Study on Impurity Removal from Lepidolite Leaching Solution and the Extraction Process of Rubidium

Abstract

Efficient removal of iron and aluminum impurities is critical for the extraction of lithium and rubidium from zinnwaldite, a lithium-bearing mineral. In this study, solvent extraction using P507 was employed to remove iron and aluminum from zinnwaldite leaching solutions. However, stripping iron from the organic phase proved challenging due to the strong interaction between iron ions and the extractant. To address this, a novel reduction stripping method was developed using ascorbic acid (AA) as a reductant. This method exploits the reduction of Fe³⁺ to Fe²⁺ in the aqueous phase, weakening the binding between iron ions and the organic phase, thus enabling efficient stripping. The optimized process achieved over 99.99% removal of iron and aluminum impurities. Subsequently, rubidium was selectively extracted using t-BAMBP, with a total recovery rate of 88.53%. Scaling-up experiments confirmed the feasibility of the process for industrial applications, demonstrating high efficiency and reagent recyclability. This study offers a promising approach for the efficient extraction and separation of valuable metals from zinnwaldite, with potential for broader applications in metal processing.

1. Introduction

Lithium is a critical metal resource widely utilized across various industries, including traditional industry, electric vehicles (EVs), biomedicine, aerospace, and others [1,2]. With the rapid growth of industries such as electric vehicles and energy storage, the strategic importance and demand for lithium have been continuously rising, and the efficient extraction of lithium resources has attracted widespread attention. Lithium resources are primarily categorized into two types: ore-based and brine-based [3,4]. Although most lithium resources are concentrated in brine, lithium extraction from brine is still restricted by factors such as the complex composition, long production cycles, and vulnerability to weather conditions. Therefore, lithium extraction from ores can provide an important source of lithium supply [5].

Currently, the primary raw materials for lithium extraction are spodumene, lepidolite, and zinnwaldite. Zinnwaldite (KLiFeAl[AlSi₃O₁₀](F,OH)₂) [6] has a similar structure to that of lithium mica, but it contains a higher iron content and is frequently associated with rubidium. As a rare metal, rubidium has important applications in electronic semiconductors, renewable energy, and catalysis [7]. Zinnwaldite is a silicate mineral with a dense mineral structure and low chemical activity, which makes it difficult to react with acids and sodium hydroxide under normal temperatures and pressure [8]. Therefore, high temperature or mechanical activation, combined with sulfuric acid leaching, is required to extract valuable metals from zinnwaldite. Rubidium in the leaching solution is primarily separated and enriched through the extraction method. Lithium ions, on the other hand, are precipitated and enriched in the form of lithium carbonate. In the leaching solution, in addition to lithium and rubidium, there are also significant amounts of iron and aluminum impurity ions [6,9,10,11,12]. To obtain qualified lithium and rubidium products, iron and aluminum impurities need to be removed first.

The removal of iron and aluminum from zinnwaldite can be achieved through several methods, including chemical precipitation and solvent extraction. Among these, the chemical precipitation method is the current mainstream process. Under appropriate conditions, although the chemical precipitation method can completely remove iron and aluminum, the recovery rates of lithium and rubidium are only 60% due to the mechanical inclusion of precipitation, and iron and aluminum cannot be recycled. Although the solvent extraction method may not achieve deep impurity removal, it has the advantages of good selectivity, high metal recovery, and recyclable reagents. To efficiently extract valuable metals from zinnwaldite, solvent extraction was selected for the iron and aluminum removal in this study [13,14].

Common extractants with good selectivity for iron and aluminum ions include 1-phenyl-3-methyl-4-benzoylpyrazolone (P204) and 2-ethylhexyl phosphoric acid (P507). Although they exhibit effective removal of iron and aluminum, the strong binding interaction between the extractant molecules and Fe³⁺ makes stripping the iron from the organic phase difficult, preventing the extractant from being recycled efficiently [15,16]. Studies have shown that using excess sulfuric acid (H₂SO₄), complexing agents, precipitants, or reducing agents can partially achieve the stripping of iron ions [17,18,19].

The acid stripping process is typically carried out using excess sulfuric acid. The water phase after stripping requires a large amount of alkali for neutralization, so this method is primarily used for treating organic phases with low iron ion concentrations. Coordination stripping involves a reaction between Fe³⁺ and the complexing agent, leading to the formation of a compound with low solubility in organic phases. Common complexing agents used in this process include hydrochloric acid and oxalic acid. However, coordination stripping is relatively costly and has limited industrial application. The reduction stripping method uses iron powder, zinc powder, or sodium borohydride to reduce Fe³⁺ in the organic phase to Fe²⁺. This reduces the binding strength between the organic phase and iron, thereby achieving the stripping of iron ions. However, this method requires harsh operating conditions and is still in the experimental research stage [20]. The precipitation stripping method involves reagents forming a precipitate with iron ions, thereby separating iron from the organic phase. This method is mainly used for deep impurity removal from the organic phase.

To fully recover rubidium and lithium from the solution, it is necessary to develop efficient impurity removal technologies and optimize the extraction and separation process for rubidium. In this study, we used the leaching solution from zinnwaldite as the experimental object. P507 was selected as the extractant, and the extraction process was optimized to achieve efficient removal of impurities, including iron and aluminum ions, from the leaching solution. To address the issue of the difficulty of stripping iron from the organic phase, we developed a new water-phase reduction stripping method based on the principles of extraction equilibrium. We chose ascorbic acid (AA), which has high solubility in water but very low solubility in the organic phase, as the reductant. Since the distribution of Fe³⁺ between the organic and aqueous phases reaches equilibrium, during the stripping process, Fe³⁺ in the aqueous phase is reduced to Fe²⁺, which promotes the transfer of Fe³⁺ from the organic phase to the aqueous phase. Based on this principle, we achieved efficient stripping of iron and aluminum from the P507 organic phase under mild conditions. Furthermore, for the leaching solution with Li⁺ and Rb⁺ after impurity removal, we developed a 4-tert-buty1-2-(α-methy1 benzy1) pheno (t-BAMBP) extraction followed by a sulfuric acid stripping process. This process takes advantage of t-BAMBP’s selective extraction properties, with Rb⁺ > K⁺ > Li⁺ in NaOH solution, enabling a complete liquid treatment process after roasting leaching. After a complete scaling-up treatment process, the removal rates of iron and aluminum exceed 99.99%, and the total recovery rate of rubidium reaches 88.53%. Furthermore, the cyclic use of P507 and t-BAMBP has been achieved. Scaling-up experiments demonstrate that the newly developed process has promising industrial application potential. We hope that this study can provide valuable insights for the extraction and separation of metal ions from ores.

2. Materials and Methods

2.1. Materials and Reagents

The solution we are dealing with in this study is the leaching solution obtained from the roasting-sulfuric acid leaching process of zinnwaldite (XRD shown in Figure S1) from a mine in Hunan, China. The concentrations of the main metal ions in the leaching solution are shown in

Table 1. Chemical composition of zinnwaldite sulfuric acid leaching solution.

Chemical Composition	Fe	Al	Li	Rb
Concentration (g/L)	1.65	4.40	0.26	0.14

. The extractants P507 and t-BAMBP (the structures of which are shown in Figure S2) used in the extraction experiments were purchased from Wuhan Huaxiang Kejie Biotechnology company(Wuhan, China), with purities of 95% and 90%, respectively. Sulfonated kerosene was purchased from Guangdong Wengjiang Chemical reagent company(Shaoguan, China). Sulfuric acid, NaOH, KSCN, and ascorbic acid were all purchased from Sinopharm chemical reagent company, with analytical reagent grade purities. The water used in the experiments was deionized water.

2.2. Extraction and Stripping Experiments

All the extraction and stripping experiments were conducted with three parallel experiments. The optimal parameters were obtained by adjusting the organic-to-aqueous (O/A) phase ratio, extractant concentration, oscillation frequency, oscillation time, and solution pH. For the extraction process, the volume ratio of the aqueous phase was fixed at 50 mL. For the stripping process, the volume ratio of the organic phase was fixed at 50 mL. The extraction was performed using a liquid separation funnel oscillator (FY-A500, Changzhou Yichen Instrument Manufacturing Co., Ltd., Changzhou, China). After the oscillation, the liquid separation funnel was placed at room temperature for static stratification. The metal ion concentration in the aqueous phase was measured by a flame atomic absorption spectrometer (FL-AAS) (AA-6880, Shimadzu China Co., Ltd., Shanghai, China). Before the measurement, the concentration of the raffinate needed to be diluted to fall within the FL-AAS range. The corresponding extraction rate was calculated using Formula (1). The concentration of specific metal ions in the organic phase during extraction can be calculated using Equation (2).

$$E=\left(1-{\displaystyle \frac{M_{Aq}{V}_{Aq}}{M_{Fd}{V}_{Fd}}}\right)\times 100\%$$

(1)

$$M_{org}=\left(M_{Fd}{V}_{Fd}-M_{Aq}{V}_{Aq}\right)/V_{org}\times 100\%$$

(2)

E is the extraction rate of the organic phase. M_Aq (g/L) is the concentration of metal ions in the aqueous phase after extraction. M_org ((g/L) is the concentration of metal ions in the organic phase after extraction. M_Fd (g/L) is the initial concentration of the solution to be extracted. V_Aq (L) is the aqueous phase volume after extraction. V_org (L) is the organic phase volume after extraction. V_Fd is the volume of the initial solution to be extracted.

In the stripping experiment, the stripping rate represents the proportion of a specific metal ion transferred from the organic phase to the stripping aqueous phase. The stripping rate can be calculated using Equation (3).

$$S=\left({\displaystyle \frac{V_{\mathit{aq}}{M}_{\mathit{aq}}}{V_{\mathit{org}}{M}_{\mathit{org}}}}\right)\times 100\%$$

(3)

In the equation, S (%) is the stripping rate. V_org and V_aq (L) represent the volume of the organic phase and aqueous phase in the stripping process, respectively. M_org and M_aq (g/L) represent the concentration of the metal ion in the organic phase and aqueous phase, respectively.

3. Results

3.1. Extraction of Iron and Aluminum

We chose P507 as the extractant to remove impurity ions of iron and aluminum from the zinnwaldite leaching solution. First, we investigated the effects of oscillation frequency, extraction time, P507 concentration, initial solution pH, and O/A ratio on the iron ion extraction efficiency, aiming to obtain the optimal extraction conditions for iron ions.

3.1.1. Oscillation Frequency

The effect of oscillation frequency on the extraction rate was investigated. Figure 1 illustrates the relationship between oscillation frequency and extraction rate under the following experimental conditions: initial pH of 2.4, P507 extractant concentration of 20%, phase ratio (O/A) of 1, and extraction time of 6 min.

As shown in Figure 1, increasing the oscillation frequency enhances the iron extraction rate. At 100 rpm, the iron extraction rate is 14.13%, whereas it increases to 93.67% when the oscillation frequency is raised to 300 rpm. The extraction rates of lithium and rubidium remain at a low level of about 5%.

In the oscillation frequency experiment, the extraction rate of iron exhibited a steep slope, indicating that iron reacts quickly with the extractant and that mass transfer plays a significant role in the extraction equilibrium. Through the experimental phenomena in the oscillation process, it is found that the change in extraction rate is related to the emulsification phenomenon at the two-phase interface, particularly at low oscillation frequencies. When the oscillation frequency is too low, incomplete mixing of the two phases may occur, resulting in a smaller interface area between the organic and aqueous phases, which can affect the efficiency of the extraction process. At high oscillation frequency, the applied force field produces a demulsification effect and makes the two phases mix evenly. As shown in the pictures of the two-phase changes after extraction (Figure 2), increasing the oscillation frequency leads to the gradual clarification of the aqueous phase and an improvement in the utilization of the organic phase. Therefore, to improve the extraction rate of iron and effectively utilize the organic phase, the appropriate oscillation frequency is 300 rpm.

3.1.2. Extraction Time

The effect of extraction time on the extraction rate was investigated at an oscillation frequency of 300 rpm, initial pH of 2.4, P507 extractant concentration of 20%, and a phase ratio (O/A) of 1. As shown in Figure 3a, the extraction rate of iron reached 86% after 2 min of extraction. After 8 min of extraction, the extraction of iron reached equilibrium. At the initial stage of the extraction process, the extraction rate is high because the concentration difference of the solute between the aqueous phase and organic phase is large, providing a strong driving force, and the transfer of the solute is rapid. Therefore, the extraction efficiency increases quickly with time. As time progresses, the extraction rate gradually slows down and eventually reaches equilibrium. At this point, the concentration difference of the solute between the aqueous phase and organic phase decreases, and the extraction process approaches saturation. The extraction rate no longer increases significantly and reaches a balanced state. Excessively long extraction times may result in energy waste and increased operational costs. As the extraction time increases, the extraction rates of lithium and rubidium remain at a relatively low level, thus the optimal extraction time is determined to be 8 min.

3.1.3. Extractant Concentration

The effect of extractant concentration on the extraction rate was investigated under the condition of oscillation frequency of 300 rpm, initial pH of 2.4, phase ratio (O/A) of 1, and extraction time of 8 min. The relationship between P507 concentration and extraction rate is shown in Figure 3b. The extraction rate increased with the increase of P507 concentration. When the concentration of P507 reaches 15%, the extraction rate of iron can exceed 97%. Further increasing the concentration leads to only a slow increase in the extraction rate. By the time the concentration of P507 reaches 25%, the extraction rate of iron can even reach 99.99%. Within the experimental range of extractant concentration, the extraction rates of lithium and rubidium remain consistently below 8%.

A higher concentration of the extractant can provide more active sites, facilitating the binding and transfer of the target metal ions to the organic phase. To reduce the usage of the organic phase or stripping aqueous phase in the subsequent extraction and stripping processes, the composition of the organic phase should be 25% P507 + 75% sulfonated kerosene.

3.1.4. Initial pH Value

Figure 3c shows the effect of the initial pH of the aqueous phase on the extraction efficiency at an oscillation frequency of 300 rpm, P507 extractant concentration of 25%, phase ratio (O/A) of 1, and extraction time of 8 min. When the initial pH is 0.6, the extraction efficiency of iron is only 37.35%. As the pH increases to 1.8, the extraction efficiency of iron significantly rises to 98.47%. Further increasing the pH does not lead to a noticeable increase in the extraction efficiency of iron. Throughout the entire pH range of the experiment, the extraction efficiencies of lithium and rubidium remain below 8%.

The initial pH mainly affects the leaching rate of iron ions by altering their solvation environment. The color of the leaching solution at different pH values was observed to be light green, light yellow, and brownish red when the pH was below 2.4, 2.4, and 3.0, respectively (as shown in Figure S3). This is attributed to the increase in pH, which triggers the hydrolysis of iron ions and the subsequent formation of iron hydroxide precipitates. The increase in the initial pH will lead to an increase in cost and aggravate the dissolution of organic phase. Hence, the initial pH value of the leaching solution was selected to be 1.8.

3.1.5. O/A Ratio

The relationship between O/A and extraction rate is shown in Figure 3d under the conditions of an oscillation frequency of 300 rpm, extraction time of 8 min, organic phase composition of 25% P507, and initial pH value of 1.8. The increase in O/A can improve the extraction rate of metal ions. When O/A is 1.5, the iron in the leaching solution can be basically removed (over 99.99%), and the loss rates of lithium and rubidium are 4.91% and 7.46%, respectively. The increase in O/A will prolong the separation time of the two items, and increase the organic phase dissolved in the aqueous phase, resulting in organic phase loss and water pollution. Therefore, a low O/A of 1.5 was used in this experiment.

Based on the above experiments, we obtained the optimal extraction conditions for iron removal: oscillation frequency of 300 rpm, extraction time of 8 min, P507 concentration of 15%, initial pH of 1.8, and O/A ratio of 1.5. Under these conditions, a 99.99% iron extraction and removal efficiency can be achieved, with lithium and rubidium extraction rates not exceeding 8%. However, by detecting the aluminum ion concentration in the raffinate under these conditions, we found that the aluminum extraction rate is only 14.15%. Therefore, a two-stage extraction process is required to further remove aluminum ions from the aqueous phase.

3.1.6. Two-Stage Extraction of Aluminum

A two-stage extraction of the raffinate is conducted to further remove the aluminum from the aqueous phase. Considering the simplicity of the process operation, only the effects of pH value and O/A ratio are considered for the two-stage extraction of aluminum, with other parameters fixed at an oscillation frequency of 300 rpm, extraction time of 8 min, and P507 concentration of 25%.

As shown in Figure 4a, at the O/A ratio of 1.5, the extraction rate of aluminum significantly increases with the rise in pH. A steeper slope indicates that pH has a significant effect on aluminum extraction. pH mainly regulates the extraction rate of aluminum by altering its degree of hydrolysis. At pH = 4.8, the extraction rates of aluminum, lithium, and rubidium were 63.78%, 2.54%, and 2.36%, respectively. Increasing the pH value will lead to more P507 loss. Therefore, pH = 4.8 was selected as the initial pH for subsequent extraction. At pH 4.8, the extraction efficiency of aluminum was enhanced by increasing the O/A ratio. The effect of the O/A ratio on aluminum extraction is illustrated in Figure 4b. With the increase in the phase ratio, the extraction rates of lithium and rubidium did not change significantly, while the extraction rate of aluminum increased progressively. When the phase ratio reached 3, the aluminum extraction rate stabilized, reaching 90.34%. Under these conditions, after two extraction stages, 91.71% of the aluminum was removed from the aqueous solution. Therefore, the extraction of aluminum should be carried out at an O/A ratio of 3.

Under the optimal conditions, after a first stage of iron removal and a second stage of aluminum removal, we removed 99.99% of Fe³⁺ and 91.71% of Al³⁺ from the zinnwaldite leaching solution, with lithium and rubidium losses not exceeding 10%. The extracted aqueous solution was adjusted to a higher pH using NaOH to further remove iron and aluminum, and to meet the pH requirements for the extraction of rubidium using t-BAMBP.

3.2. Stripping of Iron and Aluminum

3.2.1. Sulfuric Acid Concentration and A/O Ratio

The organic phases obtained from the previous two extraction steps have similar compositions. Therefore, we first mixed the organic phases obtained from the two extraction steps and then performed the subsequent stripping of iron and aluminum from the organic phase. The effects of the sulfuric acid concentration on the stripping rate were investigated under the following conditions: a phase ratio of aqueous/organic (A/O ratio) of 1, oscillation frequency of 300 rpm, and extraction time of 8 min. The results are shown in Figure 5a. The stripping of aluminum is more efficient compared to iron stripping. When the concentration of sulfuric acid is increased to 4 mol/L, the stripping rate of aluminum can exceed 98%. Unfortunately, due to the strong binding force between Fe³⁺ and P507, even when the sulfuric acid concentration is increased to 6 mol/L, the stripping rate of iron remains below 40%. Therefore, we selected a sulfuric acid concentration of 5 mol/L as the optimal condition.

The effect of the A/O ratio on the stripping rate was tested under the following conditions: an oscillation frequency of 300 rpm, extraction time of 8 min, and sulfuric acid concentration of 5 mol/L. As shown in Figure 5b, an increase in the A/O ratio can improve the stripping rate of iron. When the A/O ratio reaches 4, the stripping rate of iron approaches equilibrium, with a value of 61.85%. In Figure 5, the slope of the iron stripping rate is relatively small, indicating that the sulfuric acid concentration has a limited effect on the iron stripping equilibrium. Therefore, new methods need to be developed to break the existing equilibrium and promote iron stripping.

3.2.2. Stripping of Fe³⁺ Enhanced by Ascorbic Acid

Traditional stripping methods reduce Fe³⁺ to Fe²⁺ in the organic phase to weaken the binding strength between iron and P507, thereby achieving iron stripping. In contrast, based on sulfuric acid stripping, we improved the stripping rate of iron by adding a reducing agent to the aqueous phase. We utilized the distribution equilibrium of Fe³⁺ between the organic and aqueous phases. When Fe³⁺ in the aqueous phase is reduced to Fe²⁺, the previous distribution equilibrium of Fe³⁺ is disrupted, thus promoting the transfer of iron from the organic phase to the aqueous phase. By adding an excess of reducing agent, we can maintain a low concentration of Fe³⁺ in the aqueous phase. Therefore, through a single extraction, almost all the iron in the organic phase can be stripped into the aqueous phase. Furthermore, we selected ascorbic acid as the reducing agent. Ascorbic acid, a mature industrial product, is inexpensive. More importantly, ascorbic acid is soluble in water but has low solubility in the organic phase, ensuring that it does not contaminate the organic phase after stripping. This allows for the recycling of the extractant. The resulting aqueous solution from the stripping contains iron and aluminum ions along with reduced ascorbic acid (named dehydroascorbic acid). Iron and aluminum ions can be effectively separated by precipitation, while the organic matter can be degraded by microorganisms in the water [15].

Under the conditions of an oscillation frequency of 300 rpm, a stripping time of 8 min, and a sulfuric acid concentration of 5 mol/L, the effect of the A/O ratio on the ascorbic acid stripping system was investigated. One molecule of ascorbic acid (C₆H₈O₆) can lose two electrons and be oxidized to dehydroascorbic acid (C₆H₆O₆). Therefore, 1 mole of ascorbic acid can reduce 2 moles of Fe³⁺. To ensure the complete reduction of Fe³⁺, 1.2 times the theoretical amount of ascorbic acid was added into sulfuric acid. The results of the stripping experiment are shown in

Table 2. Effect of A/O ratio on the stripping rate of iron and aluminum.

Add Ascorbic Acid	A/O	Stripping Rate of Iron (%)	Stripping Rate of Aluminum (%)
No	4	61.85	99.99
Yes	2	66.57	99.99
Yes	3	70.18	99.99
Yes	4	99.99	99.99

. The stripping rate of aluminum remained consistent at 99.99%. After the addition of ascorbic acid, the stripping of iron was significantly enhanced. When the A/O ratio was 4, the stripping rate of iron reached 99.99%. When KSCN was added to the stripped aqueous phase, the solvent barely turned red, indicating that the iron in the solution mainly existed in the Fe²⁺ form, which demonstrates the effective reduction of Fe³⁺ by ascorbic acid. This new process effectively solved the problem of the difficult stripping of iron in the P507 system.

3.3. Extraction of Rb with T-BAMBP

The purification solution obtained after the removal of iron and aluminum contains the metal elements of Li, Rb, K, and Na. t-BAMBP, an extractant capable of extracting alkali metals in the order of Cs > Rb > K > Li > Na under sodium hydroxide conditions, which was selected for the selective separation of Rb.

3.3.1. Effect of NaOH Concentration on Rubidium Extraction

The effect of NaOH concentration on rubidium extraction was investigated under the conditions of a t-BAMBP concentration of 0.25 mol/L, O/A ratio of 1, extraction time of 2 min, and oscillation frequency of 300 rpm. The results are shown in Figure 6a. The extraction rates of iron and potassium reach their maximum when the NaOH concentration is 40 g/L. With the increase in NaOH concentration, the extraction ability of t-BAMBP improved, but the increase in Na⁺ concentration inhibited the extraction of rubidium and potassium. In contrast to rubidium and potassium, the extraction rate of lithium continued to increase with a higher NaOH concentration. Therefore, we selected the optimal NaOH concentration of 40 g/L, under which most of the Rb can be extracted, while minimizing the transfer of lithium Li to the organic phase.

3.3.2. Effect of t-BAMBP Concentration on Rubidium Extraction

The effect of t-BAMBP concentration on rubidium extraction was investigated under the conditions of an NaOH concentration of 40 g/L, O/A ratio of 1, extraction time of 2 min, and oscillation frequency of 300 rpm. As shown in Figure 6b, with the increase in t-BAMBP concentration, the extraction rates of rubidium, lithium, and potassium increased. Large amounts of potassium and lithium in the aqueous phase entering the organic phase will reduce the purity of rubidium and complicate its subsequent purification. To achieve a higher purity organic phase and a larger separation factor, the t-BAMBP concentration was selected to be 0.25 mol/L, which provided the best separation of rubidium, lithium, and potassium.

3.3.3. Effect of O/A Ratio on Rubidium Extraction

The effect of O/A ratio on rubidium extraction was investigated under the conditions of an NaOH concentration of 40 g/L, t-BAMBP concentration of 0.25 mol/L, extraction time of 2 min, and oscillation frequency of 300 rpm. As shown in Figure 6c, the extraction rate of rubidium increases significantly with the rise in O/A ratio. When O/A exceeds 2, the growth rate starts to slow down. A higher O/A ratio would increase the burden on subsequent stripping, so the optimal O/A ratio is determined to be 2. The extraction rates of rubidium, potassium, and lithium were 88.41%, 31.65% and 1.59%, respectively.

3.3.4. Effect of Extraction Time on Rubidium Extraction

The effect of extraction time on rubidium extraction was investigated under the conditions of an NaOH concentration of 40 g/L, t-BAMBP concentration of 0.25 mol/L, O/A ratio 2, and oscillation frequency of 300 rpm. As shown in Figure 6d, the extraction of rubidium reached equilibrium rapidly. Further increases in extraction time have a minimal effect on the extraction rate. Besides, extending the extraction time excessively can result in the transfer of trace amounts of potassium into the organic phase, thereby affecting the separation process. Therefore, the appropriate extraction time of 2 min is recommended.

3.3.5. Extraction Equilibrium Isotherms of Rubidium

The extraction rate of rubidium was 88.41% under the optimum single-stage extraction conditions of an NaOH concentration of 40 g/L, t-BAMBP concentration of 0.25 mol/L, O/A ratio of 2, and extraction time of 2 min. Rubidium was not completely extracted from the solution during the single-stage extraction. Under the optimal single-stage extraction conditions, an organic phase was used to extract an initial aqueous solution, and then the extracted organic phase was continued to extract a new initial aqueous solution. After repeating this process five times, we recorded the concentration of rubidium ions in both the organic and aqueous phases after each extraction and plotted the extraction isotherm, as shown in Figure 7.

After continuous extraction of the initial aqueous solution 5 times, the organic phase still did not reach the saturated equilibrium state. At this point, the organic phase can continue to be used for extraction, with an extraction capacity of 43% of that of the fresh organic phase. Based on this, we carried out two-stage counter-current extraction under optimal extraction conditions, achieving a rubidium extraction rate greater than 99.99%, and the concentration of rubidium in the leaching solution was less than 0.001 g/L.

3.4. Stripping of Rubidium from Organic Phase

3.4.1. Effect of Washing Water pH on the Washing Rate of Metal Ions

Before stripping rubidium from the organic phase, the organic phase needs to be washed to separate potassium. At an A/O ratio of 1, the effect of pH value of washing water on the washing rate of metal ions was investigated. The results are shown in Figure 8a.

With the increase in the pH of the washing water phase, the elution rate of metal ions gradually increased, but rubidium was less affected. When the pH of the washing water phase is 10, the final rubidium loss rate is only 4%, and most of the impurities of lithium and potassium in the organic phase can be transferred to the water phase, thereby achieving effective separation of rubidium, lithium, and potassium. After washing, the higher the pH value, the longer the phase separation time required. Therefore, a washing water phase pH of 10 is considered optimal.

3.4.2. Effect of H₂SO₄ Concentration on the Stripping Rate of Rubidium

The washed organic phase was stripped using H₂SO₄ with an A/O ratio of 1, and the effect of H₂SO₄ concentration on the stripping rate of rubidium was investigated. The results are shown in Figure 8b. H₂SO₄ demonstrates a significant stripping effect on rubidium in the washed organic phase. At an H₂SO₄ concentration of 0.5 mol/L, the stripping rate of rubidium reaches 91.04%. When the concentration of H₂SO₄ is increased to 1.5 mol/L, rubidium is completely stripped from the organic phase. Therefore, we chose to perform the stripping of rubidium from the washed organic phase under the conditions of an A/O ratio of 1 and a sulfuric acid concentration of 1.5 mol/L. After a single-stage extraction, 99.99% of rubidium can be stripped into the aqueous phase, thereby achieving the recycling of the extractant.

3.5. Scale-Up Experiment

Based on the optimal process parameters obtained from the above experiments, we conducted a complete scale-up experiment for the extraction of iron and aluminum impurities and the separation of rubidium from a 10 L volume of zinnwaldite leaching solution. The process flow is shown in Figure 9. The metal concentration in the purified solution after the iron–aluminum removal process is shown in

Table 3. Metal concentration in the purified solution after the iron–aluminum removal process.

Metal Ion	Fe	Al	Li	Rb	K
Original leaching solution (g/L)	5.53	10.33	0.84	0.63	29.26
Purified solution (g/L)	/	/	0.81	0.60	28.48
Proportion (%)	<0.001	<0.001	96.10	93.84	97.33

, and the metal concentration in the finally obtained Rb₂SO₄ solution is shown in

Table 4. Metal concentration in the finally obtained Rb₂SO₄ solution.

Metal Ion	Fe	Al	Li	Rb	K
Original leaching solution (g/L)	5.53	10.33	0.84	0.63	29.27
Rb2SO4 solution (g/L)	/	/	0.021	0.59	1.67
Proportion (%)	<0.001	<0.001	0.25	94.34	0.57

. In the scale-up experiment, a 25% P507 + 75% sulfonated kerosene system was employed for the removal of iron and aluminum, with an oscillation frequency of 300 rpm and an extraction time of 8 min. The initial pH of the first-stage iron extraction leaching solution was 1.8, with an O/A ratio of 1.5. For the secondary aluminum extraction, the initial pH was 4.8, and the phase ratio was 3. The stripping solution consisted of 5 mol/L H₂SO₄ and ascorbic acid, with an A/O ratio of 4. The two-stage counter-current extraction of rubidium was performed under the following conditions: 0.25 mol/L of t-BAMBP, 40 g/L of NaOH, an O/A ratio of 2, and an extraction time of 2 min.

After the scale-up experiment, the removal rates of iron and aluminum exceed 99.99%, and the total recovery rate of rubidium reaches 88.53%. Furthermore, the cyclic use of P507 and t-BAMBP has been achieved. Scaling-up experiments demonstrate that the newly developed process has promising industrial application potential.

4. Conclusions

• In this study, we removed iron and aluminum impurities and separated rubidium from the leaching solution of zinnwaldite using an extraction process. Specifically, using P507 extractant and sulfonated kerosene as the organic phase, under optimal conditions, a single-stage extraction removed 99.99% of the iron, and a two-stage extraction removed 91.71% of the aluminum. For the purified solution, t-BAMBP was used as the extractant, and under optimal conditions, a two-stage counter-current extraction achieved a 99.99% extraction of rubidium. These processes were validated through comprehensive scale-up tests. As a result, the removal rates of iron and aluminum exceeded 99.99%, and the total recovery rate of rubidium reached 88.53%.
• For the recycling of the extractants, both aluminum and rubidium can be stripped from the organic phase back into the aqueous phase using a simple sulfuric acid process. To address the difficulty of stripping iron from the P507 organic phase, we developed a novel reduction stripping process. Ascorbic acid was used to reduce Fe³⁺ in the aqueous phase to Fe²⁺, breaking the distribution equilibrium of Fe³⁺ between the organic and aqueous phases, thus promoting the transfer of iron from the organic phase to the aqueous phase. After a one-step reduction stripping, 99.99% of the iron was removed from the organic phase. We hope this study provides new insights into the efficient extraction and separation of lithium and associated rubidium resources.