Lithium extraction from low-grade brines via strain-induced electronic structure modulation of MnO₂ nanorods through Mg incorporation

Abstract

In this study, the incorporation of Mg²⁺ induced the transformation of MnO₂ from the original cubic [MnO₆] octahedron to a flattened conformation (FLA-MnO₂). The flattened conformation mitigated structural deformation of the [Mn³⁺·O₆] octahedron by inhibiting the J–T aberration during the reduction process of Mn⁴⁺, thereby improving the selective adsorption capacity of lithium. When used in hybrid capacitive deionization (HCDI), FLA-MnO₂ exhibited a high Li⁺ adsorption capacity of 30.14 mg g⁻¹ in 32.74 mg L⁻¹ Li⁺ ion solution, with low energy consumption of 0.45 Wh g⁻¹. Notably, the FLA-MnO₂‖AC HCDI cell exhibited high stability, maintaining 82% of its capacity over 100 cycles. This was evidenced by the 82% capacity retention and low Mn loss (1.3%) over 100 cycles. Finally, the excellent selective extraction performance of Li⁺ ions was demonstrated in Lop Nor, the low-grade original brine of the XieLi salt flats. The calculated separation factors for Li⁺/Na⁺, Li⁺/K⁺, Li⁺/Ca²⁺, and Li⁺/Mg²⁺ reached impressive values of 4.37, 3.75, 3.24, and 2.11, respectively, making the Lop Nor a candidate electrode for lithium extraction from low-grade raw brines.

---

1. Introduction

In the new energy era, lithium as a strategic resource has gradually been emphasized by various countries, and it has a wide range of uses in new energy batteries, chemical catalysis, ceramics, atomic energy, aerospace, and other fields.^1–3 Although lithium is abundant globally, it is not evenly distributed; it is concentrated in Bolivia (24%), Argentina (22%), Chile (11%), the USA (10%), Australia (8%), and China (6%).⁴ In recent years, the use of lithium-ion batteries has increased dramatically due to the rapid growth of the electric vehicle industry, which has increased the global demand for lithium resources.⁵ To overcome the lithium supply shortage and consider the limited reserves of lithium-bearing ores such as lepidolite and spodumene, researchers have invested significant effort in the large-scale development of water-based lithium resources (e.g., salt lakes and seawater).⁶ Currently, there are many standard methods for extracting lithium from salt lakes, among which precipitation, solvent extraction, and electrodialysis can effectively extract Li⁺ ions. However, these methods add chemical reagents that cause secondary pollution or involve high energy consumption processes, and they have other shortcomings that limit their long-term industrialization applications.^7,8 The most challenging obstacle for Li⁺ extraction from salt lakes is the selective extraction of Li⁺ from complex aqueous systems containing many competing ions (e.g., Na⁺, K⁺, Mg²⁺, Ca²⁺), especially for brines with high Mg/Li ratios, it is difficult to separate Li⁺ from Mg²⁺ due to their similar physicochemical properties and ionic radii.⁹ Therefore, it is urgently necessary to develop a more efficient, less energy-consuming, and environmentally friendly method for the highly selective and effective extraction of Li⁺ from salt lakes and seawater.

One of the most promising methods for recovering Li⁺ from salt lakes is Li⁺ ion sieve adsorption.¹⁰ In recent years, a variety of inorganic Li⁺ ion sieves (LIS), such as spinel-type aqueous manganese oxides derived from lithium-manganese oxides, have been developed and effectively used as Li⁺ adsorbents.^10–13 However, the powdery nature of LIS makes it susceptible to physical loss during recovery after adsorption. Therefore, using LIS as a powder adsorbent is not economically viable. To solve this problem, the researchers selectively extracted Li⁺ from brine using capacitive deionization (CDI). In this method, Li⁺ ion sieves are made into electrodes, and Li⁺ ions can be selectively adsorbed by driving an electric current into the LIS electrode. The λ-MnO₂//Pt cell is the first reported selective Li⁺ recovery system.¹⁴ Subsequently, researchers have conducted many studies on lithium-ion sieve electrode materials; HCDI systems such as LiCoMnO₄‖AC, rGO/NCM‖AC, and CF-NMMO‖AC have been reported for Li⁺ recovery.^12,15,16 HCDI offers several advantages, including a high rate, low energy consumption, no secondary contamination, excellent stability, and repeatability.¹⁷ In addition, the excellent Li⁺ selectivity provided by the Li⁺ ion sieve electrode ensures high capacity during lithium recovery.

λ-MnO₂ is an excellent Li⁺ selective material due to its unique tetrahedral vacant sites.¹⁸ However, its inherent poor conductivity and poor stability limit its large-scale application.^13,19 A series of methods were explored to solve these problems. In general, cationic disorder structures in spinel and heterogeneous phases or low-valent cations (e.g., Li⁺, Ni²⁺, and Al³⁺) are doped in manganese-based oxides in place of manganese atoms to increase the content of Mn⁴⁺ and thus mitigate localized J–T aberrations.^20–23 However, these methods have drawbacks in terms of their scalability and lack of a deeper understanding. The [Mn⁴⁺·O₆] octahedra are characterized by the aggregation of six symmetrically arranged oxide ligands. During the reduction of Mn⁴⁺, when the double degenerate e_g orbitals undergo single occupation, the electron density is distributed in a noncubic manner, resulting in the elongation of the two axial Mn–O bonds and contraction of the four equatorial Mn–O bonds. This sizeable structural deformation leads to undesirable intracrystalline cracks, poor structural stability, and rapid capacity decay.²⁴ Therefore, the intrinsic improvement of cyclic stability by suppressing the J–T aberration requires further exploration.

This work proposes a method for compressing [MnO₆] octahedra by incorporating Mg²⁺ ions between transition metal (TM) slabs to improve their electrochemical stability. The resulting flattened [MnO₆] configuration in λ-MnO₂ (FLA-MnO₂) avoids the formation of degenerate electronic ground states during electrochemical processes, thus suppressing the J–T aberration, effectively improving the electrochemical reversibility and further enhancing the cycling stability during Li⁺ insertion/extraction. FLA-MnO₂ as the cathode and activated carbon (AC) as the anode was investigated as an HCDI system to evaluate the effects of operating voltage and feed concentration on the performance of HCDI, cycling stability, and the performance of selective recovery of Li⁺ ions from brine. The results show that the prepared FLA-MnO₂ electrode has excellent Li⁺ ion adsorption capacity in low-concentration Li⁺ ion solutions with lower energy consumption. It should also be noted that the FLA-MnO₂ electrode was stable, as evidenced by the significant capacity retention, low manganese dissolution loss rate, and stable pH change during cycling. The FLA-MnO₂‖AC system also demonstrated excellent Li⁺ ion selectivity in Lop Nor, Xieli salt flat low-grade original brine, which also proved the potential of FLA-MnO₂ for the extraction of lithium from salt lakes.

2. Materials and methods

2.1. Materials

Anhydrous manganous sulfate (MnSO₄, 99.99%, Aladdin), ammonium persulphate ((NH₄)₂S₂O₈, AR, Sinopharm Chemical Reagent), lithium hydroxide monohydrate (LiOH·H₂O, 99.99%, Aladdin), sulfuric acid (H₂SO₄, Aladdin), magnesium nitrate hexahydrate (Mg(NO₃)₂·6H₂O, 98%, Sinopharm Chemical Reagent), carbon black (Super P Conductive, 99%, metals basis, Alfa Aesar), activated carbon (Sinopharm Chemical Reagent), 1-methyl-2-pyrrolidinone (NMP, 99.5%, Sigma-Aldrich), and absolute ethanol (Sinopharm Chemical Reagent) were used as received. All solutions were prepared using deionized water (DI H₂O) with resistivity ≥18 MΩ·cm at 25 °C.

2.2. Fabrication of electrode materials

2.2.1 Synthesis of MnO₂. Typically, 0.151 g anhydrous manganous sulfate (MnSO₄) and 0.228 g ammonium persulphate ((NH₄)₂S₂O₈) were mixed in 80 mL DI water, and then the solution was transferred into a Teflon-lined autoclave to perform the hydrothermal reaction at 120 °C for 12 h. The solid powders (β-MnO₂) were filtered, washed with deionized water and dried in a vacuum oven at 120 °C for 12 hours. Then, 0.105 g of lithium hydroxide monohydrate (LiOH·H₂O) and 0.435 g of β-MnO₂ in 10 mL of ethanol for 1 hour, drying the mixture at ambient temperature, and then calcinated at 700 °C for 10 hours. To prepare MnO₂, the obtained powder was dissolved in 0.5 M H₂SO₄ solution and subjected to a 50-minutes ultrasonic treatment, followed by washing, drying, and collecting for further use.

2.2.2 Synthesize of FLA-MnO₂. 0.1 g of the above-prepared MnO₂ and 0.0294 g Mg(NO₃)₂·6H₂O were added to deionized water (140 mL). The mixture was stirred at 60 °C for 20 h. The reaction product was washed with ultrapure water and absolute ethanol until the conductivity of the supernatant was below 30 μS cm⁻¹. Finally, the product was dried in a vacuum oven at 60 °C for 12 h. The resulting sample is defined as FLA-MnO₂.

2.3. Characterization of electrode materials

The crystallographic characteristics of the as-prepared MnO₂ and FLA-MnO₂ materials were analyzed using an XRD-6100 diffractometer (Shimadzu, Japan) at a tube voltage of 40 kV and a tube current of 30 mA with Cu Kα radiation. The XRD data were analyzed using Jade 9.0 software (Materials Data, Inc). X-ray photoelectron spectroscopy (XPS) data were recorded using an X-ray photoelectron spectrometer (PH1-5700 ESCA system, USA) equipped with a hemispherical analyzer and an aluminum anode (monochromatic Al Kα 1486.6 eV) as source (at 12–14 kV and 10–20 mA). The data processing and peak fitting were performed using Avangage software. The element content in Materials or brine was determined using an Optima 8300 (PerkinElmer, USA) inductively coupled plasma optical emission spectroscopy (ICP-OES) instrument. The morphology of MnO₂ or FLA-MnO₂ samples was examined by scanning electron microscopy (SEM) using a SU1510 microscope (Hitachi, Japan) at an accelerating voltage of 1.5 kV. Raman spectroscopy was performed using a UniRAM (UniNanoTech) instrument under excitation by a 532 nm laser.

2.4. Electrochemical measurements

Cyclic voltammetry (CV), galvanostatic charge–discharge (GCD), and electrochemical impedance spectroscopy (EIS) analyses were conducted in a 1.0 M LiCl solution using a three-electrode system, which was monitored and controlled by a CS310M electrochemical workstation. (Correst Instruments Inc., Wuhan, China). The three-electrode configuration is composed of a working electrode, a platinum foil electrode (counter electrode), and an Ag/AgCl (3 M KCl) electrode (reference electrode). The working electrodes were fabricated by combining MnO₂ or FLA-MnO₂, acetylene black (Alfa Aesar, Shanghai, China), and polyvinylidene fluoride (PVDF, Aladdin Chemical Co.) in an 8 : 1 : 1 ratio. The mixture was ground by hand and dispersed in a slurry of 1-methyl-2-pyrrolidinone (NMP, Alfa Aesar, Shanghai, China), which was then deposited onto a graphite paper with a size of 1 cm × 1 cm and dried in a vacuum at 80 °C. Approximately 1 mg of the active materials were loaded on each electrode. More details on electrochemical characterizations are presented in Text S2.†

2.5. Electrochemical HCDI cell and lithium recovery process

A two-electrode system was constructed using MnO₂ or FLA-MnO₂ as the anode and AC as the cathode in the HCDI cell (Fig. 1). The HCDI cell and other apparatus were integrated to construct the desalination platform (Fig. 1a and b), which was operated in batch mode and under constant voltage conditions for all the desalination tests. When a constant voltage is maintained between the FLA-MnO₂ (or MnO₂) and AC electrodes, Li⁺ ions in the feed migrate toward the anode and enter the lattice of spinel FLA-MnO₂ (or MnO₂) via the faradaic process, while the Cl⁻ anions were captured by the AC electrode. During this process, the lithium concentration in the feed water would be decreased. Next, pure water is passed into the cell, and the electrodes could be regenerated by reversing the cell voltage while the ions were released, producing a solution enriched with Li⁺ ions (Fig. 1c and d). A total of 40 mL of feed solution was circulated between an HCDI cell with a volume of ∼2 mL and the feed solution reservoir at a flow rate of 40 mL min⁻¹. The conductivity of the effluent was consecutively monitored using a DDSJ-307F conductivity meter (INESA Scientific Instrument Co., Shanghai, China) to calculate the Li⁺ concentration. To examine the selectivity of the FLA-MnO₂ electrodes, i.e. a series of molar ratio Mg²⁺/Li⁺ binary salt solutions and Lop Nor the Xieli salt flats low-grade original brine (Xinjiang, China, Table S5, ESI†) were employed to assess the practical applicability. The concentration of each ion before and after electrochemical lithium recovery was determined using an inductively coupled plasma atomic emission spectrometer. More details on the Li⁺ selectivity coefficient, charge efficiency and energy consumption are given in Text S2.† Furthermore, for measuring the Li⁺ concentration, the correlation between conductivity and ICP measurements is presented in Table S6 and Fig. S18 (ESI)† to demonstrate the validity of the testing method.

Fig. 1 Electrochemical HCDI cell: (a) experimental setup, (b) schematic design, (c) adsorption process, and (d) desorption process.

3. Results and discussion

3.1. Composition and morphology of MnO₂ and FLA-MnO₂ materials

Fig. S1 (ESI)† illustrates the synthetic process of the MnO₂ precursor via the traditional chemical precipitation method. Mg²⁺-confined MnO₂ was then prepared using MnO₂ nanorods as the starting material through subsequent ion exchange.²⁵ The XRD patterns of the as-prepared MnO₂ and FLA-MnO₂ nanorods (Fig. 2a) suggest that all samples exhibited the standard spinel phase (space group: Fd3m; JCPDS: PDF# 44-0992) without other impurity peaks. The sharp and well-defined peaks also indicate that these samples were well-crystallized, confirming the successful synthesis of MnO₂ and FLA-MnO₂. Further to investigate the effect of Mg²⁺ incorporation on the structure and morphology of MnO₂, SEM images in Fig. S2 (ESI)† show that both MnO₂ and FLA-MnO₂ have short nanorod structures with an average length of approximately 450 nm, and there is no significant change in morphology. This observation suggests that the appropriate amount of Mg incorporation did not alter the structure of spinel MnO₂.²⁶ The chemical compositions of the fabricated MnO₂ and FLA-MnO₂ are provided in Table S1 (ESI).† Notably, FLA-MnO₂ exhibited a right-shifted peak assigned to the (111) plane, demonstrating a contractive interlayer distance of (111). In addition, the lattice constants a and b, as well as the cell volume of FLA-MnO₂, increased with the incorporation of Mg, and the lattice constant c slightly decreased (Fig. 2b). The increased cell volume of FLA-MnO₂ can be attributed to the strong electrostatic interaction between the incorporated Mg²⁺ and adjacent O atom layers. The element mapping (Fig. S3 and S4, ESI†) shows that uniformly distributed Mn, O, and Mg elements can be observed in FLA-MnO₂, whereas only Mn and O elements can be found in MnO₂, further verifying the successful incorporation of Mg²⁺ in FLA-MnO₂.

Fig. 2 (a) XRD patterns of MnO₂ and FLA-MnO₂. (b) Lattice constants and cell volumes of MnO₂ and FLA-MnO₂. (c) XPS Mn 2p survey of MnO₂ and FLA-MnO₂. (d) Raman survey of MnO₂ and FLA-MnO₂.

To characterize the fine structures of MnO₂ and FLA-MnO₂, spectroscopic characterization was performed. Fig. S5† shows the full XPS survey of MnO₂ and FLA-MnO₂. In the case of the XPS survey of the Mg²⁺-confined MnO₂, the main one-peak located at 1303,8 eV was ascribed to the presence of Mg1s in the structure, which was in good agreement with earlier reports.²⁷ The high-resolution XPS spectra of Mn2p show two components (Mn 2p_3/2 at ∼642 eV and Mn 2p_1/2 at ∼653 eV) due to spin–orbit splitting, suggesting that the Mn oxidation states lie between +3 and +4 (Fig. 2c).¹⁷ The peak fitting for Mn 2p_3/2 demonstrated an increase in the atomic content of Mn⁴⁺ from 36.44% in MnO₂ to 49.74% in FLA-MnO₂, while the content of Mn³⁺ decreased from 63.56% in MnO₂ to 50.26% in FLA-MnO₂ due to Mg²⁺ incorporation. The intercalated Mg²⁺ into MnO₂ led to the suppression of the Mn³⁺ concentration, weakening the Jahn–Teller effect due to the higher concentration of Mn⁴⁺.²⁸ These results provide conclusive evidence for the successful incorporation of Mg atoms into the crystal structure of FLA-MnO₂ to partially replace Mn atoms in occupying octahedral sites in the lattice.^29,30 Moreover, as shown in Fig. S6 (ESI),† the O 1s spectra of the two samples deconvolution peaks at 529.8, 531.0 and 532.4 eV are ascribed to the Mn–O–Mn, Mn–O–H, and H–O–H. In comparison with MnO₂, FLA-MnO₂ presents its Mn–O–Mn peak with higher binding energy, manifesting the electron depletion feature of O atoms in Mn–O–Mn.³¹ It can be interpreted by the fact that Mg²⁺ confined in occupying octahedral sites strongly interacts with O atoms, thereby inducing electron transfer from O atoms to Mg²⁺ ions.²⁶ To probe the changes in the electronic structure of FLA-MnO₂, a Raman survey (Fig. 2d) was performed to elucidate specific changes in the [MnO₆] octahedron. The results show that both MnO₂ and FLA-MnO₂ exhibited two peaks. In particular, the peaks of FLA-MnO₂ (MnO₂) at 579.7 (586.7) and 652.4 (635.2) cm⁻¹ represent the flatting vibration of Mn–O in the [MnO₆] octahedron along the d_x²−y² orbital planes and in the d_z² orbital direction. The peak of FLA-MnO₂ at 652 cm⁻¹ is blue-shifted, indicating that the bond length of FLA-MnO₂ in the axial direction becomes shorter. In contrast, the peak of FLA-MnO₂ at 579.7 cm⁻¹ exhibit a significant redshift, suggesting that the bond length of FLA-MnO₂ in the equatorial plane increased. The changes observed in these two peaks indicate that the Mn–O bonds are contracted along the axial direction in the [MnO₆] octahedra, which confirms the flattening of the [MnO₆] octahedral characteristic in FLA-MnO₂.^32,33 In addition, XRD analyses (Fig. 2b) show that MnO₂ exhibits the same lattice constant of 8.03 Å in different orientations, while FLA-MnO₂ shows an average lattice constant of 7.76 Å in the axial direction and 8.83 Å in the other four equatorial planes. Whereas in the face-centred cubic structure, the relationship between the lattice constant and the bond length is in accordance with , the lattice constant and bond length are linearly related, so the bond length increases (decreases) as the lattice constant increases (decreases), which is in good agreement with the Raman results. The characteristic flattened [MnO₆] octahedra with electron redistribution was obtained by incorporating Mg²⁺ in FLA-MnO₂.

3.2. Electrochemical properties of MnO₂ and FLA-MnO₂ materials

To explore the effect of flattened [MnO₆] octahedron characteristics of FLA-MnO₂ on the electrochemical Li⁺ insertion/extraction, the electrochemical performance was tested in 1 M LiCl using a three-electrode system. The cyclic voltammetry (CV) curves of the MnO₂ and FLA-MnO₂ electrodes are shown in Fig. 3a. Two pairs of well-defined redox peaks appeared in the CV curves of these electrodes at scan rates of 2 mV s⁻¹ within a potential window of 0.2–1.2 V, which corresponded to the reversible Li⁺ insertion/extraction processes.³⁴ Compared to MnO₂, FLA-MnO₂ showed significantly improved peak currents, and a slight peak shift occurred. Mg²⁺ doping eliminates Mn 3d orbital degeneracy, the newly accepted electrons during Mn⁴⁺ reinduction prefer to occupy the d_x²−y² orbitals with relatively low energy levels, and the absence of the degenerate ground state avoids the occurrence of J–T distortion, which contributes to the reversibility of lithium insertion/extraction.²⁶ Additionally, FLA-MnO₂ shows its CV curve with an enlarged integral area, indicating higher specific capacitance and Li⁺ storage capacity.³⁵ This result is consistent with that of the GCD curve, where the two charge–discharge platforms correspond to the two pairs of redox peaks in the CV curve, respectively. Notably, the GCD curves of both MnO₂ and FLA-MnO₂ at different current densities (Fig. S7, ESI)† exhibit two pairs of charge–discharge platforms and symmetries, indicating that these electrodes exhibit highly reversible and Faraday behavior.³⁶ Based on the GCD results, the specific capacitance (Fig. 3c) of FLA-MnO₂ was calculated to be 619.13 F g⁻¹ at a current density of 2 A g⁻¹, which is much higher than that of MnO₂ (357.13 F g⁻¹), proving that FLA-MnO₂ has superior capacitance for flattening [MnO₆] octahedra. In addition, both electrodes maintained the profile shapes unchanged in the CV and GCD curves at different scan rates and various current densities. In particular, even at 12 A g⁻¹, a high capacity of 255.6 F g⁻¹ can be achieved for FLA-MnO₂ (Fig. 3c), demonstrating the great rate capability of FLA-MnO₂. Furthermore, there is a linear relationship between the specific current and the voltage drop (IR drop) determined from the GCD curve, which decreases in the order MnO₂ < FLA-MnO₂ (Fig. S9, ESI).† Moreover, we found that the voltage drop increased with increasing current density because it increases the difficulty for ions to enter FLA-MnO₂ and MnO₂, leading to an increase in internal resistance. This discovery demonstrates that the FLA-MnO₂ electrode performs exceptionally well as a capacitive electrode and has a lower internal charge-transfer resistance (Table S2, ESI).† Electrochemical impedance spectroscopy (EIS) (Fig. 3d) was further conducted to determine the charge transfer property in MnO₂ and FLA-MnO₂. The Nyquist plot of FLA-MnO₂ shows a smaller semicircle at a higher frequency than that of MnO₂, indicating facilitated charge transfer behavior. The fitting results in Table S2 (ESI)† indicate that FLA-MnO₂ possesses a lower charge transfer resistance (R_ct) of 0.64 Ω than that for MnO₂ (R_ct = 0.85 Ω), suggesting fast charge transfer behavior on FLA-MnO₂. In addition, the Warburg impedance values (Z_w) of these electrodes derived from the sloped portions in the intermediate- and low-frequency regions decrease in the order FLA-MnO₂ > MnO₂, indicating that FLA-MnO₂ has a more rapid interfacial ion migration rate than MnO₂. Fig.S8 (ESI)† shows the CV curves of the MnO₂ and FLA-MnO₂ electrodes at different scan rates of 2, 5, 10, 15, and 20 mV s⁻¹. Faster scan rates tend to result in higher peak currents and more severe polarization. The relevant diffusion coefficients of Li⁺ (D_Li⁺) in MnO₂ and FLA-MnO₂ were also explored based on the CV results. The calculated Li⁺ diffusion coefficient of FLA-MnO₂ was significantly higher than that of MnO₂ (Table S3, ESI),† highlighting the excellent Li⁺ diffusion transport performance of FLA-MnO₂ featured by flatten [MnO₆] octahedrons. To gain a deep insight into the Li⁺ storage process, the typical theory of Bruce Dunn was introduced, and the kinetic analysis was performed by CV evaluation, as shown in Fig. 3e and S10.† FLA-MnO₂ exhibited b values of 0.667/0.709 and 0.639/0.516, exceeding those of MnO₂ (0.609/0.658 and 0.599/0.484). These b values are typical of the dual-mode electrochemical storage behavior of capacitive contribution and faradaic intercalation, which theoretically facilitates the realization of efficient lithium extraction.³⁷ The b values of the FLA-MnO₂ electrode are more significant than those of the MnO₂ electrodes, suggesting that the FLA-MnO₂ electrode exhibits better ion transport efficiency and rate capability. Meanwhile, the capacitive contribution of FLA-MnO₂ was 46% at 20 mV s⁻¹ in Fig. 3f, higher than that of MnO₂ (40%) in Fig. S11 (ESI),† suggesting a faster electron migration rate and ion transport speed.

Fig. 3 Electrochemical characterizations of the MnO₂ and FLA-MnO₂ electrodes: (a) CV curves at 2 mV s⁻¹ in 1 M LiCl solution. (b) GCD curves at 3 A g⁻¹ in 1 M LiCl solution. (c) Specific capacitances at different specific currents. (d) Nyquist plots of EIS and the fitting curves, inset at the bottom of the figure, show the equivalent circuit. (e) Relationship between peak current and scan rate in the logarithm (FLA-MnO₂). (f) Diffusion-controlled and capacitive-controlled contributions of FLA-MnO₂. (g) Cycling performance for MnO₂ and FLA-MnO₂ at a current density of 10 A g⁻¹. (h) CV curves of the FLA-MnO₂ electrode in 1 M MCl solution (M = Li⁺, Na⁺, K⁺, Ca²⁺, and Mg ²⁺) at a scan rate of 2 mV s⁻¹.

To explore the cycling performance of FLA-MnO₂, both FLA-MnO₂ and MnO₂ were tested at 10 A g⁻¹ for 16 000 cycles. Impressively, FLA-MnO₂ could still achieve a remarkable capacity retention of 58.8% after 16 000 cycles (Fig. 3g), which was much better than that of MnO₂ at 8.4%, indicating the enhanced reversible Li⁺ insertion/extraction performance of FLA-MnO₂ with the flattened [MnO₆] unit. Fig. 3h shows typical cyclic voltammetry (CV) curves for FLA-MnO₂ in the voltage range 0.2–1.2 V at 2 mV s⁻¹ in different electrolytes. When the electrolyte was LiCl, cathodic and anodic peaks corresponding to the reversible Li⁺ insertion/extraction process appeared. However, no significant peaks were observed in electrolytes containing other Na⁺, K⁺, Mg²⁺ and Ca²⁺ cations. These results indicate that compared with other cations, FLA-MnO₂ can selectively react with Li⁺ ions, suggesting the feasibility of using FLA-MnO₂ for the selective extraction of Li⁺ ions. By analyzing the physical and thermodynamic properties of these ions, this phenomenon may be related to the diffusion coefficients, the ionic radii and the Hydration free energy (Table S4, ESI).† Na⁺, K⁺ and Ca²⁺, which have larger ionic radii than Li⁺, may be difficult to embed in FLA-MnO₂ crystals due to spatial effects. On the other hand, the active sites of Li⁺ ions are present in the material lattice, which ensures the reversible insertion/extraction of Li⁺ ions through the “memory effect”. Although the ionic radii of Mg²⁺ are similar to those of Li⁺, its higher hydration-free energy than Li⁺ makes it more energetically difficult to capture.^38–40 Therefore, benefiting from the smaller ionic radii, lower hydration-free energy and relatively larger diffusion coefficient of Li⁺ ions, FLA-MnO₂ exhibits significant selectivity for Li⁺ ions.

3.3. Lithium extraction from the HCDI device

All Li⁺ extraction experiments were conducted in batch mode on an HCDI platform with FLA-MnO₂ (or MnO₂) as the cathode and AC as the anode. Fig. 4a shows the concentration versus time curves of the MnO₂ and FLA-MnO₂ electrodes in 16.37 mg L⁻¹ Li⁺ ions (100 mg L⁻¹ LiCl) solution at 1.2 V. During a charge–discharge cycle, all electrodes exhibit two typical stages corresponding to the insertion and extraction of Li⁺.⁴¹ The FLA-MnO₂ electrode with the flattened [MnO₆] unit exhibits a more excellent Li⁺ adsorption capacity than MnO₂. According to the calculations in eqn (S8),† the FLA-MnO₂ electrode demonstrates a higher adsorption capacity of 17.24 mg g⁻¹ for Li⁺ in 16.37 mg L⁻¹ Li⁺ ion (100 mg L⁻¹ LiCl) solution, which is far superior to that of the MnO₂ electrode (5.61 mg g⁻¹). More evidence indicative of the superiority of the FLA-MnO₂ electrode with the flattened [MnO₆] unit is depicted in the Kim-Yoon plots (Fig. 4b), which reveals the fact that the closer the SAR versus SAC profiles are to the upper-right corner of the Kim-Yoon plot, the better the performance of the electrodes in both SAR and SAC.⁴² The results show that the FLA-MnO₂ electrode can capture most of the Li⁺ ions faster. Fig. 4c shows the current response, charge efficiency and faradaic efficiency of these electrodes in a solution of 16.37 mg L⁻¹ Li⁺ ion at 1.2 V. The FLA-MnO₂ electrode exhibited a higher current level than the MnO₂ electrode. According to eqn (S10),† the charge efficiency of the FLA-MnO₂ electrode (0.85) is much higher than that of MnO₂ (0.51). Flat [MnO₆] octahedra with electron redistribution were obtained by doping Mg²⁺ in MnO₂, and the Faraday efficiency of FLA-MnO₂ for the applied current efficiently used in Li⁺ intercalation was increased from 0.76 to 0.97 from eqn (S11).† The possible reasons for this phenomenon are the considerable voltage drop (IR drop) in the GCD curve of MnO₂ and the large internal resistance of the material itself, which further reduces the charging efficiency of the material. The faster charge transfer rate and lower internal resistance of FLA-MnO₂ prompt more current to be applied to lithium intercalation. These results suggest that the flattened [MnO₆] structure may improve the charge transport behavior, possibly by reducing interfacial resistance or facilitating electron mobility, and that the high Faraday efficiency suggests that a greater proportion of the applied electrical charge is efficiently used for the lithium intercalation process rather than competing side reactions.^26,43

Fig. 4 (a) Concentration versus time plots of HCDI operation with MnO₂ and FLA-MnO₂ electrodes in 16.37 mg L⁻¹ Li⁺ ion solution at 1.2 V. (b) Kim-Yoon plots. (c) Current responses, faradaic efficiency and charge efficiency (inset). (d) Plots of concentration versus time in HCDI cell with the FLA-MnO₂ electrode in 16.37 mg L⁻¹ Li⁺ ion solution at varying cell voltages, and plots of adsorption capacity versus cell voltage (line with symbols). (e) Adsorption capacity of MnO₂ and FLA-MnO₂ electrodes versus feed concentration. (f) Energy consumption of MnO₂ and FLA-MnO₂ electrodes versus feed concentration.

To investigate the extraction performance of FLA-MnO₂ and MnO₂ electrodes for Li⁺ from low-grade brines, we further investigated the effects of cell voltage and Li⁺ concentration on the Li⁺ extraction performance of these electrodes. Fig. 4d and S12† show the Li⁺ insertion/extraction process of various electrodes in the voltage window of 0.6 to 1.2 V in 16.37 mg L⁻¹ Li⁺ ion solution. The cell voltage is proportional to the adsorption capacity at the MnO₂ and FLA-MnO₂ electrodes. Notably, FLA-MnO₂ exhibited excellent cycling reversibility, with the adsorption capacity of the electrode decreasing from 17.24 mg g⁻¹ to 11.41 mg g⁻¹ when the cell voltage decreased from 1.2 V to 0.6 V, and the adsorption capacity recovered to 17.15 mg g⁻¹ immediately when the cell voltage was restored to 1.2 V. However, the electrolysis of water and the formation of a concentration-polarized layer on the working electrode at high voltages (>1.23 V) reduce the charging efficiency of the HCDI system and increase energy consumption.^44,45 Thus, the electrodes exhibited optimal Li⁺ extraction at 1.2 V. Then, we further explored the effect of the electrode on the Li⁺ concentration at 1.2 V voltage (Fig. S13, ESI).† The results showed that the higher the Li⁺ concentration in the feed water, the faster the adsorption and desorption reached equilibrium. The corresponding adsorption capacity also increased with increasing Li⁺ ion concentration (Fig. 4e), which was attributed to the higher ionic strength of the electrolyte in solution, which produced a stronger driving force.⁴⁶ For instance, as the Li⁺ ion concentration increased from 3.27 (20 mg L⁻¹ LiCl) to 32.74 (200 mg L⁻¹ LiCl) mg L⁻¹, the adsorption capacity of the FLA-MnO₂ electrode also increased from 3.52 to 30.14 mg g⁻¹. The high Li⁺ adsorption capacity of FLA-MnO₂ is mainly attributed to the fact that the flattened [MnO₆] unit structure shortens the diffusion path of Li⁺ ions inside the material and facilitates electron transfer and ion migration in the FLA-MnO₂ electrode, which helps to promote the insertion of Li⁺ ions.⁴⁷ Compared with other MnO₂ electrodes in CDI cells reported in the literature, FLA-MnO₂ can achieve higher Li⁺ adsorption capacity at lower concentrations (Table S7, ESI).† These results indicate that FLA-MnO₂ electrodes with flattened [MnO₆] units possess a feasible basis for Li⁺ ion extraction from low-grade brine. Energy consumption is one of the most important indicators for determining the magnitude of economic efficiency in HCDI cell adsorption/desorption processes. The energy consumption of the FLA-MnO₂‖AC cell in different concentrations of Li⁺ ion solution is calculated by eqn (S13)† and is shown in Fig. 4f. In addition to the adsorption amount, the energy consumption increases with the increase of Li⁺ ion concentration. When the Li⁺ ion concentration increases to 16.37 mg L⁻¹, the adsorption capacity of the MnO₂ electrode reached 5.61 mg g⁻¹, with a Li⁺ adsorption ratio of 25.4% and energy consumption of 0.46 Wh g^−1. In contrast, the adsorption amount of the FLA-MnO₂ electrode reached 17.24 mg g⁻¹, with a Li⁺ adsorption ratio of 96.5% and energy consumption of only 0.32 Wh g⁻¹. When the Li⁺ ion concentration continued to increase to 32.74 mg L⁻¹, the Li⁺ adsorption ratio of the FLA-MnO₂ electrode gradually decreased, but the energy consumption continued to rise. This may be due to the limited number of active sites in the material that allowed the adsorption of Li⁺ ions at the FLA-MnO₂ electrode to gradually reach saturation in a high concentration of Li⁺ ion solution. High energy consumption is undesirable in any practical process. Therefore, the FLA-MnO₂ electrode, which produces an energy consumption of only 0.45 Wh g⁻¹ in 32.74 mg L⁻¹ Li⁺ ion solution, has great potential for application in Li⁺ extraction from low-grade brines.

3.4. Stability of Li⁺ ion recovery system

Cyclic stability is another important indicator for evaluating the economic benefits of electrodes during long-term insertion/extraction of Li⁺ ions. In this experiment, the solution was placed at 1.2 V for 40 min (insertion phase), and then the voltage was reversed at 1.2 V for 40 min (regeneration phase). Fig. 5a and S14† compare the Li⁺ adsorption capacity and capacity retention rate of FLA-MnO₂‖AC cells compared with MnO₂‖AC cells over 100 cycles at 1.2 V in a 16.37 mg L⁻¹ Li⁺ ion solution. Fig. S15† shows the current profiles of the HCDI system with MnO₂ and FLA-MnO₂ electrodes during Li ion insertion and extraction. It was observed that the current readings (during charging) of the FLA-MnO2‖AC HCDI system were higher than those of the MnO₂‖AC system and that the FLA-MnO₂‖AC HCDI system showed faster ion insertion. As expected, the applied potential during the Li⁺ ion insertion phase reduces the concentration of the solution, leading to an increase in current due to ion migration caused by electrode polarization. The maximum current peak is driven by electrode polarization, and the ion removal peak lags because of the resistance of the non-ideal electrolyte.⁴⁸ At the start of the regeneration cycle at reverse voltage, there is a large increase in the solution concentration (ion desorption) and an increase in the negative current associated with ion desorption. At the end of the multiple regeneration phases of the MnO₂‖AC system, a gradual decrease in conductivity below the initial concentration value and a slight decrease were observed, suggesting that the application of the counter-voltage may have induced some ion crossover. This phenomenon is related to the adsorption of substances previously present on the counter electrodes and the re-adsorption of ions present in the diffusion layer.⁴⁹ After 100 cycles, the adsorption capacity of the MnO₂‖AC cell gradually decreases from the initial 6.23 mg g⁻¹ to the final 3.13 mg g⁻¹, with a capacity retention rate of 50%. In contrast, the structurally flattened [MnO₆] conformation avoids the formation of degenerate electronic ground state formation during the electrochemical process, suppressing the J–T distortion and effectively improving the electrochemical reversibility. The FLA-MnO₂‖AC cell exhibited a gradual decrease from 16.61 to 13.56 mg g⁻¹, with a capacity retention rate of 82%. Fig. S16† also provides the electrosorption/desorption kinetics of MnO₂ and FLA-MnO₂ at the 1st, 50th, and 100th cycles. FLA-MnO₂ with faster electrosorption/desorption kinetics with an increasing number of cycles can be attributed to the fact that the structurally flattened [MnO₆] conformation avoids the formation of a condensed electronic ground state during the electrochemical process, suppresses the J–T distortion, and effectively improves the electrochemical reversibility so that the ionic repulsion is minimized and most of the charge is provided to the electrode polarization. By analyzing the regeneration step, the reversed voltage resulted in the desorption of essentially all adsorbed ions from MnO₂ and FLA-MnO₂ in the removal phase, and even though Mn solvation loss and J–T distortion during continuous lithium ion insertion/extraction resulted in a decrease in lithium adsorption capacity, the rate of desorption remained comparable to the rate of adsorption as the number of cycles increased and was therefore kinetically favorable and fully reversible. To further evaluate the structural stability and reliability of the FLA-MnO₂ electrode after 100 cycles, the XRD, SEM and Raman data of the post-cycling electrode were collected and compared with those of the pristine FLA-MnO₂ (Fig. 5b–d). The XRD diffraction signals of FLA-MnO₂ after cycling experiments were consistent, indicating that the material did not undergo a phase transition after 100 high-intensity cycles.⁵⁰ The FLA-MnO₂ electrode maintains a high crystallinity, the characteristic peak is slightly shifted to a slight angle after 100 cycles, and the half-peak width in the (111) plane exhibits a minimal change (0.189°), as shown by the enlarged (111) peak in Fig. 5b. This can be explained by incompletely de-embedded lithium in the electrode. The lithium ions contained in the lattice contributed to longer bonds between the O atoms and larger lattice constants, resulting in a smaller angle of shift of the XRD peak position.⁵¹ Also, the SEM results (Fig. 5c) clearly showed that the FLA-MnO₂ electrode was sufficiently stable and no significant change in morphologies during the stability test. The Raman results (Fig. 5d) after 100 absorption/desorption cycles showed that the peak of FLA-MnO₂ at 652.4 cm⁻¹ was slightly red-shifted (647.2), indicating that the bond length of FLA-MnO₂ was slightly lengthened in the axial direction. The peak of FLA-MnO₂ at 579.7 cm⁻¹ shows a slight blue shift (585.4), indicating that the bond length of FLA-MnO₂ was slightly shorter in the equatorial plane. However, compared with MnO₂ (635.2/586.7), it also exhibits a flattened structure and stable cycling ability. In addition, the Mn dissolution loss of the MnO₂ electrode was 3.6%, whereas that of the FLA-MnO₂ electrode was only 1.3% (Fig. 5e).

Fig. 5 (a) Cycling stability and adsorption retention rate of MnO₂ and FLA-MnO₂ electrodes in 16.37 mg L⁻¹ Li⁺ ion solution applied in HCDI for 100 charge–discharge runs. (b) XRD spectra of FLA-MnO₂ before and after 100 cycles. (c) SEM pattern of FLA-MnO₂ after 100 cycles. (d) Raman spectra of FLA-MnO₂ before and after 100 cycles. (e) Mn dissolution loss ratio of MnO₂ and FLA-MnO₂ electrodes after 100 cycles. (f) Effluent pH for the FLA-MnO₂‖AC cell during 100 cycles.

In summary, the flattened [MnO₆] unit structure significantly improved the cycling stability of HCDI cells. In addition, we simultaneously monitored the pH and the faradaic efficiency change of the FLA-MnO₂‖AC cell during the whole cycling test. The pH of the effluent water varied in the range of about 1 and stabilized at ≈5.0–6.0 (Fig. 5f), indicating that although the electrode undergoes a side reaction of hydrolysis dissociation to make the pH change in the process of inserting/extracting Li⁺ ions, this side reaction is also still confined to a very small range.¹⁸ The Faraday efficiency values for each adsorption effluent were also close to 1 and stabilized at ≈0.95–1.0 (Fig. S17†). The results show that the applied current is efficiently used for Li⁺ intercalation rather than lost to side reactions, which also ensures that 1.2 V is a safe operating voltage for a two-electrode HCDI cell, which remains within the electrochemical stabilization window for water even after 100 cycles. The results show that FLA-MnO₂ electrodes with flattened [MnO₆] units are an effective method to improve the stability of Mn-based Li⁺ ion sieves with good capacity retention, stable crystal structure, and low Mn solubility loss, making it a promising method for long-term low-grade brine Li⁺ extraction.

3.5. Selective extraction of Li⁺ ions generated by FLA-MnO₂‖AC HCDI cell

Among the many competing ions, Mg²⁺ ions have ionic radii similar to those of Li⁺ and impede the selective extraction of Li⁺ ions.¹³ Therefore, we investigated the selective extraction behavior of Li⁺ ions from FLA-MnO₂‖AC HCDI cells at different Mg/Li molar ratios. As shown in Fig. 6a, the different molar ratios of Mg/Li mixed solutions were set as 1 : 1, 2 : 1, 5 : 1, 10 : 1, and 30 : 1, where “1” denotes the concentration of 5 mM ions. The results showed that the FLA-MnO₂‖AC HCDI cell exhibited excellent Li⁺ selective adsorption capacity (adsorption capacities of ≈1.49 and ≈7.65 mg g⁻¹ for Mg²⁺ and Li⁺, respectively) at the same concentration of the mixed solution (1 : 1). However, the adsorption capacity of the FLA-MnO₂‖AC for Mg²⁺ ions increased with increasing Mg²⁺ ion concentration and the adsorption capacity of Li⁺ decreased, indicating that higher concentrations of Mg²⁺ ions interfered with the selective extraction of Li⁺ ions. At the highest molar ratio of 30 : 1, the lithium selective extraction capacity was suppressed to ≈2.64 mg g⁻¹. In general, the separation factor is a key metric that can be intuitively determined for selective Li⁺ extraction from FLA-MnO₂‖AC. As shown in Fig. 6a, increased from 4.83 to 7.66 when the molar ratio was increased from 1 : 1 to 5 : 1. However, a decreasing trend of the separation factor was observed as the concentration of Mg²⁺ in the mixed solution continued to increase, and the separation factor dropped to 3.59 at a molar ratio of 30 : 1, but it was always far exceeded 1, still revealing the great potential in the selective extraction of Li⁺ ions from brine with a range of Mg/Li ratios. This excellent lithium separation factor can be attributed to the special recognition of Li⁺ by the imprinted vacancies in spinel-type manganese oxides. This allows Li⁺, which has a smaller ionic radius and hydration-free energy, to be more easily intercalated in the 3D network structure of Mn₂O₄ than Mg²⁺.⁵² The experiment was repeated five times to further explore the potential application and reproducibility of the FLA-MnO₂‖AC HCDI cell for selective lithium ion recovery using Lop Nor, the low-grade original brine of the XieLi salt flats (Xinjiang, China), as shown in Fig. 6b, c, where the ionic concentrations of Li⁺, Na⁺, K⁺, Ca²⁺, and Mg²⁺ in the low-grade original brine are shown in Table S5, ESI.† The concentrations and adsorption ratios of different ions before and after adsorption are shown in Fig. 6b. It is noteworthy that the adsorption ratio of Li⁺ at the electrode reaches 55.4% even though the concentration of interfering ions (95 366.5 mg L⁻¹ for Na⁺, 10 101.3 mg L⁻¹ for K⁺, 29 403.1 mg L⁻¹ for Mg²⁺) is much higher than that of Li⁺ ions (42.1 mg L⁻¹), while the adsorption ratio of the interfering ions is still very low (12.6% for Na⁺, 14.8% for K⁺, 17.1% for Ca²⁺, and 26.8% for Mg²⁺). This is due to the steric effect hindrance that makes it more difficult for other interfering ions with larger radii (Na⁺, K⁺ and Ca²⁺) to enter the lattice than Li⁺ ions. As shown in Table S8,† the results of the simulation calculations for the hydration energies of different ions indicate that although the radius of Mg²⁺ is similar to that of Li⁺, the higher hydration energy of Mg²⁺ (−1199.9 kJ mol⁻¹) determines the higher energy required for its dehydration (Li⁺ −424.1 kJ mol⁻¹), making its intercalation kinetics of Mg²⁺ difficult.⁵³ As shown in Fig. 6c, the adsorption capacity of Li⁺ ions in low-grade original brine was 34.96 mg g⁻¹. However, due to the high concentration of interfering ions in the brine, the adsorption (18 117.3 mg g⁻¹ for Na⁺, 2236.3 mg g⁻¹ for K⁺, 14.7 mg g⁻¹ for Ca²⁺, 11 820.5 mg g⁻¹ for Mg²⁺) of interfering ions by the FLA-MnO₂‖AC HCDI cell was much higher than that of Li⁺. However, even with such a high concentration gap, the calculated separation factors for Li⁺/Na⁺, Li⁺/K⁺, Li⁺/Ca²⁺, and Li⁺/Mg²⁺ reached impressive values of 4.37, 3.75, 3.24, and 2.11, respectively. The experiment was repeated five times and yielded satisfactory results (as indicated by the error bars). Therefore, experiments in low-grade original brine further demonstrate that FLA-MnO₂ electrodes exhibit excellent Li⁺ selective adsorption capacity in a complex coexisting cation environment, and the low energy consumption and high cycling stability exhibited by this electrode predicted the potential for long-term Li⁺ extraction from low-grade original water.

Fig. 6 (a) Li⁺ adsorption capacity and values in aqueous solutions with different Mg/Li ratios. (C_Li⁺ = 5 mM, Mg/Li ratios were 1, 2, 5, 10, and 30). (b) Adsorption ratio and concentration before and after adsorption of Li⁺, Na⁺, K⁺, Ca²⁺, and Mg²⁺ by the FLA-MnO₂ electrode in Lop Nor, the low-grade original brine of the XieLi salt flats. (c) Adsorption capacity and separation factor of Li⁺, Na⁺, K⁺, Ca²⁺, and Mg²⁺ by the FLA-MnO₂ electrode in Lop Nor the XieLi salt flats low-grade original brine.

4. Conclusions

In this work, we successfully designed and prepared a novel FLA-MnO₂ containing flattened [MnO₆] octahedra. This flattened crystalline unit avoids the formation of a degenerate electronic ground state during cycling, thus mitigating the J–T distortion and contributing to the enhanced electrochemical Li⁺ insertion/extraction reversibility. FLA-MnO₂ has a high specific capacitance of 619 F g⁻¹ and cycling stability, with capacity retention of 58.8% after 16 000 cycles compared to 8.4% for MnO₂. When used for hybrid capacitive deionization (HCDI), the FLA-MnO₂‖AC cell achieved a Li⁺ adsorption capacity of 30.14 mg g⁻¹ and low energy consumption of 0.45 Wh g⁻¹ Li⁺ in a 32.74 mg L⁻¹ Li⁺ ion solution. The FLA-MnO₂‖AC HCDI cell exhibited high stability with 82% capacity retention and a low Mn loss of 1.3% over 100 cycles. The FLA-MnO₂‖AC cell achieved excellent lithium selectivity in synthetic brine with a separation factor ≈7.66 at a Mg/Li molar ratio of 5. In addition, the FLA-MnO₂‖AC cell maintained high Li⁺ adsorption capacity with good selectivity in Lop Nor, the low-grade original brine of the XieLi salt flats. The calculated separation factors for Li⁺/Na⁺, Li⁺/K⁺, Li⁺/Ca²⁺, and Li⁺/Mg²⁺ reached impressive values of 4.37, 3.75, 3.24, and 2.11. Therefore, the FLA-MnO₂ electrode with a flattened structure has the advantages of high adsorption capacity, good cyclic stability, good selectivity, and low energy consumption, providing a useful direction and solution to solve the poor stability of manganese-based Li⁺ ion sieves for the efficient extraction of lithium from brine.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 52422406) and the Natural Science Foundation of Hubei Province (No. 2022CFA064).

References

1. U. Wietelmann and R. J. Bauer, Lithium and Lithium Compounds, Ullmann’s Encycl. Ind. Chem., 2000 DOI:10.1002/14356007.a15_393.
2. P. Greim, A. A. Solomon and C. Breyer, Assessment of lithium criticality in the global energy transition and addressing policy gaps in transportation, Nat. Commun., 2020, 11, 1–11,  DOI:10.1038/s41467-020-18402-y.
3. G. Martin, L. Rentsch, M. Höck and M. Bertau, Lithium market research - global supply, future demand and price development, Energy Storage Mater., 2017, 6, 171–179,  DOI:10.1016/j.ensm.2016.11.004.
4. J. Zhang, Z. Cheng, X. Qin, X. Gao, M. Wang and X. Xiang, Recent advances in lithium extraction from salt lake brine using coupled and tandem technologies, Desalination, 2023, 547, 116225,  DOI:10.1016/j.desal.2022.116225.
5. E. Zhang, W. Liu, Q. Liang, X. Liu, Z. Zhao and Y. Yang, Selective recovery of Li⁺ in acidic environment based on one novel electroactive Li⁺-imprinted graphene-based hybrid aerogel, Chem. Eng. J., 2020, 385, 123948,  DOI:10.1016/j.cej.2019.123948.
6. R. Zhu, S. Wang, C. S. S. Li, S. Yin, L. Zhang and X. Jiang, Lithium extraction from salt lake brines with high magnesium/lithium ratio : a review, Environ. Chem. Lett., 2023, 21, 1611–1626,  DOI:10.1007/s10311-023-01571-9.
7. Y. Xiong, J. Zhou, P. Lu, J. Yin, Y. Wang and Z. Fan, Electrochemical lithium extraction from aqueous sources, Matter, 2022, 5, 1760–1791,  DOI:10.1016/j.matt.2022.04.034.
8. X. Zhao, S. Yang, Y. Hou, H. Gao, Y. Wang, D. A. Gribble and V. G. Pol, Recent progress on key materials and technical approaches for electrochemical lithium extraction processes, Desalination, 2023, 546, 116189,  DOI:10.1016/j.desal.2022.116189.
9. Y. Mu, C. Zhang, W. Zhang and Y. Wang, Electrochemical lithium recovery from brine with high Mg²⁺/Li⁺ ratio using mesoporous λ-MnO₂/LiMn₂O₄ modified 3D graphite felt electrodes, Desalination, 2021, 511, 115112,  DOI:10.1016/j.desal.2021.115112.
10. R. Chitrakar, H. Kanoh, Y. Miyai and K. Ooi, Recovery of lithium from seawater using manganese oxide adsorbent (H_1.6Mn_1.6O₄) derived from Li_1.6Mn_1.6O₄, Ind. Eng. Chem. Res., 2001, 40, 2054–2058,  DOI:10.1021/ie000911h.
11. Z. Cai, Y. Ma, X. Huang, X. Yan, Z. Yu, S. Zhang, G. Song, Y. Xu, C. Wen and W. Yang, High electrochemical stability Al-doped spinel LiMn₂O₄ cathode material for Li-ion batteries, J. Energy Storage, 2020, 27, 101036,  DOI:10.1016/j.est.2019.101036.
12. X. Zhao, G. Li, M. Feng and Y. Wang, Semi-continuous electrochemical extraction of lithium from brine using CF-NMMO/AC asymmetric hybrid capacitors, Electrochim. Acta, 2020, 331, 135285,  DOI:10.1016/j.electacta.2019.135285.
13. X. Shang, J. Liu, B. Hu, P. Nie, J. Yang, B. Zhang, Y. Wang, F. Zhan and J. Qiu, CNT-Strung LiMn₂O₄ for Lithium Extraction with High Selectivity and Stability, Small Methods, 2022, 6(7), 2200508,  DOI:10.1002/smtd.202200508.
14. H. Kanoh, K. Ooi, Y. Miyai and S. Katoh, Electrochemical recovery of lithium ions in the aqueous phase, Sep. Sci. Technol., 1993, 28, 643–651,  DOI:10.1080/01496399308019512.
15. B. Hu, B. Zhang, Y. Wang, M. Li, J. Yang and J. Liu, Prussian blue analogue derived 3D hollow LiCoMnO₄ nanocube for selective extraction of lithium by pseudo-capacitive deionization, Desalination, 2023, 560, 116662,  DOI:10.1016/j.desal.2023.116662.
16. X. Zhao, M. Feng, Y. Jiao, Y. Zhang, Y. Wang and Z. Sha, Lithium extraction from brine in an ionic selective desalination battery, Desalination, 2020, 481, 114360,  DOI:10.1016/j.desal.2020.114360.
17. G. Tan, S. Wan, J. J. Chen, H. Q. Yu and Y. Yu, Reduced Lattice Constant in Al-Doped LiMn₂O₄ Nanoparticles for Boosted Electrochemical Lithium Extraction, Adv. Mater., 2024, 36, 1–10,  DOI:10.1002/adma.202310657.
18. K. Ooi, Y. Miyai and J. Sakakihara, Mechanism of Li Insertion in Spinel-Type Manganese Oxide. Redox and Ion-Exchange Reactions, Langmuir, 1991, 7, 1167–1171,  DOI:10.1021/la00054a025.
19. H. J. Hong, I. S. Park, T. Ryu, J. Ryu, B. G. Kim and K. S. Chung, Granulation of Li_1.33Mn_1.67O₄ (LMO) through the use of cross-linked chitosan for the effective recovery of Li⁺ from seawater, Chem. Eng. J., 2013, 234, 16–22,  DOI:10.1016/j.cej.2013.08.060.
20. K. Zhang, D. Kim, Z. Hu, M. Park, G. Noh, Y. Yang, J. Zhang, V. W. hei Lau, S. L. Chou, M. Cho, S. Y. Choi and Y. M. Kang, Manganese based layered oxides with modulated electronic and thermodynamic properties for sodium ion batteries, Nat. Commun., 2019, 10, 1–12,  DOI:10.1038/s41467-018-07646-4.
21. U. Maitra, R. A. House, J. W. Somerville, N. Tapia-Ruiz, J. G. Lozano, N. Guerrini, R. Hao, K. Luo, L. Jin, M. A. Pérez-Osorio, F. Massel, D. M. Pickup, S. Ramos, X. Lu, D. E. McNally, A. V. Chadwick, F. Giustino, T. Schmitt, L. C. Duda, M. R. Roberts and P. G. Bruce, Oxygen redox chemistry without excess alkali-metal ions in Na_2/3[Mg_0.28Mn_0.72]O₂, Nat. Chem., 2018, 10, 288–295,  DOI:10.1038/nchem.2923.
22. Z. Xiao, F. Xia, L. Xu, X. Wang, J. Meng, H. Wang, X. Zhang, L. Geng, J. Wu and L. Mai, Suppressing the Jahn–Teller Effect in Mn-Based Layered Oxide Cathode toward Long-Life Potassium-Ion Batteries, Adv. Funct. Mater., 2022, 32, 1–9,  DOI:10.1002/adfm.202108244.
23. X. Zhu, F. Meng, Q. Zhang, L. Xue, H. Zhu, S. Lan, Q. Liu, J. Zhao, Y. Zhuang, Q. Guo, B. Liu, L. Gu, X. Lu, Y. Ren and H. Xia, LiMnO₂ cathode stabilized by interfacial orbital ordering for sustainable lithium-ion batteries, Nat. Sustain., 2021, 4, 392–401,  DOI:10.1038/s41893-020-00660-9.
24. H. Yaghoobnejad Asl and A. Manthiram, Proton-Induced Disproportionation of Jahn–Teller-Active Transition-Metal Ions in Oxides Due to Electronically Driven Lattice Instability, J. Am. Chem. Soc., 2020, 142, 21122–21130,  DOI:10.1021/jacs.0c10044.
25. K. Lu, Z. Hu, Z. Xiang, J. Ma, B. Song, J. Zhang and H. Ma, Cation Intercalation in Manganese Oxide Nanosheets: Effects on Lithium and Sodium Storage, Angew. Chem., Int. Ed., 2016, 55, 10448–10452,  DOI:10.1002/anie.201605102.
26. Z. Hou, J. Wang, N. Dai, S. Yao, S. Wang, Y. Ji, X. Gao, H. Zhang, Z. Tang, Y. Sun, S. Li, Y. Liu, W. Fu, K. Nie, Y. Jiang, Y. M. Yan and Z. Yang, Eliminating the Mn 3d Orbital Degeneracy to Suppress the Jahn–Teller Distortion for Stable MnO₂ Cathode, Adv. Energy Mater., 2024, 14, 1–8,  DOI:10.1002/aenm.202302477.
27. B. Tekin and R. Demir-Cakan, Understanding the role of water-based electrolytes on magnesium-ion insertion/extraction into λ-MnO₂ lattice structure, Solid State Ionics, 2019, 335, 67–73,  DOI:10.1016/j.ssi.2019.02.015.
28. Z. Cai, Y. Ma, X. Huang, X. Yan, Z. Yu, S. Zhang, G. Song, Y. Xu, C. Wen and W. Yang, High electrochemical stability Al-doped spinel LiMn₂O₄ cathode material for Li-ion batteries, J. Energy Storage, 2020, 27, 101036,  DOI:10.1016/j.est.2019.101036.
29. Q. Li, C. Wang, Y. Zhu, W. Du, W. Liu, M. Yao, Y. Wang, Y. Qian and S. Feng, Unlocking the critical role of Mg doping in α-MnO₂ cathode for aqueous zinc ion batteries, Chem. Eng. J., 2024, 485, 150077,  DOI:10.1016/j.cej.2024.150077.
30. Y. Zhao, B. Wang, M. Shi, S. An, L. Zhao and C. Yan, Mg-intercalation engineering of MnO₂ electrode for high-performance aqueous magnesium-ion batteries, Int. J. Miner. , Metall. Mater., 2022, 29, 1954–1962.
31. S. Yao, R. Zhao, S. Wang, Y. Zhou, R. Liu, L. Hu, A. Zhang, R. Yang, X. Liu, Z. Fu, D. Wang, Z. Yang and Y. M. Yan, Ni-doping induced structure distortion of MnO₂ for highly efficient Na⁺ storage, Chem. Eng. J., 2022, 429, 132521,  DOI:10.1016/j.cej.2021.132521.
32. L. Yan, C. Shen, L. Niu, M. cheng Liu, J. Lin, T. Chen, Y. Gong, C. Li, X. Liu and S. Xu, Experimental and Theoretical Investigation of the Effect of Oxygen Vacancies on the Electronic Structure and Pseudocapacitance of MnO₂, ChemSusChem, 2019, 12, 3571–3581,  DOI:10.1002/cssc.201901015.
33. L. Liu, L. Su, Y. Lu, Q. Zhang, L. Zhang, S. Lei, S. Shi, M. D. Levi and X. Yan, The Origin of Electrochemical Actuation of MnO₂/Ni Bilayer Film Derived by Redox Pseudocapacitive Process, Adv. Funct. Mater., 2019, 29, 1–12,  DOI:10.1002/adfm.201806778.
34. N. Xie, Y. Li, Y. Yuan, J. Gong and X. Hu, Fabricating a Flow-Through Hybrid Capacitive Deionization Cell for Selective Recovery of Lithium Ions, ACS Appl. Energy Mater., 2021, 4, 13036–13043,  DOI:10.1021/acsaem.1c02654.
35. Y. Bao, J. Hao, S. Zhang, D. Zhu and F. Li, Structural/Compositional-Tailoring of Nickel Hexacyanoferrate Electrodes for Highly Efficient Capacitive Deionization, Small, 2023, 19, 1–11,  DOI:10.1002/smll.202300384.
36. A. Gong, Y. Zhao, M. He, B. Liang and K. Li, High-performance desalination of three-dimensional nitrogen-doped carbon framework reinforced Prussian blue in capacitive deionization, Desalination, 2021, 505, 114997,  DOI:10.1016/j.desal.2021.114997.
37. Y. Liu, X. Gao, K. Wang, X. Dou, H. Zhu, X. Yuan and L. Pan, Rocking-chair capacitive deionization with flow-through electrodes, J. Mater. Chem. A, 2020, 8, 8476–8484,  10.1039/c9ta14112j.
38. N. Xie, Y. Li, Y. Lu, J. Gong and X. Hu, Electrochemically Controlled Reversible Lithium Capture and Release Enabled by LiMn₂O₄ Nanorods, ChemElectroChem, 2020, 7, 105–111,  DOI:10.1002/celc.201901728.
39. J. Lee, S. H. Yu, C. Kim, Y. E. Sung and J. Yoon, Highly selective lithium recovery from brine using a λ-MnO₂-Ag battery, Phys. Chem. Chem. Phys., 2013, 15, 7690–7695,  10.1039/c3cp50919b.
40. G. Ma, Y. Xu, A. Cai, H. Mao, X. Zhang, D. M. Shin, L. Wang and H. Zhou, Binder-Free LiMn₂O₄ Nanosheets on Carbon Cloth for Selective Lithium Extraction from Brine via Capacitive Deionization, Small, 2024, 20, 1–12,  DOI:10.1002/smll.202306530.
41. X. Zhao, H. Yang, Y. Wang, L. Yang and L. Zhu, Lithium extraction from brine by an asymmetric hybrid capacitor composed of heterostructured lithium-rich cathode and nano-bismuth anode, Sep. Purif. Technol., 2021, 274, 119078,  DOI:10.1016/j.seppur.2021.119078.
42. T. Kim and J. Yoon, CDI ragone plot as a functional tool to evaluate desalination performance in capacitive deionization, RSC Adv., 2015, 5, 1456–1461,  10.1039/c4ra11257a.
43. X. He, W. Chen, F. Sun, Z. Jiang, B. Li, X. Li and L. Lin, Enhanced NH⁴⁺ Removal and Recovery from Wastewater Using Na- Zeolite-based Flow-Electrode Capacitive Deionization : Insight from Ion Transport Flux, Environ. Sci. Technol., 2023, 57(23), 8828–8838,  DOI:10.1021/acs.est.3c02286.
44. J. Jin, M. Li, M. Tang, Y. Li, Y. Liu, H. Cao and F. Li, Phase- and Crystallinity-Tailorable MnO₂ as an Electrode for Highly E ffi cient Hybrid Capacitive Deionization (HCDI), ACS Sustain. Chem. Eng. J., 2020, 10,  DOI:10.1021/acssuschemeng.0c04101.
45. M. Wang, X. Xu, Y. Li, T. Lu and L. Pan, Enhanced desalination performance of anion-exchange membrane capacitive deionization via effectively utilizing cathode oxidation, Desalination, 2018, 443, 221–227,  DOI:10.1016/j.desal.2018.06.002.
46. B. Hu, X. Shang, P. Nie, B. Zhang, J. Yang and J. Liu, Lithium ion sieve modified three-dimensional graphene electrode for selective extraction of lithium by capacitive deionization, J. Colloid Interface Sci., 2022, 612, 392–400,  DOI:10.1016/j.jcis.2021.12.181.
47. C. Zhang, X. Zhan, T. Al-Zoubi, Y. Ma, P. C. Shih, F. Wang, W. Chen, S. Pidaparthy, R. M. Stephens, Q. Chen, J. M. Zuo and H. Yang, Electrochemical generation of birnessite MnO₂ nanoflowers for intercalation of Mg²⁺ ions, Nano Energy, 2022, 102, 107696,  DOI:10.1016/j.nanoen.2022.107696.
48. J. J. Lado, R. E. Pérez-Roa, J. J. Wouters, M. I. Tejedor-Tejedor, C. Federspill and M. A. Anderson, Continuous cycling of an asymmetric capacitive deionization system: an evaluation of the electrode performance and stability, J. Environ. Chem. Eng., 2015, 3, 2358–2367,  DOI:10.1016/j.jece.2015.08.025.
49. J. J. Wouters, J. J. Lado, M. I. Tejedor-Tejedor, R. Perez-Roa and M. A. Anderson, Carbon fiber sheets coated with thin-films of SiO₂ and γ-Al₂O₃ as electrodes in capacitive deionization: Relationship between properties of the oxide films and electrode performance, Electrochim. Acta, 2013, 112, 763–773,  DOI:10.1016/j.electacta.2013.08.170.
50. Q. Qu, L. Fu, X. Zhan, D. Samuelis, J. Maier, L. Li, S. Tian, Z. Li and Y. Wu, Porous LiMn₂O₄ as cathode material with high power and excellent cycling for aqueous rechargeable lithium batteries, Energy Environ. Sci., 2011, 4, 3985–3990,  10.1039/c0ee00673d.
51. M.K. and T. H. Tsutomu Ohzuku, Electrochemistry of Manganese Dioxide in Lithium Nonaqueous Cell: III . X-Ray Diffractional Study on the Reduction of Spinel-Related Manganese Dioxide, J. Electrochem. Soc., 1990, 137, 3–9,  DOI:10.1149/1.2086552.
52. S. Kim, J. S. Kang, H. Joo, Y. E. Sung and J. Yoon, Understanding the Behaviors of λ-MnO₂ in Electrochemical Lithium Recovery: Key Limiting Factors and a Route to the Enhanced Performance, Environ. Sci. Technol., 2020, 54, 9044–9051,  DOI:10.1021/acs.est.9b07646.
53. J. G. Gamaethiralalage, K. Singh, S. Sahin, J. Yoon, M. Elimelech, M. E. Suss, P. Liang, P. M. Biesheuvel, R. L. Zornitta and L. C. P. M. De Smet, Recent advances in ion selectivity with capacitive deionization, Energy Environ. Sci., 2021, 14, 1095–1120,  10.1039/d0ee03145c.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta01941a

‡ These authors contribute equally to this work.

---