Recovery of Critical Elements (Dysprosium and Ytterbium) from Alkaline Process of Indonesian Zircon Tailings: Selective Leaching and Kinetics Study

Rare earth metals are important today, especially for heavy rare earth elements (HREEs). Dy and Yb, included in the HREE group, are widely used for permanent magnets and sensors. On the other hand, accumulated solid waste of zircon tailings is an environmental issue in mineral processing. Valuable minerals such as rare earth elements (REEs) also remain in the tailings. This study aims to determine the optimum conditions of the experimental parameters and obtain an appropriate kinetic model. The leaching process was carried out in a flat bottom three-neck flask equipped with a thermometer and reflux condenser. Observation of the effect of process conditions on recovery and kinetics analysis obtained satisfactory results. In this study, Dy and Yb leaching from solid REE-hydroxide using HCl solution afforded a total of 61.71% and 74.55% recovery of Dy and Yb, respectively, at operating conditions of 1 M HCl, 60°C, 10 g/100 mL (solid-to-liquid ratio), at 60 min (leaching duration), and 150 rpm (stirring speed). It was found that the leaching process was controlled by chemical reactions with apparent activation energies of 31.68 kJ/mol and 11.95 kJ/mol for Dy and Yb leaching, respectively.

Dysprosium; Kinetics analysis; Leaching; Ytterbium; Zircon tailings

Dysprosium (Dy) and ytterbium (Yb) are two elements found in the lanthanide group. Dy and Yb are classified as heavy rare earth elements (HREEs) based on their atomic numbers. These metals are included in the list of critical metals and are among the most widely used in the energy sector (Guyonnet et al., 2015). Dy metal is used in permanent magnets, phosphors, and metal halide lamps, while Yb is widely used for optical lenses, pressure sensors, and lasers (Lucas et al., 2014). The need for Dy and Yb metals in the future is predicted to increase (Supriadi et al., 2023).

REEs demand is predicted to increase by 2035 to reach around 374,000 metric tons, with the most used as permanent magnet products at approximately 80,000 metric tons (Yang et al., 2017). In contrast to Dy, Yb has begun to be widely used for memory devices and as a catalyst in the industry to replace other catalysts categorized as too toxic and polluting.

Currently, the primary sources of REEs are the minerals bastnaesite, monazite, and xenotime. Bastnaesite is the primary source, with REEs content of approximately 57% (Boulanger, Bazin, and Turgeon, 2019). Monazite and xenotime are often found in minerals bearing cassiterite and zircon (Trisnawati et al., 2022). The presence of this mineral is usually concentrated in the by-products of cassiterite/zircon processing (Gediga et al., 2019). REEs processing from zircon by-products solve solid waste accumulation and add value to the zircon beneficiation process.

Several studies have observed the REEs leaching kinetics from different material sources. The REEs leaching mechanism is usually influenced by the ore properties, mass transfer behavior, and reaction kinetics. In general, three models can describe the leaching process: homogeneous, shrinking core, and grain models (Petrus et al., 2022; Wanta et al., 2022; Liu et al., 2012). Shrinking-core models can be used to describe the leaching process for systems containing non-porous solid particles and reactant fluids.

Research into REEs leaching has been conducted extensively. Extraction was carried out using various secondary REE sources. Stopic and Friedrich (2016) researched the kinetics of yttrium leaching from ceramic waste dust. Under optimal operating conditions, a yttrium dissolution of 98.6% was achieved. Yttrium dissolution can be described using a kinetic model based on a chemical reaction as the limiting step. The observed activation energy was 31.2 kJ/mol. Obuz et al. (2018) studied the leaching kinetics of REEs in complex ores. The reaction constants for the solutions of Ce, La, Nd, and Pr were determined. Research on REE recovery from secondary sources from Indonesia has also been carried out. Rosita et al. (2019) observed the yttrium recovery from coal fly ash (CFA) using sulfuric acid, while Prihutami et al. (2020) also recovered yttrium from CFA using citric acid with the highest recovery of 83.35%.
        This study aims to investigate the leaching of Dy and Yb from secondary sources (zircon tailings) and determine optimal conditions for HREE leaching using HCl from REE-hydroxide. The study also aims to analyze the kinetics of the leaching process and obtain recovery data, such as a kinetic model and its activation energy value.

2.1. Roasting

     The raw material is zircon tailings from Monokem Surya, Ltd., mined and processed in Bangka Island and Central Kalimantan. The zircon tailings sample contained monazite and xenotime minerals. The Dy and Yb contents in the zircon tailings sample were 1.45% and 1.29%, respectively (Prameswara et al., 2020; Trisnawati et al., 2020). REE-hydroxide was prepared by reacting zircon tailings with NaOH flakes at a temperature of 450°C for 3 hours and a 1:1 wt ratio of zircon tailings and NaOH flakes. The roasting process was carried out in a muffle furnace. Our previous work has thoroughly described the REE-hydroxide preparation stage, sample characterization, mineral composition analysis, and elemental analysis (Prameswara et al., 2020).

2.2. Leaching experiment
    The leaching process was carried out in a flat bottom three-neck flask equipped with a thermometer, reflux condenser, and hot plate with magnetic stirrer Thermo Scientific Cimarec. The experimental schematic illustration is presented in Figure 1. Leaching was performed under several operating conditions: temperature, solid-to-liquid ratio (S/L), and acid concentration. A total of 5 g of REE-hydroxide with a particle size of 44 µm was leached for each data collection. Analytical grade HCl (37%, Merck & Co., Inc.) was used. The leaching process was carried out in 0.125–2 M HCl concentration, temperatures ranging from 30–110°C, and a solid-to-liquid (S/L) ratio of 5–25 g/100 mL, a stirring speed of 150 rpm. Based on several studies, higher stirring speed does not significantly affect REE recovery (Kurnia et al., 2016). Elemental analyses of Dy and Yb were performed on the leaching residue. Leaching recovery was calculated using Equation 1. 

where x₀ is the REE concentration in the feed multiplied by the mass of the feed (%), and X_i is the REE concentration in the residue multiplied by the mass of the residue (%). Elemental analysis was performed using an X-ray fluorescence spectrometer (Rigaku NEX QC, S/N QC1202) on the solid residue. Residue characterization at the optimum leaching process was carried out using Rigaku Miniflex 600 Benchtop X-ray diffractometer (XRD). The leaching process was repeated twice to ensure an acceptable measurement error (less than 10%).

2.3. Kinetic analysis

      Kinetic analysis was performed using a shrinking-core model. The kinetic models to be tested are the chemical reaction limiting step, diffusion through the ash layer limiting step, and the diffusion model developed by Zhuravlev, Leshokin, and Templeman (ZLT). These three models are expressed in Equations (2)–(4), respectively.

where x is the percentage leaching recovery divided by 100. kc, kd, and kz are the rate constant for the chemical reactions control kinetics model, diffusion through the ash layer kinetic model, and ZLT kinetic model, respectively. While t (min) is the leaching duration. The kinetic model selected to determine the kinetic parameters (frequency factor and activation energy) is the model that exhibits the highest R-squared value. The kinetic parameters were evaluated using the Arrhenius equation (Equations 5 and 6).

where k is the rate constant (kc/kd/kz), A is the frequency factor, E_a is the activation energy (kJ/mol), R = 8.314462 (J/(mol·K)), and T is the leaching temperature (K). The activation energy was obtained from the slope of the curve 1000/T vs. ln k, and the frequency factor was obtained from the intercept.

Figure 1 Experiment scheme

3.1. Roasting
      At this stage, REE-phosphate and zirconium silicate contained in monazite and xenotime naturally react with the NaOH flakes. As mentioned in da-Silva, Dutra, and Afonso (2012), the typical reaction is as follows (Equations 7-9). Based on our previous research, there was a decrease in the zircon concentration in the leaching residue. This indicates that zircon also reacted in the alkaline fusion process. The phosphate concentration is also reduced, indicating that most of the phosphate that binds to REE reacts with NaOH to produce Na₃PO₄ (Prameswara et al., 2020; Trisnawati et al., 2020). Separating the major and unwanted elements from the final product appears effective during the alkaline fusion process. This successful separation also optimizes the REEs leaching process later.

3.2. Effect of operating conditions on Dy and Yb leaching recovery

Observations were made on the effect of the leaching operating conditions on Dy and Yb recovery. As shown in Figure 2a, the recovery for Dy and Yb increased drastically in the HCl concentration range of 0.125–1 M. The recoveries for Dy and Yb reached 47.38% and 72.74%, respectively. The recoveries did not show any increase in HCl concentrations above 1 M. Dy and Yb recovery reach plateau shows that the limiting effects have been reached. Increasing the HCl concentration would only increase the dissolution of unwanted elements, such as Si, P, and Al (Walawalkar, Nichol, and Azimi, 2016).
Tabel 1 Thermodynamic parameter for leaching Dy(OH)₃ and Yb(OH)₃ 

The effect of temperature on the recovery of Dy and Yb is shown in Figure 2b. Increasing the temperature increased the recovery of Dy and Yb. The higher the temperature, the higher the kinetic energy of the reactants, causing the REE-hydroxide solid to collide and react faster with the HCl solution. The increase in recovery occurred in the temperature range of 30–60°C. Furthermore, there was no increase in the recovery of Dy and Yb. The increase in the leaching temperature did not increase the Dy and Yb recoveries. Temperatures above 90°C decrease REE leaching recovery by causing the solvent to evaporate, reducing the amount of solvent needed for the reaction. Additionally, high temperatures generate new issues in the reaction due to the solvent's boiling point. The exothermic nature of the REE-hydroxide leaching reaction with HCl is indicated by a negative value of enthalpy  obtained from the database and a negative value of Gibbs free energy  indicating the reaction can occur spontaneously. The low activation energy value (section 3.3) shows that the rate of heat generated during the reaction is fast.

Figure 2c indicates that an increase in recovery occurred at a solid-to-liquid (S/L) ratio of 10 g/100 mL. Subsequently, there was a decrease in the recovery of Dy. However, Yb recovery remained at 74.55% until the S/L ratio was 20 g/100 mL. Meanwhile, at an S/L ratio of 10 g/100 mL, Dy recovery reached 59.56% and then slowly decreased as the S/L ratio increased. Yb recovery is higher than Dy recovery at the same operating condition. This is because the solubility of Yb is higher than the solubility of Dy. In general, the solubility increases with increasing atomic number (Kim et al., 2018). The decrease in recovery occurs because the limit for REEs solubility has been reached, but other major elements continue to dissolve. The increase in the S/L ratio will lead to an increase in the density of the solution, which will then decrease the mobility of the Cl- ion. Therefore, it will be slowing down the rate of REE³⁺ transfer into the solution body. Thus, the product contains higher concentrations of major elements such as P, Zr, Si, and Al at a higher S/L ratio (Dupont and Binnemans, 2015). Therefore, an S/L ratio of 10g/100 mL was chosen as the optimum S/L ratio.

Figure 2 Effect of leaching operating conditions on Dy and Yb recoveries: (a) effect of HCl concentration (80°C, S/L=10 g/100 mL, 150 rpm, and 60 min); (b) effect of temperature (2 M HCl, S/L=10 g/100 mL, 150 rpm, and 60 min); and (c) effect of S/L ratio (2 M HCl, 80°C, 150 rpm, and 60 min)

The components in the leaching residues were analyzed using XRD to determine the components contained. The results of the XRD analysis are presented in Figure 3. ZrSiO₄ and SiO₂ represent the initial content of the tailings. The REE concentration is not significantly different from that of the undesirable elements. These data accurately indicate that most of the REEs are in the dissolved form in REE-Cl₃.

Elemental analysis was performed on solid product samples (residual leaching) under optimum operating conditions. Figure 4 shows the concentrations for some major impurity elements and HREEs (Dy and Yb) in the solid product. Major elements also appear to be much dissolved in the leaching process under optimum leaching conditions because the solubility of P, Zr, Si, and Al in HCl is higher than Dy and Yb.

Figure 3 XRD analysis of leaching feed and residue (solid product) at optimum operating conditions (1 M HCl, 60^°C, 10 g/100 mL, and 60 min)

Figure 4 Elemental analysis on the solid product at optimum operating conditions (1 M HCl, 60°C, 10 g/100 mL, and 60 min)

3.3. Kinetics analysis of Dy and Yb leaching process

    Kinetic analysis was performed for the Dy and Yb leaching processes at a temperature of 80–100°C. A shrinking-core model was used to describe the leaching process. The limiting step or control in this leaching kinetic model describes the most dominant process that occurs. Chemical reaction control, ash layer diffusion control, and ZLT kinetics models evaluated the REEs leaching recovery. The kinetics analysis was performed for a leaching duration of 0–7.5 minutes because there was no significant recovery increase above 7.5 minutes. Figures 5a and 5b show that the recovery of Dy and Yb constantly increased up to 7.5 min and then reached a plateau with a longer leaching duration.