『Abstract
　Pyrite is precipitated within the cells of higher plants and algae at ambient temperatures by diffusion of aqueous Fe(II) followed by reaction with microbial or inorganic S(-II). The absence of any additional oxidizing agent in sufficient concentrations to account for the mass of pyrite formed in the experimental systems suggests that the reaction involves aqueous H2S. Pyrite is formed from undersaturation relative to mackinawite, FeSm, since FeSm is not observed in the cells and it is seen to dissolve in the experimental systems. the overall pyrite forming reaction, starting with Fe(II) as a reactant, appears to be

Fe(II) + H2S→FeS^o + H2S→FeS2p

where FeS^o is the monomeric representation of an aqueous FexSx cluster or reaction intermediary. Although a series of balanced reactions could be written, none of these provides a unique description of the process and reaction mechanisms cannot be determined from this type of empirical experimentation. Pyrite formation is determined by the rate of H2S diffusion into the cells. An apparent diffusion coefficient for H2S of 10^-8 cm² s^-1 is obtained, which includes the effects of tortuosity and effective porosity. The supersaturation limit for pyrite formation in these cells approaches 10¹¹, which is lower than the value found for pyrite nucleation on pyrite seeds. Biological surfaces, such as plant cells, appear to have a greater surface activity than pyrite seeds. However, the supersaturation limit is affected by transport through the diffusion boundary layer, which is a surface-independent common rate-determining process. The rate of pyrite formation in the cells is relatively rapid, with over 50 vol.％ pyrite being formed within the parenchyma cells of Apium sp. within 2 weeks. The rapid formation of pyrite within plant cells contributes to the preservation of cells during fossilization. There is no evidence for particulate iron sulfides penetrating the cells and the nature of the cell wall would preclude this. Greigite, Fe3S4 and mackinawite, FeSm, are therefore not necessarily direct precursors to pyrite formation, which is consistent with their observed limited abundance in sedimentary systems. The direct involvement of stoichiometric quantities of other inorganic components, such as dioxygen and elemental sulfur, was precluded in the experimental design, through the chemistry of the plant substrate and through the exclusion of particulates by the cell wall system. The intracellular chemical environment is complex, however, and trace oxidants may play a role in initiating pyrite nucleation.

Keywords: Pyrite; Plant; Fossilization; Sulfide; Celery; FeS』

1. Introduction
2. Methods
　2.1. Batch experiments with Chlorella
　2.2. Batch experiments with Apium
　　2.2.1. Experiments with celery strips
　　2.2.2. Experiments with whole celery petioles
　2.3. Chemostat experiments with Apium and Platanus
　2.4. Observations of products
3. Results
　3.1. Experiments with Chlorella
　3.2. Experiments with Apium
　　3.2.1. Batch experiments
　　3.2.2. Chemostat experiments
4. Discussion
　4.1. Exclusion of precipitated FeSm
　4.2. Advection and diffusion
　4.3. Mass-volume considerations
　4.4. Fe transport in sulfidic systems
　4.5. Pyrite nucleation
　4.6. Constraints on pyrite formation
5. Conclusions
Acknowledgments
References