Late Paleozoic–Cenozoic Tectonothermal Evolution of Transbaikalia: Thermochronology of the Angara–Vitim Granitoid Batholith

Phanerozoic orogenic structures played a key role in the geological history. Reconstructing their geodynamic evolution commonly includes such important aspects as the origin of granitic batholiths, as well as their ascent, emplacement, and tectonic exhumation as petrological evidence for stacking and subsequent extension of continental lithosphere. Constraining the age of crystallization, high-temperature metamorphism, and later magmatism, metamorphism, meta-somatism, and tectonic activity, became possible due to progress in radiometric dating. Dating of minerals by U/Pb (SHRIMP, LA-ICP-MS for U-bearing minerals, primarily zircon), ⁴⁰Ar/³⁹Ar (micas and feldspar), and fission track (apatite) methods can cover the temperature range of the closure of isotope systems from crystallization of melts at ≈850 °С (U/Pb, zircon) to <100 °С (apatite fission tracks) and thus allows tracing the history of igneous bodies from crystallization to exhumation. The available ⁴⁰Ar/³⁹Ar data for amphibole, biotite, and feldspar which have different closure temperatures of the K–Ar isotope system reveal correlation between the thermal history of granitoids and tectonic regime of orogenic processes (Travin et al., 2012). That is, collisional and accretionary orogens evolve at different rates which depend more on tectonic than erosion processes. Therefore, the thermal history of granitic batholiths can provide independent evidence on the evolution of orogenic areas.

We study the thermal history of the Angara–Vitim batholith, one of world largest igneous bodies, which records the tectonic evolution of the late Paleozoic Mongol–Okhotsk orogenic belt in southern Siberia. The age constraints of thermal and tectonic events are obtained using U/Pb isotope dating of zircon, ⁴⁰Ar/³⁹Ar dating of amphibole, biotite, and feldspar, as well as apatite fission track (AFT) thermochronology. All minerals were extracted from the same samples, i.e., the results represent successive closure of isotopic systems at specific points of the geological space.

The late Paleozoic Angara–Vitim batholith was chosen for several reasons. It is an enormous granitic body reaching about 1,000,000 km³, which has been well documented in terms of petrology and geochronology. Its thermal history was affected by later Mesozoic intrusions which make up a large NE Mongol–Transbaikalia volcanoplutonic belt in the central part of the area (Dobretsov, 2003). The batholith was sampled outside the belt to minimize biases from the later intrusions.

The Angara–Vitim batholith was formed in the late Paleozoic, between 325 and 270 Ma. It originated in a typically crustal magma source (Barguzin complex), and the contribution of the mantle component increased progressively (Chivyrkui, Zaza, Early Selenga, Early and Late Kunalei complexes) toward a mixed crustal-mantle composition (Yarmolyuk et al., 1997; Litvinovsky et al., 1999, 2011; Tsygankov et al., 2007, 2017).

Magmatism in Transbaikalia, including the formation of AVB, was associated with subduction events, with possible inputs from enriched mantle sources (Zorin, 1999; Donskaya et al., 2013). On the other hand, late Paleozoic magmatism in the region was controlled by joint action of a mantle plume and processes at the final stage of the late Paleozoic–early Mesozoic orogeny (Yarmolyuk et al., 1997; Litvinovsky et al., 1999; Tsygankov et al., 2007, 2017). Magma emplaced at depths 20–13 km, during the collisional (10%), around 320–290 Ma, and postcollisional (90%) 310–280 Ma events (Litvinovsky et al., 1999; Tsygankov et al., 2007, 2017). A major role of the late Paleozoic–early Mesozoic tectonic activity in the consolidation of continental crust was proven by data on late Paleozoic activity in the Transbaikalian segment of the Central Asian orogenic belt, especially, on the 295 ± 2 Ma metamorphic complexes of Mount Mandrik (Mazukabzov et al., 2010) and three-stage folding and thrusting in the Tunka Goltsy (Bare Mountains) area from 316 to 286 Ma (Buslov et al., 2009; Ryabinin et al., 2011). Those movements were concurrent with strike-slip faulting along the Main Sayan Fault that separates the Siberian craton and the Altai–Sayan orogenic area (Savel’eva et al., 2003). In general, the late Paleozoic orogeny, which produced fold-thrust belts and related igneous and metamorphic bodies, occurred in the context of global-scale collisional interactions of East European and Siberian continents which left abundant traces in southern Siberia (Buslov et al., 2003, 2004, 2013, 2022; Dobretsov and Buslov, 2011; Buslov and Cai, 2017).

The Mongol–Okhotsk belt resulted from the closure of the respective ocean and a North China-Eurasia collision in the Early Mesozoic (Zonenshain et al., 1990; Zorin, 1999; Tomurtogoo et al., 2005; Donskaya et al., 2013; Shevchenko et al., 2014; Sorokin et al., 2020). In the late Mesozoic, a vast area of northeastern continental Asia, >3,000,000 km², including Transbaikalia, underwent crustal extension, either as a result of postorogenic collapse of thick crust in the Mongol–Okhotsk orogen, or due to a far-field effect of the Pacific subduction (Wang et al., 2012). Those events were illustrated by data on metamorphic core complexes (mostly in southern Transbaikalia) as direct indicators of extension (Sklyarov et al., 1997; Wang et al., 2012; Donskaya et al., 2008). Meanwhile, the Mesozoic events in northern Transbaikalia, including the Angara–Vitim granitoid batholith, remain poorly investigated. Reconstructing the thermal history of igneous rocks in orogenic areas with geochronological methods is a reliable tool to study the tectonic evolution of orogens. The approach is based on the use of isotopic systems that closed at different temperatures (Hodges, 2004): ~900 °С for U/Pb in zircon, ~550 to 250 °С for ⁴⁰Ar/³⁹Ar in amphibole (550 °С), biotite and muscovite (~330 °С), feldspar and plagioclase (~250 °С), and ~110 °С in the case of AFT.

The ages and closure time of isotopic systems compared in different minerals have implications for the depths where the rocks were located in different periods of time, from the onset of crystallization to tectonic exhumation, assuming an average temperature gradient of 25–30 °C/km. The ages and paleotemperatures are plotted in a series of diagrams for coeval rocks sampled on the surface, which illustrate the evolution of cooling, and the regional history as a whole. The regional thermal history is converted into denudation chronology, with regard to the geothermal gradient. The time-temperature diagrams show the amount of material denuded for certain time and, hence, the denudation rates, which accelerate during large tectonic events. We study the history of crustal blocks that accommodate the late Paleozoic Angara–Vitim batholith.

Minerals for U/Pb and ⁴⁰Ar/³⁹Ar dating (zircon, amphibole, biotite, muscovite, K-feldspar, and plagioclase), as well as apatite for fission track thermochronology, were extracted by the conventional techniques of magnetic and density separation at the V.S. Sobolev Institute of Geology and Mineralogy (Novosibirsk).

⁴⁰Ar/³⁹Ar dating of monomineral fractions was carried out at the Analytical Center for Multi-element and Isotope Studies (Novosibirsk) as in (Travin, 2016). Samples were wrapped in aluminum foil together with biotite MCA-11 and OCO 129-88 standard monitor samples, placed in quartz capsules, vacuumed, and sealed. The capsules were irradiated by fast neutrons in a Cd-lined tube of the IRT-T nuclear reactor at the Tomsk Polytechnical University, with a neutron flux gradient no more than 0.5% of the sample size. The samples were exposed to external stepwise heating in a quartz tube. The ⁴⁰Ar and ³⁶Ar blank runs (10 min at 1200 °С) did not exceed 3 × 10^–10 and 0.003 × 10^–10 ncm³, respectively. Argon cleaning was performed using ZrAl-SAES getters. The argon isotope composition was measured on a Micromass Noble Gas 5400 mass spectrometer (UK), to an accuracy of ±1σ. The contribution of interfering Ar isotopes formed together with ³⁹Ca and ⁴⁰K was estimated using the coefficients of (³⁹Ar/³⁷Ar)_Ca = 0.001279 ± 0.000061, (³⁶Ar/³⁷Ar)_Ca = 0.000613 ± 0.000084 and (⁴⁰Ar/³⁹Ar)_K = 0.0191 ± 0.0018. The plateau ages were calculated in Iso-plot-3 (Ludwig, 2003), as weighted average values over at least three successive temperature steps. The results were interpreted with reference to the conventional criteria (Fleck et al., 1977): (i) age difference between any two steps not exceeding K = 1.96 * √(σ12 + σ22); (ii) consistent Ca/K ratios (mineralogical criterion); (iii) at least 60% cumulative ³⁹Ar released at each step.

Apatite fission track (AFT) dating was carried out at the Analytical Center for Multi-element and Isotope Studies (Novosibirsk) and the Kazan Federal University, Kazan (Buslov et al., 2021), following the conventional procedure (Donelick et al., 2005; Soloviev, 2008). Apatite grains picked manually from the heavy fraction after separation in heavy liquids were mounted on epoxy resin discs, polished and etched in 5.5 М HNO₃ for 12–18 s at 21 ± 2 °C. Special preparations were made additionally for standard samples. The fission-track density of U was calculated manually using a high-resolution Olympus BX51 optical microscope, at magnification ×1250, in specified zones of apatites. The distribution of track lengths was determined using the Dpx-View software for the InfinityX camera and plotted in histograms. The contents of ²³⁸U were estimated at the sites with calculated track lengths and densities on a ThermoFisher Scientific Qc iCAP ICP-MS analyzer coupled with an NWR213 laser ablation system. The workflow included two measurements of external (Durango apatite) and control (Fish Canyon Tuff and Limberg t3 apatites) standard samples in the beginning and at the end of each run and one (external or control) measurement at every five to ten measurements of the samples. The external standard was used to correct the data for element fractionation upon laser evaporation, mass discrimination, and mass spectrometer drift, while the control sample was run to check the quality of measurements. Additionally, the NIST SRM 612 and NIST SRM 610 standards were measured in the beginning and at the end of each run. The selected external and control apatite samples are homogeneous as to the track age and U distribution and have suitable spontaneous fission-track density. The raw mass spectrometry data were processed in Iolite 3.65, with ⁴³Ca as a marker. The determined concentrations of ²³⁸U (ppm) in apatite, with the respective errors, as well as fission-track density and track lengths in the zones where U concentrations were estimated, were used to plot time-dependent temperature variations. The thermal history of rocks was modeled using the HeFTy Version 1.8.2 software (Ketcham et al., 2000; Ketcham, 2005). The time-temperature diagrams were interpreted assuming a geothermal gradient of 30–25 °С/km.

The Angara–Vitim batholith emplaced in the Late Paleozoic between 325 and 270 Ma, and its source evolved from typically crustal (Barguzin complex) to mixed crustal-mantle (Yarmolyuk et al., 1997; Litvinovsky et al., 1999, 2011; Tsygankov et al., 2007, 2017), with progressively increasing contribution of the mantle component (Chivyrkui, Zaza, Early Selenga, Early and Late Kunalei complexes). See Table 1 and Fig. 1 for the complete list of samples.

The Svyatoi Nos igneous rocks include the Barguzin and Chivyrkui granitoid complexes which intrude Early Paleozoic gneisses, amphibolites, marbles, and quartzites (Fig. 2) alternating concordantly along the intrusion margin. Gneiss granosyenites in the northern part host wallrock xenoliths.

Granites from the Barguzin complex sampled from the central part of the granite field (11-09-24/1) and in the southern part of the peninsula (SNV-9) (Fig. 2) previously gave U/Pb zircon ages of 297.5 ± 1.5 and 289.4 ± 1.1 Ma, respectively (Mikheev, 2019; Travin et al., 2020). We dated both samples by the ⁴⁰Ar/³⁹Ar method and sample 11-09-24/1 by AFT. Additionally, ⁴⁰Ar/³⁹Ar dating was applied to biotite-amphibole gneiss sample SNV-21 (wallrock metamorphics) from the western part of the peninsula (Fig. 2).

Other samples dated in this study included andradite syenite (svyatonosite) of the Markov and Escola plutons along the margins of the Svyatoi Nos intrusion (Petrova et al., 1981), with 261 ± 21 Ma, 273 ± 24 Ma (Sm/Nd), and 286 Ma (U/Pb, titanite) reported ages (Levitsky et al., 2006); Chivyrkui quartz syenite from the eastern coast of the Chivyrkui Bay dated previously by the U/Pb zircon method at 297 ± 1 and 289.4 ± 1.1 Ma (Mikheev, 2019); and Barguzin granite from the Baklany Island. The andradite syenite samples were dated by the ⁴⁰Ar/³⁹Ar method on K-feldspar from monzonite of the Markov (Em-13-16, Em-13-18) and Escola (Em-13-19, Em-13-21) plutons, and the samples from the Chivyrkui Bay (Em-14-21) and Baklany Island (Em-14-27) were analyzed by both ⁴⁰Ar/³⁹Ar and AFT methods.

Several samples of gneissic biotite granite with fragments of metamorphic rocks, massive homogeneous biotite granite, porphyritic quartz syenite, and quartz monzonite that represent the Barguzin, Chivyrkui, and Zaza complexes of the Angara–Vitim batholith were collected in the northeastern coast of Lake Baikal (Tsygankov et al., 2017; Mikheev, 2019) and dated by the U/Pb method (Table 1, Fig. 1).

The Barguzin, Chivyrkui, Zaza, Lower Selenga, and Early Kunalei granitoid complexes (Tsygankov et al., 2007) were also sampled at the 60 × 30 km Kurba site where they are exposed at the modern erosion level along the Kurba River (Fig. 2 in Litvinovsky et al. (2011)).

Zelenaya Griva pluton belongs to the Barguzin complex (autochthnonous facies of its main phase) and is composed of medium-grained often porphyritic gneissic biotite granite, with migmatite and parallel or crosscutting leucogranite veins along the contacts. Biotite gneiss granite sample Zr-65/1, an equivalent of sample Z-13-02 we analyzed, was previously dated at 325.3 ± 2.8 Ma by the U/Pb zircon method (Tsygankov et al., 2007).

Temen pluton, which likewise belongs to the Barguzin complex, has no contact with the Zelenaya Griva gneiss granite but is intruded by the 303–289 Ma Zaza leucogranite (Tsygankov et al., 2007). Age constraints (318 ± 4 Ma, U/Pb zircon) are available for sample Te-01-06 of porphyritic biotite granite (Tsygankov et al., 2010) similar to sample T-13-01 of this study.

Khasurta pluton, part of the Lower Selenga complex, consists of fine to medium-grained monzonite, quartz monzonite, and biotite-hornblende syenite and granosyenite and is intruded by small fine leucogranite stocks and dikes. Its central part is composed of homogeneous medium-grained quartz syenite and granosyenite. A U/Pb zircon age of 283.7 ± 5.3 Ma was obtained for monzonite sample Hs-59a (Tsygankov et al., 2007) equivalent to our sample H-13-03.

The thermal history of rocks was reconstructed from AFT data for several granite samples (SE1218, SE1220, SE1224) from the Western Baikal region (Fig. 1), which have U/Pb zircon ages in the range 302–320 Ma corresponding to the emplacement of the Angara–Vitim batholith (Bishaev et al., 2022).

The AVB igneous complexes were also sampled in north-western Transbaikalia (Fig. 1) southward from Lake Baunt to the Talaya River along the Tsypikan River (Antonov et al., 2016).

Baunt pluton is composed of medium to hypabyssal coarse-grained homogeneous biotite granite and leucogranite which compositionally correspond to high-K subalkaline rocks with high contents of main incompatible elements. U/Pb zircon ages are available for biotite granite sample Ant-51-1 (310.8 ± 2.8 Ma) and for gneiss granite sample Ant-54-5 from a field of autochthonous medium to coarse-grained biotite gneiss granite, inhomogeneous medium biotite leucogranite south of Mount Khapton, and fine biotite leucogranite farther southward (Antonov et al., 2016). The rocks have low-alkaline moderate to high potassic compositions with low concentrations of main incompatible elements, volatiles, and metals. The ages reported by Antonov et al. (2016) span a long interval from the Archean and Paleoproterozoic to the latest late Carboniferous, while sample Ant-54-5 showed two clusters in the concordia diagram: around 305.3 ± 3.1 Ma (11 points) and 286.7 ± 2.3 Ma (8 points).

In this study, ⁴⁰Ar/³⁹Ar dating was applied to samples Ant-51-1, Ant-54-5, and Ant-54-3.

The ⁴⁰Ar/³⁹Ar spectra of the analyzed mineral fractions (Fig. 3, Table 1) show quite prominent age plateaus, and the measured ages are correlated with the stability of isotopic systems decreasing in the series amphibole => biotite => feldspar. Thus, it is reasonable to assume that the plateau ages of magmatic minerals correspond to the closure time of their K/Ar isotopic system (Hodges, 2004).

The K-feldspar age spectrum of sample T-13-01 contains an intermediate plateau of four steps corresponding to 142 ± 2 Ma, at 25% of ³⁹Ar released in the low-temperature part (Fig. 3) and another intermediate plateau of 212.2 ± 2.3 Ma, at 50% of ³⁹Ar released, in the high-temperature part, both values being younger than the biotite age (281.4 ± 2.9 Ma) for the same sample. The young ages may be due to the presence of asynchronously closed diffusion domains of different sizes in the K-feldspar structure.

The age spectrum of plagioclase from the same sample likewise contains an intermediate plateau of two steps corresponding to 215.2 ± 2.3 Ma, i.e., similar to the high-temperature value for K-feldspar, after low-temperature steps with notably overestimated ages.

The spectra of K-feldspar from Ant-51/1 biotite granite and Ant-54/5 gneiss granite show plateaus of two steps with 153.9 ± 2.6 and 173.0 ± 3.4 Ma ages in the low-temperature portion, which presumably record the closure of the smallest diffusion domains with the lowest effective closure temperature.

The 37-44 Ma AFT ages of granite samples 11-09-24/1, Em14-21, and Em14-27 from the Chivyrkui Bay and the Svyatoi Nos Peninsula (Figs. 1, 2) are consistent with the values reported (Jolivet et al., 2009) for granite samples from three sites in the Barguzin Range. The AFT ages record the time when the rock cooled down to 110 °С corresponding to a depth of ~4 km at the average temperature gradient 30–25 °C/km.

The field of most probable cooling curves (Fig. 4) was constrained by inversion of track lengths in HeFTy Version 1.8.2 (Ketcham et al., 2000; Ketcham, 2005). The samples reached the 110° isotherm between 70 and 50 Ma, during an episode of fast cooling when the exhumed material rose to ~4 km. The cooling episode was followed by a period of relative stability till the Late Miocene and subsequent fast exhumation for the past 5 Ma. Similar curves (Fig. 4) were obtained for the Barguzin Range granite (Jolivet et al., 2009).

Granite samples from the southwestern AVB part showed much older ages (118 to 77 Ma) which approximately agree with the AFT data for their counterparts from the Western Baikal region ranging in age mostly from 140 to 100 Ma (van der Beek et al., 1996). The cooling may have occurred late during the Jurassic–Early Cretaceous orogeny associated with closure of the Mongol–Okhotsk Ocean when at least 2–3 km of rocks were denuded.

The published (Yarmolyuk et al., 1997; Litvinovsky et al., 1999; Tsygankov et al., 2007, 2017; Jolivet et al., 2009; Antonov et al., 2016; Travin et al., 2020) and new thermochronological data (Fig. 5) show that the rocks of the sampled southwestern, eastern, and northeastern parts of the Angara–Vitim batholith shared a generally similar thermal history. The thermochronological record should be interpreted taking into account that the rocks originated at different depths of 15 to 20 km for the Barguzin complex (autochthonous facies granites) to 10–15 km or shallower for allochthonous granites of the Barguzin and other complexes.

The isotope system of biotite closed at T_c ~330 °С (Hodges, 2004), and the respective biotite age can be used to estimate the time when the rocks reached the <10 km depths, assuming a temperature gradient of 25–30 °С. Correspondingly, the thickness of rocks denuded by the time of the isotope system closure varied from 15 km to 3–4 km, depending on the initial depths. The biotite age of samples from the northeastern coast of Lake Baikal coincides with the time of magmatism (Table 1, Fig. 1), i.e., the rocks in this AVB part ascended to <10 km during or soon after the emplacement.

The biotite ⁴⁰Ar/³⁹Ar ages of samples from the Svyatoi Nos, Kurba, and Baunt sites, where abyssal autochthonous granites of the Barguzin complex coexist with allochthonous granites of the same and other complexes (Table 1, Figs. 1, 2), are 177–246 Ma for the Svyatoi Nos site, 168–281 Ma for the Kurba site, and 211–238 Ma for the Baunt site. The thermal effect of later intrusions might be responsible for the younger biotite ages, but no bodies that could potentially produce a thermal aureole are evident from geological data. Most likely, the rocks, which originated at different depths but became juxtaposed at the modern erosion level, reached the 10 km depth asynchronously. The presence of older rocks within the sites indicates that at least 3 km of material had been denuded after the granite emplacement. The isotope system of biotite in the other plutons had been closing for 120 Myr, possibly, because the tectonic processes after the late Paleozoic activity were very slow.

The age of origin and early exhumation stage of the AVB rocks coincides with the late Carboniferous–early Permian age (316–286 Ma) obtained by ⁴⁰Ar/³⁹Ar dating of syntectonic minerals for the Main Sayan right-lateral strike-slip fault (Savel’eva et al., 2003) coexisting with the fold-thrust structures of the Tunka Bare Mountains in the East Sayan.

Unlike the 120 Myr long closure of the biotite isotope system in the AVB granitoids, the isotope systems of K-feld-spar/plagioclase closed at ~250–200 °C within a relatively narrow time span of 170–140 Ma (Late Jurassic–Early Cretaceous) at all sampled sites (Fig. 5). With the 25–30 °С/km geothermal gradient, these temperature values correspond to depths of 7–8 km indicating an amount of denudation about 3 km since the origin of the Mongol–Okhotsk belt (Didenko et al., 2010; Wang et al., 2012; Shevchenko et al., 2014; Arzhannikova et al., 2020).

The closure of the K-feldspar/plagioclase isotopic system in the samples from the southwestern AVB part was followed shortly after by the AFT system closure (Fig. 5 b), when the rocks reached depths of 4–5 km (110 ± 10 °C isograde). Therefore, the growth of the Mongol–Okhotsk orogen continued in the area and caused the denudation of rocks (3–4 km).

Inversion of track lengths (Fig. 4) shows that the granitoids sampled at the modern erosion level reached 3–2 km depths and temperatures of 80 °C in the time span from 80 to 50 Ma, which indicates ~1.5–1 km of denudation. Then, from 60 to 5 Ma, Transbaikalia underwent slow denudation in the conditions of tectonic stability. The thermal history curves of the AVB rocks for the past 5 Ma are steep, and the amount of denudation was 2–3 km. Rapid cooling of the rocks was possibly caused by erosion in the area subject to uplift under a far-field effect of the India–Eurasia collision (Dobretsov et al., 1996; Buslov et al., 2007; de Grave et al., 2007; Buslov, 2012).

The revealed tectonothermal history of the AVB rocks in the Western Baikal region has implications for the formation of orogens and related sedimentary basins. For instance, deposition in the Irkut, Kan-Acha, Tuva, Kuznetsk, and West Siberian basins was associated with the formation of the Mongol–Okhotsk orogen.

The tectonothermal evolution of Transbaikalia recorded in thermochronological data for the Angara–Vitim batholith included several stages of activity and ensuing denudation correlated with global-scale tectonic events:

• (1) late Paleozoic (320–280 Ma): activity concurrent with large-scale collisional processes in southern Siberia, including folding and thrusting in the Tunka Bare Mountains and denudation from 10 km (autochthonous granites) to a few km (allochthonous granites), depending on the origin depths of rocks;

• (2) Middle–Late Jurassic (170–140 Ma): Mongol–Ok-hotsk orogeny and ~3 km of denudation;

• (3) Pliocene–Holocene (5 Ma to present): activity in Transbaikalia as an echo of the India–Eurasia collision and 2–3 km of denudation.

The study was supported by the Russian Science Foundation (grant No. 22-17-00038, thermochronology) and was carried out as part of the research plan of the Geological Institute, Ulan Ude (project AAAA-A21-121011390002-2, petrology) and on government assignment to the V.S. Sobolev Institute of Geology and Mineralogy (projects Nos. 122041400057-2 and 122041400171-5, geology).