CHARGE MATERIAL SYNTHESIZED FROM RECYCLED LITHIUM-ION BATTERIES

Abstract

Lithium-ion battery (LIB) recycling is considered as an important component to industry sustainability. A massive number of LIBs in portable electronics, electric vehicles and grid storage will eventually end up in wastes, leading to serious economic and environmental problems. Hence, tremendous effort has been made to improve hydrometallurgical recycling process since it is the most promising option for handling end-of-life LIBs owing to its wide applicability, low cost and high productivity. Despite these advantages, some extra elements (Al, Fe, C, F, etc.) remain as impurities in the removal process and remain in the solution, presenting a challenge to obtaining high-quality cathode material. This approach demonstrates the improved electrochemical performance by adding potential impurities in the leaching solution.

Claims

1. A method of generating a charge material precursor having an preferred oxidation state for secondary battery charge material, comprising: forming a leach solution including target charge materials by leaching the target charge materials from a recycling stream in a leach solution; controlling a pH of the leach solution for dissolving the target charge materials in the leach solution; establishing a predetermined percentage of halides in the leach solution; and adding a strong base to the leach solution for precipitating charge material particles; the charge material particles including the target charge materials in a charge material precursor form for subsequent sintering with a Li compound.

2. The method of claim 1 wherein the target charge materials include compounds of at least one of Ni, Mn, Co, and Al.

3. The method of claim 1 wherein the halides include fluoride.

4. The method of claim 1 wherein the halides result in holes in the precipitated charge material particles, the holes increasing a ratio of 2+ to 3+ oxidation states of the charge material precursor.

5. The method of claim 3 wherein a predetermined percentage of fluoride resulting from addition of fluorine to the leach solution prior to precipitation of the charge material increases a percentage of Ni.sup.2+ ions over Ni.sup.3+ ions on a surface of the charge materials following subsequent sintering.

6. The method of claim 3 wherein a predetermined percentage of fluoride resulting from addition of fluorine to the leach solution prior to precipitation of the charge material increases a percentage of Co.sup.2+ ions over Co.sup.3+ ions on a surface of the charge material particles following subsequent sintering.

7. The method of claim 1 wherein the leach solution has a range of 0.2-5 at % fluoride prior to precipitation of the charge material and increases a percentage of Ni.sup.2+ ions on a surface of the charge material particles to between 40.1%-43.8%.

8. The method of claim 1 wherein the leach solution has a range of 0.2-5 at % fluoride prior to precipitation of the charge material and increases a percentage of Co.sup.2+ ions on a surface of the charge material particles to between 13.0%-35.2%.

9. The method of claim 1 further comprising forming the leach solution to include the target charge materials to result in a NCM622 charge material precursor and the halides consist of fluoride at a predetermined percentage.

10. The method of claim 1 further comprising forming the leach solution from a leach acid selected from the group consisting of sulfuric acid, hydrochloric acid, nitric acid, phosphoric acid, boric acid, hydrofluoric acid, hydrobromic acid, hydroiodic acid and perchloric acid.

11. The method of claim 1 wherein the strong base includes precipitation agents selected from the group consisting of sodium hydroxide or potassium hydroxide.

12. The method of claim 1 further comprising determining the predetermined percentage of halides in the leach solution for achieving an oxidation state of one or more of the target charge materials.

13. A charge material precursor having: nickel, manganese and cobalt (NMC) particles in a precursor form responsive to sintering with a lithium compound for forming an active charge material; a fluorine impurity in the NMC particles of between 0at % and 5.0at %; holes in the particles resulting from the fluorine impurity; and each of the particles defining a structure with a surface, the surface having nickel ions, at least 40.1% of the Ni surface ions having an oxidation state of +2.

14. The charge material precursor of claim 13 wherein the surface has cobalt ions, at least 13% of the cobalt surface ions having an oxidation state of +2.

15. The charge material precursor of claim 13 wherein the fluorine increases a ratio of Ni.sup.2+ ions over Ni.sup.3+ ions by 10%.

16. The charge material precursor of claim 14 wherein the fluorine increases a ratio of Co.sup.2+ ions to 35.2%.

17. A method of controlling an oxidation state in recycled secondary battery charge material, comprising: leaching NMC (Nickel, Manganese, Cobalt) charge materials from a recycling stream in a leach solution; controlling a pH of the leach solution through addition of a leach acid for dissolving the charge materials in the leach solution, the leach solution having a pH; adjusting the pH to a pH range for precipitating impurities for removal while leaving the charge materials in solution, the impurities including iron and copper; following the impurity removal, establishing a predetermined percentage of fluoride; and adding a strong base for precipitating charge material particles; the charge material particles including the NMC in a charge material precursor form for subsequent sintering with a Li compound, the precipitated charge materials having an oxidation state based on the predetermined percentage.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The foregoing and other features will be apparent from the following description of particular embodiments disclosed herein, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views.

[0013] The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention.

[0014] FIG. 1 is a context diagram of a hydrometallurgical recycling approach suitable for use with configurations herein;

[0015] FIG. 2 shows the precursor morphology at different times in which the average particle diameter and sphericity increases;

[0016] FIG. 3 shows the SEM images of synthesized NCM622 samples from precursor;

[0017] FIGS. 4A-4D show phase and structure of prepared materials are analyzed by powder X-ray diffraction with patterns;

[0018] FIGS. 5A-5D show refinement profiles of VNCM, 0.2FNCM, 1FNCM and 5FNCM cathodes of the materials of FIGS. 4A-4D;

[0019] FIG. 6A-6C show spectra of surface scans of the cathode materials;

[0020] FIGS. 7A-7D show deconvolution scans and percentages of Ni and Co surface ions on the cathode material particles;

[0021] FIGS. 8A-8C show a plot of fitting lines between peak current (Ip2) vs. scan rate (v);

[0022] FIG. 9 shows li-ion diffusion within different cathode crystal structures; and

[0023] FIGS. 10A-10C are schematic diagrams showing formed holes based on the doping/impurities.

DETAILED DESCRIPTION

[0024] Depicted below is an example method and apparatus for recycling batteries such as lithium ion batteries. The proposed approach is an example and is applicable to other lithium and non-lithium batteries for recycling spent batteries and recovering active cathode material suitable for use in new batteries.

[0025] FIG. 1 is a context diagram of a hydrometallurgical recycling approach suitable for use with configurations herein. The cathode materials widely used in commercial lithium ion batteries include LiCoO.sub.2, LiMn.sub.2O.sub.4, LiNiO.sub.2, LiNi.sub.xCo.sub.yAl.sub.z O.sub.2, LiNi.sub.xMn.sub.yCo.sub.zO.sub.2 and LiFePO.sub.4. In order to recycle lithium ion batteries effectively, it is beneficial to consider all the various battery chemistries. Thus, it is beneficial to develop a simpler and environmentally acceptable recycling process generally applicable to various widely used LIBs. Configurations disclosed herein present an example to extract compounds including the desirable elements of Co, Ni, Mn, and Li from mixed cathode materials and utilize the recycled materials to produce active materials for batteries. Alternate chemistries may be recycled using the methods disclosed.

[0026] Referring to FIG. 1, at step 1, discharged Li ion batteries are crushed/shredded. Mechanical separation processes are applied as a pretreatment to separate the outer cases and shells and the plastic fraction, as shown at step 1a. The recycling stream is a comingled mass including at least some of the charge materials, e.g. Ni, Mn, Co of the target charge materials, meaning the battery chemistry and ratio of the intended recycled battery, such as NMC 622, NMC 532, NMC 111 or NMC 811, for example.

[0027] The sieved cathode powder will be leached by 4-5M sulfuric acid (H.sub.2SO.sub.4) and a reducing agent such as 29-32% hydrogen peroxide for about 2-3 hours at 70-80° C., as depicted at step 2. Other concentrations of the leach acid may also be employed, as well as alternate reducing agents. In alternate arrangements, the leach solution may be formed from a leach acid including one or more of sulfuric acid, hydrochloric acid, nitric acid, phosphoric acid, boric acid, hydrofluoric acid, hydrobromic acid, hydroiodic acid and perchloric acid. Addition of hydrogen peroxide H.sub.2O.sub.2 changes not only Fe2+ to Fe3+, but also other metal ions Mn, Ni, Co to 2+, thus leading to separate iron by controlling pH of the solution in step 3. After filtration, residual LiFeO.sub.4 and carbon can be separated by centrifugation, as shown at step 2a. Other impurities are also removed from the surface of the solution, as shown at step 2b.

[0028] The metallic elements of interest are transferred to the aqueous solution as the crushed raw cathode materials form a granular mass used to generate the solution of aggregate battery materials from the spent cells, as depicted at step 3. This includes the target charge materials of Co (cobalt), Ni (nickel), Mn (manganese); other target charge materials may be employed using the present approach with alternate battery chemistries. The pH is adjusted to extract iron, copper and aluminum as Fe(OH).sub.3, Cu(OH).sub.2 and Al(OH).sub.3. This involves adjusting the pH to a range between 3.0-7.0. Accordingly, an NaOH solution is added to adjust pH number to deposit Fe(OH).sub.3, Cu(OH).sub.2 and Al(OH).sub.3 which have a lower solubility constant, and maintain Mn.sup.2+, Co.sup.2+, Ni.sup.2+ in the solution, then Fe(OH).sub.3, Cu(OH).sub.2 and Al(OH).sub.3 are separated by filtration. It should be noted that the above processes include maintaining the solution at a temperature between 40 deg. C. and 80 deg. C, thus avoiding high heat required in conventional approaches.

[0029] The pH adjustment for impurity removal generally adjusts the pH based on particular impurities sought for removal. A pH range of around 3-5, for example, causes iron to precipitate out of solution. Copper tends to precipitate at a pH around 5-7. Depending on the quantity and type of impurities in the recycling stream, the impurity removal phase may target different pH ranges. Once the charge material metals (Ni, Mn, Co) have been dissolved, the pH will have been reduced to around 1-3 depending on the quantity and concentration of the acid leach. Before the pH is raised to precipitate the NCM in a pure form, impurities can be precipitated, optionally through addition of a reductant, discussed further below. In other words, while adding sodium hydroxide or other strong base for raising the pH, the impurities precipitate out at a lower pH, followed by precipitation of NCM hydroxides at a higher pH. The strong base may also include precipitation agents such as mixtures of sodium hydroxide or potassium hydroxide.

[0030] The target charge materials are now dissolved in the solution. Based on the predetermined target ratio of the target charge materials, the solution is adjusted to achieve the predetermined ratio of target charge materials. In the example approach, this is a 6:2:2 combination of nickel, cobalt and manganese, although any suitable ratio could be employed. Therefore, adjusting the solution includes identifying a desired ratio of the charge materials for use in recycled cathode material resulting from the generated solution, and adding virgin materials to achieve the desired ratio. Adding the raw materials includes adding additional quantities of the target charge materials for achieving the desired ratio without separating the individual target charge materials already in solution form, therefore the mixed target charge materials (Co, Mn, Ni) do not need to be separately drawn or extracted as in conventional approaches, which usually involve high heat to break the molecular bonds of the compounds.

[0031] In other words, the pH is raised first to extract elements considered “impurities,” meaning those other than the charge materials. In an NCM chemistry, this would include Fe, Cu, Al and other trace materials. Al may in fact be part of the battery chemistry in other battery formulations. Once impurities are extracted, the pH is further increased to precipitate the target charge materials. In configurations herein, halides such as fluorine are formed from either a residual quantity from the recycling stream, and/or addition of additional materials as a doping agent.

[0032] To achieve the intended chemistry, or ratio of the target charge materials, the concentration of Mn.sup.2+, Co.sup.2+, Ni.sup.2+ in the solution is tested, and adjusted the ratio of them to 6:2:2 or other suitable ratio with additional CoSO.sub.4, NiSO.sub.4, MnSO.sub.4. A NaOH solution is added to increase the pH to around 11, usually within a range of 10.0-13, thus adjusting a pH of the solution such that the target charge materials for the new (recycled) charge materials precipitate. Ni.sub.1/3Mn.sub.1/3Co.sub.1/3(OH).sub.2 or Ni.sub.1/3Mn.sub.1/3Co.sub.1/3O(OH) or a mixture thereof can be coprecipitated such that the respective mole ratio is 1:1:1, as depicted at step 4. Ni.sub.xMn.sub.yCo.sub.z(OH).sub.2 or Ni.sub.xMn.sub.yCo.sub.zO(OH) or a mixture with different ratios of x, y, and z can also be precipitated. Na.sub.2CO.sub.3 is added in the solution to deposit Li.sub.2CO.sub.3, as depicted at step 5. Finally, the recovered Ni.sub.1/3Mn.sub.1/3Co.sub.1/3O(OH).sub.2 and Li.sub.2CO.sub.3 are sintered to produce the cathode material.

[0033] In the example arrangement, the target charge materials include manganese (Mn), cobalt (Co), and nickel (Ni) extracted from charge material of the spent battery cells, in which the target charge materials remain commingled in the solution during precipitation. Adjusting the pH includes adding a substance, such as NaOH (sodium hydroxide, also referred to as lye or caustic soda) for raising the pH such that the target charge materials precipitate, however any suitable substance for raising the pH may be employed. The end result is that adjusting the pH includes adding sodium hydroxide for raising the pH to permit precipitation of the target charge materials for use as a cathode precursor material without separately precipitating the individual compounds defining the target charge materials. The intermediate, or precursor form will result in a lithium oxide form following sintering with lithium carbonate Li.sub.2CO.sub.3.

[0034] Na.sub.2CO.sub.3 is added in the solution to deposit Li.sub.2CO.sub.3 at about 40° C. After filtrating, Li.sub.2CO.sub.3 can be recycled as the starting material to synthesize the active cathode material Ni.sub.1/3Mn.sub.1/3Co.sub.1/3O.sub.2, as shown at steps 5 and 5a. Therefore, the method adds back the lithium to the precipitated target charge materials to form active cathode material suitable for the new battery, and precipitates the target charge material in the predetermined ratio to form charge material for a new battery.

[0035] The coprecipitated materials Ni.sub.1/3Mn.sub.1/3Co.sub.1/3(OH).sub.2 or Ni.sub.1/3Mn.sub.1/3Co.sub.1/3O(OH) or their mixture and recovered Li.sub.2CO.sub.3, with additional Li.sub.2CO.sub.3 in molar ratio 1.1 of Li versus M (M=Ni.sub.1/3Mn.sub.1/3Co.sub.1/3), are mixed and grinded in mortar, as depicted at step 6. The mixture may be reformulated by any suitable processing to form the active cathode material 134 for new batteries 140. In the example approach, the mixture was sintered at 900 for 15 hours. The reaction product may be ground into powder for subsequent distribution and reformation into new cells 140. The Ni.sub.1/3Mn.sub.1/3Co.sub.1/3O.sub.2 is sintered by a high temperature solid-state method at 900° C. for 15 hours.

[0036] Extending the approach of FIG. 1, cathode materials were synthesized by co-precipitation reaction and post-sintering process, in which transition metal ions M.sup.2+ (M =Ni, Co, Mn) coprecipitate to form Ni.sub.0.6Co.sub.0.2Mn.sub.0.2(OH).sub.2 precursor in an alkaline solution, then a mixture of dried precursors and lithium salt was sintered to obtain LiNi.sub.0.6Co.sub.0.2Mn.sub.0.2O.sub.2 (NCM622) as the final cathode samples. Examples below illustrate a leach solution to include the target charge materials to result in a 622 NMC charge material precursor and the halides consist of fluoride at a predetermined percentage of substantially around 1.0 at %. Additional configurations may establish a predetermined percentage of halides in the leach solution based on an oxidation state of one or more of the target charge materials, as intended to be formed on the resulting charge material particles. The halides are based on halogens and an electronegativity of another atom. Other suitable halides may be employed, including some of the more common ions: chloride (Cl—), bromide (Br—), iodide (I—) and astatide (At—).

[0037] Metal sulfate hydrates NiSO.sub.4.Math.6H.sub.2O, CoSO.sub.4.Math.7H.sub.2O and MnSO.sub.4.Math.H.sub.2O were added into deionized (DI) water at a stoichiometric ratio of 6:2:2 to obtain 2 M metal sulfate solution. For impurity-involved sample preparation, NaF salt was mixed into the metal sulfate solution with different concentrations of 0.2 at %, 1 at % and 5 at %, and designated as 0.2FNCM, 1FNCM and 5FNCM respectively, hereinafter. Prior to commencement, 0.5 M NH.sub.3.Math.H.sub.2O was filled into a 5 L continuous stirred-tank reactor (CSTR) as a complexing agent. Then, the metal sulfate solution and ammonia solution were pumped into the reactor with a controlled flow rate. Meanwhile, 7.5 M NaOH solution was added into the reactor by a peristaltic pump to keep the reaction at a desired pH condition. The co-precipitation reaction was operated continuously for 12 hours at a steady condition of pH=11 and T=55° C. under nitrogen protection. After reaction, synthesized precursors were filtered and washed with DI water to remove residuals until pH reduced to 7, then precursor powders were dried in oven at 130° C. overnight. To obtain cathode materials, precursors were mixed with Li.sub.2CO.sub.3 in a mortar at a stoichiometric ratio of 1:1.05 (5% excess of lithium salt was used in order to compensate for the loss of lithium ions during sintering). Next, the mixture underwent a two-step sintering process: (I) heated up to 450° C. for 5 h then cooled down to room temperature, (II) heated up to 850° C. for 18 h then followed by a same cooling method (ramp rate was fixed to 2° C. per minute). In the end, a total number of four cathode materials were obtained: LiNi.sub.0.6Co.sub.0.2Mn.sub.0.2O.sub.2 (VNCM), and LiNi.sub.0.6Co.sub.0.2Mn.sub.0.2F.sub.xO.sub.2, (x=0.002, 0.01 and 0.05, named as 0.2FNCM, 1FNCM and 5FNCM, respectively).

[0038] The particle morphologies and microstructures were observed by scanning electron microscopy (SEM, JEOL JSM-7000F) coupled with energy-dispersive spectroscopy (EDS) to analyze element composition. Inductive coupled plasma mass spectroscopy (ICP-MS) was utilized to measure the precise concentrations of all metallic elements in the cathodes. The phases of each sample were identified by powder X-ray diffraction (XRD, PANalytical Empyrean) with Cu Kα (λ=1.54 Å) and a step size of 0.02° per scan. To acquire lattice parameters, the following Rietveld refinement was conducted via FullProf software and the powder structure of LiNi.sub.0.6Co.sub.0.2Mn.sub.0.2O.sub.2 (PDF #00-066-0854) was chosen as the reference model. X-ray photoelectron spectroscopy (XPS) data for all cathodes was acquired using PHI 5000 VersaProbe II system (Physical Electronics) to investigate element composition and the oxidation state of transition metals. The X-ray source was operating at 25 W equipped with monochromatic Al Kα (hυ=1486.6 eV) which was set at Ar.sup.+ ion and electron beam sample neutralization, fixed analyzer transmission mode. The XPS spectra were calibrated against C—C at 284.8 eV before subsequent deconvolution simulation (spectra were fitted to multiple Gaussian-Lorentz peaks) which was processed on XPSpeak software.

[0039] CR-2032 coin cells, which consist of cathode electrode, separator, electrolyte and lithium metal as the anode, were assembled in an argon-gas filled glovebox (H.sub.2O, O.sub.2<1 ppm) to study the electrochemical performance. The electrodes were prepared by mixing active materials (cathode powder, 80 wt %), conductive carbon (C65, 10 wt %) and polyvinylidene fluoride binder (PVDF, 10 wt %) in N-methyl-2-pyrrolidone (NMP) solvent to form a slurry. The mixed slurry was then cast on an aluminum foil by a doctor blade (MTI) and dried at 60° C. overnight. Round electrodes (Φ=12 mm) were calendared and punched from dried electrode sheet to get a final thickness˜40 μm with an active mass loading of 3.5˜4.0 mg/cm.sup.2. The electrode samples were further dried at 120° C. overnight in vacuum oven to remove residual NMP and moisture. A trilayer polypropylene-polyethylene-polypropylene membrane (Φ=16 mm) and 1 M LiPF.sub.6 in ethylene carbonate/ethyl methyl carbonate (EC/EMC, 3:7 in wt %) were used as the separator and electrolyte, respectively. The electrochemical performance was tested on a Land battery testing system (LAND, CT2001A) between 3.0V and 4.3V (vs. Li/Li.sup.+). Specifically, cells were measured under current density of 0.33 C and 5C in cycle test, while the current densities were set from 0.1 C up to 5 C in rate performance test (1C=175 mAh/g). Tests of cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were conducted on an electrochemical analyzer (Bio-Logic SAS, VMP3) using the EC-Lab program. In CV test, scanning potential was set between 3.0V and 4.5V (vs. Li/Li.sup.+) at a rate of 0.1 mV/s. In EIS test, cells before and after cycles were measured within the frequency range from 100 kHz to 10 mHz at an amplitude of 10 mv.

[0040] To monitor the morphology of hydroxide precursors during the synthesis, sample was collected every 3 hours, then analyzed with SEM. FIG. 2 shows the precursor morphology at different times in which the average particle diameter and sphericity increases as the reaction time gets longer. At each specified time, no clear difference is observed between samples from 0.2FNCM (row 201), 1FNCM, at row 202, and 5FNCM at row 203, indicating that the presence of fluoride ions within alkaline environment during co-precipitation does not form extra precipitates to inhibit particle growth as well as to impede particle sphericity. The particle morphology has a close correlation with its physical properties, powder tap density for instance. In Table I, the tap density also rises as the reaction goes and all FNCM samples share a similar density level at each time section. It is noted that compared to FNCMs, virgin group has a lower powder tap density at 3 hrs. (col. 211)., 6 hrs. (212) and 9 hrs. (213) which is about 77%, 79% and 89% of the average value of precursors affected by fluorine impurity. However, when the precipitation process reaches to the end at 12 hours (214), all samples approach a similar density level (˜1.90 g/ml). This phenomenon suggests that fluorine impurity has influenced the co-precipitation process, due to an increased powder tap density that caused by faster nucleation and particle growth at its early stages, while brings no prominent changes in the particle morphology and tap density when the reaction comes to steady state.

TABLE-US-00001 TABLE I Sample 3 hrs 6 hrs 9 hrs 12 hrs VNCM 0.68 1.15 1.52 1.85 0.2FNCM 0.94 1.47 1.73 1.93 1FNCM 0.92 1.47 1.72 1.93 5FNCM 0.78 1.41 1.65 1.88

[0041] FIG. 3 shows the SEM images of synthesized NCM622 samples from precursor. In FIG. 3, SEM images of 0.2FNCM (row 301), 1FNCM (row 302) and 5FNCM (row 303) are shown for precursors (column 211), cathodes (column 312) and cathode cross-sectionals (column 313) (scale bar 10 μm). (Insets: images at higher magnification with scale bar 2 μm). A typical near spherical secondary particle composed of numerous tiny flake-like primary particles is observed in col. 311. Primary particles are the most fundamental building blocks of synthesized materials. During the reaction, the growth, attachment and aggregation of primary particles produces the spherical secondary particles. For all cathodes shown in col. 312, the shape of the primary particles turns from flake to polygon after high-temperature precursor sintering. Both precursors and cathodes have a consistent surface morphology as well as secondary particle dimension (˜8 μm) which implies that fluorine impurity has negligible impacts on final cathode morphology. However, the particle interior is completely different. According to cross-sectional images given in col. 313, holes are found inside FNCM cathode particles (marked by circles and yellow arrows). Also, with the increase of impurity concentration from 0.2 at % to 5 at %, the hole formation within the cathode particles also rises. This demonstrates that the fluorine impurity is a key factor to the formation of holes in cathode particles and this beneficial feature could promotes the electrochemical performance enhancements.

[0042] SEM-EDS mapping results of as-prepared cathodes show that transition metal elements Ni, Co and Mn are distributed homogeneously throughout the particles. The precise composition of each element in NCM622 cathodes at different impurity level was measured by ICP-MS. As shown in Table II, the atomic ratio of Ni, Co and Mn in all samples remains close to the theoretical value (6:2:2). Specifically, exact atomic ratio among NCM is calculated to be 6:2.05:2.00, 6:2.11:2.01, 6:2.10:2.00 and 6:2.15:2.05 (Ni=6) for VNCM, 0.2FNCM, 1FNCM and 5FNCM, respectively. Additionally, the lithium composition remains stable regardless of variation in impurity concentration. Therefore, it can be concluded that fluoride ions do not occupy the sites of cations but to replace oxygens in cathode crystal during the synthesis process.

TABLE-US-00002 TABLE II Sample Li (%) Ni (%) Mn (%) Co (%) Na (%) VNCM 103.6 59.7 19.9 20.4 0 0.2FNCM 102.8 59.3 19.9 20.9 0 1FNCM 102.6 59.4 19.8 20.8 0 5FNCM 103.6 58.8 20.1 21.1 0

[0043] A typical mechanism for the synthesis of dense and spherical hydroxide particles via co-precipitation can be performed. There are two steps to form the precipitates: transition metal ions first coordinate with an ammonia complex agent (1), then slowly precipitate out of the base solution (2). Relevant two-step reactions are as follows:

M.sup.2++nNH.sub.3.fwdarw.[M(NH.sub.3).sub.n].sup.2+  (1)

[M(NH.sub.3).sub.n].sup.2++20H.sup.−.fwdarw.M(OH).sub.2↓+nNH.sub.3  (2)

At the initial stages, relatively sufficient fluoride ions could lead to the reaction:

[M(NH.sub.3).sub.n].sup.2++2F.sup.−.fwdarw.MF.sub.2+nNH.sub.3  (3)

which consumes the complex ions [M(NH.sub.3).sub.n].sup.2+ and pushes the equilibrium of the reaction (2) to the left. In this case, the dissolution rate of primary particles exceeds re-crystallization so that holes are created in precursors.

[0044] FIGS. 4A-4D show phase and structure of prepared materials as analyzed by powder X-ray diffraction with patterns. FIGS. 4A-4D show X-ray diffraction patterns of precursors with expansion of (001) plane in FIGS. 4A,4B, and cathodes with expansion of (003) plane, in FIGS. 4C and 4D. FIGS. 5A-5D show refinement profiles of VNCM, 0.2FNCM, 1FNCM and 5FNCM cathodes. All patterns in FIG. 4A-4B have the same diffraction peaks, corresponding to β-Ni(OH).sub.2 (PDF #00-059-0462), which is a layered metal hydroxide. There are nine noticeable diffraction peaks in XRD spectra of all cathodes in FIGS. 4C-4D, perfectly matched LiNi.sub.0.6Co.sub.0.2Mn.sub.0.2O.sub.2 (PDF #00-066-0854), which refers to a typical layered transition metal oxide. No extra phases or structure changes are found in both precursor and cathode patterns. Meanwhile, a pure-phase and well-layered cathode material are maintained. Nonetheless, the (001) peak of FNCM precursors broadens a bit as shown in 4A due to a decrease in primary particle size. Also, (003) plane slightly moves to lower angles for all FNCM cathodes (FIGS. 4C-4D) as a result of the expansion of c-axis in crystal lattice. It is worth noting that 5FNCM has the lowest I.sub.(003)/I.sub.(104) ratio (1.59), while other cathodes VNCM (1.81), 0.2FNCM (1.80) and 1FNCM (1.76) hold a much higher value. The higher I.sub.(003)/I.sub.(104) value indicates a lower Li.sup.+/Ni.sup.2+ mixing in cathode materials, resulting in a superior electrochemical performance. Thus, recovered cathodes under the influence of high concentration fluorine impurity (5 at %) may be unlikely to be competitive in comparison with the others.

[0045] In order to obtain lattice information of recovered cathodes in more detail, Rietveld refinement was used to calculate structure parameters and the rhombohedral crystal system with space group R3m was used as standard. FIGS. 5-5D show fitted XRD patterns of all cathode materials and refined crystallographic data are listed in Table III, showing Rietveld refinement results of the example NCM622 cathodes. As presented in FIGS. 5A-5D, a satisfactory consistency is achieved between calculated and observed patterns. VNCM is shown in FIG. 5A, and 0.2FNCM, 1FCNM and SFCNM in FIGS. 5B-5D respectively, Additionally, a relatively small χ.sup.2 is attained for each fitting operation (Table III), which guarantees all refined results in this work are of high quality. It is clear that the expansion of c-axis is related to the increase of the fluorine impurity level, which is consistent with previous studies. In fact, the inter-slab distance c of VNCM (14.214 Å) is enlarged˜0.03% to 14.218 Å for 5FNCM, accompanied with a˜0.13% expansion in cell volume. It should be elucidated that such an expanded lattice will not only reduce the resistance for Li diffusion, but also it could cut down the energy barrier of cation exchange (Li.sup.+/Ni.sup.2+) between layers. Apparently, compared to 0.2FNCM and 1FNCM, 5FNCM suffers from less cation mixing (4.41%) due to its excess lattice expansion, whereas other cathodes with minor modification in crystal structure remain in good condition. The replacement of O.sup.2− with F.sup.− is believed to be the reason to induce appreciable changes to the NCM lattice parameters. It has been reported that the inclusion of fluorine can lead to the reduction of TM ions for charge compensation. Since TM ions with lower oxidation state have a larger radius, when an increasing number of O.sup.2− is substituted by F.sup.− under the influence of higher impurity concentration (which is further proved by the decline in O.sub.Lattice value), the presence of more bigger TM ions conforms to the greater lattice expansion. These results suggest that the fluorine impurity (<1 at %) could enhance Li diffusivity in cathodes via optimized lattice modification, whereas the well-ordered structure may be damaged when excessive impurities (>5 at %) are involved.

TABLE-US-00003 TABLE III a-axis c-axis volume Ni.sub.Li Sample (Å) (Å) (Å.sup.3) O.sub.Lattice (%) χ.sup.2 VNCM 2.866 14.214 101.07 2.003 3.85 2.77 0.2FNCM 2.866 14.216 101.14 1.992 3.78 2.90 1FNCM 2.866 14.216 101.13 1.983 3.77 2.76 5FNCM 2.867 14.218 101.20 1.934 4.41 3.33

[0046] FIG. 6A-6C show spectra of surface scans of the cathode materials. XPS tests were employed to investigate materials composition and elements valence states in the prepared cathodes. A survey scan on Na is in FIG. 6A shows no signal in all samples. This ensures that Na-related substances do not exist in final recovered materials. FIGS. 6B and 6C display F is spectra of surface and bulk for all cathode samples. A strong peak is observed in surface spectra of FNCM cathodes. In FIG. 6C, on the contrary, there is no fluorine signal at all. It means that the fluoride ions are mainly clustered at the surface rather than inner part of cathode particles. In addition, the peak (˜685 eV) in FIG. 6B closely coincides with metal fluorides like LiF and TMF.sub.2 (TM=Ni, Co or Mn). Moreover, with the impurity concentration rises from 0.2 at % to 5 at %, the peak intensity increases relatively. Thus, XPS analysis verifies the inclusion of fluoride into cathode materials by the substitution for oxygen, and more importantly, fluoride ions largely reside near the particle surface to form a layer with elevated levels of low valence (+2) TM ions.

[0047] In the claimed approach, a predetermined percentage of fluoride is introduced from addition of fluorine to the leach solution, or simply from residual amounts, prior to precipitation of the charge material. The result is an increase of Ni.sup.2+ and/or Co.sup.2+ ions over respective Ni.sup.3+ and Co.sup.3+ ions on a surface of the charge materials following subsequent sintering. Other ions may also realize a preferred surface distribution based on the particle crystal structure.

[0048] FIGS. 7A-7D show deconvolution scans and percentages of Ni and Co surface ions on the cathode material particles. To clearly determine the relevant contents of Ni.sup.2+/Ni.sup.3+ and Co.sup.2+/Co.sup.3+ in cathode materials, the spectra deconvolution of 2p.sub.3/2 peak region was performed to calculate the accurate percent ratio of different TM ions. The fitted spectra of Ni 2p.sub.3/2 shown in FIG. 7A indicate a mix of Ni.sup.2+ and Ni.sup.3+ in cathodes. In the example he leach solution has a range of 0.2-5 at % fluoride prior to precipitation of the charge material and increases a percentage of Ni2+ ions on a surface of the charge material particles to between 40.1%-43.8%. Therefore, the fluorine has a quantifiable effect towards increasing at least the Ni and Co having an oxidation state of 2+(Ni.sup.2+, Co.sup.2+) on the particle surface. As demonstrated above, when the leach solution has a range of 0.2-5 at % fluoride prior to precipitation of the charge material and increases a percentage of Ni.sup.2+ ions on a surface of the charge material particles to between 40.1%-43.8%. Similarly, when the leach solution has a range of 0.2-5 at % fluoride prior to precipitation of the charge material and increases a percentage of Co.sup.2+ ions on a surface of the charge material particles to between 13.0%-35.2%. Other ions may also be favorably distributed based on the halide doping. As shown above, the Ni 2p.sub.3/2 peak can be split into Ni.sup.2+ (854.7 eV) and Ni.sup.3+ (856.5 eV). Similarly, as shown in FIG. 7B, the Co 2p.sub.3/2 peak can be divided into Co.sup.2+ (782.1 eV) and Co.sup.3+ (780.2 eV), respectively. According to FIG. 7C, the ratio of Ni.sup.2+ at particle surface has increased about 4% from 40.1% (VNCM) to an average of 43.8% for all FNCM cathodes. In contrast, the content of Ni.sup.2+ within the core is only little affected by the additional fluorine impurity, with a largest ratio change less than 1% between VNCM and 5FNCM. This further confirms the fact that more Ni.sup.2+ ions exist near surface of FNCM cathodes, which could contribute to a higher reversible capacity during cycling. FIG. 7D shows only a small amount of Co.sup.2+ appears in virgin cathode surface. The percentage of surface Co.sup.2+ increases to ˜13% for 0.2FNCM and 1FNCM and reaches as high as 35.2% for 5FNCM. It can be inferred that the reduction of cobalt ions also occurs at the particle surface, especially in the case of a high level of fluoride incorporation (5 at %) into the cathode particle. It should be noted that the large size and charge differences between Li.sup.+ and Co.sup.3+ ensure a good cation ordering, which is critical for fast lithium-ion diffusion in cathodes. Thus, with a remarkably low surface Co.sup.3+ concentration, it is not surprising that 5FNCM has a bad cation mixing as revealed previously by XRD analysis. The ratio of Co.sup.2+ can be shown to keeps stabl (˜36%) for all samples at the bulk of the particles.

[0049] FIGS. 8A-8C show a plot of fitting lines between peak current (Ip2) vs. scan rate (v) from CV curves for (8A) anodic side and (8B) cathodic side; and (8C) Calculated Li-ion diffusion coefficient. FIG. 9 shows li-ion diffusion within different cathode crystal structures. Referring to FIGS. 8A-9, a series of cyclic voltammetry (CV) tests at various scan rates were used to determine Li diffusion coefficient (Du) by using Randles-Sevcik equation:

I.sub.p.sup.2=(2.69×10.sup.5).sup.2n.sup.3A.sup.2D.sub.LiC.sup.2υ  (4)

where n is the number of electrons transferred in redox reaction (n=1); A is the electrode surface area (1.13 cm.sup.2); C is the theoretical molar concentration of Li-ion in NCM crystal (0.05 mol/cm.sup.3) and Du is the Li diffusion coefficient. The ratio of the squared peak current (I.sub.p.sup.2) to the scan rate (υ) is proportional to the lithium-ion diffusivity (D.sub.Li). The linear relationship between these two variables (I.sub.p.sup.2 vs. υ) is depicted in FIGS. 8A-8B where the slope of fitted lines directly reflects the diffusion rate of Li-ion in NCM materials. Then, the slope values are plugged into Randles-Sevcik equation to calculate Du and the relevant data are given in FIG. 8C. The anodic slopes and cathodic slopes shown in FIGS. 8A,8B confirm a strong correlation between anodic and cathodic diffusivity which represent the Li diffusion during de-lithiation and lithiation, respectively. The Du of VNCM is calculated to be 1.57×10.sup.−10 cm.sup.2/s (anodic) and 4.24×10.sup.−11 cm.sup.2/s (cathodic), which are in accordance with recent reports.

[0050] According to the slopes, 0.2FNCM and 1FNCM have a larger value of Du compared to the virgin standard, while the Li-ion diffusion condition in 5FNCM is the worst of them all. In particular, 0.2FNCM has the highest Li diffusion coefficient of 2.04×10.sup.−10 cm.sup.2/s (anodic) and 4.83×10.sup.−11 cm.sup.2/s (cathodic) which are about 30% and 14% greater than that of virgin standard, indicating that 0.2FNCM could possess the best electrochemical performance among all prepared samples. The lattice images in FIG. 9 demonstrate the effects on ionic diffusion caused by different level of fluoride inclusion in NCM crystal. It should be noted that the octahedral sites (marked by red x) between close-packed oxygen ions are for Li-ion diffusion within layers. When minor fluoride substitution occurs (<1 at %), the diffusion paths will still remain feasible. Since fluoride ions have a smaller ionic radius (˜1.3 Å) and charge number compared to oxygen ions (˜1.4 Å), such a substitution could result in an enlarged interspace and a reduced energy barrier that is responsible for the increase of Li-ion diffusivity.

[0051] However, excessive fluoride inclusion (>5 at %) will trigger a considerable deterioration in cation ordering. Therefore, more diffusion routes will be blocked by the mismatched Ni.sup.2+ ions which leads to diminished diffusion efficiency.

[0052] FIGS. 10A-10C are schematic diagrams showing formed holes based on the doping/impurities to illustrate the positive influences on the recovered NCM622 cathodes that are caused by fluorine impurity. The addition of small amounts of halides such as fluoride results in holes in the precipitated charge material particles. The internal particle charges and forces associated with the holes increases a ratio of 2+ to 3+ oxidation states of the charge material precursor. By introducing the fluorine reaction equilibrium during co-precipitation, fluoride ions promote the dissolution of primary particles and result in the formation of cavity, or hole in the precursors. FIG. 10A shows that conventional virgin materials exhibit solid particles 101 with no holes. FIG. 10B shows a˜1% fluorine content having negative F.sup.1− ions bound to Ni.sup.2+ ions, including a hole 102, and an increased density of F.sup.1− ions with 5% F in FIG. 10C, also surrounding a hole 103.

[0053] Therefore, after precursors are sintered, cathode particles with a hollow structure defined by holes are obtained in FNCM materials. The definition of a particle structure including the holes has a positive effects on cathode rate performance and cycle stability because the presence of holes reduces the overall Li diffusion distance and could improve particle stability by restraining volume change during cycling. Further, due to charge compensation, a surface layer with an increased ratio of Ni.sup.2+ is demonstrated in the FNCM cathodes, which substantially increases the reversible capacity of materials. Reducing the amount of high valence Ni.sup.3+ at near surface enhances an ability to maintain the surface stability. In addition, the Li diffusion coefficient will be improved in the FNCM cathodes (<1 at %) where a tiny portion of lattice oxygens are replaced by fluoride ions which results in a better structure and energy level for ion transport in the lithium layer. Despite all those positive roles, an undesired high level of cation disorder is found in 5FNCM cathode, which in turn brings negative impact on Li diffusion. A modest level of fluorine impurity, or doping, is beneficial on the recovered NCM622 cathodes obtained via hydrometallurgical process. Similar benefits extend to at least other NMC chemistries. In general, these results confirm the advantages of fluorine impurity on the NMC622 cathode materials.

[0054] While the system and methods defined herein have been particularly shown and described with references to embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.