The replacement of magnetite by hematite (i.e., martite) is commonly observed in various geologic systems. In contrast to the formation of martite through oxidation, numerous laboratory experiments have demonstrated that it can also occur by a redox-independent reaction. However, a refined process regarding the expected precursors and the epitaxial crystallography from magnetite dissolution to hematite precipitation remains unknown. In this work, high resolution transmission electron microscope (TEM) has been used to investigate the porous martite from the Baishiya iron skarn deposit, East Kunlun orogenic belt, building a non-classical growth model for martite formed by redox-independent reactions. Initially, Fe
2+ was leached out from magnetite by a reducing hydrothermal fluid to form ferrihydrite. Ferrihydrite was not stable, and could have been transformed to goethite and hematite nanocrystals during a gradual decrease of the local solution pH. At the dissolution front, the exposure of high-energy (110) facets of magnetite was more likely to occur, presumably providing a substrate for the nucleation of pseudocubic hematite nanocrystals confined by (012) facets in the presence of impurities (i.e., Si, Ca, and Cl). Repeated attachment of hematite nanocrystals in the exposed magnetite surface began to fuse and thus form hematite mesocrystal. Meanwhile, leaching of ferrous ions might have increased the permeability of iron ores, resulting in an increase of meteoric water flow with dissolved oxygen. This, in turn, could have oxidized ferrous ions, and subsequently reduced the permeability until an equilibrium state between porosity and oxygen flux has been achieved, reducing the volume change. With a continuous fluid supply, the reaction front of magnetite continued to migrate inward, resulting in a
coarse graining of hematite mesocrystal through chemical exchanges (i.e., recrystallization, Ostwald ripening, and intercrystalline diffusion). Finally, a bulk matite crystal could form by this non-classical reaction pathway,which is different from the monomer-by-monomer addition of ions in a classical crystal grwoth model. Therefore, our work suggests that martite formed by redox-independent reactions has a crystallographic fingerprint (i.e., (110)
magnetite || (012)
hematite) different from martite formed by redox reactions (i.e., (110)
magnetite || (110)
hematite). More importantly, due to the presence of ferrihydrite, goethite, and hematite mesocrystal, martite by particle attachment may have geochemical properties different from those formed by the monomer-by-monomer addition, which might have been overlooked in geologic systems.
发表评论