—The paper presents data on the composition and age of mafic rocks of the shoshonitic series in the Irkut block of the Sharyzhalgai uplift (southwest of the Siberian Сraton). According to the U–Pb dating of magmatic zircon, the formation of monzodiorites of the Poludennyi massif and gabbro-dolerites in the endo- and exocontact zones of the Toisuk pluton occurred at 1.87 and 1.86–1.85 Ga, respectively. The intrusion of mafic magmas and their underplating into the basement of the crust under postcollisional extension resulted in the near-coeval mafic and granitoid magmatism in the Irkut block between 1.87 and 1.84 Ga. The Paleoproterozoic mafic associations belong to the shoshonitic series and are characterized by enrichment in incompatible elements, including Zr, and low negative εNd(T) values. These geochemical and isotopic characteristics point to the derivation of magma from a long-lived enriched-mantle source, such as the subcontinental lithospheric mantle. The crystallization of zircon from the last portions of the evolved mafic melt is evidenced by low zirconium saturation temperatures (710–965 °C) and zircon enrichment in U and Th with increasing Th/U, reflecting the growth of concentrations of highly incompatible elements in the residual melt.

The final amalgamation of Archean terranes into the Siberian Craton as a result of orogenic (collisional) processes occurred in the Paleoproterozoic (Donskaya, 2020). The main markers of collisional orogeny are various Paleoproterozoic granitoids located in all exposed shields, including basement uplifts along the modern southern margin of the Siberian Platform, where their formation probably marks the craton collision with other continental blocks and assembly into the Paleoproterozoic Columbia supercontinent. A detailed analysis showed that granitoid and mafic magmatism in the South Siberian Belt occurred in the range of 1.88–1.84 Ga (Donskaya, 2020). In the western part of the belt, within the Sharyzhalgai and Biryusa uplifts, collisional granitoid magmatism corresponds to the interval of 1.87–1.84 Ga (Turkina and Kapitonov, 2019, and references therein). The heat source of widespread Paleoproterozoic granite formation is still being discussed. The key issue is the role of mantle heat and development of near-coeval mafic magmatism accompanied by underplating of mantle melts into the crustal basement. In the Sharyzhalgai uplift, data on Paleoproterozoic mafic magmatism were limited to the 1864 Ma gabbrodolerites of the Kitoi dike swarm (Gladkochub et al., 2013), the 1863 Ma gabbro of the Malyi Zadoi peridotite–gabbro intrusion (Mekhonoshin et al., 2016), and small veins of lamprophyre-like rocks from the Kitoi River area (Ivanov et al., 2019); the scales of these mafic bodies are not comparable with the volume of granitoid intrusions. Thus, the key issue remains the identification of Paleoproterozoic mafic magmatic complexes in the Sharyzhalgai uplift using geological, structural, and isotope-geochronological methods.

For dating mafic rocks, zircon, along with baddeleyite, serves as the main mineral geochronometer. At the same time, for mafic rocks, the question of magmatic vs. xenogenic origin of zircons remains debatable. In the case of zircon crystallization from mafic magma, it is important to understand what conditions lead to the Zr saturation of mafic melt, including the influence of the melt composition. Alternatively, zircon in mafic rocks may be trapped from crustal rocks in an intermediate magmatic chamber or at the level of intrusion/extrusion formation (i.e., be inherited/xenogenic).

This paper presents the results of dating zircon from two mafic complexes in the Sharyzhalgai uplift: gabbro-dolerites from small bodies and mafic inclusions in the marginal zone of the Toisuk pluton and monzodiorites of the Poludennyi massif. Characteristics of these rocks are given to reveal the relationship between mafic and granitoid magmatism and clarify the causes and conditions of crystallization of zircon from parental mafic magmas.

The Irkut block in the southeast of the Sharyzhalgai uplift (Fig. 1) is composed of a dominant association of mafic and felsic granulites, which were formed by metamorphism of magmatic protoliths in the Neoarchean at 2.70–2.66 Ga, subordinate high-alumina paragneisses, the deposition of which is constrained at ≤2.75 Ga, and Paleoproterozoic (1.95–1.84 Ga) paragneisses, marbles, and calciphyres as well as rare relics of Paleoarchean (ca. 3.4 Ga) granulites of intermediate composition (Turkina, 2022). All the Archean magmatic and sedimentary rocks underwent high-temperature metamorphism and were injected by veins of synorogenic granitoids of different scales at 2.55–2.54 Ga (Poller et al., 2005; Sal’nikova et al., 2007; Turkina et al., 2012). The second stage of folding and granulitic metamorphism is limited to the time range of 1.87–1.84 Ga and was accompanied by widespread formation of granitoids that make up large intrusions (Toisuk and Nizhnii Kitoi) along the northeastern boundary of the Irkut block and numerous layerlike and crosscutting bodies of charnockites in the south (Turkina and Kapitonov, 2019). The Paleoproterozoic granitoids intersect gneissic fabrics of the host Archean metamorphic rocks and are not deformed, i.e., are postorogenic.

Among the Paleoproterozoic mafic magmatic associations, we studied the Poludennyi gabbro–monzodiorite massif and small bodies of gabbro-dolerites at the southwestern endo- and exocontacts of the Toisuk pluton (Fig. 1). The Poludennyi massif is located in the central part of the Irkut block in the interfluve of the Poludennyi and Srednii Toisuk rivers (52°01ʹ27.6ʺ N, 103°15ʹ59.2ʺ E) and is a sill-like body with a visible size of 800 × 120 m. It is located in Archean mafic and felsic granulites, which are injected by numerous veins of Neoarchean gneissic granitoids. The monzodiorites cross host Archean rocks, and in all the parts of the massif, they have massive texture and coarse- to medium-grained structure; the monzodiorites are not deformed, i.e., are postorogenic. After Paleoproterozoic orogenic events, the largest scale manifestation of mafic magmatism was intrusion of dikes of the Neoproterozoic Nersa complex at ca. 725–715 Ma (Ernst et al., 2016), which correspond to the last stage of endogenous activity in the Sharyzhalgai uplift. Consequently, the Poludennyi massif probably formed in the Paleoproterozoic.

Gladkochub et al. (2013) first described a swarm of Paleoproterozoic gabbro-dolerite dikes on the left bank of the middle reaches of the Kitoi River. Thin (up to 0.8 m) dikes are exposed in bedrock outcrops. They are characterized by submeridional strike and steep dip; the latter differentiates them dramatically from the gently dipping Neoproterozoic dikes of the Nersa complex. The host rocks of the Kitoi dike swarm are predominantly Neoarchean heterogeneous gneissic granitoids. As will be shown below, gabbro-dolerites similar in composition are located in the southwestern exo- and endocontact zones of the Toisuk monzodiorite–granite pluton in the interfluve of the Bol’shoi Zadoi and Toisuk rivers (Fig. 1). On the left bank of the Bol’shoi Zadoi River (52°08ʹ34.8ʺ N, 103°07ʹ03.9ʺ E), separate fragments of gabbro-dolerites of visible sizes up to 0.5 m are located in bedrock outcrops composed of Archean banded gneissic granites. Insufficient exposure does not allow determining the thickness and extent of this probable dike body. Numerous elongated, irregular, and oval inclusions of gabbro-dolerites occur in the endocontact zone among coarse-grained biotite–amphibole monzodiorites and granodiorites of the Toisuk pluton (the coordinates of inclusions used for dating are 52°09ʹ40.0ʺ N, 103°14ʹ07.2ʺ E). The visible sizes of the gabbro-dolerite inclusions are 0.3–0.8 m. The gabbro-dolerites in both dike fragments and inclusions have massive textures and fine-grained structures and are not deformed. The gabbro-dolerites localized in Archean gneissic granitoids are similar to the dikes of the Kitoi swarm in structural position. In the Toisuk pluton, inclusions of gabbro-dolerites of different scales have a sharp contact with host biotite–amphibole monzodiorites and granodiorites and are intersected by thin veins of pegmatoid granites. These relations suggest that the gabbro-dolerites from the inclusions are close in time to the Toisuk pluton rocks.

The major and trace elements in the rocks were determined by XRF analysis on an ARL-9900 XL X-ray spectrometer and by ICP MS on an ELEMENT-II (Finnigan MAT) high-precision mass spectrometer with a U-5000AT+ ultrasonic spray, following the technique in (Nikolaeva et al., 2008), at the Analytical Center for Multielement and Isotope Research (Novosibirsk). Samples for the analysis were fused with extrapure lithium metaborate at 1050 °С in a platinum crucible, and then the fusion melt was dissolved in diluted HNO3. To ensure the stability of the obtained solution, the total acidity was maintained at 5% HNO3, and minor amounts of HF were added for correct determination of high-field-strength elements. Fusion at high temperature makes possible the decomposition of practically all weakly soluble minerals, which could concentrate rare-earth and high-field-strength elements, while increasing pH and adding HF as a complexing agent stabilize these solutions and minimize the loss of high-field-strength elements during hydrolysis at low pH. The detection limits of rare-earth and high-field-strength elements were from 0.005 to 0.100 ppm. The measurement accuracy on average was 2–5 rel.%.

The Sm and Nd concentrations and isotope compositions were analyzed by the TIMS method on a TRITON Plus multicollector thermoionization doubly focused mass spectrometer at the Geoanalitik Center of the Zavaritsky Institute of Geology and Geochemistry of the Ural Branch of the Russian Academy of Sciences (Yekaterinburg) in a static mode using the technique described in (Anikina et al., 2018). The total laboratory blanks were 0.07 ng for Sm and 0.40 ng for Nd. The measurement accuracies were ±1% for Sm and Nd (2σ), ±0.3% for 147Sm/144Nd (2σ), and ±0.003% for 143Nd/144Nd (2σ) (Table 2). The measured 143Nd/144Nd ratios were normalized to 148Nd/144Nd = 0.241572. The quality of the analyses was controlled using the JNdi-1 isotope standard. During the period of measurement, the weighted mean 143Nd/144Nd ratio for the JNdi-1 standard was 0.512109 ± 6 (2σ) (n = 16). The εNd values were calculated relative to CHUR (147Sm/144Nd = 0.1967; 143Nd/144Nd = 0.512638) (Jacobsen and Wasserburg, 1984).

Zircon U–Pb dating of three samples (7-21, 14-21, and 15-21) was carried out on a SHRIMP-II ion microprobe at the Center of Isotopic Research of the Russian Geological Research Institute (St. Petersburg), following the standard technique in (Williams, 1998). To choose analytical sites (spots), we used optical (in transmitted and reflected light) and cathodoluminescence (CL) images reflecting the internal structure and zoning of zircons. The intensity of a primary molecular-oxygen beam was 4 nA, and the spot (crater) was 25 μm in diameter and 2 μm in depth. The obtained data were processed using the SQUID program (Ludwig, 2005). The U/Pb ratios were normalized to a value of 0.0668 attributed to the TEMORA standard zircon with an age of 416.75 Ma (Black et al., 2004). The concentrations of Pb, U, and Th in the measured zircons were obtained using the 91500 zircon standard. Zircon U–Pb dating of one sample (22-21) was carried out at the Analytical Center for Multielement and Isotope Research using an Element XR (Thermo Finnigan) magnetic sector-field ICP MS coupled to a UP-213 (New Wave Research) excimer laser ablation system based on a Nd : YAG UV laser with a wavelength of 213 nm (LA-ICP-MS). The ICP MS parameters were optimized to obtain the maximum signal intensity of 208Pb at a minimum value of 248ThO+/232Th+ (less than 2%) using the NIST SRM 612 standard. All the measurements were performed using electrostatic scanning (E-scan) at the masses 202Hg, 204(Pb + Hg), 206Pb, 207Pb, 208Pb, 232Th, and 238U. Signals were detected in the counting mode for all the isotopes except for 238U (analog mode). The laser beam diameter was 35 μm; the pulse repetition rate was 6 Hz, and the laser radiation energy density was 3.5 J/cm2. The data of mass-spectrometric measurements were processed using the GLITTER software (Griffin et al., 2008). The data were corrected for U–Pb fractionation during laser ablation and for instrumental mass discrimination by standard bracketing with repeated measurements of the TEMORA-2 (Black et al., 2004) and Plešovice (Sláma et al., 2008) zircon reference material. The errors of single analyses (isotope ratios and ages) are given at the 1σ level, and the errors of calculation of the concordant ages and intercepts with Concordia are given at the 2σ level. The Concordia plots were constructed using the ISOPLOT 3 program (Ludwig, 2012).

The samples used for geochronological study are described in detail, taking into account the common features of the mineral and chemical compositions of the studied rocks. The Poludennyi massif is composed of dominant monzodiorites, and monzogabbro are rare. Monzodiorite (sample 22-21) is a medium-grained rock with the Pl + Cpx + Opx + Bt + Amph + Fsp + Qtz mineral association (Fig. 2a). The contents of main minerals are (in vol.%): Pl, 57; Cpx, 13; Opx, 5; Bt, 2; Amph, 3; Fsp, 10; Qtz, up to 10. Clino- and orthopyroxenes show solid solution decomposition structures as thin lamellae of ore minerals; amphibole develops after pyroxene. Alkaline feldspar with perthite structure often occurs as micrographic intergrowths with quartz. The content of apatite reaches 1%; other accessory minerals are ilmenite and zircon. The structure of the rocks is hypidiomorphic-grained, with poikilitic and micrographic structures in places.

Three samples of gabbro-dolerites (samples 7-21, 14-21, and 15-21) were used for dating; all of them have similar mineral compositions (Fig. 2bd). The rocks contain plagioclase phenocrysts. A typical groundmass mineral association includes Cpx (10–15%) + Opx (12–15%) + Pl (55–60%) + Bt (5–7%) + Fsp (4–6%) + Qtz (up to 1–2%). The pyroxenes show solid solution decomposition structures as thin lamellae of the ore minerals. Biotite forms elongated laths intergrown with pyroxenes. Alkali feldspar contains numerous inclusions of apatite and, less often, fine grains of zircon. In some gabbro-dolerites from the inclusions, secondary amphibole replaces pyroxene. The rocks show porphyritic structure; the groundmass demonstrated fine-grained dolerite and/or gabbro–ophitic structure. One gabbro-dolerite sample (7-21), with a similar mineral composition, is distinguished by an equal-grained aphyric structure and a larger grain size.

The rocks of the Poludennyi massif have a narrow range of the main-element contents. Monzodiorites (SiO2 = 53–54%) have low Mg# = 41–38 and a high content of TiO2 (1.7–1.9%) and P2O5; they are subalkaline with high K2O content (Table 1, Fig. 3). They show fractionated REE patterns ((La/Sm)n = 3.3–3.5; (Gd/Yb)n = 1.7–1.8) and have elevated concentrations of Ba, Th, light REE, and highfield-strength elements (146 ppm Zr) (Fig. 4a). The multielement spectra show enrichment in Rb and Ba as well as a distinct Nb minimum (Fig. 4b).

Gabbro-dolerites from the dike fragment (samples 5-21 and 7-21) are characterized by high Mg# (66–67), while rocks from the inclusions (samples 14-21 and 15-21) are depleted in Mg (Mg# = 39–49) (Table 1, Fig. 3). An increase in the concentrations of Fe2O3 and TiO2 (from 0.9 to 2.0–2.6%) and a slight increase in P2O5 content (0.52–0.64%) are observed with a decrease in Mg#. All the gabbro-dolerites are highly enriched in K2O, Ba (1036–1918 ppm), Th, and LREE and moderately enriched in Zr (207–238 ppm) and Nb (9.7–12.4 ppm). Gabbro-dolerites from the inclusions with a decrease in Mg# demonstrate a lower content of Ba and Th. A typical feature of these rocks is fractionated REE patterns with high (La/Sm)n (3.0–3.7) and (Gd/Yb)n (2.3–3.7) ratios (Fig. 4a). The multielement spectra of the gabbro-dolerites show variable depletion in Nb, and high-Mg gabbro-dolerites are also strongly depleted in Ti (Fig. 4). Similar composition features, including elevated concentrations of K2O, P2O5, and incompatible trace elements and similar shapes of multielement spectra, are typical of gabbro-dolerites from the Kitoi dike swarm (Fig. 4b).

Monzodiorites of the Poludennyi massif and gabbro-dolerites of the contact zone of the Toisuk pluton have similar isotope compositions (Table 2). All of them have negative εNd(T) values from –6.1 to –9.1. Their isotopic parameters suggest an enriched-mantle source, which correlates with the trace-element composition of these rocks enriched in incompatible trace elements.

Table S1 (Supplementary Material) presents the results of dating of zircon from the monzodiorite of the Poludennyi massif. More than 100 zircon grains were obtained from monzodiorite (sample 22-21). The zircons are prismatic grains 100–250 μm in length, rarely having pyramidal faces. The zircon grains have numerous solid-phase inclusions. In cathodoluminescence images, the zircons show mostly coarse zoning with alternating thick dark and light zones (Fig. 5a).

Zircon from monzodiorite is characterized by the wide range of U contents (170–1739 ppm) and Th/U (0.1–2.9) ratios, with the predominance of Th/U > 1 (Fig. 6). The weighted mean 207Pb/206Pb age of 48 zircon grains with discordance (D) ≤ 3% is 1873 ± 10 Ma (MSWD = 0.11) (Fig. 7) and, within the error, is equal to their concordant age of 1867 ± 2 Ma (MSWD = 1.9). In the calculation, seven grains were excluded, because their ages differ within the error from that of the main population of zircon (1892–1930 Ma); their older age suggests that these zircons were captured from the host Archean rocks and lost radiogenic Pb under the influence of mafic melt. Despite the wide range and high concentrations of U, there is no correlation between U content and the age of zircon, so we accept the age of ca. 1873 Ma as the best estimation of the time of monzodiorite crystallization.

The results of dating zircon from gabbro-dolerites are summarized in Table 3. In gabbro-dolerite from a dike fragment (sample 7-21), the zircons are elongated crystals 100–150 μm in length, which are mostly not zoned in CL images and, less often, show coarse banded zoning (Fig. 5b). Like zircons from the Poludennyi massif, they often contain solid-phase inclusions. There are two groups of zircon according to the U content: 384–690 and 940–3487 ppm. The dominating high-U zircons are enriched in Th (1090–5259 ppm), while the low-U zircons are depleted in Th (74–498 ppm); both have a wide range of Th/U ratios (0.2–2.8) with a tendency to grow in grains with high concentrations of U and Th (Fig. 6). The high-U zircons are characterized by increased discordance (D = 4–14%). All 16 zircon grains form a discordia with an upper intercept age of 1858 ± 7 Ma (MSWD = 0.27) (Fig. 8a).

Fine-grained gabbro-dolerites from two inclusions (samples 14-21 and 15-21) contain dominant small prismatic zircons 70–150 μm in length (Fig. 5cd). In sample 15-21, zircon is unzoned and includes low- and high-U groups with U concentrations of 208–1414 ppm and 2849–8661 ppm (Fig. 6). The high-U grains are characterized by a decrease in 207Pb/206Pb age due to increased discordance (D = 4–55%). Except for three grains with a Th/U ratio of 1.3–2.1, most zircons have Th/U ratios from 0.10 to 0.48. Nineteen zircons plot on the discordia line with an upper Concordia intercept age of 1857 ± 10 Ma (MSWD = 3.9), and for 15 spots with D ≤ 5%, the weighted mean age is 1855 ± Ma (MSWD = 1.6) (Fig. 8cd). Taking into account the best parameters of the weighted mean age, the age of 1855 Ma serves as a better estimate of the time of crystallization of zircon from gabbro-dolerite. Two ancient zircons from this sample with a length of more than 100 μm are weakly zoned in CL, have a rim with 207Pb/206Pb ages of 2531 and 1910 Ma, respectively, and are probably xenogenic zircons. In gabbrodolerite (sample 14-21), the zircons are unzoned in CL or show weak growth zoning in marginal parts (Fig. 5d). They have low concentrations of U (157–932 ppm) and Th (67–759 ppm) and a Th/U ratio of 0.44–0.91 (Fig. 6). The concordant and weighted mean 207Pb/206Pb ages of 11 zircon grains are 1850 ± 7 Ma (MSWD = 0.14) (Fig. 8b) and 1850 ± 8 Ma (MSWD = 0.21), which, within the error, is close to the age of zircon from other gabbro-dolerites.

Conditions of the zircon crystallization in the Paleoproterozoic mafic rocks. A characteristic feature of the dated zircons is wide variations in U and Th concentrations and Th/U values (Fig. 6). The Th/U ratio in the range of 1–2 is dominant for zircons from monzodiorites and gabbro-dolerites from a dike fragment, while gabbro-dolerites from inclusions in the rocks of the Toisuk pluton contain zircons predominantly with Th/U < 1. There is a trend toward an increase in Th/U with an increase in the contents of U and Th. The distribution coefficients of these elements are similar, so the wide range of U and Th contents in zircon may be due to their different enrichment in residual melt in the absence of other minerals concentrating U and Th. Since Th is a more incompatible element than U during the crystallization of mafic melt, this may be one of the reasons for the increase in Th/U in zircons enriched in these elements. Thus, the enrichment of zircon in both U and Th with an increase in Th/U testifies to the crystallization of zircon from a differentiated mafic melt and can be considered a typical feature of such zircons.

A prominent feature of the rocks of two studied mafic complexes is enrichment in incompatible elements. Gabbrodolerites have the highest Zr concentrations, including their high-Mg varieties (Mg# = 66; Zr = 207 ppm); therefore, high Zr concentrations are the initial feature of their parental magma. In the case of monzodiorites, the Zr enrichment might result from fractionation of a more magnesian parental melt. Elevated Zr concentrations in the studied rocks due to the capture of zircon from crustal rocks at the emplacement level or in an intermediate chamber are unlikely, since only two older zircon grains (ca. 2.5 and 1.9 Ga) were found among all the dated zircons in gabbro-dolerite. It should be noted that the monzodiorites of the Toisuk pluton, which contain the gabbro-dolerite inclusions, also lack Archean xenogenic zircons (Turkina and Kapitonov, 2019).

To assess the origin conditions, the zirconium saturation temperatures were calculated using the equations from (Shao et al., 2019). For gabbro-dolerites, the temperature is in the range of 708–905 °C, and for monzodiorites, the temperatures are 839–965 °C (Table 1). In both cases, these values are clearly lower than the probable temperatures of the generation of mafic melts; therefore, the formation of zircons occurred at the late stage of crystallization. This conclusion is consistent with the location of zircon in micrographic intergrowths of alkali feldspar and quartz, which suggests the zircon crystallization from the last melt drops enriched in Zr. This is also consistent with the enrichment of zircon in Th and U, which indicates the crystallization of zircon from an evolved residual melt.

Paleoproterozoic stage of mafic and felsic magmatism in the southwest of the Siberian Craton. The geochronological study of zircons shows that geochemically similar gabbro-dolerites composing bodies of different shapes in the exo- and endocontact zones of the Toisuk pluton are close in age and are coeval with monzodiorites of the Poludennyi massif. All these mafic magmatic rocks formed in the range of 1873–1851 Ma. If we sum up all the data on the age of Paleoproterozoic mafic rocks, the time of their formation is limited to the interval of 1.87–1.85 Ga (Table 4). As shown earlier (Donskaya et al., 2002; Sal’nikova et al., 2007; Turkina and Kapitonov, 2019), the Paleoproterozoic granitoids in the Sharyzhalgai uplift formed in a rather narrow time range of 1.87–1.84 Ga. In the Irkut block, this magmatic stage includes the formation of the largest Toisuk monzodiorite–granite pluton and numerous small bodies of charnockites and granitoids in the southeast of the block. Thus, geochronological data indicate that the mafic and granitoid magmatism were coeval. In addition to small massifs of gabbronorite, monzodiorites, and dikes, the Paleoproterozoic mafic rocks include the monzodiorites and granodiorites of the Toisuk pluton, which make up almost a third of this largest pluton in the Sharyzhalgai uplift and formed with a contribution of mafic magma (Turkina and Kapitonov, 2019). Mafic magmatism in the Irkut block occurred in a postcollisional extensional setting, which was favorable for underplating of mantle melts into the crustal basement. Mafic magmatism occurs throughout the Irkut block, suggesting the presence of an extensive thermal anomaly in the mantle, which might have ensured heating of the crust as a result of underplating of mantle magmas and have been the cause of both regional granulite metamorphism at 7–8 kbar and 850–870 °C (Sukhorukov, 2013; Sukhorukov and Turkina, 2018) and granite formation.

Composition and mantle sources of Paleoproterozoic mafic rocks. The studied monzodiorites and gabbro-dolerites and coeval gabbro-dolerites of the Kitoi dike swarm have some common compositional features. These are the presence of biotite and alkali feldspar and enrichment in K2O, Ba, and incompatible trace elements: Th, LREE, and Zr. All these features are typical of rocks of the shoshonite–latite series (Fig. 9). Both the monzodiorites and gabbrodolerites show highly fractionated multielement spectra due to enrichment in LREE and Th (Fig. 4). According to (Ivanov et al., 2019), the small bodies similar to lamprophyres and enriched in Ba, Th, Zr, and LREE are also the products of shoshonitic magmatism in the studied region.

The close ages and similarity in geochemical characteristics suggest that the monzodiorites and gabbro-dolerites formed from common parental melts derived from an enriched-mantle source. The indicator ratios of trace elements point to an enriched-mantle source. During melting of peridotites, the solid phases/melt distribution coefficients (Kd) are KdNb < KdY; that is, Nb is a more incompatible element than Y. According to the model calculations, melting of a source similar to primitive mantle leads to a slight increase in (Nb/Y)PM to 1.3 (Turkina et al., 2022). The monzodiorites and gabbro-dolerites show (Nb/Y)PM ratios of 1.7 and 2.0–2.8, respectively; the high ratios indicate the enriched character of the mantle sources for the rocks.

The above conclusion is consistent with the Nd isotopic composition of the studied rocks. Monzodiorites of the Poludennyi massif are characterized by εNd(T) values of –7.7 and –7.8; gabbro-dolerites have a wider range of negative εNd(T) values (from –6.1 to –9.6) (Table 2). The monzodiorites of the Toisuk pluton have a similar isotopic composition; the εNd(T) values are in the range from –5.3 to –10.2 (Turkina and Kapitonov, 2019). Such low values of εNd(T) cannot be a result of crustal contamination. According to the Nd isotopic composition of Paleoproterozoic granites, the crust of the Irkut block is characterized by εNd(T) in the range from –6 to –12 for a time of 1.86 Ga (Turkina, 2022). The contribution of the crustal component to the genesis of mafic rocks must be at least 30%, which is inconsistent with their composition, primarily high Mg content. In addition, there is no correlation between the values of εNd(T) and the main indices of crustal contamination: the content of SiO2, Mg#, and the (La/Sm)n ratio. Therefore, the isotope parameters of the Paleoproterozoic mafic rocks indicate a longlived enriched-mantle source, such as a subcontinental lithospheric mantle. Previously, it was presumed that the monzodiorites of the Toisuk pluton had been formed by differentiation of mafic magma from an enriched-mantle source, such as the Archean subcontinental lithosphere. Metasomatism and the formation of an enriched subcontinental lithospheric mantle might have occurred in Neoarchean time. The Neoarchean stage in the Irkut block corresponds to the formation of protoliths of mafic granulites (2.66 Ga) with elevated LILE concentrations and Nb depletion, which are typical of subduction-related volcanic rocks (Turkina et al., 2012). Model calculation (Turkina and Kapitonov, 2019) showed that the evolution of the Neoarchean mantle with low 147Sm/144Nd = 0.125 had led to a decrease in εNd(T) to values from –6.3 to –10.5 by 1.86 Ga; these εNd(T) values correspond to the range of εNd(T) for gabbrodolerites and monzodiorites. The studied rocks show a sharp depletion in Nb relative to Th and LREE, which is typical of basalts of subduction-related settings; this is evidence for their formation from the subcontinental lithospheric mantle enriched during metasomatism by fluid/melt derived from the subducting oceanic plate. Thus, a probable model for formation of the studied mafic rocks is the melting of the subcontinental lithospheric mantle enriched in previous Neoarchean subduction processes.

According to U–Pb dating of magmatic zircon, the formation of monzodiorites of the Poludennyi massif and gabbro-dolerites in the endo- and exocontact zones of the Toisuk pluton occurred at 1.87–1.85 Ga. The intrusion of mafic magmas and their underplating into the basement of the crust, caused by the rising mantle diapir in a postcollisional extension setting, determine the near-coeval mafic and granitoid magmatism in the Irkut block in the interval of 1.87–1.84 Ga. The parental magmas for the Paleoproterozoic mafic association are characterized by enrichment in incompatible elements, including Zr, and low negative εNd(T) values. These geochemical and isotopic characteristics point to the formation from a long-lived enriched- mantle source, i.e., a subcontinental lithospheric mantle. The crystallization of zircon from the last portion of the evolved mafic melt is evidenced by its location in micrographic intergrowths of alkali feldspar and quartz. Late crystallization of zircon is also indicated by low zirconium saturation temperatures (710–965 °С) and its enrichment in U and Th with increasing Th/U, resulting from the growth of the concentrations of these highly incompatible elements in residual melt.

The authors thank Dr. S.N. Rudnev and employees of the Analytical Center for Multielement and Isotope Research, Siberian Branch of the Russian Academy of Sciences: PhD. I.V. Nikolaeva, PhD. S.V. Palesskii, D.V. Semenova, A.V. Karpov, and N.G. Karmanova, who performed analytical works. We are also grateful to Ph.D. N.G. Berezhnaya (Center of Isotopic Research, A.P. Karpinsky Russian Geological Research Institute) for help in preparing zircon for research.

Uranium–lead isotope dating was supported by the Russian Foundation for Basic Research (grant No. 20-05-000265). Data on mafic and granitic magmatism were summarized as a part of a basic research project of V.S. Sobolev Institute of Geology and Mineralogy, Siberian Branch of the Russian Academy of Sciences (Novosibirsk).

Supplementary data