INTRODUCTION

Banded iron formations (BIFs) are layered stratigraphic units that consist of Fe- and Si-rich intercalated bands formed over most of the Precambrian. The conception of their sedimentary origin is dominant [1, 2], although it is under discussion (e.g., [3]).

The interpretation of the conditions of the early Earth is critically related to understanding the trajectory of BIF formation: are they a result of microbial activity or not? On the one hand, the evidence of the influence of microbial activity on BIF sedimentation indicates conditions favorable for life, especially nutrient-rich oceans with near neutral pH values. On the other hand, if the formation of BIFs is abiotic, the Precambrian oceans could have had more alkaline pH values and could be nutrient-poor [4]. There is also a viewpoint on acidic reducing geochemical conditions in the Archean responsible for the formation of dissolved forms of Fe and Si, which were trapped in ancient oceans and formed ferruginous–siliceous rocks [5].

The study of the Fe and Nd isotopic compositions of certain deposits showed that the BIF ores contain Fe from various sources [6, 7]. Firstly, continental Fe was mobilized at the continental margin probably by microbial dissimilation reduction. Secondly, Fe was derived from submarine hydrothermal vents. The significance of these two Fe sources could be equal, although their proportions could vary over time.

The S isotopy of BIF sulfides could be an important link in the chain of evidence of BIF formation. The S multi-isotopic composition is a powerful tool for the study of photochemical and biological processes that controlled the Archean S cycle, and allows drawing a conclusion on related atmospheric and marine conditions. Below, we present the first data on the S multi-isotopic composition of sulfides from the Neoarchean (2760–2740 Ma [8]) BIFs of the Kostomuksha greenstone belt of Karelia (Karelian Craton of the Fennoscandian Shield).

MATERIALS AND METHODS

Geological Setting

The Kostomuksha greenstone belt occurs in the western part of the Karelian Craton (Fig. 1a) at the boundary of the Central and West Karelian (Kianta) terranes [9]. Its elongated (in plan) longitudinal structure (Fig. 1b) is traced for 25 km with a width of 4.5–7.0 km and a general dip to the east. The belt contains two lithostratigraphic (stratotectonic) associations: the Kontok and Gimol groups. The Kontok Group mostly includes metamorphosed basalts–komatiites with interlayers of rhyolite and sedimentary rocks, which compose the western and central parts of the belt (Fig. 1b). The Gimol Group composes the eastern wall of the greenstone belt and consists of metamorphosed sandy-clayey flyschoid sedimentary rocks with BIF layers. Its lower part (Kostomuksha Formation) hosts the thickest BIF layers, which are the basis of iron deposits of the region and are the object of study in this paper. According to the most recent data, the Gimol Group of the Kostomuksha greenstone belt formed 2760–2740 Ma ago, probably in suprasubduction setting [8]. It is not excluded, however, that the BIFs are the fragments of an oceanic plate cover inside flyschoid rocks of the accretionary prism taking into account their similarity with banded siliceous sequences of the Phanerozoic oceans in accretionary prisms of subduction zones. The BIFs consist of intercalated thin (few millimeters to few centimeters) red, yellow, or creamy layers of cherts or jaspers containing black and dark gray Fe oxides, which are absent in siliceous sequences. They are similar in the presence of exclusively clayey clastic material and the low rate and the high duration of sedimentation, which is identified on the basis of SHRIMP U–Pb analysis of zircon, e.g., for the BIFs of Western Australia [10]. The BIFs and siliceous–jasper sequences are characterized by a high lateral extension of fine layers and abundance of biogenic sediments, which require calm sedimentation conditions.

Fig. 1.
figure 1

Scheme of the geological structure of the Fennoscandian Shield (a) and Kostomuksha greenstone belt (b), modified after [8, 9]. (a) 1, Archean crust; 2, Paleoproterozoic crust; 3, Caledonides, Baikalides, and Neoproterozoic rocks; 4, Archean greenstone and paragneiss belts. (b) 5, Neoproterozoic (Riphean) lamproite and kimberlite; 6, Paleoproterozoic (2.4 Ga) dolerite; 7, Neoarchean (2.72–2.71 Ga) granite; 8, Neoarchean (2.78 Ga) granitoid of TTG association; 9, metagraywacke (2.75 Ga) with BIFs (Kostomuksha and Surlampa formations); 10, sills and dikes (2.75 Ga) of metarhyolite (Kostomuksha Formation); 11, Mesoarchean (2.84‒2.78 Ga) basalt and basalt–komatiite (Ruvinvaar Formation); 12, Mesoarchean tuff, tuffite, and rhyolite–rhyodacite with interlayers of BIFs and carbonaceous shales (Shurlovar Formation); 13, Mesoarchean basalt and komatiite (Niemiyarva Formation); 14, faults: a, proven; b, inferred; c, thrusts; 15, sampling places.

Full size image

Samples for isotopic studies were collected from outcrops in operating open pits and drill cores of exploration wells. The mineral composition of rocks and ores was studied in polished sections using optical and electron microscopy.

S Isotopic Analysis

The S isotopic composition was determined in sulfides (mostly pyrrhotite and pyrite) at the Laboratory of Stable Isotopes of the Analytical Center, Far East Geological Institute, Far Eastern Branch, Russian Academy of Sciences (FEGI FEB RAS, Vladivostok, Russia), using a local laser method [11]. The ratio of S isotopes was measured on 127 (32SF\(_{5}^{ + }\)), 128 (33SF\(_{5}^{ + }\)), and 129 (34SF\(_{5}^{ + }\)) masses in a three-beam regime on a MAT-253 mass spectrometer. The measurement results δ33Smeas‰ and δ34Smeas‰ are given relative to the VCDT international standard. The measurement errors of the S isotopic composition in sulfide inclusions with a spatial resolution of ~100 µm were ±0.20‰ (1σ), ±0.15‰ (1σ), and no more than ±0.05‰ for δ34S, δ33S, and Δ33S, respectively.

RESULTS

Figure 2 and Table 1 show the isotopic characteristics of sulfide sulfur from the Kostomuksha iron deposit of this work. Several generations of Fe sulfide are recognized in the studied rocks and ores. Small cubic pyrite crystals in nondeformed magnetite ores are probably syngenetic with sedimentation. Pyrite is associated with fine-grained magnetite and formed involving sulfur related to primary sedimentation. This pyrite has negative δ34S and positive Δ33S values (Table 1, Fig. 2). In zones of late overprinted deformations and recrystallization, the early pyrite occurs as relics in the assemblage with large pyrrhotite grains. In these samples, the pyrite relics retain high positive Δ33S values and negative δ34S values, whereas the Δ33S values of pyrrhotite are close to 0 (Fig. 3, sample 669-5а).

Fig. 2.
figure 2

S isotope ratio of sulfides of the Kostomuksha iron deposit. (1) Pyrite from non-deformed magnetite ores; (2) pyrrhotite from stringer-disseminated, pocket, and brecciated sulfide ores; (3) pyrrhotite from vein injections. The Archean trend line (Δ33S = 0.89δ34S) is after [12].

Full size image
Table 1. Representative S isotope analysis data of sulfides from BIF deposit rocks, Kostomuksha Greenstone Belt of Karelia
Full size table
Fig. 3.
figure 3

Sulfide types, polished section in reflected light. (a) Relicts of primary pyrite (Py) and newly formed pyrrhotite (Po) in magnetite (Mgt) ore (sample 669-5a). (b) Vein-disseminated pyrrhotite ore in biotite slate (samples 1921-15.4).

Full size image

Pyrrhotite from stringer-disseminated, pocket, and brecciated sulfide ores in biotite schists is another sulfide generation (Fig. 3, sample 1921-15.4). The sulfides mainly include pyrrhotite and rare chalcopyrite. This pyrrhotite is mostly characterized by negative δ34S and Δ33S values (Table 1, Fig. 2).

Pyrrhotite from vein, cut, layer-by-layer pyrrhotite–quartz injections exhibits the narrow range of δ34S values around the meteoritic standard and Δ33S values close to zero (Table 1, Fig. 2).

DISCUSSION

Previous studies of S isotopes in Neoarchean iron deposits of the Kostomuksha greenstone belt led to the conclusion that the S source was related to volcanism. The δ34S value of sulfides from the deposits studied in this work, however, deviates significantly from the meteoritic standard. The high 32S content in carbon-bearing rocks was interpreted as the beginning of microbiological sulfate reduction [13].

Our results show that sulfur in sulfides has a polygenic source. The presence of traces of mass-independent fractionation of S isotopes (S-MIF) indicates the presence of sulfur involved in photochemical reactions in the Archean oxygen-free atmosphere of the Earth. The photolysis of volcanic SO2 yields S-MIF and forms two different reservoirs with reduced (elementary S8 with positive Δ33S and δ34S values) and oxidized (sulfate with negative Δ33S and δ34S values) sulfur.

The positive Δ33S values of first-generation sulfides of the Kostomuksha iron deposit indicate the genetic link of pyrite sulfur with photolytic elementary sulfur. After elementary sulfur was trapped by seawater, it precipitated onto the seafloor and was transformed to pyrite.

Pyrite, however, cannot directly form from elementary sulfur particles [14]. It is considered that this process requires precursors: either an Fe monosulfide similar to mackinawite (FeS) or a polysulfide similar to greigite (Fe3S4) [15]. The precursor minerals are dissolved with the formation of aqueous complexes of FeS, which further react with H2S or polysulfides with the formation of pyrite [15].

Before the reaction with the dissolved precursor of pyrite, the molecules of elementary sulfur require an intermediate stage to break-up the S8 rings [15]. It was noted that the positive Δ33S values are often related to the presence of a high Fe content in the host rock, which allows us to suggest the important role of iron in the break-up of S8 rings [12, 16]. This process occurs in sedimentary pore waters, where S8 rings and sulfur chains and compounds are transformed into H2S, e.g., by means of disproportionation [17, 18]. The resulting H2S is involved in the formation of pyrite, which is characterized by a positive Δ33S signature and negative δ34S values. The formation of pyrite by bacterial transformation of elementary sulfur explains the negative δ34S values of the studied pyrite samples.

The second-generation sulfides are characterized by negative Δ33S values indicating the genetic link of sulfide sulfur with a reservoir of photolytic sulfate sulfur. The negative Δ33S signature is typical of pyrrhotite formed at high temperature. In this case, the processes responsible for the transfer of the S-MIF signal from photolytic sulfate to pyrrhotite are related to high-temperature thermochemical reduction of photolytic sulfate in seawater. These processes can be associated with the evolution of hydrothermal activity due to volcanism in this region. It is noteworthy that, in contrast to microbial transformation of sulfur, these processes produce significantly weaker fractionation of δ34S between sulfate and sulfide. The negative δ34S values of pyrrhotite are thus in agreement with the negative δ34S values of the S source, i.e., marine photolytic sulfate sulfur.

The metamorphic–metasomatic recrystallization of BIFs and interlayers of host rocks, which was accompanied by a gain of ore components, led to the formation of dispersed dissemination and lenticular aggregates along the foliation planes of pyrrhotite and rare pyrite and arsenopyrite with low negative δ34S values and both positive and negative Δ33S values close to zero. This S isotopic composition is a result of mixing of sulfur from various sources reflecting the local conditions of ore formation, which can change over time. The S isotopic composition of some sulfides from late pyrrhotite–quartz veins in BIFs is close to troilite (Δ33S ≈ 0). There are similar veins, where pyrrhotite has Δ33S ≠ 0, i.e., the S isotopic composition is inherited indicating mobilization of matter from the host sedimentary rocks. The presence of Δ33S and δ34S values close to zero, however, does not prove the presence of the mantle source, but does not exclude it.

CONCLUSIONS

Our results show that sulfur for BIF sulfides was sourced from the atmosphere (photolytic elementary sulfur), seawater (atmospheric photolytic sulfate), and mantle (magmatic sulfur). The S isotope ratios of sulfides reflect the interaction between abiotic (atmospheric, hydrothermal) and biotic (microbial dissimilation reduction) processes during the formation of iron deposits of the Kostomuksha greenstone belt of Karelia.