Leaching Behavior of Titanium from Na₂TiO₃

Abstract

Leaching and hydrolysis of Ti from washed Na₂TiO₃ were studied. Synthetic rutile was mixed with NaOH at a weight ratio of 1:1.13 and salt-roasted at 773 K for 4 hours to synthesize Na₂TiO₃. This was washed with distilled water to obtain a leached sample. Acid leaching was carried out at 303, 323, 333, 343, 353, 363, and 373 K for 4 hours using 5 M hydrochloric acid. From 303 to 343 K, the leaching efficiency of Ti increased with an increase in time, with 96% Ti leached at 343 K. From 353 to 373 K, Ti was preferentially leached with time and precipitated by hydrolysis, and the leaching efficiency decreased. The leaching and hydrolysis of Ti were investigated using reaction equations and thermodynamic data corresponding to each leaching temperature. The kinetics of the leaching reaction were studied using kinetic equation for the chemical-reaction-controlled step, applying a shrinking core model of the heterogeneous reaction. The activation energies, derived from the Arrhenius plot, was 38.9 kJ/mol for the chemical-reaction-controlled plot, indicating that the chemical reaction was the rate-determining step. The hydrolysis reaction was dominant at temperatures above 353 K, and the residue after the reaction consisted mostly of TiO₂.

Fig. 2 Leaching efficiency of Ti by temperature.

1. Introduction

Ti is a critical metal used in various applications, including aerospace engineering, offshore plant, the weapons industry, the chemical industry, artificial joints, and catalysts. Although there exist more than 100 different minerals containing more than 1% TiO₂, ilmenite and rutile are the minerals of choice for industrial use. Rutile contains more than 95% TiO₂, but it has limited reserves and is poorly distributed. Therefore, more than 90% of the Ti metallurgy industry uses ilmenite. The well-known commercial processes for recovering TiO₂ include the UGS-Slag process, Becher process, sulfate process, and chloride process.¹⁾ Ti minerals are typical ores that are difficult to use directly in the hydrometallurgical process. Therefore, Ti solubility should be increased by pretreatment, for example, by a redox reaction in a rotary kiln or by a smelting process for separation of Ti from Fe. In particular, the direct dissolution of Ti is known to be easy if TiO₂ is converted to Na₂TiO₃ through a salt-roasting process.

F. Meng obtained the optimum conditions for alfa Na₂TiO₃ by rutile roasting with NaOH. α-Na₂TiO₃ is a metastable phase that is less stable than γ-Na₂TiO₃ and is more suitable for leaching TiO₂.^2,3) Y. Li explored the mechanism of Na₂TiO₃ formation and the influence of impurities, thermodynamically and crystallographically, in the Ti-bearing EAF slag during the NaOH treatment process.⁴⁾ Q. Zhang studied the thermodynamics and kinetics of the transformation mechanism of TiO₂ by carrying out NaOH roasting on spent selective catalytic reduction catalysts.^5,6) T. Xue investigated the kinetic mechanisms of EAF Ti slag by means of NaOH roasting.⁷⁾ Y. Wang prepared Na₂TiO₃ by NaOH roasting and obtained H₂TiO₃ by acid leaching and hydrolysis.⁸⁾

It is also possible to directly dissolve Ti minerals in acids in order to remove impurities such as Fe, Mg, and Ca to eventually increase the purity of TiO₂. For instance, F. Xu dissolved Mg, Al, and Ca from ilmenite in 20% hydrochloric acid, and the residues containing Ti and Si were added to NaOH solution at Na⁺/Ti⁴⁺ and O²⁻/Ti⁴⁺ molar ratios of 8/1 and 6/1, respectively, at leaching temperature. High-purity TiO₂ was obtained at 303 K within 30 minutes.⁹⁾ L. Zhao obtained 88.5% leaching efficiency under conditions of 80% sulfuric acid and 423 K by using Ti slag obtained from hydrochloric acid leaching of vanadium-bearing titanomagnetite and hydrolyzed the titanium sulfate solution to obtain 98.41% TiO₂.¹⁰⁾ N. Elhazek investigated the effects of variables such as hydrochloric acid concentration, liquid ratio, and temperature on ilmenite leaching and obtained 98% leaching under the optimal conditions of 12 M hydrochloric acid, pulp density of 1/20, temperature of 353 K, and reaction time of 2 hours and 30 minutes.¹¹⁾ N. Mostafa leached ilmenite with 20% hydrochloric acid at 343 K for 3 hours and obtained 94.6% Ti leaching efficiency. Fe was removed by solvent extraction, and TiO₂ was obtained by hydrolysis at 373 K.¹²⁾ G. K. Das achieved 98% Ti leaching from ilmenite under the conditions of 5.0–7.5 M hydrochloric acid, 3.3% of pulp density, 500 g/L Cl⁻, temperature of 343–353 K, reaction time of 4–6 hours. During this time, however, Fe almost completely dissolved.¹³⁾ C. Li reacted ilmenite in 20% hydrochloric acid at 373 K with a pulp density of 1 g ilmenite/5.5 g hydrochloric acid for 6 hours. Under these conditions, all of the Fe was dissolved and Ti remained in the residue. The leaching rate gradually decreased owing to hydrolysis with time. The final residue contained 92% TiO₂ and 2.1% Fe₂O₃.¹⁴⁾

F. Meng transformed tionite produced in the Ti metallurgy industry into Na₂TiO₃ by NaOH roasting and then leached Ti by water washing and sulfuric acid leaching. Ti was found to be 96.7% leached after 1 hour at 328 K in 40% sulfuric acid with a high liquid ratio of 2.6:1. At this time, the leaching rate of Ti was decreased by hydrolysis at 348 K.¹⁵⁾ S. Middlemas obtained Na₂TiO₃ by NaOH roasting of Ti slag and partially washed away the Al and Si by water washing. The residue thus obtained was leached with 5 M hydrochloric acid and a 10% pulp density at 323 K to obtain a 94.1 g/L Ti solution. The TiO₂ pigment was obtained by removing impurities such as Fe through solvent extraction and hydrolyzing the remaining raffinate at 373 K.¹⁶⁾ G. Moon studied for manufacturing high purity TiO₂ from a spent selective catalytic reduction catalyst through a sequence of process included soda-melting, acid leaching and desilication.¹⁷⁾

The kinetics of leaching of elements including Ti and the TiO₂ hydrolysis reaction in Ti solutions have been studied using Ti-containing raw materials such as ilmenite. For instance, E. Olanipekun investigated the effects of various experimental conditions on the leaching of Ti and Fe by leaching Nigeria ilmenite with hydrochloric acid. By applying the shrinking core model, the spreading of Ti and Fe through the porous product layer was shown to determine the rate step.¹⁸⁾ M. K. Sarker et al. investigated the rate of acid leaching of oxidized ilmenite and general ilmenite in hydrochloric acid. The leaching of the oxidized ilmenite best fitted a first-order reaction model for up to 50% leaching, followed by a spherical model. The leaching rate of general ilmenite was relatively slow and diffusion of the metal ions was the rate-determining step.¹⁹⁾ H. Tsughida et al. studied the kinetics of ilmenite leaching with respect to hydrochloric acid concentration and temperature. From 303 to 323 K, the surface chemical reaction, by the core model, was the rate-determining step. After 333 K, the rate-determining step changed as the diffusion rate of metal ions through the residual layer of ore slowed down.²⁰⁾

The process of dissolving Ti by leaching or direct leaching of Ti minerals may result in a different leaching rate of Ti depending on the leaching temperature and acid concentration, and the leaching rate of Ti may be decreased. This is because the dissolved Ti is hydrolyzed and precipitates as TiO₂. However, in most studies, the hydrolysis Ti, which occurs during Ti leaching, was not considered. Therefore, in this study, we aimed to elucidate the behavior of the two reactions based on leaching and hydrolysis of Ti generated during acid leaching of a salt-roasted rutile sample.

2. Experimental Procedure

2.1 Preparation of feedstock for leaching process

The leached samples were mixed with NaOH at a mass ratio of 1:1.13, followed by soda roasting for 4 hours at 773 K. Samples were then pulverized using a jaw crusher and cone crusher. The crushed sodium titrate was washed with distilled water to remove some of the Na. Table 1 shows the elemental contents of the raw samples, rutile, sodium titrate, and samples after washing with water. X-ray diffraction (XRD) peaks of these samples are shown in Fig. 1.

Table 1 Contents of rutile, soda-roasted rutile and water-washed soda-roasted rutile (mass%).

Fig. 1

XRD peaks of rutile, soda-roasted rutile, and water-washed soda-roasted rutile.

2.2 Leaching process

Leaching experiments were carried out using water-washed samples. Leaching was carried out using 5 M hydrochloric acid at 303, 323, 333, 343, 353, 363, and 373 K. Here, 500 mL of 5 M hydrochloric acid was poured into a five-necked round flask and heated in a heating mantle to maintain the leaching temperature. The pulp density was fixed at 10%, and 50 g of the sample was added when the leaching temperature was reached, with stirring at 250 rpm using a Teflon stirrer. The total leaching time was 4 hours, and sampling was performed at regular intervals. After the reaction was completed, the solid was separated through a vacuum filter. The sampled solutions and residues were analyzed using inductively coupled plasma mass spectrometry, and the leaching efficiency of Ti per hour was calculated by the following equation:   

\begin{equation} Y_{\text{Ti},\,t}(\%) = \frac{X_{\text{Ti},\,t}(\textit{mol}/L)}{X_{\text{Ti},\,i}(\textit{mol}/L) + X_{\text{Ti},\,j}(\textit{mol}/L)} \end{equation} (1)

where Y_{Ti, t} is the leaching efficiency of Ti at time t, X_{Ti, t} is the concentration of Ti at time t, X_{Ti, i} is the concentration of Ti in the final filtrate, and X_{Ti, j} represents the Ti concentration in the residue.

3. Results and Discussion

3.1 Leaching experiment with variable temperature

The leaching efficiency of Ti according to temperature is shown in Fig. 2. From 303 to 343 K, the leaching efficiency of Ti increased with increasing time for a total reaction time of 4 hours. This indicates that Ti was continuously leached during the 4 hours. The reaction can be written as follows:   

\begin{equation} \text{Na$_{2}$TiO$_{3}$} + \text{HCl} = \text{TiOCl$_{2}$} + \text{2NaCl} + \text{2H$_{2}$O} \end{equation} (2)

  

\begin{equation} \text{TiO$_{2}$} + \text{HCl} = \text{TiOCl$_{2}$} + \text{H$_{2}$O} \end{equation} (3)

Considering these two reactions, Na₂TiO₃ produced by soda roasting in the sample was not completely removed by water washing (see Table 1; the water-washed sample still contained Na). TiO₂ was determined to be formed by washing Na₂TiO₃, or by rutile that was not completely converted to Na₂TiO₃ during the soda roasting process. Gibbs free energy was calculated based on the thermodynamic data of each species in the equation. The thermodynamic data of each species and the Gibbs free energy of each reaction by temperature are shown in Tables 2 and 3, respectively. As shown in Table 3, the leaching of TiO₂ in hydrochloric acid is a thermodynamically unstable reaction. This is because the $\Delta G_{r}^{\circ} (T)$ value with regards of leaching reaction of the TiO₂ has a positive at all the leaching temperature. On the other hand, $\Delta G_{r}^{\circ} (T)$ value for the Na₂TiO₃ leaching reaction is negative in the leaching experiment temperature range. Therefore, leaching of Na₂TiO₃ was the main reaction in this study.

Fig. 2

Leaching efficiency of Ti by temperature.

Table 2 Thermodynamic data of each species at 298 K.

Table 3 Gibbs free energy of each reaction by experimental temperature.

3.2 Kinetics of leaching reaction of Na₂TiO₃

Based on the graph of leaching efficiency of Ti with time, the leaching of Ti was investigated with respect to its kinetics. A typical leaching reaction can be represented by a shrinking core model with a heterogeneous reaction between the solid sample and the leaching solution. In this case, a chemical-reaction-controlled step can be considered as rate-determining step.

The chemical-reaction controlled step for the rate-determining step can be defined by the following kinetic equation:   

\begin{equation} \mathrm{kt} = 1 - (1 - x)^{1/3}(\text{chemical-reaction-controlled step}) \end{equation} (4)

where k is the apparent reaction rate constant, t is the time, and x is the Ti leaching efficiency.

The leaching efficiencies determined in Fig. 2 were fed into this equation, and the kinetics were evaluated. The result is shown in Fig. 3.

Fig. 3

Plot of time vs. kt: chemical-reaction-controlled step.

The slope of the Fig. 3, i.e., the values of k and R-squared value, are shown in Table 4. The slope shown in Fig. 3 and Table 4 shows a good linear tendency in the chemical reaction controlled step at all temperatures.

Table 4 Values of k and R-squared of chemical reaction controlled step.

However, it was not clear from these results whether the chemical reaction for leaching of Ti was the rate-determining step or not. Therefore, we tried to evaluate the rate-determining step by calculating the apparent activation energy using an Arrhenius plot. Figure 4 shows an Arrhenius linear plot with ln(k) values and 1000/T. The R-squared values for chemical-reaction-controlled step was 0.9625. Based on these results, the activation energy was calculated using the following equation:   

\begin{equation} \mathrm{k} = \mathrm{A} \exp (-E_{a}/RT) \end{equation} (5)

where A is the frequency factor, E_a is the activation energy, R is the gas constant, T is the absolute temperature, and k is the apparent reaction rate constant. The calculated activation energy for the chemical-reaction-controlled plot was 38.9 kJ/mol. Within the range of 303 to 343 K, thus, the rate-determining step for leaching of Ti is the chemical reaction.

Fig. 4

Arrhenius plot of chemical-reaction-controlled step.

3.3 Hydrolysis reaction of TiOCl₂

As shown in Fig. 2, the leaching efficiency of Ti initially increased with time, then gradually decreased. This was because TiOCl₂ dissolved in hydrochloric acid is re-precipitated into the TiO₂·H₂O form by hydrolysis. The hydrolysis of TiO₂ is known to have a large effect on temperature.¹⁶⁾ The point at which hydrolysis begins to be observed occurs earlier as the temperature increases. Based on the sampling points used in this experiment, the hydrolysis reaction occurred 60 minutes after the initiation of leaching at 353 K, compared with 30 minutes at 363 K and 15 minutes at 373 K. Also, the leaching efficiency of Ti at the time of hydrolysis was reduced as the temperature increased. The hydrolysis reaction occurred at a leaching efficiency of 77% at 353 K, 65% at 363 K, and 60% Ti at 373 K. In addition, the conversion of Ti was derived for the 353 K, 363 K, and 373 K experiments, in which the precipitation reaction (Fig. 2) was visually confirmed. The initial Ti concentration at which hydrolysis started was also calculated as the concentration of dissolved Ti in solution.

Therefore the value of E_{Ti, t} which means the hydrolysis efficiency of Ti at time t was derived as follows:   

\begin{equation} E_{\text{Ti},\,t}(\%) = \frac{[\textit{TiOCl}_{2}]_{T}^{\textit{inital}} - [\textit{TiOCl}_{2}]_{T}^{t}}{[\textit{TiOCl}_{2}]_{T}^{\textit{inital}}} \end{equation} (6)

where T is the temperature (353, 363, or and 373 K) and t is the sampling time after which hydrolysis occurs at each temperature. The hydrolysis efficiency of Ti is shown in Fig. 5.

Fig. 5

Ti hydrolysis efficiency with time at each temperature.

The hydrolysis reaction occurs as follows:   

\begin{equation} \text{TiOCl$_{2}$} + \text{2H$_{2}$O} = \text{H$_{2}$TiO$_{3}$} + \text{2HCl} \end{equation} (7)

  

\begin{equation} \text{TiOCl$_{2}$} + \text{H$_{2}$O} = \text{TiO$_{2}$} + \text{2HCl} \end{equation} (8)

The hydrolysis reaction proceeds in a state in which Ti is dissolved. The Gibbs free energy of the above reaction was calculated based on the thermodynamic data of the reactants and the products (Table 2) and is presented in Table 3.

The reaction in which sodium titrate is dissolved in HCl is considered to occur simultaneously with hydrolysis as well as leaching. In this study, the leaching reaction was dominant at temperatures below 343 K for the first 4 hours. Nucleation in the TiOCl₂ solution owing to hydrolysis is thought to be continuously resolved in the embryo form which does not exceed the radius to be able to progress of nucleation and growth. Since nuclei of TiO₂ can progress from nucleation to growth on the other hand, Hydrolysis reaction at temperature higher than 353 K proceeds. If the temperature is higher, the activation energy value for nucleation and growth is lowered, which shortens the time for hydrolysis to start to be dominated.

Second, according to Le Chatelier’s law, the leaching reaction proceeds in the direction of consuming HCl. That is, as the leaching progresses, the acidity of the solution decreases. A decrease in acidity activated the hydrolysis reaction by Le Chatelier’s law.²⁴⁾ Furthermore, the concentration of TiOCl₂ increases with the dissolution of Ti in the solution, which accelerates the hydrolysis reaction. Comparing only the results 353, 363, and 373 K in Fig. 2, the hydrolysis reaction began to dominate when the TiOCl₂ concentration of the solution exceeded some value. The hydrolysis reaction also prevails at the fastest time at 373 K, despite the relatively lower TiOCl₂ concentration and higher acidity of the solution. These results indicate that the effect on the hydrolysis reaction is not only TiOCl₂ concentration and acidity of the solution but also temperature. Although the high TiOCl₂ concentration and low acidity in the solution, therefore, hydrolysis was not dominant at 343 K during the leaching time of this study, 4 hours.

3.4 The leaching residue with temperature

Figure 6 shows the XRD peaks of the residue after temperature-dependent leaching. Sodium titrate was no longer observed in all residues. This was attributed to Na having dissolved before Ti was dissolved. Na₂TiO₃ leaching eventually occurred because Na preferentially leached out and Na slowly diffused from the inside to the outside of the grain. At 343 K, an SiO₂ peak was observed. The higher SiO₂ peak intensity at 343 K was due to the higher Ti leaching efficiency at 343 K, resulting in almost no residue; this suggests a higher concentration of undissolved Si in the residue. A TiO₂ peak was also observed, probably due to insoluble TiO₂ or Ti hydrolysis at the end of leaching. TiO₂ peaks precipitated by the hydrolysis reaction began to appear from 353 K, as well as a peak corresponding to commercial TiO₂ pigment or synthetic rutile. As the leaching temperature increased, the shape of the XRD peak of the residue improved.

Fig. 6

XRD peaks of leaching residue.

4. Conclusions

In this work, leaching and hydrolysis reactions of Na₂TiO₃ were studied. The results can be summarized as follows.

1. (1)    Na₂TiO₃ was prepared by mixing the synthetic rutile and NaOH at a ratio of 1:1.13 and reacting at 773 K for 4 hours. Samples for leaching were obtained by washing Na₂TiO₃.
2. (2)    Leaching was carried out at various temperatures by dissolving the washed Na₂TiO₃ in 5 M hydrochloric acid with a pulp density of 10%. From 303 to 343 K, the leaching reaction of Ti was dominant. From 353 K, however, the solubilized Ti precipitated as TiO₂ owing to the hydrolysis reaction, and the leaching efficiency of Ti decreased.
3. (3)    The leaching and hydrolysis reactions were thermodynamically stable at the experimental temperatures. Above 353 K, the threshold for the activation energy required to reach the critical radius for nucleation and growth of TiO₂ was lowered and the hydrolysis reaction was activated. Also, at high leaching temperatures, the concentration of Ti ions in the leachate increased and the hydrolysis reaction irreversibly proceeded by Le Chatelier’s principle.
4. (4)    The leaching reaction was studied using a shrinking core model. The plot of reaction rate against time, using the chemical-reaction-controlled rate equation, had a linear. The k − 1/T graph gave an activation energy of 38.9 kJ/mol for the chemical-reaction-controlled step; thus, the leaching chemical reaction of washed Na₂TiO₃ was considered to be the rate-determining step.

Acknowledgements

This research was supported by the Basic Research Project (19-3212) of the Korea Institute of Geoscience and Mineral Resources, funded by the Ministry of Science, ICT and Future Planning of Korea.

REFERENCES

• 1)   W.  Zhang,  Z.  Zhu and  C.Y.  Cheng: Hydrometallurgy 108 (2011) 177–188.
• 2)   F.  Meng,  Y.  Liu,  J.  Chu,  W.  Wang and  T.  Qi: Can. J. Chem. Eng. 92 (2014) 1346–1352.
• 3)   F.  Meng,  Y.  Liu,  T.  Xue,  Q.  Su,  W.  Wang and  T.  Qi: RSC Advances 6 (2016) 112625–112633.
• 4)   Y.  Li,  H.  Yu,  Z.  Zhang,  M.  Zhang and  M.  Guo: ISIJ Int. 55 (2015) 134–141.
• 5)   Q.  Zhang,  Y.  Wu and  T.  Zuo: ACS Sustain. Chem. & Eng. 6 (2018) 3091–3101.
• 6)   Q.  Zhang,  Y.  Wu and  T.  Zuo: Metall. Mater. Trans. B 50 (2019) 471–479.
• 7)   T.  Xue,  L.  Wang,  T.  Qi,  J.  Chu,  J.  Qu and  C.  Liu: Hydrometallurgy 95 (2009) 22–27.
• 8)   Y.  Wang,  J.  Li,  L.  Wang,  T.  Qi,  D.  Chen and  W.  Wang: Chem. Eng. Technol. 34 (2011) 905–913.
• 9)   F.  Wu,  X.  Li,  Z.  Wang,  C.  Xu,  H.  He,  A.  Qi,  X.  Yin and  H.  Guo: Hydrometallurgy 140 (2013) 82–88.
• 10)   L.  Zhao,  Y.  Liu,  L.  Wang,  H.  Zhao,  D.  Chen,  B.  Zhong,  J.  Wang and  T.  Qi: Ind. Eng. Chem. Res. 53 (2014) 70–77.
• 11)   N.  El-Hazek,  T.A.  Lasheen,  R.  El-Sheikh and  S.A.  Zaki: Hydrometallurgy 87 (2007) 45–50.
• 12)   N.Y.  Mostafa,  N.H.H.  Mahmoud and  Z.K.  Heiba: Hydrometallurgy 139 (2013) 88–94.
• 13)   G.K.  Das,  Y.  Pranolo,  Z.  Zhu and  C.Y.  Cheng: Hydrometallurgy 133 (2013) 94–99.
• 14)   C.  Li,  B.  Liang and  H.Y.  Wang: Hydrometallurgy 91 (2008) 121–129.
• 15)   F.C.  Meng,  T.Y.  Xue,  Y.H.  Liu,  G.Z.  Zhang and  T.  Qi: Trans. Nonferrous Met. Soc. China 26 (2016) 1696–1705.
• 16)   S.  Middlemas,  Z.Z.  Fang and  P.  Fan: Hydrometallurgy 131–132 (2013) 107–113.
• 17)   G.  Moon,  Y.  Cho,  J.  Lee and  J.  Kang: Mater. Trans. 60 (2019) 988–996.
• 18)   E.  Olanipekun: Hydrometallurgy 53 (1999) 1–10.
• 19)   M.K.  Sarker,  A.K.M.B.  Rashid and  A.S.W.  Kurny: Int. J. Miner. Process. 80 (2006) 223–228.
• 20)   H.  Tsuchida,  E.  Narita,  H.  Takeuchi,  M.  Adachi and  T.  Okabe: Bull. Chem. Soc. Jpn. 55 (1982) 1934–1938.
• 21)  D.D. Wagman: The NBS Tables of Chemical Thermodynamic Properties: Selective Values for Inorganic and C1 and C2 Organic Substances in SI Units, Vol. 11, (American Chemical Society and the American Institute of Physics for the National Bureau of Standards, Washington DC, 1982).
• 22)   R.H.  West,  G.J.O.  Beran,  W.H.  Green and  M.  Kraft: J. Phys. Chem. A 111 (2007) 3560–3565.
• 23)  A. Roine: Sustainable Process Technology and Engineering, (A Manual on HSC Program, Continuous Research & Development, Finland, 2017).
• 24)   C.  Charbonneau,  R.  Gauvin and  G.P.  Demopoulos: J. Cryst. Growth 312 (2009) 86–94.