**2. Materials and Methods**

## *2.1. Materials and Equipment*

Similar to the composition of lithium-containing waste solution, a simulated solution was prepared using reagent-grade sulfate salt, and the composition is shown in Table 1. It contained 3.3 g/L lithium and a large amount of sodium as the remaining solution after cobalt and nickel were recovered from a waste secondary lithium battery. The simulated solutions were prepared using lithium sulfate (Li2SO4·H2O, 99%, JUNSEI) and sodium sulfate (Na2SO4, 99%, DAEJUNG). The electrodialysis equipment used to concentrate lithium was a CJT-055 electrodialyzer (Chang Jo Tech. Co., Seoul, Korea). A photograph of the equipment and the main specifications of the equipment are shown in Figure 2 and Table 2, respectively. Each chamber can be injected with 2.0 L of solution, and the power supply can operate within the DC current range of 0–30 volts and 0–5 amperes. In order to study the ion exchange membrane and the equipment, the upper limit of the current was set to 3 ampere. Titanium electrodes coated with platinum were used for the anode and cathode. The cartridge consisted of 10 pairs of NEOSEPTA (Ion exchange membrane, ASTOM Co., Tokyo, Japan) cation-exchange (CEM) and anion-exchange (AEM) membranes and an additional CEM spaced between the ion exchange membranes. The effective area of the ion exchange membranes in the cartridge is 55 cm2, and the main details of the membranes used in the study are shown in Table 3. Membrane stability is ensured below 40 ◦C, and the pH can be in a wide range from 0 to 14.

**Table 1.** Composition of Li-containing solution used in study.


**Figure 2.** Photograph of the electrodialysis equipment.


**Table 2.** Specifications of Electro Dialyzer.

**Table 3.** Characteristics of ion exchange membrane (NEOSEPTA).


\* Electric resistance: Measured on alternative current after equilibration with a 0.5 N NaCl solution at 25 ◦C; \*\* Burst Strength: Mullen Bursting strength.

#### *2.2. Methods*

The raw material solution was added to the dilution and concentration chambers at a constant volume ratio. Sodium sulfate (Na2SO4), used as the electrode solution, was prepared at a constant concentration, then 500 mL was added to the electrode chamber. The solution was sufficiently circulated for 5 min to remove residual air in the ion exchange cartridge and tube, then voltage was applied. It was applied as constant voltage of 10 volts, and the electrical conductivity of the solution in the dilution and concentration chambers over time was measured through the system's equipment. The experiment was terminated when the electrical conductivity of the solution in the dilution chamber became 1.00 mS/cm, where S (siemens) is the unit of conductivity and the reciprocal of the ohm (Ω) of electrical resistance, and 1 S = 1 A/V = 1 /Ω. At this time, there was a difference in the end time of the experiment depending on the volume ratio of the input solution and the initial concentration of ions. After the end of the experiment, samples were collected from the concentration and dilution chambers and analyzed by ICP-AES (Optima-4300 DV, Perkin Elmer Inc., Waltham, MA, US). The average energy consumption, average flux, concentration ratio, and water migration were respectively calculated by Equations (1)–(4), and the current was measured at intervals of 1 min through the device system to calculate the average current value, which was used in Equations (1) and (2).

The average energy consumption (E, kWh/m3) can be obtained by Equation (1) [18], where U(V) is applied voltage, I(A) is applied current, and V(L) is the total volume of raw material solution (concentration chamber + dilution chamber input).

$$E = \int\_0^t \frac{\mathsf{U}\varPi}{V} dt\tag{1}$$

The average flux (J, mol/m2h) can be obtained by Equation (2), where *m* (g) is the mass of Li concentrated in the concentrated chamber, *N* is the number of pairs of alternating cation and anion exchange membranes, *A* (m2) is the effective area of the exchange membrane, *M* is the molecular weight of lithium (g/mol), and *t* is the dialysis time (h).

$$J = \frac{m}{N \cdot A \cdot M \cdot t} \tag{2}$$

The concentration ratio (*CR*, %) can be obtained by Equation (3), where *Cf* is the final concentration and *Ci* is the initial concentration of solution in the concentration chamber.

$$\mathcal{C}\_{\mathbb{R}}\left(\%\right) = \frac{\mathcal{C}\_f}{\mathcal{C}\_i} \times 100\tag{3}$$

The water migration (*Wm*,%) can be obtained by Equation (4), where *<sup>V</sup> <sup>f</sup> <sup>D</sup>* is the final volume of solution and *V<sup>i</sup> <sup>D</sup>* is the initial volume of solution in the dilution chamber.

$$W\_m(\%) = \frac{V\_D^i - V\_D^f}{V\_D^i} \times 100\tag{4}$$

## **3. Results**

#### *3.1. E*ff*ects of Electrode Solution Concentration*

As a result of electrodialysis by HVRC in the authors' previous study [19], the effects of the applied voltage and the volume ratio of raw material solution entering the initial concentration and dilution chamber were examined. Prior to MSC electrodialysis, the concentration of sodium sulfate to be used as the electrode solution was adjusted to 0.1–1.0 M to examine the effect of the solution concentration. Then 500 mL of raw material solution was added to the dilution and concentration chambers, and 500 mL of electrode solution was added to the electrode solution chamber. The experiment was conducted by applying a constant voltage of 10 V. Figure 3 shows the electrical conductivity of the diluent and concentrate solutions over time. As the concentration of the electrode solution increased, less time was required for the process to decrease the electrical conductivity of the diluent to 1.00 mS/cm. When 0.1 M Na2SO4 was used, it took 3 h 30 min, with 0.7 M Na2SO4 it was 2 h 5 min, and with 1.0 M Na2SO4 it was 1 h 55 min. With 0.7 M and 1.0 M Na2SO4 compared to 0.1 M Na2SO4, the process lead time was reduced by more than 1 h. Table 4 shows the ion concentration in the concentrate solution after the end of the experiment. The same experiment was conducted three times, and the average value of each ion concentration was calculated. As the concentration of the solution increased, the concentration of Li increased. With 0.7 M Na2SO4, lithium was concentrated to 4.53 g/L, and with 1.0 M Na2SO4,

the concentration was slightly reduced to 4.19 g/L. Figure 4 shows energy consumption and average flux. As shown in Figure 4a, there was no significant difference in energy consumption from 41 to 43 kWh/m2 depending on the concentration of electrode solution. As shown in Figure 4b, as the concentration of electrode solution increased, the average flux increased, and it can be seen that the ion migration rate per unit time is high. Therefore, it was effective to use a concentration of 0.7 M Na2SO4 or higher in the ED process.

**Figure 3.** Effect of electrode solution concentration (M) on the electrical conductivity (D: Dilute solution, C: Concentrate solution, 10 V, D:C [1:1]).

**Table 4.** Analytical results of the concentrated solution after electrodialysis (ED) process (10V, D:C [1:1]).


**Figure 4.** Effect of electrode solution concentration (M) by (**a**) energy consumption and (**b**) average flux (10 V, D:C [1:1]).

#### *3.2. Concentration of Lithium by MSC Electrodialysis*

As a result of electrodialysis by HVRC in the authors' previous study [19], the highest Li concentration was obtained with applied voltage 10 V, volume ratio (diluent:concentrate) = 3:1, and 0.7 M Na2SO4 electrode solution. This was selected as an optimal condition and applied to the MSC, and 900 mL was added to the dilution chamber and 300 mL to the concentration chamber to start

the experiment. The ED process after the first stage was used as a feed solution in the second stage, and the third stage was supplied with feed solution in the same way. In order to confirm the continuous process, the raw material solution entering the second and third stages was not prepared and not used as simulation solution. The first stage was repeated 10 times, the second stage five times, and the third stage three times. The concentrate solution from each stage was collected in a bottle and used as a raw material solution in the next stage. The samples were taken from bottles to analyze the concentration of ions. The concentration of Li in each stage was compared, as shown in Figure 5. The concentration of lithium increased with the number of stages, and after the third stage, the lithium concentration was more than twice the initial concentration, at 6.95 g/L. Table 5 shows the results of ICP analysis for Li, Na, and SO4 according to the stage, lithium concentration, and process lead time. Three components increased in concentration as the number of stages increased. In the case of lithium, the concentration increased from the initial concentration or from the previous stage, but the concentration ratio (%) tended to decrease compared to the previous stage. As the ions concentration in the raw material solution increases, the amount of water hydrated and moved during the ED process also increases. Therefore, as the number of stage increases, the volume of the concentrated solution increases after the process. As shown in Table 6, the volume changes according to the water migration acted to reduce the *CR* (%) of lithium. The ion concentration in the raw material solution input at each stage increased, so the process lead time increased. The process lead time for ED in the first stage was 210 min, and in the third stage it increased significantly to 320 min.

**Figure 5.** Effect of number of concentration stage on the Lithium concentration (10 V, 0.7 M Na2SO4, D:C [3:1]).

**Table 5.** Analytical results of the concentrated solution after multistage concentration (MSC) process (10 V, 0.7 M Na2SO4, D:C [3:1]).




However, as shown in Table 5, the concentrations of sodium and sulfate in the first stage of ED were lower than those of the initial raw material solution. Repeating the first stage and storing the concentrated solution, white columnar crystals occurred, as shown in Figure 6a. These crystals were dried and crushed and recovered as a powder, and then analyzed by XRD, and were found to be Na2SO4 (thenardite, syn) (Figure 6b). Seen in the form of crystals recovered before drying, it is considered to be Na2SO4·10H2O. After the first stage, the crystals were separated into a solid–liquid, and after separation, the solution was used in a second stage ED process. When the concentration of sodium sulfate in the initial raw material solution is high, crystals may be generated inside the ion exchange cartridge during ED. These crystals can cause problems such as damage to the ion exchange membrane and reduced current efficiency, so it is necessary to remove sodium sulfate from the raw material solution before the ED process.

**Figure 6.** Photograph and result of XRD analysis of precipitate from 1st stage ED process by (**a**) precipitate and (**b**) XRD analysis of precipitate.

### *3.3. Sodium Sulfate Removal Process*

In the previous experiment, when the concentration of Na and SO4 in the raw material solution was high, a problem occurred: Na2SO4·10H2O crystals were generated during ED. Methanol was used as a precipitant to remove sodium sulfate in solution. Since sodium sulfate has low solubility in methanol, sodium sulfate may be precipitated when methanol is added to the sodium sulfate solution [20,21]. Methanol was added to the solution based on the composition of Table 1 in a constant volume ratio, followed by stirring for 30 min to induce a reaction. After the end of the reaction, filtration was performed to separate solids. Methanol was added in a volume ratio of 0.1 to 1.0 relative to the volume of the raw material solution, and as the volume of added methanol increased, the removal rate (%) of Li, Na, and SO4 increased, as shown in Figure 7. When the volume ratio of added methanol was 0.6 or more, the loss rate of lithium was 14% or more. It is preferable to proceed with the Na2SO4 removal process at a volume ratio of 0.4 or less of methanol with a lithium loss (%) of less than 0.3%. When the volume ratio of added methanol was 0.4, lithium loss was about 2%, Na removal was 65%, and SO4 removal was 51%. Figure 8a shows the powder generated in the Na2SO4 removal process, and in Figure 8b the results of XRD analysis show that it was sodium sulfate. It is considered that the recovered powder can be used to prepare the electrode solution for ED process without purification. In addition, methanol is likely to be recovered through fractional distillation after sodium sulfate filtration. The recovered sodium sulfate and methanol can be reused to reduce the overall process cost.

**Figure 7.** Effect of methanol addition on the sodium and sulfate removal (25 ◦C).

**Figure 8.** Photograph and result of XRD analysis of precipitate from Na2SO4 removal process by (**a**) precipitate and (**b**) XRD analysis of precipitate.

#### *3.4. Concentration of Lithium by MSC Electrodialysis after Sodium Sulfate Removal Process*

Sodium sulfate crystals were generated during concentration by ED using a raw material solution containing a large amount of sodium sulfate. Prior to the ED process, the MSC ED process was performed using a solution from which sodium sulfate was partially removed through a Na2SO4 removal process as a raw material. The first stage was repeated 20 times, the second stage eight times, the third stage five times, and the fourth stage three times. The concentrate solution from each stage was collected in a bottle and used as a feed solution in the next stage. Methanol was added to the raw material solution in Table 1 in a volume ratio of 0.4 to induce the reaction, and the solution from which sodium sulfate was removed was used as the raw material solution. Its composition is about 2.65 g/L Li, about 10 g/L Na, and about 71 g/L SO4. In Figure 9, the concentration of lithium increased with the number of stages; it was concentrated to 10.02 g/L after four stages from an initial concentration of 2.65 g/L. Although the concentration of lithium was lower than that of the raw material solution, which had not undergone a Na2SO4 removal process, it showed a higher concentration even in the first stage, and the process time was shortened. Table 7 shows the concentration of lithium, sodium, and sulfate in each stage, the *CR* (%) of lithium, and the process lead time. Compared to the initial concentration of lithium, the *CR* (%) tended to increase, but compared to the previous stage, the rate tended to decrease gradually. This phenomenon occurred because water was migrated from the dilution chamber to the concentration chamber as shown in Table 8. In addition, it showed a tendency to increase the process lead time compared to the *CR* (%) of lithium. However, compared to using a raw material solution that

had not undergone a Na2SO4 removal process, a high *CR* (%) is shown in a relatively short process lead time. If the concentration of components other than lithium is high in the raw material solution, it is considered efficient to perform ED after removal.

**Figure 9.** Effect of number of concentration stage on the lithium concentration after Na2SO4 removal process (10 V, 0.7 M Na2SO4, D:C [3:1]).

**Table 7.** Analytical results of the concentrated solution after Na2SO4 removal process and MSC process (10 V, 0.7 M Na2SO4, D:C [3:1]).




#### **4. Conclusions**

In this study, the valuable metals were recovered from the waste lithium secondary battery and a simulated solution was prepared with the same solution composition as the remaining filtrate. An experiment was performed to concentrate lithium from the simulated solution through a multi-stage concentration process using electrodialysis. The following conclusions were drawn:

(1) As the concentration of the electrode solution increased, the concentration ratio (%) of lithium increased and the process lead time decreased. When 0.7 M Na2SO4 was used as an electrode solution, the average flux during the process was high and the concentration ratio (%) of lithium increased, which was effective for concentration.

(2) In the case of multistage concentration electrodialysis using an initial raw material solution containing a large amount of sodium sulfate, sodium sulfate crystals were generated during the ED process. In addition, the concentration ratio (%) of lithium was low even after the third stage of ED, and a long process lead time was required.

(3) Sodium sulfate was precipitated by adding methanol to remove sodium from the raw material solution. As the volume of methanol increased, the sodium sulfate removal rate (%) increased, but the loss rate (%) of lithium also increased. When methanol as a raw material solution was added in a volume ratio of 0.4, sodium was removed by 65% and sulfate ion by 51%, and the loss rate of lithium was 0.3%.

(4) During multistage concentration (MSC) using electrodialysis with a solution from which sodium was partially removed as raw material, lithium was effectively concentrated to 10 g/L in the solution in four stages.

**Author Contributions:** Y.C., K.K., J.A., and J.L. carried out the experiments and analysis and interpretation of the data obtained. Y.C. and J.A. wrote and edited the paper. J.A. is the co-principal investigator of this work. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP), funded by the Ministry of Trade, Industry and Energy, Republic of Korea (Project No. 20185210100050).

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
