| dc.description.abstract | Ilmenite is an important raw material for titanium-containing products. Currently, it is processed using large amounts of coal but this could be substituted with natural gas. This could potentially increase the efficiency and lower CO2 emissions of the process. Technology from direct reduced iron (DRI) processes which use natural gas could be adapted for ilmenite reduction. In some DRI processes natural gas is reformed to a mixture of CO and H2, so-called synthesis gas, or syngas.
This work investigates the factors which influence the reduction rate and final products of ilmenite reduction with syngas. Previous work focuses mainly on pure CO or pure H2 gases, rarely mixtures of the two. Therefore, the previous models were based only on the single gases. These models should be extended to mixtures of CO and H2. The effects of the water-gas shift reaction (WGSR) could then be modelled. The influence of impurities, especially Mg, should also be included in the model since previous studies indicate that this could change the reduction rate. The aim of the present work is to determine which factors are most important for reduction of ilmenite in synthesis gas and then develop a model which can be used to understand the kinetics of the reactions.
Pellets of preoxidised ilmenite concentrates were reduced in batches of 200 g in a vertical retort furnace. The mass loss of each experiment was measured as a function of time to show the reaction progress. The gas composition, temperature and ore source were the main variables which were investigated. The final products were analysed for chemical and phase composition. Microanalysis was also done with an optical microscope, a scanning electron microscope and an electron probe microanalyser.
From experimental work, it was found that reduction has two stages the first is very fast and the second is noticeably slower. These stages correspond to reduction from Fe3+ to Fe2+ and Fe2+ to metallic iron. Low temperature and low H2 content in the gas results in slower reaction rates. However, investigations at low temperatures are affected by carbon precipitation in the crucible. This occurs to some degree at 880 ◦C but not at 1000◦C. Under strongly reducing conditions, there is some degree of titanium reduction in addition to iron reduction.
When 8 % CO2 and/or H2O is added to the reducing gas the second stage of reduction reaches a limit very quickly. Only a small amount of iron metallises from the reaction of ulv¨ospinel to ilmenite and metallic iron. Ilmenite is not reduced with this amount of CO2 and/or H2O.
Analysis of the final products shows that iron metallises to very finely distributed particles when H2 is present in the reducing gas. As well, an M3O5 phase containing Mg-, Fe-, and Ti-oxides forms in the final product due to the presence of Mg. This additional phase complicates the analysis of the final product since reduced titanium may also be present in the M3O5 solid solution. Formation of this phase also hinders metallisation of iron oxides. The presence of Mg and subsequent formation of the M3O5 phase decreases the reaction rate and overall conversion degree.
A modified shrinking core model was developed to better understand the kinetics of reduction. The model accounts for the presence of both CO and H2 in the gas as well as reduction of both iron and titanium oxides. The WGSR is also integrated in the model. The driving force in the model is the oxygen potential difference between the gas and the solids. Using appropriate mass transport and chemical reaction kinetic parameters the model fits very closely to the experimental results. It was found that Mg mainly affects the chemical reaction rate parameters and Fe mainly affects the mass transport parameters. The alternate reactions which include ulv¨ospinel and the M3O5 phase were not modelled explicitly but were incorporated into the kinetic parameters.
Experimental work suggests that the WGSR is at equilibrium. The model indicates that oxygen flows from the carbon species toward the hydrogen species to achieve equilibrium. However, further refinements are necessary to model the kinetics parameters of the WGSR.
This work confirms that ilmenite reduction is possible with syngas. The gas composition, temperature and chemical composition of the solids are among the most influential factors during reduction. The experimental work and subsequent model form a basis towards future work for designing an industrial process. | nb_NO |
