Crystal-size distributions and possible biogenic origin of Fe sulfides

Sedimentary greigite (Fe3S4) can form either by “biologically controlled ” or by “biologically induced min­ eralization ” (BCM and BIM, respectively). In order to identify the origin of magnetic Fe sulfides, we studied and compared the sizes and morphologies of greigite crystals produced by a magnetotactic microorganism (previously described and referred to as the “many-celled magnetotactic prokaryote ”, MMP) and Fe sulfides from two specimens of Miocene sedimentary rocks (from £±ka, in the foredeep of the Western Carpathians and from Michalovce, in the Transcarpathian Depression). Greigite grains from the MMP and the £±ka rock show nearly Gaussian crystal-siz e distributions (CSDs), whereas the CSD is lognormal for Fe sulfides from the Michalovce rock. We simulated various crystal-growth mechanisms and matched the calculated and observed CSDs; crystals from the MMP and the £±ka rock have CSDs that are consistent with random growth of crystal nuclei in an open system, whereas the CSD of the Michalovce Fe sulfides is consistent with surface-controlled growth followed by supply-controlled growth in an open system. On the basis of CSDs and characteristic contrast features in the transmission electron microscope, greigite in the £±ka rock is likely of BCM origin, whereas the Fe sulfide crystals in the other rock sample were produced by BIM processes. Our results indicate that the methods we applied in this study may contribute to the identification of the origin of magnetic Fe sulfide minerals in sedimentary rocks. Key-words: greigite, magnetotactic bacteria, biologically controlled mineralization, biologically induced mineraliza­ tion, crystal size distribution. Introduction many cases; magnetite can be produced by both biogenic and inorganic processes, whereas sedi­ Fine-grained magnetic minerals in sediments mentary greigite likely forms as a result of bio­ and sedimentary rocks include Fe oxides and sulgenic processes. Microorganisms mediate the fides such as magnetite (Fe3O4) and greigite formation of minerals in two different ways: in (Fe3S4). The origin of these minerals is unkown in “biologically controlled mineralization” (BCM) the crystals form inside the living cell, whereas in “biologically induced mineralization” (BIM) they form outside or on the cell as a result of the indi­ rect effect of the microorganism’s metabolism (Lowenstam & Weiner, 1989). Apart from a study that showed the presence of presumably BCM greigite in soil (Stanjek et al., 1994), we do not have direct evidence that greigite of BCM origin influences the magnetic properties of sediments and rocks. The goal of this study is to establish cri­ teria that could be used for identifying BCM greigite in geological specimens. BCM greigite is known to be produced by magnetotactic bacteria (Mann et al., 1990; Farina et al., 1990; Heywood et al., 1990). Such microor­ ganisms obtain an evolutionary advantage by forming intracellular, single-domain magnetic crystals that cause the bacteria to be oriented in the Earth’s magnetic field (Frankel et al., 1997). Magnetotactic bacteria generally produce either magnetite or greigite with only one morphological type described to date that produces both (Bazylinski, 1999). Although most studies on magnetotactic bacteria were performed on mag­ netite-producing species, several studies addressed the physical and chemical properties and forma­ tion mechanisms of greigite magnetosomes (Heywood et al., 1991; Pósfai et al., 1998a and b; Lins et al., 2000); from these we know that BCM greigite crystals are characterized by a size range from about 30 to 150 nm, an abundance of defects, and a typical spotty contrast in transmission elec­ tron microscope (TEM) images. Greigite can also form in sediments as a result of BIM processes. Owing to the metabolism of sulfate-reducing bacteria, H2S is present at the level of the OATZ (Oxic-Anoxic Transition Zone) and below. By the reaction of H2S and HS– with dissolved Fe, amorphous FeS precipitates. Through a series of solid-state transformations, the amorphous precipitate converts to mackinaw­ ite (tetragonal FeS) and then either to greigite or pyrite (Morse et al., 1987; Schoonen & Barnes, 1991; Rickard, 1997). Organisms play a role in creating the chemical environment that is neces­ sary for Fe sulfide formation during this process, but they do not control the growth of greigite par­ ticles. As a consequence, such particles likely have a broad size distribution, and their spatial arrange­ ment is non-specific (Bazylinski & Moskowitz, 1997). Although much research has been done to describe the conditions and kinetics of the forma­ tion of BIM Fe sulfides (Rickard, 1995, 1997; Rickard & Luther, 1997), physical properties such as crystal-size distributions and microstructural features of sedimentary BIM greigite remain unknown. Greigite has been increasingly reported as the main carrier of magnetic remanence in sedimenta­ ry rocks (Krs et al., 1992; Roberts, 1995; Jelinovska et al., 1998). For the paleomagnetic interpretation of rock magnetism it is of impor­ tance whether the greigite formed by BIM pro­ cesses or was deposited within magnetotactic bacteria. When these bacteria die, they eventually lyse and their magnetosomes become dispersed and are deposited either in random orientations or oriented by the Earth’s magnetic field. On the other hand, BIM greigite is produced at the sedi­ ment surface or at a moderate burial depth, and the question is when these crystals acquired their chemical remanence. In either case, it is not straightforward to determine whether the primary magnetic-field direction is preserved in the greig­ ite crystals. Bulk magnetic methods are useful for identify­ ing ferrimagnetic Fe sulfides, particularly greigite (the most recent compilation was published by Sagnotti & Winkler, 1999); however, these studies do not provide information regarding the origin of greigite. Bulk magnetic methods can be used to detect the presence of single-domain crystals. In the case of sedimentary magnetite, single-domain grains are usually interpreted as formed by BCM processes. Since the single-domain size range extends to quite large sizes in greigite (up to about 0.8 μm, Hoffmann, 1992), and both BIM and BCM processes produce greigite that is primarily single-domain, bulk magnetic methods do not seem to be very useful for distinguishing BCM from BIM greigite. Since the finding of nanoscale magnetite crys­ tals that were interpreted as possible traces of fos­ sil life in a Martian meteorite (McKay et al., 1996), there is strong interest in defining criteria that can be used for identifying biogenic minerals in rocks. Several characteristics of BCM mag­ netite were recently listed by Thomas-Keprta et al. (2000). These include (1) constrained size (singledomain) and aspect ratios, (2) chemical purity, (3) defect-free structure, (4) occurrence of crystals in chains, (5) unusual crystal morphology, and (6) crystal elongation along specific crystallographic directions. Thomas-Keprta et al. (2000) described a crystal population in the meteorite ALH84001 that is similar to magnetite produced by the mag­ netotactic bacterium strain MV-1 with respect to five of the above six criteria. In our view, criterion (3) may not be useful for identifying BCM mag­ netite, since several magnetotactic strains are known to produce twinned crystals (Devouard et al., 1998). Also, as Thomas-Keprta et al. (2000) noted, even bacterial crystals are not expected to occur in chains once they had been deposited in the sediment after the host bacterium died. Criteria (1), (2), (5), and (6) may be the most useful for dis­ tinguishing BCM magnetite. In addition to the above criteria, the shape of the crystal-size distribution (CSD) can be used for identifying BCM minerals. Certain crystal nucle­ ation and growth processes are known to produce distinct CSDs (Eberl et al., 1998). For example, in an open system surface-controlled growth yields lognormal distributions, whereas in a closed sys­ tem supply-controlled ripening processes produce more symmetric or even negatively-skewed CSDs (Eberl et al., 1998). Previous studies of BCM magnetite showed that the CSDs are typically neg­ atively skewed, with sharp cutoffs towards larger sizes (Devouard et al., 1998), and could be best simulated as if the crystals grew by Ostwald ripen­ ing (Eberl & Frankel, 1999). CSDs of magnetite from some magnetotactic bacteria display two maxima, indicating the role of agglomeration dur­ ing crystal growth; this process may be responsi­ ble for the occurrence of twinned crystals (Arató et al., 2000). Similar information on the CSDs of Fe sul­ fides from magnetotactic bacteria is completely lacking. Therefore, the main goals of the present paper are to characterize the sizes and morpholo­ gies of Fe sulfide crystals from a magnetotactic organism, the MMP, and to compare the results with CSDs obtained from Fe sulfides that occur in sedimentary rocks. The analysis of CSDs provides information regarding the formation mechanism of Fe sulfides in magnetotactic bacteria and can be helpful for identifying BCM greigite in sedimen­ tary rocks.