Production of synthetic rutile from tin ore beneficiation byproduct through preoxidation and reductive leaching in hydrochloric acid

This paper examines the effectiveness of the method for producing synthetic rutile from ilmenite through pre-oxidation and reductive leaching of pre-oxidized ilmenite in hydrochloric acid. Thermodynamic simulation of the pre-oxidation of ilmenite concentrate was performed to evaluate the phases formed during the process as a function of temperature. The pre-oxidation experiments were performed at different temperatures between 700 and 1000 °C in a muffle furnace for 6 h. The optimum temperature of pre-oxidation was revealed to be at 700 °C where ilmenite transformed into hematite and rutile, which is in accordance with the result of the thermodynamic simulation. Series of the leaching experiments were carried out under variations of HCl concentration (5–8 M), leaching temperature (70–100 °C), solid/liquid ratio (1/5–1/20 g/mL), ilmenite ore particle size distribution, and duration of leaching (6–12 h). Taguchi method utilizing L16 orthogonal array was adopted in the leaching step to design and reduce the required number of experiments. Analysis of variance (ANOVA) indicated that the temperature and solid/liquid (S/L) ratio were the most influential leaching parameters for the dissolution of iron and titanium. The optimum conditions for maximising the dissolution of iron, while minimizing the dissolution of titanium were at a temperature of 80 °C, HCl of 6 M, S/L ratio of 1/20 g/mL, ore particle size distribution of 44–77 µm (-200 + 325 mesh), and leaching duration of 6 h. The leaching experiment conducted under these conditions resulted in iron extraction of 98.07% with co-extraction of titanium of 11.35%. The leach-residue contains 92.6% rutile, 2.9% hematite, and 2.5% cassiterite which can be classified as synthetic grade rutile.

The mineral cassiterite (SnO 2 ), the primary source of tin metal, is often associated with ilmenite (FeTiO 3 ) in nature. Hence, the processing of cassiterite often produces a by-product that has high contents of ilmenite, which is the primary source of titanium dioxide (TiO 2 ) and titanium metal. Zglinicki et al. 1 reported that the TiO 2 concentration in the by-product of tin ore beneficiation in Bangka Island, Indonesia, can be as high as 51%. This suggests its potential as a resource for TiO 2 production since the concentration is within the range of a commercial ilmenite concentrate, which is 40-65% TiO 2 .
The pigment industry consumes more than 90% of the global TiO 2 production due to its high refractive index and inertness. The commercial production of TiO 2 pigment has been carried out by two distinct methods, namely the sulphate and chloride process. The latter is considered cheaper since the hydrochloric acid is recycled and is considered cleaner since it produces lower volumes of waste and less invasive residues than the former 2, 3 . Unlike the former, however, the latter requires higher feed grades of about 90% TiO 2 4, 5 . Therefore, an ilmenite concentrate needs to be upgraded before it can be fed into the chloride process. It is notable that the growing titanium metal industry also relies on such a high-grade feed material.
There are several processes to upgrade ilmenite for producing the so-called synthetic rutile, the mineral obtained through chemical alteration of ilmenite, as a feed for the chloride process. The commercial processes to produce synthetic rutile include the Becher process 6 , Benelite process 7 , Murso process 8 , and Austpac process 9 , which all involve roasting and leaching. In the Becher process, the ilmenite is oxidised with air to convert the Leaching experiment. The pre-oxidized ilmenite at the optimum condition was then treated in reductive leaching experiments using hydrochloric acid and iron powder as reducing agent to selectively dissolve iron while minimizing co-dissolution of titanium and concentrating TiO 2 in the leach residue. The experimental design of the leaching of pre-oxidized ilmenite was established by Taguchi Method. Series of the leaching experiments were carried out under variations of five variables, namely hydrochloric acid (HCl) concentration, tem-    Table 3. Based on the variables and their level of variations, the experimental program was determined, and the detail is presented in Table 4. The experiments were conducted in duplicate for each experimental condition to obtain a high confidence level of the results. The leaching experimental set-up is shown schematically in Fig. 2. The leaching experiment was carried out in a 500-mL three-necks flask with a condenser attached to one of the necks to condense the evaporated water, thus a constant volume of the solution during the leaching can be maintained. Agitation of the slurry was done by using a magnetic stirrer which was integrated with a hot plate. The temperature of the solution during the leaching was measured by a thermocouple immersed in the solution. The thermocouple was connected to the hotplate for adjustment of the heating level to keep the leaching temperature at ± 0.5 °C around the targeted temperature. The volume of the initial HCl solution was 200 mL. The stirring speeds during leaching were kept constant at 700 rpm. For each experiment, iron powder as a reducing agent (i.e., reduces Fe 3+ to Fe 2+ ) was added to the solution 20 min after the leaching started with a constant iron/ilmenite mass ratio of 1/5. Filtration of the slurry by filter paper was carried out upon completion of the leaching experiments. The filtrates were analysed by atomic absorption spectroscopy (AAS, Shimadzu AA-6300, Japan) for measurements of dissolved iron and titanium concentrations, while the solid residues were washed, dried, and digested using concentrated hydrofluoric acid for determining the mass of iron and titanium remained in the leach residue. The solution obtained by the digestion of the leach residue was also analysed by AAS for measurements of iron and titanium concentrations in the leach residue for each experiment. The percentages of extracted iron and titanium were calculated based on the mass ratio of the dissolved metals in pregnant leach solution (PLS) and the sum of the dissolved metals in the PLS and those remaining in the solid residue as formulated by the following equation: in which m 1 is the mass of dissolved metals in PLS (i.e., Ti for the calculation of Ti extraction and Fe for Fe extraction) and m 2 is the mass of metals remaining in the solid residue. As previously mentioned, the objective of the

Result and discussion
Thermodynamic simulation of ilmenite pre-oxidation. The thermodynamic aspect of the pre-oxidation of ilmenite was evaluated by performing a calculation of the equilibrium reaction between ilmenite concentrate and air atmosphere (pO 2 = 0.21 atm) at temperature intervals between 700 and 1000 °C using the FactSage 7.2 thermochemical package. The thermodynamic simulation was performed based on the ilmenite concentrate composition provided in Table 1  Determination of optimum temperature of ilmenite pre-oxidation. The pre-oxidation experiments of the ilmenite concentrate were performed at 700 °C, 800 °C, 900 °C, and 1000 °C in a muffle furnace for 6 h. The objective of the pre-oxidation process is to convert the ilmenite into rutile (TiO 2 ) and hematite (Fe 2 O 3 ), while minimizing the formation of pseudobrookite (Fe 2 TiO 5 ). At the leaching stage, the iron in the hematite is expected to be dissolved as much as possible, with minimum co-dissolution of titanium from the rutile. Through this mechanism, synthetic rutile can be concentrated in the leach residue.
XRD spectra of the products of pre-oxidation at various temperatures are depicted in Fig. 4. According to the XRD analysis of the pre-oxidation product, at temperatures of 700 °C dan 800 °C, ilmenites (FeTiO 3 ) were mostly converted to rutile (TiO 2 ) and hematite (Fe 2 O 3 ), while at 900 °C dan 1000 °C a new phase called pseudobrookite (Fe 2 TiO 5 ) was also formed along with rutile. Cassiterite (SnO 2 ) was detected in the oxidised products at the entire temperature variations. However, the peak intensity related to cassiterite is diminished in the products formed at 900 and 1000 °C. Meanwhile, the peak intensities for quartz (SiO 2 ) were diminished completely      www.nature.com/scientificreports/ Vásquez and Molina 12 , who also studied pre-oxidation of ilmenite, reported that the hematite formed during oxidation is concentrated on the outer layer of the oxidized particle, while rutile is enriched in the inner layer. This was suggested to be associated with the diffusion of iron cation to the region which has high oxygen potential leading to the formation of hematite at the surface and purer rutile at the core of the particle 19 . The enrichment of hematite at the surface of the particle would facilitate faster iron dissolution under reductive conditions and on the other hand, titanium that is contained in rutile at the inner parts of the particle will be less amenable to acid dissolution and will be concentrated in the leach residue. Hence, the phenomenon that takes place during the preoxidation allows a better leaching selectivity of iron toward titanium at subsequent reductive acid leaching of the pre-oxidized ilmenite. The oxidation of ilmenite to hematite and rutile occurs through the following reaction 20 : Meanwhile, the formation of pseudobrookite (Fe 2 TiO 5 ) takes place through the following reaction: The formation of the pseudobrookite phase in the pre-oxidation products at 900 and 1000 °C are in agreement with the findings of Vásquez and Molina 12 and Zhang and Ostrovski 20 . It was reported that the presence of pseudobrookite in the pre-oxidation product of ilmenite slows down the dissolution of iron and reduces total dissolved iron in the subsequent leaching stage 12,21 . Noticeably, the peaks of XRD spectra associated with hematite at 700 °C are stronger than those formed at 800 °C, which is in agreement with the findings of Vásquez and Molina 12 .
Based on the thermodynamic simulation and XRD analysis results of the pre-oxidation of ilmenite at various temperatures, the temperature of 700 °C was considered as the optimum temperature since it can produce the best conversion of ilmenite to hematite and rutile without the formation of pseudobrookite.
Effect of leaching variables on extracted iron and titanium analysed by ANOVA. Results of leaching experiments are shown in Table 5. The leaching experiment results showed that extracted iron was in the range of 94.24-97.54%, while co-dissolved titanium varied from 3.31 to 23.15%. Analysis of variance (ANOVA) was used to determine the level of influence of each leaching variable on the extraction levels of iron and titanium. This analysis involved the calculation of the parameters of the sum of square (SS), mean square (MS), and F-distribution. The value of standard F-distribution for iron and titanium dissolutions was then obtained from the F-distribution table using α = 0.05, df 1 = 3, and df 2 = 16, which is equal to 3.24. By comparing the value of the F-distribution from each variable to the standard F-distribution, the significance of the influence of each variable on the extractions of Fe and Ti can be assessed. The results of ANOVA which shows the influence of each variable represented by the contribution percentage of the variable on the extracted Fe and Ti are presented in Tables 6 and 7, respectively. By using a standard F-distribution value of 3.24, it can be concluded that HCl concentration, temperature, S/L ratio, and ilmenite particle size distribution have a significant influence on the extraction of iron. The temperature was found to have the highest influence on the iron extraction (42.98%), followed by S/L ratio (30.37%), acid concentration (15.81%) and ilmenite particle size distribution (4.21%). Generally, the dissolution rate of metals in an aqueous solution is enhanced when the leaching temperature is higher. It was also indicated that the leaching of ilmenite in hydrochloric acid is mass-transfer controlled,  www.nature.com/scientificreports/ therefore, the temperature is one of the most decisive parameters that affect the diffusion rate of reacting species and determines the extracted iron and titanium at a certain duration of the leaching 21 .
Similarly, all variables also exhibited a significant influence on the dissolution of titanium. Different to the case for iron dissolution, the S/L ratio was found to have the highest influence on the titanium extraction (59.84%), followed by acid concentration (25.51%), temperature (5.04%), ilmenite particle size distribution (4.77%) and duration of the leaching (2.76%). As has been previously mentioned, during the pre-oxidation stage, Fe 2 O 3 is concentrated on the outer layer of the oxidized particle, while TiO 2 is enriched in the inner layer of the oxidized particle. Titanium present in rutile at the inner parts of the particle is less amenable to acid dissolution in comparison to iron which is present at the outer layer. At a smaller solid/liquid ratio, more HCl is available in the solution which in turn will have a higher probability to diffuses to the inner layer and reacting with TiO 2 . In this regard, the variable S/L ratio eventually gives the highest influence on the dissolution of titanium-based on the experimental design by Taguchi L16: 4 5 and data analysis by ANOVA.

Signal to noise ratio analysis and the optimum condition evaluation for the iron and titanium extractions.
A statistical correlation between the experimental factors (variables) and the experimental results (outputs) was further evaluated. Signal-to-noise (S/N) ratios, which is a measure of the effect of the experimental variations on the outputs of the experiment by minimizing the effect of noise factors, for each of the control parameters were calculated. The degrees of influence of variables, represented by the parameter of signal to noise (S/N) ratios, to the experimental outputs, i.e., the dissolutions of iron and titanium, are given in Figs. 5 and 6, respectively. Since the target of the leaching process is to maximize the dissolution of iron and minimize the co-dissolution of titanium, the desired condition for the iron dissolution is indicated by the highest S/N value in Fig. 5, while the desired condition for titanium dissolution is indicated by the lowest S/N value in Fig. 6. Based on the profiles of S/N ratios for Fe and Ti extractions versus leaching variables, the condition with the highest S/N ratio for iron extraction and lowest S/N ratio for titanium co-extraction is considered at HCl concentration of 8 M, the temperature of 90 °C, S/L ratio 1/20 g/ml, the particle size distribution of 44-74 µm, and leaching duration of 10 h.
The optimum condition was further evaluated through simple mathematical models that describe the extractions of Fe and Ti as a function of variables of the leaching process based on the experimental results in Table 5. The models were established by using Minitab software. The models suggested that the combined optimum condition for high iron extraction and low titanium co-extraction is at an HCl concentration of 6 M, a temperature of 80 °C, an S/L ratio of 1/20 g/ml, and the particle size distribution of 44-74 µm, and leaching duration of 6 h. HCl concentration of 6 M was considered as an optimum level to balance the two objectives (i.e., maximizing Fe extraction, while minimizing co-extracted Ti). The increase of HCl concentration above 6 M did not significantly promote the extraction percentage of iron which is indicated by relatively flat S/N by the increase in HCl concentration from 6 to 8 M. Ramadan et al. 22 stated that the dissolution of iron in ilmenite linearly increases to a concentration of 20% HCl (around 6.5 M) and the further increase of HCl concentration beyond this level did not significantly affect the extracted iron percentage. On the other hand, higher HCl concentration also Leaching experiment at optimum condition. The extracted Fe and Ti at the optimum condition from the mathematical models were predicted to be 98.22% and 17.03%, respectively. To verify the predicted Fe and Ti extractions, a leaching experiment was carried out at the optimum condition, namely at 6 M HCl concentration, the temperature of 80 °C, S/L 1/20 (g/mL), ilmenite size fraction of 44-74 µm, and leaching duration of 6 h with iron powder ratio to ilmenite of 1/5 added in the solution and stirring speed 700 rpm. During the leaching experiment, slurry samples were collected at a pre-determined volume and time interval (i.e., 30, 60, 120, 240 and 360 min). The slurry samples were filtered through Whatman filter paper, and the filtrates were subsequently analysed by AAS to determine the concentrations of dissolved Fe and Ti. The result of the leaching experiment at optimum condition is shown in Fig. 7. Based on the experimental results presented in Fig. 7, the highest iron extraction level of 98.07% was obtained after 360 min of leaching. This value was very close to that calculated by the mathematical model for the iron extraction (98.22%). Meanwhile, the co-extraction of titanium after 360 min was 11.35% which is 5.68% lower than that calculated by the mathematical model for the titanium extraction (i.e., 17.03%). This difference is suggested to be associated with the different extents of titanium hydrolysis which cannot be controlled during the experiments.    The dominant reaction taking place during the leaching of the pre-oxidised ilmenite is the dissolution of iron. The dissolution kinetics of iron from the pre-oxidised ilmenite at the optimum condition (Fig. 7) has been analysed to determine its controlling step. The pre-oxidised particles are assumed to have a relatively uniform particle size, dense structure, and spherical geometry. In the case where hematite occupied the outer layer of the pre-oxidised particle, there was no barrier between the unreacted hematite and the leaching medium. In addition, the external radius of the pre-oxidised particle gradually decreased during the dissolution of iron. This situation can be evaluated using the shrinking core models for the shrinking sphere 23 . Two rate-controlling steps were considered, i.e., diffusion of reactant from the bulk fluid to the solid surface through the film layer and chemical reaction on the solid surface. The kinetic analysis based on the selected models indicated that the dissolution kinetic of iron in this study was controlled by the chemical reaction on the solid surface.
Based on the data of iron and titanium extractions, iron extraction relative to titanium was determined by comparing the extraction percentage of iron with the sum of iron extraction percentage and titanium extraction percentage. The Fe extraction relative to Ti was 0.90 which indicates that co-extracted Ti was relatively low and most of the titanium was concentrated in the leaching residue.
Mechanism of the leaching of pre-oxidized ilmenite in hydrochloric acid in the presence of iron powder. Prior to the pre-oxidation process of ilmenite ore, Fe exists as Fe(II) or FeO as well as FeTiO 3 and distributed evenly throughout the ilmenite ore particle. The dissolution of iron and titanium during direct leaching of ilmenite without pre-oxidation treatment takes place simultaneously through the following reaction 10 : Basically, both of iron and titanium are dissolved and consume the leaching agent, although part of dissolved titanium is slowly hydrolysed and releases hydrochloric acid according to the following reaction: As previously mentioned, during the pre-oxidation stage of ilmenite at a temperature range of 700-1000 °C, the Fe cations diffuse to the outer layer of ilmenite particle and form hematite which will be more prone to the leaching by HCl. The dissolution rate of Fe during the leaching of pre-oxidised ilmenite is fast in the first few hours of the leaching which slows down by the longer time of the leaching 21 .
Since rutile occupies the inner part of the pre-oxidized ilmenite, the leaching of Ti is minimized and controlled by the diffusion of reactant and reaction product through the solid particle 21 . In the absence of a reducing agent, the leaching of iron from hematite in pre-oxidized ilmenite by hydrochloric acid occurs through the following reaction 10 : In the presence of Fe powder as a reducing agent, iron(II) chloride and H 2 gas will be formed as the product of the reaction between Fe powder and HCl (Reaction 7). Iron(III) ions released from the leaching of hematite (Reaction 6) will be readily reduced by the hydrogen gas (H 2 ) and Fe powder simultaneously (Reactions 8 and  Analysis of the leach residues. The residue obtained from the leaching experiment of pre-oxidized ilmenite at optimum condition (i.e., HCl concentration of 6 M, temperature 80 °C, solid/liquid ratio 1/20, size fraction 44-74 µm, and 6 h of leaching duration) was analysed by scanning electron microscope (SEM, JEOL JSM-6510A, Japan), XRD and XRF. The SEM images of the surface morphologies of ilmenite and leaching residue particles are shown in Fig. 8. The ilmenite sample consisted of particles with a wide range of sizes from a few microns up to 149 microns, similarly, the leaching residue had a broad particle size distribution. Based on the SEM examination, the larger ilmenite particles had relatively smooth surface morphology while the larger leaching residue particles had rougher surface morphology which is possibly due to reaction with the HCl. The XRD analysis result of the leaching residue is shown in Fig. 9. The XRD analysis result shows that the leaching residue consists of mostly rutile and cassiterite and indicates that most of the hematite was effectively dissolved during the leaching process. The sharp peak of rutile in the XRD pattern of leaching residue is related to the rutile's low solubility in the leaching stage. The XRD spectrum shows only 1 peak of hematite which indicates a minor amount of the hematite in the leaching residue. The chemical composition of the leaching residue was determined by XRF analysis, and the result is shown in Table 8. The leaching residue contains 92.6% TiO 2 and 2.9% Fe 2 O 3 . Other metal compounds which have remarkable percentage in the leaching residue are SnO 2 (2.5%), V 2 O 5 (0.9%), MnO 2 (0.3%) and SiO 2 (0.3%). The SnO 2 is present in the leaching residue since the ilmenite was a by-product of the tin ore beneficiation process. The residue obtained from the leaching of pre-oxidised ilmenite ore under the optimum condition can be considered as synthetic grade rutile which can be utilized for the manufacturing of pure TiO 2 pigment. The concentration of TiO 2 of 92.6% is remarkably higher than the rutile commonly used for pigment manufacturing (7) Fe (s) + 2HCl (aq) = FeCl 2(aq) + H 2(g) (8) 2FeCl 3(aq) + H 2(g) = 2FeCl 2(aq) + 2HCl (aq) (9) 2FeCl 3(aq) + Fe (s) = 3FeCl 2(aq)

Conclusion
In this study, the production of synthetic rutile from ilmenite as a by-product from tin ore beneficiation through pre-oxidation and reductive leaching of the pre-oxidized ilmenite in hydrochloric acid is discussed. The objective of the pre-oxidation of ilmenite concentrate is to convert the ilmenite into rutile (TiO 2 ) and hematite (Fe 2 O 3 ), while minimizing the formation of pseudobrookite (Fe 2 TiO 5 ). At the leaching stage, the iron in the hematite is expected to be dissolved as much as possible, with minimum co-dissolution of titanium from the rutile. Using this method, synthetic rutile can be concentrated in the leach residue. Thermodynamic evaluation and experimental investigation indicated that the optimum pre-oxidation temperature was 700 °C, which converted ilmenite to hematite and rutile. Temperature and solid/liquid ratio (S/L) were found to be the most influential leaching parameters for the dissolutions of iron and titanium. The optimum conditions for maximizing iron solubility, while minimizing titanium dissolution were obtained at 80 °C, HCl = 6 M, S/L ratio 1/20 g/mL, ore particle size distribution 44-74 µm, and leaching duration of 6 h. Leaching experiments carried out under these conditions resulted in iron extraction of 98.07% and titanium co-extraction of 11.35%. The leaching residue contains 92.6% rutile, 2.9% hematite, and 2.5% cassiterite which can be classified as synthetic grade rutile. The investigation results show the utilization prospect of ilmenite as a by-product from tin ore beneficiation for producing synthetic rutile.