be unaffected by partial melting, and the palaeosome can be distinguished from the neosome. Palaeosome is a rock, which preserved its structures (such as foliations, folds, layering), and the microstructures are either unchanged or slightly coarsened (Sawyer 2008). Melanocratic material in the E6-1 drill core preserved gneissose structure, and it has much smaller grain size compared to massive and coarser grained leucocratic parts. Some of the plagioclase crystals that occur within quartz are euhedral, which is interpreted to indicate crystallization from a melt. Thus, melanocratic material belongs to a palaeosome, while leucocratic material belongs to a leucosome. Although the obtained section of the E6-1 drill core did not have a definitive melanosome, we cannot exclude presence of one in the migmatite. However, lack of melanosome and some sharp contacts between melanocratic and leucocratic domains may also suggest that the leucosome was formed in an open system (in-source leucosome) implying that it has migrated within the margins of its source layer. The term “melanosome” will be used in the discussion.

Geochemistry

Rock pieces representing both the E6-1 palaeosome and the leucosome were separated and analyzed for major and trace elements (Table 1; Figs. 4, 5, 6, 7, 8, 9). For comparison, the E6-1 data are plotted altogether with major and trace element data for the I-type TIB 0 Askersund granites from the Bergslagen lithotectonic unit in central Sweden (Persson and Wikström 1993; publicly available database of the Geological Survey of Sweden, SGU database) and granodiorites intruded during the Pilsotas phase of the Kuršiai batholith in western Lithuania (Motuza et al. 2008), the A-type late Svecofennian Fellingsbro and Vallentuna granites from the Bergslagen lithotectonic unit in central Sweden (Öhlander and Zuber 1988; SGU database) and the S-type late Svecofennian Nuuksio granites from the LSGM zone in southern Finland (Nironen and Kurhila 2008).

Fig. 4

Normative feldspar classification of palaeosome and leucosome according to O’Connor (1965)

Fig. 5

Harker diagrams showing the major element composition of the E6-1 migmatitic orthogneiss (palaeosome and leucosome) compared with the TIB 0 Askersund granites (Persson and Wikström 1993; SGU database), granodiorites of the Kuršiai batholith (Motuza et al. 2008), late Svecofennian Fellingsbro and Vallentuna granites from the Bergslagen lithotectonic unit (Öhlander and Zuber 1988; SGU database) and late Svecofennian Nuuksio granites from the LSGM zone in southern Finland (Nironen and Kurhila 2008). a FeOt vs. SiO₂; b MgO vs. SiO₂; c K₂O vs. SiO₂; d Na₂O vs. SiO₂; e TiO₂ vs. SiO₂; f P₂O₅ vs. SiO₂. Blue circle—E6-1 palaeosome; red circle—E6-1 leucosome; green triangles—TIB 0 granites; pink squares—granodiorites from the Kuršiai batholith; grey squares—Vallentuna granites; empty triangles—Nuuksio granites; blue shaded area—Fellingsbro granites

Fig. 6

Harker diagrams showing selected trace element abundances of the E6-1 migmatitic orthogneiss (palaeosome and leucosome). a Sr vs. SiO₂; b Zr vs. SiO₂; c Rb vs. SiO₂. Data sources are as for Fig. 5

Fig. 7

Primitive mantle-normalized trace element abundances. a E6-1 palaeosome and leucosome compared with the TIB 0 Askersund granites and granodiorites of the Kuršiai batholith; b E6-1 palaeosome and leucosome compared with the late Svecofennian Fellingsbro granites from the Bergslagen lithotectonic unit and late Svecofennian Nuuksio granites from the LSGM zone in southern Finland. Blue line—E6-1 palaeosome; red line—E6-1 leucosome; green area—TIB 0 granites; pink area—granodiorites from the Kuršiai batholith; blue area—Fellingsbro granites; grey area—Nuuksio granites. Both diagrams after McDonough and Sun (1995)

Fig. 8

Selected tectonic discrimination diagrams for the E6-1 migmatitic orthogneiss (palaeosome and leucosome). a Rb vs. Ta + Yb diagram, b Ta vs. Yb diagram. Blue circle—E6-1 palaeosome; red circle—E6-1 leucosome; solid line—TIB 0 granites, dashed line—granodiorites from the Kuršiai batholith. Both diagrams after Pearce et al. (1984)

Fig. 9

Primitive mantle-normalized REE abundances. a E6-1 palaeosome and leucosome compared with the TIB 0 Askersund granites and granodiorites of the Kuršiai batholith; b E6-1 palaeosome and leucosome compared with the late Svecofennian Fellingsbro granites from the Bergslagen lithotectonic unit and late Svecofennian Nuuksio granites from the LSGM zone in southern Finland. Blue line—E6-1 palaeosome; red line—E6-1 leucosome; solid green line—TIB 0 granites; pink dashed line—granodioties from the Kuršiai batholith; blue dashed line—Fellingsbro granites; solid grey line—Nuuksio granites. Both diagrams after McDonough and Sun (1995). Data sources are as for Fig. 5

In the normative feldspar classification diagram after O’Connor (1965), palaeosome has granodioritic composition and the leucosome is compositionally granitic (Fig. 4). The palaeosome tends to have higher contents of FeOt, MgO, Na₂O, TiO₂ and P₂O₅ than the leucosome. This pattern is interrupted in the K₂O vs. SiO₂ diagram (Fig. 5c) by a much higher K₂O content (~ 7 wt%) in the leucosome compared to the palaeosome (~ 3.5 wt%). Selected trace element variations are plotted in the SiO₂ variation diagrams in Fig. 6 and in the primitive mantle-normalized plots in Fig. 7. The leucosome is enriched in incompatible large ion lithophile elements (LILE) Sr (209 ppm), Rb (224 ppm), Ba (876 ppm) and depleted in high field strength elements (HFSE) e.g. Zr (105 ppm) and P. Although Nb belongs to the HFSE, its content in the leucosome is extremely high (39 ppm) compared to that in the palaeosome (9 ppm). Ta bears similar chemical characteristics to Nb, but Ta abundance both in the leucosome and palaeosome is similar (0.7 and 0.6 ppm accordingly).

Both palaeosome and leucosome have volcanic-arc granite features in the tectonic discrimination diagram by Pearce et al. (1984; Fig. 8). The REE patterns are shown in Fig. 8. Both the palaeosome and leucosome are characterized by comparatively low REE content and display two different REE patterns. The palaeosome has a fractionated REE pattern with LREE enrichment, low HREE content and a negative Eu anomaly. In contrast, the leucosome has lower Sm and HREE contents and shows slightly positive or lack of Eu anomaly.

U–Pb zircon geochronology

Zircon crystals range ~ 100–200 μm in length and are pink to dark red in color. They are stubby to short-prismatic (length to width ratios < 1.5) and rounded. In BSE images (Fig. 10), zircons are largely homogeneous with faint nebulous or minor sector zoning in some grains. Such textures are common to recrystallized metamorphic zircons, particularly those formed under high-grade conditions and coexisting with anatectic melts (Pidgeon 1992; Pidgeon et al. 1998; Hoskin and Black 2000; Rubatto et al. 2001; Möller et al. 2007). Some zircon grains have dark inherited cores which are surrounded by alteration or inclusions. Oscillatory zoning is not observed in cores or in the recrystallized domains. All grains are fractured, with greater fracture density in the inherited cores (Fig. 10). Inclusions (garnet, apatite and K-feldspar) are rare and mostly appear within or immediately surrounding cores. Although zircons were separated from a composite sample containing both palaeosome and leucosome, the only evidence of separate zircon generations observed are the inherited cores in some grains.

Fig. 10

Selected BSE images of U–Pb dated zircon grains from the E6-1 granodioritic orthogneiss. Spot sites, corresponding analysis numbers and ²⁰⁷Pb/²⁰⁶Pb ages are indicated (see Table 2 for individual U–Pb data)

Table 2 SIMS zircon U–Pb dating results from the E6-1 migmatitic orthogneiss

Twenty-three zircon spots from 16 zircon crystals were analyzed (Table 2). Post-analysis inspection of the analytical spots showed that five of these points are on cracks. As it is impossible to quantify potential contamination within cracks, these points have all been excluded from calculations, even if they do not present as outliers. Two spots from inherited zircon cores have low Pb (47 and 54 ppm) and low U (113 and 130 ppm) contents, as well as Th/U ratios of 0.45 and 0.47 (Fig. 11a). Their ²⁰⁷Pb/²⁰⁶Pb apparent ages are 1859 and 1844 Ma, respectively. All analysed spots are ∼100% concordant, as can be seen in the Table 2. The remaining data show a wide spread in ²⁰⁷Pb/²⁰⁶Pb apparent ages spanning from ca. 1845 to ca. 1801 Ma (Fig. 11b). These zircons are characterized by higher Pb (224–606 ppm) and higher U (580–1556 ppm) contents as well as lower Th/U ratios ranging from 0.08 to 0.19. Metamorphic zircon, including that in equilibrium with anatectic melt, characteristically has distinctly lower Th/U ratios than that of magmatic zircon (e.g. Rubatto et al. 2001; Hoskin and Schaltegger 2003).

Fig. 11

Th-U systematics in the zircons from the E6-1 migmatitic orthogneiss. a Th vs. U concentrations for all analyzed zircons; b Th/U_meas. vs. ²⁰⁷Pb/²⁰⁶Pb ages. Blue circles—magmatic zircon cores (Th/U > 0.4); red squares—metamorphic zircon domains (Th/U < 0.2). Data points not used into calculations are not shown

Discussion

Geochemical correlation

The southern part of the Svecofennian Domain has been intruded by various felsic rocks (I-, A- and S-type granites), which have different origin and different chemical composition. When looking for a possible protolith correlation, the E6-1 migmatite was compared with the I-type TIB 0 Askersund granites and granodiorites of the Kuršiai batholith and the S-type late Svecofennian granites (Nuuksio granites) from the LSGM zone in southern Finland. In general, the late Svecofennian magmatism in the Bergslagen lithotectonic unit is younger than within the LSGM zone, but the 1.80 Ga A-type late Svecofennian Fellingsbro and Vallentuna granites were also chosen for comparison.

Variation diagrams (Fig. 5), demonstrate good correlation between the E6-1 palaeosome and the Askersund granite fields, expressed in a decrease of FeOt, MgO, Na₂O, TiO₂ and P₂O₅ with increasing SiO₂ contents. The Kuršiai granodiorites have similar trends in the above-mentioned diagrams, but shift towards higher contents of FeOt (8–10 wt%), MgO (1.7–2.3 wt%) and P₂O₅ (0.4–0.8 wt%); the Na₂O content in the Kuršiai granodiorite is also lower (2.0–2.5 wt%). All late Svecofennian granites scatter with more felsic compositions compared to the E6-1 palaeosome. Other major element compositions (except K₂O) are similar to the Fellingsbro granite; however, FeOt, MgO, TiO₂ contents in the Vallentuna and Nuuksio granites are much lower than in the E6-1 palaeosome. The late Svecofennian Fellingsbro granites have a large scatter in Fig. 5, which overlaps the field of the Askersund granites. In spite of this overlap, the palaeosome composition plots always close to the Askersund granites.

Harker diagrams for trace elements (Fig. 6) demonstrate similarities between the E6-1 palaeosome and the Askersund granites, as well as in some cases (Sr, Rb) with the Kuršiai granodiorites. There is also considerable overlap between the palaeosome and the Askersund granites in the multi-element plot (Fig. 7a). By contrast, the Fellingsbro and Nuuksio granites have much higher contents of Ba, Th, U, Ta and Nb, whereas the Nuuksio granites have stronger troughs for P, Zr and Ti (Fig. 7b).

In spite of a shift towards higher REE abundances, the Kuršiai granodiorites and the Askersund granites have similar patterns with negative slopes and small negative Eu anomalies (Fig. 9a). REEs in the palaeosome plot mostly within the lower part of the TIB 0 field, however, the E6-1 palaeosome is depleted in HREE (Er, Tm, Yb and Lu) relative to TIB 0. Besides a higher variability in the Eu abundances, the REE pattern for the Fellingsbro granites is also similar to the Kuršiai granodiorites (Fig. 9b). The E6-1 palaeosome does not correlate with the Nuuksio granites, which have a pronounced negative Eu anomaly. Previous correlations have been made between TIB 0 and the Kuršiai granodiorites (e.g. Motuza and Motuza 2011); the data presented here support this conclusion. The E6-1 palaeosome shows greater similarity to the Askersund (TIB 0) granites, but in general is more fractionated and with lower HREE than in these granites, indicating a different source or different PT conditions of melting.

In the discrimination diagrams after Pearce et al. (1984), the Askersund granites plot transitionally between the fields for the volcanic-arc granites (VAG) and within-plate granites, but dominantly within the VAG field (Fig. 8). The same characteristics are observed for the E6-1 palaeosome, albeit with the lower HREE. In turn, the Kuršiai granodiorites and the late Svecofennian Fellingsbro granites systematically have an A-type affinity shown by higher Ga/Al ratios (not plotted). This conforms with the within-plate granite signatures seen for the Kuršiai granodiorites in Fig. 8.

In the variation diagrams for major and trace elements (Figs. 5, 6), the leucosome granite shows an affinity towards the late Svecofennian granites in Bergslagen and the LSGM zone. The high K₂O content (Fig. 5c) in the E6-1 leucosome can be explained by melting of the K-rich phases (e.g. biotites and K-feldspars).

The leucosome granite plots within the field of the late Svecofennian Fellingsbro granites in the Sr and Rb vs. SiO₂ diagrams (Fig. 6a, c). The Zr content of the former is, however, closer to the late Svecofennian Nuuksio granite with Zr compositions ranging from 50 to 200 ppm (Fig. 6b). Although similar to the Fellingsbro granites, the E6-1 leucosome granite differs by the high Nb content. This can be explained by biotite melting involved in the leucosome formation: decomposing biotite released Ti, which was consumed for rutile crystallization, in turn Nb was partitioned.

Melt formation

Geochemical data for the E6-1 migmatite show that the leucosome is characterized by a more felsic and alkalic composition relative to palaeosome (Fig. 5), as well as higher contents of Rb, Ba, Sr, U and Th, which tend to be enriched in the melt, and lower contents of immobile elements (Zr, Ti, P and HREE; Figs. 6, 7), which stay in the restite. Together with thin section studies, this suggests that the granite was formed due to partial melting of the palaeosome granodiorite. This is also inferred from the (La/Yb)_PM ratios (Table 1, Fig. 9). The LREE (La) are more incompatible compared to HREE (Yb), which causes enrichment of the LREE and depletion of HREE in the melt relative to the restite (Rollinson 1993). Thus the rock formed from the melt has higher (La/Yb)_PM ratios than the rock formed from the restite. The leucosome granite studied here has higher (La/Yb)_PM ratios compared to the palaeosome granodiorite. This, combined with the slight depletion of HREE in the palaeosome relative to TIB 0, strongly supports a model of formation of the granitic leucosomes from a TIB 0 type granodioritic protolith.

The observed mineral assemblage of the palaeosome suggests that the anatectic melt was produced due to fluid-absent biotite dehydration melting as a source of the leucosome granite (e.g. Patiño Douce and Beard 1995; Patiño Douce and McCarthy 1998). Although leucosome compositions usually do not match a typical melt composition, formation of leucosomes gives evidence for partial melting and the presence of a melt phase at peak metamorphic conditions (e.g. Nehring et al. 2009, 2010).

As the palaeosome represents the parent rock and the leucosome represents melt, but no restite (melanosome) is observed, modelling is challenging. Moreover, the E6-1 migmatite has undergone sericitic alteration, which may have caused significant change in trace element patterns. However, as it was mentioned above, leucosome mineralogy evidences peak metamorphic conditions, thus restite composition can be assumed based on existing biotite dehydration reactions (e.gPatiño Douce and Beard 1995; Sawyer 1998). During experimental fluid-absent melting of a biotite gneiss, the biotite breakdown proceeds according to the following reaction (Patiño Douce and Beard 1995):

$${\text{Bt }} + {\text{ Pl }} + {\text{ Qtz}} \rightarrow {\text{granitic melt }} + {\text{ OPx }} + {\text{ Fe oxide }}\left( {\text{magnetite and ilmenite}} \right),$$

which occurs at T∼850–930 °C and P = 3–15 kbar. Dehydration-melting reaction changes at P > 10 kbar leading to crystallization of garnet between 10 and 12.5 kbar. Among other minerals at P = 12.5 kbar and T = 930 °C, quartz, biotite and rutile occur, however clinopyroxene appears while magnetite disappears. Melting of a synthetic granodioritic biotite gneiss in all experiments described by Patiño Douce and Beard (1995) produced a granitic melt in the normative Ab-Or-An classification, which is consistent with the formation of the E6-1 granitic leucosome from the granodioritic protolith (Fig. 4).

The high Rb, Ba and K contents of the leucosome granite indicate that there was insignificant melt loss, thus melting process was isochemical and equilibrium batch melting equation can be applied (Sawyer 1991). Concentration of a trace element in the melt (C_L) can be calculated from C_L = C_O/[d_RS + F(1 − d_RS)], where C_O is a trace element concentration in the source rock, d_RS is the bulk partition coefficient of the same element in the residual solid, and F is a degree of partial melting. Mineral-melt partition coefficients (D) are presented in Table 3.

Table 3 Mineral-melt partition coefficients in granodiorites

The observed mineral assemblage with coexisting rutile and garnet in the leucosome indicates relatively high-pressure and high-temperature melting, which can be modelled using the palaeosome granodiorite as the protolith composition. A theoretical restite consisting of 32% plagioclase, 33% quartz, 6% orthopyroxene, 16% biotite, 5% garnet, 3% titanite, 2% K-feldspar and 1% Fe-Ti oxides (ilmenite, magnetite, and rutile) was used based on fitting the palaeosome mineralogy to the experimental results from Patiño Douce and Beard (1995). Concentrations of P and Zr in the model were controlled by adding trace amounts (less than 1%) of apatite and zircon, respectively. In a model producing 40% melt (Fig. 12), the trace element patterns are close to the observed leucosome composition, although certain elements still show significant deviation as is discussed below.

Fig. 12

Modelled trace element composition for formation of the leucosome from 40% partial melting of the palaeosome granodiorite

The modelled melt compositions are distinctly lower in K₂O than the leucosome, which is to be expected because of the evidence of subsequent K metasomatism in the E6-1 samples. The modelled melt composition also has significantly lower Ba and Sr than the observed leucosome composition. If the K metasomatism is related to the intrusion of the nearby lamprophyre dyke, it would be reasonable to assume the same process could be responsible for the elevated Ba and Sr concentrations observed. The inverse is true with Cs, where modelled concentrations are considerably higher than those observed in the leucosome. This can be explained by retention of Cs in the restite due to the strong adsorption potential of biotite, which at these low molar concentrations would be more significant than partitioning (e.g. Lehto et al. 2019). The elevated Nb concentrations and Nb/Ta ratios observed in the leucosome could be produced by kinetic rather than equilibrium fractionation. For example, the diffusion rate of Nb in rutile is much greater than that of Ta (Dohmen et al. 2019), therefore rutile in the restite could have lost disproportionately greater amounts of Nb to the melt, increasing both Nb and Nb/Ta.

Deformation

Thin sections observation suggests that the rock was subject to subsolidus deformation which was sufficient to completely re-crystallize and re-align the quartz (mosaic texture, appearance of triple boundaries, no undulatory extinction) in the leucosome. At microscale, ductile deformation is also evidenced by the dynamic recrystallization of feldspars (perthitic textures in K-feldspars and myrmekitic textures in plagioclases), as well as deformed plagioclase and kinked biotite grains in palaeosomes and leucosomes.

A later deformation event accompanied by K-metasomatism resulted in the calcite + quartz veins and fractures filled with adularia and hematite; this is a common feature within deformation zones (e.g. Barton and Sidle 1994; Ukar and Cloos 2019). These deformation features could be a distal effect from some undetermined adjacent shear zone. The connection between this deformation and the intrusion of the nearby lamprophyre dyke is presently unclear.

Geochronology

A concordia age of 1854 ± 15 Ma was calculated from the two magmatic cores (2σ; Fig. 13). While this determination from limited data is not on its own particularly robust, it coincides with the widespread variably deformed and metamorphosed TIB 0 intrusive rocks in southeastern Sweden (e.g. Åberg and Wikström 1983; Wikström 1991; Persson and Wikström 1993; Andersson et al. 2006), southern Gotland (Salin et al. 2019) and NW Lithuania (Motuza et al. 2008). As such this, combined with the geochemical correlation presented in the previous section, indicates that this represents the igneous protolith age.

Fig. 13

U–Pb inverse concordia (Tera-Wasserburg) diagram for the inherited magmatic cores (blue ellipses; Th/U > 0.4) and the younger group of metamorphic zircon (red ellipses; Th/U < 0.2). Grey ellipses are not used in the concordia calculations; dashed ellipses are the spots on cracks and are excluded from all calculations. Error ellipses are at the 2σ level. MSWD and probability are calculated for combined concordance and equivalence

The remaining data spread downward along or close to the concordia curve from the concordant magmatic cores. These data are concordant at low end of ²⁰⁷Pb/²⁰⁶Pb ratio range but show increasing reverse discordance (older U/Pb ages than ²⁰⁷Pb/²⁰⁶Pb ages) with increasing apparent age. The large scatter in the data along concordia can be explained by incomplete resetting of U–Pb compositions due to varying degrees of partial recrystallization of zircons in both palaeosomes and leucosomes. According to Hoskin and Black (2000), the youngest age represents more complete recrystallization, and gives the best estimate for a recrystallization-inducing event. From the four youngest and concordant points, a concordia age of 1812 ± 5 Ma was calculated (2σ, MSWD = 1.6; Fig. 11), which is interpreted as the age of the migmatization. This age correlates with the age of the metamorphic event M2 recorded in the Bergslagen lithotectonic unit of the Svecofennian Domain.

The decreased Th/U ratios of metamorphic zircons are commonly explained by the preferential loss of Th⁴⁺ over U⁴⁺ during recrystallization (e.g. Hoskin and Black 2000). As the recrystallized zircons have higher U and Pb contents than the inherited cores, such a simple explanation does not apply. During anatexis, the melt would be expected to be enriched in both U and Th relative to the protolith. This, combined with the higher compatibility of U⁴⁺ over Th⁴⁺, would likely result in an increase in concentrations of both, with a corresponding decrease in Th/U during recrystallization, as is observed here. Similar characteristics were also observed by Möller et al. (2007) in zircons from migmatites from southwestern Sweden.

Reverse discordance is often produced artificially by differential sputtering during SIMS analysis, where it may be easily identified as a horizontal vector in Tera-Wasserburg concordia diagrams (Wiedenbeck 1995; Romer 2003). If the reverse discordance observed here is such an analytical artifact, these horizontal vectors may be disguised by the variation in ²⁰⁷Pb/²⁰⁶Pb that result from partial resetting of recrystallized zircons. Other studies have indicated that such discordance only becomes statistically significant in zircons with high U concentrations (> 2500 ppm; Williams and Hergt 2000; White and Ireland 2012), which is not the case here.

Alternatively, reverse discordance can be the result of a natural process involving gain of radiogenic Pb or removal of U (e.g. McFarlane et al. 2005; Corfu 2013; Kusiak et al. 2013). In such cases, data tend to form a linear discordia array similar to that observed in regular discordant data, in which intercepts indicate the timing of the crystallization (upper) and of disturbance (lower) events (e.g. McFarlane et al. 2005). The upper intercept of this discordia is at 1814 ± 40 Ma (95% confidence; MSWD = 1.02; Fig. 14), or very similar to the calculated concordia despite the much larger uncertainty in age. Using this unconstrained regression, the lower intercept is at 1016 ± 860 Ma. By assuming that the concordia age reflects the crystallization of these zircon domains, a discordia may be anchored to 1812 ± 5 Ma, in which case a lower intercept of 1066 ± 5 Ma is produced. This result likely grossly underestimates the error in this age population; therefore, a Monte Carlo solution was calculated giving a lower intercept of 1073 + 230/ − 210 Ma (95% confidence; MSWD = 0.92). While this is a very large range of uncertainty, these results consistently indicate that disturbance occurred during the early part of the Sveconorwegian orogeny at 1.1–1.0 Ga (Bingen et al. 2005).

Fig. 14

U–Pb inverse concordia (Tera-Wasserburg) diagram for the reverse discordant zircon population. Grey dashed ellipses are the spots on cracks and are excluded from calculations. Error ellipses are at the 2σ level. See text for discussion

As a consequence of recognizing that the geochemical pattern and the age of the protolith in the E6-1 drill core is comparable with the TIB 0 granitoids, the host rock into which this protolith intruded must have been older than 1854 Ma. Since there are no indications of any rock units older than the Svecofennian in the Baltic Sea region, the only realistic alternative is thus that the host rock to the E6-1 protolith belongs to the Svecofennian Domain. This means that the southwestern border of the Svecofennian Domain must be located between the site for the E6-1 drill hole and the site for the E7-1 drill hole, where a granitoid belonging to the TIB 1b generation has been identified (Salin et al. 2019).

Conclusion

Analyses of major and trace element compositions in the E6-1 migmatitic orthogneiss reveal that the protolith composition correlates with the TIB 0 Askersund granites of the Transscandinavian Igneous Belt in southeastern Sweden. Both E6-1 protolith and the TIB 0 Askersund granites show affinity towards the volcanic-arc setting. The granitic leucosome was formed due to fluid-absent biotite dehydration melting of the granodioritic protolith represented by the palaeosome. It shows geochemical similarities to other late Svecofennian anatectic melts.

Inherited zircon cores from the the E6-1 migmatitic orthogneiss show that the age of the magmatic protolith is 1854 ± 15 Ma. This age fits well with the ages of both variable deformed and metamorphosed TIB 0 granitoids and corroborates the geochemical correlation with the TIB 0 Askersund granites. The leucosome developed via partial melting of the TIB 0 protolith at 1812 ± 5 Ma, which correlates with the M2 metamorphic event in the Bergslagen lithotectonic unit. Reverse discordance may be the result of minor U–Pb disturbance occurring during the early part of the Sveconorwegian orogeny.

Geochemical and geochronological evidence suggests that the E6-1 drill site is located within the Svecofennian Domain. Combined with the evidence from the E7-1 drill site ca. 55 km to the southwest and outside the Svecofennian Domain (Salin et al. 2019), the southwestern border of the Svecofennian Domain must lie between these sites.