English

• Currently, lithium-ion batteries (LIBs) are widely used in many fields due to their high energy density, good cycling performance, and high safety^[1-3]. Consequently, a substantial volume of waste batteries is anticipated in the future^[4]. Hazardous substances in them can significantly impact the environment, and valuable metal resources may be wasted if the spent batteries are not handled promptly^[5-6]. Therefore, it is important to recycle waste batteries to guarantee sustainable resource management and reduce environmental impact.

  Hydrometallurgy, as currently the most popular strategy, provides a cost-effective and rapid method, making it an easy, safe, and eco-friendly option for extracting and recycling metals from used LIBs^[7]. It includes pretreatment, acid leaching, and metal ion recovery processes. Pretreatment mainly includes battery discharge, disassembly, and collection of cathode materials. Inorganic acids like H₂SO₄, HCl, and HNO₃ are widely employed as leaching agents due to their high efficiency and affordability, but they release toxic gases such as Cl₂, SO₃, and NO_x, as well as wastewater and waste residues^[8-11]. As a result, many current studies favor the use of easily degradable organic acids, such as citric acid, succinic acid, and oxalic acid^[12-14]. Finally, the metal ions are transformed into metal compounds by evaporation, precipitation reaction, electrodeposition, etc^[15-17].

  Although the above method can recycle waste LIBs, it is challenging to extract high-purity products from them because of the similarity of the chemical properties of the elements in them. In recent years, numerous studies have proposed more cost-effective recycling methods, which can directly regenerate cathode materials from leaching solution containing metal ions or waste cathode materials, achieving closed-loop recycling of waste LIBs. Montoya et al. used a sintering method to successfully recycle the waste NCM111 (LiNi_0.33Co_0.33Mn_0.33O₂) material and restore the material properties to the original level^[18]. He et al. efficiently leached metals from waste NCM523 (LiNi_0.5Co_0.2 Mn_0.3O₂) materials using deep eutectic solvent (DES), and successfully regenerated the materials through co-precipitation and high-temperature solid-phase techniques^[19]. Shi et al. studied the hydrothermal regeneration method, and the regenerated materials showed excellent performance^[20]. Although these methods have achieved a closed-loop regeneration of waste LIBs, there are still some problems. The sintering method is simple to operate, but the impurities in the material are difficult to remove and the mixing is uneven. The hydrothermal method is low-cost and simple to implement, but its strict equipment requirements limit large-scale application. Additionally, the high loss rate and insufficient performance limit its wide application.

  Herein, DL-tartaric acid (C₄H₆O₆) and formic acid were used as leaching agents, and oxalic acid was used as a co-precipitant, followed by regeneration of the material through heat treatment. DL-tartaric acid is a naturally occurring organic acid frequently found in plants. It has significant industrial applications in the food, pharmaceutical, and manufacturing sectors. DL-tartaric acid is more cost-effective and has higher acidity than DL-malic acid, ascorbic acid, L-aspartic acid, and glycine^[21]. Compared with other organic acids, DL-tartaric acid can form strong polydentate coordination complexes with metal ions and has different complexing abilities with different metal ions, so it can selectively leach valuable metals^[22]. Formic acid can not only be used as a leaching agent in the reaction process, which can ionize H⁺ to improve the leaching rate, but also has similar reducing properties to aldehydes, and can be used as a reducing agent to convert high-valence metal ions into low-valence metal ions, which then react with DL-tartaric acid to form highly soluble complexes^[23-24]. The combination of the two organic acids can greatly improve the leaching efficiency and achieve efficient recovery of metal ions. After that, the materials were efficiently regenerated by co-precipitation and high-temperature solid-state method, and the regenerated materials showed good crystal structure and excellent electrochemical properties.

  1.   Experimental

  1.1   Materials and reagents

  The waste NCM523 batteries were collected from the Farasis Energy Co., Ltd. The commercial NCM523 cathode powder was bought from Ronbay New Energy Technology Co., Ltd. Ketjen black (KB, ECP-600JD) was obtained from Lion Corporation. Polyvinylidene difluoride (PVDF, ≥99.5%) was obtained from Arkema Co. Lithium carbonate (Li₂CO₃, ≥99.5%), DL-tartaric acid (C₄H₆O₆, ≥99%), and absolute ethanol (C₂H₅OH, ≥99.5%) were bought from Adamas Reagent Ltd. Formic Acid (CH₂O₂, ≥98%) was bought from Shanghai Titan Scientific Co., Ltd. Oxalic acid (C₂H₂O₄, 98%) and N-methyl-2-pyrrplidone (NMP, ≥99.0%) were purchased from Shanghai Aladdin Bio-Chem. Ultrapure water (18.2 MΩ·cm of resistivity at 25 ℃) was employed in all experiments. All of these were analytical-grade reagents and could be used without further treatment. The glassware used in the experiment was cleaned with deionized water and ultrasonic cleaning solution and was used after drying.

  1.2   Experimental procedures

  The waste NCM523 batteries were fully discharged to less than 1.5 V using discharging equipment. The dismantling process was carried out in a glove box filled with argon, and the cathode material, anode material, separator, and plastic were collected separately. The obtained cathode material was heated to 500 ℃ in an air atmosphere for 5 h, and then the cathode material and aluminum foil were separated by ultrasonication, and the obtained sample was named W-NCM.

  W-NCM was placed in a mixed solution of DL-tartaric acid and formic acid, and the mixture was heated and stirred until the solid was completely dissolved. The leachate was poured into an oxalic acid solution and stirred at 50 ℃ for 12 h to carry out the co-precipitation reaction. The formed precipitate was dried and mixed with Li₂CO₃, then heated to 500 ℃ in an air atmosphere for 5 h. A gradient experiment was performed on the second calcination temperature to determine the optimal calcination temperature during the regeneration process. These calcination temperatures include 850, 900, and 950 ℃. The range of materials obtained was named R-NCM-X, where X ℃ represents the calcined temperature. Commercial NCM523 was used as a control in comparative tests, where polycrystalline NCM523 was designated C-NCM and single crystal NCM523 was designated CS-NCM.

  1.3   Material characterizations

  X-ray diffraction (XRD) patterns were obtained on a Bruker X-ray diffractometer (Bruker D8 Advance) with Cu Kα radiation (λ=0.154 nm), operating at 40 kV and 150 mA, a scan range of 10° to 80°, and a scan rate of 10 (°)·min^-1. The amounts of Li, Ni, Co, and Mn on the materials were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES, PerkinElmer 8300). X-ray photoelectron spectroscopy (XPS, Escalab 250XI, Al Kα radiation, Thermo Fisher, 150 W) was used to characterize the elemental composition and chemical valence changes on the surface of the materials. The results were calibrated with the C1s peak at 284.8 eV. The morphology of the samples was characterized using scanning electron microscopy (SEM, Hitachi S4800). The microstructure and elemental distribution of the samples were characterized using high-resolution transmission electron microscopy (HR-TEM, FEI TALOS 200X) with an operating voltage of 200 kV.

  1.4   Electrochemical measurement

  The cathode consisted of cathode material, KB, and PVDF in a mass ratio of 8∶1∶1. The mixture was blended using ball milling to form a slurry, which was then evenly coated onto aluminum foil and dried overnight at 80 ℃ in a vacuum. The assembly of the coin cell batteries was performed in an argon-filled glove box. The battery case model was CR2025 with lithium metal as the anode. The electrolyte was 1 mol·L^-1 LiPF₆ dissolved in a mixed solution of ethylene carbonate (EC) and diethyl carbonate (DEC) with a volume ratio of 1∶1. After the battery was assembled, it was tested using the Neware Battery Testing System (BTS) with a voltage range of 2.5-4.3 V. Cyclic voltammetry (CV) (2.5-4.3 V, 0.1 mV·s^-1) and electrochemical impedance spectroscopy (EIS) (0.01-10⁵ Hz) were performed using a CHI760E electrochemical workstation.

  2.   Results and discussion

  2.1   Regeneration process

  The flow chart of the regeneration of waste NCM is shown in Fig. 1. After a long period of cycling, the capacity of lithium-ion batteries was reduced due to lithium deficiency and structural damage. Through high-temperature calcination and ultrasonic treatment, the conductive carbon and PVDF in the material were removed, and the cathode powder and aluminum foil were effectively separated. The DL-tartaric acid-formic acid system effectively leached the transition metal ions in the spent cathode material, and the leaching solution reacted with oxalic acid to form oxalate compounds. The obtained oxalate compound was evenly mixed with lithium salt, and then the regenerated NCM material was obtained by heat treatment. This process enabled efficient leaching of waste materials and restored the crystal structure of the materials through co-precipitation and heat treatment, thereby recovering the electrochemical properties of the material.

  Figure 1

  

  Figure 1.  Diagram of the waste NCM regeneration process

  下载: 全尺寸图片 幻灯片

  2.2   Exploration of optimal leaching conditions

  As demonstrated in Eq.1 and 2, DL-tartaric acid has two carboxyl groups per molecule, and its dissociation reaction is illustrated below:

  $ \mathrm{H}_2 \mathrm{C}_4 \mathrm{H}_4 \mathrm{O}_6 \rightleftharpoons \mathrm{HC}_4 \mathrm{H}_4 \mathrm{O}_6^{2-}+\mathrm{H}^{+} $

  (1)

  $ \mathrm{HC}_4 \mathrm{H}_4 \mathrm{O}_6{ }^{-} \rightleftharpoons \mathrm{C}_4 \mathrm{H}_4 \mathrm{O}_6{ }^{2-}+\mathrm{H}^{+} $

  (2)

  During the leaching process, numerous reactions can occur, involving various ionic valence states. Adding formic acid is crucial for enhancing the leaching reaction rate, as it not only reduces high valence ions due to its properties as a reducing organic acid but also contributes acidity, supplying hydrogen ions that boost leaching efficiency. As illustrated in Eq.3 and 4, formic acid first reduces the high valence transition metals (M³⁺ and M⁴⁺, which M includes Ni, Co, and Mn) to lower valence states. Subsequently, these ions form complexes with dissociated tartaric acid, as depicted in Eq.5.

  $ 2 \mathrm{M}^{3+}+\mathrm{HCOOH} \rightarrow 2 \mathrm{M}^{2+}+\mathrm{CO}_2 \uparrow+2 \mathrm{H}^{+} $

  (3)

  $ \mathrm{M}^{4+}+\mathrm{HCOOH} \rightarrow \mathrm{M}^{2+}+\mathrm{CO}_2 \uparrow+2 \mathrm{H}^{+} $

  (4)

  $ \mathrm{M}^{2+}+\mathrm{C}_4 \mathrm{H}_4 \mathrm{O}_6^{2-} \rightarrow \mathrm{M}\left(\mathrm{C}_4 \mathrm{H}_4 \mathrm{O}_6\right) $

  (5)

  To assess the influence of different DL-tartaric acid concentrations on the metal leaching rates from spent cathode materials, the concentration was progressively increased from 0.2 to 0.6 mol·L^-1, under conditions of a 2% volume fraction of formic acid, a reaction temperature of 90 ℃, a reaction time of 70 min, and a solid-liquid ratio of 13 g·L^-1. Results are illustrated in Fig.S1a (Supporting information). The efficiency of the leaching process was calculated using Eq.S1. At a DL-tartaric acid of 0.2 mol·L^-1, the leaching rates for Li, Ni, Co, and Mn were 54.25%, 55.03%, 54.77%, and 55.57%, respectively, indicating that this concentration was insufficient for effective acid leaching. After increasing the DL-tartaric acid concentration to 0.5 mol·L^-1, the leaching rates exceeded 97% for each element. Increasing the DL-tartaric acid concentration beyond this point did not notably improve the leaching efficiency. To balance leaching efficiency with operating costs, a DL-tartaric acid concentration of 0.5 mol·L^-1 was deemed optimal for acid leaching.

  To investigate the impact of formic acid content on the metal leaching rate from spent cathode materials, the formic acid volume fraction was progressively increased from 0 to 3% under conditions including a DL-tartaric acid concentration of 0.5 mol·L^-1, a reaction temperature of 90 ℃, a reaction time of 70 min, and a solid-liquid ratio of 13 g·L^-1. Results are presented in Fig.S1b. Without formic acid, the leaching rates for Li, Ni, Co, and Mn were 91.90%, 89.80%, 90.86%, and 90.30%, respectively. Increasing formic acid volume fraction to 2% increased the leaching rates of these metals to approximately 98%, demonstrating that formic acid′s reducing properties contributed to the reduction of the valence of the metals, thereby enhancing overall leaching efficiencies. Upon increasing the formic acid volume fraction to 3%, the enhancement in leaching efficiency became marginal. Considering economic factors, an optimal formic acid volume fraction of 2% was established.

  To examine how the solid-liquid ratio affects metal leaching rates from spent cathode materials, the ratio was gradually increased from 10 to 20 g·L^-1, with 0.5 mol·L^-1 DL-tartaric acid, 2% formic acid, a reaction temperature of 90 ℃, and a reaction time of 70 min. Results are illustrated in Fig.S1c. As the solid-liquid ratio increased, the metal leaching rate continuously decreased, achieving over 99% leaching efficiency at a ratio of 10 g·L^-1. At a solid-liquid ratio of 20 g·L^-1, the leaching rates for Li, Ni, Co, and Mn decreased to 98.71%, 94.47%, 92.26%, and 95.44%, respectively. This decrease is attributed to the larger ratio reducing the convection and diffusion of the solution, thereby diminishing metal leaching efficiency. With a solid-liquid ratio of 13 g·L^-1, leaching rates for all metals stabilize around 98%. Considering both operating costs and leaching efficiency, an optimal ratio of 13 g·L^-1 was established.

  To assess how reaction temperature influences the metal leaching rates from spent cathode materials, the temperature was gradually increased from 50 to 90 ℃, maintaining 0.5 mol·L^-1 DL-tartaric acid, 2% formic acid, a reaction time of 70 min, and a solid-liquid ratio of 13 g·L^-1. Results are presented in Fig.S1d. At a reaction temperature of 50 ℃, the leaching rates for Li, Ni, Co, and Mn were 66.95%, 46.86%, 32.89%, and 45.51%, respectively. As the reaction temperature increased, the activated molecules′ content increased, accelerating the reaction and significantly enhancing the efficiency of metal leaching. When the temperature reached 90 ℃, leaching rates for each metal increased to approximately 98%, establishing 90 ℃ as the optimal reaction temperature.

  To assess how reaction time affects metal leaching rates from spent cathode materials, the reaction time was progressively extended from 20 to 80 min, under conditions including 0.5 mol·L^-1 DL-tartaric acid, 2% formic acid, a reaction temperature of 90 ℃, and a solid-liquid ratio of 13 g·L^-1. Results are depicted in Fig.S1e. At a reaction time of 20 min, leaching rates for Li, Ni, Co, and Mn were 87.22%, 88.23%, 77.26%, and 85.50%, respectively. As reaction time increased, the molecules gained more kinetic energy, facilitating more effective intermolecular collisions and significantly boosting leaching efficiency. Upon reaching a reaction time of 70 min, leaching rates for each metal increased to approximately 98%. Further extension of reaction time did not significantly enhance leaching efficiency, establishing 70 min as the optimal duration.

  Drawing from the experimental findings above, the optimal acid leaching conditions have been established as follows: 0.5 mol·L^-1 DL-tartaric acid, 2% formic acid, a reaction temperature of 90 ℃, a reaction time of 70 min, and a solid-liquid ratio of 13 g·L^-1. As illustrated in Fig.S1f, the leaching rates for Li, Ni, Co, and Mn achieved 98.03%, 98.69%, 97.59%, and 98.5%, respectively. This demonstrates that the leaching conditions effectively leach the spent NCM materials and establish a solid foundation for subsequent regeneration steps.

  To investigate the reaction control mechanism of metals in spent cathode materials using DL-tartaric acid and formic acid, a kinetic analysis of the leaching process was conducted, examining leaching rates at various times (20-70 min) and temperatures (50-90 ℃). Results are illustrated in Fig.S2. The optimal leaching results are consistent with the previous experimental findings, demonstrating a gradual increase in metal leaching rates with rising temperature and extended reaction times.

  The shrinkage kernel model describes the characteristics of the acid leaching process well, and the leaching of spent cathode materials is divided into five key steps: (1) transport of leaching agent molecules to the liquid-liquid interface; (2) movement of leaching agent molecules through the diffusion layer to the reaction interface; (3) leaching agent molecules react with spent cathode materials at the solid-liquid interface, causing the dissolution of metal ions; (4) diffusion of the resultant metal ions through the diffusion layer to the liquid-liquid interface; (5) release of metal ions into the solution^[25]. As detailed in Eq.6, the Avrami equation model, commonly applied in leaching kinetics studies, demonstrates fitting results depicted in Fig.S3^[26].

  $ [-\ln (1-x)]^2=k t $

  (6)

  where k is the reaction rate constant; x is the metal ion leaching rate (%); and t is the leaching time (min). The confidence level (R²) of the Avrami equation model, as shown in Table S1, indicates that the model provided a good fit. Consequently, the Avrami equation effectively describes the leaching process of metal ions from spent materials using DL-tartaric acid and formic acid. In addition, the apparent activation energy (E_a) of the metal ions can be calculated by the Arrhenius equation as shown in Eq.7 which describes the reaction rate constant as a function of reaction temperature.

  $ \ln k=\ln A-E_{\mathrm{a}} /(R T) $

  (7)

  where k is the reaction rate constant (min^-1), T is the reaction temperature (K), R is the ideal gas constant (taken as 8.314 5 J mol^-1·K^-1), and A is the preexponential factor.

  The values of the leaching rate constant for the metal leaching process at different temperatures were brought to fit in the Arrhenius equation and the results are shown in Fig.S4. The apparent activation energies of Li, Ni, Co, and Mn were calculated as 65.06, 107.36, 121.35, and 109.65 kJ·mol^-1, respectively. According to the literature, the leaching process is controlled by diffusion when E_a < 20 kJ·mol^-1, by diffusion and surface chemical reaction when 20 kJ·mol^-1 < E_a < 40 kJ·mol^-1, and by surface chemical reaction when E_a > 40 kJ·mol^{-1 [13]}. In this experiment, the leaching process of the four metal ions was controlled by the surface chemical reaction, and the metal leaching rate varied greatly with temperature in this mode, which is consistent with the results of previous experiments with a single control variable.

  2.3   Physical characterization analysis of materials

  After the acid leaching process, the metal ions were recovered in the form of oxalate dihydrate by co-precipitation. The reactions that occur during co-precipitation are shown in Eq.8:

  $ \mathrm{M}\left(\mathrm{C}_4 \mathrm{H}_4 \mathrm{O}_6\right)+\mathrm{H}_2 \mathrm{C}_2 \mathrm{O}_4 \rightarrow \mathrm{MC}_2 \mathrm{O}_4+\mathrm{C}_4 \mathrm{H}_6 \mathrm{O}_6 $

  (8)

  Fig.S5a and S5b show the SEM images of the co-precipitated sample, indicating that the material is composed of ellipsoidal particles of about 4 μm with a typical oxalate crystal structure. XRD tests were performed on the samples, as shown in Fig.S5c, which are in good agreement with the characteristic diffraction peaks of the standard card PDF No.25-0582, indicating that the precursor formed by co-precipitation is oxalate dihydrate. The image inserted in Fig.S5c is a physical image of the precursor, which was light green due to the high nickel content in the material.

  To obtain the optimal calcination temperature for the preparation of R-NCM, the recycled material prepared at different regeneration temperatures was tested, and the results are shown in Fig.S6 and S7. Lower calcination temperature will lead to lower crystallinity of the material and unable to form a good layer structure, while higher calcination temperature will lead to a longer diffusion path and slower ion transport rate for lithium ions^[27]. Fig.S6 shows the morphology of the recycled material at different temperatures, and it can be seen that the particle size of the recycled material increased with the continuous increase of temperature. Fig.S7 shows the charge-discharge performance of the recycled material at different calcination temperatures. The discharge-specific capacities of recycled materials at 0.1C (18 mA·g^-1) under different calcination conditions at 850, 900, and 950 ℃ were 148.5, 168.5, and 161.3 mAh·g^-1, respectively; the first Coulombic efficiencies were 78.0%, 85.2%, and 85.2%, respectively; the first discharge-specific capacities of the recycled materials at 0.5C were 141.3, 153.9, and 149.6 mAh·g^-1, and the capacity retention rates after 50 cycles were 97.0%, 99.0%, and 90.0%, respectively. The regenerated material showed the best discharge specific capacity and cycle stability at a calcination temperature of 900 ℃, so the calcination temperature was determined as 900 ℃. Unless otherwise noted, R-NCM below refers to the sample prepared at optimal temperature (900 ℃).

  The XRD patterns of C-NCM, R-NCM, and W-NCM are shown in Fig. 2a.The diffraction peaks of these three samples correspond to the standard card PDF No.09-0063, indicating that these materials possessed a hexagonal crystalline system α-NaFeO₂-type layered structure with a R3m space group^[20]. Although the XRD patterns of W-NCM exhibited the same α-NaFeO₂ structure as C-NCM and R-NCM with no additional peaks observed, slight variations can be seen. These variations primarily stemmed from the oxidation of Ni²⁺ to Ni³⁺, a response that compensated for lithium deficiency in W-NCM, leading to changes in the bond distances and ionic radius. Zooming the plots in the range of 18.0°-20.0° and 64.0°-66.0°, as illustrated in Fig. 2b and 2c. The (003) plane characteristic peak of W-NCM was observed to shift to the left, which was caused by electrostatic repulsion between the oxygen layers resulting from lithium deficiency^[28]. Meanwhile, the (108)/(110) peak splitting degree of W-NCM was not obvious, indicating that the material had poor layered properties^[29]. In contrast, the distinct peak splitting observed in C-NCM and R-NCM indicated well-developed layered structures, implying that the crystal structure of the material had been repaired after regeneration.

  Figure 2

  

  Figure 2.  (a) XRD patterns of C-NCM, R-NCM, and W-NCM, and (b, c) their enlarged views in detail; High-resolution XPS spectra of (d) Ni2p, (e) Co2p, and (f) Mn2p of C-NCM, R-NCM, and W-NCM

  下载: 全尺寸图片 幻灯片

  After refining the XRD pattern, the lattice parameters of the materials are listed as shown in Table S2. The c/a value is crucial in characterizing the layered structure of lithium-ion battery cathode materials. A higher c/a value indicates a more pronounced layered structure, and it is widely accepted that a c/a value above 4.9 indicates that the material has a good layered structure. The c/a value of W-NCM was 4.933, and the value of the R-NCM was 4.965, which means that the main part of W-NCM maintained a well-layered structure, and only part of it was destroyed throughout the cycling process, and the structure of the material was completely repaired after the regeneration process. The intensity ratio of the (003) plane diffraction peak to the (104) plane diffraction peak (I₀₀₃/I₁₀₄) is widely recognized as a crucial index for assessing the extent of cation mixing in NCM materials. A lower I₀₀₃/I₁₀₄ value signifies a higher degree of cation mixing^[30]. The I₀₀₃/I₁₀₄ value for R-NCM was 1.87, significantly higher than the 1.66 ratio for W-NCM, suggesting that the cation mixing was considerably reduced after regeneration. The ratio of the sum of the intensities of the (006) and (102) plane diffraction peaks to the intensity of the (101) plane diffraction peak (I_006+102/I₁₀₁) can be used to determine the structure of the layered oxides in the material, a smaller value indicates a more orderly arrangement of the material′s crystal structure, which is generally less than 0.5^[31]. The I_006+102/I₁₀₁ ratio of R-NCM was 0.45, significantly lower than that of W-NCM (1.07) and comparable to C-NCM (0.42). This indicates that the regenerated materials have a more ordered crystal structure^[32]. Both XRD pattern analysis and the comparison of refining lattice parameters indicate that the regenerated materials have restored their crystal structure, showing a good layered structure and high crystallinity.

  The survey XPS spectra of C-NCM, R-NCM, and W-NCM are shown in Fig.S8. The Ni2p spectra of C-NCM, R-NCM, and W-NCM are displayed in Fig. 2d. Ni2p_3/2 and Ni2p_1/2 spectra exhibited binding energy peaks at approximately 854.2 and 873.3 eV, respectively. The Ni2p_3/2 binding energy peak was divided into two distinct fitted peaks (Ni²⁺ at 854.6 eV and Ni³⁺ at 855.7 eV). Based on the peak area, the Ni³⁺ content in W-NCM was 32.9%, which was much lower than that of 38.7% in C-NCM. The decrease in Ni³⁺ content worsens cation mixing, reduces electrochemical performance, and seriously affects the charging and discharging efficiency of the battery^[33]. The Ni³⁺ content in the regenerated material was 36.8%, indicating that the regeneration process effectively reduced the degree of cation mixing. Fig. 2e presents the Co2p spectra for the three materials. The Co2p_3/2 and Co2p_1/2 binding energy peaks were approximately 780.0 and 795.1 eV, respectively. The fitting results revealed that the valence state of Co remained unchanged, aligning with previous reports^[34]. The Mn2p spectra of C-NCM, R-NCM, and W-NCM are shown in Fig. 2f. Peaks around 642.7 and 653.8 eV represent Mn⁴⁺ in the 2p_3/2 and 2p_1/2 orbitals, respectively. The valence state of Mn does not change and does not participate in electrochemical reactions.

  The structure characterizations of W-NCM, C-NCM, and R-NCM were investigated by SEM. As shown in Fig. 3a, the waste material had a rough surface with particle agglomeration, as well as many small and irregularly shaped particles. It is mainly due to mechanical crushing during spent battery pretreatment and damage to the material during battery charging and discharging. Fig. 3c shows the regenerated sample in an irregular agglomerated form, which is mainly because the sample was prepared by co-precipitation^[27]. After the regeneration process, the surface of the sample was smooth and free of impurities, similar to the new material (Fig. 3b). To further investigate the crystal structure of the regenerated material, the morphological properties of R-NCM were analyzed by TEM. As can be seen from Fig. 3d and 3e, the regenerated material was 1-2 μm particles, with a distinctly ordered lattice fringe and a good layered structure, with no remaining impurity phases. As shown in Fig. 3f, a lattice stripe spacing of 0.204 nm was obtained after image analysis and processing, corresponding to the (104) plane of the NCM material.

  Figure 3

  

  Figure 3.  SEM images of (a) W-NCM, (b) C-NCM, and (c) R-NCM; (d-f) TEM and HR-TEM images of R-NCM; (g) HADDF-TEM and elemental mapping images of R-NCM

  下载: 全尺寸图片 幻灯片

  Fig. 3g shows the elemental mappings of R-NCM. It could be seen that Ni, Co, and Mn were evenly distributed after regeneration. For a more accurate understanding of the elemental distribution, the elemental content of the material was determined by ICP-OES. The results are illustrated in Table S3, the molar ratio of Li, Ni, Co, and Mn in W-NCM was 0.7∶0.51∶0.19∶0.3, indicating an obvious lithium deficiency in W-NCM. After regeneration, the lithium content in R-NCM reached the normal value of 1.04∶0.51∶0.19∶0.3. The primary cause of the capacity decline in spent cathode materials is the deficiency of Li, and R-NCM has regained the missing lithium. These results demonstrated that the regeneration process restored the crystal structure and elemental ratio of the waste NCM523 material, strongly confirming the feasibility of the recycling process.

  2.4   Electrochemical performance of materials

  Fig. 4a demonstrates the initial CV curves for C-NCM, R-NCM, and W-NCM. The CV curves of all three samples exhibited a single oxidation peak and reduction peak, corresponding to the Ni²⁺/Ni³⁺ redox process. The redox potential differences for the three materials in the first cycle were 0.144, 0.149, and 0.340 V, respectively. Compared to C-NCM and R-NCM, W-NCM exhibited higher electrode polarization and a slower transmission rate of lithium ions. In addition, cation mixing in the material also leads to increased polarization^[35]. Meanwhile, R-NCM and C-NCM exhibited similar peak shape and peak current, indicating that the regeneration process increased the lithium-ion diffusion rate and reduced the electrode polarization. The EIS of the three materials is shown in Fig. 4b. The charge transfer resistance (R_ct) was approximately 84 Ω for C-NCM and 124 Ω for R-NCM, significantly lower than the 265 Ω observed for W-NCM, indicating that regeneration effectively restored the conductivity of the material. According to the lithium-ion diffusion coefficient equation, as shown in Eq.S2 and S3, the lithium-ion conductivities for C-NCM, R-NCM, and W-NCM of 6.24×10^-14, 5.32×10^-14, and 1.91×10^-14 cm²·s^-1, respectively, and the results are shown in Fig. 4c and Table S4. The ionic diffusion coefficient of R-NCM was significantly higher than that of W-NCM. This improvement is due to the structural repair of the material, which reestablishes the pathways for Li⁺ transport^[36].

  Figure 4

  

  Figure 4.  Electrochemical performance: (a) CV curves; (b) EIS and the equivalent circuit diagram (Inset); (c) linear fit of Z_re vs ω^-1/2; (d) charge-discharge curves at 0.1C; (e) rate performance at varied current rates; (f, g) cycling performance at 0.2C and 0.5C and schematic diagram of charge and discharge (Inset); (h) capacity retention rates at 0.5C

  In b: R_s: solution resistance, CPE: constant phase angle element, R_ct: charge transfer resistance, Z_w: Warburg impedance.

  下载: 全尺寸图片 幻灯片

  The electrochemical properties of the materials were tested to further explore their potential applications in lithium-ion batteries. We compared the charge-discharge performance of the recycled materials with that of commercial materials and waste materials, where commercial materials include polycrystalline NCM523 materials and single crystal NCM523 materials. Fig. 4d displays the first charge-discharge curves of materials at 0.1C. The discharge curve of W-NCM was obviously steeper, featuring a lower discharge voltage platform and reduced specific capacity. This is due to lithium loss over prolonged cycling and destruction in the material′s crystal structure. The charge-discharge curve of the regenerated material became longer and smoother, resembling that of the new material. The regenerated material achieved a discharge-specific capacity of 168.51 mAh·g^-1, nearly matching the new material of 174.57 mAh·g^-1, and significantly exceeding the spent material′s original discharge-specific capacity of 91.58 mAh·g^-1. This indicates the effective regeneration and restoration of the crystal structure. Fig. 4e displays the discharge performance of the materials at different rates. The discharge-specific capacities of the regenerated materials were significantly higher than W-NCM and comparable to C-NCM, recording discharge-specific capacities of 168.27, 165.83, 155.93, 143.60, and 127.30 mAh·g^-1 at 0.1C, 0.2C, 0.5C, 1C, and 2C, respectively. Upon returning to a rate of 0.1C, the discharge-specific capacity of R-NCM stabilized at about 163.53 mAh·g^-1, demonstrating the regenerated material had good rate performance.

  Fig. 4f demonstrates the cycling stability of C-NCM, W-NCM, and R-NCM at the rate of 0.2C. Initially, R-NCM showed a discharge-specific capacity of 159.19 mAh·g^-1, comparable to C-NCM of 167.34 mAh·g^-1 and significantly exceeding the spent material of 63.00 mAh·g^-1. After 100 cycles, both C-NCM and R-NCM maintained high-capacity retention of 94.30% and 93.55%, respectively, whereas the spent material retained only 49.53%. The Coulombic efficiency of both C-NCM and R-NCM consistently exceeded 97.0%, except during the first cycle. Conversely, the Coulombic efficiency of W-NCM was unstable, dropping to as low as 92.5%. To further explore cycling performance at a higher rate, Fig. 4g illustrates the results at a rate of 0.5C. Initially, R-NCM and C-NCM had discharge capacities of 153.89 and 161.45 mAh·g^-1, respectively. After 100 cycles, these capacities were maintained at 143.26 and 152.47 mAh·g^-1, respectively. In contrast, W-NCM started with a discharge-specific capacity of only 35.89 mAh·g^-1, with a capacity retention rate of 23.32% after 100 cycles. Notably, R-NCM demonstrated higher specific capacity and superior cycling stability at various rates compared to W-NCM. Fig. 4h shows the capacity retention rate of the three samples after different cycles at 0.5C. After 25, 50, and 75 cycles, R-NCM showed capacity retention rates of 99.78%, 99.02%, and 96.29%, respectively, significantly higher than those of W-NCM. After 100 cycles, the capacity retention of R-NCM reached 93.09%, slightly below C-NCM (94.43%). In addition, we tested the charge-discharge performance of commercial single crystal NCM523 and compared it with R-NCM. The SEM and selected area electron diffraction (SAED) results of CS-NCM are shown in Fig.S9, and the charge-discharge performance is shown in Fig.S10. In summary, the discharge specific capacity, cycle stability, and rate performance of the recycled material were well recovered compared to the commercial NCM523, which is related to the low cation mixing and good layered structure order of the recycled material.

  2.5   Comparison with other regeneration methods

  To highlight the advantages of this recycling method, outcomes were compared with other regeneration processes based on specific capacity, leaching rate, long cycle, low-cost, and environmentally friendly. These comparisons are depicted in Fig. 5a. As shown in Fig. 5b and Table S5, cathode materials regenerated through this process demonstrate stronger cycle stability and specific capacity compared to those treated with other regeneration methods. From a cost perspective, this process uses DL-tartaric acid and formic acid as leaching agents. These are not only more cost-effective but also more acidic than other commonly used leaching acids, as shown in Table S6. Fig. 5c, Table S7 and S8 show the cost of raw materials required to regenerate one ton of waste NCM and compare it to other methods^{[26-27, 37]}. The results show that the main cost is the leaching agent, and this process has the advantage of lower cost compared with other organic acid leaching processes due to the small amount of leaching agent. From an environmental perspective, the process uses organic acids that are less corrosive compared to inorganic acids, making them more environmentally friendly. The organic acids used are more biodegradable than inorganic acids, which can significantly contaminate water and soil if not properly managed. A comparative analysis of various leaching processes is detailed in Fig. 5d and Table S9. The analysis shows that the leaching efficiency of this process is comparable to that of inorganic acids, both of which are above 97%, which is slightly higher than that of some other similar organic acid leaching processes.

  Figure 5

  

  Figure 5.  (a) Comparison diagram of this work with other regeneration processes; Comparison of (b) electrochemical performance, (c) cost, and (d) leaching rate in this study with those reported in the literature

  下载: 全尺寸图片 幻灯片

  3.   Conclusions

  In conclusion, a sustainable, cost-effective, scalable, and eco-friendly regeneration method has been developed for recycling spent NCM cathode materials. This regeneration process includes leaching in a DL-tartaric acid-formic acid system, co-precipitation with oxalic acid, and a high-temperature solid-phase synthesis. Using 0.5 mol·L^-1 DL-tartaric acid and 2% formic acid, leaching efficiencies achieved were 98.03%, 98.69%, 97.59%, and 98.50% for Li, Ni, Co, and Mn, respectively, under optimal conditions of a 90 ℃ reaction temperature, a 13 g·L^-1 solid-liquid ratio, and a 70-minute reaction time. Characterizations demonstrated that the regenerated NCM cathode materials retained their layered structure exhibited high crystallinity and uniform particle distribution, and were essentially free of impurities. The first discharge capacity of the regenerated NCM cathode material reached 168.5 mAh·g^-1 at 0.1C and the capacity retention rate after 100 cycles was 93.09% at 0.5C. This is an efficient and cost-effective solution for the recycling of waste LIBs.

  Supporting information is available at http://www.wjhxxb.cn

  
  Acknowledgments: This work was supported by the National Natural Science Foundation of China (Grant No.52201258), the Natural Science Foundation of Jiangsu Province (Grants No.BK20210650, BK20210651) and the Natural Science Foundation of Jiangsu Colleges and Universities (Grant No.21KJB430003). JIN Yachao and SONG Li were supported by the Startup Foundation for Introducing Talent of NUIST (Grant No.2021r047). At the same time, this work has been supported by the Cyan Project of Jiangsu Colleges and Universities, the Key Discipline Construction Project of Jiangsu Universities, the Joint Laboratory of Air Pollution Control of Jiangsu Province, and the Engineering Technology Research Center of Environmental Clean Materials of Jiangsu Province.
• 1. [1]

     GOODENOUGH J B, PARK K S. The Li-ion rechargeable battery: A perspective[J]. J. Am. Chem. Soc., 2013, 135(4):  1167-1176. doi: 10.1021/ja3091438

  2. [2]

     MÄRKER K, REEVES P J, XU C, GRIFFITH K J, GREY C P. Evolution of structure and lithium dynamics in LiNi_0.8Mn_0.1Co_0.1O₂ (NMC811) cathodes during electrochemical cycling[J]. Chem. Mater., 2019, 31(7):  2545-2554. doi: 10.1021/acs.chemmater.9b00140

  3. [3]

     ZHANG W H, LIANG L W, ZHAO F, LIU Y, HOU L R, YUAN C Z. Ni-rich LiNi_0.8Co_0.1Mn_0.1O₂ coated with Li-ion conductive Li₃PO₄ as competitive cathodes for high-energy-density lithium ion batteries[J]. Electrochim. Acta, 2020, 340:  135871. doi: 10.1016/j.electacta.2020.135871

  4. [4]

     NATARAJAN S, BORICHA A B, BAJAJ H C. Recovery of value-added products from cathode and anode material of spent lithium-ion batteries[J]. Waste Manage., 2018, 77:  455-465. doi: 10.1016/j.wasman.2018.04.032

  5. [5]

     AHMED S, NELSON P A, GALLAGHER K G, SUSARLA N, DEES D W. Cost and energy demand of producing nickel manganese cobalt cathode material for lithium ion batteries[J]. J. Power Sources, 2017, 342:  733-740. doi: 10.1016/j.jpowsour.2016.12.069

  6. [6]

     WANG W Q, ZHANG Y C, ZHANG L, XU S M. Cleaner recycling of cathode material by in-situ thermite reduction[J]. J. Clean. Prod., 2020, 249:  119340. doi: 10.1016/j.jclepro.2019.119340

  7. [7]

     HE L P, SUN S Y, SONG X F, YU J G. Leaching process for recovering valuable metals from the LiNi_1/3Co_1/3Mn_1/3O₂ cathode of lithium-ion batteries[J]. Waste Manage., 2017, 64:  171-181. doi: 10.1016/j.wasman.2017.02.011

  8. [8]

     YANG Y, XU S M, HE Y H. Lithium recycling and cathode material regeneration from acid leach liquor of spent lithium-ion battery via facile co-extraction and co-precipitation processes[J]. Waste Manage., 2017, 64:  219-227. doi: 10.1016/j.wasman.2017.03.018

  9. [9]

     JOULIE M, LAUCOURNET R, BILLY E. Hydrometallurgical process for the recovery of high value metals from spent lithium nickel cobalt aluminum oxide based lithium-ion batteries[J]. J. Power Sources, 2014, 247:  551-555. doi: 10.1016/j.jpowsour.2013.08.128

  10. [10]

      LEE C K, RHEE K I. Reductive leaching of cathodic active materials from lithium ion battery wastes[J]. Hydrometallurgy, 2003, 68(1):  5-10.

  11. [11]

      CHEN X P, FAN B L, XU L P, ZHOU T, KONG J R. An atom-economic process for the recovery of high value-added metals from spent lithium-ion batteries[J]. J. Clean. Prod., 2016, 112:  3562-3570. doi: 10.1016/j.jclepro.2015.10.132

  12. [12]

      LI L, GE J, WU F, CHEN R J, CHEN S, WU B R. Recovery of cobalt and lithium from spent lithium ion batteries using organic citric acid as leachant[J]. J. Hazard. Mater., 2010, 176(1):  288-293.

  13. [13]

      LI L, QU W J, ZHANG X X, LU J, CHEN R J, WU F, AMINE K. Succinic acid-based leaching system: A sustainable process for recovery of valuable metals from spent Li-ion batteries[J]. J. Power Sources, 2015, 282:  544-551. doi: 10.1016/j.jpowsour.2015.02.073

  14. [14]

      ZENG X L, LI J H, SHEN B Y. Novel approach to recover cobalt and lithium from spent lithium-ion battery using oxalic acid[J]. J. Hazard. Mater., 2015, 295:  112-118. doi: 10.1016/j.jhazmat.2015.02.064

  15. [15]

      YU J C, MA B Z, QIU Z J, WANG C Y, CHEN Y Q. Separation and recovery of valuable metals from ammonia leaching solution of spent lithium-ion batteries[J]. ACS Sustain. Chem. Eng., 2023, 11(26):  9738-9750. doi: 10.1021/acssuschemeng.3c01714

  16. [16]

      GUPTA S, PANT K K, CORDER G. An environmentally benign closed-loop process for the selective recovery of valuable metals from industrial end-of-life lithium-ion batteries[J]. Chem. Eng. J., 2022, 446:  137397. doi: 10.1016/j.cej.2022.137397

  17. [17]

      CHENG J M, ZHENG C, XU K, ZHU Y C, SONG Y, JING C Y. Sequential separation of critical metals from lithium-ion batteries based on deep eutectic solvent and electrodeposition[J]. J. Hazard. Mater., 2024, 465:  133157. doi: 10.1016/j.jhazmat.2023.133157

  18. [18]

      MONTOYA A T, YANG Z Z, DAHL E U, PUPEK K Z, POLZIN B, DUNLOP A, VAUGHEY J T. Direct recycling of lithium-ion battery cathodes: A multi-stage annealing process to recover the pristine structure and performance[J]. ACS Sustain. Chem. Eng., 2022, 10(40):  13319-13324. doi: 10.1021/acssuschemeng.2c02643

  19. [19]

      HE T, DAI J J, DONG Y T, ZHU F S, WANG C, ZHEN A G, CAI Y R. Green closed-loop regeneration of ternary cathode materials from spent lithium-ion batteries through deep eutectic solvent[J]. Ionics, 2023, 29(5):  1721-1729. doi: 10.1007/s11581-023-04967-3

  20. [20]

      SHI Y, CHEN G, LIU F, YUE X J, CHEN Z. Resolving the compositional and structural defects of degraded LiNi_xCo_yMn_zO₂ particles to directly regenerate high-performance lithium-ion battery cathodes[J]. ACS Energy Lett., 2018, 3(7):  1683-1692. doi: 10.1021/acsenergylett.8b00833

  21. [21]

      HE L P, SUN S Y, MU Y Y, SONG X F, YU J G. Recovery of lithium, nickel, cobalt, and manganese from spent lithium-ion batteries using L-tartaric acid as a leachant[J]. ACS Sustain. Chem. Eng., 2017, 5(1):  714-721. doi: 10.1021/acssuschemeng.6b02056

  22. [22]

      TIAN Y L, AO X Q, YANG M, YANG Y C, WEI J Y, WANG F Y. Study on the leaching of lithium from lithium-poor clay-type ore using tartaric acid by calcination and water leaching[J]. Hydrometallurgy, 2024, 227:  106335. doi: 10.1016/j.hydromet.2024.106335

  23. [23]

      ZHENG Y, SONG W, MO W T, ZHOU L, LIU J W. Lithium fluoride recovery from cathode material of spent lithium-ion battery[J]. RSC Adv., 2018, 8(16):  8990-8998. doi: 10.1039/C8RA00061A

  24. [24]

      GAO W F, ZHANG X H, ZHENG X H, LIN X, CAO H B, ZHANG Y, SUN Z. Lithium carbonate recovery from cathode scrap of spent lithium-ion battery: A closed-loop process[J]. Environ. Sci. Technol., 2017, 51(3):  1662-1669. doi: 10.1021/acs.est.6b03320

  25. [25]

      LI L, BIAN Y F, ZHANG X X, GUAN Y B, FAN E S, WU F, CHEN R J. Process for recycling mixed-cathode materials from spent lithium-ion batteries and kinetics of leaching[J]. Waste Manage., 2018, 71:  362-371. doi: 10.1016/j.wasman.2017.10.028

  26. [26]

      NING P C, MENG Q, DONG P, DUAN J G, XU M L, LIN Y, ZHANG Y J. Recycling of cathode material from spent lithium ion batteries using an ultrasound-assisted DL-malic acid leaching system[J]. Waste Manage., 2020, 103:  52-60.

  27. [27]

      REFLY S, FLOWERI O, MAYANGSARI T R, SUMBOJA A, SANTOSA S P, OGI T, ISKANDAR F. Regeneration of LiNi_1/3Co_1/3Mn_1/3O₂ cathode active materials from end-of-life lithium-ion batteries through ascorbic acid leaching and oxalic acid coprecipitation processes[J]. ACS Sustain. Chem. Eng., 2020, 8(43):  16104-16114.

  28. [28]

      QIN Z Y, WEN Z X, XU Y F, ZHENG Z C, BAI M L, ZHANG N, JIA C K, WU H B, CHEN G. A ternary molten salt approach for direct regeneration of LiNi_0.5Co_0.2Mn_0.3O₂ cathode[J]. Small, 2022, 18(43):  2106719.

  29. [29]

      JUNG S K, GWON H, HONG J, PARK K Y, SEO D H, KIM H, HYUN J, YANG W, KANG K. Understanding the degradation mechanisms of LiNi_0.5Co_0.2Mn_0.3O₂ cathode material in lithium ion batteries[J]. Adv. Energy Mater., 2014, 4(1):  1300787.

  30. [30]

      JIANG G H, ZHANG Y N, MENG Q, ZHANG Y J, DONG P, ZHANG M Y, YANG X. Direct regeneration of LiNi_0.5Co_0.2Mn_0.3O₂ cathode from spent lithium-ion batteries by the molten salts method[J]. ACS Sustain. Chem. Eng., 2020, 8(49):  18138-18147.

  31. [31]

      HE L P, SUN S Y, YU J G. Performance of LiNi_1/3Co_1/3Mn_1/3O₂ prepared from spent lithium-ion batteries by a carbonate co-precipitation method[J]. Ceram. Int., 2018, 44(1):  351-357.

  32. [32]

      CHEN Z, WANG J, CHAO D L, BAIKIE T, BAI L Y, CHEN S, ZHAO Y L, SUM T C, LIN J Y, SHEN Z X. Hierarchical porous LiNi_1/3Co_1/3Mn_1/3O₂ nano-/micro spherical cathode material: Minimized cation mixing and improved Li⁺ mobility for enhanced electrochemical performance[J]. Sci. Rep., 2016, 6(1):  25771.

  33. [33]

      GAO H P, YAN Q Z, TRAN D, YU X L, LIU H D, LI M Q, LI W K, WU J L, TANG W, GUPTA V, LUO J, CHEN Z. Upcycling of spent LiNi_0.33Co_0.33Mn_0.33O₂ to single-crystal Ni-rich cathodes using lean precursors[J]. ACS Energy Lett., 2023, 8(10):  4136-4144.

  34. [34]

      HUANG C, XIA X, CHI Z W, YANG Z H, HUANG H J, CHEN Z X, TANG W J, WU G Q, CHEN H Y, ZHANG W X. Preparation of single-crystal ternary cathode materials via recycling spent cathodes for high performance lithium-ion batteries[J]. Nanoscale, 2022, 14(27):  9724-9735.

  35. [35]

      WANG J, LI D Q, ZENG W H, CHEN X Y, ZHANG Y X, ZHANG S J, LI Z P, LI C H, MU S C. Degradation mechanism, direct regeneration and upcycling of ternary cathode material for retired lithium-ion power batteries[J]. J. Energy Chem., 2025, 102:  534-554.

  36. [36]

      QI C, WANG S H, ZHU X K, ZHANG T, GOU Y J, XIE Z X, JIN Y C, WANG Y, SONG L, ZHANG M D. Environmental-friendly low-cost direct regeneration of cathode material from spent LiFePO₄[J]. J. Alloy. Compd., 2022, 924:  166612.

  37. [37]

      ZHENG R J, WANG W H, DAI Y K, MA Q X, LIU Y L, MU D Y, LI R H, REN J, DAI C S. A closed-loop process for recycling LiNi_xCo_yMn_(1-x-y)O₂ from mixed cathode materials of lithium-ion batteries[J]. Green Energy Environ., 2017, 2(1):  42-49.

• 

  Figure 1  Diagram of the waste NCM regeneration process

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) XRD patterns of C-NCM, R-NCM, and W-NCM, and (b, c) their enlarged views in detail; High-resolution XPS spectra of (d) Ni2p, (e) Co2p, and (f) Mn2p of C-NCM, R-NCM, and W-NCM

  下载: 全尺寸图片 幻灯片
  

  Figure 3  SEM images of (a) W-NCM, (b) C-NCM, and (c) R-NCM; (d-f) TEM and HR-TEM images of R-NCM; (g) HADDF-TEM and elemental mapping images of R-NCM

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Electrochemical performance: (a) CV curves; (b) EIS and the equivalent circuit diagram (Inset); (c) linear fit of Z_re vs ω^-1/2; (d) charge-discharge curves at 0.1C; (e) rate performance at varied current rates; (f, g) cycling performance at 0.2C and 0.5C and schematic diagram of charge and discharge (Inset); (h) capacity retention rates at 0.5C

  In b: R_s: solution resistance, CPE: constant phase angle element, R_ct: charge transfer resistance, Z_w: Warburg impedance.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Comparison diagram of this work with other regeneration processes; Comparison of (b) electrochemical performance, (c) cost, and (d) leaching rate in this study with those reported in the literature

  下载: 全尺寸图片 幻灯片
•