Nd-Sr Isotopic Study of Magmatic Rocks and 40 Ar/ 39 Ar Dating of the Maﬁc Dike of the Proterozoic Ulan-Sar’dag Ophiolite M é lange (Southern Siberia, East Sayan, Middle Belt, Russia)

: We report the ﬁrst radiogenic Nd-Sr isotope data in the magmatic rocks island-arc ophiolite assemblage from the middle branch of the East Sayan ophiolite complexes to better constrain geodynamic processes in this segment of the CAOB in southern central Siberia. The magmatic rocks belong to the following geochemical types: (1) Ensimatic island-arc boninites; (2) island-arc assemblage; (3) enriched basalts of mid-ocean ridges; and (4) oceanic island-like basalts. The boninites have a positive value ε Nd (T), which is generated from a depleted mantle source (N-MORB). The island-arc assemblage has negative or slightly positive ε Nd (T) and was formed from an enriched mantle source due to the subduction of terrigenous rocks. The source of the terrigenous material was most likely the rocks of the Archean TTG (Trondhjemite Tonalite Granodiorite) complex of the Gargan block. Isotopic ratios for E-MOR and OIB-like basalts are characterized by positive or slightly negative values of ε Nd (T). The maﬁc dike, which crosscut ophiolite rocks, corresponds to OIB-like basalts. The values of ε Nd (T), measured 87 Sr/ 86 Sr and I (Sr), in the maﬁc dike correspond to the EM I mantle source. The E-MOR and OIB-like basalts appear to be formed in late-stage asthenospheric mantle melting via the decompression melting processes. The obtained isotope geochemical data for the E-MOR and OIB-like basalts probably indicate the mixing of island-arc melts with asthenospheric melts. We undertook 40 Ar/ 39 Ar dating of the maﬁc dike, which crosscut the ophiolite unit. The maﬁc dike has a whole-rock 40 Ar/ 39 Ar weighted mean plateau age of 799 ± 11 Ma. The dating constrains the minimum age of the ophiolite and island-arc magmatism in the region.

The Ulan-Sar'dag ophiolite mélange (for simplicity called Ulan-Sar'dag massif-USM) belongs to the middle and possibly independent ophiolitic branch [34]. Magmatism of the studied area has been controlled by the evolution of the island arc and subduction zone. Isotope geochemical studies of the Ulan-Sar'dag massif will provide important information about mantle sources, mantle-crust interaction, and the evolution arc magmatism in the studied segment of the CAOB.
Based on the geochemical data, we recently discussed the Ulan-Sar'dag massif composed of magmatic rocks with different geochemical signatures [26], which formed due to multistage magmatism, according to the evolution of the subduction zone in the studied segment of CAOB.
This work aimed to study the Sr-Nd isotope characteristics for each geochemical type's rocks of the Ulan-Sar'dag massif. Radiogenic Sr-Nd isotopic and trace elements geochemistry should display contributions of different magmatic sources and melts, mantle-crust interaction, and metamorphic processes. We also determined the ages of the mafic dike, which crosscuts the ophiolites, by the 40 Ar/ 39 Ar geochronology method that constrains the minimum age of the ophiolite in the region implying that all accretion ended in the early Neoproterozoic (late Tonian). The isotope geochemical and some geochronological data offer the opportunity to better constrain geodynamic processes in this segment of the CAOB in southern central Siberia.

Geological Settings
The Ulan-Sar'dag massif is located in the inner part of the Gargan block ( Figure 2). The north ophiolite branch includes the Dunzhugur massif, situated at the Gargan block s western margin, and the northern part of the Ospa-Kitoi massif ( Figure 2). The Dunzhugur ophiolite assemblage is associated with the same name Dunzhugur island-arc complex and two generations of basic dikes, intruding of the Dunzhugur ophiolite [20,29]. The plagiogranite and volcaniclastic rocks in the Dunzhugur island-arc complex yielded an age of 1020-850 Ma [4,20,21] using U-Pb and Pb/Pb zircon dating. The evolution of the Dunzhugur island-arc complex continued during the period of 850-760 Ma [4,20,21]. A collision of the Dunzhugur island arc and the continental margin of the Gargan block occurred at about 810 Ma [4].
The south ophiolite branch includes the southern part of the Ospa-Kitoi massif. At the Ospa-Kitoi, ophiolitic serpentinite mélange is unconformably overlain by dolomites of the Gorlyk suite, with a late Neoproterozoic-early Cambrian age with the contact between them marked by ophiocalcites [4,6,8]. It is the only stratigraphic evidence for a late Neoproterozoic age of the ophiolitic assemblage.
The ophiolite assemblage of the Ulan-Sar'dag massif (USM)) is associated with the volcanic-sedimentary sequence (Ilchir suite), Proterozoic Carbonate Cover (Irkut suite), syncollisional Sumsunur tonalite-trondhjemite complex with an age of 790 Ma [19] (Figure 3). The ophiolite assemblage and volcanic-sedimentary sequence are presented as independent tectonic slices. The units are tectonically obducted on the Gargan block. The age of the trachyandesite from the volcanogenic sequence of the USM determined by U-Pb dating of the zircon is 833 ± 4 Ma [26]. The Ulan-Sar'dag massif is a lenticular body elongated in an eastwest direction with dimensions of 2 × 5 km 2 . It occurs as an ophiolitic mélange, composed of various structurally mixed different ages, origins, lithostratigraphic units. These units are represented by ophiolitic altered ultramafic-mafic rocks with podiform chromitites, foliated serpentinites, deep sea sedimentary cherts, shaled volcanic-sedimentary (Ilchir suite) rocks with MORB, island-arc and OIB-like geochemical affinities, mafic dikes, and limestones (Irkut suite) (Figures 3 and 4) [24][25][26]. The volcanic-sedimentary slice is composed of two units: lower-metavolcanic unit and upper-black, black-quartz-carbonate schists and marbles. The volcanic-sedimentary slice is thrust on the cover limestones of the Irkut suite. The ophiolite slice is thrust on the volcanic-sedimentary units. The rocks at the base of the tectonic ophiolite slice are intensively deformed and have numerous shear zones and sliding surfaces. The thrust zone is marked by foliated serpentinite (Figure 4). On the eastern flank of the Ulan-Sar'dag massif, the mafic dikes crosscut the ophiolite unit ( Figure 4). Dikes have a northeastern extension, with a length of 650-700 m and a thickness of up to 10 m (azimuth 320-330 • , angle 75 • ). At the border with the dunite, chilled margins exist with a thickness of up to 1.5 m.

Sample Description
The ophiolite rocks and the lower units of the volcanic-sedimentary sequence (Ilchir suite) were studied in term geochemistry [26] and isotope characteristics.
The analyses of the whole-rock major, trace, and rare-earth element compositions were carried out at the Analytical Center for Multi-Elemental and Isotope research (VS Sobolev Institute of Geology and Mineralogy, Novosibirsk, Russia). According to major and trace element compositions, four geochemical types of the magmatic rocks were identified in the area of the USM. The types are (1) boninites; (2) island-arc assemblage; (3) E-MOR basalts; (4) oceanic island-like basalt. Each type is characterized by its petrography, concentration of major, LILE, REE, and HFSE elements, and trace element ratios. Detailed petrographic and geochemical characteristics of rocks are given in the previous work [26]. For isotopic characterization, we selected samples from each rock group (Table 1). Abbreviations: Ol-olivine, Hyp-hypersthene, Di-diopside, Pl-plagioclase, Or-orthoclase, Npn-nepheline, Qz-quartz, Ap-apatite, Zrn-zircon, Chr-chromite, Mag-magnetite, Ilm-ilmenite, Ttn-titanite.

40Ar/39Ar Isotope Analytical Method
The 40Ar/39Ar dating was carried out at the Analytical Center for Multi-Elemental and Isotope Research (VS Sobolev Institute of Geology and Mineralogy, Novosibirsk, Russia). The 40Ar/39Ar dating of a ground fraction of the whole-rock sample VSK-17-19 was carried out using the step-heating technique [38]. The samples were irradiated in the nuclear BWR reactor at the Tomsk Polytechnic University. A standard biotite sample MSA-11 (OSO No. 129-88) with an age of 311.0 ± 1.5 Ma, validated against international standard samples Bern-4M (muscovite) and LP-6 (biotite), was used as the fluence mon-itor.
The age of the Bern 4M and LP-6 were assumed to be 18.51 and 128.1 Ma, respec-tively [39]. The Ar isotopic composition in the irradiated samples was measured using a Noble Gas 5400 mass-spectrometer [38] Noble Gas 5400 (Micromass, United Kingdom, Manchester) mass-spectrometer S.N. SV022). The separation of gas fractions and the isotopic analysis of the argon in them was carried out in a temperature range from 500 to 1200 • C using a quartz furnace with external heating. The raw data were corrected for procedural blank contributions, mass discrimination by analysis of atmospheric Ar, de-cay of radioactive 37Ar and 39Ar isotopes produced by irradiation, and interferences of 36Ar, 39Ar, and 40Ar produced from 40Ca, 42Ca, and 40K, respectively. When correcting for isotopic mass fractionation, a value of 40Ar/36Ar in atmospheric argon was used equal to 298.6 [40]. For correction, the following coefficients were used: (39Ar/37Ar)Ca = 0.001279 ± 0.000061, (36Ar/37Ar)Ca = 0.000613 ± 0.000084, (40Ar/39Ar)K = 0.0191 ± 0.0018. The amount of 40Ar, 39Ar, 38Ar, 37Ar, 36Ar in the blank was normally about 3 × 10−10, 2 × 10−12, 2 × 10−11, 1 × 10−12, 9 × 10−12 cm3 STP, correspondingly. The errors in individual measurements for each of the steps in the spectrum are given taking into account the error, including both the analytical error and the error in determining the decay constant K. The isotopic ratio of 40Ar/36Ar characterizing instrument fractionation in a standard portion of air argon for the measurement period was 298 ± 2. The plateau ages were calculated using the Isoplote 3.0 program (the program was created in 2006 by K.R. Ludwig at the Berkeley Geochrono-logical Center, San Francisco, USA) if 60% or more of the 39Ar was released in three or more contiguous steps, and the apparent ages agree to within 3σ of the integrated age of the plateau segment.

Sm-Nd Isotope Analytical Method
The whole-rock isotope research was carried out in the Collective Use Centre of the Kola Science Centre RAS (Apatity, Russia). In order to define concentrations of samarium and neodymium, the sample was mixed with a compound tracer 149 Sm/ 150 Nd prior to dissolution. It was then diluted with a mixture of HF + HNO 3 (or + HClO 4 ) in Teflon sample bottles at a temperature of 100 • C until complete dissolution. Further extraction of Sm and Nd was carried out using standard procedures with two-stage ion exchange and extractionchromatographic separation using ion-exchange tar «Dowex» 50 × 8 in chromatographic columns employing 2.3 N and 4.5 N HCl as an eluent. The separated Sm and Nd fractions were transferred into nitrate form, whereupon the samples (preparations) were ready for mass-spectrometric analysis. The isotope Nd composition and Sm and Nd contents were measured on a 7-channel solid-phase mass-spectrometer (Finnigan MAT-262 (RPQ), "Thermo Fisher Scientific", Waltham, MA, USA), in a static double-band mode, using Ta + Re filaments. A mean value of 143 Nd/ 144 Nd ratio in a JNd i -1 standard was 0.512081 ± 13 (N = 11) in the test period. The error in 147 Sm/ 144 Nd ratios was 0.3% (2σ), which is a mean value for 7 measurements in a BCR-2 standard [41]. The blank intralaboratory contamination was 0.3 ng for Nd and 0.06 ng for Sm. The accuracy of estimation of Sm and Nd contents was ± 0.5%. Isotope ratios were normalized per 146 Nd/ 144 Nd = 0.7219, and then recalculated for 143 Nd/ 144 Nd in J Ndi -1 = 0.512115 [42]. Values of ε Nd (T) and T (DM) model ages were estimated using present-day values of CHUR after [43] at ( 143 Nd/ 144 Nd = 0.512630, 147 Sm/ 144 Nd = 0.1960) and DM after [43] ( 143 Nd/ 144 Nd = 0.513151, 147 Sm/ 144 Nd = 0.21).

Rb-Sr Isotope Analytical Method
The whole-rock isotope research was carried out in the Collective Use Centre of the Kola Science Centre RAS (Apatity, Russia). Distilled acids HCl, HF, HNO 3 , and H 2 O (distillate) were used to decompose samples. Then, the produced solution was divided into two aliquots to estimate the isotope composition and contents of Rb and Sr. The latter were estimated by isotope dissolution using an 85 Rb/ 84 Sr mixed tracer. Rb and Sr were extracted by elution chromatography on Dowex resin 50 × 8 (200-400 mesh) and 1.5 N and 2.3 N of HCl were used as eluates. The amount of resin in the used columns was 7 cm 3 and~4 cm 3 . Extracted Rb and Sr concentrates were evaporated until dry and then treated with several drops of HNO 3 . The isotope composition of Sr and contents of Rb and Sr were estimated on an MI-1201-T mass spectrometer MI-1201T "Electron", Sumy, Ukraine in a double-band mode using Ta + Re filaments. Prepared samples were placed onto the bands in a nitric form. The isotope composition of Sr in all of the measured samples was normalized per 0.710248 [44]. Errors in the estimation of the Sr isotope composition (95% confidence interval) were not higher than 0.04%; errors in the estimation of Rb-Sr ratios were 1.5%. The blank intralaboratory contamination in Rb and Sr was 2.5 ng and 1.2 ng, respectively. Correction of mass fractionation for Sr isotopic ratios was based on an 88 Sr/ 86 Sr value of 8.37521 using a power law.

40 Ar/ 39 Ar dating of the Mafic Dike
In order to constrain the minimum age of the Ulan-Sar'dag ophiolite mélange, a whole rock sample (VSK- [17][18][19] was taken from a gabbro-diabase dike by cutting it for 40 Ar/ 39 Ar step-heating dating. The dating was performed by 40 Ar/ 39 Ar by step heating. The obtained age and Ca/K spectra of the sample are shown in Figure 6. In the age spectrum, a conditioned plateau was distinguished, characterized by 63% of the cumulative 39 Ar and a weighted mean plateau age (WMPA) value of 799 ± 11 Ma. Among the minerals that make up the dike rock, amphibole was the main potassium-containing mineral, with a minimum amount of biotite. Analyzing the nature of radiogenic argon release during step-heating, it can be noted that minimal amounts were released at lowmedium temperatures (600-700 • C), at which biotite decomposition usually occurs in a vacuum. Most of the radiogenic argon is released at high temperatures (930-1100 • C), at which the crystal lattice of the amphibole is destroyed. Thus, it is logical to assume that the plateau observed in the age spectrum corresponded to the closure of the amphibole K-Ar isotope system. Considering that the introduction of dikes took place in shallow conditions and was accompanied by their rapid cooling, the dating obtained by the plateau method corresponded to the age of formation of the mafic dike. The 40 Ar/ 39 Ar weighted mean plateau age of 799 ± 11 Ma constrains the minimum age of the ophiolite mélange in the region, implying that all accretion ended in the early Neoproterozoic (late Tonian).

Geochemical Features and Nd-Sr Isotopes
The main and rare chemical composition of the studied rocks is presented in Table  S1. A detailed geochemical characteristic of the magmatic rocks of the Ulan-Sar'dag ophiolite mélange is given in a previous study [26]. According to a previous study, the magmatic rocks belong to the following geochemical types (Figure 7): (1) ensimatic island-arc boninites; (2) island-arc assemblage; (3) tholeiitic basalts of mid-ocean ridges; and (4) oce-

Geochemical Features and Nd-Sr Isotopes
The main and rare chemical composition of the studied rocks is presented in Table S1. A detailed geochemical characteristic of the magmatic rocks of the Ulan-Sar'dag ophiolite mélange is given in a previous study [26]. According to a previous study, the magmatic rocks belong to the following geochemical types (Figure 7): (1) ensimatic island-arc boninites; (2) island-arc assemblage; (3) tholeiitic basalts of mid-ocean ridges; and (4) oceanic island-like basalts. The main and rare element chemical composition of the studied rocks is given in Table S1. Further, we present the isotopic characteristics for each group of rocks. Sixteen whole-rock samples were analyzed for Sr and Nd isotopic compositions. The results are given in Table S2 and plotted in Figure 8. For all samples, except for the mafic dikes, the ε Nd (T) and I Sr (T) ratio was calculated to the date 833 Ma, and for the mafic dikes-799 Ma.    [26]; normalizing values are from [45]: (a,b) Group I-boninite USM and literature data [46]: Low calcium (LCB), medium calcium (ICB), high calcium (HCB) boninites; (c,d) group II-island-arc assemblage, Sk-Dunzhugur ophiolite massif, G-gabbro, LG-leyko gabbro, literature data [29]; (e,f) group III geochemical type-E-MOR basalts, Sk-Dunzhugur ophiolite massif, mafic dike literature data [29]; (g,h) group IV-OIB-like type geochemical group.  [26]; normalizing values are from [45]: (a,b) Group I-boninite USM and literature data [46]: Low calcium (LCB), medium calcium (ICB), high calcium (HCB) boninites; (c,d) group IIisland-arc assemblage, Sk-Dunzhugur ophiolite massif, G-gabbro, LG-leyko gabbro, literature data [29]; (e,f) group III geochemical type-E-MOR basalts, Sk-Dunzhugur ophiolite massif, mafic dike literature data [29]; (g,h) group IV-OIB-like type geochemical group. The current ratios for the depleted mantle and CHUR were used to calculate ε Nd (T) and T (DM) [43]. The first geochemical type-ensimatic island-arc boninites (La/Sm)n = 1.

Discussion
Based on the isotope geochemistry data, four geochemical types of rocks were distinguished: (1)-boninites, (2)-island-arc assemblage, (3)-EMOR basalts, (4) transitional to basalts of oceanic islands (OIB-like) [26] in the magmatic rocks of the Ulan-Sar'dag massif. According to the study of 147 Sm/ 144 Nd and 87 Rb/ 86 Sr isotopic systems, igneous rocks of the Ulan-Sar'dag massif are characterized by a significant dispersion of the values of isotopic parameters. The values of 147 Sm/ 144 Nd varied from mantle (0.18) [47,48] to crustal (0.10-0.13) [49] values, and ε Nd (T)-from +0.17 to −12 (Table S2). Most of the rocks of the Ulan-Sar'dag massif had crustal values of 87 Sr/ 86 Sr (Table S2) and I Sr (T) calculated to the date 833 Ma, except a few samples (gabbro, basalt, andesite) and samples from the mafic dike with mantle (but elevated) values, variations I Sr (T) from 0.703 to 0.705 (Table S2)

Constraints on Mantle Source
High-magnesium basaltic-andesite of the first geochemical type (boninite) has an REE distribution corresponding to boninites (Figure 7a,b), with low Ta/Yb, Th/Yb, (La/Sm) n ratios close to N-MOR values, high Cr content > 500 ppm, (Th/La)PM < 1 [26]. Low REE concentrations and negative anomalies of HFSE (Nb, Ta, Ti) are typical for boninites [46]. In the discrimination diagrams, high-magnesium basaltic-andesite lies in the field of islandarc rocks and boninites [26]. Isotopic geochemical characteristics of the boninite: mantle value 147 Sm/ 144 Nd and positive value ε Nd (T) = +5.23 ( Figure 8, Table S2) confirm formation from a depleted mantle source with a long prehistory for this group of rocks. Model age T (DM) 1.4 Ga can be accepted as the age of the mantle source and possible as the forming age of an ensimatic island arc. Adjacent age was obtained for plagiogranites of the Dunzhugur ophiolite massif-1020 Ma and 1035 Ma-for Ilchir gabbro [20,33]. The 87 Sr/ 86 Sr values in the boninite were very high; both measured (0.717) and calculated to the date 833 Ma I Sr (T) = 0.7157 and are comparable to the values in seawater, shales, gneiss [49][50][51]. The values 87 Sr/ 86 Sr in seawater NP, vary from 0.705 to 0.709 range [52]. The values 87 Sr/ 86 Sr in protoplatform carbonate sediments-dolomites of Irkut suite, range from 0.705 to 0.709 [53]. The value of 87 Sr/ 86 Sr in the boninite of the USM was close to 87 Sr/ 86 Sr in boninites from the Ospa-Kitoi and Dunzhugur massifs with 87 Sr/ 86 Sr = 0.713 [32].
Gabbros, basaltic-andesite, andesites, dacites, and rhyolites of the second geochemical type have a slightly pronounced negative slope in the REE pattern (Figure 7c,d) and the distributions of REE, HFSE, and LILE, which fully corresponded to the distributions of the island-arc rocks. All rocks from this group had a negative anomaly for Ta, Nb, and Ti. In several samples, positive anomalies in Sr, Ba, Rb, and Zr were presented. The second group had a wide spectrum of island-arc geochemical affinities [26]. High Cr content > 800-500 ppm was established in the rocks of mafic composition. In andesites, the Cr content was much lower. Relationships Sr/Y, (Th/La) PM , which reflects the involvement of a subduction (crustal) component in the melt [54], varied from 5 to 47. It is supposed that there was no subduction component at Sr/Y < 25 (Table S1). Relation (Th/La) PM > 1 had values 1.4-5 (Table S1), except one peridotite sample. The rocks of the second group were characterized by an increased content of Th/Yb = 2-4 (Table S1), which is typical for modern subduction basalts [55]. Rocks of the second geochemical group had signs of participation of a crustal component in the magmatic source. The inverse relationship between (La/Sm) PM -(Nb/La) PM (Figure 9), as for (Th/La) PM > 1, testified in favor of the enrichment of the mantle source due to the subduction of metaterrigenous rocks (crustal material). Participation of the crustal material in the magmatic source has also been confirmed by Sm-Nd data of NP terrigenous rocks [56] and granite gneisses of the Gargan block (unpublished data by Turkina). For almost all samples from the second group, crustal values of 147 Sm/ 144 Nd of 0.10-0.12 were observed, and only in the gabbro sample, this value was 0.15. Model age T (DM) varied 1.5-2.3 Ga. This group was characterized by negative ε Nd (T) and high crustal values ( 87 Sr/ 86 Sr) and I Sr (T) having a large spread (Figure 8). The effect of the mature continental crust on the geochemical and isotopic compositions is established in many ophiolitic complexes in orogenic belts and magmatic rocks in igneous provinces [57,58], the Altai-Sayan region [59], Central Asian Fold Belt [60,61]. At the mixing of two magmatic sources and with crustal component participation, a negative correlation between (La/Sm)-εNd(T) was established ( Figure 10) [62]. Magmatic rocks of the second group have lain in the area of the mixing line of the two magmatic sources in the diagram ε Nd (T)-(La/Sm) ( Figure 10). Boninite was used as the last member, which is supposed to have a depleted mantle source. In addition, granite gneisses TTG with a low ε Nd (833 Ma) = −23 value (unpublished data by Turkina O.M.), which provides a decline in the ε Nd (T) value for magmatic rocks at the low contribution of sediment matter was used. The contribution of various sources to the formation of igneous rocks was calculated using the binary mixing model [63]. The proportion of crustal components for the second group of rocks varied from 14 to 45% (Figure 11).  The high values of 87 Sr/ 86 Sr in all samples of the second group (gabbro, andesites), except for basalt and mafic dike, were comparable with such values in seawater, shales, gneiss. On the one hand, the 87 Rb/ 86 Sr isotopic system was more susceptible to the influence of metamorphogenic processes, and the safety of the system may be disturbed. On the other hand, the wide range and high values of 87 Sr/ 86 Sr may reflect the contribution of several sources: altered oceanic crust, terrigenous sediments of oceanic crust, mantle wedge above the subducting plate, sialic rocks of continental crust, or volcanic strata bases in island arcs [51,64,65]. During the ascent, the parental magmas could interact with rocks of the continental crust and sediments of the subducting slab. In the diagrams, 87 Sr/ 86 Sr-143 Nd/ 144 Nd, εNd 0 -87 Sr/ 86 Sr (Figures 12 and 13), almost all samples fell into the field of continental crust, except for one sample, Nd-Sr isotopic ratios of which reflect either the characteristics of a magmatic source or its isotopic system were influenced by a younger mafic dike. The basalt (eastern flank of the massif, near the area of the mafic dike occurrence) from the second geochemical group in the 87 Sr/ 86 Sr-143 Nd/ 144 Nd diagram lies in the field of the EM I source (Figure 12). In diagram εNd 0 -87 Sr/ 86 Sr (Figure 13), basalt lies on the mixing line of the Prevalent mantle, upper continental crust, and Pacific pelagic sediments. Positive values of ε Nd (T) in the basalt indicate a juvenile source of these rocks. The ε Nd in comparison with the value of ε Nd (833 Ma) for DM = (+8.2), may indicate the participation of melts from enriched sources in the generation of these rocks, or the influence of the introduced mafic dike (the fourth geochemical group) on the isotopic system of the basalt. On the tectonic discrimination diagram, the rocks of the second group are plotted in the fields of a volcanic arc, island-arc tholeiite, calc-alkaline basalts, and continental arc [26]. According to geochemical and isotopic parameters, rocks of the second group were generated from an enriched magmatic source due to a subduction component. The source of the terrigenous-sedimentary material was most likely rocks of the Archean TTG complex of the Gargan block ( Figure 11).  The geochemical characteristics of the basalts and one gabbro sample from the third group of rocks corresponded to E-MOR basalts, except for LILE elements (Cs, Rb, Ba, and U) [26]. Sr/Y values < 25, (Th/La) PM < 1, suggested the absence of a subduction component in the melt (Table S1), although elevated concentrations of Rb, Ba, Sr (Table S1) indicated the enrichment of the mantle source with incoherent elements. In general, the rocks had hybrid characteristics of N-MOR, E-MOR basalts. Basalt from this group had mantle values of 147 Sm/ 144Nd = 0.17, unlike basalt from the second geochemical group. The value of ε Nd (T) was positive (+3) (Table S2), which may indicate the origin of this group of rocks from a poorly enriched mantle source. The positive ε Nd (T) implied an injection of the upwelling asthenosphere into the arc segment, which would have provided a fertile mantle. A high value of 87 Sr/ 86 Sr = 0.710, I Sr (T) = 0.709, was recorded in basalt, which was comparable to the values in Proterozoic shales, gneiss. On the tectonic diagrams, rocks of this group fell into the fields of MOR and E-MOR basalts [26].
According to geochemical characteristics, the fourth geochemical type corresponded to ocean island-like basalt [26] (Figure 7g,h; Table S1). The rocks were enriched with both LREE and HREE, HFSE. The Sr/Y < 25, (Th/La) PM < 1, had elevated concentrations of Ta, Th, Yb, which may indicate the involvement of astenospheric melts [26,[72][73][74][75]. According to 147 Sm/ 144 Nd systematics, the fourth group was close to the second group, having island-arc geochemical characteristics. The 147 Sm/ 144 Nd value had crustal values (0.13) in both trachybasalt and the mafic dike. The values of ε Nd (T) for trachybasalt were close to chondrite (+ 0.11), for mafic dike, calculated to the date 799 Ma ε Nd (T), and fit into a small range from close to chondrite to weakly negative values (+0.14)-(−2.2). The calculated model age for trachybasalt T(DM) was 1.5 Ga, for mafic dike, 1.5-1.8 Ga (Table S2). In the diagram (La/Sm) PM -ε Nd (T), the rocks of the third and fourth geochemical groups lie between boninite and island-arc assemblage ( Figure 10). The values of I Sr (T) in the fourth group had a mantle but elevated values of 0.703-0.706. Isotope ratios in the mafic dikes (the fourth geochemical group) in the diagrams, 87 Sr/ 86 Sr-143 Nd/ 144 Nd, εNd 0 -87 Sr/ 86 Sr plotted in the field of the mantle source EM I are shown in Figures 12 and 13. Isotope geochemical characteristics of the third and fourth groups are probably likely indicative of a mixture of island-arc melts with asthenospheric melts. Geodynamic processes could be associated with the rollback of the subducting oceanic lithosphere, slab breakoff, upwelling asthenosphere, and decompression melting. It would produce E-MORB and OIB-like magma with high 87 Sr/ 86 Sr and low 143 Nd/ 144 Nd values [73][74][75]. Further, these processes could lead to an extension of the Dunzhugur island arc that can be attributed to intra-arc rifting and subsequent formation the Shishkhid back-arc basin [4].

Comparison Isotope Data of Ulan-Sar'dag Massif with Ospa-Kitoi and Dunzhugur Massifs
Comparing the isotopic and geochemical characteristics of the ophiolite rocks and associated island-arc assemblage of the Ulan-Sar dag massif and the mafic rocks closest to them in geochemical characteristics (island-arc type) from the ophiolite association of the Dunzhugur and Ospa-Kitoi ophiolite massifs, it can be seen that the gabbro of the Dunzhugur massif had increased values (La/Sm) n from 1.7 to 4, (Th/La) n > 1 (1.2-3). It was similar to the rocks in the second group of the Ulan-Sar dag massif array, indicating a crustal component contribution in the result of subduction settings. The calculated ε Nd (T) for the age accepted for the Dunzhugur island arc of 1020 Ma had a small spread of values, fitting into the range (+1)-(−1) and, accordingly, a smaller range of values of T (DM) 1.7-2.2 Ga, which also confirmed the contribution of the crustal component (Figures 9 and 10). The geochemical characteristics of the Dunzhugur massif diabase dikes were similar to the geochemical characteristics of basalts group III. They were characterized by low, but positive values of ε Nd (T). 147 Sm/ 144 Nd ratios had both crustal (0.12), and mantle (0.17) values [29]. At the same time, the mafic dike of the USM, in contrast to the diabase dikes of the Dunzhugur massif, contained high concentrations of REE, HFSE, Th.
Cumulative and volcanic rocks of the Ospa-Kitoi massif had an ε Nd (1020) from +4.3 to −2 [32] comparable to the isotopic parameters of magmatic rocks of the USM but differed from them in geochemical affinities.

Isotopic Evidence of Crust-Mantle Interaction
Isotopic data obtained for various complexes of the Eastern Sayan of the Central Asian Folded Belt [60] (Table 2) indicated a significant contribution of ancient crustal material to the genesis of the ophiolite and island-arc rocks of the Eastern Sayan, suggesting a model of mixing of the Neoproterozoic juvenile source and Archean crustal source (tonalitetrondhjemite gneisses of the Gargan block), which will give a wide range of ε Nd (T) values from negative to positive values, depending on the proportion of crustal material.  A perfect example of contamination crustal matter and its influence on the isotopic characteristics of rocks formed from mantle sources is the Dovyren ultrabasite-basite stratified massif (quite well preserved) [65] which has an extensive range of values of I Sr (T) 0.709-0.715. Isotopic characteristics for peridotites and basic rocks of the Dovyren massif [65] [65], the change processes for the following reasons do not cause all variations of 87Sr/86Sr: in peridotites, there is more than a 30-fold variation of Sr, but they show close I Sr (T) values. If one supposes variations in 87 Sr/ 86 Sr ratios are the result of changes, then a hypothetical mechanism should provide a certain fluid-rock ratio. The concentration of Sr in the fluid is proportional to the concentration of Sr in the rock with which the fluid interacts [65].

The Stages of Magmatism of the Ulan-Sar'dag Massif
According to the isotope geochemical data of the Ulan-Sar'dag massif, we propose three stages of magmatism, reflecting the evolution of the arc magmatism and subduction zone in the studied segments of CAOB ( Figure 14).
Stage I. In the Meso-Neoproterozoic, the intra-oceanic ensimatic island-arc system was formed. Boninite has positive ε Nd (T) and is formed from a depleted mantle source. The high value of I Sr (T) is due to the disturbance of the closure 87 Rb/ 86 Sr isotopic system: interaction with seawater and metamorphic processes.
Stage II. During the development of an ensialic island arc, boninite magmatism turned to calc-alkaline andesite magmatism at Neoproterozoic, which occurred with the involvement of subducted sediments both oceanic and terrigenous (Proterozoic gneiss of the Gargan block) to the magma generation chamber, which leads to a dispersion of isotopic parameters, negative values of ε Nd (T), and the ancient of model ages, and high I Sr (T).
Stage III. The subducting oceanic lithosphere had been proceeded by rollback, slab breakoff, upwelling asthenosphere, and decompression melting to produce E-MOR and OIB-like basites. Mixing asthenospheric melts with island-arc magmas can lead to enrichment of melts with radiogenic Nd and Sr and correspond to EM1 mantle source. The Ulan-Sar'dag Massif is a member of the Proterozoic ophiolite and island-arc systems in the CAOB. They have subduction geochemical characteristics and were formed on the Neoproterozoic active continental margin of the Siberian paleocontinent. Proterozoic ophiolite and island-arc systems are associated with the evolution of the Paleo-Asian ocean in the period 1050-790 Ma. The island-arc assemblage of the Ulan-Sar'dag massif has geochemical characteristics of both ensimatic and ensialic island arcs. The stage of the initiation of ensimatic island arc magmatism in the Paleo-Asian ocean occurred in the time fence of 1050-790 Ma and is most widely developed in the Eastern Sayan, Northern and Eastern Baikal Lakeside, Mongolia [2][3][4]6,7,9,20,79,80]. The initiation of ensialic arc magmatism on the Neoproterozoic active margin occurred 830-780 Ma (Eastern Sayan, Central Transbaikalia, Mongolia) [21,56,60,77]. The formation of plumes associated with the melting of slabs plunging into the subduction zone and the participation of asthenospheric melts in island-arc magmatism occurred at the boundary of 820-770 Ma in almost all Neoproterozoic island-arc systems with the formation of subalkaline-alkaline magmatic series [79,80]. The collisional granites formed at the 750 Ma boundary recorded collisionalorogenic events and accretion of island arcs at the Proterozoic stage [4,19].

1.
The Ulan-Sar'dag massif consists of various structurally combined material blocksophiolite ultramafic-mafic rocks with podiform chromitites, cumulative gabbro, foliated serpentinites, island-arc assemblage, mafic dikes. Four isotope geochemical types of the magmatic rocks of the USM have been established.

2.
The ophiolites of the Eastern Sayan belong to the supra subduction type. The first isotopic data on the magmatic rocks of the USM was obtained. Isotope geochemical characteristics of the magmatic rocks of the USM reflect the evolution of island arc magmatism and subduction zone. The boninite was formed from a depleted mantle source, and the island-arc assemblage was originated from an enriched mantle source which results from the subduction of terrigenous matter. The calculated proportion of the crustal component was from 14% to 45%. The source of the terrigenous material was most likely the rocks of the Archean TTG complex of the Gargan block. The E-MOR and OIB-like basalts and mafic dike were probably generated by mixing island-arc melts with asthenospheric melts produced from upwelling asthenosphere and decompression melting.

3.
A whole-rock 40 Ar/ 39 Ar weighted mean plateau age of 799 ± 11 Ma constrains the minimum age of the ophiolite mélange in the region, implying that all accretion ended in the early Neoproterozoic (late Tonian).

4.
The Ulan-Sar'dag Massif is a member of the Proterozoic ophiolite and island-arc systems in the CAOB. They have subduction geochemical characteristics and were formed on the Neoproterozoic active continental margin of the Siberian paleocontinent. Proterozoic ophiolite and island-arc systems were associated with the evolution of the Paleo-Asian ocean in the period 1050-790 Ma. The island-arc assemblage of the Ulan-Sar'dag massif has geochemical characteristics of both ensimatic and ensialic island arcs. The stage of the initiation of ensimatic island arc magmatism in the Paleo-Asian ocean occurred in the time fence of 1050-790 Ma and is most widely developed in the Eastern Sayan, Northern and Eastern Baikal Lakeside, Mongolia. The initiation of ensialic arc magmatism on the Neoproterozoic active margin occurred 830-780 Ma (Eastern Sayan, Central Transbaikalia, Mongolia). The formation of plumes associated with the melting of slabs plunging into the subduction zone and the participation of asthenospheric melts in island-arc magmatism occurred at the boundary of 820-770 Ma in almost all Neoproterozoic island-arc systems with the formation of subalkaline enriched alkaline magmatic series. The collisional granites formed at the 750 Ma boundary recorded collisional-orogenic events and accretion of island arcs at the Proterozoic stage.
Supplementary Materials: The following supporting information can be downloaded at https:// www.mdpi.com/article/10.3390/min12010092/s1, Table S1: Chemical composition of the magmatic rocks of the Ulan-Sar'dag ophiolite mélange used for isotope study, Table S2: The Nd-Sr isotope data of the magmatic rocks from the Ulan-Sar'dag ophiolite mélange.