Energy Drive for the Kiruna Mining District Mineral System(s): Insights from U-Pb Zircon Geochronology

Abstract

The Kiruna mining district, Sweden, known for the type locality of Kiruna-type iron oxide–apatite (IOA) deposits, also hosts several Cu-mineralized deposits including iron oxide–copper–gold (IOCG), exhalative stratiform Cu-(Fe-Zn), and structurally controlled to stratabound Cu ± Au. However the relationship between the IOA and Cu-systems has not been contextualized within the regional tectonic evolution. A broader mineral systems approach is taken to assess the timing of energy drive(s) within a regional tectonic framework by conducting U-Pb zircon geochronology on intrusions from areas where Cu-mineralization is spatially proximal. Results unanimously yield U-Pb ages from the early Svecokarelian orogeny (ca. 1923–1867 Ma including age uncertainties), except one sample from the Archean basement (2698 ± 3 Ma), indicating that a distinct thermal drive from magmatic activity was prominent for the early orogenic phase. A weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1877 ± 10 Ma of an iron-oxide-enriched gabbroic pluton overlaps in age with the Kiirunavaara IOA deposit and is suggested as a candidate for contributing mafic signatures to the IOA ore. The results leave the role of a late energy drive (and subsequent late Cu-mineralization and/or remobilization) ambiguous, despite evidence showing a late regional magmatic-style hydrothermal alteration is present in the district.

1. Introduction

The northern Norrbotten ore province, Sweden, is broadly described as an iron oxide–copper–gold (IOCG) district (cf. [1,2,3,4,5]), and the mining activities related to iron oxide–apatite (IOA) deposits and Cu ± Au mineralization are important for supplying metals to Europe. The large volume of Na-(Ca) alteration (manifested as albite, scapolite, and actinolite) and abundance of low-Ti iron oxide ores (Kiirunavaara, Malmberget) together with structurally controlled hydrothermal Cu ± Au occurrences are shared characteristics with other IOCG districts worldwide [3,6,7]. Despite the growing body of literature on these deposits, the nature and genesis of IOCG deposits and their relationship to IOA deposits are still of major debate. IOA deposits have been suggested to represent the deeper-set roots of an IOCG system where the two deposit styles represent a single evolving magmatic–hydrothermal system (e.g., in the Great Bear Magmatic Zone and the Chilean Iron Belt; [8,9,10,11]). On the other hand, the classification of IOA and IOCG as two groups with distinct geneses or as two end members that occur with separate ore-forming components and conditions is preferred [4,5,6,7,12]. For example, some authors adopt a purely orthomagmatic genesis for IOA deposits where volatile- and iron-rich melts form from magma immiscibility [13,14,15,16], whereas in IOCG deposits, a direct link between magmatic fluids and metal sources for the ore can be ambiguous in some localities (cf. [4]). For example, surface- or basin-derived brines driven by igneous or crustal heat and metamorphic–hydrothermal fluids have been postulated fluid sources for IOCG-style deposits (cf. [4]), revealing the complications in grouping IOCG and IOA systems under one genetic model.

Due to the large and complicated nature of these ore deposit systems, investigating the possible relation between IOA and IOCG deposits utilizing a broader-scale mineral systems approach can be important for understanding the genetic processes. A mineral systems approach assesses essential ore-forming ingredients on a variety of scales such as energy drive, fluid and metal sources, transport pathways, and traps [17]. However, even if certain mineral system components are shared, if the timing of the two systems is fundamentally distinct, they cannot be attributed to forming from the same hydrothermal system, as emphasized by Groves et al. [6] and Skirrow [7].

An added element of complexity to the mineral systems approach occurs in regions with long and complicated tectonic histories, such as the Norrbotten province, which has experienced several phases of extension and deformation resulting in compound structural patterns and overprinting magmatic and metamorphic events (cf. [2,5,18,19,20,21,22]). Furthermore, the landscape has been shaped by Quaternary glaciation leaving extensive till coverage and poor outcrop exposure. Radiometric age constraints (generally U-Pb zircon and titanite and Re-Os molybdenite) indicate a significant IOA and IOCG mineralization period occurred during the early Svecokarelian (ca. 1880–1860 Ma; [5,23,24,25,26,27,28,29]). However, increasing structural and geochronology data link IOCG-style mineralization to a late phase of the tectonic evolution (ca. 1.82–1.78 Ga; [5,18,19,20,21,27]). These data raise an important question about the timing of introduction and/or remobilization of Cu ± Au when assessing the genetic relationship between IOA and IOCG systems in Norrbotten.

In this study, the energy drive of the Kiruna mining district mineral systems is investigated. Energy drives can include, for example, metamorphic or magmatic events, or sedimentary thickness [30]. Metamorphic events correlating to the Svecokarelian orogeny in Norrbotten have been dated by several authors (cf. [31] and references therein), however, ages related to the early Svecokarelian metamorphic event are rare in literature [25,32], with results more often yielding late orogenic or younger ages (cf. [31] and reference therein, [33]). In the Kiruna mining district, structural evidence suggests the area was at a higher crustal level and too shallow to develop ductile fabrics during the early Svecokarelian [21]. Moreover, metamorphic fingerprints from the early (M₁) phase are either absent, masked, or unidentified (cf. [21]). Consequently, this study focuses on the absolute timing of magmatic intrusions as a more robust way to elucidate the timing of energy driving processes. Both the early and late Svecokarelian orogenic phases overlap with magmatic events and highly saline fluids affected the district rocks and local ore deposits, interpreted to reflect mobilized evaporite components from elevated heat and fluid flow from magmatism [5,26,34,35,36,37]. The samples collected for the study were taken from an area situated 5 km south of the giant Kiirunavaara IOA deposit in spatial proximity to the Rakkurijärvi (IOCG-style) deposit and Pahtohavare epigenetic Cu (Fe-Au) deposits. Importantly, while the Rakkurijärvi deposit has been constrained radiometrically to the early Svecokarelian mineralizing period (Re-Os molybdenite, U-Pb allanite, rutile; [5,26,27]), the age of the Pahtohavare deposits is unknown. However, previous studies have found evidence for a partial contribution of magmatic fluids on the ore formation at both localities studied (fluid inclusions [5,34,36,37], Cl/Br ratios and δ³⁷Cl on scapolite [38], C and O isotopes on carbonate [26], Sm-Nd on titanite [39]), indicating the magmatism likely served as an energy driver and possibly a metal and fluid supplier for the mineral systems.

The focus of most geochronology studies in the Kiruna mining district has been on the plutonic and volcanic rocks within the vicinity of the Kiirunavaara IOA deposit, which has indicated a dominance of early Svecokarelian igneous activity [23,24,28]. One study [24], however, showed that a late Svecokarelian intrusion (U-Pb ICP–MS on zircon and titanite) occurs approximately 10 km southwest of Kiirunavaara implying that both regional magmatic events affected the Kiruna mining district. With two overprinting magmatic episodes potentially driving two time-separated mineral systems in Kiruna, it is important to illuminate the ages of the plutonic activity to holistically understand the timing of ore formation/remobilization. The samples taken for this study are representative of the intrusive bodies exposed at the current erosional level in Kiruna. Although they are from a relatively small sampling area, the data from this study taken with previous geochronological results contribute to the understanding of IOA and IOCG mineral systems and are placed into context of the regional tectonic framework.

2. Geological Setting

The nomenclature of the rocks used in this paper follows that described by Martinsson [40] where the “meta” prefix is excluded for clarity and emphasis on the original protolith (as all rocks are variably metamorphosed to greenschist or lower amphibolite facies).

2.1. Regional Geologic Setting

The underlying basement rocks of the Norrbotten craton are exposed just north of Kiruna in the Norrbotten province (Figure 1) and consist of Neo-Archean (ca. 2.9–2.6 Ga) gneissic granitoids of tonalitic to granodioritic composition, amphibolites, and paragneisses [2,41] formed from continental buildup during the Lopian orogeny [2]. During the early Paleoproterozoic, rifting of the basement (ca. 2.5–2.0 Ga) generated a prominent greenstone belt of tholeiitic volcanic and volcaniclastic rocks, and rift-related sedimentary successions [42,43,44,45] spanning from Norway to Russia Karelia [43].

A shift from rifting and basin development to a subduction system marked the beginning of the polyphase Svecokarelian accretionary orogenic cycle (ca. 1.90–1.78 Ga) with related arc and back-arc magmatism [2]. Approximately three to four successive mid- to late-Paleoproterozoic deformation events (described below) are recorded in the Norrbotten lithotectonic unit and have been constrained by ages of magmatic rocks, hydrothermal alteration, and structural characterization (cf. [2] and references therein; [5,18,19,20,21,49,50]).

2.1.1. Extension and Early Orogenic Crustal Shortening

At ca. 1.90 Ga, early NE-directed subduction and back-arc extension in northern Norrbotten commenced, followed by the accretion of the Skellefte volcanic arc in the south onto the Archean craton to the north (termed D₁ in Norrbotten, but D₂ in the Skellefte district and Överkalix lithotectonic unit; Figure 1, [31,46,51,52]). Recent geochronological–structural investigations have shown that the timing of early back-arc extension, volcanism, and associated synvolcanic faulting can be bracketed within 1887 ± 3 Ma (syn-basin deposition volcanic, U-Pb zircon; [22]) to a least ca. 1874 ± 9 Ma in the Kiruna area (age of the hanging wall of the Rektor deposit, U-Pb zircon; [28]). The igneous intrusive–extrusive suites associated to this early period of the Svecokarelian orogeny include the co-magmatic, calc-alkaline to alkali-calcic, synorogenic Haparanda suite–Porphyrite group and Perthite monzonite suite (PMS)–Kiirunavaara group [2]. Radiometric dating (U-Pb zircon) of these units indicate intrusion from ca. 1.89–1.86 Ga, with the Haparanda suite magmatism and Porphyrite group volcanics occurring slightly earlier (with overlap) than the PMS–Kiirunavaara group rocks [2]. Furthermore, Haparanda suite–Porphyrite group rocks are generally predominant in the eastern part of the Norrbotten province while the PMS–Kiirunavaara group rocks dominate in the west. Early district-scale Na-Ca metasomatism occurred during this period with the formation of albite, marialitic scapolite, magnetite, amphibole, and carbonate [2,5,20,35].

Subsequent crustal shortening (D₁) formed a penetrative tectonic fabric from an inferred NE–SW shortening direction [18,19,20,21,49,50]. The fabric development overlapped with peak metamorphic conditions (lower amphibolite) recorded by syntectonic growth of hornblende [20]. Regionally, amphibolite facies metamorphism is associated to D₁, but in the Kiruna mining district the metamorphism is of upper greenschist facies.

2.1.2. Late Orogenic Crustal Shortening

The second phase of the Svecokarelian orogeny (D₂; ca. 1.82–1.78 Ga) is characterized by E–W crustal shortening and basin inversion as recorded in the central Kiruna area [21]. The deformation is associated with brittle–plastic conditions producing spaced S₂ cleavages and strong strain partitioning [18,20], in contrast to the mostly plastic conditions of D₁. The timing of the late orogenic period is broadly constrained to ca. 1.82–1.78 Ga from radiometric ages from metamorphic minerals and plutonic rocks and by using cross-cutting relationships of tectonic fabrics with known plutonic ages [2,20,32,33,53]. However, a direct geochronological constraint on deformation has recently been obtained on syntectonic titanite (U-Pb LA-ICP-MS) approximately 5 km east of the Kiruna district, which indicates the hosting shear zone was active around 1812 Ma during an event lasting up to 20 m.y. [22].

Widespread anatectic S-type, silica-rich granites of the Lina suite [2,53,54] and quartz-poor monzonitic A-/I-type granitoids, syenites, and gabbros of the Transscandinavian Igneous Belt (TIB, e.g., Edefors suite) intruded in northern Sweden during D₂ [32,53,55,56]. The associated metamorphic conditions have been suggested to be of high temperature, low pressure characteristics related to the abundant magmatism [2,18]. Commonly, a potassic–ferroan alteration is associated to D₂ structures in Norrbotten [19,20]. For example, calcic and potassic–ferroan metasomatism is observed in late orogenic structures in the Nautanen deformation zone with the assemblage K-feldspar–quartz–biotite–amphibole–magnetite–hematite–epidote–scapolite–sulfide commonly spatially associated to syn-D₂ granitoids and pegmatites [18,19]. However, K-bearing alteration is also observed related to early Svecokarelian ore-forming events (e.g., Rakkurijärvi, Per Geijer iron ores; [5,26]). In addition, radiometric dating suggests that a late sodic–calcic alteration style also developed during D₂ [22], implying that the hydrothermal alteration associated to the early- and late-cycle tectonomagmatic events in Norrbotten was of similar character, complicating the usage of mineral parageneses as vectors toward specific ore deposits.

The latest structures recorded in Norrbotten reflect the last stages of the Svecokarelian orogeny and the transition into a post-orogenic collapse [19,21]. A D₃ event records a clockwise-rotation of the stress field into a NNW–SSE- to N–S-orientation with crustal shortening that resulted in a gentle refolding of pre-existing fabrics and crenulation of white mica and chlorite domains [21]. The fourth event is brittle in nature and cuts earlier fabrics [21]. An example includes a late calcite–quartz hydraulic brecciation at the Luossavaara and Rektor IOA deposits that cross-cuts an S₂ fabric [21]. Additionally, hydrothermal U-Pb monazite ages from the Rektor and Kiirunavaara deposits [57,58] and U-Pb ages of stilbite, titanite, and apatite from open fractures at the Malmberget deposit [59] indicate two disturbance events at ca. 1.74 Ga and 1.62 Ga that are possibly coeval with the hydraulic fracturing and constrains the latest tectonothermal events identified so far in northern Norrbotten.

2.2. Local Geology

2.2.1. Stratigraphy and Local Intrusive Rocks

The stratigraphy in the Kiruna area, here called the Kiruna mining district (defined by Figure 2), is composed of well-preserved rocks and comprehensive overviews of the stratigraphy have been made by Lundbohm [60], Frietsch [61], Bergman et al. [2], Martinsson et al. [5], and Andersson et al. [21].

In the Kiruna mining district (Figure 2) the stratigraphic units young to the east. The lowermost units unconformably overly the Archean basement and belong to the Kovo group, deposited during to the earliest episodic rifting period in the Paleoproterozoic (ca. 2.5–2.3 Ga, [43]). The Kovo group is a 1–2 km thick package [41] of conglomeritic and volcaniclastic rocks, tholeiitic lavas, and calc-alkaline mafic to intermediate volcanics [43]. The best-preserved stratigraphy of this rock succession outcrops north of Kiruna in the Kovo area. Overlying the Kovo group is the Kiruna greenstone group, which hosts the Viscaria and Pahtohavare syngenetic and epigenetic Cu (±Au ± Fe ± Zn) deposits [43]. The Kiruna greenstone group consists of a series of basaltic lavas, conglomerates, and dolostones at the base with overlying successions of komatiitic to tholeiitic volcanics, calc-alkaline volcanosedimentary formations, thin layers of graphitic schists, and pillow basalts. These greenstones are also intruded by mafic sills related to the upper pillow basalt subunit [43].

The Rhyacian Kiruna greenstone group is unconformably overlain by the mid-Orosirian Kurravaara conglomerate, a polymictic, poorly sorted conglomerate composed of rounded to subrounded clasts of approximately 0.5 to >10 cm (locally up to 50 cm) in size [61] (Figure 2). The Kurravaara conglomerate is overlain by trachyandesitic volcanic rocks of the Hopukka formation that make up the footwall of the Kiirunavaara IOA deposit. The hanging wall to the deposit consists of dacitic–rhyolitic volcanic rocks of the Luossavaara formation (ca. 1890–1874 Ma) [23,28,40,63]. The Luossavaara formation is overlain by units of rhyolitic tuffs, basaltic lavas, breccia conglomerates, and graywackes of the Matojärvi formation [40]. Finally, these formations are overlain by the Hauki quartzite, composed of arenites and quartz–arenites that display well-preserved cross-bedding, and breccia–conglomerates. The unit represents the youngest rocks of the stratigraphy. The eastern boundary of the Hauki quartzite marks a tectonic repetition where a greenstone unit is juxtaposed next to the Hauki quartzite [64].

Both regionally and locally, the designation of Svecokarelian intrusions to a specific time period has been made based on the foliation development and chemistry (e.g., foliated calc-alkaline Haparanda suite versus weakly foliated alkali–calcic PMS), mineralogy (e.g., less mafic minerals belonging to Lina suite equivalents), or with magnetic field patterns [2]. However, these methods may result in misleading classifications where strain partitioning is prevalent and where mineralogic compositions overlap between igneous suites. Geochronological investigations have been carried out on the intrusive bodies in direct vicinity to the IOA deposits in the central Kiruna area [23,24,28], but are sparse in other parts of the Kiruna mining district, such as in the south where the Pahtohavare and Rakkurijärvi deposits occur in spatial proximity to igneous intrusions (Figure 2). Currently, most intrusions are classified as PMS suite and are suggested to represent the deeper magma chambers that produced the Kiirunavaara group volcanic rocks that host the IOA deposits [40,65]. In the southern part of the district, only one intrusion has been dated (1792 ± 4 Ma, U-Pb inductively coupled plasma mass spectrometry (ICP–MS) intercept age from zircon and titanite fractions; Figure 2; [24]) and is the only known igneous body of ca. 1.80 Ga (D₂) age in the area. For a summary of geochronology data on the plutonic and volcanic rocks in the Kiruna mining district, see Figure 3.

2.2.2. Ore Deposits and Alteration

A variety of ore deposit types are present in the Kiruna mining district including the Kiirunavaara and other IOA deposits, stratiform–stratabound Cu (Fe-Zn) deposits (i.e., Viscaria, Eastern Pahtohavare), epigenetic Cu ± Au deposits (Pahtohavare), and IOCG-style deposits (Rakkurijärvi; [26,43]). Of these deposit types, the IOA, the Pahtohavare epigenetic Cu ± Au, and the Rakkurijärvi IOCG deposits are interpreted to have formed during the Svecokarelian orogeny (ca. 1.90–1.78 Ga, cf. [5,26,28]).

The Kiirunavaara and other IOA deposits (Figure 2) in the Kiruna mining district have been studied for over a century. The Kiirunavaara deposit occurs within the Kiirunavaara group volcanics as a tabular low-Ti magnetite–apatite orebody with magnetite–actinolite ore-breccias developed at both the hanging wall and footwall contacts [2]. Albitization occurs in the footwall and varying amounts of actinolite, albite, biotite, and scapolite occur as common alteration minerals around the deposit [2,5]. Various hypotheses about the genesis of the Kiirunavaara deposit include magmatic intrusive [67,68,69] to extrusive [13,70] to sedimentary–exhalative [71], and hydrothermal [72]. The timing of IOA formation in Kiruna has been constrained by radiometric dating of both host rocks and associated alteration and ore minerals [5,23,24,27,63], which illuminates the close temporal relation of the PMS–Kiirunavaara group and the IOA deposits. The closest direct age determination has been done by Westhues et al. [28] on zircon grains extracted directly from the Kiirunavaara and Nukutus magnetite ores that yielded ages of 1874 ± 7 Ma and 1877 ± 4 (U-Pb secondary ion mass spectrometry (SIMS)), respectively, which overlap with a titanite age from a decimeter-wide calc-silicate alteration zone at the footwall contact to Kiirunavaara (1878 ± 4 Ma, U-Pb thermal ion mass spectrometry (TIMS); [5]).

The Rakkurijärvi IOCG ores occur approximately 5 km SSE of the Kiirunavaara IOA (Figure 2) and are hosted in the Kurravaara conglomerate and Kiirunavaara group volcanics. Rakkurijärvi is described as an IOCG deposit with the ore occurring primarily in a range of sulfide-mineralized breccia styles [26]. The host rocks are brecciated to varying degrees and the main mineralization occurs as magnetite breccias with chalcopyrite–calcite matrix infill. Additionally, within the Kurravaara conglomerate Cu mineralization occurs and selective–pervasive magnetite replacement of conglomerate clasts suggests some mineralization to be a result of hydrothermal replacement [26]. Na-Ca-Fe alteration (albite–scapolite–actinolite–magnetite) accompanied the early magnetite precipitation in the rocks, which was overprinted by K-alteration (K-feldspar–scapolite–biotite) and a later low-temperature epidote–hematite–muscovite–chlorite assemblage [26]. C and O isotopes of calcite from veins, breccia matrices, and marble indicate that the Rakkurijärvi ore-related calcite derived isotopic signatures from both magmatic and marine carbonate sources [26]. Furthermore, εNd isotope data from titanite at Rakkurijärvi indicate an overlap with the regional values for granitoids from the early Svecokarelian [26,39]. The deposit is dated by Re-Os (molybdenite: 1862 ± 6 Ma and 1853 ± 6 Ma) and U-Pb (laser ablation (LA)–ICP–MS on allanite: 1854 ± 18 Ma, and TIMS on rutile: 1859 ± 2 Ma) constraining the timing of formation to ca. 1.86 Ga [5,26,27], toward the end of the early Svecokarelian cycle (late D₁) in northern Norrbotten.

The deposits at Pahtohavare are situated approximately 5 km SW of the Kiirunavaara IOA deposit (Figure 2) and are hosted in mafic to intermediate tuffites, mafic sills, and graphite schist horizons of the Kiruna greenstone group. One ore body is interpreted to be syngenetic and to have formed with the deposition of the greenstones, whereas three deposits are described as epigenetic [73]. The epigenetic ores are structurally and lithologically controlled and have previously been suggested to be of similar age as the Rakkurijärvi deposit [5]; however, no direct age determination has been carried out. The mineralization assemblage consists of pyrite and chalcopyrite occurring disseminated and as breccia infill in pervasively albite-altered graphite schist and tuffite horizons. Additionally, pyrite and chalcopyrite occur in quartz–carbonate (calcite and ferrodolomite) veins. Thin magnetite-rich horizons occur within the hosting tuffites [34,73]. Fluid inclusion studies indicate a highly saline and hot (>500 °C) fluid was responsible for early quartz and pyrite vein precipitation [34], with Cl/Br ratios on fluid inclusion leachates reflecting mixed magmatic–hydrothermal and evaporitic fluid sources [37]. Cu contents in hypersaline brines from Pahtohavare range from ~100–500 ppm and have been compared to saline magmatic fluids from the Cloncurry District, Australia [37]. Furthermore, Cl/Br ratios and δ³⁷Cl isotope signatures of marialitic scapolite from scapolite-altered host rocks near the Pahtohavare deposit records mixed magmatic and residual brine/halite signatures [38]. Together, both direct and indirect evidence suggest that magmatic fluids were involved in the formation of scapolite alteration and mineralizing veins at Pahtohavare.

3. Materials and Methods

3.1. Geologic Sampling

Sampling was conducted both in the field and at the Geological Survey of Sweden’s National Drill Core Archive in Malå during 2019. Samples for geochronology were taken as fresh specimens from outcrops for analysis. Thin sections for microscopic investigations were prepared by Precision Petrographics Ltd. in Vancouver, Canada, for microscopic investigation. See

Table 1. Sample information.

Sample Name	Alias	Association	Relative Timing	Description	Location	SWEREF99 TM		Age (Ma, This Study)
JBHA19005	005	PMS	Early orogenic	Aplitic granite	Kalixforsbron	N 7523019	E 719786	Early orogenic
JBHA19008	008	PMS	Early orogenic	Magnetite–ilmenite gabbro	Kalixforsbron	N 7522725	E 720539	1877 ± 10
JBHA19017	017	PMS	Early orogenic	Granite	Saarijärvi	N 7527323	E 711237	1887 ± 4
JBHA19025	025	PMS	Early orogenic (Late orogenic in [24])	Quartz syenite	Saarijärvi	N 7527359	E 709173	1880 ± 7
LLPAH19065 (Drill hole PAH92002, 19 m)	065	PMS	Early orogenic	Granite	Saarijärvi	N 7528947	E 711824	1880 ± 3
LLSAR19097	097	Neoarchean	Archean	Granite	Saarijärvi	N 7526604	E 709743	2698 ± 3
LLPAH19098	098	Porphyrite group?	Early orogenic	Porphyritic quartz andesite	Pahtohavare	N 7527381	E 713693	1911 ± 12

for additional sample information.

3.2. Lithogeochemistry

Samples were crushed and powdered following conventional methods at ALS Minerals in Piteå, Sweden. Lithogeochemical total whole rock characterization data were obtained from Bureau Veritas Laboratories, Vancouver, Canada. Major elements and oxide concentrations were determined through lithium borate fusion dissolution coupled with ICP–atomic emission spectroscopy analysis, whereas trace elements (including rare earth elements (REE)) were measured with ICP–MS. Complementary elements (Mo, Cu, Pb, Zn, Ni, As, Cd, Sb, Bi, Ag, Au, Hg, Tl, Se) were processed by digesting the sample with a modified aqua regia solution (1:1:1 HNO₃:HCl:H₂O) and measured by inductively coupled plasma (ICP)–emission spectroscopy (ES) or mass spectroscopy (MS) analysis. FeO speciation was determined separately by titration. QA/QC was carried out by submitting blind duplicates and certified standards from the United States Geological Survey (AVG-2 and GSP-2) to assess reproducibility and accuracy, respectively.

3.3. U-Pb Zircon Geochronology

Samples were mechanically reduced to a grain size of ≤ 1.3 mm using a tabletop digital BB-50 Fritsch jaw crusher and an agate mill. Both the jaw crusher and the mill were subjected to a strict cleaning protocol. Heavy mineral concentrates were obtained from a 70–90 µm fraction via a hydroseparation system at Luleå University of Technology (LTU) through a two-step series. A computer-controlled CNT HS 11 device [74] pulsed water from a high potential energy source firstly through a 30 cm glass separation (J-shaped) tube to prepare a pre-concentrate. The pre-concentrate was pulsed through a 10 cm glass separation tube to obtain a final concentrate. Heavy minerals were panned in a watch glass and picked under water aided by a binocular microscope. For magnetic samples, magnetic minerals were removed in advance using a hand magnet and a Frantz Isodynamic Magnetic Separator. Zircon grains were mounted in epoxy resin with the Geostandards 91500 (²⁰⁷Pb/²⁰⁶Pb age of 1065 Ma, ²⁰⁶Pb/²³⁸U age of 1062 Ma; [75]) and the M257 zircon reference material (²⁰⁶Pb/²³⁸U age of 561.3 ± 0.3 Ma; [76]), polished, and imaged by scanning electron microscopy (SEM) for both backscatter electron (BSE) and cathodoluminescence (CL) images. Post-analysis BSE and CL imaging was also undertaken to assess the spot locations.

Samples were gold-coated and analyzed at the NordSIMS Laboratory at the Swedish Museum of Natural History using a CAMECA IMS 1280 ion microprobe. A 10 nA oxygen primary beam was generated using an Oregon Physics Hyperion H201 high-brightness, radiofrequency plasma source and focused in Gaussian mode onto the sample in approximately a 15 µm spot diameter (a 5 µm raster was retained during analysis to homogenize the focused beam over a larger area). A single ion counting electron multiplier measured the secondary ions in the mass spectrometer over 10 cycles of data. Details of the analytical procedure and calibration strategy closely follow those outlined by Whitehouse and Kamber [77] and Jeon and Whitehouse [78]. The ²⁰⁶Pb/²³⁸U ratios were calibrated using the zircon Geostandards 91500, with the session reproducibility propagated in quadrature onto the individual run uncertainties, along with an additional 0.5% uncertainty to take account of the slight discordance reported from 91500 [75]. Given the very high stability of the RF plasma primary beam and consequently restricted range in UO₂/U ratio compared to duoplasmatron sessions, a one-dimensional calibration scheme in which Pb/U = k × Pb/UO₂ [79] was utilized, with k derived from the average of the 91500 analyses in each session. The ²⁰⁷Pb/²⁰⁶Pb ratios require no calibration due to the inherently small mass bias at high mass [80] that is well within analytical uncertainty on this ratio. Quality control was monitored using regularly interspersed analyses of the secondary standard M257, which yielded Pb/U ages within 1% of the accepted value from each of the three analytical sessions relevant to this study (Table S1). All data were reduced using in-house software and common Pb corrections were made following the Stacey and Kramers [81] model for terrestrial present-day ²⁰⁴Pb composition. Corrections were made on analyses where the ²⁰⁴Pb count proved statistically significant compared to the session average detector baseline [82]. U-Pb isotopic data were reduced using the Excel Isoplot add-in [83]. Decay constants follow the recommendations of Steiger and Jäger [84].

4. Results

4.1. Sample Descriptions

4.1.1. Kalixforsbron Area—Samples 005, 008

Sample 005 was collected from a granitic outcrop adjacent to a large gabbroic pluton in the Kalixforsbron area (Figure 2). The granite is fine-grained with a sugary texture (Figure 4a), and in thin section, the sample is hypidiomorphic with subequal quantities of quartz, plagioclase, and feldspar, and minor biotite, chlorite, and muscovite. Accessory minerals include zircon and apatite. Grain boundaries of quartz and feldspar crystals are serrated, and quartz shows bulging recrystallization textures (Figure 4b). The contact zone between the granite and the gabbroic pluton is characterized by hydraulic breccias associated with K-feldspar + quartz + epidote potassic alteration affecting gabbroic rocks within a structure showing both plastic and brittle deformation (Figure 4c). The zircon grains for sample 005 are pink, weakly prismatic or subhedral, and show oscillatory to irregular zonation patterns and occasional resorption embayments (Figure 4d). Many zircon grains are highly fractured with textures showing rehealed fracture planes and these areas were avoided for the analyses. Several zircon grains show an inclusion-rich “spongy” oscillatory growth zone toward the edge of the grains.

Sample 008 is a magnetite–ilmenite gabbro (Figure 4e) from within the gabbroic pluton (Figure 2). In thin section, the sample shows a weak foliation fabric of subhedral to anhedral grains of intermediate plagioclase and clinopyroxene of approximately equal proportions (Figure 4f), and interstitial magnetite–ilmenite (15%–20% modal abundance). Magnetite inclusions can be found in clinopyroxene and plagioclase. Fine-grained reaction coronas occur between magnetite–ilmenite and plagioclase grains (Figure 4f). Ilmenite exsolutions occur in magnetite, and apatite is abundant in the sample (Figure 4f). Seven zircon grains were collected from 5 kg of sample. These were transparent and subhedral to rounded, oscillatory zoned, and with rims that show a very bright CL signal (Figure 4g).

4.1.2. Saarijärvi Area—Samples 017, 025, 065, 097

Sample 017 from the Saarijärvi area (Figure 2) was sampled from a strongly magnetic (magnetite-bearing), medium-fine-grained granitic rock rich in potassium feldspar, with patchy areas of coarse quartz–potassium feldspar alteration (Figure 5a). Optical microscopy reveals hypidiomorphic equigranular with myrmekitic to granophyric textures of quartz and albite occurring interstitially between feldspar grains and as grain boundary replacement textures (Figure 5b). K-feldspar replaces plagioclase grains in patchy textures. Zircon grains from the sample are pink and have narrow prismatic, broad prismatic, and stubby shapes (Figure 5c). Internal textures are generally homogenous with limited irregular to concentric internal zonation. The zircon grains have a moderate crack density and are often broken (Figure 5c).

Sample 025 was collected from a pink, potassic-altered, medium-coarse-grained syenite outcrop (Figure 5d) in the Saarijärvi area close to the locality previously dated to be 1792 ± 4 Ma (syenite from drill core; U-Pb ICP–MS intercept age from zircon and titanite fractions) by Romer et al. [24] (Figure 2), and is the only known intrusion of late orogenic age in the district. The sample in thin section shows a relict hypidiomorphic equigranular texture of mostly alkali feldspar and minor quartz, with serrated to lobate grain boundaries between alkali feldspar grains and partial conversion to a fine-grained mineral phase along cleavage planes and cracks (Figure 5e). The sample has brittle fractures filled with chlorite, rutile, and titanite. The zircon grains from the sample are transparent or with tints of yellow, orange, and red. They are prismatic with the majority showing complex zoning patterns and few showing oscillatory zoning (Figure 5f). Resorbed textures indicate multiple generations of zircon stability which were targeted in the radiometric analysis. Fracture density is moderate, and it is common to see radial fractures around cores that are dark in CL imagery, suggesting high-U cores and differential volume expansion during metamictization [85,86].

Sample 065 was taken from an intersection of fine-grained light pink granitic rock in drill core (Figure 5g), drilled north of Saarijärvi, at approximately 19 m depth (Figure 2). The granitic intersection is approximately 25 m long and has fine veinlets of actinolite, biotite, and quartz with faint red alteration halos. In thin section, the sample has a porphyritic texture with 100–200 μm alkali feldspar and plagioclase laths surrounded by a matrix of fine-recrystallized quartz and fine-grained feldspar (Figure 5h). The zircon grains from sample 065 are orange to pink and are regular prismatic to subhedral prismatic with a moderate fracture frequency. The grains show weak oscillatory zoning, sector zoning, and typically have a thin dark and partially digested rim of <10 µm (Figure 5i).

Sample 097 was collected from a medium-grained quartz-rich granitic rock (Figure 5j) to the south of Saarijärvi (Figure 2) that yielded an Archean age (2710 ± 4 Ma) in a previous study [87]. The sample shows strong deformation textures in quartz and feldspars, based on irregular extinction patterns, embayed to lobate grain boundaries, and quartz bulging quartz recrystallization textures. The sample is highly fractured with infill of biotite–chlorite, feldspar, and titanite (Figure 5k). Zircon grains from the sample have gray and pink coloration, prismatic shapes, and moderate crack density. CL and BSE imagery indicates that a range between homogenous (lacking zones) to oscillatory zonation occurs within the grains, with some grains showing thin (1–5 µm) overgrowth rims that tend to be inclusion-rich (Figure 5l).

4.1.3. Pahtohavare Area—Sample 098

Sample 098 is a porphyritic quartz–andesite (based on modal composition in thin section) with cm-scale plagioclase laths encompassed by a fine-grained biotite-rich matrix (Figure 6). It was sampled ca. 300 m SW of Pahtohavare on the south side of the NW–SE trending shear zone (Figure 2). In outcrop, the fabric varies from randomly oriented laths to small zones of fluidal texture (Figure 6a). In thin section, rounded and slightly deformed masses of quartz with biotite and minor scapolite, epidote, magnetite, and muscovite are interpreted to be relict amygdales. The coarse porphyritic grains of plagioclase are situated within a spaced and unevenly developed fabric defined by biotite. Epidote and fine scapolite porphyroblasts (100–300 µm) occur. Plagioclase grains show minor alteration to epidote, white mica, biotite, and fine-grained magnetite (Figure 6b). Zircon grains for sample 098 from Pahtohavare are transparent and prismatic with oscillatory zones (Figure 6c). Some grains occasionally record an inner zone of inclusions. Weak radial fracture patterns are occasionally observed around cores.

4.2. Lithogeochemistry

Lithogeochemistry results are shown in Figure 7a–e and full geochemistry results are presented in the Supplementary Table S2. On the Na- and K-metasomatism plot [88] (Figure 7a), all samples except two (008 and 017) indicate significant Na-alteration. The potassic-altered granite from Saarijärvi (sample 017) plots in the igneous spectrum, likely indicating a K-overprint of earlier Na-alteration, whereas the magnetite–ilmenite gabbro from Kalixforsbron (008) is interpreted as relatively unaltered. Accordingly, chemical classification plots were chosen to reduce the reliance on alkali concentrations and mobile elements. The R1–R2 chemical variation diagram [89] indicates that the felsic igneous intrusions are chemically granitic in character (granite to alkali granite) whereas the magnetite–ilmenite gabbro (008) plots in the theralite–gabbroic fields (Figure 7b). Sample 008 has anomalously low SiO₂ content (38.1 wt. %), possibly due to dilution by the high abundance of Fe-oxides. In the Yb vs. Ta discrimination plot [90] all the felsic plutonic samples fall within the volcanic-arc field (Figure 7c), though bordering with the syncollisional (005, 065, 017) and within-plate domains (017, 097). In the Ta/Hf vs. Th/Hf diagram [91], two samples (017, 065) plot on the border between within plate volcanic zone (WPVZ) and active continental margin fields (Figure 7d). The Archean sample (097) plots in the active continental margin field together with the sample 005. The spider diagram (Figure 7e) of REE concentrations normalized to C1 chondrite [92] indicates fractionation patterns with elevated light REE (LREE) and depleted heavy REE (HREE) and a strong negative Eu anomaly for the felsic intrusive samples 017 and 065 (Saarijärvi). However, sample 005 (Kalixforsbron) shows a strongly enriched LREE and strongly depleted HREE pattern with only a weak negative Eu anomaly. The magnetite–ilmenite gabbro (008, Kalixforsbron) shows a weak LREE enrichment (Figure 7e).

4.3. U-Pb Zircon Geochronology

The U-Pb data for all samples were assessed for concordia based on the relative discordia percent (percent difference between ²³⁸U/²⁰⁶Pb ± 2σ of the data point and the concordia curve at a fixed ²⁰⁷Pb/²⁰⁶Pb value, i.e., assessed by the closest approach of the error ellipse on a 2σ level to concordia, Tables S3–S9) combined with the presence or absence of textural evidence (e.g., cracks, metamict domains, inclusions). Spots where discordant values correlated with textural evidence were removed from the calculations (Tables S3–S9). Weighted average ²⁰⁷Pb/²⁰⁶Pb ages were calculated after identification of a statistically coherent data population in concordia space. Weighted average ages are presented as the preferred crystallization age of the samples to account for possible sensitivity of the concordia ages to Pb/U inaccuracies. Uncertainties reported are 95% confidence a priori errors. For details of the U-Pb datasets, post-analysis spot descriptions, and details on data used or omitted in the age calculations, refer to Supplementary Tables S3–S9.

4.3.1. Kalixforsbron Area—Samples 005, 008

Thirty-four analyses were taken from zircon grains from sample 005 that yield U concentrations between 300 and 9440 ppm, Th/U ratios between 0.1 and 1.4, and common Pb (f₂₀₆ %) values between 0.0 and 1.8% (Table S3). Neither a concordia nor weighted average age could be calculated for sample 005 due to most of the dataset yielding > 5% discordance. A lightly trimmed dataset removing the highly discordant points with correlating textural evidence (e.g., cracks) produces a weak discordia trend (MSWD = 5.2, n = 26) between ~1875 and ~470 Ma (Figure 8a). Although the results cannot be used to assign a statistically robust crystallization age for this sample, it can be concluded that this pluton intruded during the early orogenic phase of igneous activity in the Kiruna area.

Twelve analyses were taken from zircon grains from the magnetite–ilmenite gabbro (sample 008), which yield U concentrations between 20 and 840 ppm, Th/U ratios between 0.9 and 1.9, and common Pb (f₂₀₆ %) values between 0.0 and 1.2% (Table S4). The data yield a mostly concordant U-Pb dataset (Figure 8b). Two discordant ellipses (defined by ±1% age discordance) correlating to spots overlapping with cracks were removed along with one concordant ellipse that had a high uncertainty (defined by the large ellipse size, Figure 8b). The remaining concordant data yield a weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1877 ± 10 Ma (95%, MSWD = 1.3, n = 9; Figure 8b).

4.3.2. Saarijärvi Area—Samples 017, 025, 065, 097

Thirty-two analyses were taken from zircon grains from sample 017 that yield U concentrations between 250 and 7500 ppm, Th/U ratios between 0.4 and 2.0, and common Pb (f₂₀₆ %) values between 0.0 and 1.2% (Table S5). The U-Pb dataset for sample 017 is relatively discordant, and a regression line through the spots with the least textural defects (n = 20) yields a weak discordia trend between ~1885 and ~477 Ma (Figure 8c). Eight ellipses (2σ) overlap the concordia curve; however, one analysis plots younger than the concordant cluster with a ²⁰⁷Pb/²⁰⁶Pb age of 1848 ± 24 Ma (2σ) and is interpreted to possibly represent a late metamorphic zircon growth at the end of the early Svecokarelian crustal shortening phase (ca. 1.86 Ga). The remaining seven analyses yield a weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1887 ± 4 Ma (95%, MSWD = 0.25, n = 7, Figure 8c), which is proposed to best represent the crystallization age of the sample.

Twenty-six analyses were taken from zircon grains from sample 025 which yield U concentrations between 20 and 510 ppm, Th/U ratios between 0.1 and 1.8, and common Pb (f₂₀₆ %) values between 0.0 and 3.5% (Table S6). The U-Pb analyses for sample 025 show evidence of recent Pb loss with most of the discordant data deriving from analyses that intersected fractures. Three concordant analyses plot significantly away from the concordant cluster and were excluded from the age calculation. These spots each exhibited either textural evidence (e.g., interpreted mixed hydrothermal and igneous domains, cracks) or analytical uncertainty for exclusion. A regression line through the remaining dataset after trimming one additional discordant analysis intersecting a large crack yields an upper intercept age of ~1879 Ma (n = 22). Twelve concordant analyses from the sample yield a weighted average ²⁰⁷Pb/²⁰⁶Pb of 1880 ± 7 Ma (95%, MSWD = 0.9, n = 12, Figure 8d) and is interpreted as the most reliable crystallization age for the sample.

Thirty-eight analyses were taken from zircon grains from sample 065 that yield U concentrations between 160 and 4370 ppm, Th/U ratios between 0.3 and 1.2, and common Pb (f₂₀₆ %) values between 0.0 and 1.6% (Table S7). The U-Pb data from sample 065 yield a weak discordia trend between ~1884 (n = 32) and ~447 Ma (Figure 8e) with minimal trimming of discordant data. Of the data overlapping the concordia curve, two analyses were removed from the age calculation because the 2σ error ellipse did not correlate statistically with the concordant cluster and have slightly younger ²⁰⁷Pb/²⁰⁶Pb ages. The resulting weighted average ²⁰⁷Pb/²⁰⁶Pb age is 1880 ± 3 Ma (95%, MSWD = 0.9, n = 14, Figure 8e) and is taken as the crystallization age of the sample.

Twenty-four analyses were taken from zircon grains from sample 097 that yield U concentrations between 120 and 740 ppm, Th/U ratios between 0.3 and 1.1, and common Pb (f₂₀₆ %) values between 0.0 and 0.6% (Table S8). The U-Pb data for sample 097 yield an Archean age from both the regression line and the concordia cluster. Two different regression lines (Figure 8f) can be observed reflecting a disturbance in the U-Pb isotopic system at ca. 1.78 and 0.7 Ga. Both regression lines show an upper intercept age of about 2700 Ma. A concordant cluster (n = 7) yields a weighted average ²⁰⁷Pb/²⁰⁶Pb age of 2698 ± 3 Ma (95%, MSWD = 0.8, n = 7, Figure 8f), which is taken as the best estimate for the age of crystallization.

4.3.3. Pahtohavare Area—Sample 098

Sixteen analyses were taken from zircon grains from sample 098 which yield U concentrations between 190 and 2060 ppm, Th/U ratios between 0.1 and 1.3, and common Pb (f₂₀₆ %) values between 0.0 and 2.6% (Table S9). The U-Pb data from sample 098 indicate that two zircon grains are inherited with apparently different ²⁰⁷Pb/²⁰⁶Pb ages (one discordant at 2688.8 ± 4.9 and one concordant at 2210.2 ± 7.2 Ma, Figure 8g). A regression line through the data, excluding the inherited points and other spots with textural defects, results in a weak discordia trend between ~1910 and ~418 Ma (Figure 8g). Only a few data points (n = 4) are concordant, which yield a weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1911 ± 12 Ma (95% confidence, MSWD = 1.9, n = 4, Figure 8g). The weighted average ²⁰⁷Pb/²⁰⁶Pb age is interpreted to best represent the age of the sample.

5. Discussion

Assessing the energy drive of the mineral systems in the Kiruna mining district is one of the critical aspects for a holistic understanding of IOA and IOCG genesis and relationships. While the timing of the IOA deposits regionally shows a strong temporal affinity to the early Svecokarelian orogeny [5,23,24,25,27,28,63], increasing geochronological evidence indicates that Cu mineralization occurs both early and late (cf. [5,27]), drawing into question the relative timing of the largely undated Cu-mineralization in the Kiruna mining district. Regionally, at least two distinct periods of magmatotectonism occurred during the Svecokarelian orogeny [2], and in the Kiruna district, the question remains unresolved for what role these time-separated events played on the IOA and IOCG mineral systems. The following sections discuss the geochronology and lithogeochemical data and implications these results have on understanding the energy drive for the mineralization in Kiruna.

5.1. Early Igneous Activity

The oldest sample collected in the study is an Archean granite (sample 097) in the Saarijärvi area, and the age obtained (2698 ± 3 Ma, Figure 8f) confirms a previous Archean age determination for this intrusive body (2710 ± 4; [87]). It is the only known exposure of Archean rock in the Kiruna mining district. The sample shows strong Na-alteration in the lithogeochemistry data (Figure 7a) whereas evidence for potassic alteration is lacking.

Sample 098 from the southwestern side of the NW–SE trending shear zone in Pahtohavare was taken from an area mapped as the Porphyrite group [64]. The large plagioclase phenocrysts and amygdales found in the sample (Figure 6) agree with previous descriptions of Porphyrite group volcanics [40,93]. The sample yields a weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1911 ± 12 Ma (Figure 8g), which is older than the typical age for the group (ca. 1.89–1.88 Ga, [56]) but consistent with being slightly older than the Kiirunavaara group (e.g., Hopukka and Luossavaara formations, ca. 1.88–1.86 Ga [28,93], Figure 3). This age, together with a recent age determination by Andersson et al. [22] on a volcanic intercalation in the Kurravaara conglomerate (1887 ± 3 Ma), mark the earliest Orosirian volcanic activity determined in the Kiruna district.

5.2. Timing of the Magnetite–Ilmenite Gabbro Intrusion and Possible Relation to IOA Deposits

The age obtained from the magnetite–ilmenite gabbro (1877 ± 10 Ma, sample 008, Figure 8b) is the first age determination of a mafic pluton in Kiruna. The gabbro is enriched in iron oxides with approximately 15%–20% magnetite–ilmenite modal abundance, and ca. 23% Fe₂O_3(total) content. Apatite also occurs abundantly in the sample and was noted as the primary nonmagnetic heavy mineral in the final concentrates of mineral extraction for age dating. It is also SiO₂-poor (38.1 wt. %) and falls below the typical gabbroic silica content in classification diagrams (45%–52%, [94]), which may be explained by dilution due to the high amount of magnetite and ilmenite. The zircon grains from the sample are rounded with bright rims in CL imagery (Figure 4g), and consequently the possibility that these zircon grains are inherited based on the rounded shape and bright rims cannot be excluded. Therefore, the 1877 ± 10 Ma age obtained is conservatively interpreted as the maximum age of the gabbro rather than the absolute age of crystallization. However, the deformation zone that cross-cuts the gabbro, which is characterized by brecciated and potassic-altered gabbroic clasts with quartz infill, suggests that the pluton was emplaced prior to brittle–ductile D₂ deformation, tentatively placing it into the early Svecokarelian timeframe.

Although no direct genetic link can be made, the approximate timing of formation for this intrusion (1877 ± 10 Ma) overlaps with the zircon U-Pb SIMS age of the IOA emplacement in Kiruna (1877 ± 4 Ma, 1874 ± 7 Ma), determined from zircon grains extracted from magnetite ore samples [28]. The zircon grains dated by Westhues et al. [28] show complex textural characteristics including thin oscillatory zoning in the cores, irregular patchy textures, and rims with very bright CL responses. The zircon grains from the gabbro in this study do not show irregular patchy textures but share the characteristic of thin oscillatory zoning and very bright rims in CL. Troll et al. [95] obtained δ⁵⁶Fe and δ¹⁸O isotopic data from the Kiirunavaara IOA deposit that overlap with high temperature igneous and volcanic reference materials, supporting the influence of magmatic processes for the iron oxide–apatite ore. The age-dated magnetite ores from Kiruna show a juvenile signature in both εHf_i-(zircon) and εNd_{i-(whole rock)} data, which were interpreted to indicate that the ore components derived from the mafic volcanic rocks of the Kiruna greenstone group [58,72]. Leaching of the greenstones in the Kiruna mining district as a mechanism to provide iron during the IOA ore formation was argued for by Parák [96], but the model is not appropriate for other IOA deposits in Norrbotten, such as in the Gällivare district where no greenstone stratigraphy has been recorded. However, the Gällivare district also hosts a large magnetite-bearing [97] ultramafic–mafic layered complex (Dundret) that was dated to be 1883 ± 5 Ma [33], overlapping with the back-arc extension phase of the Svecokarelian orogeny and timing of IOA emplacement. The results of the current study show that both the Kiruna and Gällivare mining districts (host to the giant Kiirunavaara and Malmberget IOA deposits, respectively) have contemporaneous relationships with mafic Fe-bearing magmatism. The intrusion of mafic magma into a crustal setting hosting evaporitic sequences and a consequential unmixing of iron-rich and volatile-rich melts has been suggested as a model for IOA deposits in Norrbotten [5] and elsewhere [15,98,99]. In the Kiruna mining district, the U-Pb geochronology data from this study suggest that the generation of mafic Fe-rich magmatism seen in the Kalixforsbron area can be an alternative candidate for contributing to the juvenile εHf_i-(zircon) and εNd_{i-(whole rock)} signatures presented by Westhues et al. [58,72], which is supported by U-Pb geochronology data.

5.3. Energy Drive Insights for the Mineral System

Overall, the zircon U-Pb age results show that abundant thermal activity was occurring in the Kiruna mining district during the early stage of the Svecokarelian orogeny from ca. 1923–1867 Ma (including age uncertainties, Figure 8). The Ta/Hf and Th/Hf ratios of the samples reflect an active continental margin bordering to within-plate environments for the felsic early orogenic magmatism (Figure 7d). Furthermore, Yb vs. Ta values show the intrusions fall within the volcanic arc field and border the within-plate granite fields (Figure 7c). Though relatively few data points exist in this study, the bordering tendency of tectonic environment agrees with the observed trend for the Kiirunavaara group intermediate–felsic volcanics in the Gällivare district [33] and for PMS intrusive rocks regionally (cf. [31], Figure 7c). The volcanic arc signatures, the mafic to felsic compositional range of the intrusions, the distance to the coeval subduction zone [100], the occurrence of thick basaltic layers in the Kiirunavaara group west of Kiruna [2], and the basin stratigraphy of the Kiruna mining district [21], support an extensional back-arc environment for the northern Norrbotten region at the onset of the Svecokarelian orogeny.

Sodic alteration is recorded in all of the felsic plutonic samples (Figure 7a) reflecting the early regional Na-Ca metasomatism and additionally bracketing the maximum age of the metasomatism to syn- to post-magmatic activity. The alteration likely resulted from the elevated heat and fluid flow caused by magmatism, and the hydrothermal fluids subsequently mobilized components from highly saline sources and channeled them into large-scale shear zones [35] facilitating the widespread distribution. Only sample 017 (from Saarijärvi) shows evidence for a K-overprint indicated by coarse patchy quartz–K-feldspar alteration and elevated K₂O content. K-alteration manifests as local hydrothermal alteration proximal to ore deposits within the Kiruna mining district (e.g., hanging wall to the Per Geijer IOA ores, Rakkurijärvi IOCG; [5,26,28,101]); however, it is also observed affecting other rocks in the district (e.g., Kurravaara conglomerate, along brittle fracture planes, in brittle fault structures) indicating that a regional potassic alteration also affected the district.

The model age of sample 025 (1880 ± 7 Ma) from an outcrop in the Saarijärvi area contradicts the much younger age (1792 ± 4 Ma; U-Pb ICP–MS titanite and zircon) of the same locality (from drill core) presented by Romer et al. [24] and agrees with the other early Svecokarelian ages from the Saarijärvi area obtained in this study. Our results indicate that the crystallization of these igneous bodies occurred at the transition from extension to early D₁ compression stage [2,20,21,102] of the tectonic evolution and question the extent of late orogenic magmatism in the Kiruna area. However, it is clear that the Svecokarelian orogeny affected the zircon grains as recorded by zonation, dissolutions, and regrowth textures (Figure 5f), which occasionally yield ²⁰⁷Pb/²⁰⁶Pb individual spot ages around 1.80 Ga (Supplementary Tables S3–S7), in agreement with tectonothermal activity in a nearby shear zone [22].

The geochronology results of this study unanimously point to abundant early Svecokarelian intrusive activity and thermal drive in the Kiruna mining district. Despite this, a late thermal drive and magmatic influence cannot be excluded when other factors are considered. First, though K-alteration exhibits large variation regionally, a characteristic potassic–ferroan (e.g., K-feldspar, quartz, biotite, magnetite, epidote) alteration occurs in relation to granitoids and pegmatites of the late orogenic Lina suite [18,20]. In the Kiruna mining district, K-feldspar coatings are often observed in late structural features. Additionally, patchy quartz–K-feldspar alteration is observed affecting rocks without any known proximity to mineralization such as at sample locality 017 (Figure 5a). At various localities in the Kurravaara conglomerate (e.g., at the contact with the Kiruna greenstone group in the Pahtohavare area and to the west of Luossavaara; Figure 2), a feldspar–epidote–magnetite hydrothermal alteration assemblage occurs in late structural features, in patchy textures, and around clast rims. These observations suggest a late orogenic K-alteration affected the district, possibly linked to late magmatic fluid circulation. Second, the late orogenic magmatic rocks are regionally voluminous and outcrop extensively just outside of the Kiruna area (Figure 1). Tavakoli et al. [103] recently showed potential field modeling results that indicate these intrusions have sheetlike geometries and it may be possible that similar sheetlike intrusions underly parts of the Kiruna area that are inaccessible to sampling. Faults and shear zones in the Kiruna mining district likely played an important role for the transport of fluids and heat, and recent work has shown that basin inversion during D₂ affected the district with many fault zones being reactivated, including the northern continuation of the shear zone hosting the Rakkurijärvi IOCG deposit [21]. Andersson et al. [21] further noted that Cu-Fe sulfides occur in brittle structures kinematically constrained to this late orogenic phase. This observation, when placed in a tectonic context and viewed from a mineral systems perspective, argues that some late mineralization or remobilization occurred in the Kiruna mining district, even though the geochronology data from this study indicate a single thermal driving event tied to the early phase. It is possible that the results do not apply completely to all other parts of the district, as deeper and unexposed intrusions may exist. However, more geochronological and mineral systems studies should be done to further understand the nature and timing of the IOA and IOCG systems.

6. Conclusions

This study used a mineral systems approach constrained by a regional tectonic framework to investigate the timing of energy drives related to the Kiruna mining district mineral systems. The U-Pb zircon geochronology conducted provides valuable new age data for intrusions in the Kiruna mining district and insight into the thermal drive for the mineral systems. The dated intrusions were sampled from the southern part of the district where Cu ± Au and IOCG mineralization occur in spatial proximity to plutonic bodies. New ages were obtained on the Archean basement (2698 ± 3 Ma) and a volcanic rock (1911 ± 12 Ma) interpreted to belong to the Porphyrite group. Results from mafic to felsic plutonic rocks yield ages that overlap with the early Svecokarelian magmatic period (ca. 1923–1867 Ma including age uncertainties) and illustrate that abundant magmatic activity was occurring in the Kiruna mining district during that time, which likely provided heat (and possibly metals and fluids) to drive the early IOA- and IOCG-related mineral systems. A weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1887 ± 10 Ma for a magnetite–ilmenite gabbroic pluton overlaps in age with the Kiirunavaara IOA deposit and marks the second district (along with the Gällivare district) that has giant IOA deposits forming coevally with large Fe-oxide-rich mafic magmatism. No late Svecokarelian ages were obtained in this study, despite sampling a syenitic pluton previously dated to have crystallized in the late phase of the Svecokarelian orogeny (D₂). Our sample yielded a weighted average ²⁰⁷Pb/²⁰⁶Pb age of 1880 ± 7 Ma. However, the extensive and voluminous nature of the late magmatism regionally and observation of potassic alteration in the district suggest that a deeper or more distal thermal influence of late orogenic magmatism still played a role in the late mineral system.