SHRIMP U-Pb Geochronology of the Sakatti and Kaarrekumpu Cu-Ni-Platinum Group Element Deposits—Constraints on the Timing of Fertile Magmatism in the Central Lapland Greenstone Belt

The northern part of the Fennoscandian Shield hosts the Central Lapland greenstone belt, which is one of the largest known Paleoproterozoic greenstone belts in the world and an emerging region for the production of Cu-Ni-platinum group elements (PGEs). The Central Lapland greenstone belt is exposed over an area of ~80,000 km² and stretches across northern Norway, northern Sweden, and northern Finland. It consists of supracrustal sequences of metavolcanic and metasedimentary rocks that were deposited between 2.5 and 1.9 Ga within an intracratonic rift basin in the Archean Karelian craton (Hanski and Huhma, 2005; Köykkä et al., 2019). Much of the bedrock in the Central Lapland greenstone belt is covered by a veneer of Quaternary glacial till and numerous mires, which obscures the geology. Base-of-till sampling through the overburden has contributed to the discovery of the Kevitsa Cu-Ni-PGE deposit by the Geological Survey of Finland in the 1980s and the Kaarrekumpu and Sakatti deposits in the 2000s. The world-class Sakatti Cu-Ni-PGE sulfide deposit is located approximately 130 km north of the Arctic Circle, halfway between the currently mined Kevitsa Cu-Ni-PGE deposit and the municipality of Sodankylä. Sakatti was discovered by Anglo American in 2009 (Coppard, 2011) and is one of the most significant metal discoveries in Europe in over a century. The deposit contains resources of 44.4 million metric tonnes (Mt) at 1.9% Cu, 0.96% Ni, 0.05% Co, 0.64 g/t Pt, 0.49 g/t Pd, and 0.33 g/t Au (Anglo American, 2022). The nearby Kaarrekumpu Ni-Cu-PGE deposit (Harju, 2018) was discovered by Anglo American four years earlier in 2004 and is located ~5 km northwest of Sakatti.

In this paper, we provide high-precision SHRIMP U-Pb zircon and titanite ages that constrain the ages of crystallization of the host rocks and associated Cu-Ni-PGE sulfide mineralization at the Sakatti and Kaarrekumpu deposits. Neither of the deposits was previously dated by U-Pb method. The only available age estimation for the Sakatti deposit is an Re-Os errorchron age of 2063 ± 35 Ma fitted for the host rocks and sulfides (Moilanen et al., 2021), which showed open-system behavior as indicated by the mean squared weighted determination (MSWD) value of 2076. We discuss here the significance of the new U-Pb ages in the context of the timing of mafic-ultramafic magmatism and formation of coeval Cu-Ni-PGE deposits in the Central Lapland greenstone belt region and the relationship of magmatism and ore formation to regional metamorphism and deformation. Direct dating of mafic-ultramafic rocks can be challenging because of common low quantities of U-bearing accessory minerals, and therefore samples were carefully selected from the large drill core archive for successful dating.

The Central Lapland greenstone belt is flanked by rocks of the Karelian craton to the east and south, the Norrbotten craton to the west, and the Kola craton and the Archean Belomorian belt to the northeast (Hanski and Huhma, 2005; Hanski and Melezhik, 2013). The volcanic and sedimentary sequences were deposited between 2.52 and 1.88 Ga on Archean basement of the Karelian craton in a long-lived intracontinental rift setting (Hanski and Huhma, 2005; Köykkä et al., 2019) (Fig. 1). A tectonic switch from lithospheric extension to compression resulted in convergent tectonics between 2.05 and 1.92 Ga and to continental collision at 1.92 to 1.87 Ga (Lahtinen et al., 2005; Luukas et al., 2017; Köykkä et al., 2019). Following the lithostratigraphic classification by Hanski and Huhma (2005) and Köykkä et al. (2019), the supracrustal succession of the Central Lapland greenstone belt is divided into five groups (Fig. 1): Salla (2.52–2.44 Ga), Kuusamo (2.44–2.38 Ga), Sodankylä (2.38–2.15 Ga), Savukoski (2.20–2.05 Ga), Kittilä (2.15–1.92 Ga), and Kumpu (1.92–1.88 Ga). These groups are cut by four major stages of mafic-ultramafic magmatism at ca. 2.44, ca. 2.21, ca. 2.15, and ca. 2.05 Ga.

The Salla and Kuusamo Groups (ca. 2.52–2.44 and 2.44–2.38 Ga, respectively) constitute magmatism related to the initial rifting of the Archean craton with the extrusion of felsic and intermediate volcanic rocks (Salla Group) and subsequent widespread subaerial to subaqueous mafic volcanic rocks (Kuusamo Group) to an early synrift basin (Köykkä et al., 2019). The Kuusamo group also hosts sparse komatiitic volcanic rocks (previously Onkamo Group), which are partly deposited onto Archean craton and partly onto earlier volcanic rocks of the Salla Group (Hanski and Huhma, 2005; Köykkä et al., 2019). The volcanic sequence is intruded by ca. 2.44 Ga layered mafic igneous complexes (e.g., Koitelainen and Akanvaara; Mutanen, 1997), which mark the minimum depositional age for the Salla Group (Hanski and Huhma, 2005; Köykkä et al., 2022). This early-rift magmatism was followed by a period of increasing sedimentation that characterizes the Sodankylä Group (ca. 2.38–2.15 Ga), for which a minimum depositional age is constrained by ca. 2.2 Ga mafic differentiated sills (Hanski and Huhma, 2005; Huhma et al., 2018). The Sodankylä Group comprises mainly epiclastic sediments including arkosic quartzites with siltstone and mudstone interbeds, conglomerates, and carbonate rocks, which were deposited in synrift to early postrift shallow marine environments at ca. 2.38 to 2.22 Ga and were followed by subaerial mafic tholeiitic volcanism at ca. 2.22 to 2.15 Ga (Köykkä et al., 2019). Gradual deepening of the marine basin is marked by deposition of graywackes and graphitic-sulfidic schists of the Savukoski Group (ca. 2.15–2.05 Ga), which are intruded by ca. 2.05 Ga mafic-ultramafic complexes such as Kevitsa, giving a minimum depositional age for the Savukoski Group sediments (Mutanen and Huhma, 2001; Hanski and Huhma, 2005; Köykkä et al., 2019). The Savukoski Group sedimentation was followed by vast subaqueous volcaniclastic komatiitic-picritic magmatism at ca. 2.06 to 2.05 Ga (Saverikko, 1985; Hanski et al., 2001), indicating plume impingement beneath the rift valley and marking the onset of breakup of the Archean craton with consequent formation of a failed rift passive margin (Luukas et al., 2017; Köykkä et al., 2019). The rocks of the three lower supracrustal successions (Kuusamo, Sodankylä, Savukoski) are in tectonic contact with the Kittilä Group (ca. 2.15–1.92 Ga), interpreted as an allochthonous oceanic plateau (Rastas et al., 2001; Hanski and Huhma, 2005), which was obducted during the juxtaposition of the Norrbotten and Karelian cratons at ca. 1.92 Ga (Lahtinen et al., 2005; Luukas et al., 2017; Köykkä et al., 2019). The unconformably overlying Kumpu Group (<ca. 1.88 Ga) comprises a thick sedimentary package of sandstones, conglomerates, and siltstones with minor felsic volcanic rocks (Rastas et al., 2001), which were deposited in a foreland basin (Köykkä et al., 2019) during the early stages of the assembly of the Columbia supercontinent between 1.94 and 1.76 Ga (Zhao et al., 2004; Lahtinen et al., 2005, 2023; Daly et al., 2006; Lahtinen and Huhma, 2019). In the Central Lapland greenstone belt, thrusting, folding, and back thrusting is well-documented during these times (Luukas et al., 2017; Sayab et al., 2021). Regional metamorphism reached greenschist to lower amphibolite facies conditions with local variations in metamorphic grade and strain intensity (Tyrväinen, 1983; Hölttä et al., 2007; Hölttä and Heilimo, 2017). The geologic setting and lithostratigraphy of the greenstone belt succession and in particular the Savukoski Group komatiitic-picritic magmatism bear similarities with that of Paleoproterozoic large igneous provinces (LIPs) such as the ca. 1980 Ma Pechenga-Onega LIP (Pechenga and Imandra-Varzuga belts; Melezhik and Sturt, 1994; Lubnina et al., 2016; Köykkä et al., 2022) and the ca. 1880 Ma Circum-Superior LIP (Raglan and Thompson belts, e.g., Machado et al., 2010; Ciborowski et al., 2017).

The supracrustal succession in the Central Lapland greenstone belt is associated with four generations of mafic-ultramafic intrusions, some of which may be feeders to the greenstone belt volcanic rocks:

1. A first event includes large, layered intrusions such as Koitelainen and Akanvaara, which comprise pyroxenites and gabbros with minor peridotites and dunites (Mutanen, 1997). These intrusions cut the Salla Group volcanic rocks. Gabbroic rocks from the two intrusions yielded zircon ages of 2439 ± 3 and 2436 ± 6 Ma, respectively (Mutanen and Huhma, 2001). Both intrusions contain reef-style Cr-V-Ti and PGE mineralization.

2. A second event includes the ca. 2.21 Ga Haaskalehto diabase sills. These differentiated sills intrude the Sodankylä Group (Tyrväinen, 1983; Hanski and Huhma, 2005; Hanski et al., 2010) and form part of the craton-wide gabbro-wehrlite association magmatism (e.g., Vuollo and Huhma, 2005). The gabbro-wehrlite association sills comprise mostly gabbros with limited ultramafic phases and contain distinct albitized plagioclase and are therefore referred to as the albite diabases (e.g., Vuollo and Huhma, 2005) or karjalites (e.g., Vuollo and Huhma, 2005; Davey et al., 2020).

3. A third event comprises differentiated mafic-ultramafic intrusions in the eastern parts of the Central Lapland greenstone belt. These intrusions commonly show basal ultramafic olivine cumulates that grade to olivine gabbros, gabbros, and granophyres in upper and marginal zones. They are ca. 2.15 Ga in age and characterized by depleted mantle sources indicated by elevated ε_Nd values (>2), contrasting to near chondritic (~0) values in the ca. 2.21 Ga gabbro-wehrlite association sills (e.g., Huhma et al., 2018). The tholeiitic volcanic rocks of the Sodankylä Group show similar ε_Nd values (Huhma et al., 2018), and although they are considered to be ca. 2.25 Ga (Silvennoinen, 1992; Huhma et al., 2018), they may be younger and represent volcanic counterparts to the ca. 2.15 Ga intrusions (Köykkä and Luukas, 2021). Numerous tholeiitic volcanic sequences, intrusions, and dikes spanning ages between 2.15 and 2.12 Ga and showing depleted mantle ε_Nd signatures (~3) similar to these ca. 2.15 Ga intrusions in the Central Lapland greenstone belt are found at the southern and eastern Karelian craton and adjacent Belomorian province (cf. Vuollo and Huhma, 2005; Hanski and Huhma, 2005; Stepanova and Stepanov, 2010; Stepanova et al., 2014, 2022; Huhma et al., 2018). The ca. 2.15 Ga Rantavaara intrusion located ~5 km to the north of Sakatti hosts minor contact-type Cu-Ni mineralization (Mutanen, 2005), and recently a Co-Cu-(Au) mineralization has been discovered from ca. 2.15 Ga intrusions in the Tanhua area, ~35 km east from Sakatti (Konnunaho et al., 2022). The 2.15 Ga intrusions cut the upper parts of the Sodankylä Group and lower parts of Savukoski Group sediments and define their maximum and minimum ages, respectively (Räsänen and Huhma, 2001; Mutanen, 2005; Huhma et al., 2018; Konnunaho et al., 2022).

4. A fourth event took place at ca. 2.05 Ga and includes most of the ultramafic magmatism and related Cu-Ni-PGE deposits in the Central Lapland greenstone belt (cf. Peltonen et al., 2014). The magmatism includes intrusions and subvolcanic olivine-pyroxene cumulate bodies, which are thought to be feeder channels of the komatiitic-picritic volcanism of the Savukoski Group (Räsänen, 2008; Virtanen et al., 2024). One of the most studied of these is the Kevitsa intrusion, which contains a disseminated Cu-Ni-PGE sulfide deposit hosted by olivine pyroxenites (Mutanen, 1997; Santaguida et al., 2015; Luolavirta et al., 2018a, b). Dating of skeletal zircon grains in olivine pyroxenite near the basal contact of the Kevitsa intrusion yielded conventional U-Pb zircon ages of 2058 ± 4 and 2054 ± 5 Ma for a crosscutting diorite dike (Mutanen and Huhma, 2001). The diorite dike furthermore yielded a titanite age of 2044 ± 4 Ma, which provides a minimum age for the intrusion. Other mafic-ultramafic intrusions in the vicinity of Kevitsa yielded U-Pb zircon ages between 2.06 and 2.03 Ga, such as Puijärvi (2035 ± 8 Ma), Satovaara (2025 ± 8 Ma), Moskuvaara (2039 ± 14 Ma) (Huhma et al., 2018), and Rovasvaara (2055 ± 5 Ma; Rastas et al., 2001). The Savukoski Group komatiitic-picritic lavas yielded Sm-Nd (clinopyroxene-whole rock) ages of 2056 ± 25 Ma (Hanski et al., 2001) and 2064 ± 38 Ma (Huhma et al., 2018) and are thought to be coeval with the intrusions. Further evidence of this age range is provided by a quartz-phyric volcanic rock at Matarakoski, 10 km southwest of the Kevitsa intrusion, which gives a U-Pb age of 2048 ± 5 Ma and sets a lower age limit for the deposition of pelitic-graphitic sediments and komatiitic-picritic magmatism of the Savukoski Group (Mutanen and Huhma, 2001). The ca. 2.05 Ga rocks are characterized by ε_Nd values ranging from coeval depleted mantle values (>2) to crustal values (<–3) indicating contributions from variable crustal components (Hanski et al., 2001; Peltonen et al., 2014; Huhma et al., 2018; Puchtel et al., 2020). The ca. 2.05 Ga metallogenic event can be widely correlated in the Fennoscandian Shield (e.g., Turchenko, 1992; Hanski and Melezhik, 2013; Orvik et al., 2022; Hansen et al., 2023; see discussion in this paper) but also around the globe with overlapping ages in the Bushveld and Uitkomst Complexes in South Africa (Zeh et al., 2015; Maier et al., 2018), Mirabela-Palestina Complex in Brazil (Lazarin, 2011; Barnes et al., 2011), and Elanskii Complex in the Voronezh Massif in Ukraine (Chernyshov et al., 2012).

The Sakatti deposit is hosted by three olivine cumulate bodies, referred to as the main body and the northeast and southwest satellite bodies (Fig. 2). Nickel, copper, and PGEs reside in massive, semimassive, disseminated, and stockwork vein sulfide ore types. The massive ore can be divided into Ni- and Cu-dominant types, whereas all the other ore types are Cu dominated. In the Main ultramafic body, the mineralization types vary from vein stockworks at shallow levels to discretely stacked massive sulfide lenses accompanied by disseminated sulfides at deeper levels. The massive and vein stockwork ore types typically have sharp contacts with the host rocks (Brownscombe, 2016). The ore mineral assemblage comprises chalcopyrite, pyrrhotite, pentlandite, pyrite, magnetite, and PGE-Ni tellurides. The host rocks are olivine ortho-, meso-, and adcumulate rocks, komatiitic lavas, and (olivine) gabbronorites. The cumulate unit, which hosts most of the massive ore, is a 1,500-m-wide, 1,000-m-long, and on average 300-m-thick unit that typically dips to the north-northwest at 40° to 70°. The cumulate rocks have mostly been altered to serpentinites and serpentine-tremolite rocks with olivine and pyroxene pseudomorphs, but particularly the adcumulates contain pristine magmatic olivine, orthopyroxene, and clinopyroxene as well as chromite. Interstitial plagioclase is present in mesocumulates and orthocumulates (Brownscombe, 2016). Cumulates show evidence of internal compositional variation and layering. The unit is flanked by a komatiitic lava unit (part of the Savukoski Group), which shows a variety of rapid quenching and undercooling textures, including flow top/bottom breccias, hyaloclastite, hopper olivine, and spinifex, as well as thin basal cumulates. The contacts between the lavas and the olivine cumulates are either characterized by flow top/bottom breccias or by a gradational increase in olivine content toward coarser-grained olivine cumulates with poikilitic pyroxene. The lavas bear high alkali (Na₂O + K₂O) and plagioclase contents (Brownscombe, 2016), which may be the result of wall-rock alteration. Nevertheless, the high MgO, Ni, and Cr and low TiO₂ contents point toward a komatiitic origin. In the shallow parts of the deposit, the komatiitic lavas partly host Cu-rich stockwork vein mineralization and Cu-rich massive ores (Fröhlich et al., 2021). Coarse-grained mineralized (olivine) gabbronorites and leucogabbronorites occur as inclusions or as randomly oriented dikes or sills within the cumulate package and komatiitic lavas. They have thicknesses of between ~0.2 and 10 m and may represent an evolved melt phase. In contrast to the other host rocks of the deposit, these rocks bear magmatic brown amphibole and apatite. They show sharp, commonly chilled margins with the cumulate unit and lack spatial continuity in adjacent drill holes and therefore could represent hybrid rocks (autoliths). The olivine gabbronorites frequently contain massive, semimassive, disseminated, or interstitial net-textured sulfides, which are dominated by chalcopyrite.

Wall rocks of the mafic-ultramafic package include a heterogeneous carbonate breccia, which marks the transition zone from the footwall to the hanging wall of the deposit. The breccia unit is characterized by intense dolomite-albite-talc alteration and consequent reddish tint due to decomposition of ferrous minerals. The unit also encloses a bedded massive dolomite unit. Neighboring the breccia are volcaniclastic sediments, which are finely laminated tuffs or phyllites, with bedding defined by differing mineralogy (rutile + phlogopite vs. quartz + plagioclase: Brownscombe, 2016). The basal contact of the cumulate package is characterized by a major fracture zone, known as “the basal thrust,” described in detail by Vuorisalo (2021), which contains ~5- to 50-m intersections of anhydrite-bearing carbonate breccia that grades downward to argillaceous carbonate rocks.

As outlined by Brownscombe (2016) and Fröhlich et al. (2021), two different interpretations have been put forward for the emplacement of the Sakatti igneous bodies, namely an extrusive origin and an intrusive origin. The extrusive model postulates a large high-Mg (komatiitic) lava river system (cf. Hill, 2001) including lava tubes and differentiated cumulate rocks in basal axial depressions. The intrusive model postulates a chonolith-like conduit that formed at shallow depths, with the olivine gabbronorites reflecting late-stage melt differentiation. In both models the cumulate rocks and komatiitic volcanic rocks are part of the Savukoski Group stratigraphy (including coeval mafic-ultramafic intrusions) and are related by a common plumbing system.

The Kaarrekumpu intrusion, described in detail by Harju (2018), is hosted by laminated sediments and tuffs of the Savukoski Group and is described as a bowl-shaped intrusion that plunges to the northeast at 15° (Fig. 3). The mineralized intrusion has a strike length exceeding 2 km and is composed of a marginal series of chilled fine-grained pyroxenite (marginal pyroxenite unit) that is found in the base of the intrusion and a layered series of plagioclase-bearing lherzolite (peridotite unit). These are overlain to the north by plagioclase-bearing (olivine) websterite (upper pyroxenite unit). The base of the intrusion displays a basal reversal, where units of the intrusion do not systematically become more primitive downward, as exemplified by the contaminated marginal pyroxenite, which shows more evolved compositions compared to the peridotite unit in the base of the layered series (Harju, 2018). The marginal pyroxenite unit is pervasively altered to actinolite and contains abundant xenoliths of the Savukoski Group volcano-sedimentary rocks. The peridotite unit is locally highly altered to serpentine and talc, whereas the upper pyroxenite unit is more pristine. The layered series is cut by later gabbroic dikes of undefined age and poorly defined orientation. Based on lithogeochemistry, these dikes can be related to the Fe tholeiitic dikes, many of which bear ages between 2.05 and 2.0 Ga in the Central Lapland greenstone belt (cf. Tyrväinen, 1983; Räsänen and Huhma, 2001; Rastas et al., 2001; Vuollo and Huhma, 2005; Harju, 2018). The sulfide mineralization occurs as disseminated, blebby/patchy and net-textured types (pyrrhotite-chalcopyrite-pentlandite) and less common semimassive and massive types (pyrrhotite-chalcopyrite). The mineralized zone generally occurs in a 5- to 10-m-thick zone at the base of the intrusion.

A total of eight samples, four from each intrusion, were collected from seven diamond drill holes, with the most evolved rock types of each deposit targeted to obtain enough zircon, baddeleyite, and titanite as well as rutile for analysis. Sample descriptions have been provided in the supplementary material (App. 1). Many of the studied holes have previously been sampled for assay, and as such continuous samples of quarter core (BQ 40.2 mm diam or NQ 50.6 mm in diam) were collected across single evolved units, not overlapping changes in lithology. A thin section and sample for whole-rock analysis were also taken for comparison.

Drill holes from the central and eastern parts of the Sakatti main body contain mineralized (olivine) gabbronorites within the cumulate unit, which were sampled. The primary mineralogy of gabbronorites consist of plagioclase, clinopyroxene, orthopyroxene, olivine, chromite, and brown amphibole (probably hornblende) (Fig. 4). Some samples have been altered to scapolite (after plagioclase), tremolite-actinolite and phlogopite-chlorite (after clino- and orthopyroxene), serpentine (after olivine), and magnetite (after chromite), often preserving their original habit as pseudomorphs. Chalcopyrite, pyrrhotite, pyrite, and minor pentlandite are present as interstitial sulfide disseminations in all analyzed samples. Presence of interstitial magmatic sulfides in both the olivine gabbronorites and cumulate rocks strongly suggests that the two are cogenetic.

In the Kaarrekumpu intrusion more evolved gabbroic zones in the upper pyroxenite unit were sampled. These gabbroic zones show gradational contacts with the plagioclase websterite evidencing a cogenetic relationship. These rocks are relatively unaltered and preserve their magmatic mineralogy and textures, although some localized pervasive alteration to serpentine and talc (after olivine), tremolite-actinolite and chlorite (after pyroxenes), and clay minerals (after plagioclase) can be observed (Fig. 4). Sulfides occur as disseminated pyrrhotite, pentlandite, chalcopyrite, and pyrite. Samples selected show a gradational contact with the associated ultramafic unit or a shared sulfide mineralogy and type of mineralization.

Mineral separation was performed in the laboratories of the Research School of Earth Sciences at The Australian National University (RSES, ANU). Samples were crushed and milled and the fines washed off in a settling beaker. Magnetic minerals were separated using a hand magnet and a Frantz isodynamic separator. Heavy minerals were concentrated using tetrabromoethane and methylene iodide. Concentrated zircons and other minerals were handpicked under a binocular microscope and mounted in epoxy, together with the relevant standards. Temora II (416.8 ± 1.3 Ma; Black et al., 2004) is the primary U-Pb standard and SL13 (Claoué-Long et al., 1995) is a chip of a single crystal with a uniform U content and is used to calibrate U, Th, and Pb concentrations. For titanite, the standard used was BLR-1 (1050.5 ± 0.9 Ma; Aleinikoff et al., 2007), and for rutile the Wodgina standard (2845.4 ± 0.5 Ma; Ewing, 2011) was used. The grains were polished to half the thickness of the average grain in the mount to expose any complex internal structures. All grains were photographed in transmitted and reflected light and these, together with scanning electron microscopy (SEM) cathodoluminescence (CL) images (or backscattered electron [BSE] images for titanite and rutile). These were used to decipher the internal structures of the sectioned grains and to target specific areas for SHRIMP spot analysis.

Uranium-lead analyses were done using the SHRIMP II instrument. The data have been reduced in a manner described by Williams (1998, and references therein), using the SQUID 2 Excel macro of Ludwig (2009). The decay constants recommended by the IUGS Subcommission on Geochronology (as given in Steiger and Jäger, 1977) were used in the age calculations.

Uncertainties given for individual U-Pb analyses (ratios and ages) are at the 1σ level; however, uncertainties in the calculated weighted mean ages are reported as 95% confidence limits (2σ). For the age calculations, corrections for common Pb were made using the measured ²⁰⁴Pb and the relevant common Pb compositions from the Stacey and Kramers (1975) model. The ²⁰⁷Pb/²⁰⁶Pb ratios and ages were monitored for instrumental mass fractionation by repeated analysis of the zircon standard OG1 (Stern et al., 2009). No corrections were made for this fractionation.

All samples contained zircon in various quantities with minor titanite also recovered from seven out of eight samples, as well as rutile in one sample. The latter two minerals were difficult to date as they tend to have low U contents and high common Pb. No baddeleyite was recovered. Analytical results are shown in Table 1. Descriptions for each sample and full SHRIMP U-Pb data tables are included in supplementary materials (App. 2). The concordia plots were produced using IsoplotR software (Vermeesch, 2018).

One zircon and a few titanite and rutile grains were found in the heavy mineral concentrate of this sample. Two spots were analyzed on the zircon grain, giving 2% discordant data with a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2054.1 ± 7 Ma (Fig. 5A). Two of the six analyses done on the titanite grains showed large variations in common Pb concentrations and U-Pb systematics during analysis, and these two results were too compromised to be of any use for geochronology. The other four analyses plot on or slightly below the concordia and give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2030.9 ± 14.6 Ma (Fig. 5B). Titanite grain 5 has a very low common Pb content indicating Pb loss and, excluding that, produces a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2043 ± 20 Ma (MSWD = 0.77). A regression gives a similar, but less precise, upper intercept age of 2056.6 ± 30 Ma. Rutile was also found in the heavy mineral concentrate of this sample. Four U-Pb analyses were done on these, and three of these give a late Neoproterozoic concordia age of 600.1± 19.2 Ma (Fig. 5C).

Six zircon grains were found in the heavy mineral concentrate of this sample. All are subhedral to anhedral with weak or no zoning and dull CL response. Eight analyses were done on the six grains and the data are plotted in Figure 6A. The results are up to 10% discordant, but obviously this is a uniform, coeval population with a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2052.9 ± 2.4 Ma. Regression of the same data gives an upper intercept age of 2056.2 ± 5 Ma.

The zircons from this sample are dark brown and largely opaque and have poor form. The CL imaging shows no zoning, with the CL response mostly dark and/or mottled in areas of alteration. Some bright-CL veins and patches record zones of fluid-driven alteration. Uranium and thorium concentrations are highly variable but are generally high, with maxima of 3,922 ppm U and 8,941 ppm Th measured in spot 18.2. The high Th/U contents (up to 2.5) are typical of zircons crystallized from mafic melts. Despite these high radiogenic element contents and the poor state of the zircons, the data plot on or close to concordia (<10% discordant). There is no obvious correlation between high U or Th content and discordance. Excluding two analyses (14.1 and 16.1) that appear to have suffered significant intermediate Pb loss, the remaining 21 analyses combine to give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2052.9 ± 1.4 Ma (Fig. 6B). An attempt was made to date the titanite from this sample, but the results are difficult to interpret. Two of the four analyses had unacceptably high common Pb contents and extremely low U (<1 ppm)—a combination that renders the results too imprecise to be meaningful. The remaining two analyses plot within error of concordia but without a uniform ²⁰⁷Pb/²⁰⁶Pb age. These data are plotted and tabulated together with the titanite data from the other 12MOS8091 sample (Fig. 7E).

Both zircon and titanite were extracted and dated from this sample. The zircon crystals are dark brown and heavily opaque and have anhedral or rare subhedral form. The CL image shows bright patches and veins of fluid-driven alteration, occasional relict magmatic zoning, and some complex internal structures that suggest that there might be two generations of zircon growth. The younger generation appears to have more crystal form either as discrete grains (e.g., grain 5, Fig. 7) or as embayments or rims around the cores. In most cases the older grains tend to be uniformly dark, with flat and featureless CL response, but with strong modification around the edges (e.g., grain 10, Fig. 7). Twenty-eight spot analyses were done on 21 different grains to unravel the full history of these complex zircons. Two of these were discarded as they had extreme common Pb contents. The remaining analyses form two distinct age groups (Fig. 7A). The weighted mean ²⁰⁷Pb/²⁰⁶Pb ages calculated are 2049.2 ± 23.8 Ma for the older group comprising 10 analyses (<10% discordant) and 1779.2 ± 3.6 Ma for the younger group (10 analyses, <10% discordant). Upper intercept ages for the two groups are 2054 ± 3.8 (Fig. 7B) and 1788.1 ± 7.8 Ma (Fig. 7C), respectively. Lower intercepts indicate recent Pb loss. As shown in Figure 7A, there are several analyses that are either highly discordant or are scattered between the two groups. The majority of these are spot analyses in zircons that also show the older ages. The U and Th contents are highly variable both within and between age populations. The younger generation does, however, tend to have much lower Th/U values (<0.10). Given the complexity of the zircon data, some dating of the titanite was undertaken. The titanite grains are dark brown and have uniform BSE emissions apart from a few grains that do show lighter mottling in some areas. Nine analyses of different grains were done. Common Pb and U contents are reasonable for geochronology. The concordia plot (Fig. 7E) shows all eight spots aligning along a short discordia to give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1755 ± 11.6 Ma. Regression of these eight spots gives an age of 1785.9 ± 14.2 (Fig. 7D).

Although this sample produced a good yield of zircons, the concentrate is heterogeneous in terms of shape, color, and probably origin. The zircons range from subhedral, elongate, gray-brown crystals to anhedral, altered grains with variable zoning. Some grains have small overgrowths. To cover the full complexity of this zircon population, 25 analyses were done on 21 different grains. As expected, the results show a range of ages but with the main population defining a coherent concordant or near-concordant group with more falling away from concordia showing significant Pb loss (Fig. 8A). Most of the zircons in this group are subhedral, clear grains, but even in this group CL imaging shows some grains have very small overgrowths and have highly variable zoning and internal structures. A weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2053.8 ± 5 Ma is calculated from the most concordant analyses (n = 7) as shown in Figure 8B. Other analyses are both older and younger than this coherent group. Two grains (13.1 and 19.1) are clearly inherited from an Archaean source. A significant number of analyses plot close to concordia indicating apparent Mesoproterozoic and lower dates (ca. 1900–500 Ma). Two analyses of titanite from this sample were discordant and yielded a combined weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 1883.1 ± 32 Ma (Fig. 9).

One large dark-brown, opaque subhedral zircon grain (450 µm in length) was recovered from this sample. Most of the heavy mineral concentrate was made up of similarly large, brown titanite grains with ragged and variable shape. The zircon is highly altered and is crosscut and riddled with a bright-CL phase of what appears to be secondary or recrystallized zircon (Fig. 10A). BSE imaging of the titanite grains shows mostly uniform, unzoned internal structures apart from some subtle or patchy gray shade changes in some grains that could record different phases of growth. Four analyses were done on different parts of the large zircon grain (avoiding the altered areas), and they all plot as a coherent, slightly discordant group from which a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2049.2 ± 6.2 Ma can be calculated (Fig. 10A). Given the scarcity of zircon in this rock, most of the dating was done on titanite. This shows two distinct ages (Fig. 9). An older group comprises four analyses that are discordant but give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2107 ± 130.6 Ma (Fig. 10). A younger group also shows some significant discordance, but the six analyses are relatively tightly clustered and give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1781.6 ± 9.2 Ma (Fig. 10C). These two generations of titanite generally correlate with the structures observed through BSE imaging, with the brighter areas tending to be the younger generation. The two groups also have distinct geochemical characteristics with, for example, the younger group having lower Th/U contents.

Zircon and titanite are relatively abundant in this rock, with the zircon grains showing a wide range of crystal forms from occasional acicular crystals to more common blocky, small brown crystals that are ~50 to 70 μm in diameter. Zoning is generally weak but, where present, comprises mainly sector zoning and faint oscillatory zoning toward the margins of the grains. Some grains have very small rims and embayments (e.g., grains 1 and 9, Fig. 11). Titanite crystals are large (up to 400 μm in the longest dimension), dark brown and semiopaque. BSE imaging shows blotchy areas, but the images are generally uniform in color and structure. The zircons have moderate to low U and Th contents apart from one analysis (13.1). All 22 spots analyzed combine to give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2053.6 ± 2 Ma (Fig. 11A). The errant analysis is discordant and appears to have been affected by an intermediate event, likely the ca. 1920 to 1890 Ma event recorded elsewhere in this study. An attempt was made to date the titanites from this sample as well, with the aim being to identify and date younger events that may have affected the sample. Only one of the three analyses done had high enough U to date but gave an imprecise age that suggests a complex Pb loss history (Fig. 9).

This sample yielded some excellent quality zircon crystals that are ideal for dating. The crystals are clear, light brown, and commonly euhedral and take a range of forms from oblate to highly acicular grains with length/width aspect ratios up to 8:1 (e.g., grain 40, Fig. 11). Chunky, dark-brown titanites were also extracted and analyzed. Most of the U-Pb zircon data plot as a group that straddles the concordia curve (Fig. 11B) with all 24 analyses plotting within 10% of the curve and giving a well-constrained weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2054.5 ± 2 Ma. Regression of all the data gives an upper intercept age of 2054.9 ± 2.2 Ma. Three analyses of the dark-brown titanite from this 11MOS7028 sample did not produce a consistent age in terms of their measured ²⁰⁷Pb/²⁰⁶Pb dates, and it is probable that they are recording the effects of intermediate Pb loss as mentioned in the results from the previous sample. These data are plotted in Figure 9.

The zircons analyzed show a variety of sizes, forms, and U and Th concentrations (App. 2). They tend to have little internal structure, apart from occasional weakly developed and poorly preserved sectors or oscillatory zoning. Subtle rims, embayments, or overgrowths are observed with CL imaging, but these tend to be too small to analyze with the 25-μm spot size used. In almost all cases, it is considered reasonable to interpret the zircons and their ca. 2056 to 2053 Ma U-Pb ages as igneous in origin. The generally high ²³²Th/²³⁸U (up to 2.5) measured in these zircons is characteristic for zircons that crystallized from mafic magmas. The age of both Sakatti and Kaarrekumpu is thus well constrained by the data from this suite of samples. The magmatic U-Pb age of Sakatti also gives a new minimum age estimate for the Savukoski Group komatiites and supports the Sm-Nd ages obtained by Hanski et al. (2001; 2056 ± 25 Ma) and Huhma et al. (2018; 2064 ± 38 Ma) as some of the dated gabbronorite samples are associated with the komatiitic lavas.

The magmatic ages of titanites show more discordant data and larger errors than zircons, which is due to their low U contents and elevated common Pb. Titanite is vulnerable to postcrystallization modification, even low-grade metamorphism, as indicated by high discordance in many of the titanite ages in this study. When compared to zircon data, it can be assumed that titanites showing ages around 2050 Ma, e.g., Sakatti sample 10MOS8039 (2056.6 ± 30 Ma, Fig. 5) and Kaarrekumpu sample 05MOS7008 (2107.2 ± 130.6 Ma, Fig. 10), yield magmatic ages. There is no known regional tectonic event in northern Fennoscandia at ca. 2050 Ma that would promote metamorphic age of these titanites. Titanite typically has estimated closure temperatures ca. 100° to 300°C lower (Cherniak, 1993; Hartnady et al., 2019) compared to that of zircon (974°C: Cherniak and Watson, 2000). In the case of Sakatti and Kaarrekumpu, all zircon and titanite ages are within error limits of each other, reflecting closure of zircon and titanite at the same time and emphasizing coeval and rather rapid cooling for these units at around 2056 to 2053 Ma. As the gabbronorites in Sakatti show chilled contacts with the komatiitic lavas and cumulate, the result is considered as the minimum age of the host rocks and Cu-Ni-PGE mineralization. However, olivine data (H. Höytiä, unpub. data, 2023; Virtanen et al., 2024) indicate that the gabbronorites bear Ni contents similar to that of Sakatti olivine cumulates, strongly suggesting a similar, coeval origin for these rocks.

Increased lithospheric thinning at ca. 2.15 Ga most likely marks the onset of a Paleoproterozoic plume event in northern Fennoscandia (Hanski and Huhma, 2005; Vuollo and Huhma, 2005; Stepanova et al., 2014, 2022; Huhma et al., 2018; Hansen et al., 2023). The first magmas that migrated to the lithosphere from the coeval depleted mantle (ε_Nd ~2–5; Stepanova et al., 2014; Huhma et al., 2018) were the ca. 2.15 Ga mafic (mostly tholeiitic) intrusions, dikes, and their volcanic equivalents, e.g., the Jatuli continental flood basalts of the Karelian province that are represented in Central Lapland greenstone belt by the Sodankylä Group tholeiitic volcanic rocks (cf. Lehtonen et al., 1998; Rastas et al., 2001; Hanski and Melezhik, 2013; Stepanova et al., 2014; Huhma et al., 2018; Köykkä and Luukas, 2021). This short-lived magmatism is defined as the Rantavaara LIP by Davey (2019). Initiation of a contemporaneous plume event is also recorded in North America, where Davey et al. (2022) defined a three-stage model for formation of mafic dikes and volcanic rocks—early (2135–2130 Ma), middle (2130–2113 Ma), and late (2113–2101 Ma)—based on data from Superior (Marathon LIP), Hearne (Griffin-Kazan-Chipman LIP), Wyoming (Bear Mountain LIP), and Karelia and Kola cratons (Tohmajärvi-Pirtguba LIP), which were closely positioned during the Paleoproterozoic (Davey et al., 2020). Davey et al. (2022) considered these events as precursor events of the final-stage breakup at ca. 2.07 Ga represented by dike swarms in Fort Frances and Palomaa LIPs (Huhma et al., 2018; Davey et al., 2020) in the Superior and Karelian cratons, respectively. Coeval mafic-ultramafic magmatism and volcanism is also recorded in the North China craton, where dike swarms, sills, intrusions, and coeval volcanic rocks of Hengling, Haicheng-Xiliu, Zanhuang, and Yixingzhai areas record ages of ca. 2.15, 2.12, 2.09, and 2.05 Ga, respectively (Peng et al., 2012; Peng, 2015), all regarded as separate LIPs. In the Central Lapland greenstone belt, the final stage of extension and plume migration beneath the rift valley is marked by short-duration magmatism that included the komatiitic-picritic volcanism and mafic-ultramafic intrusions (e.g., Kevitsa) of a common plumbing system (Hanski et al., 2001; Virtanen et al., 2024) at ca. 2.06 to 2.05 Ga. Coeval volcanic and igneous activity is also recorded in the Pechenga and Imandra-Varzuga belts, where alkaline picrites and basalts of the Kuetsjärvi, Kolosjoki, and Umba Formations yield ages of ca. 2.06 to 2.05 Ga (Melezhik et al., 2007; Martin et al., 2013). This distinct event of ca. 2.06 to 2.05 Ga comprising several intrusions and volcanic sequences related to the final breakup of supercontinent Kenorland in northern Fennoscandia fulfills the criteria of an LIP. We suggest the name “Sápmi LIP” to this event, referring to the name of the region used by the Indigeneous Sámi people.

Kaarrekumpu sample 04MOS7001 has two grains (13.1 and 19.1) that are clearly inherited from an Archaean source, yielding discordant ²⁰⁷Pb/²⁰⁶Pb ages of ca. 2758 Ma (grain 13, 7% discordant) and ca. 3071 Ma (grain 19, 40% discordant) (Fig. 8). They are within the ballpark range of ages of orthogneisses found across the Karelian craton (ca. 3.5–2.7 Ga; e.g., Slabunov et al., 2006; Hölttä et al., 2012), and the presence of Archean inherited zircons suggests that the Kaarrekumpu magma assimilated Archean basement material, directly or indirectly, evidence of which is also provided by negative Nb and Ta anomalies in primitive mantle-normalized plots of the Kaarrekumpu ultramafic host rocks (Harju, 2018). The ca. 2.05 Ga intrusions typically digested Sodankylä and Savukoski Group sedimentary and volcanic rocks upon ascent and emplacement (e.g., Mutanen, 1997; Brownscombe, 2016; Luolavirta et al., 2018a, b), and therefore we assume that the Kaarrekumpu Archean zircons are inherited detrital zircons incorporated by assimilation of these Paleoproterozoic sediments, which further reinforces the sulfur-source scenario considered below.

Some of the magmatic zircons from Sakatti yield discordant ages that fall between ca. 2050 and 1850 Ma (Figs. 6–8). Populations of titanites from Kaarrekumpu give concordia ages of ca. 1.92 Ga (Fig. 10) and a combined regression age of 1937.6 ± 54.2 Ma for three samples (Fig. 9). We interpret that those zircons and titanites with lowest common lead contents (App. 2) record an inversion event that has been widely documented in the Central Lapland greenstone belt at ca. 1.94 to 1.89 Ga. This event has been ascribed to the collision of the Karelian, Kola, and Norbotten cratons during the broadly coeval Lapland-Kola and Svecofennian orogenies (Daly et al., 2006; Lahtinen and Huhma, 2019), the south-southwestward thrusting of the Lapland Granitoid Complex and resultant obduction of the Kittilä Group allochthon (Lahtinen et al., 2005, 2018). The zircons yielding these ages have strongly modified edges and complex, fine-scale structures (e.g., grains 6 and 18; Fig. 7), indicating modification by fluid-driven activity. The ca. 1.9 Ga ages were previously documented in proximity to the internal zones of the orogens in the northeastern and southwestern parts of the Central Lapland greenstone belt (Lahtinen et al., 2018; Lahtinen and Huhma, 2019), which underwent higher-grade metamorphism, in comparison to the estimated metamorphic conditions in the Sodankylä area that range from greenschist facies (ca. 0.5–8 kbar, 300°–400°C) to middle amphibolite facies (ca. 3.5–8.5 kbar, 250°–500°C) (Tyrväinen, 1983; Hölttä et al., 2007; Hölttä and Heilimo, 2017). Our ages demonstrate variable intensity of postmagmatic modification both at a regional scale and at a very local scale in the Central Lapland greenstone belt, which may have affected Re-Os systematics of the Sakatti sulfide samples that yielded a combined isochron age of 1928 ± 16 Ma (Moilanen et al., 2021).

A late Paleoproterozoic (Statherian) ca. 1.78 Ga age is defined by a population of 10 concordant or semiconcordant zircons from Sakatti (Fig. 7) that give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1779.2 ± 3.6 Ma. Similar ages are recorded in Sakatti and Kaarrekumpu titanites, which give a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1755.6 ± 11.6 Ma (Fig. 7) and a regression age of 1781.6 ± 3.6 Ma (Fig. 10), respectively. One grain (7, Fig. 7) records both ca. 2050 and ca. 1790 Ma ages, strongly suggesting that the ca. 1.78 Ga event is metamorphic and caused variable Pb loss. Many zircons yielding this age are characterized by relatively low Th/U ratios (<0.2) and have very small rims or overgrowths in CL imaging (e.g., 7 and 8 in Fig. 7) but some are also euhedral and unaltered (grain 5, Fig. 7) and therefore probably grew during this stage. This ca. 1.78 Ga age group is characteristic for the late stages of tectonic evolution in northern Fennoscandia, including orogenic collapse and locally high temperatures during the aftermath of the Svecofennian orogeny such as recorded in the postcollisional A-type Nattanen Granite Suite (Heilimo et al., 2009), alkaline plutonism of the Lohiniva appinite suite and some of the granitoids in Central Lapland Granitoid Complex (Rastas et al., 2001; Lahtinen et al., 2018). Similar U-Pb ages are found in numerous Au(-Cu) and iron oxide copper-gold (IOCG) deposits in the Central Lapland greenstone belt, which are thought to have formed by late-stage hydrothermal events (Mänttäri, 1995; Niiranen et al., 2007; Molnár et al., 2018). The Orosirian-Statherian boundary age group (1790–1780 Ma) is recorded globally in a number of plume-generated LIPs (see compilation by Ernst et al., 2021), evidencing the close association of Baltica to several other cratons during the assembly of the supercontinent Columbia (Nuna) (e.g., Shumlyanskyy et al., 2021; Lahtinen et al., 2023).

A third population of concordant or semiconcordant zircons (Fig. 8) shows late Mesoproterozoic to Neoproterozoic ages, between 1.1 and 0.48 Ga. These ages are thought to record Neoproterozoic events related to the assembly of Rodinia during the Neoproterozoic and perhaps an Ordovician fingerprint of the Caledonian orogeny. Similar U-Pb zircon and baddeleyite ages were recorded by Hanski et al. (2010) from ca. 2.21 Ga mafic differentiated sills in the Central Lapland greenstone belt, which have been attributed to circulating low-temperature hydrothermal fluids during rapid burial following the Caledonian orogeny (Larson and Tullborg, 1998).

Rutile can be a good mineral for establishing the age of mineralization or low-temperature metamorphism, with a closure temperature range between 400° and 600°C (Mezger et al., 1989; Cherniak, 2000). A late Neoproterozoic (early Ediacaran) concordia age of 600 ± 20 Ma is defined by three of four analyzed rutile grains (Fig. 5), which record a very late thermal event related to the final stages in the breakup of supercontinent Rodinia and the formation of the Iapetus Ocean magmatic province. A similar age has been recorded by kimberlites in eastern Finland (O’Brien, 2015) as well as by dolerite dike swarms and alkaline magmatism related to intraplate rifting in Norway and Sweden (Torsvik et al., 1996; Nystuen et al., 2008) that are part of the Baltoscandian-Egersund LIP. In the Central Lapland greenstone belt this age group may also be recorded in the U-Pb lower intercept regression ages in zircons of ca. 2.2 gabbro-wehrlite association intrusions (Hanski et al., 2010), but its relationship to any later sulfide ore remobilization/upgrade process cannot be established with the available data.

The 2056 to 2053 Ma ages obtained for the world-class Sakatti deposit and Kaarrekumpu occurrence are key markers for a short-lived and significant Cu-Ni-PGE–mineralizing event in northern Fennoscandia. Other localities showing similar ages in the Central Lapland greenstone belt include the Cu-Ni-PGE–mineralized Kevitsa and Puijärvi intrusions and the Lomalampi komatiites, which have been discussed by other authors (Peltonen et al., 2014; Makkonen et al., 2017; Moilanen et al., 2019). Mineralized ca. 2.05 Ga mafic-ultramafic rocks are also encountered farther north and west in the Karasjok and Pulju greenstone belts, respectively, demonstrating a wide-area extent of this fertile magmatic event. In the Karasjok greenstone belt Cu-Ni-PGE–mineralized mafic intrusions such as Gállojávri (2051 ± 4 Ma, Orvik et al., 2022) and Porsvann (2061 ± 4 Ma, Hansen et al., 2023) have yielded similar ages. Farther west in the Pulju greenstone belt, the Hotinvaara Ni-Cu-PGE deposit (Papunen, 1998; Konnunaho et al., 2015) as well as the Hietakero Co-Cu-Ni deposit in Lätäseno belt (Konnunaho et al., 2018) are associated with komatiites that have been stratigraphically correlated to the Savukoski Group. The timing of such a distinct metallogenic event in northern Fennoscandia evidently assures the exploration potential of these ca. 2.06 to 2.05 Ga units in the Karelia, Kola, and Norrbotten cratons but also the potential for more widely for synchronous magmatism in contemporaneously proximal crustal blocks. With reference to coeval deposits in the ca. 2.06 Ga Bushveld LIP in South Africa, high-precision U-Pb dating of the Rustenburg Layered Suite of the Bushveld Complex (Mungall et al., 2016) and the Nkomati Ni-Cu-PGE deposit in the Uitkomst intrusion in the Transvaal Supergroup (Maier et al., 2018) was the key to establish complex-scale geochronology and a petrogenetic link between the ca. 2055 Ma Bushveld Complex and its satellite intrusions, which, coupled with geochemical modeling, creates a strong exploration tool in the area.

Combining U-Pb (or other high-precision age dating method) and Sm-Nd systematics may aid in correlating genetically similar intrusions, dikes, and extrusions, as well as in discriminating between potentially barren and mineralized intrusions. The Cu-Ni-PGE systems at ca. 2.05 Ga in northern Fennoscandia show disturbed Sm-Nd systems and considerable variation in initial ε_Nd values: most mineralized intrusions record coeval depleted mantle values (ε_Nd ~2–5) and negative crust-affected values (ε_Nd ~–5 to –3), indicating significant crustal contamination by assimilation (cf. Mutanen, 1997; Hanski and Huhma, 2005; Peltonen et al., 2014; Huhma et al., 2018; Orvik et al., 2022).

The introduction of external sulfur and timing thereof are important in the formation of economic Ni-Cu-PGE ores (e.g., Ripley and Li, 2013; Barnes et al., 2016), as they can trigger early sulfide saturation in feeder systems of the ca. 2050 Ma komatiitic-picritic volcanic rocks. Ingestion of sulfur-bearing units in mafic-ultramafic magmas is recorded in all known mineralized deposits (e.g., Sakatti, Kevitsa, Puijärvi, Kaarrekumpu, Hotinvaara, Gállojávri). Considering the lithostratigraphic constraints for the area, plausible upper crustal sulfur sources include the graphitic-sulfidic schists (Lahti et al., 2012) and anhydrite-bearing metaevaporites (Haverinen, 2020), which predate the timing of Savukoski Group komatiitic-picritic magmatism and coeval mafic-ultramafic intrusions. In this regard, the deposits of the Central Lapland greenstone belt show similarities with, e.g., the Norilsk-Talnakh deposits in Russia, in terms of widespread extrusive and intrusive mafic-ultramafic magmatism, presence of both mineralized and barren feeder systems, and presence of evaporites as crustal sulfur sources (e.g., Arndt et al., 2003; Malitch et al., 2014; Iacono-Marziano et al., 2017; Barnes et al., 2023).

We also note that the ca. 2.15 Ga mafic intrusions can be fertile for Ni-Cu-Co-PGE deposits, as they are coeval/slightly younger than the plausible sulfur sources, and contact-type Cu-Ni-Co mineralization has been found in the Rantavaara and Tanhua intrusions (Mutanen, 2005; Konnunaho et al., 2022). Mineralization in the ca. 2.15 Ga intrusions, however, appears localized, and the intrusions show coherent ε_Nd values of ~2 to 5, indicating limited crustal assimilation (Huhma et al., 2018).

In the light of Paleoproterozoic metamorphic ages (ca. 1.92 Ga, ca. 1.78 Ga) recorded in both the Sakatti and Kaarrekumpu deposits, postmagmatic modification and remobilization of some of the sulfide ores is plausible and may be dependent on variable intensity of alteration with localization of fluids along structures controlling the diffusivity of elements in some of the titanites and zircons. For example, Sakatti is directly associated with the roughly E-W–striking Viiankiaapa thrust fault (“the Basal thrust”) that was activated several times during the 1.96 to 1.76 Ga time span with mineralogical evidence of circulating carbonaceous saline fluids (Vuorisalo, 2021). Estimations of pressure-temperature (P-T) conditions at Sakatti vary but may have locally achieved 6 kbar and 650°C, which could have partly reset the titanite system.

A Cu-rich intermediate solid solution (iss) sulfide melt can be liquid at temperatures below <850°C (Barnes et al., 2016; Saumur and Cruden, 2017), and sulfide-telluride assemblages can record equilibration temperatures as low as 200° to 250°C (Helmy et al., 2007). Given that temperature range, the Cu-sulfide-PGE telluride stockwork veins may have formed due to remobilization of the Cu-rich portion of the ore from the cumulates to the surrounding lavas. A coexistent fluid phase can significantly facilitate remobilization, as for instance, PGEs are mobile under varying hydrothermal conditions (<500°C), depending on $fO2,$ pH, and chlorine and sulfur contents of the fluid (Sassani and Shock, 1998; Hanley, 2005; Barnes and Liu, 2012). Copper is commonly more mobile than PGEs in hydrothermal and hydrothermal-magmatic systems, whereas Ni and Co would have been less mobile. Patten et al. (2023) have concluded that the metamorphic devolatilization of the Savukoski Group rocks at ca. 1.9 to 1.8 Ga could have mobilized metals, such as S, Cu, and Te. The Cu-rich nature of the ore at Sakatti, variable Pt/Pd ratios in ore types, and mere association of PGEs with tellurides suggest that in addition to comagmatic sulfide melt fractionation, later remobilization of metals may have been important in the genesis of some of the Sakatti ore types. Noteworthy is that the numerous ca. 1.8 Ga orogenic Au(-Cu) and IOCG deposits in Central Lapland greenstone belt were formed by oxidizing and saline metalliferous fluids that migrated through major lineaments, as shown by intense sodic and calcic alteration of the host and wall rocks of these deposits. Fluid temperature estimates in these deposits range from 300° to 550°C (Niiranen, 2005; Molnár et al., 2018), capable of redistributing Cu and PGEs. The Sakatti host and wall rocks share alteration assemblages (carbonate-albite-scapolite-biotite; P. Eilu and T. Törmänen, unpub. report, 2018) similar to those of the ca. 1.8 Ga hydrothermal deposits. Unpublished Pb-Pb model ages of chalcopyrite and galena in Sakatti (Y. Lahaye and H. O’Brien, unpub. report, 2018) record events at ca. 1.8 Ga, therefore raising additional potential for hydrothermally upgraded Cu-Ni-PGE deposits. The Sakatti δ³⁴S define a narrow range with most data falling between 2 to 4‰ in all sulfides (chalcopyrite [ccp], pyrrhotite, pyrite, pentlandite); the values in chalcopyrite, however, also show a distinct group of negative values δ³⁴S_ccp ranging from –7 to –2.5‰ (cf. Brownscombe, 2016; Y. Lahaye and H. O’Brien, unpub. report, 2018), indicating modification by oxidizing crustal fluids. Similar broad variations in δ³⁴S (up to ±10‰) ranging from positive values of crustal origin to negative values of fluid origin are also observed in Au(-Cu) and IOCG deposits (cf. Niiranen et al., 2007; Molnár et al., 2019; Vasilopoulos et al., 2024) but not in the conventional orthomagmatic ca. 2.05 Ga Cu-Ni-PGE deposits in the Central Lapland greenstone belt (cf. Grinenko et al., 2003; Konnunaho et al., 2018; Luolavirta et al., 2018c). Therefore, the nature, extent, and significance of postmagmatic remobilization of metals by hydrothermal fluids at Sakatti Cu-Ni-PGE deposit requires further study.

The Sakatti and Kaarrekumpu Cu-Ni-PGE sulfide deposits share a common magmatic age of ca. 2056 to 2053 Ma and belong to the ca. 2.05 Ga age group of Cu-Ni-PGE–endowed mafic-ultramafic rocks in the Central Lapland greenstone belt. The Cu-Ni-PGE–mineralized Kevitsa intrusion shares a coeval age of 2058 ± 4 Ma (Mutanen and Huhma, 2001). The ages obtained from Sakatti lend credence to a minimum age of extrusion of the Savukoski Group komatiites to the same age bracket. We conclude that northern Fennoscandia bears a well-defined short-lived magmatic event at ca. 2060 to 2050 Ma, which can be regarded as an LIP, for which we propose the name of Sápmi LIP. This distinct event emphasizes targeting Cu-Ni-PGE exploration to the ca. 2060 to 2050 Ma magmatism in Fennoscandia but in the adjacent contemporaneous cratonic blocks. Younger ages from both deposits indicate that the region was affected by a thermal event at ca. 1920 and ca. 1790 Ma, which can be related to major tectonic events and indicates that the sulfide ores in the deposits may have seen remobilization and metal upgrading. Regarding implications for exploration, high-precision age dating of undefined mafic-ultramafic bodies coupled with Sm-Nd systematics and geochemical modeling may help identify mineralized bodies within the magmatic plumbing system at ca. 2060 to 2050 Ma and even more widely at ca. 2150 to 2050 Ma. Metal upgrade of preexisting massive Cu-Ni-PGE ores is thought to have occurred in proximity to major structures that control percolation of saline fluids during postemplacement collisional and exhumation events between ca. 1920 and ca. 1790 Ma.

The authors thank Jukka Jokela (previously Anglo American) and Dave Braxton for approval of this research project. The analytical work was financed by Anglo American (AA Sakatti Mining Oy). H. Höytiä thanks the K.H. Renlund Foundation for a personal research grant. Petri Peltonen is thanked for comments on the manuscript. Richard Ernst, an anonymous reviewer, and Editor Larry Meinert are thanked for insightful and constructive reviews that improved the manuscript and brought it to a global context. Timo Saarimäki (Geological Survey of Finland) is thanked for thin-section preparation.

Henri Höytiä is an exploration geologist with a strong combined industry and academia background. He obtained his M.Sc. degree from the University of Helsinki in 2019, majoring in petrology and economic geology. His work history includes positions in major and junior mining and exploration companies as well as academic projects with the Geological Survey of Finland and the University of Helsinki. Currently, he is working on metallogeny and exploration of Ni-Cu-PGE deposits in northern Fennoscandia for his Ph.D. project, and he works for Anglo American in the Sakatti Cu-Ni-PGE mine development project and in regional exploration as a geologist.