The Production of Titanium Dioxide Pigment from Waste N.Z. Steel-Making Slag

Abstract

Waste slag produced by the New Zealand Steel Ltd Glenbrook mill represents a significant potential resource of titanium. This report presents the groundwork for the development of a process to recover this titanium as TiO2, which is used extensively as a white pigment. The solubility of the slag in various mineral acids was investigated and an acid sulphation technique developed that will extract up to 90 % of the Ti, Mg, and A1 from the slag in water-soluble form. The slag is sulphated at 250° C and ambient pressure with a 3x stoichiometric excess of 90 % H2SO4 in a constant volume slurry. The metallic elements initially dissolve but are rapidly precipitated as strong-acid insoluble sulphates. The filtered precipitates are leached with water at 70° C to dissolve the metal values. The leachate contains more acid than is desirable for immediate thermal hydrolysis of the titanium. Evaporating the solution to a boiling point of 150° C, equivalent to a sulphuric acid concentration of 65 - 70 % w/w, precipitates 90 % of the titanium as essentially acid-free titanium oxysulphate. The acid : TiO2 mole ratio is reduced from > 4 in the feed solution to 0.5 in the filtered titanium precipitate. The precipitate dissolves completely in water at 70° C Thermal hydrolysis studies were undertaken to determine the optimum conditions for the recovery of titanium dioxide from this solution. The effect of the analytical composition of the titanium bearing liquor on the kinetics of hydrolysis and the particle morphology of the precipitated hydrate was determined. The hydrates display a characteristic dual size distribution, consisting of spherical 0.03 - 0.09 μm diameter primary particles flocculated into large irregular agglomerates 1 - 5 μm in size. For successful calcination to a pigment it is essential that the primary particles are only loosely coagulated. This quality is promoted by hydrolysis from concentrated (150 g/kg TiO2) liquors which are also low in acidity (< 10 % w/w H2SO4), and the use of adequate seed material. The sum of the free and potential acidity of the liquor should not exceed 24 % H2SO4; this is a practical equilibrium limit above which little hydrolysis occurs. The resulting hydrates all possess the anatase crystal structure due to the sulphate content of the precipitating liquor. The kinetics of hydrolysis shows two distinct stages; an initial slow induction period followed by rapid precipitation of the hydrate. The kinetics of the main precipitation stage is first order with respect to the liquor TiO2 concentration and very negatively affected by the concentration of free sulphuric acid. It is also proportional to the cube root of the number of seed nuclei employed suggesting the rate determining step starts at the surface of the hydrate particles. The initial induction period appears to be associated with the establishment of equilibria between dissolved titanium aquo and sulphate complexes. Accompanying calcination trials demonstrate that a mono-dispersed particle size distribution appropriate to a quality titanium dioxide pigment is achievable with suitable hydrates. The pigment crystal dimensions can be regulated by the temperature and introduced potassium level chosen for calcination. Anatase pigments with the optimum optical diameter of 0.25 μm are favoured by calcination temperatures of < 930° C and relatively large (0.7 %) quantities of K2O addition. The colour of the anatase pigments produced are comparable to equivalent commercial samples with L* (whiteness) values of 98.5. The kinetics of particle growth during calcination is more rapid than the transformation of anatase hydrate to the rutile crystal structure. Complete conversion to rutile results in calcined crystals typically larger than 1 μm, far exceeding the optimum optical diameter. Promoter sols are therefore required if rutile pigment is desired. Experiments with the formation and efficacy of such sols are described. A schematic for the proposed process is included along with a mass balance for titanium and other major components of the slag. Suggestions for further research are considered.