With the growing focus on energy and environmental issues, lithium has become an increasing concern since it provides a key part in green energy technology [1–3]. The widespread adoption of lithium-ion batteries, particularly in electric vehicles, has had a substantial positive impact on the lithium market [4–8]. Consequently, ensuring a reliable supply of lithium is crucial for the advancement and sustainability of green technologies [9].

Despite the increasing demand and limited stocks, researchers have been striving to increase the lithium source [10, 11]. While the majority of the world’s lithium deposits are found in salt lakes, efforts to secure lithium resources have led to extensive exploration of seawater as a potential source [3, 7]. Seawater contains a staggering 16,000 times more lithium (230,000 MT) than salt lakes or ores (14 MT), making it an attractive option [12]. However, the process of extracting lithium from seawater is a complex challenge due to the low concentration of lithium ions and the presence of numerous other cations with similar properties [7, 13–15]. To increase the lithium-ion concentration, the approach involved attempting to extract lithium from concentrated seawater that is a by-product from salt manufacturing companies, which typically contains lithium-ion concentrations between two and one hundred times higher than seawater [16]. These solutions contain ions such as Ca²⁺, Mg²⁺, K⁺, and Na⁺, especially highly concentrated Li⁺ (approximately 20 mg/L) compared to ordinally seawater (0.17 mg/L) [17]. Therefore, the solution could be considered one of the potential lithium sources.

However, the solution is typically discharged or used in food production and discarded approximately 100,000 tons of concentrated seawater annually in South Korea [18]. Therefore, an efficient method of recovering lithium from concentrated seawater is required. Some previous research has proposed absorption methods for recovering lithium from seawater [19–25]. The spinel lithium manganese oxide or analogue adsorbents have been reported to extract lithium from source water via ion exchange [26]. Although these adsorbents have high selectivity for lithium, the method is time-consuming and hard to reuse since manganese ions can be released during acid treatment [27].

Recently, an electrochemical lithium recovery (ELR) system using λ-MnO₂ has been considered a promising technology due to its rapid, energy saving, and eco-friendly characteristics [28, 29]. However, λ-MnO₂ electrodes have limited stability, resulting in rapid electrode degradation from manganese dissolving after repeated charging and discharging processes [30, 31]. In particular, recovery lithium at extremely low lithium concentration significantly reduces the stability of the electrode, so it is a big challenge to extract lithium from concentrated seawater by the ELR system [15, 32]. In this study, we present a method to improve the stability of ELR systems using a λ-MnO₂ electrode to increase the efficiency of lithium recovery from the actual concentrated seawater from the salt manufacturing process (Fig. 1). In order to demonstrate the lithium-ion recovery performance of the LiMn₂O₄ electrode, we employed an Ag/AgCl electrode as the counter electrode due to the high capacity of the Ag electrode (theoretical value: 250 mAh/g), which is greater than that of the LiMn₂O₄ electrode (theoretical value: 148 mAh/g) and its ability to capture chloride ions effectively. By applying the state-of-charge (SoC) level during the operation, we effectively improved lithium-ion selectivity and cycling ability of the LiMn₂O₄ electrode. This approach has significantly enhanced both the stability and efficiency of the system [33–35]. The results of the energy efficiency and lithium-ion selectivity were confirmed by conducting lithium recovery tests in actual concentrated seawater under different SoC level operations. Furthermore, constant current cycling experiments were conducted using the SoC control method for evaluating the system’s stability.

Cyclic voltammetry (CV) was performed to confirm the electrochemical properties of LMO at different cations. Fig. 2 shows the results of CV curves according to the potential change from 0 V to 1.2 V at a scan rate of 0.5 mV s⁻¹. Under LiCl 1.0 M conditions, the two redox peaks attributed to the spinel structure of LMO can be verified with clarity [37–40]. In contrast, the significant peak could not be identified in the other four solutions, indicating that the LMO electrode has a strong affinity for Li ion and is capable of recovering lithium under a cation-mixed electrolyte, as shown in previous research [34].

Table 1 shows the chemical composition of concentrated seawater obtained from a salt manufacturing company (Hanjusalt). The lithium-ion concentration is 18 mg L⁻¹, which is about 100 times higher than seawater (0.17 mg L⁻¹); however, the solution is highly concentrated with ions other than lithium. For example, the mass ratios of sodium, potassium, and magnesium ion over lithium ion are 2000, 2570, and 2210, respectively. When comparing the concentrations of other ions focusing on the lithium-ion concentration, the concentrations of other ions were remarkably mixed at a high concentration compared to the lithium-ion concentration.

Fig. 3(a) shows the potential change with charging time for LMO electrodes during the lithium recovery process. The potential curves show that the charging time is similar for 100% SoC and 60% SoC; however, the voltage is gradually increasing at 60% SoC. This result indicates that 60% SoC has a more stable charging process, extending operation time and increasing the amount of lithium recovery. Fig. 3(b) shows the concentration changes of each ion in the recovery solution after the recovery process. The concentration changes of other ions, except lithium, were similar between 100% SoC and 60% SoC. However, focusing on the change in lithium ions, it indicates that using SoC control at 60% (1.99 mM) had a better recovery rate compared to the general method at 100% SoC (0.23 mM). In terms of energy, the energy consumption to recover 1 mole of lithium with 60% SoC is 27 Wh. Compared to 100% SoC (243 Wh/Li⁺mole), this value is about 89% lower, suggesting that SoC control significantly reduces the energy required for lithium recovery. In the case of 100% SoC (cell voltage range: 0 to 1.0 V) during a charging step, an oxygen evolution reaction (OER, 0.817 V vs. NHE at pH 7) may occur more actively compared to 60% SoC level operation, and hydrogen ions generated by the reaction (2H₂O → O₂ + 4H⁺ + 4e⁻) create an acidic condition. In particular, the LiMn₂O₄ electrode would be damaged by an acidic electrolyte as Mn³⁺ convert to Mn²⁺owing to the disproportionation process [41]. As a result, the degradation of the LiMn₂O₄ electrode accelerates under 100% SoC operation, leading to changes in the crystal structure, which decreases the lithium-ion selectivity and reduces the repetitive stability. As shown in Fig. 3(c) , the operation at 60% SoC leads to significantly reduced OER and Mn dissolution compared to 100% SoC during a charging step. This reduction in undesirable side reactions is a key reason why we concluded that the 60% SoC operation is more advantageous for the stability and efficiency of the system. By limiting the SoC to 60%, the side reactions are minimized, reducing the structure change of the LiMn₂O₄ electrode and enhancing the overall lithium recovery performance. Consequently, the SoC control strategy enables more stable and energy-efficient lithium recovery from concentrated seawater.

A galvanostatic charge-discharge method (GCD) was performed with a two-electrode cell in concentrated seawater, and the stability of the electrode was checked during charging and discharging for 30 cycles. Fig. 4(a) shows the results of the experiment under general conditions (100% SoC) as a charge/discharge curve. When it was 100% SoC, the experiment was performed at 0.05 A/g from 0 V to 1.0 V. The discharge capacity was 45.7 mAh/g in the first cycle, however, the capacity decreased sharply to 10.8 mAh/g after 30 cycles.

Fig. 4(b) shows the 60% SoC result. The 60% SoC control was performed in a voltage range from 0 V to 0.8924 V, using a voltage value when the experimental value was 52.5 mAh/g obtained at 100% SoC. The discharging capacity was 40.5 mAh/g in the first cycle and decreased to 16.3 mAh/g after 30 cycles. Although the capacity decreased in both cases, comparing the capacity difference between the first and last cycles showed that 60% SoC (24.2 mAh/g) reduced the capacity difference less than 100% SoC (34.9 mAh/g). This indicates that 60% SoC provides better cyclic stability, demonstrating that 100% SoC causes more damage to the electrode and less stability than 60% SoC.

Fig. 4(c) shows the specific capacity of the electrodes of SoC 100% and SoC 60% after 30 cycles of charge and discharge in concentrated seawater. Generally, SoC 100% usually shows a higher specific capacity because the battery can store more energy due to the longer charging time, however, this figure shows that SoC 60% has a higher capacity. This can be attributed to the fact that the overall capacity of SoC 100% has decreased over multiple charge and discharge cycles, and the multiple side reactions that occur at full charge have further reduced the capacity of the battery. Furthermore, the average coulombic efficiency at 100% SoC was 88.7%, whereas at 60% SoC, it recorded a higher average value of 98.5%. These results indicate that maintaining a 60% SoC contributes to reduced side reactions and improved stability.

The crystal structure of the LMO electrodes before and after 30 cycles was analyzed using X-ray diffraction (XRD) to confirm the changes in electrode structure, and the results are shown in Fig. 5. After 30 cycles were performed under the same conditions as in Fig. 4, surface analysis of pristine, 100% SoC, and 60% SoC was conducted. LiMn₂O₄ XRD analysis from 100 to 800 showed characteristic peaks of [111], [311], [222], [400], [331], [511], [440], and [531] due to the spinel structural characteristics [42, 43]. Comparing the 511 and 440 peaks between 58° and 66° with pristine, there was a noticeable change in the peak positions under full charge/discharge conditions (100% SoC). However, after controlling the SoC level, the peak of the electrode did not change significantly compared to pristine electrode. According to the results of our previous studies and the experiment, controlling the SoC level improves the stability of the LMO electrode for lithium recovery from concentrated seawater.

This research developed a process technology to enhance the stability of lithium recovery from concentrated seawater using an ELR system by incorporating state-of-charge (SoC) control. The feasibility of lithium recovery was verified under the actual concentrated seawater conditions, where various ions are present in high concentrations. In the lithium recovery tests, the 60% SoC configuration selectively recovered more lithium and consumed 89% less energy per mole of lithium recovered (27 Wh/Li⁺mole) compared to the 100% SoC configuration (243 Wh/Li⁺ mole). Further stability analysis by GCD results showed superior cyclic stability and higher capacity at 60% SoC. When comparing the average Coulombic efficiency over 30 charge and discharge cycles, the 60% SoC maintained a remarkable efficiency of 98.5%, compared to 88.5% for the 100% SoC. Furthermore, the XRD analysis revealed that the 60% SoC minimized structural changes in the LMO. According to these results, the SoC control operation in the ELR system has the potential to improve both the stability and efficiency of lithium recovery from concentrated seawater.