The Source of Ore Fluids and Sm–Nd Age of Siderite from the Largest Bakal Deposit, Southern Urals

Based on Sm–Nd data, a crustal source of iron-ore fluid was substantiated and the probability of age estimation for hydrothermal–metasomatic siderite of the Bakal Group, Southern Urals, was shown for the first time. The εNd (Т) values of siderite (from –13.4 to –17.6) plot in the field of Riphean shale and not the Precambrian rift gabbro and granite of this region. The obtained Sm–Nd age of the Bakal siderite is 970 ± 40 Ma, which is consistent with the Pb–Pb age of siderite from the major ore phase (~1000 Ma). The established age boundary coincides with tectonic restructuring, including the formation of a number of barite–polymetallic deposits, as well as ferruginous magnesite and fluorite in the Riphean deposits on the western slope of the Southern Urals.


INTRODUCTION
The geological study of the Bakal siderite deposits, which are a typical object of siderite hydrothermalmetasomatic deposits in sedimentary carbonate rocks (Smirnov, 1976), is considered in a huge number of papers and of monographs (Yanitskii and Sergeev, 1962;Timeskov, 1963;Varlakov, 1967;Anfimov et al., 1984;Kuznetsov et al., 2005;Kholodov and Butuzova, 2008;Krupenin, 2017). The group of Bakal iron deposits in the Southern Urals is a convenient testing ground for solving the problem of the genesis of iron ore formations in Precambrian sedimentary rocks. In explored reserves (more than 1.2 bln t, Yanitskiy and Sergeev, 1962) and production volume, the Bakal ore field is on par with the well-known siderite giants, such as Erzberg in Austria, Bilbao in Spain, and Quenza in Algeria (Pohl et al., 1986;Frimmel, 1988).
The Bakal deposits have been mined for more than 250 years, since Demidov's time. The long-term controversy about the genesis of Riphean siderite in the Urals during the 20th century ruled out the idea of D.V. Nalivkin (1931) and his supporters (Malakhov, 1957;Druzhinin, 1971;Borshchevskii et al., 1978;Dunaev, 1983) on the lagoonal origin of layered bodies. The set of features makes it possible to classify these ore deposits as hydrothermal-metasomatic (Zavaritskii, 1939;Davydenko, 1956;Yanitskii and Sergeev, 1962;Anfimov, 1997;Krupenin, 1999, etc.). The source of ore fluids remains controversial. A.N. Zavaritskii (1939) and his followers (Yanitskii and Sergeev, 1962;Timeskov, 1963;Varlakov, 1967;etc.) suggested the removal of iron from the magma chamber after large-scale intrusion of gabbroids (doleritic dikes of the Bakal syncline up to 200 m thick; the Kusa-Kopan intrusion with a length of up to 90 km) and granitoids (Berdyaush and Ryabinov massifs) (Timeskov, 1963). The ore-forming process was associated with the Paleozoic Uralian orogen. Later L.V. Anfimov proposed a catagenetic source of iron mobilized by elision solutions from the host shale rocks in the Riphean (Anfimov et al., 1984;Anfimov, 1997), while V.N. Kholodov suggested the input of iron into sedimentary strata from Early Precambrian ferruginous quartzite of the crystalline basement (Kholodov and Butuzova, 2008). According to a study of the Pb-Pb systematics, the age of siderite mineralization was estimated as 1010 ± 100 Ma (Kuznetsov et al., 2005). It is suggested that ore formation is associated with the migration of iron-bearing solutions upsection into carbonate reservoirs along decompaction zones and faults at the tectonic activation stage at the Middle-Late Riphean boundary. As evident from the Rb-Sr systematics of carbonate rocks and ores (Kuznetsov et al., 2005), ore solutions interacted with shale of the sedimentary basin. Chromatographic study of the composition of fluid inclusions, as well as carbon, hydrogen, and oxygen stable isotopes in siderite and host carbonate rocks, has demonstrated the important role of buried brines for the fluid formation in the catagenetic basin (Prochaska and Krupenin, 2013). This paper presents the results of a study of the Sm-Nd isotopic systematics of siderite and host carbonate rocks of the Bakal ore field for the first time, showing the probability of using these data to determine the source of iron ore fluid and to refine the age of metasomatic siderite.

GEOLOGY OF THE DEPOSITS
The world's largest accumulation of siderite iron ores (Bakal deposits, Bakal ore field) is located on the western slope of the Southern Urals and is confined to Lower Riphean carbonate-terrigenous deposits, which formed as a part of the Upper Precambrian sedimentary sequence within the Cis-Uralian pericratonic depression the of eastern part of the continent Baltica (Puchkov, 2010). A thick (more than 15 km) sequence of volcanosedimentary deposits accumulated there over more than 1 Ga (1750-640 Ma). In the Vendian and Late Paleozoic, the Riphean sedimentary rocks of the eastern part of the Cis-Uralian pericratonic depression were involved in fold zones and became part of the Uralian Orogen, forming the Bashkir Meganticlinorium (BMA), which is a large structure within the Central Ural Uplift.
The modern erosion level in the structure of the BMA makes it possible to study the section of typical Riphean sediments (Stratotip rifeya, 1982;Maslov et al., 2001;Semikhatov et al., 2009Semikhatov et al., , 2015. It includes four strats ( Fig. 1): lower (Burzyanian, 1750-1400 Ma), middle (Yurmatinian, 1400-1030± 30 Ma), upper (Karatavian, 1030, and end (Arshinian, 770-640 Ma) (Puchkov, 2010;Puchkov et al., 2017). The Vendian sediments overlying the Riphean rocks in the composition of tillites and terrigenous Asha Group were deposited 640-548 Ma (Grazhdankin et al., 2011;Zaitseva et al., 2019). The Riphean series has a regular structure with a coarseclastic base and terrigenous-carbonate fill in the middle and upper parts. In addition, the Burzyanian and Yurmatinian rocks contain rift volcanics at the base (Ai and Mashak formations, respectively); the Arshinian volcanic rocks are confined to the middle part of the terrigenous section (Igonin Formation). A number of anticlinoria and synclinoria with a northeastern strike are distinguished in the tectonic structure of the BMA; the boundaries between them are complicated by thrusts and upthrusts. Burzyanian deposits of the Ai Formation overlie AR-PR 1 crystalline schist and gneiss of the Taratash Complex in the northern part of the BMA (Taratash Anticlinorium). Igneous formations in the BMA are mainly confined to the zone of the Mashak rift east of the Zyuratkul-Karatash Thrust and are represented by both volcanic rocks (trachybasalt and trachyrhyolite of the Mashak Formation) and more widespread dike swarms in the Burzyanian deposits, a submeridional Kusa gabbroid intrusion consisting of several massifs up to 90 km long, as well as associated plagiogranite (Guben and Ryabinov massifs), and the Berdyaush Pluton of rapakivi granite (Ernst et al., 2006;Kholodnov et al., 2010;Larin, 2011). The volcanic rocks of the Mashak Formation are observed in the eastern part of the BMA from north to south at a distance of up to 200 km, forming a linear rift complex together with dike swarms in sedimentary rocks and intrusive bodies of Kusa gabbroid and granitoid rocks (Parnachev et al., 1986, etc.).
Siderite is associated with Burzyanian deposits (Lower Riphean): the Bakal ore field (near the city of Bakal) in the Bakal Formation and Akhta deposit (20 km east of the city of Kusa) in the Satka Formation. Magnesite of the South Ural Province, including the Satka deposits, the largest in Russia (Krupenin and Kol'tsov, 2017), as well as a number of deposits of polymetals, barite, and fluorite, are associated with Burzyanian and Yurmatinian carbonate deposits. Riphean rocks are altered at the deep catagenesis level, while in the eastern part of the BMA, east of the Zyuratkul'-Karatash Thrust, at the metagenesis and greenschist metamorphism level (Anfimov, 1997).
The sediments of the Zigal'ga Formation are underlain by the volcanogenic-terrigenous Mashak Formation with a thickness of up to 3 km to the east from the Bakal ore field and the regional Zyuratkul'-Karatash Fault traced by the Kusa Intrusion in the Taganaisko-Iremel'skii Anticlinorium. The age of Mashak rhyodacite is 1383 ± 3 Ma (Krasnobaev et al., 2013). Thus, the Bakal ore field is located on the western side of the Mashak rift. its of the Malobakal Subformation is up to 700 m. The ore field is complicated by folds of different orders and is intersected by numerous upthrusts and shears with an amplitude of up to 500 m and a prevailing SW-NE orientation (more than 20), serving as natural boundaries of individual deposits (Yanitskii and Sergeev, 1962). A multi-level structure of ore deposits is observed near the steeply dipping faults: mineralization occurs in 2-3 units below the surface of stratigraphic unconformity within a single deposit. For example, this is typical for the Shikhanskoe and Rudnichnoe deposits developed by the mines, as well as the Irkuskan and Novobakal'skii quarries. Thus, steeply dipping faults are the ore-supplying zone, while the surface of the interformational Prezigal'ga unconformity is an ore-distributing zone (Yanitskii and Sergeev, 1962;Krupenin, , 2017. Large siderite deposits have a sheetlike shape; their thickness reaches tens of meters, and the length is hundreds of meters along strike. The smaller bodies are pocket-and stocklike. Some bedlike deposits reach a length of 2-3 km; their maximum thickness is determined by the thickness of the ore-bearing carbonate unit and may reach 80 m (Shikhanskoe Deposit, Berezov Unit, Irkuskanskoe Deposit, and Gaev Unit). All orebodies have much evidence for superimposed formation after sedimentary carbonate rocks: the boundaries of siderite bodies intersect elements of layering and stromatolite textures. Siderite intersects the crystalline magnesite deposits occurring in the Shuyda Unit (RF 1 b 2 3 ) with an age of ~1370 Ma (Ovchinnikova et al., 2018). Dolomite relics have been recorded in siderite bodies; there are the signs of prefold ore formation; siderite pinches out near marble formation zones, at the contacts with large doleritic dikes and sills; less frequently, sideritization of mafic dikes with preservation of ophitic texture is observed. In addition, there have been observations of individual postore doleritic dikes that intersect preore dikes and form an exomorphic contact zone of finely disseminated magnetite up to 1 m thick in siderites .
Siderite deposits in limestone always occur in a thin zone of ankerite and a wide environ of Fe-dolomite; dolomite relics are observed in siderite bodies. Large siderite deposits are composed of monomineral sideroplesite containing >30% FeO; up to 3-12% MgO, 1.5-2% MnO, and up to 1.5-2% CaO are registered in the form of isomorphic impurities. Bi-and polymineral ores predominate on the flanks of large deposits and in the small bodies: sideroplesites with an admixture of dolomite-ankerite and even calcite (Timeskov, 1963). According to the huge microprobe data obtained for ten different localities in the western and central parts of the Bakal ore field, the contact ankerite shows variations in the average FeCO 3 concentrations of 14.2 and 20.8 wt %, respectively (Krupenin, 2017). At the same time, no pronounced differ-ences were detected in the composition of the contact sideroplesite in the same objects; the content of the FeCO 3 end-member varies in the range of 67.8-80.4 wt %. Considering the huge scale of hydrothermal metasomatism in the Bakal area, we may assume that the mineral formation was of an equilibrium character. The use of the ankerite-siderite (Annovitz and Essene, 1987) and ankerite-breinerite (Martynov, 1990) geothermometers for the study of the temperature limits of metasomatism showed that the temperature of sideroplesite formation did not depend on the position of the ore deposit in the stratigraphic section of the Bakal Formation, but is determined by the position in the structure of the Bakal ore field. The average calculated temperatures of metasomatism in the middle part (Vostochno-Bulandikhinsky and Irkuskan quarries) are 250-270°С, while in the peripheral western part (Novobakalskii quarry) they do not exceed 190-220°С (Krupenin, 2017).
The content of most of trace elements in siderite is at the clarke level. The ores do not contain the high concentrations of elements indicating a link to igneous mafic or felsic rocks (Ellmies et al., 1999). The higher contents in siderite relative to host dolomite are detected for Ba (the average contents are 59 and 23 ppm, respectively), Rb (1.98 and 0.32), and Cs (0.22 and 0.08); the lower contents, for Sr (12 and 34; in limestone they increase to 2000 ppm). The REE distribution in siderite differs from that in the host carbonate rocks. Chondrite-normalized spectra of the host carbonate rocks, free of terrigenous-shale admixtures, show a gradual decrease in the concentrations from La to Lu (Fig. 3, curves 1 and 2). In siderite the trend changes to the opposite with a gradual increase in the concentration from La to Lu (see Fig. 3, curve 3), which supports the mineralogical control, since the ionic radius of Fe 2+ is closer to that of Lu than of La. The depletion of siderite in LREEs is a favorable factor for the accumulation of samarium (an incoherent element) relative to neodymium. Shale carbonates (Fig. 3, curves 4 and 5) demonstrate the same tendency towards decrease in La/Yb is observed in the limestone-dolomite-siderite series, but in the latter, a subhorizontal distribution of REE is observed (Fig. 3, curve 6).

RESEARCH METHODS
The experience in studying Riphean siderite using microprobe analysis and Rb-Sr and U-Pb isotopic studies showed that the most complete metasomatic alteration occurred in large bodies within the central part of the Bakal ore field (Kuznetsov et al., 2005;Krupenin, 2017). Therefore, we selected mainly orebodies in the center of the ore field for the Sm-Nd systematics: in the Shuida and Gaev units in the Irkuskan quarry and in the Verkhnebakal unit in the Vostochno-Bulandikhinskii quarry. Siderite, dolo-  (Table 1).
At the first stage, field samples collected from the walls of quarries were subjected to comprehensive physicochemical study in order to identify the rocks purest from terrigenous admixtures and susceptible to sideritization processes. Chemical silicate analysis (X-ray fluorescence method, a SRM-35 spectrometer, Table 2), X-ray diffraction analysis (an XRD-7000 diffractometer, Shimadzu), and a study of the concentration of more than 40 trace elements, including REE (Perkin Elmer Elan-9000 ICP-MS) were performed at the Geoanalitik Collective Use Center, Yekaterinburg) using standard techniques. Some of the chemical analyses were obtained by X-ray fluorescence at the Berlin Technical University, Germany.
Selective dissolution of the carbonate component of the sample for isotope studies was carried out according to the methodology of (Kuznetsov et al., 2005;Ovchinnikova et al., 2018) in 1 N HCl without heating. The Nd isotopic composition in the samples was measured on a Neptune Plus mass spectrometer by MC-ICP-MS (Geoanalitik Collective Use Center). The Sr isotopic composition of siderite from the Irkuskan deposit was measured on a Finnigan MAT 261 mass spectrometer (Institute of Precambrian Geology and Geochronology, Russian Academy of Sciences, St. Petersburg), while the other samples were measured on a Triton Plus mass spectrometer (Geoanalitik Collective Use Center). The uncertainty in measuring the 147 Sm/ 144 Nd ratio was 0.4%; 143 Nd/ 144 Nd, -0.01%; 87 Rb/ 86 Sr, 0.5%; and 87 Sr/ 86 Sr, 0.002% (2σ). The average value in a standard Merck sample (Nd 2 O 3 ) during the operating period was 0.51172 ± 0.00003 (2σ, n = 3); for NIST SRM 987, 0.71026 ± 0.00002 (2σ, n = 3) at the Geoanalitik Collective Use Center and 0.71025 + 0.00001 (2σ, n = 16) at the Institute of Precambrian Geology and Geochronology, Russian Academy of Sciences. Table 3 presents the data from studying the Rb-Sr and Sm-Nd isotopic systematics.

RESULTS
In the central part of the ore field, the siderite deposits were studied in the Irkuskan and Vostochno-Bulandikhinskii quarries. Siderites on the northern flank of the Irkuskan quarry formed after limestone stromatolite bioherm in the Gaiev Unit: the pattern of metasomatic zoning of carbonates with stromatolite texture is clearly visible on the northern flank (Fig. 4a). All carbonate rocks contain an admixture (2-10%) of quartz, hydromica, and chlorite. To study the Sr-Nd isotope systematics, rock varieties purest from terrigenous admixtures were selected (Tables 1-3, samples of the X-… series). Siderite forms sheetlike bodies in the host shallowly bedded dolomite, which are practically devoid of terrigenous admixture in the stratigraphically lower Shuida Unit in this quarry (samples K-1/13 and K-1/5); however, upsection, the content of terrigenous admixture in dolomite increases (sample A-2) and the siderite mineralization is poorer. GEOLOGY    Siderite samples with massive structures (453-1 and Bul-4) were collected in the Vostochno-Bulandikhinskii quarry (Fig. 4b) from a thick ore bed at the 750 m horizon. Six samples of siderite with a fine-layered structure (17-S-2 to 17-S-7) were taken from one stratigraphic layer at a distance of 1 to 9 m from the intersecting metasomatic contact with finely bedded dolomite at the 790 m horizon. No sedimentary limestones are known in this stratigraphic horizon. Siderite of this layer contains 3-6% of quartz and an insignificant admixture of hydromica, chlorite, and sometimes pyrite (up to 2%). Near the contact, at a distance of up to 1.5 m, siderite contains an admixture of dolomite, the content of which decreases deeper into the deposit from 7 to 2%. Further from the endomorphic contact, the deposit is represented by monomineralic sideroplesite with a stable composition: 41.4-43.7% FeO, 8.8-9.4% MgO, 1.6-1.9% MnO, and 0.28-0.69% CaO ( Table 2).
For comparison, samples of siderite (y-5) quite pure of the terrigenous-shale admixture and near-ore dolomite (y-7) were collected from the western part of the ore field (Novobakalskii quarry) in the Berezov Unit, at the base of the Malobakal Subformation. The rocks had a vein-lenticular structure due to the occurrence of secondary veins, which have a yellowish creamy color in siderite and grayish white in dolomite. As in other carbonate units, sideroplesites in these Fe O rocks contain minor admixtures of quartz, dolomite, muscovite, and pyrite.
All of the studied siderites have very low contents of Sr (2-3 ppm) and Rb (0.2-1.46 ppm), which is typical of the Bakal Deposit (Kuznetsov et al., 2005). The strontium content in siderite increases (samples K-1/5, 39.41 ppm, and S-3, 15.05 ppm) in the contact zones and approaches that in dolomites. The measured 87 Sr/ 86 Sr ratio in siderite of the main ore phase is 0.7328-0.7385, but in siderite of the Vostochno-Bulandikhinskii quarry, it decreases to 0.7177-0.7334 (Table 3). The measured 87 Sr/ 86 Sr ratio in stromatolite dolomite is 0.7084. This ratio in clayey limestone, dolomite, and siderite near shale rocks increases to 0.7153, 0.7294-0.7419, and 0.7442, respectively (Table 3). Sedimentary limestone, with a minimum content of terrigenous admixture, were detected only on the eastern periphery of the ore field at Mt. Berezovaya in the Berezov Unit. They have the highest concentrations of strontium (up to 2000 ppm) and low 87 Sr/ 86 Sr values (0.7046-0.7048), characteristic of the Early Riphean Ocean (Semikhatov et al., 2009;Kuznetsov et al., 2018).
The Sm and Nd concentrations in siderite vary from 0.13-0.96 to 0.23-2.86 ppm, respectively; in dolomites, from 0.05-0.84 to 0.20-4.84 ppm; and in clayey stromatolite limestone, from 0.36 to 2.02 ppm (Table 3). The Sm and Nd concentrations in sedimentary limestone of the Bakal Formation are even lower (at the level of 0.06 ppm) and are not suitable for isotopic studies. The Sm/Nd ratio in siderite is significantly higher (0.30-0.66) than that in dolomite (0.20-0.25) and limestone (0.18), which is explained by the high concentrations of samarium and closeness of its ionic radius to that of iron versus neodymium (mineralogical control). The Sm-Nd characteristics of Riphean sedimentary limestone ( 147 Sm/ 144 Nd = 0.1297-0.1501, 143 Nd/ 144 Nd = 0.51157-0.51179) of the underlying Satka Formation (Krupenin et al., 2016) are close to those of limestone and dolomite of the Bakal Formation, but differs from siderite ( 147 Sm/ 144 Nd = 0.1797-0.4016, 143 Nd/ 144 Nd = 0.51179-0.51322) (Table 3).

DISCUSSION
The Sm-Nd systematics of siderite of hydrothermal genesis has to date not been studied. The only one was performed with an example of siderite veins hosting the Ag-polymetal mineralization of the Jebel Avam Deposit in Late Paleozoic shale-carbonate rocks of the Atlas Mountains (Castorina and Masi, 2008). In this deposit, the Sm and Nd concentrations in hydrothermal vein siderite are higher by an order of GEOLOGY    magnitude vesus Bakal siderite. The ε Nd (Т) values in vein siderite were calculated for the probable age of the mineralization (280 Ma): -5.7 ± 0.8. These values turned out to be intermediate between the ε Nd (Т) values in the host shale (-6 to -11) and intruding granite (-1.5 to -3.0). This allowed the authors to suggest the hydrothermal postmagmatic origin of the ore fluid, genetically related to collisional processes in Alpine folding. In this case, the Nd isotopic composition of the ore fluid formed with the participation of solutions from two sources: the host shale and intruding granite. This conclusion is additionally supported by data on the isotopic composition of strontium, carbon, and oxygen, and the presence of a sharp positive Eu anomaly in siderite (Castorina and Masi, 2008).
For all intents and purposes, we have studied the Sm-Nd systematics of hydrothermal-metasomatic siderite for the first time. Variations in the Sm/Nd ratio recorded in siderite of the Bakal Deposit make it possible to consider them a potential target for Sm-Nd dating. The correlation for all siderite samples in coordinates 147 Sm/ 144 Nd-143 Nd/ 144 Nd yields a line with an age of 998 ± 130 Ma (MSWD = 19). The broad scattering of points is explained by inclusion in the calculation of samples from contact zones and subjected to postore processes. This age is 987 ± 430 Ma (MSWD = 18) for siderite from the central parts of orebodies in the Irkuskan quarry. The minimum deviation is typical of sideroplesite samples from the profile in the orebody of the Vostochno-Bulandikhinskii quarry (samples S-2, …, S-7; see Fig. 4b), which do not contain dolomite admixtures and plot in the Sr-Nd isotope field of the Bakal shale (Fig. 5). If we exclude sample S-3 from the set for calculating the age, which has a high strontium content (15.05 ppm) associated with incomplete isotope-geochemical alterations in the metasomatic product, then we obtain a more accurate age of 970 ± 40 Ma, MSWD = 1.6 (Fig. 6). These samples have the purest composition without dolomite admixtures or residual calcium and strontium concentrations of the limestone protolith. Complete rearrangement of the Sm-Nd isotope system occurred in these samples, and their age most closely corresponds to the time of metasomatic recrystallization, previously estimated by the Pb-Pb method as 1010 ± 100 Ma (Kuznetsov et al., 2005).
Thus, the two isotopic systematics yield close ages of metasomatic ferruginous carbonates for the Bakal deposits, which significantly verifies the age boundary  of the formation of siderite mineralization, close to 1000 Ma.
Study of the age of metasomatic siderite using the Sm-Nd systematics is not a routine method. We know of only one successful case of such application for estimating the age of metasomatic siderite of the Erzberg Deposit, the largest in Western Europe. This siderite deposit, with total reserves of up to 400 mln t, is confined to the Paleozoic volcanosedimentary sequence, structurally located in the Greywacke Zone of the Eastern Alps (Styria, Austria), and has been mined for more than 1300 years (Redlich, 1916;Obruchev, 1935). Orebodies of coarse-crystalline siderite with complex shape, up to 200 m thick, metasomatically replace fine-grained Devonian limestone with the formation of a thin contact zone of ankerite and preservation of sedimentary structures in all varieties of carbonate rocks and ores. There is no wide zone of ferruginous dolomite around siderite deposits, like what we observe at the Bakal deposit. The age of siderite is 208 ± 22 Ma (MSWD = 4.6) and is confined to the divergent phase at the beginning of the Alpine tectogenesis cycle (Prochaska, 2016). Unfortunately, the author of the cited publication did not provide the initial Sm-Nd isotopic data, except for a graph with the Sm-Nd isochron.
The ε Nd (Т) value calculated for the Bakal ore field, taking into account the estimated age of siderite metasomatism (1000 Ma), varies from -13.4 to -17.6 in siderite of the central parts of ore units, and from -10.5 to -12.6 in siderite of the peripheral zones. The Nd (Т) value in most dolomites and limestone, is significantly higher, from -6.8 to -10.9, and in one dolomite near a clay member, it decreases sharply to -18.4 (Fig. 5). The Nd (1430) values for host shale of the Bakal Formation are -8 . The ε Nd (Т) value for shale of the Bakal Formation recalculated for an age of 1000 Ma, is -13.6, which is very close to the corresponding ε Nd (1000) value for siderite. The low ε Nd values are typical of rocks formed upon alteration or assimilation of older crustal formations with a Sm/Nd ratio lower than in the original chondrite reservoir (Fore, 1989). It is highly probable that the neodymium isotopic composition of orebearing fluids was formed as a result of intense interaction of buried solutions with shale of the Riphean sedimentary basin. The data of the Sm-Nd systematic reflect the crustal nature of the iron-bearing fluid from which the Bakal siderite formed.
The position of siderite on the 87 Sr/ 86 Sr orig. -ε Nd (T) plot differs significantly from the field of rift gabbrogranite intrusions of the Middle Riphean occurring in the area of the Bakal deposits (Fig. 5). Even in comparison with the host Riphean marine limestone (0.7046-0.7048 and -10.7 to -12.3), siderite is significantly enriched in crustal Sr and Nd. At the same time, siderite plots in the field of shale of the Bakal Formation (εNd (T) values are recalculated for 1000 Ma). This is consistent with the geochemical model of the formation of iron ore fluid due to the interaction of catagenetic solutions with silty-shale rocks in the Riphean sedimentary basin (Anfimov, 1997;Krupenin, , 2017Kuznetsov et al., 2005).
The idea of a geological model for the formation of such large accumulation of sideroplesite in carbonate rocks of the Bakal Formation is due to a favorable combination of a number of factors: lithogenetic, geotectonic, and structural.
Lithogenetic factor. The buried brines were the basis for the formation of high-Mg fluids, which contributed to both dolomitization of limestone and the formation of metasomatic magnesite deposits under the specific rift conditions of heating of carbonate layers upon migration in catagenesis (Krupenin and Kol'tsov, 2017). Magnesium is exchanged for iron in the process of interaction of aggressive chloride brines with shale under the thermodynamic conditions of deep catagenesis (Ellis, 1968;Drever, 1971;Maynard, 1985). Long-term interaction of brines with terrigenous-shale rocks in a catagenetic basin leads to a change in their composition to the ferruginous-magnesian. As a result, the formation of metasomatic Femagnesites and breinerites is probable (Krupenin and Kol'tsov, 2017;Krupenin et al., 2019).
Based on the data of (Yardley and Bodnar, 2014, Fig. 2.11), the metal/chloride ratio changes by five orders of magnitude between the low-and high-temperature hydrotherms (60-400°С). In other words, the amount of Fe 2+ mobilized by brines under the reducing conditions of the interiors increases sharply with increasing temperature of the brine fluid. This is important in the case of the formation of brines in the deep horizons of the super-rift depression. The magnesian-ferruginous chloride brine is formed in the high-temperature deep part of the rock basin. The tectonic activation stimulates its migration and discharge into the colder side parts of the former rift. Iron precipitates under the conditions of a decrease in solubility, when a brine fluid is introduced into carbonate rocks, and thus a mode of maximum favoring of Femetasomatism with the formation of sideroplesite is provided.
The chemical composition of fluid inclusions in Bakal siderite, corresponding to strong evaporite brines, was confirmed by special studies using the method of ion chromatography (Bottrell et al., 1988), by the high concentrations of chlorine and sodium, as well as bromine, which accumulates in solution under the conditions of evaporite thickening of seawater only (Prochaska and Krupenin, 2013). With the further evolution of the brine composition, the bromine concentrations are retained, being a stable tracer of the genetic nature of fluids (McCaffrey et al., 1987;Kesler et al., 1996;Leach et al., 2010). Such brines have a high potential for the leaching of metals (in particular iron) from shale upon migration, (Savard et al., 1998 Chaudhuri and Clauer, 1993;Kharaka and Thordsen, 1992). The interaction of fluids with sandy shale rocks leads to an increase in the proportion of radiogenic strontium and REE in solutions (Gorokhov et al., 1982. Epigenetic dolomite in the Riphean section of the Southern Urals usually has a higher 87 Sr/ 86 Sr ratio than that in the host limestone (Kuznetsov et al., 2008;Semikhatov et al., 2009;Kuznetsov et al., 2017). This ratio in metasomatic magnesite, ferruginous magnesite, and siderite becomes even higher (Frimell, 1988;Cortecci and Frizzo, 1993;Kuznetsov et al., 2007;Krupenin and Kuznetsov, 2009 (Table 3).
According to a number of researchers, the evolution of magnesian brines into Mg-Fe-bearing fluids was the reason for the formation of ferruginous carbonates after magnesian ones at deposits in the Paleozoic terrigenous-carbonate rocks of Central Europe (Frimmel, 1988;Radvanec et al., 2004;Prochaska, 2016). The geochemical evidence presented by Prochaska (2016) unequivocally indicates the origin of the ore solutions of the Erzberg siderite deposit from the brines of the Triassic evaporite basins submerged in the Paleozoic strata at the stage of rift extension. Hydrothermal iron ore fluids were formed during the interaction of aggressive brines with volcanogenicterrigenous host rocks under the redox conditions of catagenesis or low-temperature metamorphism (Prochaska, 2016).
The similar conditions probably occurred during the formation of Fe-Mg metasomatic carbonates in the Riphean rocks of the Southern Urals. Using the example of the same Bakal ore field, we see that lowiron magnesite in the Shuydinskaya Unit of the Bakal Formation was controlled by the Mashak riftogenesis at the beginning of the Middle Riphean, while siderites were formed much later, in the Late Riphean Ovchinnikova et al. , 2018). Another example is the ferruginous magnesite Ismakaevskoe Deposit (up to 5% FeO), which has a total thickness of up to 400 m in the Early Riphean deposits. However, according to the Sm-Nd data, the formation of magnesite occurred at the end of the Middle Riphean, 1250 Ma (Krupenin et al. , 2016). In addition, a breinerite body (up to 20.5% FeO) was found in the same limestone unit, 10 km to the south from Ismakaevo (Krupenin et al., 2019).
The geotectonic factor in the formation of siderite mineralization is controlled by the position of the ore field in the super-rift depression, namely, on the western flank of the Mashak rift, where iron-bearing solutions could be discharged from deeply submerged areas of the rift basin (super-rift depression) at the Grenville stage of tectonothermal activation. The age of 1000 Ma is borderline for the geological evolution of the eastern margin of the Baltic. At this time, inversion and a change in the tectonic regime occurred. The evolution of the intracratonic sedimentary basin (superrift depression) after this boundary gave way to the formation of a large pericratonic basin, which occupied a wide area in the north and east (in modern coordinates) of Baltica (Maslov, 1997). There was a long hiatus in sedimentation more than 250 Ma ago in the sedimentary sequence between the Middle Riphean Avzyan Formation and the Late Riphean Zil'merdak Formation (Maslov et al., 2018;Kuznetsov et al., 2017).
The deposits of the Bakal Formation subsided under the Middle Riphean deposits to depths of no more than 2.5-3 km by the time of the ore formation. Accordingly, the host rocks were heated to 120-150°С under the influence of the geothermal gradient. At the same time, the distribution of REE in ores and siderite-ankerite thermometry yield of ore deposition temperatures of 250-260°C in the central parts of the ore field and 180-220°C in the peripheral zones (Krupenin, 2017). The relatively high temperatures of the hydrothermal process indicate that the Bakal ore field became an discharge area of intraformational fluids coming from deeper and, accordingly, more heated parts of the paleohydrogeological basin at the Middle-Late Riphean boundary. The fluid formation zone was located in Lower Riphean sediments, presumably east of the ore field in the Mashak rift zone, where the thickness of the overlying Middle Riphean sediments sharply increased during active volcanosedimentary accumulation (Mashak Formation), and reaching 6-7 km.
Studies of the oil-and-gas-bearing basins show that the discharge of catagenetic fluids occurs at the stages of tectonic activation, when gas-water solutions accumulated within thick shale units, which originated upon the formation of hydromica after smectites and caused abnormally high reservoir pressures, which penetrated into the upper horizons of the sedimentary crust. These solutions, mixed with infiltration waters, release gases and metals dissolved in them at the stage of temperature and pressure drop (Lebedev, 1992). We suggest that the relaxation of hydrothermal Mg-Ferich fluids occurred at the regressive stage of the evolution of the Middle Riphean sedimentary basin. At that time, a series of faults formed under the influence of upward movements at the end of the Avzyan (the pre-Zilmerdak sedimentation hiatus), which facilitated the migration of elision fluids from the deep parts of the basin to low pressure zones on its periphery.
The structural factor controls the conditions for the formation of deposits with unique reserves of siderite ores. This became possible due to the focusing of the fluid flow in the zone of the tectonic block, uplifted, inclined, and partially eroded in the pre-Zigal'ga (the initial period of formation of the Mashak graben), consisting of rocks of the Bakal Formation. Ascending solutions were preserved beneath the overlying quartzite sandstone of the Zigal'ga Formation, which were already significantly compacted and had negligible permeability by the end of the Middle Riphean ; they were also relatively resistant to brittle deformations. Hydrothermal iron-magnesian chloride solutions with low pH produced metasomatic replacement in carbonate rocks of the Bakal Formation in the adjacent zone along a relatively permeable interformational unconformity surface and formed siderite-ankerite-dolomite zoning.
The formation of the world's largest accumulation of hydrothermal-metasomatic siderite ores at the 1000 Ma boundary in Riphean deposits of the Southern Urals was a major event in the formation sequence of other deposits of this time interval, which fell on the Middle-Late Riphean boundary. Most of the deposits are confined to the western side of the Mashak graben and are localized in the deposits of the Early and Middle Riphean. Two stages of mineragenic activity at this time may be distinguished: 1250-1200 and 1000 Ma.
The first stage is associated with the formation of barite-polymetallic deposits and manifestations of the SEDEX-type in the deposits of the Avzyan Formation, which completes the section of the Middle Riphean, fluorites, and ferruginous magnesites. The Sm-Nd age of Fe-magnesite of the Ismakaevskoe Deposit in the Early Riphean rocks is 1250 ± 130 Ma, despite too huge error due to the incomplete cogeneticity of magnesite samples, clearly tends to the same stage (Krupenin et al., 2016). In addition, vein and hydrothermal-metasomatic fluorites of the Suranskoe Deposit with an age of 1230 Ma (Sm-Nd and Rb-Sr methods) are confined to the same Early Riphean carbonate rocks in this area (Krupenin et al, 2012). A detailed study of the formation sequence of fluorite generations and their geochemical patterns indicates that fluorine was extracted from felsic igneous rocks of the Mashak graben upon interaction with catagenetic brines. This is supported by the brine nature of fluid inclusions, as well as by the high REE concentrations and negative Eu anomaly in the major fluorite generations. The barite-polymetallic mineralization forming stratified sulfide ores is widespread in the Avzyan Formation of the Middle Riphean (Kuzhinskoe and Verkhne-Arshinskoe deposits and a number of ore occurrences). The sulfur isotopic composition of sulfides suggests primary sedimentary ore accumulation due to the cyclic sulfate reduction upon discharge of metalliferous thermal exhalations in the seafloor depressions (δ 34 S is +21-26‰) (Krupenin, 2004).
The second stage (~1000 Ma) includes the formation of siderite of the Bakal ore field. Most likely, the deposits of metasomatic ankerite were formed in the Avzyan Formation at the same stage.
It is important to note that, despite the extensive development of thick carbonate and terrigenous units (which can be both reservoirs and a source of catagenic fluids), not a single magnesite and siderite deposit is observed in the overlying deposits of the Upper Riphean. This emphasizes the importance of the Middle-Late Riphean boundary for the formation of the minerogenic character of the region.

CONCLUSIONS
(1) The first study of the Sr-Nd isotope systematics of hydrothermal-metasomatic siderite was carried out based on the example of the Bakal deposits, the world's largest accumulation of this type of iron ore.
(2) The source of the hydrothermal fluid from which the metasomatic siderite of the Bakal deposits formed is supported by data on the isotopic composition of strontium, indicating interaction of iron-bearing fluids with terrigenous-shale rocks of the Riphean sedimentary basin, as well as by data of the Nd-systematics, indicating the crustal nature of the fluid.
(3) The formation model for Precambrian siderite at the Grenville stage of tectonothermal activation, rather than during the Mashak rifting event at the beginning of the Middle Riphean, has been supported for the first time by Sm-Nd isotopic data and refines the first estimate of the age of mineralization by the Pb-Pb method. Application of two isotopic systems for the direct study of the age of siderite significantly verifies the age of the siderite mineralization of the Bakal group of deposits.
(4) The identified age limit of ~1000 Ma is associated with tectonic restructuring in the region (the eastern margin of Baltica) at the Middle-Upper Riphean boundary, which was of decisive importance for the formation of a number of stratiform and hydrothermal-metasomatic deposits in the Southern Urals. It was at this time that the minerogenic character of the region was formed; further tectonic restructurings only complicated the orebodies.