AU752851B2 - Production of synthetic rutile by low temperature reduction of ilmenite - Google Patents

Description

P/00/01il Regulation 3.2

AUSTRALIA

Patents Act 1990

ORIGINAL

COMPLETE SPECIFICATION eve* 0..

0000S .00.

STANDARD PATENT Invention Title: ilmenite Production of synthetic rutile by low temperature reduction of The following statement is a full description of this invention, including the best method of performing it known to us: FHPMELC6991 75028.8 C6/98147003.1 1A "Production of Synthetic Rutile by low temperature reduction of ilmenite" Field of the Invention This invention is concerned with the production of synthetic rutile by low temperature reduction of pre-oxidised titaniferous material, typically but not exclusively ilmenite.

Background Art Ilmenite is the most commonly occurring titanium-containing mineral. It is an oxide of titanium and iron, most simply represented as FeTiO 3 but often also written as FeO.TiO 2 Apart from iron, there are typically a number of other 10 impurities in much smaller amounts which vary from deposit to deposit, including silicon, manganese, magnesium, aluminium and vanadium. In altered or weathered ilmenites, thorium can also be present at levels which are now considered unacceptable for downstream processing.

A number of processes have been proposed for upgrading ilmenite to 15 synthetic rutile, typically having a TiO 2 content greater than 90%. Synthetic rutile is a suitable feed for the chloride process route to the production of pure TiO 2 a feedstock for pigment and other valuable commodities.

The most widely practised synthetic rutile process in Australia, a major commercial source of ilmenite, is commonly known as the Becher process. In this process, ilmenite concentrate is reduced in a rotary kiln, with coal as reductant and at a temperature typically in the range 1120 11600°C. The iron in the ilmenite is converted largely to metallic iron. The resultant reduced ilmenite is then subjected to aeration leaching, in which the metallic iron is oxidised to a readily separable form in the presence of a dilute ammonium chloride solution with air sparging. A sulphuric acid leach then removes some of the residual iron, as well as other impurities including manganese and silicon.

In the earlier practice of the Becher process, it was thought beneficial to C6/98147003.1 P: I 2 oxidise the ilmenite in air at a temperature around 10000C but this pre-oxidation step was omitted by all or most processors some years ago. Such pre-oxidation is detailed, for example, in Bracanin et al, 'The Development of a direct reduction and leach process of ilmenite upgrading" 101st AIME Meeting San Francisco, 1972. This paper reported that pseudobrookite was the principal component after a sufficient duration of oxidation, eg. 3-4 hours.

Another group of known processes involve a reduction step using a gaseous reductant, typically hydrogen, at a relatively lower temperature, usually around 750-9000C. For example, US patent 4097574 discloses preoxidation in air in a rotary kiln or fluid bed furnace at a temperature in the range of 592-8700C, followed by hydrogen reduction of the iron largely to metallic iron, preferably at a temperature in the range of 787-8451C. An aeration leach is optionally followed by an acid leach.

The Murso process is described, eg, in British patent 1225826, and 15 involves a pre-oxidation step similar to that in US4097574, but the gaseous ":reduction at 850-900oC is predominantly to ferrous iron and is followed by a single hydrochloric acid leach. Another process involving partial reduction is disclosed in Australian patent application 64383/96.

In these low temperature processes with gaseous reductant (in comparison 20 to the Becher process with solid reductant), and indeed in the older form of the Becher process, the pre-oxidation has been thought to be beneficial, especially with primary as distinct from altered or weathered ilmenites, by breaking down the larger ilmenite sub-grains in a way which renders the mineral more amenable to subsequent reduction and leaching. It is, furthermore, conventional wisdom in low temperature processes, notwithstanding the higher pre-oxidation temperature known to have been adopted in the earlier practice of the Becher process, that the maximum temperature in the oxidation kiln is 9000C if sintering effects are to be avoided: both US4097574 and GB1225826 state an upper temperature limit of 8700C for the pre-oxidation step. A similar process disclosed in example 6 of Australian patent 639089 mentions pre-oxidation "at 900C" and a reduction 094047262 3 temperature "less than 900°C".

An objective of all synthetic rutile processes is to maximise the TiO 2 content of the end-product. Once that content is greater than 90%, even incremental improvements in the TiO 2 proportion can be reflected in a much higher product price. It is an object of the present invention to obtain a high grade, ie very high TiO 2 content, synthetic rutile from a primary ilmenite.

Disclosure of the Invention The invention entails the realisation that, contrary to the prevailing view in relation to low temperature synthetic rutile processes, the pre-oxidation step can and should be carried out at a higher temperature, specifically above 900°C and preferably in the range 925 to 1000°C, with measurable effect in terms of TiO 2 content of the synthetic rutile product.

The invention accordingly provides in one aspect a process for upgrading an iron-containing titaniferous material to produce a synthetic rutile, which includes 15 roasting the titaniferous material at an elevated temperature in an atmosphere and under conditions such that iron in the material is oxidised largely to a ferric state whereby to produce an oxidised titaniferous material. This oxidised titaniferous material is treated, in an atmosphere containing a reducing gas, at a temperature in the range of 600-900°C so as to reduce the iron in the material largely to 20 metallic iron, whereby to produce a reduced titaniferous product. The reduced titaniferous product is cooled and the cooled product is subjected to further treatments including one or more leaching treatments for separating out metallic iron and other impurities, whereby to produce a synthetic rutile residue. The oxidation roasting is carried out at a temperature sufficiently above 9000°C for the titaniferous material to be converted by the roasting from a structure in which M 2 0 3 is the major phase to a structure in which M 3 0 5 is the major phase, whereby to enhance the amenability of the reduced titaniferous product to separation of 0 RA impurities by the leaching treatment(s), and to increase the TiO 2 content of the synthetic rutile residue.

004047262 3a Preferably, the temperature of the titaniferous material during said roasting is selected so that the titaniferous material is converted by said roasting from a structure in which M 2 0 3 is the major phase to a structure in which M 3 0 5 s is the ego 9 06/98147003.1 4 major phase.

It is believed that prior pre-oxidation treatments in lower temperature synthetic rutile processes resulted in a structure in which M 2 0 3 predominated relative to M 3 0 5 ie. in effect oxidation was only partially effected. The aforementioned Bracanin et al paper described pre-oxidation in the context of high temperature processes using coal as reductant, and did observe that pseudobrookite (ie M 3 0 5 was the predominant species after 3-4 hours oxidation.

It was not, however, appreciated by those carrying out low temperature synthetic rutile processes with gaseous reductant that the pre-oxidation in these processes was producing a predominant M 2 0 3 structure. The present inventors have appreciated this fact, and that it is significant, and have further realised that a relatively high temperature pre-oxidation has a beneficial effect on the subsequent •low temperature reduction step with gaseous reductant. It is thought that the temperature above which M 3 0 5 becomes dominant varies with the mineral source, 15 but it may typically be in the vicinity of 920-9300C.

According to a second aspect of the invention, therefore, there is provided a process for upgrading an iron-containing titaniferous material to produce a synthetic rutile, which includes roasting the titaniferous material at an elevated temperature above 9000C in an atmosphere and under conditions such that iron in the material is oxidised largely to a ferric state whereby to produce an oxidised titaniferous material. The oxidised titaniferous material is treated, in an atmosphere containing a reducing gas, at a temperature in the range of 600- 9000C so as to reduce the iron in the material largely to metallic iron, whereby to produce a reduced titaniferous product. The reduced titaniferous product is cooled and the cooled product is subjected to further treatments including one or more leaching treatments for separating out metallic iron and other impurities, whereby to produce a synthetic rutile residue. The oxidation roasting is carried out at a temperature selected whereby the titaniferous material is converted by the oxidation roasting from a structure in which M 2 0 3 is the major phase to a structure in which M 3 0 5 s is the major phase.

Subsequent disclosure is applicable to both aspects of the invention.

06/98147003.1 Preferably, during the oxidation roasting the temperature of the titaniferous material is above 9250C, but less than 10000C. A higher temperature increases the likelihood of sintering.

It is found that the optimum temperature varies with the source of the feed material, but is generally in the range 925-9750C.

The oxidation mechanism results in the presence of a number of different oxide species, but may be essentially represented by the following equation: 2FeTiO 3 1202 Fe 2 TiO 5 TiO 2 (1) which may be written: 10 2M 2 0 3 V20 2

M

3 0 5 TiO 2 (2) where M represents any metal species such as Fe, Mn, Mg, Al or Ti, and in general includes impurities.

Preferably, the oxidation roasting is carried out under conditions to discourage accretion formation. A de-agglomerating agent may be added to the oxidation roast step. This may be effected by utilising a circulating fluidised bed for the roasting step. It is believed that the high level of turbulence, and the near material circulation, characteristic of a circulating fluidised bed assist in countering the formation of accretion deposits.

In general, either or both of the oxidation and reduction steps are carried out in respective circulating fluidised beds.

The oxidation step is preferably carried out in an oxygen containing atmosphere, more preferably air for which steps are taken to control its oxygen potential, eg. either by oxygen enrichment and/or by relying on a raised pressure, eg 10 bar or so, and/or adding diluents.

The preferred reductant for the reduction step is an atmosphere which is substantially a hydrogen atmosphere, preferably substantially pure hydrogen.

Carbon monoxide is less suitable because iron carbides may be formed during the reduction stage, which adds to leaching costs. The hydrogen may be supplied mixed with a relatively inert gas.

Typically with hydrogen reduction a pressure above atmospheric is required, eg. about 10 bar, but possibly lower or higher, to improve process economics.

The reduction reactions result in a variety of oxide species. However, the 10 key representative reactions are as follows: TiO 2 Fe 2 TiO 5

H

2 2FeTiO 3

H

2 0 (3) Fe TiO 3

H

2 Fe TiO 2

H

2 0 (4) The first of these may be written: TiO 2

M

3 0 s H 2 2M 2 0 3

H

2 0 where M represents any metal species, and in general includes impurities.

The preferred temperature range for the reduction step is 7500 to 9000C. In general, it should be as low as achievable to optimise the leaching steps, but, as is known, the lower the temperature the lower the water vapour tolerance for the reduction reactions. The selected temperature will therefore be a compromise between these opposing effects.

The metallisation of the iron achieved in the reduction step is preferably greater than 80%, more preferably greater than It is preferred that the leaching treatments include, firstly, an aeration leach to oxidise the iron metal to a readily separable oxide, and then an acid leach, eg.

C6/98147003.1 7 with a mineral acid, to remove a significant proportion of the residual iron and other impurities. The acid may be, e.g. sulphuric acid, depending on the composition of the ilmenite. Alternatively, e.g. hydrochloric acid might be employed. The aeration leach may be conventional eg. similar to that used in the Becher process, employing ammonium chloride as catalyst.

The process parameters are preferably such that the TiO 2 content of the synthetic rutile produced is greater than 90%, more preferably about While not wishing to be bound by the following, which is presently merely a theory, applicant's present understanding is that the effect of the higher 10 temperature pre-oxidation, and the predominant M 3 0 5 structure of the oxidised titaniferous material, may be a finer, grain structure which is conducive to enhance the reduction reactions, especially with a gaseous reductant. On conversion as

M

2 0 3 in the reduction step to rutile and metallic iron (equation 4 above), the result is a more open rutile lattice with a finer metal phase, both aspects enhancing metal oxidation and separation in the aeration leach and enhanced access to impurities and sites during the acid leaching process. These effects may arise because the residual M 2 0 3 phase is more finely dispersed within the reduced ilmenite grain, which aids in the removal of impurities by acid leaching.

It is further thought that the oxidation mechanism results, after reduction, in ee0o 20 a titanate lattice of improved integrity and uniformity, giving reduced production of TiO 2 fragments during reduction, aeration and leaching, and so higher TiO 2 recovery into the end-product.

Example 1 Samples (120g) of three different ilmenite concentrates were oxidised at temperatures of 8500C, 9000C, 950'C and 10000C in a bubbling fluidised bed utilising a 12.5% 02 and 87.5% N 2 gas mixture as fluidising/oxidising gas, at a total flow rate of 6 1/min at STP. The gas velocity was 0.8 m/sec. Analyses (wt%) of the samples are set out in Table 1.

C6/98147003.1 8 The reaction was allowed to run to completion of oxidation. PXRD patterns were recorded for grab samples from the oxidation runs. Quantitative phase analyses were made in the final oxidation runs, and the results are presented in Figures 1, 2 and 3.

It will be seen that, in each case, there is a sharp onset of a conversion from a structure in which M 2 0 3 dominates relative to M 3 0 5 and the converse. This occurs at a temperature between 900°C and 950 0 C. The proportion of TiO 2 also declines. The fourth phase depicted is an intermediate species indicated as "HT239", ie. Fe 2 Ti 3

O

9 This is largely eliminated by oxidation at 10000C.

The figures reveal how prior pre-oxidation carried out at temperatures typically around 8500°C in fact produced little conversion of the M 2 0 3 structure to

M

3 0 5 and that temperatures above 900°C, and preferably in the region of 9500C, are needed to ensure predomination of M 3 0 5 relative to M 2 0 3 It might be expected, then that previously predicted benefits of pre-oxidation to the reduction step would not be seen unless pre-oxidation was effected above 9000°C. This is indeed demonstrated by the next example.

Table 1 Analyses of Ilmenite Samples Capel primary Old Hickory A CRL Analysis TiO 2 54.3 56.7 50.1 Fe(t) 30.1 29.1 32.7 FeO 22.7 20.4 25.3 MnO 1.50 1.67 1.56 MgO 0.21 0.11 0.94

AI

2 0 3 0.84 0.53 1.09 SiO 2 0.93 0.32 0.63 Cr 2 0 3 0.04 0.06 1.10 Th ppm 131 41 27 C6/98147003.1 9 Example 2 The oxidation products (65g samples) of Example 1 for the Capel primary and CRL concentrates were mixed with limestone (2g) and reduced with hydrogen at 8500°C in a bubbling fluidised bed and at a linear gas density of 0.6m/sec. Grab samples were taken at 5, 10, 15 and 30 minutes. The lime was separated from the reduced ilmenite (RI) by magnetic separation and metallic iron content of the RI was measured by Satmagan. The results are set out in Table 2.

The results in Table 2 show that the pre-oxidation temperature has an effect on metallisation rates. The influence is greater for the CRL concentrate, where an increase in pre-oxidation temperature from 850 to 950°C gives a 25% increase in average metallisation rate. Further increase of the oxidation temperature to 1000°C causes the metallisation rate to decrease slightly, possibly due to sintering effects.

Table 2. Metallic iron in Capel and CRL grab sample Rl's 15 corresponding to different pre-oxidation treatments Reduction Preoxidation Reduction Time (min) Run no. Temp. °C 5 10 CRL conc.

FB323 850 5.7 16.0 26.2 FB316 900 6.8 16.1 27.3 FB317 950 7.4 19.9 31.4 FB318 1000 7.5 18.0 30.0 Capel conc.

FB319 850 5.3 14.7 23.5 FB320 900 6.2 15.9 24.9 FB321 950 7.8 17.0 26.7 FB322 1000 7.1 18.0 26.5 C6/98147003.1 <d, Example 3 The reduced ilmenites (RI's) from the eight reduction runs of Example 2 were demetallised in 5% H 2

SO

4 and then leached in refluxing 15% HCI for 2 h, with regular withdrawal of solution aliquots for analysis. The procedure entailed heating a starting volume of 85 ml of the chosen strength acid to boiling point in a round bottom 250 ml glass reactor fitted with a mechanical stirrer (run at 500 rpm) and a condenser, then adding 10 g of demetallised RI sample. After different selected time intervals, the stirrer was turned off, the slurry was allowed to settle for a few seconds, and then a 5 ml aliquot of solution was withdrawn using a pipette. The solution was transferred to a volumetric flask and made up to 250 ml with dilute nitric acid, then analysed for Fe, Ti, Mg and Mn by ICP-AES (atomic emission spectroscopy) analysis. Calculations of the metal atom removal into solution after each time interval were corrected for the aliquots previously removed. At the end of the leaching (2 to 8 the remaining leach solution with suspended solids was decanted and filtered. The fine solids on the filter paper were washed with 3 x 100 ml water and 2 x 50 ml ethanol and dried at 1000C.

The coarse solids were washed progressively with 2 x 50 ml ethanol 10% HCI, 3 x 400 ml water and 2 x 50 ml ethanol, then dried at 100°C and combined with the fines for XRF analyses. The XRF analyses of RI's, DP's (demetallised products), and leach products (ie synthetic rutile SR) for both Capel and CRL samples are given in Table 3. Results on the extraction of Fe, Mn, Mg and Ti from the solution analyses are reported in Table 4.

The grade increases with increasing pre-oxidation temperature, from 95.6 to 96.3% TiO 2 for the CRL samples and from 96.7 to 98.0% TiO 2 for the Capel samples. The higher grades at the higher pre-oxidation temperatures are due mainly to lower residual iron levels. This was evident already in the demetallised products, where a decrease or more than 1% in the total residual iron was observed in response to decrease in the pre-oxidation temperature. An inspection of the XRD patterns of the demetallised products showed that this is due to a decrease in residual metallic iron. XRD patterns of the SR products also showed more than twice as much residual metallic iron in samples derived from the lower C6198147003.1 0 0 0 go .00 00 *50 0 go 11 temperature pre-oxidations. Thus the lower pre-oxidation temperatures promote the metallisation of some of the iron in regions inaccessible by the leaching solution.

The effect of pre-oxidation temperature on leaching kinetics was also plotted for these experiments. The time to achieve 50% removal of both Mg and Mn was found to be approximately halved by increasing the pre-oxidation temperature from 850 to 100000C.

Table 3. XRF analyses on Capell and CR1 RI's, DP's and 15% HCII SR's.

Run. no. Pre-oxid. TiO 2 Fe 2

O

3 Mn 3

O

4 Si02 A1 2 0 3 MgO Cr 2 0 3 (00) CaDell FB319 RI 850 62.3 48.1 1.91 0.31 0.62 0.24 0.09 FB320 RI 900 62.1 48.0 1.90 0.30 0.62 0.25 0.09 FB321 RI 950 62.3 48.1 1.92 0.33 0.61 0.24 0.12 FB322 RI 1000 62.3 48.1 1.91 0.32 0.62 0.24 0.14 FB319 DIP 850 93.8 2.08 2.68 0.37 0.84 0.33 0.10 FB320 DIP 900 93.9 2.14 2.66 0.36 0.86 0.32 0.09 FB321 DIP 950 94.7 1.06 2.54 0.34 0.89 0.31 0.15 FB322 DIP 1000 94.7 0.81 2.57 0.39 0.87 0.33 0.19 FB319 SR 850 96.7 1.11 0.35 0.34 0.79 0.10 0.09 FB320 SR 900 97.0 0.98 0.30 0.33 0.84 0.09 0.10 FB321 SR 950 97.5 0.53 0.21 0.32 0.84 0.09 0.13 FB322 SR 1000 98.0 0.39 0.18 0.35 0.82 0.08 0.13

CRIL

FB323 RI 850 57.6 53.4 1.88 0.38 0.52 1.05 0.35 FB316 RI 900 57.7 53.5 1.87 0.36 0.56 1.04 0.38 FB317 RI 950 57.5 53.4 1.90 0.45 0.60 1.05 0.40 FB318 RI 1000 57.4 53.4 1.89 0.43 0.55 1.04 0.45 FB323 DIP 850 91.3 2.51 2.82 0.47 0.72 1.52 0.54 FB31 6 DIP 900 90.7 2.88 2.78 0.48 0.77 1.49 0.62 FB317 DIP 950 90.8 2.47 2.87 0.54 0.83 1.59 0.66 FB31 8 DIP 1000 91.9 1.50 2.67 0.58 0.79 1.49 0.62 FB323 SR 850 95.6 1.35 0.29 0.50 0.81 0.46 0.63 FB316 SR 900 95.7 1.27 0.26 0.54 0.88 0.41 0.68 FB317 SR 950 95.9 0.95 0.17 0.59 0.90 0.38 0.65 FB318 SR 1000 96.3 0.71 0.11 0.59 0.86 0.29 0.64 C6/98147003.1 12 Table 4. Element extraction in 15% HCI leaching or CRL and Capel DP's Element removal as of total Pre-oxid Mg Mn Fe Ti Temp (oC) Capel 850 29.8 54.2 27.1 1.8 850 44.2 67.9 36.2 2.3 850 57.7 81.9 44.0 2.7 120 850 71.2 89.8 49.8 2.9 900 33.9 54.0 33.2 1.9 30 900 50.5 74.0 49.0 2.6 900 64.2 85.0 52.3 120 900 71.2 90.8 56.8 3.1 15 950 40.6 66.9 31.9 2.1 30 950 54.0 80.4 40.1 60 950 65.2 87.0 46.4 120 950 74.9 92.4 51.4 2.2 15 1000 43.0 74.0 32.5 2.4 30 1000 56.0 87.4 40.2 2.7 1000 67.5 91.8 45.1 2.2 120 1000 74.0 95.2 49.3 1.8

CRL

850 19.8 38.9 23.1 1.8 30 850 34.6 59.5 33.0 2.9 60 850 52.4 75.8 42.3 3.6 120 850 74.4 85.3 50.2 3.6 900 25.0 43.6 25.9 900 39.6 62.8 37.6 900 61.6 80.2 52.1 120 900 76.7 89.0 60.4 3.2 950 25.7 55.0 27.7 950 38.5 70.3 40.8 3.2 950 59.9 83.0 53.3 2.6 120 950 78.6 91.1 62.6 2.3 1000 31.5 67.5 30.0 1000 48.7 82.0 39.6 3.1 1000 69.6 90.5 47.6 120 1000 86.0 95.5 53.6 13 Example 4 The experiments of Examples 1, 2 and 3 were repeated on samples from the Monto Goondicum (abbreviated as Monto ilmenite deposit in Queensland.

The analysis of the samples is set out in Table The Monto Goondicum concentrate has a particularly high magnesia content. The Mg substitutes for Fe in the ilmenite lattice. Relatively pure MgTiO 3 with the ilmenite-type structure occurs naturally as the mineral geikelite. An additional set of tests was therefore added for a pre-oxidation temperature of 10500°C, to assist in the breakdown of the Mg-rich ilmenite.

The metallic iron analyses (Satmagan) on grab samples for the reduction step are given in Table 6.

A comparison with Table 2 shows that the rates of metallisation of the Monto G.

concentrate are twice as high as those obtained for the Capel and CRL samples.

This is due to the effect of the higher hydrogen flow (2 m/s c.f. 0.6 m/s) in a flow 15 regime where gas starvation is important.

The significant effect of pre-oxidation temperature on metallisation kinetics is seen to be at temperatures between 900 and 9500°C. XRD patterns on the oxidation products showed that the greatest change in the oxidation phase assemblage occurs in this temperature interval. AT 900°C, ilmeno-hematite plus rutile were the major oxidised phases while at 9500°C, M 3 0 5 (pseudobrookite) became the major oxidation product. An XRD pattern of the oxidised product obtained at 1050°C showed almost pure pseudobrookite, with only small amounts of haematite, rutile and anatase.

The demetallised products from the fluid bed reductions were leached in refluxing 15% HCI, with withdrawal of solution samples for analysis as described in Example 3. Total leaching times of 4 h were used, compared to only 2 h for the CRL and Capel samples. XRF analyses on the Rl's, DP's and SR products, as well as on the original concentrate, are reported in Table 7. The DP analyses in C6/98147003.1 14 Table 7 show a high level of residual iron expressed as Fe20 3 in the sample from the 900°C pre-oxidation, with a progressive decrease in residual iron as the pre-oxidation temperature is increased. This is consistent with lower degrees of metallisation in RI's obtained from the lower-temperature pre-oxidised samples, and is confirmed by the metallic iron contents of the Rl's, shown in Table 7. The lower metallisation levels for the 900 and 950°C pre-oxidation samples in Table 7 compared to those given for the final grab samples in Table 5 are due to the slower reduction kinetics in the runs where grab samples were not removed and thus the average hydrogen-to-solids ratio remained lower.

10 The XRF analyses of the SR products in Table 7 show that the process of the invention is effective for treating a high-magnesia concentrate. Grades of TiO 2 were obtained and the grade improved significantly (95% to 97% TiO 2 with increase in the pre-oxidation temperature used. The leaching curves had not plateaued after 4 h leaching with refluxing 15% HCI and thus higher grades could 15 be obtained with further leaching.

Table 5. Analysis of Monto Goondicum concentrate 'o TiO 2 Fe20 3 Mn 3

O

4 SiO 2 A1 2 0 3 MgO CaO 50.4 49.0 0.92 0.46 0.27 2.97 0.18 In table 5, total iron has been reported as Fe20 3 and hence the total is greater than 100%.

C6/98147003.1 Table 6. Metallic iron in grab samples of Monto G. fluid bed Rl's corresponding to different pre-oxidation conditions FB339 FB340 FB341 FB342 Pre-oxid. Temp (OC) 900 950 1000 1050 Reduction Time (min) 14.7 18.9 17.3 17.9 27.2 32.1 30.5 31.8 33.6 36.0 36.1 37.9 37.6 37.5 38.4 39.3 25 39.1 38.9 39.4 39.6 30 40.1 39.9 39.8 40.0 Table 7. Analyses of Monto G. RI's, DP's and SR's from 15% HCI reflux 5 leaching Run Pre-ox TiO 2 Fern Mn 3 0 4 SiO 2 A1 2 0 3 MgO CaO No. (oC) 338 RI 900 55.3 37.2* 1.03 0.43 0.31 3.25 0.23 331 RI 950 55.6 39.4* 1.02 0.47 0.28 3.28 0.25 332 RI 1000 55.4 40.4* 1.02 0.50 0.31 3.26 0.25 333 RI 1050 55.4 40.6* 1.03 0.44 0.32 3.27 0.25 338 DP 900 85.6 7.74 1.45 0.67 0.43 4.65 0.11 331 DP 950 88.8 4.55 1.43 0.73 0.46 4.74 0.13 332 DP 1000 90.4 2.41 1.39 0.72 0.47 4.69 0.11 333 DP 1050 91.3 1.52 1.36 0.72 0.46 4.60 0.10 338 SR 900 95.1 1.59 0.21 0.62 0.36 0.76 0.05 331 SR 950 95.30 1.97 0.26 0.62 0.37 0.90 0.06 332 SR 1000 96.6 1.00 0.16 0.65 0.36 0.58 0.05 333 SR 1050 97.00 0.62 0.10 0.70 0.41 0.37 0.05 C6/98147003.1 16 Example A sample of 100 grams of Boodanoo (WA) ilmenite was oxidised in a laboratory reactor using a 12.5% 02 and 87.5% N 2 gas mixture as fluidising/oxidising gas, at a total flowrate of 6 I/min at STP. The oxidation conditions are detailed in Table 8 below. At the end of oxidation, the gas mixture was replaced with N 2 to prevent further reaction. 5 grams of limestone (CaCO 3 was then added. The N 2 was then replaced by a 2:1 mixture of H 2

:N

2 at a total flowrate of 12 I/min at STP. Reduction was carried out at 850°C for the times shown in Table 8 below.

10 The resultant product was then leached in dilute 5% sulphuric acid to remove the metallic iron that was formed on reduction (this was used to simulate the effect of an aeration leach). The demetallised product was then leached using HCI for 2 hours under reflux conditions to produce the SR product. Table 9 below contains the analysis for the four samples after the demetallisation and final 15 leach stages.

DP and SR indicate demetallised and post leaching synthetic rutile product respectively.

The higher Ti0 2 grade is apparent in comparing the two pairs of results for 900°C, then 1000°C pre-oxidation.

Table 8 Oxidation and Reduction of Boodanoo Ilmenite Run Number Oxidation Temp Oxidation Time Reduction Time (min) (min) FB47 900 30 FB48 1000 30 FB49 900 120 1000 120 C6/98147003.1 17 Table 9 XRF Results for Leaching of Reduction Products Test Sample TiO 2 Fe 2 03 Mn 3 04 SiO 2 Al 2 03 Mg0 CaO FB47 DP 91.09 4.83 2.48 1.02 0.71 0.15 0.07 SR 95.32 1.64 0.41 1.06 0.71 0.05 0.03 FB48 DP 93.24 1.99 2.27 1.09 0.78 0.14 0.06 SR 96.95 0.94 0.15 1.09 0.73 0.03 0.03 FB49 IDP 92.94 1.44 2.34 1.00 0.69 0.14 0.07 SR 95.64 0.78 0.27 1.03 0.69 0.04 0.03 FB50 DP 94.16 0.43 2.28 0.93 0.67 0.15 0.06 SR 96.05 10.49 10.10 0.92 10.64 10.02 0.037 4

S*

a.

4' 4 54 a 44 a Se