Water‐facilitated targeted repair of degraded cathodes for sustainable lithium‐ion batteries

Directly repairing end‐of‐life lithium‐ion battery cathodes poses significant challenges due to the diverse compositions of the wastes. Here, we propose a water‐facilitated targeted repair strategy applicable to various end‐of‐life batches and cathodes. The process involves initiating structural repair and reconstructing particle morphology in degraded LiMn2O4 (LMO) through an additional thermal drive post‐ambient water remanganization, achieving elemental repair. Compared to solid‐phase repair, the resulting LMO material exhibits superior electrochemical and kinetic characteristics. The theoretical analysis highlights the impact of Mn defects on the structural stability and electron transfer rate of degraded materials. The propensity of Mn ions to diffuse within the Mn layer, specifically occupying the Mn 16d site instead of the Li 8a site, theoretically supports the feasibility of ambient water remanganization. Moreover, this method proves effective in the relithiation of degraded layered cathode materials, yielding single crystals. By combining low energy consumption, environmental friendliness, and recyclability, our study proposes a sustainable approach to utilizing spent batteries. This strategy holds the potential to enable the industrial direct repair of deteriorated cathode materials.


INTRODUCTION
Lithium-ion batteries (LIBs) are increasingly common as they approach the end of useful lives, raising concerns about environmental impact and the raw material supply risks. 1 Every component within LIBs has the potential to produce pollutants through chemical interactions with environmental substances, resulting in significant contamination. 2 Furthermore, LIBs contain significant amounts of lithium, cobalt, nickel, copper, aluminum, and manganese, with enrichment effects, and their accumulation in the food chain poses risks to human health. 3The International Council on Clean Transportation estimates that new energy vehicles (EVs) emit approximately 130 g CO 2 per kilometer over their entire lifecycle 4 Recycling and reapplying end-of-life LIBs could reduce associated carbon emissions by 4 g CO 2 per kilometer.This reduction would have a substantial impact on carbon emissions not only during the battery manufacturing process but also throughout the entire lifecycle of EVs.To safeguard human health, promote effective energy conservation, and emission reduction, and support environmentally sustainable development, it is imperative to adopt a centralized and secure approach for the treatment of end-of-life LIBs.
6][7] However, this method demands significant energy consumption 8 and/or intricate hydro-chemistry processes, 9 rendering the end products economically less viable. 10Consequently, research has shifted toward direct repair techniques with lower energy consumption and cost inputs, 11 wherein the degraded cathode material undergoes complete relithiation 12 or remanganization, 13 accompanied by the rearrangement of the degraded crystal structure. 14Key techniques in this regard include the solid-phase method, 15 molten-salt method, 16 and hydrothermal annealing repair method. 17The solid-phase method entails the quantitative addition of a single lithium salt, with the material's structural transformation achieved through heat treatment in a specific atmosphere. 18Molten salt restoration typically involves an excess of mixed lithium salts, sintered at high temperatures to effect structural change after being heated for a period at low temperatures to establish a homogeneous molten salt system. 160][21] However, the charge state, level of metal element deficiency, and crystalline structure vary depending on the aging state of the wastes. 20,22The limited scope of direct repair technology lies in the inhomogeneity of the wastes, hindering the repaired material from promptly regaining the appropriate stoichiometric composition and optimum crystal structure.While a precise Li/TM ratio is not critical for hydrothermal relithiation under aqueous conditions, the repair process is timeconsuming and presents safety issues in a high-pressure environment. 17,23,24Additionally, distinct recycling techniques are necessary for different cathode systems due to variations in physical/chemical features, elemental composition, and lithium storage properties. 25Therefore, it is essential to develop a universally applicable and safe repair method that does not require predetermined recyclables.
Herein, we propose a water-facilitated targeted repair strategy applicable to various end-of-life batches and cathodes.The incorporation of Mn 2+ into the vacancies of degraded LiMn 2 O 4 (LMO) material is achieved through ambient water remanganization, thereby completely restoring the composition of LMO.Subsequent heat treatment induces the oxidation of Mn 2+ and Mn 3+ to Mn 4+ , resulting in the restoration of the crystal structure and microscopic morphology.The electrochemical and kinetic properties of the repaired LMO obtained through this strategy surpass those repaired by the solidphase repair method.This repair method can also be applied to highly deteriorated transition metal oxidebased cathodes, such as LiCoO 2 , LiNi 0.5 Co 0.2 Mn 0.3 O 2 , and LiNi 0.6 Co 0.2 Mn 0.2 O 2 , facilitating their relithiation.Notably, the ambient water-facilitated repair strategy proves to be significantly cost-effective and competitive and introduces a fresh perspective on the short-range repair of spent LIBs.

Ambient water-facilitated remanganization
The degradation of LMO materials has been established to be primarily caused by Mn deficiency due to the dissolution of Mn. 13,26 Therefore, it is imperative to precisely address the Mn vacancy for the repair of spent LMO (S-LMO) powder.Weak electrolyte manganese acetate (Mn(CH 3 COO) 2 ) can undergo dissociation in aqueous solution, as demonstrated by Equation (1), resulting in the presence of Mn 2+ .In the water, numerous microscopic interfaces exist between the manganese acetate solution and S-LMO particles.When an external force is applied to stir the LMO, the forces acting on the surface's atoms, ions, and molecules become unbalanced, leading to the continuous formation of new surfaces.These newly formed surfaces are highly unstable.On one hand, the host LMO particles possess numerous reactive groups and a broad surface that allows them to electrostatically attract Mn 2+ in an aqueous solution to the particle surface.On the other hand, the LMO particles reduce surface energy through mutual reaction with Mn 2+ , as illustrated in Equation (2).
(2) The S-LMO powder, collected for analysis, underwent a 24-hour remanganization process in Mn-containing water under ambient conditions (experimental details can be found in the supporting information).Subsequently, the remanganizated LMO (RM-LMO) underwent thorough washing with deionized water, followed by annealing repair, resulting in the designation AR-LMO.The outcomes of this process demonstrate that ambient remanganization enables Mn 2+ in the solution to fill the Mn vacancies in the S-LMO powder, achieving the elemental repair of the degraded material.This conclusion finds support in the content analysis conducted using the Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES), the electrochemical performance assessment of different LMO powders, and X-ray photoelectron spectroscopy (XPS).
The Li and Mn molar ratios in RM-LMO and AR-LMO were approximately 1:2.13 and 1:2.15, respectively, whereas the Mn deficiency in S-LMO was around 25% (Table S1).In contrast, for the material remanganized under identical conditions without stirring (WS-LMO), the Li to Mn molar ratio was 1:1.84.These findings suggest that ambient water remanganization with a certain external force field is an effective method.RM-LMO outperformed S-LMO by 80% at 1C (1C = 148 mA/g), even delivering 150 mA h g −1 after 100 cycles (Figure 1A).Thermal analysis indicated that both S-LMO and RM-LMO powders experienced minimal weight loss within 30-800 • C, 0.75%, and 1.15%, respectively (Figure 1B).A distinct endothermic peak was observed within 590-720 • C and both materials exhibited relatively similar mass losses of 0.84% and 0.85%.The intrinsic residual Mn 3 O 4 entering the Mn vacancies in the S-LMO material caused the heat absorption reaction, while the surface-adsorbed Mn 2+ entering the Mn vacancies at high temperatures may be the cause of the RM-LMO material.XPS analysis revealed that the chemical states of C1s and O1s in different LMO samples were not significantly altered (Figure 1C,D), whereas the valence states of Mn varied significantly, as indicated by the results of the nonlinear least-squares fitting (NLLSF) of Mn2p (Figure 1E). 27The Mn 2+ contents in S-LMO, RM-LMO, and WS-LMO were 0.00%, 14.06%, and 3.06%, respectively.This suggests that Mn 2+ entered into the interior of degraded LMO particles after ambient water remanganization, affirming the profound impact of external force on the effectiveness of this process.Moreover, after heat treatment, the Mn valence state in the AR-LMO samples was 4+, indicating that only Mn 2+ was adsorbed on the LMO powder after ambient water remanganization.To complete the entire repair process, the RM-LMO powder must undergo annealing following ambient water remanganization.
The remanganization process is notably influenced by the concentration of the Mn 2+ -containing solution at ambient temperature.Consequently, we investigated the impact of manganese acetate solution concentration (0.1-0.5 g/mL) on the valence state of the repaired crystals, characteristic elements, and the electrochemical properties of the repaired materials.Notably, only the color of the solution changed, while the pH of the aqueous solution remained constant before and after ambient remanganization (Figure S1).The AR-LMO repaired with various solution concentrations exhibited the distinctive peaks of LMO crystals, displaying good symmetry and sharp peak peaks, indicating that the repaired materials possessed excellent crystallinity (Figure 1F).According to the results of NLLSF of Mn2p (Figure 1G), the valence state of Mn was consistently 4+ in all samples, except for the AR-LMO-2 sample, where an experimental error might be the cause.This confirms that the valence state of Mn ions in the AR-LMO material has been successfully repaired.The characteristic peak spacing of the Mn 3s orbital ranged from 4.93-5.46eV (Figure 1H and Table S2), further indicating that the elemental valence state of the LMO material was restored to 4+ after repair with different solution concentrations.The repaired material demonstrated a substantially higher discharge capacity, even double that of the S-LMO material (Figure S2).The most notable specific discharge capacity was observed in the material restored in a 0.1 g/mL solution, which could discharge up to 237.8 mA h g −1 following activation.However, the stability was lower than that of the repair materials at other concentrations (0.2-0.5 g/mL).The discharge-specific capacity of AR-LMO-1 is only 169.1 mA h g −1 after 60 cycles.In contrast, a stable contribution of 181.1 and 182.5 mA h g −1 specific discharge capacities was observed from AR-LMO-3 and AR-LMO-4.This indicates that the screening effect induced by the elevated Mn ion concentration in the water hinders charge transfer between the water and particle contact. 28After careful evaluation, the repaired materials at a concentration of 0.3 g/mL were selected for the subsequent study.
It is imperative to delve into the evolution of the crystal structure and microscopic morphology of degraded LMO during the repair process.All samples validated the typical spinel phase pattern with the Fd-3m space group (Figure S3). 29Notably, a characteristic peak of Mn 3 O 4 impurity in S-LMO and an impurity peak not associated with Mn 3 O 4 at 28 • was observed, the latter disappearing after repair (Figure S4).Additionally, the lattice parameter increased from 8.2378 to 8.2466 Å, and the cell volume increased from 559.03 to 560.83 Å 3 (Table S3).Neutron diffraction refinement findings indicated the same variations in the cell characterization parameters before and after repair, providing evidence that Mn defects in S-LMO were successfully rectified.An impurity peak not belonging to Mn 3 O 4 also appeared in the neutron diffraction pattern of S-LMO (Figure S5), and this peak disappeared in the repaired AR-LMO sample (Figure S6), further affirming the successful repair of the degraded LMO cathode in ambient aqueous solution.
To observe the structure and morphology of different LMO powders, we employed scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HRTEM).During the charge/discharge cycle, continuous stresses were generated due to anisotropic lattice changes.The accumulation of stresses caused expansion and contraction of particle volume, leading to internal cracks and voids due to cracking of particle interfaces and grain boundaries after reaching a certain threshold.The subsequent obstruction of the Li + transport channel resulted in the progressive degradation of the LMO capacity.Three distinct morphologies were identified in S-LMO, including normal octahedrons with pores and microcracks, microcracked blocks with smooth surfaces, and porous microspheres (Figure S7).The cracks and gaps in the particles could not be filled after manganese replenishment at ambient temperature, but a significant amount of Mn 2+ microparticles was attached to the surface of RM-LMO crystal particles (Figure S8).The surface of the AR-LMO particles was smooth and angular, with diameters of about 1 µm, and most of them were octahedral (Figure S9), indicating the completion of structural repair of the failed material.The crystal face spacing of AR-LMO was 0.25 and 0.29 nm, corresponding to the orientation of the (311) and (220) atomic faces in typical LMO, respectively (Figure S10B).The selected area electron diffraction pattern of AR-LMO displayed regularly arranged dots (Figure S10C), confirming that the material was a single crystal.Therefore, the HRTEM images further validate the reconstruction of the spinel structure in AR-LMO.

Dynamical characteristics of repaired LMO particles
Elemental and structural repair of deteriorated LMO materials can be achieved by adding a specific volume of manganese salts after high-temperature conversion.To evaluate the effective repair effect of the ambient manganese supplementation-annealing technique, we repaired the degraded LMO materials using the conventional solidphase technique, obtaining the repaired materials labeled as SR-LMO.A two-phase transition near 4 V is observed in the SR-LMO material at scanning rates of 0.1-0.4mV/s (Figure 2A).Our previous research uncovered Li/Mn disorder in degraded LMO materials. 26The electrochemical characteristics of the repaired LMO materials differ significantly from those of commercial LMO materials.One reason is the potentially reversible de-embedding of Li at the Li-16d location during charging/discharging, contributing some capacity.Consequently, a redox peak near 2 V appears in the cyclic voltammetry (CV) curve.Eventually, the CV curve displays three pairs of redox peaks.Secondly, the solid solution reaction at 4 V transforms the two-phase transition process into a solid solution reaction, enhancing the material's stability during charging/discharging.All these changes are observable in the electrochemical responses during CV studies from 0.1 to 0.4 mV/s for SR-LMO materials (Figure 2B).This suggests that the two-phase transition can be successfully suppressed by the solvent effect-oriented repair of AR-LMO materials, improving the materials' electrochemical stability.
As the scan rate increases, the polarization phenomenon causes the reduction peak to shift to the left and the oxidation peak to shift to the right at each reaction voltage.Additionally, the redox peak broadens as the scan rate increases from 0.1 to 0.4 mV/s, indicating a capacitive contribution role in lithium storage.The redox peak current exhibits a power-law dependence on scan rate and a roughly linear proportionality of "I = av b " (b = ∼ 0.5) (Figure 2C), indicating that the electrochemical reaction of AR-LMO and SR-LMO is primarily a diffusion-controlled process.All peak currents were linearly related to the square root of scan rates after evaluation using the standard Randles -Sevcik equation ). Figure 2D shows that AR-LMO had a Li + diffusion rate near 4 and 3 V more consistent with that of SR-LMO.The oxidation peak at 4.00 V and the reduction peak at 3.95 V have faster Li + expansion rates, while the oxidation peak at 3.10 V and the reduction peak at 2.60 V have slower Li + expansion rates.
The current at a fixed potential originates from the combination of pseudo-capacitive current and diffusion current (i = k 1 v + k 2 v 1∕2 ), where k 1 v and k 2 v 1∕2 denote the pseudo-and diffusion-controlled contributions, respectively. 30,31The capacitive contribution as a function of the voltage can be obtained with i = k 1 v at a fixed potential.The volumetric diffusion-controlled reaction capacity accounts for 71.67% of the total reaction capacity of the AR-LMO electrode at 0.1 mV/s (Figure 2E).The contribution of the pseudo-capacitance steadily grows as the scan rate increases from 0.1 to 0.4 mV/s, peaking at 35.15% at this rate.The SR-LMO electrode also exhibits an increase in pseudo-capacitance with an increasing scan rate, reaching a share of 46.90% at 0.4 mV/s.The galvanostatic intermittent titration technique (GITT) method was employed to accurately assess the Li + diffusion performance of several LMO electrodes.The D Li + of the electrode was calculated according to equation 3 32,33 : where n B denotes the molar mass, V M denotes the molar volume of the electrode materials, ΔE s = E 0 −E 3 and ΔE  = E 1 −E 2 , S denotes the cell interfacial area, and τ represents the time duration of the pulse.The AR-LMO electrode exhibited the same Li + diffusion rate before and after charge/discharge, showing good cycle stability (Figure 2F).The distinct charge/discharge stages of the SR-LMO electrode displayed Li + diffusion efficiencies consistent with those of the AR-LMO electrode, as shown in Figure 2G.
The AR-LMO electrode had a considerably greater Li + diffusion efficiency than the SR-LMO electrode.The significantly differing lithium release kinetics at the end of charge/discharge of the SR-LMO electrode leads to capacity loss.

Electrochemical response of repaired LMO particles
We assembled half-cells to conduct in situ X-ray diffraction (XRD) investigations, aiming to comprehend the relationship between structural stability and AR-LMO and SR-LMO.The spinel skeleton of AR-LMO remained wellmaintained throughout two successive charge/discharge processes, as depicted in Figure 3A.Characteristic peaks of AR-LMO reverted to their initial discharge positions, signifying the stability and inhibitory effect on the synergistic Jahn-Teller of AR-LMO.Similar structural reversible phase transitions were observed in SR-LMO, as shown in Figure 3B.However, the angular shift of characteristic peaks in SR-LMO was more pronounced than in AR-LMO during Li + de-embedding.For instance, the shift of the AR-LMO (111) peak was 0.84 • (Figure 3C), while that of the SR-LMO (111) peak was 0.96 • (Figure 3D).This implies that AR-LMO exhibits greater structural stability with less deformation.Furthermore, Figure 3E demonstrates a high overlap of CV curves at a scan rate of 0.1 mV/s, confirming the excellent reversibility of AR-LMO.
Figure 3F illustrates the rate capabilities of AR-LMO and SR-LMO materials.Following a high current density cycle of 5 C, AR-LMO and SR-LMO materials retained 233.5 and 224.1 mA h g −1 at 0.2 C, respectively.This aligns with the initial data and even surpasses expectations with high current density cycling.Additionally, Figure 3F demonstrates that the capacity of solid-phase repaired materials was significantly higher after ambient manganese supplementation-annealing at the same current density, especially at 1 and 2 C (better performance than commercial LMO, Figure S11).These two samples exhibited initial capacities of 184.5 and 149.3 mA h g −1 at 1 C following activation (Figure 3G).The SR-LMO remained stable at 138.5 mA h g −1 after 100 cycles, with a capacity retention of 92.8%.The discharge-specific capacity of AR-LMO remained constant at 169.5 mA h g −1 after 100 cycles.Figures S12 and S13 display the differential capacity plots (dQ/dV) curves and charge/discharge curves of AR-LMO and SR-LMO after various cycles.The retention intensity and shape of characteristic peaks at each voltage indicate that the AR-LMO cathode material did not significantly deteriorate and exhibited outstanding capacity retention.Furthermore, the pseudo plateau near 3 V was notably narrower compared to SR-LMO, suggesting the suppression of the first-order phase transition.These findings suggest that appropriate treatment can partially restore the electrochemical performance and cycle stability of degraded LMO materials.Additionally, compared to solid-phase repaired samples, those treated with ambient water remanganization exhibited improved electrochemical performance and cyclic stability.

First-principles calculations
Taking LMO as the central focus of this investigation, the structure and charge distribution of degraded LMO deviate from the ideal, impacting both its structural stability and ion diffusion within the material. 34Consequently, a profound understanding of the characteristics of waste materials becomes imperative for the advancement of recycling technologies.Our initial exploration involved an examination of the electronic structure of degraded LMO through first-principles calculations.Previous research has underlined that significant compositional changes in degraded LMO are attributed to Mn defects. 26Accordingly, we devised LMO structures with various defect states, encompassing an ideal spinel structure (LMO@Mn0, Figure 4A), a structure with one missing Mn atom (LMO@Mn1, Figure 4B), a structure with two missing Mn atoms (LMO@Mn2, Figure 4C), and a structure with four missing Mn atoms (LMO@Mn4, Figure 4D).Theoretical studies reveal that different degrees of lattice defects in LMO induce alterations in the spatial geometry and chemical environment of atoms.As delineated in Table S4, various defect states led to reduced crystal structure sizes.With the exception of LMO@Mn0 and LMO@Mn4, the crystal cell parameters a, b, and c were diminished and no longer identical.Moreover, the crystal structures lacking 1 and 2 Mn atoms exhibited distortion, while LMO@Mn4 reverted to the standard cubic crystal system.This distortion results from the flaws altering the Li-O and Mn-O chemical bonds between atoms.
The density of electronic states (DOS) among different structures (Figure 4E, more information in Figure S14) illustrates distinct electronic structures for Mn and O. Consequently, varying levels of Mn deficiency influence the material's conductivity and its ability to conduct electrons.With increasing deficiency, the DOS of Mn and O atoms in LMO@Mn0 and LMO@Mn4 exhibits substantial values at the Fermi energy level (E f ), signifying conventional metal-like characteristics in these systems.In contrast, the Mn and O atoms in LMO@Mn1 and LMO@Mn2 display negligible DOS values at E f , indicating semiconductor or insulator characteristics.The peaks of the occupied states for Mn and O are smaller, and the occupied states become larger for LMO@Mn1 and LMO@Mn2 compared to LMO@Mn0 and LMO@Mn4.Conversely, from LMO@Mn4 to LMO@Mn0, the peaks of the occupied states for Mn and O are marginally greater, while the occupied states are smaller.The projected density of states (PDOS) of the t 2g and e g orbitals in different configurations vary significantly between −0.5 and 0.5 eV.The e g orbital occupied state is larger than the t 2g orbital occupied state, as evidenced by the detailed PDOS of the t 2g and e g orbitals of Mn in Figure 4F.This suggests that the t 2g orbital of Mn bonds with the O 2p orbital in various structures.The occupied states of the O 2p orbital and the occupied state of the t 2g orbital of Mn nearly overlap in LMO@Mn0 and LMO@Mn4.PDOS differences between atoms impact interactions between adjacent atoms.Therefore, the Mn-O interaction is significantly affected by the apparent PDOS difference near E f between LMO@Mn0 and LMO@Mn4.
The defect formation energy served as a metric to assess the challenge of Mn vacancy defect formation, providing insights into the potential for spontaneous defect remediation.Among the different structures, LMO@Mn1 exhibited the lowest formation energy (1.82 eV), followed by LMO@Mn2 (5.84 eV) and LMO@Mn4 (24.48 eV), suggesting the theoretical feasibility of spontaneous repair for Mn vacancy defects in these structures.Theoretical calculations were conducted using the LMO@Mn1 structure to demonstrate the diffusion of Mn within the LMO framework.By simulating two Mn diffusion pathways in LMO@Mn1, encompassing the Mn-O layer and the Mn layer, we compared the diffusion energy barriers.The energy barriers for Mn diffusion in the Mn-O and Mn layers were 5.22 and 3.34 eV, respectively, as depicted in Figure 4G.These findings provide theoretical support for solution Mn replenishment (Figure 4H), as Mn tends to diffuse in the Mn-O layer and occupy the Mn16d site rather than the 8a site of Li.

Extension of the repair strategy
We investigated the reusability of the manganese solution through multiple treatments.The solution, initially at a concentration of 0.7 g/mL, remained clear and free of impurities, albeit with a diminishing volume over successive treatments (Figure S15).Throughout multiple treatments, the solution's pH consistently ranged between 6.61 and 6.62.This stability indicates that Mn 2+ in the solution effectively infiltrated the Mn vacancies in the S-LMO lattice.The initial high concentration of the designed water led to excessive adsorption on the material's surface, covering some Mn vacancy active sites.Consequently, the discharge capacity of the first sample was low, with a discharge-specific capacity 160.8 mA h g −1 after 60 cycles (Figure S16).Subsequent repairs (4th and 5th) resulted in consistently poor sample capacities, with the 5th sample exhibiting the lowest capacity and stability.This outcome is attributed to the low concentration of Mn 2+ in the water, leading to the persistence of certain Mn vacancies in the repaired material.
In contrast, the 2nd and 3rd repaired materials exhibited the best discharge-specific capacity and stability, indicating that the solution concentration was optimal for Mn 2+ to effectively occupy Mn vacancies.Furthermore, the discharge-specific capacities of the 3rd and 4th samples were comparable to those treated in 0.3 and 0.4 g/mL solutions, suggesting that this concentration range was the most effective for repairing S-LMO.These findings underscore the critical role of Mn solution concentration in regulating the repair effect.Adjusting the solution's concentration with water or Mn(CH 3 COO) 2 allows for the repeated use of the same manganese solution in repairing S-LMO.Moreover, we extended the application of the waterfacilitated effect targeted repair strategy to other systems, such as LiNi 0.5 Co 0.2 Mn 0.3 O 2 (NCM) and LiCoO 2 (LCO) cathodes.The widespread recognition of Li deficit as the primary characteristic of degraded NCM 35 and LCO 36 materials prompted our exploration.We utilized lithium acetate dihydrate (CH 3 COOLi 2H 2 O) as the Li source and adjusted the aqueous solution concentration to 0.3 g/mL for ambient lithium replenishment.The spent NCM (S-NCM) exhibited a mixture of large and small particles, with some agglomeration phenomena observed (Figure S17).Continuous Li + de-embedding during charging/discharging led to prominent holes in the material.The Energy Dispersive Spectrometer revealed a homogeneous distribution of Ni, Co, and Mn in S-NCM (Figure S17).The layer structure was evident, and the surface of the repaired NCM (R-NCM) particles appeared remarkably smooth (Figure 5A).Importantly, the agglomeration of R-NCM particles weakened due to the effective dispersion in the aqueous solution, displaying single-crystal properties and a uniform distribution of Ni, Co, and Mn components (Figure 5B,C).Similarly, in the LiNi 0.6 Co 0.2 Mn 0.2 O 2 (NCM622) system, polycrystalline conversion to single crystal was also observed (Figure S18), indicating that the morphology of NCM materials can be effectively through relithiation in ambient solution.R-NCM exhibited significantly improved cycling performance (Figure 5D-F).Additionally, the cycle performance and stability of the repaired LCO were notably enhanced using this technique (Figure S19).These results underscore the universal applicability and suitability of the water-facilitated targeted repair strategy for various cathode systems.

CONCLUSION
We present a novel method for the targeted repair of degraded cathode materials in lithium-ion batteries (LIBs) through the use of ambient water.Elemental repair of degraded LMO can be achieved via ambient-temperature water remanganization, while structural repair can be accomplished through thermal treatment.The resulting repaired LMO materials exhibit superior electrochemical and kinetic properties compared to materials repaired using conventional solid-state methods.Our proposed strategy offers several advantages over traditional repair processes: (1) Versatility: Our method can safely repair ion vacancies in positive electrode materials at normal temperature and pressure, without the need for precise knowledge of the waste's composition and structural integrity.This makes it suitable for the synchronous and accurate repair of different material batches.(2) Cost-effectiveness: The repair agent employed in our method is reusable, ensuring zero loss and low economic cost.
(3) Wide applicability: Our process is broadly applicable to relithiate LiCoO 2 and NCM materials, promoting the conversion of various polycrystalline NCM structures into single crystal materials.This efficiently facilitates the upcycling and recycling of cathode materials.In summary, our work introduces a novel strategy for the directional water-facilitated repair of waste LIBs cathode materials.This approach demonstrates the efficient in situ repair of materials with varying defect degrees, ensuring accurate restoration of battery materials in terms of morphology, structure, and electrochemical properties.

Direct repairing of scrap LMO powder
The S-LMO powder was pretreated in the same way as reported previously. 26The powder was placed in aqueous manganese acetate (MnCH 3 (COO) 2 ) solution at different concentrations, stirred for 24 h at ambient temperature, filtered and dried in a quartz crucible, warmed to 700 • C at a rate of 5 • C/min in air and kept warm for 15 h, and then slowly cooled to ambient temperature at a rate of 5 • C/min.The annealed repaired samples with different concentrations were labeled as AR-LMO-1, AR-LMO-2, AR-LMO-3, AR-LMO-4, and AR-LMO-5, while the samples after remanganization treated by drying only were labeled as RM-LMO.S-LMO powder was mixed well with a certain amount of manganese acetate at 700 • C and kept for 15 h.This is a solid-state repairing process, so the obtained samples were labeled as SR-LMO.For the samples repaired by five times ambient temperature remanganization annealing in the same solution, they are marked as 1st, 2nd, 3rd, 4th, and 5th.For NCM and LCO wastes, the samples were stirred at ambient temperature for 24 h in a 0.3 g/mL solution of lithium acetate.A certain amount of Li 2 CO 3 was added to make sure the Li/TM ratio was 1.05:1.After evenly mixed, heated the mixed powder to 350 • C in the air at 5 • C/min, held for 5 h, and then heated to 850 • C at 5 • C/min, held for 12 h, and then slowly cooled to ambient temperature.Use the same method for spent NCM622 material.

Materials characterization
Cu Kα radiation ex-situ X-ray diffraction spectrometer (X'pert PRO, PANalytical) was used to characterize the crystal structure and phase composition of the pow-der materials.In-situ XRD at ambient temperature was employed to analyze the chemical transformations occurring during the charging/discharging of the cells.The data were collected using a step scan with a scan speed of 1.5 • /min and a scan angle (2 θ) of 15-80 • .XPS (ESCALAB 250Xi spectra-meter; Thermo Scientific) was used to study the phase formation of solid powders.The excitation source is an Al Kα X diffraction source with a power of 150 W. The binding energy can be calibrated with the C 1s peak of hydrocarbon at 284.8 eV.The microstructures and surface morphology of solid powders were characterized by field emission SEM (FESEM).A high-resolution transmission electron microscopy (HRTEM, JEM-2100F) was used to detect the lattice fringe of samples (Tecnai G2 F30 S-TWIN).The content of all metals in the powders was determined by inductively coupled plasma emission spectroscopy (ICP-OES, iCAP 6300 Radial; Thermo Scientific).
The powder samples needed to be completely dissolved in aqua regia (HCl: HNO 3 = 3:1, v/v) before analysis.

Electrochemical conversion and measurements
A 2025 button cell with Li metal as a counter electrode, 1 M LiPF 6 in EC: DEC: EMC (volume ratio: 1:1:1) as the electrolyte, and a three-layer membrane (Celgard 2320) as separator was used for powder electrochemical testing.The recycled cathode electrode was prepared by coating a mixture of 80 wt% cathode powder, 10 wt% polyvinylidene fluoride, and 10 wt% carbon acetylene black on an aluminum foil current collector.The prepared electrode sheets were then dried overnight in a vacuum oven at 80 • C until the solvent evaporated.The electrode loading was about 3.5 mg/cm 2 .The cells were assembled in an argon atmosphere glove box, where both oxygen and water concentrations were below 0.1 ppm.A NEWARE battery test system was used to perform constant current charging and discharging at 30 • C. Cells were galvanostatically charged and discharged at different current densities for a cut-off voltage of 1-4.7 V (vs.Li/Li + ) at ambient temperature.In particular, for NCM and LCO, the test conditions were as follows: voltage regulation range of 2.7-4.3V. CHI 660C (CH Instruments) was used to record CVs in a coin cell with different scan rates.A GITT test was performed on a NEWARE battery test system with a lithium reference adjusted at a voltage range of 1-4.7 V at 0.1 C. The GITT test was performed in a second cycle.Each current pulse lasted for 10 min at 0.1 C and was allowed to equilibrate for 180 min before starting the next pulse.The collected data were analyzed to determine the diffusion coefficient of the material.

F I G U R E 2
Cyclic voltammetry (CV) curves with different scan rates for (A) SR-LMO electrode and (B) AR-LMO electrode.(C) Log-scale plot of peak current dependence on the scan rate.The b value indicates the slope of the linear fit curve.(D) Plot of CV peak current versus square root scan rates for AR-LMO electrode and SR-LMO electrode.(E) The pseudo and diffusive current in the experiment CV image voltage at various scan rates.Galvanostatic intermittent titration technique (GITT) curves and the corresponding DLi+ of (F) AR-LMO electrode and (G) SR-LMO electrode.

F I G U R E 3
The structural evolution of (A) AR-LMO and (B) SR-LMO cathode during the two successive charge/discharge cycles (Tested at 0.1 C).In situ X-ray diffraction (XRD) patterns of (C) AR-LMO and (D) SR-LMO cathode from 18.5 • to 19.5 • .(E) CV curves at 0.1 mV/s for AR-LMO electrode.(F) Rate performance of SR-LMO and AR-LMO samples.(G) Cycling performance of SR-LMO and AR-LMO samples at 1 C.

F I G U R E 4
Theoretical calculations of LiMn 2 O 4 (S-LMO) structures with different degrees of Mn defects.(A) structure without Mn atom missing (LMO@Mn0), (B) structure with 1 Mn atom missing, (LMO@Mn1) (C) structure with 2 Mn atoms missing (LMO@Mn2), (D) structure with 4 Mn atoms missing (LMO@Mn4).(E) The density of states (DOS) and (F) projected density of states (PDOS) values of different structures.(G) Diffusion energy barriers of Mn in LMO@Mn1 along different paths: diffusion in an Mn-O layer and diffusion in the Mn layer.(H) Schematic illustration of water-facilitated targeted repair strategy for degraded LMO material.

F
I G U R E 5 (A, B) Scanning electron microscopy (SEM) images of the repaired LiNi 0.5 Co 0.2 Mn 0.3 O 2 (R-NCM) powder.(C) Energy dispersive X-ray spectroscopy (EDS) mapping of R-NCM powder.(D) Cycling performance of different spent NCM (S-NCM) and R-NCM powder at 1 C. (E) Charge/discharge curves of R-NCM at different rates.(F) Electrochemical properties of R-NCM at different rates.