Ooidal ironstones in the Meso-Cenozoic sequences in western Siberia:assessment of formation processes and relationship with regional and global earth processes

Maxim Rudmin,Santanu Banerjee,Elshan Abdullayev,Aleksey Ruban,Ekaterina Filimonenko,Elena Lyapina,Roman Kashapov and Aleksey Mazurov

Abstract This study investigates the process of formation of ooidal ironstones in the Upper Cretaceous-Paleogene succession in western Siberia. The formation of such carbonate-based ironstones is a continuing problem in sedimentary geology, and in this study, we use a variety of data and proxies assembled from core samples to develop a model to explain how the ooidal ironstones formed. Research on pyrite framboids and geochemical redox proxies reveals three intervals of oceanic hypoxia during the deposition of marine ooidal ironstones in the Late Cretaceous to the Early Paleogene Bakchar ironstone deposit in western Siberia; the absence of pyrite indicates oxic conditions for the remaining sequence. While goethite formed in oxic depositional condition, chamosite, pyrite and siderite represented hypoxic seawater. Euhedral pyrite crystals form through a series of transition originating from massive aggregate followed by normal and polygonal framboid. Sediments associated with goethite-chamosite ironstones,encompassing hypoxic intervals exhibit positive cerium, negative europium, and negative yttrium anomalies.Mercury anomalies, associated with the initial stages of hypoxia, correlate with global volcanic events. Redox sensitive proxies and ore mineral assemblages of deposits reflect hydrothermal activation. Rifting and global volcanism possibly induced hydrothermal convection in the sedimentary cover of western Siberia, and released iron-rich fluid and methane in coastal and shallow marine environments. This investigation, therefore, reveals a potential geological connection between Large Igneous Provinces (LIPs), marine hypoxia, rifting and the formation of ooidal ironstones in ancient West Siberian Sea.

Keywords: Ooidal ironstone,Pyrite framboids, Late Cretaceous, Early Paleogene, Western Siberia,Bakchar ironstone deposit, Redox conditions

1 Introduction

Ooidal ironstones represent non-siliceous sandy and clayey sediments consisting of at least 5% iron oolith/ooid and 15% iron (Young 1989; Petranek and Van Houten 1997),and form abundantly during the Ordovician-Devonian, the Jurassic-Cretaceous and the Paleogene (Van Houten and Bhattacharyya 1982; Van Houten and Arthur 1989; Young 1989) but are rare in modern sediments (Kimberley 1994;Heikoop et al. 1996; Sturesson et al. 2000). The origin of ooidal ironstones remains controversial as there is no unanimity amongst researchers regarding the source of iron.Several researchers considered that iron is available by the weathering of continents and it is delivered into the ocean by river or ground water (Strakhov 1947; Castano and Garrels 1950; Huber and Garrels 1953; Kholodov 2014).Others suggested submarine hydrothermal exhalative source of iron (Pavlov et al. 1991; Kimberley 1979, 1989,1994; Pavlov 1996; Rudmin et al. 2019; Todd et al. 2019).Modern ooidal ironstones developed in the Mahengetang Island in Indonesia (Heikoop et al. 1996) support the“volcanogenic” hypothesis (Sturesson et al. 2000; Sturesson 2003). Proponents of the “exhalation” hypothesis consider the transportation of hydrothermal Fe2+along with phosphorus and manganese by anoxic seawater and upwelling of the same on oxygenated continental shelf (Todd et al.2019). Subsequently the Fe2+gets oxidized by biological or abiological ways resulting in formation of ooids and peloids(Dahanayake and Krumbein 1986; Burkhalter 1995; Taylor and Konhauser 2011; Todd et al. 2019). Several studies indicated that mineralogical and chemical compositions of ironstones implicate the mode of origin as well as depositional conditions influence (Van Houten and Purucker 1984; Maynard 1986; Kimberley 1989; Siehl and Thein 1989;Sturesson et al.2000).

Recent investigations explore the relationship between periods of intense volcanism (Large Igneous Provinces,LIPs), global ocean anoxic events and the formation of ooidal ironstone deposits (Percival et al. 2015; Ernst and Youbi 2017; Scaife et al. 2017; Keller et al. 2018).Although Van Houten (1986) and Van Houten and Arthur (1989) indicated a relationship between ooidal ironstones and anoxic events, this is yet to be investigated thoroughly (Turgeon and Creaser 2008; Jenkyns 2010; Raven et al. 2018). While, the origin of the giant ironstone in the Meso-Cenozoic Bakchar deposit has been debated, but no study attempts to relate hypoxia,global volcanism and ironstone deposits, because the integration of data related to mineralogical and chemical composition of ironstones with proxies for redox and volcanism is still lacking.

The Bakchar ironstone deposit has been studied since 1960s, and the early studies have provided the overall geological structure from core data (Belous et al. 1964;Nikolaeva 1967). Subsequently Pavlov (1989) has proposed the origin of West Siberian iron ore basin by considering spatial association of ironstone deposits with oil and gas deposits. A detailed core drilling has made it possible for extensive studies on the origin of the ironstone (Podobina and Kseneva 2005; Asochakova 2014;Rudmin et al. 2014; Gnibidenko et al. 2015; Rudmin and Mazurov 2016).

Fig. 1 (a) Location map of the Bakchar ironstone deposit in western Siberia; (b) Simplified geological map of the Bakchar ironstone deposit showing areas of three main ironstone(Rudmin et al.2019)

The study of pyrite framboids is useful for determination of hypoxia during marine deposition (Wilkin et al.1996; Wignall et al. 2005; Wei et al. 2015). Further, concentrations of trace metals such as molybdenum, vanadium, uranium are widely used as proxies for reflecting depositional redox condition (Brumsack 2006; Tribovillard et al. 2006; Algeo and Tribovillard 2009;Lebedel et al. 2013). Mercury anomaly traces global volcanism in a few studies (Scaife et al. 2017;Sabatino et al. 2018; Them et al. 2019). An integrated study involving pyrite framboids, trace element concentrations and mercury anomaly of the Bakchar ironstone deposit is likely to provide a detailed perspective for establishing the relationship between hypoxia, global volcanism, rifting and the origin of ooidal ironstones. The aim of this work is to (1) infer the origin of iron in the Bakchar ironstone deposit in western Siberia; (2) understand variations in redox conditions across the ironstone deposit; and, (3)examine the relationship between LIPs, anoxia and the timing of ironstone formation. We assess redox conditions of the depositional setting based on several proxies including diameter of pyrite framboids, concentrations of trace elements and enrichment factors of these elements in available core samples across the Late Cretaceous-Early Paleogene section in Bakchar ironstone deposit. And on the basis of mercury anomalies in sediments, we attempt to record contemporary volcanic events.

2 Geological setting

The Bakchar ironstone deposit develops in the southeastern part of western Siberia (52°01′45″N; 82°07′20″E) situated 200 km northwest of Tomsk (Fig. 1). The ironstone occurs as 150 km wide and about 2000 km long deposit in the West Siberian iron ore basin along eastern and south-eastern borders of the West Siberian Plate (Fig. 1a). This Upper Cretaceous-Palaeogene iron ore deposit occurs in subsurface at depth ranging from 165 m to 230 m (Belous et al. 1964; Podobina and Kseneva 2005; Lebedeva et al. 2013, 2017). The ore-bearing strata in Bakchar deposited in transgressive coastal and shallow marine environments, attaining a maximum thickness of 80 m (Fig. 2; references for shallow marine).The deposit occurs at the northern end of a Mesozoic dome, also known as Bakchar tectonic uplift (Belous et al. 1964; Rudmin et al. 2019). The gently dipping (less than 3°) ore horizons occur at the north-west and east of watershed axes of Rivers Galka and Bakchar. The roof of the upper ore horizon in Bakchar ironstone lies subsurface at the depth ranging from 157 m to 160 m at the western sector of the deposit, whereas at the eastern sector, the same deposit occurs at depth ranging from 170 m to 175 m.

Fig. 2 Lithostratigraphic columns representing two drilling wells, 570 and 803, through the Bakchar ironstone deposit (for locations see Fig. 1b),with relevant stratigraphic details,occurrence of ironstones,and sample positions

Ooidal ores occur at three stratigraphic levels in this deposit, viz. the Narym, the Kolpashevo and the Bakchar(Fig. 2; Belous et al. 1964; Rudmin and Mazurov 2016;Rudmin et al. 2019). The Narym ironstone occurs at the top part of the Ipatovo Formation, which, in turn, overlies the Kuznetsovo Formation. The Ipatovo Formation consists of fine-grained sandstones and greyish-green siltstones with interlayers of clay that passes upward to the Narym ironstones. The Kolpashevo ironstones occur in association with glauconitic sandstones, siltstones and silty clays of the Santonian Slavgorod Formation and the Campanian-Maastrichtian Gan’kino Formation. The lower part of the Lyulinvor Formation consists of medium-grained sands and sandstones which pass upward to the Bakchar ironstones and claystones.

3 Materials and methods

Eighty-seven core samples were collected from the Bakchar ironstone deposit with an interval from 0.5 m to 2 m from two wells, 570 and 803 at eastern and western sectors respectively (Fig. 2). Core samples were powdered using an agate ball mill for geochemical studies.68 thin sections were investigated for pyrite framboids by scanning electron microscope (SEM). Diameters of pyrite framboids were measured in polished sections using a TESCAN VEGA 3 SBU SEM equipped with an OXFORD X-Max 50 EDS analyzer and a Si/Li crystal detector. An accelerating voltage of 20 kV with a beam current between 3.5 nA and 15 nA was used for SEM observations. For each thin section, more than 100 pyrite framboids were measured primarily in backscatter mode.

Fig. 3 SEM images showing different morphological forms of pyrite in Bakchar ironstone deposit. (a) A huge aggregate of massive pyrite (maspyr) with normal framboids (norm-fr) and macro framboids (macro-fr) in sample 5702189; (b) Nodule of massive pyrite with normal framboids in sample 8032047; (c) Macro framboid and normal framboids in cement in sample 8032360; (d) Overgrowth or annular framboids (overgr-fr) in sample 5702195; (e)Polygonal framboids(polyg-fr)in sample 5702286;(f)Massive and euhedral(euh-pyr)pyrite in sample 5702189

Table 1 Statistical parameters of pyrite framboids from three intervals in well 570(size in μm)

Major element concentrations of the powdered samples were estimated by HORIBA XGT 7200 X-ray fluorescence microscope (XRF) operated at a tube current of 1 mA, beam diameter of 1.2 mm, and a voltage of 50 kV.Fused pellets were prepared by pressing and melting sample powder for analysis using XRF. The detection limit for major elements was better than 0.01 wt.%.Concentrations of trace elements (TEs), including rare earth elements (REEs), were estimated using inductively coupled plasma-mass spectrometry (ICP-MS) at the Hydrogeochemistry Research Laboratory, Tomsk Polytechnic University. About 0.5 g of powdered sample was fused using 0.8 g of LiBO2/Li2B4O7at 1050°C for about 15 min. The glass beads were dissolved in a mixture of 5:4:1.5 HF, HNO3, and HClO4at 120°C in a platinum crucible for 6 h. The acid mixture was allowed to evaporate at 160°C. The sample residue was dissolved in 10 ml of 5HNO3. The resultant solutions were filtered and analyzed for REEs (see El-Habaak et al. 2016). Elemental concentrations were normalized to Al content to remove the effect of variable terrigenous input (Brumsack 2006; Tribovillard et al. 2006). Enrichment factor (EF) was calculated for each sample following Tribovillard et al. (2006). The Al EF was calculated as: Al EF=AlSample/AlPAAS.While EF values exceeding 1 represent a detectable enrichment of the element above the average shale concentration, those in excess of 10 indicate a moderate to strong enrichment (Tribovillard et al. 2006,2012; Nú?ez-Useche et al. 2016).

Pyrolysis was performed on samples using a model 6 turbo ROCK-EVAL analyzer (Vinci Technologies) at the Arctic Seas Carbon Research International Laboratory, Tomsk Polytechnic University. Around 100 mg samples were heated initially at 300°C to release volatile hydrocarbons, and subsequently at 600°C to release pyrolytic hydrocarbons. The CO/CO2released during the analysis was registered with infra-red source cell detector to quantify the total organic carbon content (TOC). Mercury concentrations were measured using 50 mg of powdered sample with a Lumex RA-915 Portable Mercury Analyzer fitted to the PYRO-915 Pyrolyzer.

Fig. 4 SEM images showing evolutionary series of pyrite formation from framboid to euhedral crystal: (a) Massive pyrite microcrystals; (b) Normal framboids; (c, d)Polygonal framboids;(e)Euhedral pyrite crystal

4 Results

4.1 Morphology and size distribution of pyrite framboids

Pyrite occurs as various forms including framboids(Fig. 3a–e), irregular and subhedral masses (Fig. 3a,b), and euhedral crystals (Fig. 3f). The last variety of pyrite is the least abundant. Pyrite framboids are further divided into several types, such as normal,macro, overgrowth and polygonal (Fig. 3; cf. Sawlowicz 1993; Wei et al. 2015). Mean diameter, standard deviation and other statistical parameters for normal pyrite framboids are presented in Table 1.

Fig. 5 Cross plot between mean diameter and standard deviation of pyrite framboids. Oxic-anoxic field boundaries are adapted on the basis of published data (cf. Wilkin et al. 1996; Bond and Wignall 2010; Merinero and Cárdenes 2018). Most of the samples of the Slavgorod Formation plot within the anoxic field.Form.=Formation

Fig. 6 Lithostratigraphic column for well 570 showing distribution of pyrite framboids, geochemical profiles of TOC, Hg, Hg/TOC, elemental enrichment factors (EFs). Correlation between small sizes of pyrite framboids (as per Fig. 5), relatively high content of TOC, Hg/TOC, and elevated enrichment factors of Mo,U, V indicate three hypoxic intervals(marked by light grey shade).Form.=Formation; ir. =ironstone

Table 2 Major oxides for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit

Table 2 Major oxides for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit(Continued)

Table 2 Major oxides for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit(Continued)

Normal framboids (Fig. 3a–c) appear as regular spheres with diameters varying from 2 μm to 20 μm.Each framboid consists of numerous microcrystals,the diameter of which varies from 0.7 μm to 1.3 μm.The diameter of macro-framboids (Fig. 3a, c) varies from 21 μm to 60 μm. The diameter of octahedral to dodecahedral micro-crystals within macro-framboids may be up to 2 μm. The overgrowth (also known as annular) variety of framboids may exhibit a distinctive outer ring of elongated crystals (Fig. 3d; cf. Merinero et al. 2009, 2017; Merinero and Cárdenes 2018). The diameter of this rare variety of framboids varies from 6 μm to 20 μm. The polygonal framboid often consists of densely-packed microcrystals and has long dimension varied from 5 μm to 12 μm (Fig. 3e). Pyrite also occurs as massive aggregates consisting of tiny crystals (Fig. 3a, b, f). Diameter of individual crystal in these aggregates varies from 0.5 μm to 1.0 μm. Euhedral pyrite may occur either as solitary crystal or as clusters; pyrite in these clusters often exhibits dodecahedral form with average long dimension up to 10 μm(Figs. 3f and 4e).

Normal framboids with size varying from 2 μm to 20 μm, with a geometric mean of 6.1 μm, appear to be the first formed pyrite. The aggregation of iron sulfide microcrystals or the crystallization from sulfide-saturated colloids forms normal framboids(Berner 1984). The filling of the intercrystalline space in normal framboids with ferrisulfide and subsequent recrystallization possibly forms euhedral pyrites (cf.Zhao et al. 2018). Thus, the evolution of the pyrite reflects a transition from normal framboid to euhedral crystal through polygonal variety (Fig. 4).

Small pyrite framboids occur within shale, siltstone and ironstone beds in the three intervals in well 570 corresponding to the middle Coniacian, the Santonian and the early Campanian ages (Figs. 5 and 6). The average diameter of normal framboids in these intervals varies from 5.1 μm to 12.0 μm. The Santonian interval exhibits the smallest pyrite framboid with an average diameter of 5.6 μm (standard deviation:2.5 μm) (Table 1). The mean diameter of framboids in the Coniacian and Campanian intervals varies from 6.4 μm to 12.0 μm (Fig. 5). Pyrite framboid does not occur in the Maastrichtian and Paleocene formations.

The occurrence of small variety of pyrite framboids in the three intervals in well 570 indicates oxygendepleted depositional conditions (cf. Berner 1984;Wilkin et al. 1996; Wei et al. 2015; Merinero et al.2017; Rickard 2019). The abundance and size of framboids indicate their formation in the hydrogen sulfide-rich bottom water column. The cross plot between mean diameter and standard deviation of framboids further indicates anoxic depositional conditions for two intervals: one at the lower part of the Slavgorod Formation and another encompassing the upper part of the Slavgorod Formation and the bottom part of the Gan’kino Formation (Figs. 5 and 6). Fe mineral assemblage for these intervals consists of chamosite peloids and ooids, siderite microcrystals and pyrite framboids. Rare authigenic minerals include barite and galena. Statistical parameters of pyrite framboids indicate dysoxic depositional condition for the lower part of the Ipatovo Formation (Figs. 5 and 6). Mineral assemblage of this interval comprises chamosite ooids and peloids, siderite-berthierine cement and pyrite besides rare pyrrhotite, barite, galena, sphalerite and arsenopyrite. Pyrite framboids less often occur within siltstone-claystone interlayers in the Lyulinvor Formation hosting the Bakchar ironstones.

4.2 Geochemistry

Concentrations of major and trace elements of ironstones and associated rocks are provided in Tables 2,3, 4, and 5. Variations in trace element enrichment factors across the stratigraphic succession are provided in Figs. 6 and 7. While Mo, U, and V are proxies for sedimentary redox conditions, concentrations of P relate to primary productivity in marine environments, and contents of Ti and Zr indicate clastic influx (Brumsack 2006; Tribovillard et al. 2006; Algeo and Tribovillard 2009; Lebedel et al. 2013).

High values of enrichment factors (EFs) coincide with three hypoxic intervals containing small diameter of pyrite framboids in the lithostratigraphic column of well 570 (Fig. 6). Both Mo EF and V EF are higher than the median (10.9 and 18.2, respectively)in the lower and upper hypoxic intervals in well 570(Fig. 6). U EF (from 2.1 to 3.3) in these intervals is higher relative to the rest of the sedimentary sequence in well 570 (Fig. 6). P EF has several peaks,which are synchronous with Fe EFs. P EF correlates positively with Fe EF (r2=0.8). Values of both P EF and Fe EF are higher within ironstone intervals than the rest of the sequence (Figs. 6 and 7). High valuesof EFs, especially the U EF, recognize two hypoxic intervals in well 803 (Fig. 7).

Table 3 Trace elements for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit

Table 3 Trace elements for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit(Continued)

Table 3 Trace elements for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit(Continued)

Concentrations of rare earth elements (REEs) vary from 124.5 ppm to 652.4 ppm (av. 348.7 ppm) in the Bakchar ironstone deposit. The ratio of light to heavy rare earth elements (LREE/HREE) varies from 5.4 to 28.8 with an average of 14.1. The hypoxic intervals exhibit positive Ce anomalies from 0.96 to 1.40. High values of Ce/Ce* (varying from 1.2 to 1.4)are characteristic for the Coniacian ooidal ironstones, the early Santonian siltstones and ooidal ironstones, the Campanian-Paleocene ooidal ironstones and glauconitic rocks. Eu anomalies range between 0.8 and 1.2 in the Bakchar ironstone deposit.Hypoxic intervals show negative Eu anomalies from 0.7 to 1.0. Most samples exhibit positive Ce and negative Eu anomalies in the Coniacian-lower Santonian rocks, accompanied by depletion in HREE(Fig. 8). Most ironstone samples show negative Y anomalies (Fig. 8), while those associated with siderite exhibit sharp negative Y anomalies. The top part of the Narym ironstone, the bottom part of the Santonian sandstones and the lower part of the siderite-goethite ore of the Bakchar ironstone exhibit negative Y anomalies up to 0.75 (Fig. 8).

The total organic carbon (TOC) content of rocks varies widely from 0.06 wt.% to 5.71 wt.% (Table 4;Figs. 6 and 7) within the Meso-Cenozoic Bakchar ironstone deposit. The TOC content is slightly higher(av. 0.36 wt.%) in sediments of the early Coniacian,the early Santonian, the early Campanian, and the late Paleocene-early Eocene (Figs. 6 and 7). The TOC content of the hypoxic intervals varies between 0.1 wt.% to 0.7 wt.%. However, the TOC content at the bottom part of the Ipatovo Formation is very high(5.7 wt.%). The Hg content varies from 1.8 ppb to 146.7 ppb (av. 38 ppb) in sediments; and varies between 4.0 ppb and 70.3 ppb within hypoxic intervals.High Hg content value (up to 110 ppb) characterizes the layer immediately below the hypoxic intervals in well 803 (Fig. 7). Hg concentrations and Hg/TOC values vary slightly from one well to another. Hg/TOC values are higher in sediments of the early Coniacian, the early Paleocene and the early Eocene in well 570; and in sediments of the late Turonian, the late Santonian, the Maastrichtian, the early Paleocene and the early Eocene in well 803 (Figs. 6 and 7).

5 Discussion

5.1 Hypoxia during the formation of ooidal ironstone deposit

Size of pyrite framboids as well as trace element enrichment factors distinguishes three hypoxic intervals within the Meso-Cenozoic Bakchar ironstone deposit (Figs. 6 and 7). Absence of pyrite framboids as well as the mineralogical assemblage dominated by goethite ooids with early diagenetic illite-smectite and lepidocrocite cement indicates oxic depositional conditions (Aller et al. 1986;Kimberley 1994; Heikoop et al. 1996; Sturesson et al.2000; Di Bella et al. 2019) for the Maastrichtian and Paleocene intervals (Figs. 6 and 7; see also Rudmin et al.2019). High values of Mo EF (above 11.3), U EF (above 2) and V EF (above 16.2) as well as occurrence of tiny(mean diameter ＜6.3 μm) pyrite framboids indicate hypoxic shelf waters during the Coniacian (Figs. 5, 6, and 7).The Coniacian deposits encompass the time interval corresponding to OAE-2 (ocean anoxic event 2; Schlanger and Jenkyns 1976; Arthur et al. 1988; Jenkyns 2010).Widespread anoxic depositional condition during the Santonian results in the formation of tiny (mean diameter ＜5.9 μm) pyrite framboids at the water column and/or on the sea floor (Figs. 5, 6, and 7). This Santonian hypoxic interval is marked by high values (from 11 to 28) of Mo EF, U EF, and V EF (Figs. 6 and 7), and corresponds to the OAE-3 known for the accumulation of organic-rich marine sediments (Wagreich 2012; Jones et al. 2018). The TOC content of sediments remains low notwithstanding the isolation of West Siberian Sea.Since P correlates well with Fe, the enhanced P EF indicates co-precipitation of chamosite and phosphate. A high iron flux possibly controls the precipitation of phosphorus in the form of phosphate-bearing rare earth elements. The lack of correlation between Fe EF and Ti EF and between Fe EF and Al EF (Figs. 6 and 7) rules out continental source of iron. Dysoxic bottom seawater promoted the formation of large framboids (mean diameter from 6.2 μm to 12.0 μm) and macro framboids(mean diameter from 20 μm to 46.4 μm) as well as large number of micro-aggregates of pyrite below the sediment-water interface during the early Coniacian and the late Campanian.

Bottoms of Kolpashevo and Bakchar ironstones in well 803 are marked by sharp swing in Mo EF, U EF and V EF (Fig. 7). Similar sharp swing in Mo EF and V EF (34.9 and 30.4 respectively) marks the Santonian hypoxicinterval underlying the Kolpashevo ironstone. The Coniacian Narym ironstone in well 570 (Fig. 6) exhibits a sharp increase in U EF from 2.1 to 2.6. Because of the greater depth in the eastern part compared to that of the western part of the ironstone deposit (Fig. 2), the hypoxic seawater was more extensive in the former.

Table 4 Enrichment factors (EFs), Hg, TOC and Hg/TOC for samples of two sections in wells 570 and 803 from eastern and western(respectively) corners of the Bakchar ironstone deposit

Table 4 Enrichment factors(EFs), Hg,TOC and Hg/TOC for samples of two sections in wells 570 and 803 from eastern and western(respectively) corners of the Bakchar ironstone deposit(Continued)

Table 4 Enrichment factors(EFs), Hg,TOC and Hg/TOC for samples of two sections in wells 570 and 803 from eastern and western(respectively) corners of the Bakchar ironstone deposit(Continued)

While the accumulation of chamosite-goethite ooids and different cements reflect a change in the geochemical environment from oxic to anoxic conditions during formation of ironstones (Rudmin et al. 2019). It is characterized by sequential formations of goethite-chamosite ooids, glauconite, berthierine or chamosite, pyrite and/or siderite. Narym and Kolpashevo ironstones encompassing hypoxic intervals consist of different mineral assemblages such as chamosite/glauconite + berthierine +goethite ± siderite ± occasional large pyrite and chamosite + glauconite + siderite + tiny pyrite, respectively.

Total content of REEs increases in ironstone intervals and decreases slightly in hypoxic intervals in both wells.The Coniacian and Santonian sediments usually exhibit higher LREE/HREE ratios (average of 15.3) relative to those in the Campanian and Paleocene (average of 13.3)(Figs. 6 and 7). Sediments representing hypoxic intervals as well as goethite-chamosite ironstones exhibit positive Ce anomaly, caused by the oxidation of cerium and its precipitation as iron oxyhydroxide and/or aluminosilicate (Bau 1999; Kraemer et al. 2017). Hypoxic intervals in the Narym and Kolpashevo ironstones, however, exhibit negative Eu anomaly.The negative Eu and Y anomalies are similar to those exhibited by low-temperature hydrothermal fluids (Bau and Dulski 1999; Sylvestre et al. 2017). Enhanced hydrothermal activity possibly led to the extensive emission of iron and methane saturated fluids and induced hypoxia on shelf of the West Siberian Sea. Redox sensitive proxies, REEs patterns, pyrite framboids as well as chamosite-pyrite-siderite mineral assemblages of ironstones indicate hypoxic bottom water condition. Hypoxic intervals in the Narym and Kolpashevo ironstones correspond to global geological events such as OAE-2 and OAE-3, respectively. Therefore, the formation of ooidal ironstones is closely associated with periods of oceanic hypoxia.

5.2 Relationship between ooidal ironstone formation,hypoxia, global volcanism and rifting

Several intervals such as in the Turonian, the early Santonian, the Santonian-Campanian boundary, the early Maastrichtian and the Maastrichtian-Paleocene boundary in the sections of the western and eastern parts of the deposit are marked by high contents of mercury (both Hg and Hg/TOC). Peaks of Hg and Hg/TOC patterns precede or coincide with the Coniacian and Santonian hypoxic intervals. Higher values of Hg and Hg/TOC in well 803 than in well 507 relates to closeness of the exhalation site in case of former. Ore bodies in the western part of the deposit (well 803) occur close to the tectonic uplift and therefore remain close to the exhalation site. Hg anomalies in the Meso-Cenozoic sediments of the West Siberian Sea correlated with the volcanisms of LIPs (such as Greater Ontong-Java Plateau, Karoo-Ferrar Province, High Arctic Province, and Deccan Traps) occurred in the Cenomanian-Turonian,the Campanian, the Maastrichtian-Paleocene (Ernst and Youbi 2017). The close correspondence between Hg anomalies and hypoxic intervals within the Meso-Cenozoic ironstone deposit in West Siberian Sea establishes the relationship between LIPs and OAEs (cf.Percival et al. 2015, 2018; Ernst and Youbi 2017; Scaife et al. 2017; Them et al. 2019). These large volcanic events possibly contributed to the formation of extensive ironstone ore deposits on ocean floor (cf. Van Houten and Arthur 1989; Percival et al. 2015; Ernst and Youbi 2017; Song et al. 2017).

Intense accumulation of ooidal ironstone might take place soon after an oceanic hypoxia (cf. Van Houten and Arthur 1989; McLaughlin et al. 2012; Bekker et al. 2014).Underwater exhalations of iron-saturated fluids would intensify during global volcanic events. Increase in bioproductivity and related marine hypoxia results in abundant ferrous iron in shelf waters (cf. Homoky 2017;Konhauser et al. 2017). The immobilization of this iron as silicates causes the accumulation of chamosite/glauconite-siderite-pyrite ironstones, as in the Narym and Kolpashevo ironstones. While the exhalation of fluids in oxic conditions leads to the rapid precipitation of iron as oxyhydroxides (goethite, lepidocrocite) as in the Bakchar ironstone deposit (Kimberley 1989; Sturesson et al. 2000;Rudmin et al. 2019).Global volcanic events possibly triggered the hydrothermal convection in continental rift systems of the West Siberian Plate (Surkov 2002; Saraev et al. 2011; Vibe et al. 2018). Metal-saturated aqueous fluids moved into the near-bottom waters through convection, similar to the continental rift system of the Dead Sea (Dill et al. 2010), or through underwater exhalation similar to the system of Mahengetang (Heikoop et al. 1996), the Hawaiian Islands (Hearty et al. 2010)and Trinidad Island (Kimberley 1994). The methane fluid degassing possibly accompanied the intensive removal of iron and the formation of ironstones in the geological past (cf. Rudmin et al. 2018).

Table 5 Rare earth elements + Y (both in ppm) and geochemical parameters for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit

Table 5 Rare earth elements + Y(both in ppm) and geochemical parameters for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit (Continued)

Table 5 Rare earth elements + Y(both in ppm) and geochemical parameters for samples of two sections in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit (Continued)

6 Conclusions

Based on the studies of pyrite framboids and major and trace elements concentrations in core samples collected from the two wells penetrating the ironstone deposits in western Siberia, we can draw the following conclusions:1) Pyrite of different sizes occurs in various morphological forms such as framboids, irregular and subhedral masses, and euhedral crystals. Variation in diameter of pyrite framboids reveals three hypoxic intervals within the Meso-Cenozoic Bakchar ironstone deposit in western Siberia.

2) Occurrences of ooidal ironstone correspond to the three hypoxic intervals in the middle Coniacian, the Santonian, and the early Campanian. Trace element enrichment factors exhibit sharp swings in these intervals.Narym and Kolpashevo ironstones encompassing the hypoxic intervals reflect to different mineral assemblages and geochemistry. Therefore the formation of ooidal ironstone is closely linked to the hypoxic shelf water.

Fig. 7 Lithostratigraphic column for well 803 showing geochemical profiles of TOC, Hg, Hg/TOC, elemental enrichment factors (EFs). Correlation between high content of TOC, Hg/TOC, and elevated enrichment factors of Mo, U, V indicates two hypoxic intervals (marked by light grey shades).Paucity of pyrite framboids rules out hypoxia in the Maastrichtian-Paleocene sequences.Form.=Formation;ir.=ironstone

Fig. 8 Variations of REEs patterns through the stratigraphic sequences in wells 570 and 803 from eastern and western (respectively) corners of the Bakchar ironstone deposit. Hypoxic intervals in the Narym and Kolpashevo ironstones exhibit negative Eu and Y anomalies, indicating influence of low-temperature hydrothermal fluids

3) Enhanced hydrothermal activity resulted in lowtemperature fluid emissions and induced oceanic hypoxia during the formation of ironstones in Bakchar ironstone deposit. Redox sensitive proxies, pyrite framboids and ore mineral assemblages record the hydrothermal activation.

4) Mercury anomalies characterize hypoxic intervals,which were contemporaneous with global volcanic events. Thus, a potential geological connection is established between LIPs, OAEs and ooidal ironstone deposits. Hydrothermal exhalation associated with rifting contributes iron for the ooidal ironstone deposits.

Abbreviations

EF: Enrichment factor; ICP-MS: Inductively coupled plasma-mass spectrometry; LIPs: Large Igneous Provinces;OAE: Ocean anoxic event; REEs: Rare earth elements; SEM: Scanning electron microscope; TEs: Trace elements;TOC: Total organic carbon;XRF: X-ray fluorescence microscope

Acknowledgements

The authors thank the anonymous reviewers and editors for their suggestions and revisions, which improved the paper.

Authors’ contributions

MR and AM conceived and designed the study. MR and AR collected the materials. MR, EF, EL, RK, and AR performed the laboratory investigations. MR,SB, EA and AM analyzed the data. SB and EA helped modify the manuscript.MR and SB wrote the paper. All authors read and approved the final manuscript.

Funding

Laboratory geochemical investigations were carried out at National Research Tomsk Polytechnic University within the framework of a Competitiveness Enhancement Program Grant (Project VIU-OG-61/2019). The mineralogical part of research was funded by Russian Foundation for Basic Research and Tomsk Region (19–45-703002).

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

Competing interests

The authors declare that they have no competing interests.

Author details

1Division for Geology, School of Earth Sciences and Engineering, Tomsk Polytechnic University, 634050 Tomsk, Russia.2Department of Earth Sciences,Indian Institute of Technology Bombay, Powai, Mumbai 400076, Maharashtra,India.3Department of Geoscience, French-Azerbaijani University (UFAZ),AZ1000 Baku, Azerbaijan.4Institute of Monitoring of Climatic and Ecological System of SB of RAS, 634055 Tomsk,Russia.

Received: 1 May 2019 Accepted: 11 November 2019