Hematite ore reduction to magnetite with CO/CO2 – kinetics and microstructure

Uncertainties and discrepancies are common in the first step of the reduction of hematite ore to magnetite, especially as regards the experimental conditions governing the transition from porous to lamellar magnetite. Hence the problem has been reinvestigated with an ore from Itabira (Brazil). This ore has been carefully characterized with X-ray diffraction, X-ray photoelectron spectroscopy, optical and scanning electron microscopy and electron microprobe analysis. It has been established that α-quartz and goethite are present, and iron combined on the surface with S or Si; most particles, in the range 50–200 μm, are hematite single crystals with 0.1% Al as the only substitution; there are also smaller crystals agglomerated by the gangue. Numerous experiments have shown that the transition from porous to dense lamellar magnetite is enhanced by an increase of temperature and a decrease of CO %. This is explained by a competition between the chemical reaction and solid state diffusion of ions along the hematite-magnetite interface. The shrinking-core model has been applied to the domain of porous magnetite, yielding a chemical rate constant ? = 15.5 exp (?69500/RT), roughly twice lower than in the reduction of hematite synthetic single crystals. The nucleation frequency increases sharply with temperature to ～700°C where it reaches ～10⁹ m^?2·s^?1, as with pure hematite.