Technical Comments

Response to Comment on “Abiotic Pyrite Formation Produces a Large Fe Isotope Fractionation”

See allHide authors and affiliations

Science  03 Feb 2012:
Vol. 335, Issue 6068, pp. 538
DOI: 10.1126/science.1213032


Czaja et al. assert that Guilbaud et al. claim that “the geologic record of Fe isotope fractionation can be explained by abiological precipitation of pyrite.” At no point did we suggest this. We reported a previously underestimated Fe isotope fractionation that contributes to the sedimentary Fe isotope signal.

Czaja et al. (1) underline the danger of using Fe isotopes as a unique tool to interpret ancient ocean biochemistry without consideration of geological and petrographic contexts. This philosophical position, with which we entirely agree, inspired our original work (2) when we noted that interpretations of the Fe isotope composition of sedimentary pyrite lacked any experimental calibration of the Fe isotope fractionation possible during pyrite formation. Indeed, the assumption that pyrite is a passive recorder of the Fe isotope reservoirs from which it was formed is tacit throughout much of the published literature.

The discussion by (1) regarding independent evidence for microbial activity in banded iron formations and Ca-Mg carbonates is entirely relevant for a review on sedimentary Fe isotope signatures, but our paper focused specifically on the process of pyrite formation. In light of our data, 56Fe-depleted pyrite is not sufficient evidence on its own for past microbial activity and therefore emphasizes the necessity to use diverse geochemical and geological tools. Czaja et al. note that in bulk black shales with high [Fe] and a low degree of pyritization, negative pyrite δ56Fe signal would generally be diluted by other Fe-bearing phases. We entirely agree, and it seemed intuitive that where pyrite is a minor species, the bulk composition will not reflect the fractionation associated with its formation but another mechanism. The Fe isotope values plotted in figure 3 of our paper represent pyrite grains and nodules from (3) and not bulk black shales. We noted that the difference in Fe isotope composition between Archean seawater and the mean sedimentary pyrite grains is similar to the fractionation we find in our abiotic experiments. Although this does not preclude the involvement of any other processes, it makes it difficult to deconvolve each of the contributing Fe isotope signatures. Furthermore, studies that have isolated each different highly reactive Fe pool (4, 5) support that δ56Fe pyrite signatures are dependent upon the degree of Fe utilization as pyrite. We strongly encourage researchers to continue such signal extraction before interpreting bulk black shale compositions.

Czaja et al. show some degree of confusion over pyrite-formation mechanism and the presence of FeS in our experiments when they assert that our results are only valid under conditions in which the degree of pyritization is “~2%...of the initial marine Fe(II)aq inventory.” Initially, there is the kinetic [not equilibrium, as Johnson et al. (6) incorrectly suggest] isotope fractionation associated with the exchange of water and sulfide ligands and shift from octahedral to tetrahedral coordination as Fe2+aq reacts to form FeSm. This fractionation produces a 56Fe-depleted condensed phase. Where ΣFe(II)aq exceeds ΣS(-II)aq, or where the Ion Activity Product only marginally exceeds KspFeSm, there is the opportunity for progression toward isotope exchange equilibrium, for which a small 56Fe enrichment is associated with the condensed phase (7). However, the kinetics of this Fe isotope exchange are asymptotic to a metastable disequilibrium value in which the condensed phase is still 56Fe- depleted compared with its aqueous counterpart (8). Czaja et al. take great issue with this fractionation process. In fact, compared with the fractionation associated with pyrite, it is a minor contributor to the overall fractionation, and the difference between the initial FeSm precipitate and one that has undergone Fe isotope exchange is on the order of 0.5 to 0.6‰. The third and most important contributor to Fe isotope fractionation is pyrite formation with a mean fractionation factor of Δ56FeFe(II)RES-pyrite ~ 2.2‰. Therefore, the amount of Fe use as pyrite should not be <2% to produce 56Fe-depleted pyrite, as Czaja et al. assert, but can be as large as 50% (2). We should clarify that in figure 2 in (2), the 10% in column B does not refer to the requirements of a model but refers to the fact that the fractionation presented was recorded in experiments in which 10% of Fe was precipitated by sulfide (8, 9).

Czaja et al. contend that measured fractionations are experimentally induced kinetic effects and note that associated fractionations may be rate dependent. The experimental methodology used has been applied previously to investigations of the kinetics and mechanism of FeS2 formation (10) and has been shown to be robust. The example quoted is for a reversible isotope exchange reaction and is invalid for the unidirectional pyrite-forming reaction. It is unfortunate that Czaja et al. choose a paper documenting “anomalous” FeS accumulations to support their assertion that pyrite formation in sediments is uniformly slow. They do not discriminate between the distinct processes of pyrite nucleation, which initiates pyrite formation, and subsequent pyrite growth, which accounts for most of pyrite formation. The most recent authoritative critical review of the chemistry of the Fe-S system (11) notes that whereas pyrite crystal growth is relatively fast, pyrite nucleation is slow and rate limiting. The causes of variable pyrite nucleation rates in sedimentary environments are diverse and have been reviewed in detail by (12). It is plain from these detailed critical reviews that pyrite-formation rates in modern sediments cannot be summed up by the simple flat assertion that (1) make. Given that, similar speculation on rates in the Archean and Palaeoproterozoic are conjectural.

We conclude that our experimental approach is valid and the results pertinent to the question of Fe isotope fractionation during sedimentary pyrite formation. We do not attribute the geologic Fe isotope record to pyrite formation alone, and we have always recognized that in combination with other isotope data, there is evidence for redox and microbial involvement in the early sedimentary Fe cycle. We reemphasize that a large fractionation can be associated with pyrite formation, and this potential fractionation pathway must be taken into account when interpreting Fe isotope signatures of sedimentary pyrite.

References and Notes

  1. Acknowledgments: This work was funded by an ECOSSE Ph.D. studentship to R.G. and Natural Environment Research Council research grant N E/E003 958/1 to I.B.B.
View Abstract

Navigate This Article