Abstract
The chlorination of Fe203 and Fe304 has been investigated in a fluidised
bed in the range 850- 950°C in order to elucidate the role of
iron in the segregation process for nickeliferous laterites. The
kinetics of Fe203 chlorination in the range using 20% HCl - 80% N2
gas mixtures is consistent with a single- stage starvation control
model, whereas Fe304 exhibits a two- stage reaction mechanism. Oxygen
potential measurements using a CSZ probe located above the bed and
X-ray powder results confirm the proposed reaction scheme.
When a serpentitic nickeliferous laterite (1.03% Ni - 11.7% Fe) was
chlorinated at 900°C with a 2o% HCl - 80% N2 gas mixture it was found
that the oxidic iron in the ore was extracted in a similar manner to
the synthetic iron oxides. Al so, it was found that the nickel
extraction from the ore paralleled that of the iron and the tests
indicated that for the major part of the extraction, the solid nickel
species were in equilibrium with the gas phase.
The results obtained from the chlorination of Fe203 and Fe304 with
2o% HCl - 5% CO - 75% N2 at 900°C indicated that a starvation control
model should again be applicable, but, no completely satisfactory
model was devised. However, when Fe3o4
was chlorinated in more
reducing conditions with 20% HCl - 10% CO - 70% N2 or 10% HCl - 1o%
CO - 80% N2 at 90cPC it was found that the rates of iron extraction
were consistent with those predicted by a starvation control model.
This model assumed equilibrium between Fe3O4/FexO and the gas phase.
On extension of these tests to the nickel laterite or e it was found
that the iron oxides in the ore reacted in an analogous manner to
the synthetic oxides. However, these free iron oxides quickly
disappeared from the ore not only due to the rapid rates of chlorination
but the interaction between iron species and the silicate minerals also.
Hence , initially there was an increase in the iron content of the
silicate minerals, but, after all of the iron oxides had been removed
the iron in the silicate minerals reacted with the gases.
Although increasing the CO content of the reacting gas mixture
resulted in a significantly greater rate of iron extraction, there
was only a slight effect on the nickel chlorination. This can be
explained by applying to the ore a model similar to that developed for
the pure iron oxides.
CO when added to the reacting gas mixture appeared to lower the
ultimate recoveries of both nickel and iron. This apparent effect
is not well understood but could be associated with solid state
changes in the ore.
It has been found that the models devised for chlorination of the
iron oxides and ore can be used to explain the fluidised bed results
of Webb (40) and Brittan and Liebenberg (7). Furthermore, it was
shown by thermodynamic calculations based on the results of Brittan
and Liebenberg (7) that the stability of the iron deposited by
reduction of the fluidised bed effluent gas could be explained by
considering equilibrium of simple gaseous species.