Provenance, Age, and Tectonic Settings of Rock Complexes (Transangarian Yenisey Ridge, East Siberia): Geochemical and Geochronological Evidence

Abstract

The tectonic evolution of the Siberian Cratonic margins offers important clues for global paleogeographic reconstructions, particularly with regard to the complex geological history of Central Asia and Precambrian supercontinents Columbia/Nuna and Rodinia and its subsequent breakup with the opening of the Paleo-Asian Ocean. Here, we present an overview of geochemical, petrological, and geochronological data from a suite of various rocks to clarify the age, tectonic settings, and nature of their protolith, with an emphasis on understanding the tectonic history of the Yenisey Ridge fold-and-thrust belt at the western margin of the Siberian Craton. These pre-Grenville, Grenville, and post-Grenville episodes of regional crustal evolution are correlated with the synchronous successions and similar style of rocks along the Arctic margin of Nuna-Columbia and Rodinia and support the possible spatial proximity of Siberia and North Atlantic cratons (Laurentia and Baltica) over a long period ~1.4–0.55 Ga.

1. Introduction

The Yenisey Ridge fold-and-thrust belt is located along the western margin of the Siberian Craton, forming part of the Central Asian Orogenic Belt (CAOB), and is hence a key region to gain insights into the Precambrian tectonic evolution of the Siberian Craton and the crustal growth of one of the world’s largest Phanerozoic accretionary orogenic belts [1,2]. It is one of the few regions of Siberia where Paleoproterozoic, Mesoproterozoic, and Neoproterozoic magmatic and metamorphic rocks are closely associated [3,4]. Understanding the tectonic evolution of the Siberian Cratonic margins is also important for global paleogeographic reconstructions and deciphering the complex tectonic structure of Central Asia [5,6]. The geological and tectonic history of the Cratonic margins of Siberia is also crucial for tracing the history of the Precambrian supercontinents Columbia/Nuna and Rodinia and its subsequent breakup with the opening of the Paleo-Asian Ocean [7,8,9,10,11,12,13,14,15,16,17].

However, there are ideas that the time period of the Earth’s geological evolution between 1.8 and 0.75 Ga ago spanning the middle Proterozoic eon, known in the literature as the boring billion, is characterized by a lesser extent of magmatic and tectonic activity induced by changes in mantle convection and plume activity, which contrasts with the dramatic changes in preceding and succeeding eras [18,19]. On the basis of this assumption, several recent studies suggested the absence of the Mesoproterozoic and, in particular, Grenvillian collisional events within the Yenisey Ridge, and this gave rise to contradictory interpretations of the major tectonic framework of this region.

In recent years, there is increasing evidence that the change from Nuna to Rodinia was one transition in which Siberia, Laurentia, and Baltica formed a stable core that assembled during the construction of Nuna by 1.8–1.7 Ga and continued, with little change in configuration, until the Rodinia breakup at 0.75 Ga [20]. This contrasts with the significant changes in continental configurations that occurred in the initial assembly of Nuna or in the transition from the Rodinia breakup to the Pangea assembly. Extension-related magmatism at ca. 1.4–1.25 Ga is widespread around parts of Laurentia, Baltica, and Siberia and is ascribed to Nuna breakup [20]. However, the general paucity of passive margins through the Mesoproterozoic and the absence of any evidence for 1.1–1.0 Ga orogenesis in blocks such as Siberia [21] argue for no significant breakup and reassembly of these continental fragments in the transition from Nuna to Rodinia [22]. The Siberian craton could have been an important part of these supercontinents. Studies of the geological history of the Meso-Neoproterozoic continental margins of Siberia are crucial for solving the problem of the possible incorporation of the Siberian craton into the Columbia/Nuna and Rodinia supercontinents, an issue that has been widely debated for more than 20 years [23].

The aim of this paper in this context is to re-examine the igneous and metamorphic complexes and sedimentary suites of the western margin of Siberian Craton, in view of recently published data, and to re-assess the role of Siberia in the Nuna and Rodinia supercontinents.

One of the most effective methods for determining the nature and origin of protoliths and tectonic settings is the analysis of rare-earth element (REE) spectra and indicator ratios between other trace elements [24]. Recent advances in understanding the distribution and geochemistry of REEs have primarily contributed to a better understanding of the evolution of igneous and sedimentary rocks. The major, trace, and rare-earth element behavior in rocks is believed by many geologists to be the best indicators for such reconstructions because they are not significantly redistributed during weathering, transport, diagenesis, and metamorphism [25]. As a result, their abundance in altered rocks generally reflects the composition of the source rocks [26]. Thus, metamorphic rocks derived from different provenances preserve a geochemical record of the composition and evolution of the upper continental crust and, consequently, allow us to judge their protolith origin [27].

The total REE contents in metamorphic rocks have received comparatively little attention. For example, there is virtually no data on the distribution of REEs in Fe- and Al-rich pelitic schists, the metamorphism of which leads to the stability of rare mineral assemblages comprising chloritoid, Fe-cordierite, and other minerals, such as chloritoid + biotite, chloritoid + biotite + andalusite, and cordierite + garnet + muscovite [28,29]. The unusual bulk rock composition is interpreted to have been a distinction between typical metapelite and Fe- and Al-rich metapelite topologies in P–T diagrams [30,31]. In the literature, such a specific Fe- and Al-rich bulk rock composition is attributed to lateritic weathering processes [32]. However, full-profile lateritic weathering crusts are often absent in Precambrian sections [33], thus arousing interest in the origin of these rocks.

On the other hand, this study area has recently become a focus of growing interest because it contains large resources of gold, manganese, lead, zinc, niobium, antimony, iron, etc. The formation of the most promising deposits and occurrences of the Al-rich rocks in the Transangarian segment of the Yenisey Ridge are genetically related to the And–Sil and Ky–Sil types of metamorphism. In the 1979–1980s, the PGO “Krasnoyarskgeologiya” carried out prospecting and mining works in 20 areas within the central part of the Yenisey Ridge. As a result, the prospects of the Teya and Mayakon deposits were confirmed with calculated reserves of ~90 and ~75 million tons of aluminum oxide, respectively [34]. In recent years, andalusite, kyanite, and sillimanite, which are constituents of high-alumina schists, have become increasingly important in practical terms for the production of aluminum oxide, silumin, and aluminum for the expansion of the raw material base in industry [35,36].

The geochemical features, petrogenesis, and precise dating of the magmatic, metamorphic, and sedimentary events are important in this content. Here, we present an overview of the geochemical, petrological, and isotope geochronological data from a suite of various rocks to clarify tectonic settings and the nature of their protolith, with an emphasis on understanding the tectonic history of this region.

2. Geological Background

2.1. Overview of the Regional Setting and Main Structural Elements of the Yenisey Ridge

The Yenisey Ridge is a complex Precambrian fold-and-thrust belt along the western margin of the Siberian Craton, striking 700 km northwestwards along the Yenisey River (Figure 1a). Geophysical data indicate transpression and vertical crustal thickening in this belt [37]. The regional structural framework of the Yenisey Ridge has traditionally been interpreted as a system of NW–SE-oriented crustal tectonic units characterized by large-scale isoclinal folds dipping to NE and accompanied by regionally developed schistosity. The thrusts follow the general structural trend of the Yenisey Ridge.

According to structural and seismic data [38], the thrust systems are generally W-vergent dipping 40–60° to NE. The NW-trending structures of the Yenisey Ridge are divided into two segments, separated by the ENE-trending strike-slip Nizhneangara Fault [39]. Two allochthonous units have been recognised south of the Nizhneangara Fault: the Paleoproterozoic granulite-amphibolite facies Angara-Kan block and the Neoproterozoic, mainly island arc-type, Predivinsk terrane, which lies along the western end of the Angara-Kan block [40]. North of the Nizhneangara Fault, the Transangarian Yenisey Ridge is mainly composed of west-verging thrust sheets that contain Meso-Neoproterozoic rocks, including the East and Central continental margin blocks, and the Isakovka island arc terrane [41]. All the tectonic blocks represent crustal segments ranging in length from 300 to 500 km, and with a width of 50–80 km. These are separated by the largest known thrust faults in the region.

This study focuses on the northern segments of the Yenisey Ridge, partly within YRSZ (Figure 1b). This shear zone forms an extension of the Baikal-Yenisey fault within the right bank of the Yenisey River and can be traced for about 200 km along the western flank of the Yenisey Ridge. This largest deformation structure, which occurs as a prominent metamorphic lineament, divides the region into Craton blocks and island arc terranes. It is well traced along the Yenisey Ridge and within the Turukhansk-Norilsk tectonic zone by the disappearance of several seismic discontinuities and deepens with the fault plane dipping to the east [42,43]. Within the Yenisey Ridge, YRSZ is clearly distinguished by gravity anomalies and is also prominently imaged on seismic profiles. This YRSC defines a system of contiguous subparallel faults or shears, with uplift and thrust kinematics, and focus shear deformations, characterized by near-fault cataclasis, mélange, and dynamic metamorphism (Figure 2). It is several hundred kilometers long, and the width of the zones of stress-metamorphism runs to a few tens of kilometers. These zones serve as the seams dividing the tectonic blocks of the region and are the areas of active interaction of these blocks [44].

Blastomylonites are developed over these rocks adjacent to the fault zones [45]. Overall, the shear zones mainly display dextral and, to a lesser extent, sinistral kinematics, with blastomylonites showing elements of laminar flow. Boundaries between the rock complexes are marked by the detachment and bedding surfaces [46,47].

The regional faults (Yenisey, Tatarka-Ishimba, and others) are often accompanied by a series of higher-order splay faults, the activity of which resulted in overthrusting of smaller blocks [48,49]. These processes brought about regionally heterogeneous metamorphic pressure conditions, leading to two distinct facies series: andalusite-sillimanite (low-pressure) and kyanite-sillimanite (medium-pressure) [50,51,52].

The mechanism of origin and timing of regional deep faults in the region are still unclear. However, interpretation of the geochronological data on the heterogeneous monazite populations in tectonites subjected to recrystallization during the consecutive deformational events indicates a repeated reactivation of YRSZ in the region at 1.54–0.6 Ga [53,54,55].

For a detailed review of the geochronology, tectonic setting, and geodynamic history of the various complexes in this region, including a summary of the chronological sequence of the main stages and events in the Yenisey Ridge, we refer our readers to a review paper by Likhanov et al. [56].

The present study area includes the paleocontinental and paleooceanic blocks. The Garevka Complex (GC) comprises the oldest rocks in the Transangarian Yenisey Ridge and is located in the western part of the Central block, within the YRSZ. In the eastern part of the Central block, the GC is succeeded by the Teya complex (TC), with the Garevka Group overlain by the lower Proterozoic Teya Group. Tectonically, TC is located within the axial part of the Central block, which is bounded by a set of extensive faults along the Tatarka-Ishimba suture zone. All the rocks display prominent blastomylonitic fabrics produced by multistage brittle and ductile deformation along faults [12,39].

2.2. Geology of the Studied Areas, Petrography, Textural Observations, and P–T Conditions of Metamorphism

2.2.1. Metabasite Tectonites

In our study area along the north-western domain of the Transangarian Yenisey Ridge, three major lithological units define the YRSZ structure, and from east to west these are: gneiss-amphibolite suite, metabasite-ultramafic suite, and volcanic-plutonic suite (Figure 3). The continental (gneiss-amphibolite) block (zone 1) consists of the Garevka Complex. The Garevka Complex is essentially composed of metamorphosed terrigenous- and carbonate-terrigenous rocks of the Nemtikha and Malaya Garevka Groups [42]. The Nemtikha Group is largely plagiogneiss granites and metapelite gneisses. Porphyroblastic and blastomylonitic deformed granitogneisses, A-granites, rapakivi-type granites, amphibolites, metapelitic schists, anatectic migmatites, and subordinate amphibolite and metaterrigenous-carbonate rocks (quartzite, calciphyre, marble) are the dominant rock types in the Malaya Garevka Group [45]. The latter two blocks (zone 2 and 3) are constituents of the Isakovka terrane, carrying ophiolites composed of mélange sheets and lenses of amphibolitized tholeiitic metabasalt and metabasite and metaultramafics (antigorite, dunite, and harzburgite with minor antigoritized pyroxenite), which are related to the Surnikha Complex. The volcanic-plutonic block consists of metadacite-andesite-basalt association metamorphosed under greenschist facies conditions and related to the Kiselikha Group [42].

We studied in detail the five types of rock associations from within the suture zone. Samples of metabasite tectonites were collected along the right bank of the Yenisey river (locations include Osinovsk rapids (samples 8–14), Borisikha (sample 1), Verhnyaya Surnikha (sample 7), Proklyataya (samples 2–5), and Garevka Rivers (sample 6)) from the northern segment of the YRSZ (Figure 3). These samples differ both in their nature and intensity of deformations and in the heterogeneity of their compositions. In the northernmost part of the study area, in the region of the Osinovsk rapids, the typical units of the mélange are boudinaged relic steep (85–90°) blocks and sheets of Paleoproterozoic Nemtikha Group metapelite gneisses characterized by an assemblage of Grt + Bt + Ms + Pl + Qz ± Kfs ± Chl together with boudins of amphibolite composed of Grt + Amp + Pl + Ep + Chl + Ph + Spn + Rt. The mineral abbreviations are from [57]. The lenticular lamellar structure in the mélange sequence characterized by the alternation of blastomylonite and garnet amphibolites with an assemblage of Grt + Amp + Pl + Ph + Pg + Ep + Spn + Cb + Chl + Rt in association with boudinaged and highly serpentinized dunite and subordinate harzburgite and pyroxenite are also identified in the V. Surnikha river and upstream of its estuary in the right bank of the Yenisey river area. The Borisikha area rocks are represented by banded amphibolites with coarse-crystalline relic lenses of Pl-Amp compositions and a gabbroid appearance, dated as 680 Ma [58]. Fine-grained amphibolites have locally developed isometric and deformed garnet grains, which form local thin (<1–3 cm) bands and lenses in association with the Amp + Ab + Chl + Ep + Spn assemblage.

The rocks described above constitute part of the subduction-accretion complex, where they appear as tectonic layers, lenses, or blocks within serpentinite mélange. They display shear flow structures, which are widely developed both on meso- and micro-levels. These include kinematic indicators of shear or transport during a deformation event [59]: linear deformational features, ordered structures of plastic flow, stretching and rupture of en echelon folds, kink banding in micas, “pressure shadows” (strain shadows) of recrystallized quartz, S-shaped and deformed garnet grains having “snowball” textures, fractures in mineral grains showing patchy banding, deformation of lamellar twinning in plagioclase, parallel distribution of fine-grained lenticular mineral aggregates, and schistosity, cataclasis, and boudinage [60].

The typical mineral assemblage of the metabasite tectonites is hornblende + garnet + epidote + chlorite + phengite + zoisite + plagioclase + quartz + titanite + rutile + ilmenite ± carbonates. The microtextural features and the modal proportion of metabasite minerals indicate two stages of their development, culminating in highly deformed blastomylonite minerals (Figure 4). Grains of garnet and titanite often contain relic inclusions of glaucophane, albite, phengite, epidote, and chlorite, indicating the involvement of these minerals in the early pre-peak association (Figure 4a,c). The inclusions show no significant preferred orientation; thus, evidence for an early foliation is not preserved in the metabasites. The discovery of relic minerals of glaucophane from Yenisey suture zone tectonites in our study suggest a “Cordilleran-type” convergent boundary in the western Siberian Craton, related to Oceanic crust subduction under the continental margin.

The tectonites define the general structure of the NW-trending suture zone, with a thickness of about 15–20 km between the continental and volcanic-plutonic blocks (Figure 2a). The tectonic mélange comprising mildly to highly metamorphosed rock blocks of different ages, scales, and compositions within the serpentinite matrix indicates repeated reactivation of YRSZ during Neoproterozoic at 800–600 Ma [53,61]. The formation of the most intensely deformed tectonites is inferred to be related with the final accretion of the Isakovka terrain to the west margin of the Siberian Craton at about 630–600 Ma [62]. These results are consistent with the ages of the NW Sayan metamorphic rocks, which are included together with the Yenisey Ridge in the Sayan-Yenisey accretionary belt [39].

Results from field-based observations, geochronology, microtexture analysis, and mineral chemistry, coupled with P–T path calculations derived from conventional geothermobarometry and pseudosections, provide evidence for the polymetamorphic history of the metapelite and metabasite tectonites in the Yenisey Ridge, which is recorded by the overprinting of early mineral assemblages during subsequent events as a result of different geodynamic processes. The P–T calculations suggest two superimposed metamorphic events. During the first stage, glaucophane schists formed at around 640–620 Ma at P–T conditions of 8–10 kbar and 400–450 °C. In the second stage, the rocks experienced dynamic metamorphism (c. 600 Ma) at 11–15 kbar/550–640 °C. The formation of the bulk of blastomylonites in the northern segment of the shear zone in the Northern Yenisey Ridge occurred with a 1.5–3 kbar increase in the pressure along with an insignificant increase in the temperature and low metamorphic gradient dT/dZ < 10 °C/km compared to the background values of earlier regional metamorphism. The maximum excess values of the thermodynamic parameters of metamorphism were established in metabasite tectonites of the suture zone with relict glaucophane schist associations that underwent metamorphism, with a simultaneous significant increase in the pressure by 3–5 kbar and temperature by 180–240 °C with a higher dT/dZ gradient of 15–20 °C/km. The intense deformation of the rocks led to a virtually complete recrystallization of the substrate and the formation of new high-pressure mineral parageneses. Such excess P–T values could be due to progressive metamorphism, complicated by local heating of rocks during viscous deformations and excess oriented tectonic pressure over lithostatic in ductile shear zones.

2.2.2. Granites

The granite samples used in this study were collected from the GC within the upper reaches of the Chapa River and its right-hand tributary: Koloromo River (1) (Figure 1), in the lower course of the Tis and Garevka rivers and from outcrops on the right bank of the Yenisey River (2), including granites from a small rocky island in the Yenisey River in front of the Vyatka mouth and Ostrovok island (3), from Chernorechensky and Osinovka massif (4–5), and in the lower-middle course of the Garevka River (6–8) (Figure 5).

1. The block of the plagiogneiss granites from the Nemtikha Group of about 10 km length and a width of 4 to 5 km is exposed in the study area on the right bank of the Koloromo River, 7 km upstream of its mouth. In hand specimen, the plagiogneiss granites from the Nemtikha Group are medium-grained and inequigranular, light-grey, with well-developed foliation and pseudo-porphyritic texture (Figure 6a). Under the microscope, all examined samples exhibit a blastocataclastic texture. Porphyroclasts are generally cataclastic grains of zoned igneous plagioclase, lenticular in shape and up to 3–4 mm in size. The lepidogranoblastic matrix consists of a medium-grained foliated aggregate (1–2 mm) of granoblastic-textured quartz and plagioclase, and muscovite flakes oriented along the main foliation. The distinctive feature of all the rocks is the presence of thin ribbon-like quartz-plagioclase bands (0.5–1.3 mm) aligned along the foliation direction, which are draped over the plagioclase porphyroblasts and contain small grains of secondary epidote and fine scaly chlorite and muscovite. Relics of a primary igneous texture are found in quartz-feldspar lenses. The matrix, composed of quartz + plagioclase + epidote + muscovite + chlorite, indicates lower greenschist-facies metamorphism [63].

2. Rocks from steeply dipping, sheeted, coarse-grained intrusive bodies of granites and gabbros crosscut the migmatized and intensely deformed gneisses and schists of GC (Figure 5). They are mapped within narrow near-fault zones as small intrusions and a dyke swarm composed of metamorphosed granites and gabbros (Figure 6c), which are confined to faults within YRSZ. The dykes predominantly trend northwest (340–350°), dip steeply to subvertically (80–90°), and vary from several cm to 5–10 m in thickness [64].

The felsic dykes are mostly composed of grey and pink coarse-grained porphyroblastic granites showing evidence of internal differentiation, often with a foliated texture formed by augen orthoclase granite-gneisses containing orthoclase porphyroblasts 5–7 mm to 2 cm in size. Orthoclase ovoids exhibit exsolution (perthitic plagioclase and graphic quartz) and myrmekitic textures, indicating their igneous origin. The matrix is composed of quartz, albite, microcline, and biotite [9].

3. The separate granite massif in the tectonic melange is represented by a rock ridge on the Yenisey River near Ostrovok Island, opposite to the Vyatka River mouth (Figure 6d). Granites are represented by pink-colored coarse-granular varieties with weak foliation defined by biotite, which surround K-feldspar porphyroblasts (latticed microcline) with a size of up to 1.5–2.0 cm and lens-like quartz segregations of the same size (relics of large cataclazed quartz grains) together with granoblastic quartz–plagioclase–microcline aggregate. Accessory minerals are represented by well-formed crystals of titanite with a size of up to 3 mm and small crystals of zircon and apatite [65].

4. The Chernorechensky massif (CM) is located on the right side of the Yenisey River, in the Razgromnaya River basin (Figure 1), and is located in the same tectonic zone as the small massifs of rapakivi-type granites [66] (Figure 6b) and bimodal complex of granite and mafic dykes [5] (Figure 5). The intrusive body of a lens-like shape and an area of ~140 km² occurs at large submeridional faults. Based on geophysical data, it is suggested that the massif has a subvertical dip, “plate-like” shape, and tectonic contacts with host rocks [67].

Rocks of the main intrusive phase comprise ~90% of the CM area and are represented by coarse-granular granite and granite–gneiss with ocellar and gneissoid structures (Figure 6e,f). Phenocrysts with a size up to 3 cm are represented by isometric grains of K-feldspar, and rarely by smaller plagioclase segregations. The marginal texture and large-ovoid structure of the primary magmatic rocks are well preserved in some areas of rocks. The groundmass is composed of a medium-granular aggregate of isometric and anhedral grains of quartz and subhedral oligoclase (An_17–21). Biotite forms elongated fibrous aggregates or glomeroblastic segregations. Among the accessory minerals are discrete grains of garnet, apatite, monazite, zircon, and ilmenite.

5. The Osinovka massif is an intrusive body that intrudes a Late Neoproterozoic metavolcanic-terrigenous stratum of the island-arc type, predominantly composed of rocks of calc-alkaline dacite-andesite-basaltic volcanic series. The intrusive body is accompanied by a series of dikes composed of quartz porphyries that developed in the Yenisey River Valley and on Dyadya Island, 0.6–1 km south of the tectonic contact with the massif (Figure 3). The massif is composed of coarse- and medium-grained leucocratic light pink granites. The quantitative mineralogical composition is as follows (vol.%): 15–25% plagioclase, 35–40% K-Na feldspar, 35–40% dark morion-like quartz, and 1–5% biotite. The accessory minerals are zircon, apatite, and magnetite (rarely, monazite, garnet, fluorite, and pyrite). K-Na feldspar forms large (up to 15–20 mm in size) elongated grains with a characteristic perthitic structure. Plagioclase appears in the form of idiomorphic tabular grains (5–10 mm in size) represented by fine polysynthetic twins of intensely pelitized oligoclase, often with peripheral rims of albite free of inclusions. Quartz is observed in the form of individual idiomorphic units and aggregates grown with feldspars. Biotite is pleochroic in reddish brown tones and forms fine (1–3 mm) scattered tablets, often replaced by chlorite to various degrees [68].

6. The dominant lithology of GC in the lower-middle course of the Garevka River (Figure 5) is represented by granite-gneisses of the Malaya Garevka Group with a distinct primary coarse-grained porphyroblastic texture and a lenticular-spotty structure formed by K-feldspar grains oriented parallel to the foliation plane (Figure 7a). These rocks often exhibit a well-developed ductile deformation texture and contain a recrystallized Pl–Qz–Bt–Ms matrix with minor Grt and Ilm and accessory Ap, Tur, Zrn, and orthite (Figure 7b). They contain zones with relic migmatitic textures, showing gradual transitions to the host rocks and indistinct contacts with zones of hybrid rocks [46].

7. Migmatites contain abundant leucocratic veins and rare nebulites in more melanocratic rocks (Figure 7c). Grt–Bt–Ms schists occur locally and contain zoned Grt grains. They display subvertical to vertical schistosity and metamorphic banding formed by Bt–Ms aggregates. Plagiogneisses of the Nemtikha Group occur as subhorizontal bodies and have a fine-grained lepidogranoblastic texture and are essentially composed of Bt–Pl–Qz with minor amounts of epidote group minerals (Figure 7d) [46].

8. Metamorphosed rapakivi-type Bt granites occur as small boudinaged pods and bodies within migmatized gneisses with porphyroblasts of microclinized orthoclase (Figure 7f). They are characterized by well-preserved primary mineral compositions, Carlsbad twinned orthoclase megacrysts (Figure 7g,h), marginal structure (Figure 7k), and indistinct magmatic foliation.

Microcline–zoisite orthoamphibolites, metamorphosed equivalents of mafic rocks (gabbro and monzodiorite), are present locally as small pods and bodies consisting of nematoblasts of Amp and Pl and fine prismatic crystals of Zo and Mc oriented parallel to the foliation plane. Secondary alteration resulted in the development of Chl, Ep, and Cal [46].

The results of the pressure and temperature calculations for the samples from metamorphic rocks are summarized in [45]. A distinctive feature of the studied granitoids (1–4) compared to those from the porphyroblastic granite-gneisses, schists, migmatites, and rapakivi-type granites (6–8) of GC is their crystallization from relatively hotter magmas (T = 820–840 °C vs. T = 650–750 °C) at temperatures exceeding those of the collisional metamorphism for the enclosing blastomylonitized gneisses (T = 625–630 °C at P = 7.7–7.9 kbar), which is supported by calculations of the melt saturation with zirconium. Independent estimations of the depths and temperatures of consolidation of some massifs in the Garevka Complex including CM obtained by projection of the normative compositions of granitoids on the ternary Ab–Kfs–Qz diagram provide the pressure range of 6–8 kbar at temperatures of 680–720 °C [45].

2.2.3. Basites

1. Microcline–zoisite orthoamphibolites of the Garevka Complex (samples 6, 9, 10) (Figure 5), metamorphosed equivalents of mafic rocks (gabbro and monzodiorite), are present locally as small pods and bodies consisting of nematoblasts of Amp and Pl and fine prismatic crystals of Zo and Mc oriented parallel to the foliation plane. Secondary alteration resulted in the development of Chl, Ep, and Cal [45].

2. The metagabbros of the bimodal dike belt occur as strongly differentiated bodies of dark medium-grained rocks (Figure 6c) with variable proportions of major minerals dominated by amphibole and oligoclase, sometimes in association with garnet. Secondary phases are chlorite and calcite. The association of opaque and accessory minerals includes titanite and magnetite. Fine-grained aphanitic varieties are common in the peripheral parts of the dikes. Calculations show that the P–T estimates vary significantly between different rocks from T = 695–770° C and P = 8.4–9.0 kbar for the mafic dikes to T = 580–630 °C and P = 7.1–8.6 kbar for other mafic GC rocks at a distance of 40 km from the narrow YRSZ with a width of about 20 km (samples 6, 9, 10; Figure 5), within the analytical error [5].

3. The section of the metapicrite–basalt sequence of the Angara region is exposed in the Nizhnyaya River area and downstream of its mouth, and on the left and right banks of the Yenisey River in front of the Belokopytovskie Islands (Figure 1). This section has the following composition (in vol.%): metapicrites 15, metapicrobasalt 48, metabasalt 20%, and carbonate terrigenous rocks 17%. The coastal exposures recover deformed, highly cleavage metamorphosed rocks of the basalt–picrobasalt–picrite composition, which are transformed into tremolite–actinolite, serpentine–actinolite, and chlorite–actinolite schists. This section shows mainly the lower part of the sequence, with most mafic rocks subjected to the greenschist to epidote–amphibolite facies metamorphism [69].

2.2.4. Mafic and Felsic Metavolcanites

1. Amygdaloidal basalts were found on the right bank of the Yenisey River (Figure 3). The basaltic body underwent greenschist metamorphism from the periphery. The central part of the body (from which sample 15-14 was collected) is represented by unaltered massive rocks of amygdaloidal structure, containing rare phenocrysts (up to 5–7 mm in size) of pelitized andesine. At the periphery of the basaltic body, amygdales are composed of ore mineral, zeolites, and chlorite, whereas in the central part, they consist of chalcedony and carbonate. The matrix of basalts (99 vol.%) is represented by greenish brown undecomposed isotropic volcanic glass (60–70%) with admixtures of chlorite (10–15%), carbonate (4%), and microleists of secondary albite (15–20%). Ore mineral forms uniform fine impregnations in the groundmass and characteristic seesaw skeleton crystals at the boundaries between amygdales and the groundmass. Zircon is an accessory mineral [10].

2. Metadacites of the metadacite–andesite–basaltic association are located on the right bank of Yenisey River (Figure 3). Metadacites comprise finely interbedded gray horizons with a massive and schistose structure, which have different mineral compositions. They are composed of microgranoblastic quartz aggregate (60–70%); deformed albite–oligoclase phenocrysts up to 2 mm in size (10–20%); and muscovite, biotite, and chlorite (up to 20%), which emphasize the lens–schistose structure of metavolcanic rocks. Secondary alterations are seen in epidote developed on plagioclase. Accessory minerals are represented by zircon or, rarely, apatite. For our studies, we collected sample 15-07 from massive interbeds of metadacites, which were the least schistose and metamorphosed [10].

2.2.5. Metapelites of the Garevka and Teya Complexes

The metapelite samples used in this study were collected within GC in the lower course of the Tis (sample 58–59°36′21.9 N, 91°09′42.99 E) and Garevka River (sample 27–59°51′10.5 N, 90°49′36 E) and from outcrops on the right bank of the Yenisey River (sample 56—59°16′48.5 N, 91°20′29.2 E) (Figure 5). The metapelitic rocks of the study area are Grt + Bt + Ms + Pl + And + Ky ± Sil + Qz ± St ± Ilm ± Chl ± Ep intensely deformed gneisses and schists crosscut by steeply dipping, sheeted, coarse-grained felsic dykes [5].

Unlike other mineral assemblages of GC, these metapelites contain zoned garnet porphyroblasts [13]. Garnet up to 2–3 mm in diameter is wrapped by an S₂ foliation defined by biotite and muscovite, which together with quartz, plagioclase, and ilmenite form the rock matrix. Most of the garnet porphyroblasts show three texturally distinct zones (core, mantle, and rim) (Figure 8). The cores are composed of rounded garnet admixed with randomly oriented mineral inclusions. The mantle of a strongly deformed garnet grain is crowded with oriented matrix inclusions and contains abundant ilmenite and graphite inclusions that are concentrated at the boundary between the core and the mantle. The inclusions within the mantle define a straight S₁ fabric. Garnets of sub-idiomorphic shapes exhibit thin rims (up to 0.4 mm). These types of garnets often have discontinuous edges and contain recrystallized quartz-rich strain shadows. Along with abundant inclusions of the matrix minerals, almost all garnet zones contain monazite, zircon, apatite, and tourmaline and rare epidote and xenotime. The latter two minerals are found only in the core, whereas monazite is present in all three zones [70].

The in situ U-Th-Pb geochronology of monazite and xenotime from different growth zones of the garnet porphyroblasts coupled with P–T path calculations derived from garnet zoning patterns records the three superimposed metamorphic events. The first stage is correlated to the Grenville-age orogeny during late Mesoproterozoic–early Neoproterozoic (1050–850 Ma) and included low-pressure metamorphism at c. 4.8–5.0 kbar and 565–580 °C under a metamorphic gradient of dT/dZ = 30–40 °C/km typical of orogenic belts [50,51,52]. During the second stage in middle Neoproterozoic (800 Ma), the regionally metamorphosed rocks experienced collision-related medium-pressure metamorphism at c. 7.7–7.9 kbar and 600–630 °C with low dT/dZ ≤ 10 °C/km associated with the Valhalla orogenic events [71]. The final stage evolved as a synexhumation retrograde metamorphism (785-776 Ma) at c. 4.8–5.4 kbar and 500 °C with dT/dZ ≤ 14 °C/km and recorded uplift of the rocks to upper crustal levels in shear zones The final stage of the Neoproterozoic evolution of the Yenisey Ridge was marked by the latest metamorphic event in the late Neoproterozoic (Ediacaran) between 615 and 603 Ma, related to the accretion of the Isakovka and Predivinsk island-arc terranes to the western margin of the Siberian Craton [17,39,62]. The three discrete stages define a counter-clockwise P–T path involving initial prograde low-pressure heating followed by near isothermal medium-pressure compression and post-peak retrograde decompression and cooling (Figure 9) [60].

The study area of the Teya Complex is essentially composed of metamorphosed metacarbonate-terrigenous rocks of the Lower- and Middle-Proterozoic Teya, Sukhoy Pit, and Korda Groups intruded by c. 850–860 Ma calc-alkaline granites of the Kalama massif [72]. The emplacement of granitic intrusions produced an extensive thermal contact aureole, up to 1 km in width, with a distinct metamorphic zonation from chloritoid to sillimanite-K-feldspar zones, indicating that the conditions of metamorphism were along a high metamorphic field gradient with dT/dZ ≥ 100 °C/km [28]. The regional folded structure is complicated by faults of NW strikes attributed to the Uvolga fault zone in the northern sector of the Tatarka deep fault (Figure 1).

The metapelite samples used in this study were collected from four locations (Mayakon, Polkan, Chapa, and Teya areas) along the Tatarka-Ishimba regional fault in the eastern segment of the Central Transangarian block (Figure 1). They all represent characteristic examples of metamorphic superimposed zoning of low- and medium-pressure facies series. One of the most illustrative and shining examples of a combination of metamorphic zoning of the two facies series is the Teya area in the middle reaches of the Teya River, in the interfluve of the Kurepa and Uvolga Rivers, north of the Teya granitoid massif (Figure 10). This region is composed of Early–Middle Proterozoic metacarbonate-terrigenous rocks of the Teya and Sukhoy Pit groups intruded by the granitoids of the Kalama massif. Four regional metamorphic zones defined by isograds were mapped from SW to NE (Figure 10). The metamorphic grade increases in the same direction, giving rise to the following sequence of mineral assemblages: (1) Bt + Ms + Chl + Qz + Pl (Bt zone), (2) Grt + Bt + Ms + Chl + Qz + Pl (Grt zone), (3) St + Grt + Bt + Ms + Chl + Qz + Pl + Crd ± And (St-And zone), and (4) Sil + St + Grt + Bt + Ms + Qz + Pl ± And ± Crd (Sil-St zone). In the St-And zone, Ged and Cum are stable with Grt and locally with Crd in low-aluminous, K₂O-poor metaterrigenous rocks of the Ryazanov Group. Regionally metamorphosed metapelites have a fine- to medium-grained lepidogranoblastic Ms-Chl-Bt-Qz-Pl matrix with Grt, St, And, and Sil porphyroblasts [51,52].

The metamorphic zonation exhibits a roughly symmetric pattern around the Teya anticline with the metamorphic grade increasing from the Bt to Sil zones toward the anticline core. This prograde metamorphic zoning suggests a shallow-level, LP/HT metamorphism of the andalusite-sillimanite type, intermediate between the Pyrenean- and Michigan-type facies series of Hietanen [73]. The P–T values correspond to the transition from the greenschist facies to the boundary between the epidote-amphibolite and amphibolite facies.

Close to the overthrust, the rocks of the St-And and Sil zones are affected by metamorphic overprint. This transition is recognized through the simultaneous appearance of kyanite and fibrolite (kyanite isograd) and the presence of the limiting assemblage Ky + St + Grt + Ms + Bt + Qz + Pl + Sil with andalusite relics. The rocks of the kyanite-staurolite subfacies of the kyanite schist facies occur within a 4–5-km-wide zone bounded by an NW-trending fault, which implies, together with intersecting isograds, the local character of the overlapping metamorphic event (Figure 8). Adjacent to the thrust, this metamorphic overprint was accompanied by the development of blastomylonites, which differ in the proportions of relicts and newly formed phases and the degree of rock deformations. Blastomylonites of the outer zone contain partially resorbed andalusite porphyroblasts, which are locally replaced by Ky-St-Bt aggregates. The square and prismatic cross-sections of the cataclased grains of andalusite acquire oval or rhombic shapes, with the simultaneous appearance of fibrolite. The inner zone is characterized by complete recrystallization of minerals and the absence of And relics and mostly contains Ky as the Al₂SiO₅ polymorph with subordinate Sil. Here, adjacent to the thrust, rhombic pseudomorphs are ubiquitously transformed into lenses, with their long axes oriented parallel to the schistosity, and strain shadows of recrystallized quartz are present in the matrix. The appearance of Ky and deformation microstructures indicate a medium-pressure (Barrovian-type) kyanite-sillimanite series metamorphic overprint.

In all studied areas (Figure 11), rocks with “triple point” assemblages show similar sequences in the changes in the mineral assemblages in metamorphic zones produced at the same temperature, but the proportions of relict and newly formed minerals are different, as are the intensity of rock deformations. The mineralogical differences include the absence of cordierite (Chapa and Polkan areas) and the development of chloritoid (Polkan and Mayakon areas) at the lowest metamorphic grades. The occasional occurrence of rare mineral assemblages (Cld + Bt and Cld + Bt + And) and changes in the character of zoning in the metapelites of the And–Sil type in the Polkan and Mayakon areas may be explained by the Mn-bearing garnet being more stable at medium metamorphic grades [31].

The microtextural relations between Al₂SiO₅ polymorphs testify that andalusite, sillimanite, fibrolite, and kyanite grew successively during metamorphism, with the dominance of various scenarios of reaction replacements between these minerals (Figure 12). The Mayakon area is characterized by the reaction relations And → Ky → Sil ± Fi; and the relations at the other areas are And → Ky → Sil in the Polkan area; And → Sil + Ky in the Chapa area; and And → Sil → Ky + Fi in the Teya area. These features were controlled by the complicated metamorphic history of the rocks in relation to changes in the tectonic environments.

In all the studied examples, Neoproterozoic Ky–Sil medium-pressure metamorphism was overprinted on regionally metamorphosed low-pressure andalusite-bearing rocks [47]. Zonal low-pressure LP/HT metamorphic complexes (P = 3.9–5.1 kbar, T = 510–640 °C) of the andalusite–sillimanite type were produced at a metamorphic field gradient typical of orogenesis: dT/dZ = 25–35 °C/km (Figure 9). In the rock succession of Ky–Sil metamorphism of the Teya complex, the highest pressure and temperature metapelites of the Chapa (P = 5.8–8.4 kbar, T = 630–710 °C, dT/dZ = 12–14 °C/km), Teya (P = 5.65–7.15 kbar, T = 650–700 °C, dT/dZ = 10–12 °C/km), and Polkan (P = 5.0–7.3 kbar, T = 575–645 °C, dT/dZ = 8–10 °C/km) areas were found in the northern part of the territory and are constrained to older Early Proterozoic units of the Teya Group (Likhanov et al., 2009). The metapelites of the Mayakon area crop out south of these rocks, among younger Middle Riphean rocks of the cordierite suite, which are characterized by slightly lower P–T parameters and metamorphic field gradients (P = 4.5–6.7 kbar, T = 560–600 °C, dT/dH = 6–7 °C/km) (Figure 9). The highest P–T parameters are typical of metapelites in the Chapa area, which ubiquitously contain sillimanite in practically all metamorphic zones. In other metamorphic complexes of the Ky–Sil type, sillimanite occurs more rarely [74,75].

The older metamorphic complexes of Ky-Sil type (Mayakon, Teya, and Chapa areas) were formed due to thrusting rock blocks from the side of the Siberian Craton onto the Yenisey Ridge at ~850 Ma, which is confirmed by geophysical data and study of the nature and age of source areas. The late, repeated collisional metamorphism with an age of ~800 Ma was caused by opposite movements of east-directed small blocks in the subsidiary fault zone of the higher order (Garevka, Yenisey, and Tis areas) owing to the accretion–collisional events of the Valhalla folding [56].

3. Materials and Methods

3.1. Whole-Rock and Mineral Chemistry

Major and trace element compositions were determined by XRF using a VRA-20R energy-dispersive spectrometer (Carl Zeiss, Jena, Germany) and by ICP-MS using a VG Element PQ2 (Finnigan Mat, Mainz, Germany) analyzer at the Analytical Center of the Sobolev Institute of Geology and Mineralogy (Novosibirsk) and at Common Use Center “Geoanalyst” of Zavaritsky Institute of Geology and Geochemistry (Ekaterinburg, Russia), following the procedures described elsewhere [76]. The detection limits were 0.02–0.005% for most major oxides. For MgO and Na₂O, the detection limits were 0.05 and 0.1%, respectively. The detection limits of REE and HFSE were 0.005–0.1 ppm. The average accuracy was 2–7 rel.%.

The chemical compositions of all minerals were analyzed with a Jeol JXA-8100 electron microprobe at the Sobolev Institute of Geology and Mineralogy, Novosibirsk following the procedures described elsewhere [77]. The operating conditions were 40 nA beam current, 20 kV accelerating voltage, and 20 s counting time for all elements, with an electron beam diameter of 3–4 μm. PAP correction was applied to the data. Natural and synthetic silicate and oxide standards were used for calibration.

3.2. Geothermobarometry

The metamorphic conditions of the metabasites and metapelites were deduced on the basis of pressure estimates derived from net transfer equilibria and temperature estimates obtained using exchange thermometers. For each thin-section domain, a pressure-temperature estimate was calculated by simultaneous solution of these equilibria by means of the algebraic techniques of MATHEMATICA 5.0, using the built-in function NullSpace [78]. To test the reliability of these P and T estimates, the same analytical and assemblage data were processed with the P–T calculations, which were made using the computer software [79]. The pseudosections were computed with THERMOCALC 3.33 [80] and the internally consistent thermodynamic dataset 5.5 [81] in the system Na₂O-CaO-K₂O-FeO-MgO-Al₂O₃-SiO₂-H₂O.

The temperatures of initial granitoid melt crystallization were calculated from the equilibrium Zr content in zircon-saturated melt as a function of the temperature and composition [82].

3.3. SHRIMP U-Pb Zircon Dating

In situ U-Pb zircon analyses were performed using SHRIMP II at the Center of Isotopic Research of the Russian Geological Research Institute, St. Petersburg (CIR VSEGEI). The selected zircon grains were mounted in epoxy resin together with chips of TEMORA and 91,500 reference zircon. The grains were approximately sectioned in half and polished. The choice of analytical sites was based on optical and cathodoluminescence images (CL). The U-Pb measurements were carried out using a technique described by Williams [83]. The intensity of the primary O₂ beam was 4 nA, and the spot was 25 μm in diameter. The obtained results were processed with the SQUID program. The U/Pb ratios were normalized to the value of TEMORA standard zircon, 0.0668, corresponding to the age of 416.75 Ma [84]. The errors of single analyses (U/Pb ratios and ages) are given at the 1δ level, and the errors of the calculated concordant ages and intercepts with the Concordia are at the 2 δ level. The concordia plots were constructed using the ISOPLOT/Ex program.

3.4. U-Th-Pb Monazite and Zircon Ages

The proposed stages of the metamorphic history of the rocks were dated using the in situ ages of monazite and xenotime inclusions in different generations of garnet (Figure 8). Quantitative analyses of U-Th-rich minerals were performed on a Cameca SX 100 electron microprobe equipped with five WD spectrometers at the Institute of Geology and Geochemistry, Ural Branch, Russian Academy of Sciences. Monazite and xenotime grains were located and characterized in polished thin sections by BSE imaging and energy-dispersive spectroscopy. The elemental mapping was performed by measuring the critical intensities of the X-ray lines across the monazite grains. The operation conditions were 15 kV accelerating voltage and 270 nA beam current for monazite and 200–250 nA for xenotime, and a 2–5 µm spot size. The counting times were 400 s for Th, U, and Pb, and 10–20 s for all other elements on peaks and twice as low at each of the two background positions [85]. The standards used for electron microprobe analysis included ThO₂, UO₂, Pb₂P₂O₇, synthetic LREE (for monazite) and HREE (for xenotime) phosphates, and diopside. The average detection limits for Th, U, and Pb were 147, 83, and 65 ppm in monazite, and 156, 90, and 40 ppm in xenotime. The analytical errors of the EMP analysis were ΔTh/Th = 0.9% rel., ΔU/U = 1.7% rel., ΔPb/Pb = 3.2% rel. for monazite, and ΔTh/Th = 9.5% rel., ΔU/U = 1.2% rel., and ΔPb/Pb = 1.9% rel. for xenotime (average of 16 spot analyses). Errors in the absolute ages in the range of 11–30 Ma for each run are typically within the error limits obtained worldwide, suggesting a reasonable treatment scheme [86]. Ages were calculated using a modified version of Isoplot 3.66 [85] and following the two alternative approaches: the individual ages calculated from the Th, U, and Pb concentration for individual spot analysis within a single grain [87] and the isochron ages on the ThO₂*/PbO (for monazite) and UO₂*/PbO (for xenotime) coordinates calculated from the populations of grains [88]. The age error on the ThO₂*/PbO diagram was reduced by the inclusion of a plotting element [89] with coordinates equivalent to the absolute error of the Th* and Pb content.

3.5. ⁴⁰ Ar/³⁹Ar Mica Dating

⁴⁰Ar/³⁹Ar measurements were performed using conventional techniques previously described by [90,91]. The extracted mineral fractions with a size no less than 0.15 mm together with the MCA-11 and LP-6 biotite standards as monitors were wrapped in Al foil and vacuum sealed in fused silica tubes. The samples were irradiated in a VVR-K cadmium-lined research reactor at the Nuclear Physics Research Institute of the Tomsk Polytechnical University (Russia). The neutron flux gradient was no greater than 0.5%. ⁴⁰Ar/³⁹Ar step heating experiments were performed using a quartz heated by an external furnace. Typical ⁴⁰Ar blank values of 5 × 10^–10 ncm³ were achieved at 1200 °C for 10 min. Released argon was doubly purified by exposure to Ti and ZrAl SAES getters. The argon isotope composition was measured on a Micromass 5400 noble gas mass spectrometer (UK). The apparent ages are quoted at the 1σ (sigma) level, whereas all plateau ages are reported at the 2σ level and do not include the errors on the age of the monitor.

In order to reproduce the thermal history, we conducted numerical simulation of the ⁴⁰Ar/³⁹Ar stepwise-heating data by solving the inverse problem based on the theory of multiple diffusion domains [92]. The Ar behavior in feldspars was described by a set of discrete diffusion domains of different sizes that do not interact with one another. The diffusion coefficients were calculated under the assumption of a tabular geometry. The parameters characterizing individual diffusion domains (activation energy, frequency factor, and volumetric fraction) were selected based on the kinetics of neutron-induced ³⁹Ar in vacuum. Then various variants of the thermal history were chosen using the least-squares procedure by fitting the model age spectra for those obtained experimentally.

3.6. Sm-Nd Dating

The isotopic compositions and concentrations of Sm and Nd were measured at the Geological Institute of the Kola Scientific Center, Russian Academy of Sciences (Apatity) using a Finnigan MAT-262 (RPQ) mass spectrometer with seven collectors operating in static mode using the procedure described in [93]. The blanks for laboratory contamination for Sm and Nd were 0.06 and 0.3 ng, respectively. The precision for the Sm and Nd concentrations was ±0.2% (2σ); the reproducibility of the measured ¹⁴⁷Sm/¹⁴⁴Nd ratios was ± 0.2% (2σ), and the precision of the ¹⁴³Nd/¹⁴⁴Nd measurement was ± 0.003 (2σ). The measured ¹⁴³Nd/¹⁴⁴Nd ratios were normalized to ¹⁴⁸Nd/¹⁴⁴Nd = 0.251578, which corresponds to ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219, and then re-normalized to ¹⁴³Nd/¹⁴⁴Nd = 0.511860 using the La Jolla Nd standard (California, USA). The εNd and T_Nd(DM) values were calculated relative to the present-day chondrite uniform reservoir (CHUR) values of ¹⁴⁷Sm/¹⁴⁴Nd = 0.1967 and ¹⁴³Nd/¹⁴⁴Nd = 0.512638 [94], and depleted mantle (DM) values of ¹⁴⁷Sm/¹⁴⁴Nd = 0.2136 and ¹⁴³Nd/¹⁴⁴Nd = 0.51315 [95].

4. Results

4.1. Major and Trace Element Geochemistry

4.1.1. Metabasite Tectonites

On the basis of the silica content (SiO₂ = 41.8–48.7 wt.%), the studied metabasites correspond to mafic rocks. They are characterized by a moderate content of alkali (Na₂O + K₂O < 3 wt.%) with a predominance of Na₂O over K₂O and substantial variations of Fe₂O₃ (8–18 wt.%) and MgO (6–10 wt.%), TiO₂ (0.9–1.4 wt.%), and P₂O₅ (0.07–0.23 wt.%). The total iron content varies in a narrow range from 0.53 to 0.64. The higher magnesia (MgO > 8 wt.%), lower alumina, and low content of K₂O (0.05–0.3 wt.%) suggest that the majority of the studied rocks are olivine basalts with a transition from picrobasalts to basalts. Calculated on the volatile-free basis, the major elements of most samples correspond to metabasites, with SiO₂ ranging from 45 to 50 wt. % and the rocks plot close together in the basaltic/andesitic field on the classic ACF diagram (Figure 13) [96]. Sample 4, however, is richer in iron, with 43 wt.% of SiO₂, and corresponds to an ultramafic composition. The average rock composition of the mafic tectonites is close to the composition of unaltered ocean-floor basalts (OFB) in the system NCKFMASH-Ti [97]. It should also be noted that the geochemical composition of most rocks varies considerably even within one outcrop, which is confirmed by direct observations of the mineralogical heterogeneity of the tectonized metabasites. The most significant differences between these rocks are in the rare-element composition. More common low-titanium rocks are characterized by reduced contents of REE (total REE = 37–39 ppm) and an oblique concentration distribution profile with (La/Yb)_n < 0.5, which is characteristic of normal basalts of mid-Oceanic ridges (N-MORB). The REE spectra of the less common high-titanium metabasites are enriched in light lanthanides (total REE = 43–54 ppm) and have a flat distribution profile ((La/Yb)_n >1), which is typical of the enriched ocean-floor basalts (E-MORB) (Figure 14a). Common to most multielement spectra is the depletion with respect to LILE (Rb and Ba) and distinct K and Sr minima (Figure 14b).

4.1.2. Granites of the Transangarian Yenisey Ridge

1. The major element chemistry of the gneissic and cataclastic rocks of the Nemtikha Group corresponds closely to that of calc-alkaline plagiogranites (up to 73.6 wt.% SiO₂), which are sodic (Na₂O/K₂O ≈ 2) and highly peraluminous (up to 14.6 wt.% Al₂O₃), with low CaO (<1 wt.%) and MgO (0.89–1.32 wt.%), moderate total alkali (up to 6.5 wt.% K₂O + Na₂O), and a relatively high FeO*/(FeO* + MgO) ratio (>0.7), where FeO* = 0.9Fe₂O₃ + FeO (Figure 15). They are consistently classified as trondhjemites with 4% normative corundum [100] based on the normative feldspar ratios (albite 68%, anorthite 8%, K-feldspar 24%). Chondrite-normalized REE patterns in all the studied samples are characterized by a negative Eu anomaly (Eu/Eu* = 0.53–0.64) and higher ratios of (La/Yb)n = 13.5–22, (Gd/Yb)n (up to 2.15), and LREE/HREE = 17.7–24.9 (Figure 16). The average trace element compositions in these rocks are usually enriched in Ba (538 ppm), Th (25 ppm), Nb (39 ppm), Y (46 ppm), Ga (23.2 ppm), and Ce (154 ppm) and depleted in Rb (58 ppm), Sr (67 ppm), U (2 ppm), V (26 ppm), and Sc (5.2 ppm) as compared to the M-, I- and S-types of granitoid suites, which is indicated by their low Sr/Y (<2) and high Th/U (>12) ratios. Despite the minor variations in the major and trace element contents, the studied rocks show a remarkable compositional homogeneity. A statistical analysis comparing the mean values and dispersion of the major and trace elements between rocks by Student’s t- and Fisher’s F-tests reveals that these rocks have nearly identical major and trace element compositions (overlap within errors) at the 95% significance level.

2. On the basis of their petrochemical characteristics, the major felsic units from the dyke complex can be classified as granites or, more rarely, alkaline granites with f = (FeO + 0.9Fe₂O₃)/(FeO + 0.9Fe₂O₃ +MgO) = 0.82 − 0.66 (Figure 15). They are, in general, peraluminous to metaluminous (ASI = 1.09–1.24) high-K (K₂O + Na₂O = 7.8–10.7 wt.% at K₂O/Na₂O = 2.3–2.8) rocks of mostly calc-alkaline and less commonly alkaline series (Figure 6), with high contents of Y, Th, U, Zr, Hf, Nb, and Ta; and moderate to high REE contents (106–253 ppm). The chondrite-normalized REE patterns of all samples are characterized by a negative Eu anomaly (Eu/Eu^* = 0.33–0.7) and high ratios of (La/Yb)_n = 2.9–10.6, (Gd/Yb)_n = 0.7–1.2, and (LREE/HREE)_n = 3–6 (Figure 16).

3. The main varieties of granitoids of Ostrovok correspond to granites with moderate concentrations of SiO₂ (up to 70 wt.%) and relatively low concentrations of Al₂O₃ (usually < 14 wt.%) with a high iron mole fraction f = FeO/(FeO* + MgO) = 0.82, where FeO* = FeO + 0.9 × Fe₂O₃. The average concentrations of major oxides and minor elements are given hereinafter. As a whole, these rocks are peraluminous (ASI = 1.03) subalkaline and high in potassium (K₂O + Na₂O = 8.2 wt.%, K₂O/Na₂O > 2) and belong to the calc-alkaline series (Figure 15). The normative composition of the granite is as follows (vol.%): quartz 30, orthoclase 35, and plagioclase 25; the relationship between normal feldspars (albite 33, anorthite 9, and orthoclase 58) allows us to identify these rocks as subalkaline granites with normal corundum, enstatite, and titanite (~2%). They are characterized by high concentrations of REEs (254 ppm), Y, Zr, Hf, Nb, and Ta; a clearly reflected Eu anomaly (Eu/Eu* = 0.40); and a poorly differentiated spectrum of the REE distribution (La/Yb)_n= 8.4, (Gd/Yb)_n= 1.7 (Figure 16).

4. On the basis of their petrochemical characteristics, the main varieties of Chernorechensky massif granitoids correspond to granites with medium or high concentrations of SiO₂ (68–76 wt.%) and significantly variable concentrations of Al₂O₃ (12–18 wt.%). These rocks are enriched in iron (FeO/(FeO + MgO) = 0.80–0.84) in comparison with S-type granites formed under the syn-collisional settings. In general, these iron-rich leucogranites are peraluminous (ASI = 1.1–1.4) subalkaline and high-potassium rocks (K₂O + Na₂O up to 9 wt.% at K₂O/Na₂O = 2.4–3.6) of the calc-alkaline, rarely alkaline, series (Figure 15). Their origin is usually explained by an undepleted magmatic source or fractional crystallization of crustal melts. The normative composition of the leucogranite is as follows (vol.%): quartz 25–30, orthoclase 30–40, and plagioclase 25–30 (albite 32%, anorthite 10%, orthoclase 58%), which allows us to identify this rock as subalkaline granite with normative corundum (2–4.4%), enstatite (2%), and ferrosilite (1–4.7%). It is characterized by a high concentration of REE (up to 390 ppm), Rb, Th, U, Y, Zr, Hf, Nb, Ta, and Sc; and a low concentration of Sr in comparison with other associations of GC; a clear Eu anomaly (Eu/Eu* = 0.1–0.35); and poorly fractionated REE spectrum, with (La/Yb)_n= 6.8–10 and (Gd/Yb)_n= 1.1–1.8 (Figure 16).

5. Compared with other mineral assemblages of GC, the rapakivi-type granites exhibit more pronounced Eu anomalies (Eu/Eu* = 0.29–0.35); higher K₂O, FeO, Y, Th, U, Zr, Hf, Nb, Ta, and REE; and lower transition element contents (V, Ni), along with the high La/Sc and La/Th and low Co/Th ratios (Figure 16).

6. The petrochemical composition of granites from the Osinovka massif corresponds to subalkaline two-feldspar leucogranites, and that of quartz porphyries corresponds to trachyrhyolites. These are generally weakly aluminous (Al₂O₃/(Al₂O₃ + K₂O + Na₂O) = 0.95–1.11) high-potassium (K₂O + Na₂O = 7.6–9.1 wt.% at K₂O/Na₂O = 1.2–1.8) rocks referring to predominantly calc-alkali and rarely alkali series (f = (FeO + 0.9Fe₂O₃)/(FeO + 0.9Fe₂O₃ + MgO) = 0.80–0.86). The aggregated REE distributions for granites, normalized to the chondrite composition, are characterized by clearly expressed large Eu anomalies (Eu/Eu* = 0.18–0.52) and have a significant negative inclination on the concentration profile in the zone of light lanthanides (light and heavy REE are hereinafter LREE and HREE, respectively), which is indicated by the higher values of certain ratios: (La/Yb)n = 6.3–12.5 and LREE/HREE = 7.7–14.6 (Figure 17). In the zones of medium and HREE (from Gd to Lu), these spectra are characterized by a predominantly flat (plateau) distribution. However, we found slightly different behavior of the REE distribution spectra (La/Yb)n = 2–3 and (Gd/Yb)n = 0.86–0.89 for quartz porphyries (samples 393 and 409) (Figure 6a), which differ from granites by an anomalously negative Eu anomaly (Eu/Eu* = 0.03–0.05), lower values of certain ratios (La/Yb)n = 2.1–3.3 and LREE/HREE = 3.9–5.4, and for HREE concentrations exceeding those in chondrites by a factor of 30. Like other minor elements, quartz porphyries are significantly depleted in Ba and Sr, whereas they are enriched in Y, Nb, and Ta compared to granites (Figure 17).

4.1.3. Basites of the Transangarian and Angara Parts of the Yenisey Ridge

The metagabbros from the dike belt are classified as subalkaline and moderately alkaline basalts, with SiO₂ = 45.6–47.5 wt.%, (Na₂O + K₂O) up to 3.6 wt.%, moderate TiO₂ contents of up to 2.1 wt.%, low and moderate Fe₂O₃ contents of 10.8–13%, and elevated P₂O₅ contents of 0.22–0.32 wt.%. The Fe/(Fe + Mg) ratio varies from 0.61 to 0.63.

Compared to the associated felsic rocks, the metagabbros are enriched in elements showing affinity to plagioclase (up to 1200 ppm Sr and Eu), and Ni, V, Co, and Sc but depleted in all other compatible and incompatible elements. Their chondrite-normalized REE patterns are characterized by a weak negative Eu anomaly (Eu/Eu* = 0.85–0.92) and slight variations in their indicator ratios: (La/Yb)_n of 2.5–6.4 and (Gd/Yb)_n of 1.6–2.1 (Figure 18a). Compared to the other GC metagabbros, they are strongly enriched in Sr, Y, and Ni; depleted in Cs, Rb, and Th; and have high Co/Th and La/Th and low Th/U. Their multielement patterns are characterized by the absence of Nb and Ta depletion with respect to Th and LREE (Figure 18b).

The metapicrites of the Angara region have the high contents (in wt.%) of MgO (23.7–26.8) and Fe₂O₃ (12.6–14.6); elevated TiO₂ (1.2–1.4); and low SiO₂, CaO, and alkalis. The metapicrobasalts also have elevated contents of these oxides. Based on the seven analyzed samples, their average composition (wt %) is as follows: TiO₂ = 2.1; Fe₂O₃ = 11.8; MgO = 14.2. In addition, they have much higher contents of CaO = 9.5% and alkalis (Na₂O = 2.3; K₂O = 0.37%). Similar petrochemical features are typical of metabasalts and metagabbros associated with metapicrite–picrobasalts in the same lower part of the sequence. The metabasalts are enriched in Ti and Fe and have an elevated Mg content (in wt.%): TiO₂ = 1.94; Fe₂O₃ = 12.9; MgO = 8.9. They also show a higher alkalinity: Na₂O + K₂O up to 4.2%, P₂O₅ up to 0.5%. In terms of total alkalinity, the metabasalt–picrobasalts deviate from the normal petrochemical series toward moderately alkaline, subalkaline series, which is consistent with their enrichment in Ti and Fe. Picrites show the highest Cr, Co, and Ni contents, which subsequently decrease in picrobasalts and basalts, correlating with the MgO content (Figure 19). The picrites have the lowest contents of incompatible elements: Rb, Cs, Ba, Sr, Zr, Hf, Nb, Ta, Th, and U. The value of their admixture significantly increases in picrobasalts, basalts, and gabbros. In addition, the metagabbros show significant (2–4 times) enrichment in V and Sr. High-K high-Ti subalkaline metavolcanic rocks (trachybasalts) sharply differ from metabasalts of normal series with 4–8 times higher Rb and Cs contents, and 1.5–3 times higher contents of Ba and high-field strength elements (Zr, Hf, Nb, Ta, Th, and U) (Figure 19). They also have elevated contents of Cr, Ni, and Co. The paragenetic relationships of metamorphosed picrites, picrobasalts, basalts, and gabbro of the considered association follows from the REE abundance and distribution patterns (Figure 19). The REE contents in them are close and are usually only 1.5–2 times higher in metabasalts compared to metapicrites. Noteworthy is the remarkable similarity of the REE distribution in rocks of the given association. The (La/Yb)n ratio in them varies from 6 to 7, (La/Sm)n, from 1.6 to 2.2, (Gd/Yb)n, and from 2 to 3 while the Eu/Eu* anomaly is 0.9–1.0.

4.1.4. Mafic and Felsic Metavolcanites

In the petrochemical composition, metadacites (sample 15-07) correspond to felsic volcanites and aluminous (ASI = 1.23) Na-K (K₂O + Na₂O = 7.55 wt.% at K₂O/Na₂O = 0.9) magnesian (f = 0.62) granitoids of the calc-alkali series. The REE distributions for metadacites, normalized to the chondrite composition after [99], are characterized by clearly expressed europium anomalies (Eu/Eu* = 0.45) and have a significant negative slope of the concentration profile in the zone of light lanthanides, as indicated by the higher values of certain ratios: (La/Yb)n = 10.5 and LREE/HREE = 10.7. In the zone of medium REE and HREE (from Gd to Lu), these spectra have a flat distribution pattern (Figure 20). In the contents of most chemical elements, they are comparable to low-alkaline plagiogranite porphyries of the Porozhnaya massif of an arc nature [100]. Substantial differences in the composition are observed when they are compared with subalkaline leucocratic granites from the Osinovka massif (Figure 17). Metadacites are characteristic of much lower concentrations of radioactive (U, Th, and K), high-charge (Nb, Ta), and rare-earth elements (Figure 17).

Amygdaloidal basalts (sample 15-14) (SiO₂ = 47.24 wt.%) are characterized by moderate alkali contents (Na₂O + K₂O = 3.3 wt.%), with a significant predominance of Na₂O above K₂O, and Fe₂O₃ (9.79 wt.%), MgO (5.41 wt.%), TiO₂ (1.53 wt.%), and P₂O₅ (0.31 wt.%) contents. Their REE spectra are enriched in light lanthanides ((La/Yb)n = 4.9; the total REE is 105 g/t) compared to N-MORB and E-MORB and has a high distribution profile (Gd/Yb)n = 1.32) in the zone of medium and heavy elements (Figure 14). They are also characteristic of higher concentrations of Ti, P, alkaline elements (especially Sr), LILE (Rb, Ba, K), radioactive elements (Th, U), LREE, and high-charge elements (Nb, Ta, Zr, Hf) compared to the N-MORB and E-MORB varieties, normalized to the chondrite composition after [101], and basalts from the Isakovka terrane (Figure 14).

4.1.5. Metapelites of the GC and TC

A selection of whole-rock data for metapelites from the studied areas reveals that the most prominent compositional difference among the samples involves variations in SiO₂ (58.6–61.2 wt.%), K₂O (2.6–3.7 wt.%), Na₂O (0.2–0.5 wt.%), and CaO (0.45–1.0 wt.%). Such changes are probably related to local variations in the relative abundance of quartz-rich and extremely mica-rich layers within the samples, and in the modal amounts of plagioclase, quartz, and micas. The studied rocks are homogeneous metapelites with a narrow range of high Fe- (X_Fe = FeO/(FeO + MgO + MnO) = 0.65–0.8 whole rock on a mole basis) and high Al- (X_Al = Al₂O₃−3K₂O/(Al₂O₃–3K₂O + FeO + MgO + MnO) = 0.30–0.5) bulk compositions with respect to the average pelite whole-rock compositions (X_Fe = 0.52 and X_Al = 0.13) reported by Shaw [102]. In the AFM projection of Thompson [103], they are plotted above the garnet-chlorite tie-line (Figure 20). The bulk-rock Mn content of the pelitic schists (X_Mn = MnO/(MnO + FeO + MgO) varies from 0.01 to 0.03, consistent with the whole-rock MnO content of normal pelitic schists [104]. Compared to the average pelitic rock compositions reported by Ague [105], the investigated rocks are more depleted in K₂O, CaO, and Na₂O. The chondrite-normalized REE patterns in all samples are characterized by a negative Eu anomaly (Eu/Eu* = 0.4–0.8) and higher ratios of (La/Yb)_n = 3.7–14.5, (Gd/Yb)_n = 0.7–2, and (LREE/HREE)_n = 3–6 (Figure 21).

4.2. Geochronological Data

4.2.1. Metabasite Tectonites

We analyzed zircons from rock sample 14, in which the garnet grains carry relict Gln-Ab-Ph-Ep association. In the cathodoluminescent image, the zircons of the high-pressure metabasites are characterized by a long- and short-prismatic habit with sector zoning in the cores of grains, suggesting their magmatic origin. Rarely, some zircon grains display poor zoning patterns due to recrystallization processes. In general, zircon grains have normal Th/U ratios in the range of 0.21–1.02, which, taking into account the features of the morphology and internal structure, also indicate the magmatic nature of the grains. The analysis of 10 spots from the core and rim portions of the zircon grains define spot ages in the range of 671–719 Ma, and fall along a concordia with a weighted mean age of 701.6 ± 8.4 Ma, calculated with an error of 2σ (Figure 22a). The similarity in ages from the cores and the rims of the zircon grains probably indicates the recrystallization of cores with the loss of radiogenic lead as a result of metamorphism.

To determine the timing of the deformation–metamorphic transformations of the basement in the YRSZ, we analyzed ⁴⁰Ar/³⁹Ar isotope analysis of mica fractions from the most intensely deformed samples of Grt-Pl-Bt-Ms-Kfs-Qz tectonites. Sample 5 is characterized by laminar flow textures with elements of rotation of porphyroblasts and porphyroclasts. The blastomylonites of sample 6 are distinguished by large lenticular microcline porphyroblasts, which underwent rotation during the shear flow. Sample 8 is a muscovite-bearing fine-grained mylonite. The studied minerals are represented by tabular, crystals in the rock matrix showing an orientation parallel to the general strike of the shear zone. The plateau ages obtained from ⁴⁰Ar/³⁹Ar dating represent the time of the latest phase of the metamorphic-structural-material transformations within the YRSZ. Our results show that these processes occurred in the Ediacaran (Vendian), in a narrow time range of 595–608 Ma (Figure 22b).

4.2.2. Granites

1. Zircon grains separated from the plagiogneissic granite sample 306 are generally 100–250 µm in size and are mainly pale-pink transparent prismatic crystals with an aspect ratio of 2–4. The morphological characteristics and internal structures of the analyzed zircons, together with the Th/U values (0.25–0.54) observed in the zircon cores and rims, testify to their magmatic origin. The data points using zircon cores and rims fall along a discordia that has an upper intercept at 1381 ± 22 Ma (Figure 23a). Four analyses on the zoned cores and a rim are concordant and yield a well-constrained age of 1361.5 ± 8.7 Ma, which overlaps within the error (Figure 23a, inset).

The Sm-Nd analyses yielded ¹⁴⁷Sm/¹⁴⁴Nd = 0.086924 and ¹⁴³Nd/¹⁴⁴Nd = 0.5116344, which are almost similar to the average values of post-Archean pelites from the Yenisey Ridge [106]. The Sm-Nd isotopic signatures of these rocks suggest derivation from a crustal source mostly comprising crystalline and sedimentary rocks of the Siberian craton with the Paleoproterozoic model age (T_Nd(DM-2st) = 2442 Ma, ε_Nd = −5.8).

2. Zircons from two samples of granites (samples 57 and 61) of the bimodal dyke belt are generally 150–600 μm in size and are mainly euhedral and subhedral transparent pale-pink prismatic crystals with an aspect ratio of 2–5. The CL images of most zircon crystals reveal the absence of sector zoning and, although in some cases, weak or indistinct zoning due to recrystallization is present. Zircons with thin oscillatory zoning and melt inclusions in the cores indicative of magmatic origin are rare. Ten spot analyses of the zircon cores and rims (sample 61) define a discordia with an upper intercept of 797 ± 11 Ma (Figure 23c). Nine concordant zircon cores and a rim yielded an identical (within error) age of 791 ± 5.6 Ma (Figure 23c, inset). Eleven spot analyses of cores and rims of zircons from granite sample 57 fall along the concordia and define a weighted mean age of 792 ± 6 Ma calculated at the 2σ error (Figure 23d). The similar ages of the zircon cores and rims may indicate recrystallization of the cores and the loss of radiogenic Pb during subsequent tectonothermal events. Considering the morphological characteristics and internal structures of the analyzed zircons, the ages obtained can be interpreted as emplacement ages for the dyke complex within YRSZ.

3. Zircons from sample 4 of Ostrovok granitoids collected from the tectonic melange are mostly subhedral and euhedral pale-pink transparent prismatic crystals with an aspect ratio of 2–6 and a size of 100–250 μm. The morphological features and internal structure, and the Th/U values (0.23–0.55) in the cores and margins of the grains, provide evidence for the magmatic nature of zircon. Twelve spots from the cores and mainly marginal areas of zircons fall along a discordia with an upper intercept at 756 ± 20 Ma (Figure 23b). A similar age within the error (754 ± 6 Ma) was obtained for nine concordant cores and the rims (Figure 23b, inset).

4. Zircon grains of sample 14 from the Chernorechensky granitoid massif display a long- and short-prismatic habit with thin sector zoning in the cores of grains, suggesting their magmatic origin. Rarely, some grains display poor zoning patterns due to recrystallization processes. In general, zircon grains have Th/U ratios < 1, which, together with the morphology and internal structure, indicate the magmatic nature of the grains. Analysis of 12 spots from the core and rim portions of the zircon grains defines the spot ages in the range of 692–740 Ma and fall along a concordia with a weighted mean age of 723 ± 6 Ma, calculated with an error of 2σ (Figure 24a). The similarity in ages from the cores and the rims of the zircon grains probably indicates the recrystallization of cores with the loss of radiogenic lead as a result of metamorphism. The Late Neoproterozoic ages may be interpreted as the time of consolidation of subalkaline leucogranites of CM in YRSZ.

A similar timing of magmatism was determined through in situ dating of monazite inclusions in minerals of the Chernorechensky granitoids. Monazite in the examined rocks is rich in Ce₂O₃ (28.43–30.86 wt.%) and ThO₂ (1.97–4.83 wt.%). The monazite (number of analyses = 29) grains selected for dating represent elongated to isometric inclusions with irregular shapes, ranging in size from 50 to 250 μm along the long axis, and often show pronounced core to rim zoning patterns for Y, Th, and U. The Th-U-Pb weighted mean age from single data points in monazite is 721 ± 20 Ma, MSWD = 0.19 (Figure 24b), and the isochron age is 728 ± 51 Ma when plotted in the ThO*₂–PbO diagram (Figure 24c). These data are generally comparable within errors with the concordant ages from the major zircon population. The CHIME ages calculated using two different approaches are in agreement within the errors of the method (≤1.5 rel.%), which is comparable with the more precise ages from isotopic dating.

5. In terms of morphological types, four varieties of zircons were distinguished for the sample 400 from Osinovka massif. The first and second are predominantly fragments of nontransparent brown crystals; the third and fourth are transparent light yellow prismatic crystals. All zircon grains have a glassy luster and corroded surfaces; in alcohol, they weakly demonstrate zonality. In the U–Pb diagram (Figure 25a), zircons of all morphological types occur at concordia. The U–Pb age obtained for zircons is 540–550 Ma and corresponds to the time of magmatic crystallization of subalkaline leucogranites of the Osinovka massif. These results agree well with the dates (555 ± 5 Ma) obtained for postcollisional granitoids from the Upper Kan massif (Kan block) [107].

Important additional information about the age of crust-forming events within the limits of the Isakovka terrane and the melt source of granites was obtained in Sm–Nd isotope studies. Given the isotope composition of leucogranites, they formed during melting of a sialic crustal source of Early Riphean (Early Mesoproterozoic) age, 1.5–1.65 Ga. Note that low-alkaline plagiogranite porphyries of the Porozhnaya massif with a U–Pb age of about 700 Ma, associated with arc volcanic rocks, were formed from a juvenile crustal source (εNd(T) = +1.6) of relatively younger, Middle Riphean age (TNd(DM)-2st = 1272 Ma) [100], and this significantly differentiates them from later postcollisional subalkaline leucogranites.

6. In cathodoluminescence, the zircon grains from the porphyroblastic plagiogneisses of the Malaya Garevka Group have light-colored long-prismatic cores with a fine sectorial zoning and dark unzoned rims. Normal Th/U ratios (0.14–0.29) in most of the cores, along with the fine zoning and melt inclusions, indicate the magmatic origin of zircon. The rims show smaller Th/U ratios (0.03–0.07) as is typical of metamorphic zircon. Eleven cores and rims plot along the concordia within a range of 860–900 Ma with an average of 881.5 ± 7.5 Ma ((Figure 25b).

7. Eleven spot analyses of the cores and rims of zircons from the rapakivi-type granites of the GC define a discordia with an upper intercept of 871.4 ± 9.7 Ma (Figure 25c)). The analysis of eight zoned concordant zircon cores and a rim yielded an identical (within error) age of 873.1 ± 7.1 Ma (Figure 25c, inset). The U–Pb data on the rapakivi-type granites show good agreement with previously published U–Pb zircon and ⁴⁰Ar–³⁹Ar hornblende ages for the blastomylonitized porphyroblastic granite-gneisses and microcline–zoisite orthoamphibolites of GC. These events were nearly synchronous with the retrograde metamorphism of migmatites.

8. Eight spot analyses of the cores and rims of zircons from the migmatites of the GC plot near the concordia at 796–878 Ma identify a mean age of 845 ± 15 Ma (Figure 25d). This age is coeval with the timing of collisional medium-pressure metamorphism of the kyanite–sillimanite type due to crustal thickening and thrusting of Siberian cratonic blocks on the Yenisey Ridge.

4.2.3. Basites

1. The zircon crystals from the metagabbro of GC (Sample 63) are generally 200–300 μm in size and euhedral elongate columnar. Most of them exhibit typical oscillatory zoning in the CL images, suggesting a magmatic origin. The Th/U ratios are greater than 0.5 (0.62–1.06), indicating that these are magmatic zircons. The U-Pb diagram shows that eleven data points from the ten zircon crystals are clustered and plot along the concordia curve (Figure 26a). The weighted mean ²⁰⁶Pb/²³⁸U age for 11 spots is 795 ± 5 Ma (MSWD = 0.18), which can be interpreted as the formation age of the mafic dike complex.

The zircon U-Pb ages for the felsic and mafic rocks in this study are generally consistent within the analytical error [5]. The mean ²⁰⁶Pb/²³⁸U zircon age for mafic rocks is slightly older (at 4–5 Ma) than the mean data obtained from ⁴⁰Ar–³⁹Ar incremental heating of the metagabbroic rocks. These results together indicate near-synchronous Middle Neoproterozoic ages of magmatism and metamorphism in a tight cluster of 787 to 797 Ma. The minor difference in the absolute ages between the emplacement of dyke rocks and their cooling below the closure temperature of the K–Ar system in hornblende can also be explained as a result of the rapid exhumation from the depths of their formation to upper crustal levels in shear zones with a duration of about 5–7 Myr [108].

The age spectra of hornblende from samples 63 and 29 show well-defined plateaus of 787 ± 8 and 794 ± 8 Ma (mean 790.5 Ma), respectively (Figure 26b,c), which are interpreted to correspond to cooling below the closure temperature of the K–Ar system in hornblende (500 °C).

2. The age spectrum of hornblende from the orthoamphibolites of the Garevka Complex has a distinct plateau of 902 ± 12.8 Ma, which is interpreted as a period of cooling below 500 °C, the closure temperature of the K–Ar system in hornblende (Figure 26d).

3. The metapicrite–basalt sequence and overlying carbonate rocks of the Gorevka Formation are considered as a single volcanic–sedimentary complex formed at the late Mesoproterozoic–Early Neoproterozoic boundary. The Pb–Pb ages of 1020 ± 20 Ma obtained for carbonate rocks of the Gorevkaa Formation of the Shirokaya Group. Based on the correlation of Riphean carbonate–terrigenous rocks of the Yenisey Ridge with sediments of the Riphean hypostratotype from other Siberian regions, the geochronological age of the Middle Riphean (Mesoproterozoic)–Upper Riphean (Neoproterozoic) boundary is determined as 1030 Ma [109]. The chronology of tectonomagmatic events in this region is as follows: formation of the Rybinsk–Panimba volcanic belt in a continental rift setting at the base of the Sukhoi Pit Group, accumulation of the metapicrite–basalt sequence, and their subsequent regional metamorphism at ~1.1 Ga [4]. Hence, the age of the picrite–basalt sequence could be close to the Meso–Neoproterozoic boundary.

4.2.4. Mafic and Felsic Metavolcanites

U–Pb dating of zircons was carried out for samples 15-07 (metadacites) and 15-14 (amygdaloidal basalts). In cathode luminescent images, zircons are characterized by a prismatic habit with a sectorial structure and normal Th/U ratios (<1), which indicates the magmatic nature of zircon. The figurative points in the nine rims and central parts of the zircon grains from sample 15-07 are located along concordia in the range of 674–713 Ma, with an average age value of 691.8 ± 8.8 Ma (Figure 27a). The results of the U–Pb dating of metadacites are close, within the accuracy limits of the method, to the dates of zircons from arc plagiogranites of the Porozhnaya massif (697.2 ± 3.6 Ma) [100] and zircons from metamorphozed gabbros of the Borisikha ophiolitic massif (682 ± 13 Ma) [58], which are parts of the Isakovka terrane.

The figurative points of eight zircon grains from sample 15-14 are arranged along concordia in the range of 556–587 Ma, with an average age value of 572.9 ± 6.5 Ma (Figure 27b), which corresponds to the time of magmatic crystallization of amygdaloidal basalts. These events were slightly older compared to the Late Vendian dates of granitoids from the Osinovka massif (540–550 Ma), located within the Isakovka terrane, and with the dates for the upper time limit of the formation of metaterrigenous-carbonate rocks of the Sayan series (Derba block, Eastern Sayan) [107].

4.2.5. Metapelites

1. Unlike xenotime, monazite appears to be stable in all garnet generations from the metapelites of the Garevka Complex, but the middle growth zone in garnet indicates that monazite is involved in a reaction relationship with apatite (Figure 8), which is a typomorphic indicator of metamorphic overprinting. The monazite grains selected for dating represent elongated to isometric inclusions with irregular shapes, ranging in size from 10 to 250 μm on the long axis, and often show pronounced core to rim zoning patterns for Y, Th, and U. The number of monazite analyses varied from 9 to 43. The calculation results reveal three populations of U-Th-Pb ages for different garnet generations. These age populations were obtained using different approaches and are generally consistent within the analytical error. The isochron ages range from 1056 ± 44 (sample 56) to 852 ± 37 Ma (sample 58) for cores (a, b), 801 ± 34 (sample 56) to 793 ± 23 Ma (sample 58) for middle zones (c, d), and 785 ± 30 (sample 56) to 776 ± 32 (sample 58) Ma for rims (e, f) (Figure 28). Therefore, the initial growth of garnet occurred at c. 1050-850 Ma, as recorded by the monazite and xenotime inclusions located in the core of the large porphyroblastic garnets. Monazites within a compositionally distinct second shell of garnet have an age in a narrow range from 801 to 793 Ma. A third garnet zone that forms the rim contains monazite with an age of 785–776 Ma.

The metapelite samples used for geochronological determinations in the Polkan area were collected at equal distances along the thrust from four localities in the outer (samples 284 and 252) and inner (samples 250 and 244) collisional metamorphic zones. The ⁴⁰Ar–³⁹Ar data for mica from different samples collected in the TC are given in Figure 11b. All age spectra for micas exhibit a distinct plateau segment at 772.9 ± 8.3 (sample 284), 782.6 ± 8.4 (sample 252), 786.8 ± 8.2 (sample 250), and 794.8 ± 8.8 MA (sample 244) (Figure 29). All ages calculated by the plateau method are interpreted to reflect cooling below the closure temperature of the K-Ar system in biotite and muscovite (330–360 °C), which suggests much lower temperatures than the peak metamorphic conditions during collision. The plateau ages were calculated using the criteria presented in [110]. It should be noted that the plateau ages for samples collected in the outermost zones of the study area (samples 284 and 244) do not overlap within the error and intermediate values (samples 252 and 250) lie in this range in accordance with the distance from the thrust.

The closure temperature of the K-Ar system corresponds to a crustal depth of 15 km for the geotherm calculated for the thickened crust (given the normal lithostatic pressure–depth distribution of 1 kbar/3.5 km for continental crustal rocks). This implies that the amount of exhumation for these metapelites buried to different depths during the post-collisional stage could not be less than 10.5 (sample 284), 6.9 (sample 252), 4.3 (sample 250), and 2.5 km (sample 244). Using the difference between the K-Ar closure ages (Δt) and depths (ΔZ) for deeply and shallowly buried blocks, which were either closest to or most distant from the thrust (present-day erosion surface), we obtain the rate of exhumation: V = ΔZ/Δt = 8.05 km/21.9 Myr = 370 m/Myr or 0.37 mm/yr. Taking into account our exhumation rate, we can then calculate the time it took metapelites to pass to the 330–360 °C isotherm: t = 10.5 km/0.368 km/Myr = 28.7 Myr for sample 284, 6.9/0.368 = 18.7 Myr for sample 252, 4.3/0.368 = 11.5 Myr for sample 250, and 2.5 km/0.368 km/Myr = 6.8 Myr for sample 244. Collisional metamorphism in these metapelites preceded uplift and cooling, as evidenced by the ⁴⁰Ar–³⁹Ar results. The age of the peak of metamorphism is then the sum of these values and the K-Ar closure age of mica. These ages for samples 284, 252, 250, and 244 are no older than 801.4, 801.3, 798.3, and 801.5 Ma, respectively. These results altogether indicate Middle Neoproterozoic ages in a tight cluster of 798 to 802 Ma.

2. The cathodoluminescence images of zircons from regional LP/HT metapelitic rocks of TC reveal the presence of long- and short-prismatic cores with a thin sector zoning and dark unzoned rims (Figure 30a). The zircon cores typically have low U (75–321 ppm) and Th (46–175 ppm) contents and normal Th/U ratios in the range 0.26–0.64, which, along with their morphology and internal structure, suggest a magmatic origin. The zircon rims differ significantly in their Th–U chemistry from the magmatic cores. They are characterized by marked enrichment in U (680–1608 ppm), depletion in Th (5–11 ppm), and low Th/U (0.01) typical of metamorphic zircons. Two concordant zoned zircon cores (spots 2.1 and 5.1) yielded an age of 1.94-1.86 Ga (Figure 29a), which can be interpreted as an approximate age for the protolith of the Teya Group. The source of these rocks could be the granitic gneisses of the Siberian craton metamorphosed to granulite facies in the Early Proterozoic (~1.9 Ga) [111]. Three weakly zoned zircon cores (spots 1.1, 4.1, and 7.1) yielded ages of 2.77–2.49 Ga, which agree well with the Late Archean (2.8-2.5 Ga) U–Pb age of the protolith of Al-rich orthogneisses (Kan and Yenisey Groups, Yenisey Ridge) that underwent granulite metamorphism at 1.9 Ga [112]. Two spots in the metamorphic zircon rims (2.2 and 5.2) plot along the concordia between 973 and 953 Ma (Figure 29a), which allow us to consider this interval as the age of LP/HT regional metamorphism.

These results are in close agreement with data obtained from ⁴⁰Ar–³⁹Ar incremental heating experiments (Figure 30b). The age spectrum of biotite from the same metapelite (Sample 155) has a well-defined plateau of 948 ± 21 Ma, which is interpreted to date the cooling below the closure temperature of the K–Ar system in biotite (330–360 °C). This temperature corresponds to a crustal depth of 14–15 km for a lithostatic pressure–depth profile of 1 kbar/3.5 km. A record of the thermal history of this rock formed at P = 4.8 kbar corresponding to a depth of about 17 km can be used to determine the age of metamorphism. This implies that the amount of exhumation for these metapelites could not be less than 2–3 km at the postmetamorphic stage. Given the maximum duration of exhumation, this age is certainly no older than 956 ± 21 Ma, which agrees within the error with the U–Pb zircon ages. These results are in good agreement with previous U–Th–Pb and Rb–Sr data for the Teya granite-gneiss dome (1000–900 Ma) and Ar–Ar ages for hornblende (955 ± 18 Ma) from metabasites interbedded with the Teya schists, which were obtained by other dating methods [56].

The ³⁹Ar–⁴⁰Ar analysis of biotite from sample 169 collected at the contact with the Kalama granitoids yielded a plateau age of 857 ± 7 Ma (Figure 30c). Taking the calculated exhumation rate into account, these data suggest that the time of the contact metamorphic event cannot be older than 862 Ma, which is consistent with the U–Pb age constraints on the emplacement of the Kalama granitoids.

The ³⁹Ar–⁴⁰Ar data for biotite from different samples collected from Angara, Chapa and Mayakon areas in the TC are given in Figure 31a–d. The age spectra of biotite from the metapelites collected in the Teya area exhibit well-defined plateaus of 826.1 ± 8.4 and 829.3 ± 8.4 Ma (Figure 31e,f), which may record the cooling below the closure temperature of the K–Ar system in biotite. These biotite ages agree well within the error, which testifies to the apparent synchrony of exhumation and cooling of rocks that underwent collisional metamorphism. The interpretation of the above model of crustal thickening due to thrusting demonstrates that the amount of exhumation of the metapelites subducted to different depths (25.0 km for sample 163, P = 7.1 kbar and 22.4 km for sample 161, P = 6.4 kbar) could not be less than 10.0 and 7.4 km, respectively, at the postcollisional stage. For the previously calculated exhumation rate, the time during which the metapelites (samples 163 and 161) reached the 330 °C isotherm was determined to be 27 and 20 Myr, respectively. The Upper Riphean age of peak collisional metamorphism calculated from the sum of biotite ages is no older than 853–849 Ma, which agrees with the ages of Ky–Sil metamorphism (851–863 Ma) obtained for other thrust areas of TC of the Transangarian Yenisey Ridge.

The age of the proposed eroded material that produced the protolith of the metapelites was evaluated by U–Pb zircon dating of aluminous granite gneisses sample 106 collected from an outcrop in the Nemtikha Group rocks. The zircons were transparent prismatic crystals of hyacinth habit 80–200 μm in size. Their cathodoluminescence images exhibit long-prismatic cores with sectorial zoning and outer rims with weakly pronounced zoning. The discordia plot based on seven data points for the cores and margins of the zircons has an upper intercept with the concordia corresponding to an age of 2043 ± 8.1 Ma (Figure 32). A somewhat younger age of 1962 ± 21 Ma was obtained on three zonal concordant cores and a margin (Figure 32, inset).

5. Discussion

5.1. Tectonic Settings and Nature of Protolith

Petrological, geochemical, and geochronological studies of rocks from the Transangarian Yenisey Ridge were used to identify the nature of tectonics, metamorphism, and magmatism at the western margin of the Siberian Craton and to constrain the timing of the formation of different structures. Identifying the composition of the protolith is based on evidence of the predominantly isochemical character of metamorphism with respect to the main petrogenetic components and on studies that show very limited differential mobility of elements [113].

5.1.1. Metabasite Tectonites

In general, the studied metabasites have weakly fractionated multielement spectra occupying an intermediate position between the spectra of N- and E-MORB basalts (Figure 14). According to their geochemical features, these rocks are comparable with the metabasites of the Rybinsk-Panimba (R-P) volcanic belt in the Angara region [4]. This distinguishes them from the older gabbroic rocks of the dike belt, picrite-basalts of the Angara region and the amphibolites of the GC, which are related to within-plate basalts and predominantly island-arc tholeiites (Figure 33).

The nature of the rare-earth spectra, the different values of (La/Sm)_n and (La/Yb)_n ratios, and Hf, Zr, Nb, and Ta contents reflect the different degree of depletion of the composition of mantle sources or the degree of melting. The protoliths of the low-titanium metabasites of the Isakovka terrane with reduced Nb/Y and Zr/Y ratios could have formed by melting the mantle source depleted by incompatible elements and compared with the geochemical characteristics of the upper mantle producing N-MORB basalts. The relatively high Nb/Y and Zr/Y ratios for higher-titanium rocks indicate the enriched nature of their mantle components responsible for the formation of predominantly E-MORB basalts. The protoliths of the high-pressure tectonites correspond to N-MORB and E-MORB basalts. Formation of the more primitive N-MORB components occurred at the initial stages of spreading at 701.6 ± 8.4 Ma, through melting of the upper horizons of the depleted mantle. The age of blueschist facies metamorphism corresponds to an interval from 640 to 620 Ma, corresponding to the time of the stages of the paleo-ocean subduction along the western margin of the Siberian Craton [12]. During exhumation, the glaucophane-bearing rocks were brought up into the Yenisey shear zone, where they underwent intense deformation with recrystallization and the formation of new high-pressure mineral parageneses. The formation of high-pressure tectonites in the suture zone at around 600 Ma marks the final stages of accretion of the Isakovka block to the western margin of the Siberian Craton [10].

5.1.2. Granitoids

The primitive mantle-normalized multi-element diagram and chondrite-normalized REE patterns (Figure 16), and the high contents of alkalis, mostly HFSE, indicate that the studied granitoids plagiogneiss granites from the Nemtikha Group, granites from the dike belt, granites of Ostrovok and from Chernorechensky massif, and rapakivi-type granites of the Garevka Complex are geochemically similar to A-type granites. These features and their high Ba contents (up to 1220 ppm) are broadly similar to those of Proterozoic A-type granites [114]. On the Nb–Y–Zr/4 discrimination diagram [115] for A-type granites, all of these rocks are plotted in the field of crust-derived granites (Figure 34d). Their anorogenic nature is also further corroborated by plots of the granitoids mostly clustering in the intraplate-type A-granite field on the discrimination diagrams FeO*/MgO–Zr + Nb + Ce + Y, (K₂O + Na₂O)/CaO–Zr + Nb + Ce + Y [116], Rb–Hf–Ta [117], Rb–Y + Nb, and Nb–Y [118] (Figure 33), indicating their formation in an intracontinental extension setting. We assume that this extension could have been caused by a rising mantle plume.

The geochemical features of rocks from the Osinovka massif show that they correspond to intermediate characteristics between fractionated and nonfractionated S- and M-type granites in the (FeO*/MgO–Zr + Nb + Ce + Y, (K₂O + Na₂O)/CaO–Zr + Nb + Ce + Y) classification diagrams [116] (Figure 34a,b). The domains of their compositions overlap with migmatites and granitoids of potassic series from the Garevka Complex, in contrast to rapakivi-type granites, whose figurative points are located in the zone of A-type granites [46]. However, in general, leucogranites of the Osinovka massif show depletion in Nb and Ta (Figure 17). This was inherited from their crustal source because the continental crust is characterized by the presence of a Ta–Nb minimum in the multielement spectrum [119]. The abrupt depletion in Eu, Sr, and Ti can be related to fractional crystallization of plagioclase and ilmenite. In the Rb–Hf–Ta [117] and Rb–Y + Nb [118] (Figure 34c,e) discriminant diagrams, the figurative points of the studied rocks fall within the boundary zones between arc and intraplate granites (Figure 34c,e). Intrusive rocks with these geochemical characteristics usually form in postcollisional settings, and this is supported by the location of figurative points of granitoids in the field of significantly postcollisional granites.

A common property of all varieties of studied granites is that the magmas they originated from were separated from the continental crust, as is illustrated by the Nb–Y–Zr/4 diagram [115] (Figure 34d). Given the Sm–Nd isotope composition of the studied leucogranites, they formed during melting of a sialic crustal source (εNd(T) ranges from –2.5 to –4.4) of Early Riphean (Early Mesoproterozoic) age (TNd(DM)-2st = 1491–1648 Ma). Note that some granites from the near-contact zone have clear signatures of assimilation by mafic (basic) material of the host rocks because they are characterized by the least values of the model age and higher εNd values [10].

A distinctive feature of the GM rocks is the overlapping of high-potassic A-type and S-type granitoids on the discrimination diagram of Whalen [116]. The migmatites plot exclusively in the S-type field, as opposed to the rapakivi-type granites and gneissic plagiogranites of the Na-series, which fall into the A-type granite field (Figure 34a,b). The main types of potassic rocks (porphyroblastic granite-gneisses) have geochemical traits intermediate between within-plate and orogenic granites and fall within the post-orogenic granite field, whereas all rapakivi-type granites and some Na-series granitoids plot in the field of A-type within-plate and anorogenic granites on most discrimination diagrams, e.g., Rb–Hf–Ta [117], Rb–Y + Nb, and Nb–Y [118] (Figure 34).

This indicates that the GC rocks were derived from several different sources and, possibly, under differentiation conditions. This is also consistent with variations in their Y/Nb ratio, which reflects the composition of the source. One common feature of all the high-potassic granites of GC is their origin from magmas derived from continental or underplating crustal sources, as suggested by the Nb–Y–Zr plot [115]. It is noteworthy that, based on the above petrogeochemical characteristics, all rapakivi-type granites of GC are similar to the typical rapakivi granites (vyborgites) of the Baltic shield [120].

In general, the A-type granites investigated in this study comprise a broad compositional spectrum of granitoids that crystallized from relatively high-temperature and anhydrous (water unsaturated) magmas with high concentrations of alkalis and mostly REE and HFSE. Their characteristic features also include the high-Fe content of silicates (garnet with X_Fe = 0.99; biotite with X_Fe = 0.87), the presence of ilmenite as the only Fe–Ti oxide, and the low values of LOIs, all of which offer evidence for the formation of these rocks from water-poor melts under reduced conditions below the FMQ buffer [5]. Melts derived from the basement gneisses of the Siberian Craton may be considered as a potential source for these least oxidized rock types. They are generally post-orogenic and were formed mostly in hypabyssal environments during lithospheric extension and were not directly related to lithospheric convergence. Thus, we suggest that the emplacement of these A-granites took place in an intracontinental extensional setting, caused by a single rising mantle plume.

The nature of extensional tectonics and magmatism within the study area are also in good agreement with the metamorphic P–T trajectory calculated from samples of adjacent (gneissic) host rocks [49]. The observed counter-clockwise P–T path usually reflects an overall tectonic process involving initial crustal compression immediately followed by extension, and this process is normally accompanied by intrusions of deep-derived magma or thinning of the lithospheric mantle [60].

5.1.3. Basites

Multielement patterns of the metagabbros from the dike belt are characterized by the absence of Nb and Ta depletion with respect to Th and LREE (Figure 18b), which is typical of intraplate basalts, including continental rift basalts showing distinct signatures of deep mantle sources and plumes [121,122]. Based on their intraplate affinity confirmed by plots in the Zr–Nb–Y [123] and Hf–Th–Ta [124] diagrams, these rocks differ from the other GC gabbroic rocks with island-arc tholeiite affinity (Figure 33).

In general, the rocks of the metapicrite–basalt sequence of the Angara region have well-fractionated multielement patterns that are transitional between OIB and E-MORB and show no depletion in Nb and Ta relative to Th and LREE (Figure 19), as typical of within-plate basalts, in particular, basalts of continental rift zones [122], whose relation with deep-seated mantle sources and plumes is substantiated. The geodynamic affiliation of these rocks to within-plate basalts, mainly to WPAB and WPTB, is confirmed by classification based on Zr–Nb–Y [123] and Hf–Th–Ta [124] diagrams (Figure 33). This makes them different from other mafic rocks of the Transangarian region: metabasites of the Rybinsk–Panimba volcanic belt [4] and metavolcanic rocks of the Isakovka terrane [125]. The rift nature of the volcanic rocks is also confirmed by the Zr/Y–Nb/Y diagram [126] (Figure 35), where all data points of the metapicrite–basalt sequence fall in the composition field of within-plate basalts between E–MORB and OIB. The metapicrite–basalt sequence and overlying carbonate rocks of the Gorevskaya Formation are considered as a single volcanic–sedimentary complex formed at the late Mesoproterozoic–Early Neoproterozoic boundary in the rift-related marginal–continental paleobasin on the southwestern Siberian Craton.

5.1.4. Mafic and Felsic Metavolcanites

Substantial differences in the composition of metadacites are observed when they are compared with subalkaline leucocratic postcollisional granites from the Osinovka massif: the melt source of these rocks was highly differentiated continental crust from the western margin of the Siberian Craton. Metadacites are characteristic of much lower concentrations of radioactive (U, Th, and K), high-charge (Nb, Ta), and rare-earth elements (total REE is 130 g/t versus 200–216 g/t in granites) (Figure 17). In the Rb–Hf–Ta and Nb–Y discriminant diagrams, these differences are caused by the locations of figurative points of metadacites and leucogranites: the former are plotted among arc granites, and the latter, among postcollisional and intraplate granites (Figure 34).

Amygdaloidal basalts have very well-fractionated multielement distributions located near the spectral zones of oceanic arc basalts typical of rifting structures. In certain petrochemical and geochemical parameters, these rocks clearly differ from metabasites of the Rybinsk–Panimba volcanic belt and the Isakovka terrane, which pertain to groups of normal and enriched basalts. This has been verified by diagnostic diagrams based on the relationships of the Zr–Nb–Y [123], TiO₂–MnO–P₂O₅ [127] (Figure 33), P₂O₅–TiO₂ [128], and Zr/Y–Nb/Y ratios [126] (Figure 35), where the points corresponding to the compositions of amygdaloidal metabasites are plotted predominantly among intraplate basalts, E-MORB and OIB, or volcanic rocks of oceanic arcs. In contrast to classical arc basalts, they differ by lower contents of Al₂O₃ (14.8 wt %), Ba (250 ppm), and Sr (455 ppm) but higher TiO₂ and P₂O₅ contents. However, it is possible that amygdaloidal basalts also inherited some geochemical features of the hosting arc rocks of the Isakovka terrane, as reflected in the Hf–Th–Ta diagram [124] (Figure 33). The obtained data, together with the discovered large gap in time (~120 Ma) between the formation of arc volcanic rocks and ocean-floor basalts, and with the fresh appearance of rocks almost completely unaltered by metamorphism, point to a relationship of the studied amygdaloidal basalts with the post-accretionary stage of crustal extension or rifting.

5.1.5. Metapelites of the GC and TC

Specific features for metapelites are generally similar those of the PAAS (post-Archean Australian schists) [27] and can be attributed to the presence of erosion products of felsic igneous rocks in the detritus. The inheritance of the primary magmatic composition is confirmed by a high positive linear correlation between the contents of HFSEs, such as Zr, Hf, Y, Ta, and Nb. The erosion of felsic igneous rocks is also inferred by the higher Th/U (6–7.5) and lower La/Th (0.14–0.2) ratios relative to average PAAS compositions.

We used a system of petrochemical modules and genetic diagrams [129,130] with the well-known ratios: Chemical Index of Alteration (CIA = [A1₂O₃/(A1₂O₃ + CaO + Na₂O + K₂O)]·100), [131]; Chemical Index of Weathering (CIW = [A1₂O₃/(A1₂O₃ + CaO +Na₂O)]·100), [132]; Index of Compositional Variability (ICV = (Fe₂O₃ + K₂O + Na₂O + CaO + MgO + TiO₂)/Al₂O₃), [133]; and Plagioclase Index of Alteration (PIA = [(A1₂O₃ − K₂O)/(Al₂O₃ + CaO + Na₂O + K₂O)]·100), [134]. Metapelites of the GC and TC are characterized by the following values of specific geochemical features: the hydrolyzate module, HM = (Al₂O₃ + TiO₂ + ΣFe₂O₃ + MnO)/SiO₂, ranges from 0.51 to 0.54; the iron module, IM = (ΣFe₂O₃ + MnO)/(TiO₂ + Al₂O₃), varies from 0.32 to 0.39; the Al-Si module, ASM = Al₂O₃/SiO₂, ranges from 0.36 to 0.38; and the femic module, FM = (ΣFe₂O₃ + MnO + MgO)/SiO₂, varies from 0.15 to 0.19, suggesting that they are an association of normal siallites and supersiallites (clay rocks) according to the classification of Yudovich and Ketris [129]. The alkalinity module, AM = Na₂O/K₂O, varies from 0.10 to 0.33 while the potassium module, PM = K₂O/Al₂O₃, ranges from 0.10 to 0.15. These two values suggest that the primary sediments were enriched in hydromica and chlorite. However, in the FM-NAM [129] and FAK [130] diagrams, these rocks fall within the compositional field of kaolinite (Figure 36a), which may have been caused by paleogeographic conditions during sedimentation. The quartz (24–27 vol.%)–illite (19–28%)–montmorillonite (8–15%)–kaolinite (19–31) normative composition of metapelites calculated by the MINLITH program [135] corresponds to mature sedimentary rocks (pelites and subgraywacke pelites) (Figure 36b). Kaolinite clays may have accumulated near the continental-margin zone while the more fine-grained clay material with a chlorite-hydromica composition was deposited in offshore marine basins. The lower alkalinity (NAM = 0.12–0.17; where NAM stands for the normalized alkalinity module, i.e., NAM = (Na₂O + K₂O)/(Al₂O₃) [129]), along with the low MgO values (<2.45 wt%) and higher K₂O contents (>2.28 wt%), results from the presence of K-rich granitoids in the detrital material of erosion products and is indicative of no admixture of basic volcanic rocks. The values of the titanic module (TM = TiO₂/Al₂O₃ = 0.03–0.05) are typical of initial sediment accumulation in shallow nearshore basins under humid climate conditions, thus closely matching the lithofacies data [136,137].

The plots in the FM-NAM and FAK diagrams, and the high CIA, CIW, and PIA values (82.9–95.7) and low ICV values (<1), indicate that these sediments were derived from an intensely weathered source area of the studied outcrops. The ferruginous-aluminous metapelites primarily represent redeposited and metamorphosed products of Precambrian kaolinitic, rather than lateritic, weathering crusts. In the Yenisey Ridge, the chemical weathering of rocks during the Paleo-Mesoproterozoic did not reach the lateritization stage, with the final formation of decomposition products of aluminosilicates. Weathering was mainly limited by the predominant formation of kaolinite-chlorite-hydromica rocks.

Such features are typical of post-Archean clay shales and are caused by the fact that the detrital material contained erosion products of granitoids, whose generation was associated with Eu²⁺ depletion during the sedimentation of residual plagioclase. The inheritance of the primary composition of the magmatic protolith also follows from the strong positive correlation between the concentrations of HFSE (Zr, Hf, Y, Ta, and Nb). The erosion of acid rocks also follows from the elevated values of Th/Sc = 0.69–1.33 and Th/U = 8–12.8 and from the lower values of La/Th = 1.40–3.36 and Co/Th = 0.17–0.18 than in the average PAAS composition, which is also confirmed by the occurrence of granite erosion products in the basal units of the studied metapelites. In the Eu/Eu*–(Gd/Yb)_n (Figure 37a) and La–Th (Figure 37b) diagrams, the composition data points of the metapelites are plotted mostly within the field of post-Archean cratonic deposits. An analogous conclusion follows from the arrangement of the compositional data points of these rocks in the (La/Yb)_n–Yb_n diagram, in which these points swarm within the field of post-Archean granitoids enriched in HREE and depleted in Co and Ni compared to Archean acid magmatic rocks (Figure 37c). This conclusion is in agreement with the geological history of the Transangarian Yenisey Ridge in the Precambrian: the Neoproterozoic stage of its development was preceded by the continental regime, with peneplanation and weathering crust formation [56,137]. The Early–Late Precambrian boundary corresponded to the subplatform regime, with the accumulation of Fe- and Al-rich deposits of Teya, Sukhoy Pit, and Korda Groups in shallow-water basins, which were presumably derived from the erosion products of Paleoproterozoic (c. 2000 Ma) granite-gneisses of Siberian Craton [138,139]. Detailed reconstructions of the composition and nature of their protolith based on whole-rock major and trace element geochemistry have demonstrated that these rocks represent a re-deposited and metamorphosed product of Precambrian kaolinite weathering crusts. The formation of the protolith of these rocks is attributed to the erosion of the Lower Proterozoic granite gneisses of the Siberian Craton with ages in the range 1962–2043 Ma (U-Pb SHRIMP by zircons).

5.2. Tectonic Implications

The state of our understanding of Precambrian supercontinent reconstructions is very much at a preliminary stage, with large uncertainties in the positioning of most blocks in the various proposed supercontinents [140,141]. The matching of ages of geological events between crustal blocks is proving to be a useful new tool for constraining Precambrian supercontinent reconstructions. The results presented in this study provide a reliable geochronological base to evaluate the different stages of their evolution along the western Siberian margin. Our focus in the present paper is not global reconstructions; however, we discuss the specific implications of the magmatic and metamorphic events profiled herein for our understanding of the position of the Siberian Craton in the Columbia (Nuna) and Rodinia supercontinents. Time–space plots for various fragments of the crust of the western margin of the Siberian Craton for the period 1.38-0.56 Ga are shown in Figure 38.

Age Correlations between Siberia–Laurentia–Baltica and Tectonic Interpretations

• Mesoproterozoic stage

The age of the Mezoproterozoic event suggests that the extensional regime is close to that of rifting and within-plate magmatism in the North Yenisey Ridge. The time interval obtained for the Nemtikha Group rocks (1380–1360 Ma) could mark the onset of the final stage of Mesoproterozoic rifting along the western margin of the Siberian Craton and related eruptions of picrobasaltic-basaltic tuffs and lavas, and intrusion of the associated gabbrodolerite and picritoid subvolcanic bodies of the Rybinsk-Panimba volcanic belt [72]. This age range (1380-1360 Ma) also coincides with the age of rapakivi-like granites from Garevka Complex of the Transangaria, and for other rocks that form large igneous provinces worldwide [6]. Recent data highlight the widespread nature of 1380–1360 Ma magmatic events into the other continental blocks of the Columbia (Nuna) supercontinent [6,142]. It is noteworthy that a broadly similar age interval (Mashak event that may represent the 20 Myr rift phase) obtained by U-Pb dating of rapakivi granites from the Berdyaush massif and rhyolite-dacites of the Mashak Formation was proposed as the Early–Middle Riphean boundary within the Bashkirian mega-anticlinorium of the Urals [143,144]. Magmatism and possible concomitant rifting at the Lower–Middle Riphean boundary along the western margin of the Siberian craton were coeval with mafic dykes and sills in western Laurentia (Victoria Land dykes, Hart and Salmon River Arch sills) and North Greenland (Midsommerso sills and Zig–Zag volcanics), Anabar Shield (Chieress dyke), and eastern Siberian craton (Sette–Daban sills) [145,146,147].

A slightly younger age for mafic magmatism in Siberia was revealed by the ages of two Mesoproterozoic Listvyanka (1350 ± 6 Ma) and Goloustnaya (1338 ± 3 Ma) mafic dyke swarms located in the Irkutsk promontory of the southern part of the Siberian craton [148], and this is synchronous with the 1353 ± 2 Ma Barking Dog gabbro sill from the Wellington Inlier of Victoria Island, northern Canada [149].

We therefore believe that our new age data add to the areal extent of the 1380-1360 Ma magmatism and confirms that it is a major event that reflects the extensional regime in the Yenisey Ridge probably associated with its breakup from other blocks [142]. However, this is not direct evidence for the time of formation of a continental margin and new oceanic crust. Such processes during the Proterozoic history of this region are genetically and spatially related to the origin of the Paleo-Asian Ocean [150], which formed between the North American and Siberian cratons, resulting from the breakup of the supercontinent of Rodinia [151]. The most ancient structures are represented by fragments of the oceanic crust and island arcs in the Isakovka terrane (700–620 Ma in age). However, we cannot rule out the possibility that some ophiolites are older. Therefore, the present data may admit a link between these magmatic events at ca. 1380–1350 Ma and breakup of the proposed late–Paleoproterozoic Columbia (Nuna) supercontinent along the western margin of the Siberian craton, but do not prove the link.

2.

Late Meso–Early Neoproterozoic stage

The early stage occurred as a result of the Latest Grenville orogeny during late Meso–early Neoproterozoic (1050-850 Ma) and was marked by low-pressure zoned metamorphism, with a metamorphic field gradient of dT/dZ = 25–35 °C/km typical of orogenic belts. The link between these processes and the Grenville-age orogeny was strongly supported by the earlier U-Th-Pb, Rb-Sr, and K-Ar data on granite-gneiss domes (1100-950 Ma) and the more recent single-zircon (U-Pb SHRIMP II) and ⁴⁰Ar–³⁹Ar dating of metapelites, metabasites, and rapakivi granites from the TC and GC (1140-870 Ma). These processes occurred at approximately the same time in several other lithospheric blocks of the Asian continent [152,153]. During the second stage in middle Neoproterozoic (850-800 Ma), the regionally metamorphosed rocks experienced collision-related medium-pressure metamorphism with a local pressure increase near the thrust faults and only minor heating, suggesting a low metamorphic field gradient of dT/dZ = 7–14 °C/km. The final stage evolved as a synexhumation retrograde metamorphism (790-780 Ma) with dT/dZ ≤ 15 °C/km and recorded uplift of the rocks to upper crustal levels in shear zones.

A similar style of tectonic processes involving deformation and metamorphism with broadly similar thermodynamic regimes and metamorphic field gradients, which reflect the contrasting tectonic settings during burial and exhumation in shear zones together with igneous activity, some of which show calc-alkaline and intraplate affinity, have been recorded in rocks of the Valhalla orogen. According to Cawood et al. [154,155], this orogen developed on the north-eastern Laurentian substrate around the North Atlantic borderlands and is currently exposed in Scotland, Greenland, Svalbard, and Norway, prior to the final late Neoproterozoic rifting and continental dispersal. These processes associated with the Valhalla (Knoydartian) orogenic event are assumed to have been attributed to the far-field effects of accretion around the periphery of the Rodinia supercontinent [155]. The timing of the 800-720 Ma metamorphism and magmatism in the Yenisey Ridge is similar to the ages recorded elsewhere in the North Atlantic, indicating the existence of a large-scale and long-lived, early to mid-Neoproterozoic orogenic system that formed after the Grenville orogeny and assembly of Rodinia [156]. In particular, several coeval post-Grenville metamorphic events have also been reported from Laurentian metasedimentary sequences in East Greenland and the Moine Supergroup of NW Scotland [157], the Soroy succession of northern Norway (Baltica) [158], and Svalbard [159]. Therefore, the sequence of tectonothermal events in the Valhalla orogen, which are believed to have occurred nearly synchronously along the margins of large Precambrian cratons such as Laurentia, Baltica, Siberia, and the Svalbard microcraton, suggests that they may have been in proximity to one another at that time.

3.

Middle- and late Neoproterozoic stage

Evidence for mid–Neoproterozoic rifting and related ~800 Ma bimodal magmatism during this stage, which were possibly associated with mantle plume activity and the breakup of Rodinia, were recorded in large igneous provinces from other regions of the world [160]. Such processes responsible for the destruction of the lithosphere are recognized by dyke swarms and intraplate volcanic provinces [161]. There are two well-documented LIPs in Laurentia that fall within the time interval from 800 to 720 Ma. An older event of 780 Ma for the Gunbarrel is marked by the emplacement dyke swarms in western Laurentia (~780 Ma) [162,163]. The younger, c. 725 Ma Franklin–Irkutsk plume correlates with the Franklin LIP of northern Laurentia and the Irkutsk LIP of southern Siberia [148]. The 725-715 Ma Franklin LIP extends over an area of >3 M km² and consists of units such as the Coronation gabbro sills, Minto Inlier basalts and sills, and Mount Harper volcanics in northern Canada and equivalent units in reconstructed northwestern Greenland [148,164]. Most prominently, the radiating Franklin dolerite dyke swarm converges towards the northern margin of Laurentia (north of Banks Island), marking a probable mantle plume centre and potential Neoproterozoic breakup margin. The Irkutsk LIP of southern Siberia consists of dunite–peridotite–pyroxenite–gabbro complexes: Dovyren intrusive massif (724 ± 3 Ma) and associated volcanic rocks from the Inyaptuk Formation (729 ± 14 Ma) [165], Verkhnii Kingash (726 ± 18 Ma), and the Tartai intrusion (713 ± 6 Ma) [166].

Intrusive complexes with broadly similar ages (800–740 Ma) and geodynamic settings are also reported from the Sayan–Baikal dyke belt in the southern part of the Siberian craton [21,145]. Hundreds of mafic dykes here are traditionally attributed to the Nersa Complex. The ages decrease from 787 ± 21 Ma in the Kocheriki area (northern Baikal) through 758 ± 4 and 743 ± 47 Ma in the Kitoi River area to 741 ± 2 Ma in the Biryusa area [146].

Rifting and intraplate magmatism of this age range (800–720 Ma) produced rhyolites of the rhyolite–basaltic association (753 ± 6 Ma), and subalkaline leucogranites of the Verkhnekutukas and Khariusinsk (753 ± 4 Ma), Garevka (752 ± 3 Ma), and Lendakh (749 ± 5 Ma) massifs [167] and related A-type granites of Ostrovok in the Transangarian Yenisey Ridge. The youngest age of the Chernorechensky granites coincides with intraplate granitoids from the Glushikhinskii (731 ± 5 Ma) and Strelkovskii (719 ± 5 Ma) massifs along the western margin of the Siberian Craton [167]. A similar style of tectonic processes, which reflect the contrasting tectonic settings together with igneous activity, some of which show calc-alkaline and intraplate affinity, has been recorded in rocks of the Valhalla orogeny [151,155]. The features recorded from the rocks in this study, and the spatial proximity of Laurentia and Siberia permit us to assume that the Baikal–Yenisey dyke belt was one of the peripheral branches of a hypothetical giant dyke swarm possibly related to a mantle plume centered in northern Laurentia.

Therefore, the multiple rifting and within-plate magmatism at ~800–720 Ma in the southwestern margin of Siberian Craton can be tentatively correlated to the above-mentioned plume activity with two peaks. These events may be responsible for the early stages of the Siberian separation from Rodinia and the initiation of the breakup of Siberia and Laurentia. Since the rifting event in the South China and Australia continents also occurred at ca. 800 Ma [168,169], it can be reasonably inferred that the breakup of the Rodinia supercontinent started at ca. 800 Ma globally.

The chronological sequence of development of the Transangarian Yenisey Ridge through five main phases: 1.4–1.1, 1.1–0.9, 0.9–0.85, 0.8, and 0.79–0.54 Ga, is shown in Figure 39. Our results clearly define the evolutionary trend of the lithosphere in the study area, which appears to have been controlled by changes in the geodynamic settings (rifting of the continental plates—formation of an oceanic realm—oceanic-continental plate convergence—orogeny and formation of the new continental crust). These results also provide support for the hypothesis of supercontinent cycles [170], which begin with the opening of an ocean basin preceded by a rifting stage and operated on a larger time scale of 400–650 Ma as compared to the typical Wilson cycles (ca. 200 Ma).

The Siberian Craton is usually assumed to have been part of the Mesoproterozoic [171] and Neoproterozoic supercontinents [172], but its Precambrian position in Nuna-Columbia and Rodinia reconstructions is highly controversial [173,174]. The occurrence of Siberian equivalents of the Middle Proterozoic and Late Proterozoic events, coupled with previous evidence of the Grenville-age orogenic events in the Yenisey Ridge [56], provide an important test for paleocontinental reconstructions.