REGENERATION OF LITHIUM CATHODE MATERIALS

Abstract

Regeneration of degraded cathode particles in lithium-ion batteries is achieved using a combination of hydrothermal treatment of cycled electrode particles followed by short thermal annealing. The methods provide for direct regeneration of Li-ion cathode materials including LiCoO.sub.2, LiMn.sub.2O.sub.4, LiFePO.sub.4, and Li.sub.xNi.sub.y Mn.sub.z Co.sub.1−y−zO.sub.2, in an economical and environmentally-friendly process.

Claims

1. A method for regenerating degraded lithium-ion battery cathode material, the method comprising: hydrothermally treating the cathode material in a Li-containing salt solution at a treatment temperature within a range of about 160° C. to about 220° C. for a treatment period of from 1 to 6 hours; separating the treated cathode material from the salt solution; and annealing the separated cathode material for an annealing period of from 1 to 6 hours to produce a relithiated material.

2. The method of claim 1, wherein the cathode material is one or more of LiCoO.sub.2, LiMn.sub.2O.sub.4, LiFePO.sub.4, and Li.sub.xNi.sub.y Mn.sub.z Co.sub.1−y−zO.sub.2 (0<x,y,z<1).

3. The method of claim 1, wherein the salt solution is one or more lithium salt selected from lithium hydroxide (LiOH), lithium carbonate (Li.sub.2 CO.sub.3), lithium sulfate (Li.sub.2SO.sub.4), lithium chloride (LiCl), and lithium nitrate (LiNO.sub.3).

4. The method of claim 1, wherein the salt solution includes one or more of sodium hydroxide (NaOH), potassium hydroxide (KOH), and ammonium hydroxide (NH.sub.4OH).

5. The method of claim 4, wherein the salt solution comprises a mixture of a lithium salt and KOH to yield approximately 0.1 M to 4 M OH.sup.−.

6. The method of claim 5, wherein the lithium salt is LiOH and has a concentration of approximately 0.1 M.

7. The method of claim 1, wherein the Li-containing salt solution is recycled from at least one prior use.

8. The method of claim 1, wherein the treatment temperature is approximately 220° C.

9. The method of claim 1, wherein the treatment period is approximately 2 to 4 hours.

10. The method of claim 1, wherein the treatment temperature and treatment period are selected to refill lithium deficiencies in a bulk crystal structure of the cathode material.

11. The method of claim 1, wherein annealing is performed at an annealing temperature within a range of 550° C. to 950° C.

12. The method of claim 1, wherein the salt solution is a mixture of LiOH and KOH, and wherein annealing is performed at an annealing temperature of approximately 750° C.

13. The method of claim 1, wherein annealing is performed in at least partial oxygen pressure.

14. The method of claim 1, wherein annealing is performed in an air or oxygen environment at approximately 750° C. to 850° C.

15. The method of claim 1, wherein annealing further comprises mixing the treated cathode materials with an excess amount of a lithium source.

16. A method for regenerating degraded lithium-ion battery cathode material comprising: refilling lithium deficiencies in a bulk crystal structure of the cathode material by hydrothermally treating the cathode material in a Li-containing salt solution for a treatment period of from 1 to 6 hours at ambient pressure; and annealing the treated cathode material at an annealing temperature for an annealing period of from 1 to 6 hours to produce a relithiated material.

17. The method of claim 17, wherein the cathode material is one or more of LiCoO.sub.2, LiMn.sub.2O.sub.4, LiFePO.sub.4, and Li.sub.xNi.sub.y Mn.sub.z Co.sub.1−y−zO.sub.2 (0<x,y,z<1).

18. The method of claim 16, wherein the salt solution is one or more of lithium hydroxide (LiOH), lithium sulfate (Li.sub.2SO.sub.4), lithium chloride (LiCl), lithium carbonate (Li.sub.2 CO.sub.3), and lithium nitrate (LiNO.sub.3).

19. The method of claim 16, wherein the salt solution further comprises one or a combination of sodium hydroxide (NaOH), potassium hydroxide (KOH), and ammonium hydroxide (NH.sub.4OH).

20. The method of claim 16, wherein hydrothermally treating the cathode material comprises exposing the salt solution to a treatment temperature of approximately 160-220° C.

21. The method of claim 16, wherein the annealing temperature is within the range of 550-850° C. and the annealing period is from 2 to 4 hours.

22. The method of claim 16, wherein the salt solution comprises a mixture of a lithium salt and KOH to yield approximately 0.1 M to 4 M OH.sup.− and the annealing is performed at approximately 750° C.

23. The method of claim 22, wherein the lithium salt is LiOH and has a concentration of approximately 0.1 M.

24. The method of claim 1, wherein the Li-containing salt solution is recycled from at least one prior use.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] FIGS. 1A-1C are alternative graphical depictions of a recycling and regeneration procedure for a battery according to an embodiment of the invention; FIG. 1D plots cell capacity retention of a recycled battery after multiple cycles.

[0020] FIGS. 2A-2B are SEM images (upper panel) and plots of particle size distributions (lower panel) of fresh and cycled LiCoO.sub.2 particles and samples regenerated under two different conditions, according to one embodiment of the invention;

[0021] FIGS. 2C-2D show SEM images and plots of particle size distributions of LiCoO.sub.2 particles regenerated at other conditions.

[0022] FIGS. 3A-3C provide a comparison of XRD patterns of fresh, cycled and regenerated LiCoO.sub.2 materials under different conditions.

[0023] FIGS. 4A-4B illustrate Raman spectra of a cycled and regenerated LiCoO.sub.2 cathode, respectively.

[0024] FIG. 5 illustrates the cycling performance of LiCoO.sub.2 regenerated using several different methods.

[0025] FIG. 6 compares rate performance of LiCoO.sub.2 regenerated using different methods.

[0026] FIG. 7 illustrates Nyquist plots of regenerated LiCoO.sub.2 after 100 cycles by different methods.

[0027] FIG. 8A illustrates cycling and rate performance of recycled LiCoO.sub.2 treated with pure LiOH and mixed Li salt;

[0028] FIG. 8B illustrates XRD patterns of cycled and regenerated LiCoO.sub.2;

[0029] FIGS. 8C and 8D compare cycling performance of pristine and hydrothermal-regenerated LiCoO.sub.2 powders (FIG. 8C) and pristine LiCoO.sub.2 and pristine LiCoO.sub.2 sintered with excess Li (FIG. 8D);

[0030] FIG. 8E illustrates XRD patterns of the pristine and regenerated NCM;

[0031] and FIG. 8F illustrates the cycling performance of the pristine and regenerated NCM.

[0032] FIG. 9 illustrates voltage capacity profiles and cycling performance of a pristine and regenerated LCO−NCM mixed cathode (upper panel) and HR-TEM images and FFTs of cycled and regenerated LiNi.sub.1/3 Co.sub.1/3Mn.sub.1/3O.sub.2 (NCM111) particles.

[0033] FIG. 10 is a flow diagram of an exemplary method of regenerating degraded NCM cathode materials using a particle-to-particle approach according to an embodiment of the invention.

[0034] FIG. 11 is a plot of cell capacity retention of cycling performance of a LiNi.sub.0.5 Co.sub.0.2Mn.sub.0.3O.sub.2 (NCM523) pouch cell.

[0035] FIG. 12 shows SEM images (Panels (a),(c)), particle size distributions (Panels (b), (d)) and high resolution-transmission electron microscope (HR-TEM) images (Panels (e), (f)) of pristine NCM523 particles and cycled/degraded NCM523 particles, respectively.

[0036] FIG. 13 shows SEM images (upper panel) and particle size distributions (lower panel) of NCM111 particle under different conditions.

[0037] FIG. 14 shows HR-TEM and Fast Fourier Transform (FFT) images of pristine NCM111 particles.

[0038] FIG. 15 shows HR-TEM and FFT images of cycled NCM111 particles.

[0039] FIG. 16A illustrates a sample hydrothermal lithiation process of degraded LIB cathode particles during hydrothermal treatment;

[0040] FIG. 16B illustrates hydrothermal lithiation kinetics of degraded LIB cathode particles during hydrothermal treatment.

[0041] FIG. 17 illustrates a comparison of XRD patterns of pristine, cycled, and regenerated NCM111 and NCM523 materials under different conditions.

[0042] FIG. 18 compares Rietveld refinement of XRD patterns of pristine (Panel (a)), cycled (Panel (b)), and regenerated (Panels (c)-(e)) NCM111 cathodes.

[0043] FIG. 19 compares XRD patterns of pristine (Panel (a)), cycled (Panel (b)), and regenerated (Panels (c)-(e)) NCM523 cathodes.

[0044] FIG. 20 shows SEM images (Panel (a)), particle size distribution (Panel (b)), HR-TEM images (Panels (c)-(e)) and X-ray photoelectron spectroscopy (XPS) spectra (Panel (f)) of regenerated NCM523 cathodes by different approaches.

[0045] FIG. 21 shows HR-TEM and FFT images of cycled NCM523 particles after hydrothermal treatment only.

[0046] FIG. 22 shows HR-TEM and FFT images of cycled NCM111-hydrothermal-short annealing (HT-SA) particles.

[0047] FIG. 23 shows HR-TEM and FFT images of cycled NCM111-solid state sintering in air (SS-air) particles.

[0048] FIG. 24 shows HR-TEM and FFT images of cycled NCM111-solid state sintering in oxygen (SS-02) particles.

[0049] FIG. 25 illustrates XPS patterns of pristine NCM111, cycled NCM111, NCM111-HT-SA and NCM111-SS—O.sub.2 samples.

[0050] FIG. 26 provides comparisons of cycling performance, rate performance, voltage profiles and crystal structure of pristine, non-treated and regenerated NCM111 and NCM523 samples.

[0051] FIG. 27 illustrates Nyquist plots of regenerated NCM523 cathodes after 100 cycles.

[0052] FIG. 28 compares the cycling stability of hydrothermally relithiated NCM111 at 160° C., 200° C., and 220° C. for different durations ranging from 1 h, 2 h, 4 h and 6 h.

[0053] FIG. 29A compares XRD patterns of samples relithiated at temperatures of 160° C., 200° C., and 220° C., along with control samples (T-NCM111 and D-NCM111);

[0054] FIG. 29B shows unit cell parameters of a and c;

[0055] FIG. 29C shows XPS spectra in Co 2p and Ni 2p region of samples relithiated at different temperatures.

[0056] FIGS. 30A-30D compares results of varied solution composition for hydrothermal relithiation including XRD patterns (FIG. 30A), XPS spectra (FIG. 30B), cation concentration (FIG. 30C), and cycling stability (FIG. 30D).

[0057] FIGS. 31A-31B provides voltage profiles and cycling stability, respectively, of a relithiated sample annealed at 850° C. with and without lithium source.

[0058] FIGS. 32A-32D illustrate results of evaluation of compensation for lithium loss, where FIGS. 32A & 32B show XPS spectra of relithiated samples at different annealing temperatures with Li.sub.2 CO.sub.3 and LiOH, respectively, as the Li source; FIGS. 32C & 32D plot cycling stability of relithiated samples at different annealing with Li.sub.2 CO.sub.3 and LiOH, respectively, as the Li source.

[0059] FIGS. 33A-33B show XPS survey spectra and high-resolution XPS spectra of Ni 2p, respectively; FIG. 33C shows XRD patterns of cycled T-NCM R-NCM.

[0060] FIGS. 34A & 34B are XRD patterns and Ni 2p XPS spectra, respectively of NCM111 cathodes regenerated at 1 g and 10 g scales.

DETAILED DESCRIPTION OF EMBODIMENTS

[0061] In an illustrative implementation of the inventive system and method, a simple yet efficient non-destructive cathode recycling approach is provided for generating high-capacity and high-rate active particles using LiCoO.sub.2 as the model material. Hydrothermal reaction is widely used in the synthesis of various cathode materials and has the capability of generating particles with high crystallinity and desired stoichiometry. Here, we took the advantage of this process to pre-dose Li into Li-deficient cathode particles without concern about the Li/Co ratio. Then we combined hydrothermal treatment with simple thermal annealing to regenerate LiCoO.sub.2 with desired microstructure and composition, which led to outstanding electrochemical performance. Compared with the previous approaches, this strategy shows several major advantages: i) it does not require tedious chemical analysis to determine the amount of Li.sup.+ loss, and is compatible with batteries at different capacity fading conditions; ii) it does not require long-time, energy-consuming sintering treatment since Li.sup.+ is dosed with the correct stoichiometry during the hydrothermal process; and iii) the regenerated active particles have high capacity and improved rate capability compared with solid-state synthesis approach.

[0062] For testing, different forms of cells were evaluated. The first type of cell was commercial LiCoO.sub.2 cells. Pouch cells with LiCoO.sub.2 as the cathode and graphite as the anode were purchased from MTI Corporation (www.mtixtl.com) (2 Ah, EQ-PL-605060-2 C). The pouch cells were cycled in the voltage range of 3-4.5 V using a LAND battery tester for 200 cycles and discharged to 2 V at C/10 (1 C=150 mA g.sup.−1) before disassembly. The cathode strips were harvested from the pouch cells, by thoroughly rinsing with dimethyl carbonate. After drying, the cathode strips were soaked in N-Methyl-2-pyrrolidone (NMP) for 30 min followed by sonication for 20 min. The LiCoO.sub.2 powders, binder and carbon black were removed from the aluminum substrates. The obtained suspension was centrifuged at 3500 rpm for 5 min and LiCoO.sub.2 powders were precipitated, separated and dried for regeneration. Fresh pouch cells were directly discharged to 2 V at C/10 without any cycling before disassembly and the harvested LiCoO.sub.2 material serves as the reference material for comparison.

[0063] A second cell type was “home-made” LiCoO.sub.2 cells, constructed from pristine LiCoO.sub.2 powders (MTI Corporation) to perform cycling and harvesting active cathode particles after capacity fading. To fabricate thick electrodes, LiCoO.sub.2 powders were mixed with polyvinylidene fluoride (PVDF), and carbon black (Super P65) in NMP at a mass ratio of 93:4:3 to form homogenous slurries. Then the slurries were cast on aluminum foil using a doctor blade and dried in vacuum at 80° C. for 6 h. Circle electrodes were cut and compressed by rolling mill. The active mass loading was about 28 mg/cm.sup.2. 2016-type coin cells were assembled with Li metal disc (thickness 1.1 mm) as anode, 1 M LiPF.sub.6 in ethylene carbonate and diethyl carbonate (EC:DEC 1:1 wt.) as the electrolyte, and trilayer membrane (Celgard 2320) as the separator. The cells were cycled in the voltage range of 3-4.5 V to gain >50% capacity loss and then discharged to 2 V at C/10. The following harvesting procedure was the same as stated previously. These materials will be used to validate the regeneration procedure, and also compare the electrochemical performance of the pristine LiCoO.sub.2 powder and regenerated cathodes from commercial pouch cells.

[0064] A third cell type was home-made LiNi.sub.1/3 Co.sub.1/3Mn.sub.1/3O.sub.2 (NCM) and mixed pouch cells. To demonstrate that this approach can be potentially used in mixed cathode chemistry, we built home-made pouch cells from pristine NCM powder (Toda America) and mixed pristine LiCoO.sub.2-NCM powders with an active mass ratio of 1:1. The electrodes fabrication procedure, electrolyte and separator are the same as in the home-made LiCoO.sub.2 cells. Pouch cells were assembled with Li metal foil (thickness 0.75 mm) as anode, and a typical pouch cell had an electrode area of 20 mm×55 mm. The cells were cycled in the voltage range of 3-4.5V to gain >40% capacity loss and then discharged to 2 V at C/10. The following harvesting procedure was the same as stated previously. These materials will be used to demonstrate the feasibility of this technique to regenerated mixed cathode materials.

[0065] Two different cathode regeneration methods were compared in terms of their operation characteristics and electrochemical performance of the regenerated products:

[0066] 1. Combined hydrothermal treatment and short annealing: For hydrothermal treatment, LiCoO.sub.2 powders harvested from cycled cells were loaded into a 100 mL Teflon-lined autoclave filled with 80 mL of 4 M lithium hydroxide (LiOH) solution, or a mixed solution of 1 M LiOH and 1.5 M Li.sub.2SO.sub.4. In some embodiments, a lithium containing solution can be made by a small concentration (e.g., 0.1 M) of lithium salt (e.g., LiOH, Li.sub.2SO.sub.4, LiCl, LiNO.sub.3), and may also be mixed with an alkaline solution from NaOH, KOH, NH.sub.4OH or their mixture. The autoclave was kept at a wide range of temperatures and times, at ambient pressure. The treated LiCoO.sub.2 powders were washed thoroughly with deionized water, dried and then annealed at different temperatures with a ramping rate of 5° C./min.

[0067] 2. Solid-state synthesis: For LiCoO.sub.2 regeneration from solid-state synthesis, the compositions of Li and Co of cycled cathode materials were first measured by an inductively coupled plasma optical emission spectrometer (ICP-OES, Perkin Elmer Optima 3000 DV). The harvested cathode powders were mixed with Li.sub.2 CO.sub.3 with agate mortar and pestle. The amount of Li.sub.2 CO.sub.3 was calculated to obtain the mixture with a Li/Co ratio of 1.05. The 5% excess amount of Li was added to compensate the Li evaporation during the sintering process. The mixtures were sintered at different temperatures and times with a ramping rate of 5° C./min.

[0068] The morphology of LiCoO.sub.2 powders was observed by Ultra High-Resolution Scanning Electron Microscope (UHR SEM, FEI XL30). The particle size distribution is analyzed with Nano Measurer. The crystal structure of the powders was examined by X-ray Powder Diffraction (XRD) employing Cu K.sub.α radiation. Raman spectroscopy was recorded using a Renishaw inVia confocal Raman Microscope. Spectra were collected between 300 and 900 cm.sup.−1 with a 633-nm laser. The crystal structure of the cycled and regenerated NCM materials were examined by Transmission Electron Microscopy (TEM) (FEI Titan 80-300 kV S/TEM).

[0069] To evaluate electrochemical performance, the regenerated and non-treated LiCoO.sub.2 powders were mixed with PVDF and Super P65 in NMP at a mass ratio of 8:1:1. The resulted slurries were cast on aluminum foils followed by vacuum drying at 80° C. for 6 h. Circle electrodes were cut and compressed, with controlled active mass loading of about 3 mg/cm.sup.2. Cells were fabricated by the same procedure as thicker electrodes mentioned above. Galvanostatic charge-discharge was carried out in the potential range of 3-4.3 V. The electrochemical impedance spectroscopy (EIS) tests were performed at discharged state in the frequency range of 10.sup.6 Hz to 10.sup.−3 Hz with signal amplitude of 10 mV by a Metrohm Autolab potentiostat.

[0070] The overall recycling and regeneration procedure are illustrated in FIGS. 1A and 1C. In general, cycled cells with significant capacity degradation were disassembled and electrodes strips were harvested. The cathodes were washed by NMP and active material powders were collected. The obtained powders were subject to different regeneration treatments either by hydrothermal treatment with short annealing or solid-state synthesis. The regenerated cathode materials were made into slurries to fabricate new cells. In this work, coin cells were assembled to evaluate the electrochemical performance of the regenerated cathode material for the sake of convenience and consistence.

[0071] FIG. 1B illustrates an exemplary method of regenerating cathode materials, according to one embodiment of the invention. As mentioned above, in step 120, degraded cathode materials must first be collected from cycled cells. In step 122, lithium (Li) is pre-dosed into Li-deficient cathode particles via a lithium hydroxide solution and the cathode particles with the lithium hydroxide solution are hydrothermally treated for a specified time and at a specified temperature. In step 124, thermal annealing is performed on the hydrothermally treated cathode particles to complete the regeneration of the cathode particles. In step 126, the new regenerated cathode particles are formed into a slurry, which, in step 128, can be used to fabricate new battery cells.

[0072] To produce faded cathode materials, commercial pouch cells with LiCoO.sub.2 as the cathode active material were cycled in the voltage range of 3-4.5 V to speed up the capacity fading. The resulted cell capacity retention was 74% after 200 cycles, as illustrated by the discharge capacity retention graph in FIG. 1D. The cycled cathode serves as a proper material for recycling and regeneration purpose. Two methods were used to regenerate cycled LiCoO.sub.2. For hydrothermal treatment combined with short annealing, it is speculated that the following reaction happens during the hydrothermal step:

[00001]$\begin{array}{cc}{\mathrm{Li}}_x{\mathrm{CoO}}_2+\left(1-x\right)\mathrm{LiOH}+\frac{x-1}{4}{O}_2.fwdarw.{\mathrm{LiCoO}}_2+\frac{1-x}{2}{H}_{2}O& \left(1\right)\end{array}$

[0073] The following short annealing step can increase the crystallinity and eliminate structure defects. For the solid-state synthesis approach with long-term sintering, the following reactions happen:

[00002]$\begin{array}{cc}{\mathrm{Li}}_x{\mathrm{CoO}}_2.fwdarw.\frac{1-x}{3}{\mathrm{Co}}_3{O}_4+{\mathrm{xLiCoO}}_2+\frac{1-x}{3}{O}_2& \left(2\right)\\ 2{\mathrm{Co}}_3{O}_4+3{\mathrm{Li}}_2{\mathrm{CO}}_3+\frac{1}{2}{O}_2.fwdarw.6{\mathrm{LiCoO}}_2+3{\mathrm{CO}}_2& \left(3\right)\end{array}$

[0074] where x (0.5<x<1) is the mole number of Li in the cathode. LixCoO.sub.2 starts to release oxygen at 220° C. and forms Co.sub.3O.sub.4, which is later reacted with Li.sub.2 CO.sub.3 to form LiCoO.sub.2 again.

[0075] The morphology and particle size of the LiCoO.sub.2 powders remained nearly the same after cycling and regeneration under different conditions. FIGS. 2A and 2B show SEM images and size distributions of fresh and cycled LiCoO.sub.2 particles, as well as samples regenerated at two different conditions. The SEM images and size distributions of LiCoO.sub.2 particles regenerated at other conditions are shown in FIGS. 2C and 2D. All the samples show similar morphology and particle size distribution, suggesting that the regeneration approach does not affect their morphology and sizes.

[0076] It has been demonstrated that the capacity fading of LiCoO.sub.2 is mainly contributed by the Li.sup.+ loss and the increases of solid-electrolyte interface (SEI). The Li.sup.+ loss of the cycled pouch cell was evidenced by the Inductively Coupled Plasma (ICP) result, showing a Li/Co ratio decreased from 0.99 to 0.80 after 200 cycles. For the regeneration process, the target ratio of Li/Co is 1 and the compositions of the cathode materials regenerated by different processes are listed in Table 1, which lists ICP results of regenerated cathode materials from different conditions.

TABLE-US-00001 TABLE 1 Hydrothermal No annealing 700° C. 4 h 800° C. 4 h Composition Li.sub.0.98CoO.sub.2 Li.sub.0.98CoO.sub.2 Li.sub.0.98CoO.sub.2 Solid-state synthesis 750° C. 12 h 850° C. 12 h 650° C. 12 h Composition Li.sub.0.99CoO.sub.2 Li.sub.0.98CoO.sub.2 Li.sub.0.96CoO.sub.2

[0077] For the hydrothermal treatment with an approximately 4 mole (M) lithium hydroxide (LiOH) solution, an excess amount of Li source is provided in the aqueous phase, which promotes the lithiation of cycled LiCoO.sub.2 to reach targeted stoichiometry without controlling the ratio between Li and Co. One major advantage of this step is that degraded LiCoO.sub.2 with any Li/Co ratio can be processed together. The Li/Co ratio remains the same in the regenerated particles after short annealing process. For comparison, in solid-state synthesis, the ratio between Li.sub.2 CO.sub.3 and the cycled LiCoO.sub.2 (Li-deficient) needs to be carefully controlled to reach a desired composition after long-time sintering. Also, it is shown that the final Li/Co ratio obtained after solid-state synthesis decreases as the sintering temperature increases due to Li evaporation during long-term sintering, which is consistent with literature.

[0078] Phase transitions also take place in Li.sup.+-deficient LixCoO.sub.2 during cycling (x is the mole number of Li). According to the phase diagram of LixCoO.sub.2,.sup.26 for 0.5<x<1, layered LiCoO.sub.2 transforms to cubic spinel LiCo.sub.2O.sub.4, and for x<0.5, spinel Co.sub.3O.sub.4 is also formed besides LiCo.sub.2O.sub.4. Another major degradation mechanism of LiCoO.sub.2 is that the low Li.sup.+ conductivity of the spinel LiCo.sub.2O.sub.4 and Co.sub.3O.sub.4 phases increases the polarization which also contributes to the extra capacity loss. The XRD patterns of fresh, cycled and regenerated LiCoO.sub.2 materials are compared in FIG. 3A. All the LiCoO.sub.2 materials have very high intensity ratio I.sub.(003)/I.sub.(104), which corresponds to well-defined layered structure. For better comparison, enlarged views of the patterns in the range of 35-55° are displayed in FIG. 3B. For LiCoO.sub.2 after 200 cycles, a small diffraction peak of Co.sub.3O.sub.4 can be detected, which is consistent with literature. The signal of LiCo.sub.2O.sub.4 cannot be observed separately due to its similar peak positions with LiCoO.sub.2. After regeneration from both processes, the peak of Co.sub.3O.sub.4 all disappears, indicating a successful re-construction of the layered structure.

[0079] The XRD patterns are further analyzed by Rietveld refinement and the results are displayed in Table 2, which lists the lattice parameters of fresh, cycled and regenerated cathode materials.

TABLE-US-00002 TABLE 2 Sample a/Å c/Å R.sub.B R.sub.wp Fresh 2.8114(8) 14.010(9) 2.60% 2.37% After 200 cycles 2.8103(4) 14.118(5) 0.89% 1.86% 750° C. 12 h 2.8119(9) 14.039(3) 2.29% 1.95% 850° C. 12 h 2.8136(7) 14.014(1) 1.52% 2.43% 950° C. 12 h 2.8141(1) 14.043(8) 1.53% 2.07% Hydrothermal 2.8150(3) 14.053(2) 1.15% 2.02% Hydro 700° C. 4 h 2.8146(7) 14.044(2) 0.97% 2.73% Hydro 800° C. 4 h 2.8133(1) 14.040(0) 1.12% 3.47%

[0080] The Bragg factor (RB) and weighted profile R-factor (R.sub.wp) are both below 4%. The Li.sup.+-deficient cathode after 200 cycles has decreased lattice a and increased c values compared with the fresh cathodes, which are attributed to smaller ionic radius of CO.sup.4+ than Co.sup.3+, and stronger electrostatic repulsion between the layers due to loss of Li.sup.+, respectively.

[0081] For the hydrothermal treatment approach, several regeneration conditions were tested: some cycled cathode materials were regenerated only through the hydrothermal treatment for comparison, while some materials were annealed at 700° C. and 800° C. after the hydrothermal treatment to increase the crystallinity. The increased crystallinity is confirmed by the decreased lattice a and c parameters after annealing due to a tighter pack of atoms, as well as the decrease in full width at half maximum (FWHM), as illustrated in FIG. 3C, which is consistent with literature. For the solid-state synthesis approach, the regenerated LiCoO.sub.2 has larger lattice parameters a and c at higher sintering temperature which leads to the lattice expansion.

[0082] Raman spectra further prove the existence of Co.sub.3O.sub.4 phase in the cathode after 200 cycles and its conversion to LiCoO.sub.2 after regeneration, as illustrated by the spectral charts in FIGS. 4A-4B. For the cathode materials after 200 cycles, the bands at 480, 521, 617 and 686 cm.sup.−1 are attributed to Co.sub.3O.sub.4. In the regenerated cathode through hydrothermal approach with short annealing, these bands disappeared and only two bands at 483 and 592 cm.sup.−1 are observed, which correspond to E.sub.g and A.sub.lg modes of LiCoO.sub.2. These results further confirm that Co.sub.3O.sub.4 is fully converted to LiCoO.sub.2 after regeneration.

[0083] The cycling performance of the regenerated LiCoO.sub.2 through both hydrothermal treatment with short annealing and solid-state synthesis is illustrated by the Panel (a) in FIG. 5. Reference cells with non-treated cathode directly harvested from the cycled pouch were also assembled. The coin cells were cycled at C/10 (1 C=150 mA g.sup.−1) for the first cycle as activation, and 1 C for the following cycles. FIG. 5 compares the results for regenerated LiCoO.sub.2 through hydrothermal treatment, with different cycling performance under different conditions. The sample that received hydrothermal treatment alone exhibits poor cycling stability, even worse than the non-treated material. It is noted that in the XRD patterns the I.sub.003/I.sub.104 ratio of the cathode after only hydrothermal treatment is smaller than the cycled non-treated cathode (FIG. 3A), which indicates a poorer hexagonal ordering. It is known that LiCoO.sub.2 processed at low temperature (below 400° C.) often exhibit cation disorder, which cannot keep structural stability that guarantees repeated insertion and removal of Li.sup.+. The annealing at appropriate temperature greatly enhanced the capacity retention due to increased cation ordering. After annealing at 700° C. for 4 h, the cycling stability is improved but still not ideal. After slightly increased the temperature to 800° C., the regenerated LiCoO.sub.2 after 4 h annealing achieves a significantly improved cycling performance, with a discharge capacity of 153.1 mAh g.sup.−1 in the initial cycle at C/10. After increasing the rate to 1 C, the capacity is 148.2 mAh g.sup.−1 and it can be still maintained on 135.1 mAh g.sup.−1 after 100.sup.th cycles, corresponding to a retention rate of 91.2%. An annealing temperature higher than 800° C. may also be applied, while 800° C. is sufficient to guarantee good cycling performance. The hydrothermal treatment parameters were further optimized by significantly decreasing the treatment time from 12 h to only 4 h, and increasing the hydrothermal temperature from 180° C. to 220° C. while maintaining the same short annealing step. The regenerated material under this improved condition still maintains the same stable cycling performance, as shown in Panel (a) of FIG. 5.

[0084] The regenerated LiCoO.sub.2 from solid-state synthesis under different conditions are also compared in Panel (b) of FIG. 5. The initial discharge capacities for LiCoO.sub.2 sintered at 750° C., 850° C., and 950° C. are 152.0, 152.1 and 151.1 mAh g.sup.−1, with discharge capacities on the 100.sup.th cycle of 133.2, 135.4 and 136.5 mAh g.sup.−1, respectively, which are similar to the capacity and cycling performance of samples regenerated by hydrothermal treatment with short annealing at 800° C. The slight difference in capacity retention can be due to variations in cell assembly. The solid-state synthesis at 700° C. was also performed as a comparison and the discharge capacity is 149.8 mAh g.sup.−1 in the initial cycle and 111.1 mAh g.sup.−1 in the 100.sup.th cycle, which shows poor cycling stability. This is due to the fact that 700° C. is lower than the melting point of Li.sub.2 CO.sub.3 (723° C.) and the melting of Li.sub.2 CO.sub.3 is required for its reaction with Co.sub.3O.sub.4. This result is in coincidence with the annealing experiment for the hydrothermal samples. Since solid-state synthesis at 850° C. delivers the optimum electrochemical performance, we have performed 4 more set of experiments with different sintering time of 4 h, 8 h, 18 h, and 24 h at 850° C. The resulted electrode cycling performance is shown in Panel (c) of FIG. 5. Solid-state sintering at 8 h, 12 h and 18 h deliver better cycling performance than at 4 h and 24 h. This can be understood by the following: by sintering for a short time of 4 h, the Li.sub.2 CO.sub.3 may not fully react with the degraded cathode material to compensate the Li.sup.+ loss, which leads to worse cycling stability; if the sintering time is as long as 24 h, there could be severe Li evaporation and particle agglomeration, which deteriorates the electrochemical performance. Overall, the specific capacity and cycling stability of LiCoO.sub.2 regenerated by hydrothermal approach with short annealing is similar to the best sample regenerated by the solid-state approach shown in Panel (d) of FIG. 5, while the former process is much easier to operate and scale up.

[0085] More interesting results were found in the evaluation of rate capability. The regenerated LiCoO.sub.2 materials with similar cycling stability show different rate performance in Panel (a) of FIG. 6. The LiCoO.sub.2 cathodes regenerated by two hydrothermal treatment conditions with short annealing have the best rate capabilities, while the sample treated at 220° C. for 4 h performs better. For the solid-state synthesis approach, the cathodes regenerated at 850° C. for 8 h, 12 h and 18 h show similar cycling performance to that which was illustrated in Panel (c) of FIG. 5, while different rate performance. The sample sintered for 8 h has worse rate capability than the samples sintered for 12 h and 18 h, and the latter two samples show similar rate capability. Therefore, the optimum condition to regenerate cycled cathode material through solid-state approach is identified as 850° C. for 12 h. Compared with the best rate capability in the solid-state approach, the hydrothermal treatment with annealing approach delivers much better rate performance. The cathode treated at 180° C. for 12 h has a discharge capacity of 138.0 mAh g.sup.−1 at 2 C and 124.7 mAh g.sup.−1 at 5 C, and the cathode treated at 220° C. for 4 h has higher discharge capacity of 141.9 mAh g.sup.−1 at 2 C and 130.3 mAh g.sup.−1 at 5 C. Panels (b) and (c) in FIG. 6 demonstrate the voltage-capacity profiles of LiCoO.sub.2 regenerated through hydrothermal approach at 220° C. and solid-state synthesis at 850° C. The hydrothermal sample has smaller voltage drops at the beginning of each discharging, which indicates a smaller polarization. For a clearer comparison, the voltage-capacity profiles at 5 C for all the samples are plotted in Panel (d) of FIG. 6, and the hydrothermal sample shows the highest discharge plateau. To quantify the polarization, Panel (e) of FIG. 6 displays the differences between the average charge and discharge voltages of all the samples, in which a larger difference means larger polarization due to lower electronic and Li.sup.+ conductivity. For a direct comparison between the two approaches, LiCoO.sub.2 is also regenerated through hydrothermal treatment followed by annealing at 750° C. for 12 h, and compared with the regenerated sample by solid-state synthesis at 750° C. as shown in Panel (f) of FIG. 6. By sintering at the same temperature and time, the material generated by hydrothermal approach has similar cycling performance, while much better rate performance than the solid-state approach. The hydrothermal sample has a discharge capacity of 122.6 mAh g.sup.−1 at 5 C while the solid-state sample only has 115.1 mAh g.sup.−1 at 5 C.

[0086] To investigate the mechanism for the different electrochemical performance of LiCoO.sub.2 regenerated under different conditions, EIS measurement was performed on the LiCoO.sub.2 at discharged state after 100 cycles at 1 C. Panels (a) and (b) of FIG. 7 show the Nyquist plot of regenerated LiCoO.sub.2 cathode through solid-state and hydrothermal approaches, respectively. The inset picture in Panel (a) displays the equivalent circuit to fit the Nyquist plots and get quantitative value of resistances. R.sub.s, R.sub.sei and R.sub.ct are the Ohmic resistance (electrolyte contact, electrode), SEI film resistance, and charge-transfer resistance, respectively. W is the Warburg impedance related to the Li.sup.+ diffusion. Table 3 provides the resistance values according to the fitting results from the equivalent circuit.

TABLE-US-00003 TABLE 3 R.sub.s/Ω R.sub.sei/Ω R.sub.ct/Ω D.sub.Li.sup.+/cm.sup.2 s.sup.−1 750° C. 12 h 10.21 33.16 204.7 8.01 × 10.sup.−13 850° C. 12 h 9.83 16.38 104.9 3.87 × 10.sup.−12 950° C. 12 h 7.852 17.09 129.4 1.23 × 10.sup.−12 Hydrothermal 7.492 7.835 3292 Hydro + 700° C. 4 h 10.94 36.33 198.8 1.51 × 10.sup.−13 Hydro + 800° C. 4 h 7.028 16.09 97.44 9.03 × 10.sup.−12
The LiCoO.sub.2 cathode regenerated by sintering at 850° C. has lower SEI and charge transfer resistances than the cathodes sintered at 750° C. and 950° C. The cathode regenerated by hydrothermal treatment followed by 800° C. annealing has the lowest resistances among all the regenerated cathodes, with a SEI resistance of 16.09Ω and charge transfer resistance of 97.44Ω. The superior rate performance of the hydro 800° C. cathode is attributed to its lowest charge transfer resistance which favors the charge transfer reaction for Li.sup.+ intercalation.

[0087] The linear part of Nyquist plot in the low frequency range is directly related to Li.sup.+ diffusion in electrode, and Li.sup.+ diffusion coefficient could be calculated using the following equation.

[00003]$\begin{array}{cc}D=\frac{R^2{T}^2}{2A^2{n}^4{F}^4{C}^2{\sigma }^2}& \left(4\right)\end{array}$

[0088] R is the gas constant, T is the absolute temperature, A is the interface between cathode and electrolyte (A=1.6 cm.sup.2), n is the number of electrons involved in reaction (n=1), F is the Faraday constant, C is the concentration of Li.sup.+ in the electrode (=ρ/M) based on the molecular weight of LiCoO.sub.2 (M) and density (ρ), and σ is the Warburg factor. The Warburg factor can be obtained from the slope of Z′ vs. ω.sup.−1/2 plots (co is the angular frequency) in the Warburg region. The results of the Z′ vs. ω.sup.−1/2 for regenerated cathodes after 100 cycles, along with the linear fitting curves, are shown in Panel (c) of FIG. 7. After the slopes are obtained, the Li.sup.+ diffusion coefficients for all the samples are calculated and shown in Table 3 (above). The hydro −800° C. cathode has the largest Li.sup.+ diffusion coefficient of 9.03×10.sup.−12 cm.sup.2 s.sup.−1, which corresponds well with its smallest polarization shown in Panel (e) of FIG. 6. Since the post-annealing could increase the crystallinity of cathode, annealing at 800° C. results in a higher crystallinity than at 700° C. (FIG. 3C), which favors the Li.sup.+ diffusion by providing a perfect Li.sup.+ diffusion path inside the particle. For the solid-state synthesis, 850° C. is determined to be the optimum condition to obtain good diffusion property, which is consistent with literature but it is still inferior to hydrothermal sample after 800° C. annealing.

[0089] The hydrothermal treatment was first carried out using 4 M LiOH solution with a high pH of 14.6. Developing more environmental benign operation is always desired for large-scale industrial processes. Therefore, we replaced the 4 M LiOH solution with a mixed solution containing 1 M LiOH and 1.5 M Li.sub.2SO.sub.4 to reduce the pH to 12.3. With all the other conditions the same, the cycled LiCoO.sub.2 treated with the mixed Li salt solution shows similar cycling stability and rate performance with the LiCoO.sub.2 treated with 4 M LiOH solution, as shown in FIG. 8A. The successful regeneration of LiCoO.sub.2 with mixed Li salt with decreased pH provides more flexibility and reduces the operation cost.

[0090] To further verify the effectiveness of this recycling and regeneration approach, the same hydrothermal treatment with short annealing procedure is applied to recycle LiCoO.sub.2 from the home-made cells. To have enough LiCoO.sub.2 cathode material for recycling, thick electrode with a high mass loading of 28 mg/cm.sup.2 were made. After cycling in the voltage range of 3-4.5 V to gain >50% of capacity fading, the cells were discharged to 2 V at C/10 and disassembled. ICP was performed on the harvested LiCoO.sub.2 and the composition was determined to be Li.sub.0.59 CoO.sub.2. The Li.sup.+ loss is higher compared with the commercial cathode cycled after 200 cycles, which is believed to result from the Li.sup.+ loss in the Li metal anode due to the SEI formation. It should be noted that due to the processing limitation of the lab-scale coin cells, their resistance was higher than the commercial pouch cells. Therefore, Li.sup.+ may not fully get back to the cathode even though the cell was discharged to 2 V. The LiCoO.sub.2 regeneration was performed by hydrothermal process with 4 M LiOH solution followed by short annealing at 800° C. for 4 h. The XRD patterns of cycled and regenerated LiCoO.sub.2 illustrated in FIG. 8B suggest that the regenerated materials have reproduced well-defined layered structure, similar to pristine LiCoO.sub.2 powder. The cycling performance of pristine and regenerated LiCoO.sub.2 are compared below. It is clear that the regenerated LiCoO.sub.2 could fully recover the specific capacity and cycling stability of the pristine sample, as shown in FIG. 8C. The performance is also similar to LiCoO.sub.2 regenerated from cycled commercial pouch cells. This experiment indicates that this method based on hydrothermal treatment with short annealing is effective in regenerating cycled cathodes with various degradation conditions.

[0091] Considering mixed cathode chemistry is widely used in LIBs, it is ideal that the recycling approach could regenerate mixed cathode material. Mixed LiCoO.sub.2 (LCO) and NCM cathode is used as the model material to demonstrate the feasibility of this approach to process mixed cathode material. We assembled two types of cells: 1) pure NCM-based pouch cells and 2) pouch cells using LCO−NCM mixed cathode with an active mass ratio of 1:1. After the pouch cells were cycled in the voltage range of 3-4.5 V until the capacity was decayed more than 40%, those pouch cells were disassembled and the cathode material was harvested. To confirm that the NCM cathode material had suffered from Li.sup.+ loss, ICP measurement was performed on pristine and cycled NCM cathode. The composition of the NCM cathode is changed from Li.sub.1.005Ni.sub.0.331 Co.sub.0.341Mn.sub.0.330O.sub.2 to Li.sub.0.796Ni.sub.0.326 Co.sub.0.342Mn.sub.0.329O.sub.2, which means the NCM cathode has about 20% Li.sup.+ loss after cycling.

[0092] Through this further study, one difference we found between NCM and LCO cathode material is their tolerance of underdosing and overdosing of Li. During the short annealing process, the evaporation of Li can lead to a slight underdosing of Li. Under such conditions NCM cathode material has the problem of cation mixing between Li and Ni ions, which means Ni.sup.2+ might take the place of Li.sup.+, deteriorating the electrochemical performance. Therefore, although the hydrothermal treatment can result in stoichiometric Li concentration in regenerated NCM cathode, a small amount of excess Li source (e.g., 5% Li.sub.2 CO.sub.3) is added to compensate the Li loss for regenerating LCO and NCM mixed cathode material. Considering the optimum temperature for the solid-state reaction to introduce Li was 850° C. according to experimental results, the short annealing temperature was increased from 800° C. to 850° C. for the processing of mixed cathode material. To prove that LiCoO.sub.2 has tolerance for slight overdosing of Li, pristine LiCoO.sub.2 was sintered with a small amount of Li source at 850° C. for 4 h. The cycling performance of the slightly overdosed LiCoO.sub.2 was similar with the pristine material, as illustrated in FIG. 8D. In addition, pure cycled NCM cathode from the home-made pouch was regenerated through hydrothermal treatment at 220° C. for 4 h, followed by short annealing with small amount of excess Li.sub.2 CO.sub.3 at 850° C. for 4 h. The XRD patterns of pristine and regenerated NCM powders are shown in FIG. 8E, and the regenerated material has well-defined layered structure, similar to pristine NCM powder. The cycling performance of pristine and regenerated NCM powders is shown in FIG. 8F, and the regeneration process fully recovered the capacity and cycling stability of pristine NCM cathode. Based on the above experimental results, the mixed cathode material was regenerated through hydrothermal treatment at 220° C. for 4 h, followed by short annealing with a small amount of Li sources at 850° C. for 4 h.

[0093] Two batches of cycled mixtures were regenerated by this approach. The first batch was the mixture of degraded cathode materials from cycled LCO pouch cells and cycled self-made NCM pouch cells (denoted as cycled LCO+NCM). The second batch is the degraded cathode material from cycled self-made LCO−NCM mixed pouch cell (denoted as LCO−NCM mixed pouch). The mass ratio between LCO and NCM in these two batches was both 1:1. Pristine mixed LCO and NCM material in a mass ratio of 1:1 serves as the reference for its comparison with the regenerated materials. The voltage-capacity profiles of pristine and regenerated mixed cathode materials illustrated in Panel (a) of FIG. 9 show two large plateaus at 3.9 V and 3.75 V, agreeing well with the charge/discharge behavior of LCO and NCM, respectively. As shown in Panel (b) of FIG. 9, the regenerated mixed cathode material could fully recover the original capacity and cycling stability. The initial discharge capacity for the pristine mixed material is 156.1 mAh g.sup.−1, and maintains a discharge capacity of 129 mAh g.sup.−1 after 100 cycles. After regeneration of the cycled LCO+NCM and LCO−NCM mixed pouch cathodes, their initial discharge capacities are 159 and 156.8 mAh g.sup.−1, with discharge capacities of 128.1 and 127.3 mAh g.sup.−1 in the 100.sup.th cycle, respectively. These results suggest the successful regeneration of mixed cathode materials using the aforementioned process. It is known that the phase change in the crystal structure is one important reason for the capacity degradation in NCM cathode. The migration of transition metals to the Li layers leads to phase change from layered to spinel structure. The difficulty to regenerate a NCM cathode is that transition metals and Li need to be re-constructed. To more directly prove that the cycled NCM we obtained is representative of degraded electrode with transition metal/Li disorder, and such disorder can be re-constructed after regeneration, TEM images of the degraded and regenerated electrode particles were taken. More than 20 particles were observed. Panels (c) and (d) of FIG. 9 show the representative images at the zone axis of [−1-21], from which any existing layered, spinel and rock salt structures can be observed. In the cycled electrodes (Panel (c)), the spinel phase was observed at the surface of active particles, evidenced by the existence of (−220) plane, which is the proof of migration of transition metal to the Li layers, consistent with literature report. After regeneration (Panel (d)), no spinel phase can be observed at the surface. Therefore, it is proved that: (1) the sample is representative of degraded electrodes with transition metal/Li disorder; and (2) transition metal and Li are successfully re-constructed after regeneration. More systematic research can be performed in the future to further investigate the influence of different recycling conditions on the electrochemical performance of regenerated mixed cathode material. Nevertheless, this set of experiments is a strong demonstration of the feasibility of this approach on processing mixed cathode materials. The phase change in the crystal structure is one important reason for the capacity degradation in LiCoO.sub.2 and NCM cathode materials.

[0094] Overall, this cathode regeneration strategy offers significant advantages. The common acid leaching approach requires the usage of corrosive acids, as well as complicated neutralization and precipitation steps to recover metals and reduce waste, while the solid-state synthesis requires chemical analysis and accurate control of Li.sup.+ dosage which makes large-scale operation difficult. Compared with both approaches, the hydrothermal approach with short annealing requires neither complicated chemical processing nor tedious elemental analysis to determine the Li.sup.+ loss, and can readily process batteries with different conditions of capacity degradation. The alkaline solution after processing cathode material can be reused, since the hydrothermal process will only slightly change the concentration of Li.sup.+ in the solution. Considering that the solid particles can be easily separated from the solution, we could recycle and reuse the alkaline solution after processing a batch of cathode material. Extra LiOH can be added to the used alkaline solution to compensate the decreased LiOH concentration, and the adjustment could be easily done by measuring the pH value of the alkaline solution. In addition, only a short period of annealing step is needed, rather than the long-time sintering required in solid-state synthesis approach to allow slow solid-state diffusion. Therefore, the hydrothermal approach both increases the ease of operation and decreases the energy cost for processing.

[0095] Besides the NMP dissolution approach used in this study to separate active materials, there are other solvents for PVDF to replace the toxic NMP, such as acetone. In addition, the dissolution process is not the only way to separate active materials, binder and conductive additives. Thermal treatment, for example, could be another choice to separate different components in electrodes. Researchers have successfully used thermal treatment to liberate the electrode particles from the current collectors by a vibrating screening, which can be easily performed in large-scale industry process. The details on the separation methods are not further discussed because the focus of the current study is the regeneration of the active material, rather than the separation of different components in electrodes.

[0096] To further understand the energy efficiency of this approach, we have compared the energy consumption of this process with the solid-state synthesis approach. Due to the complexity of the energy consumption calculation regarding various instrument and operation efficiency, we simplified the calculation by considering only the energy consumption required to heat the material and keep it at the desired temperature.

[0097] According to the equation:

[00004]$\begin{array}{cc}Q={\int }_0^t\left[{\mathrm{mC}}_P\frac{\mathrm{dT}}{\mathrm{dt}}+\mathrm{Ah}\left(T-T_{\mathrm{amb}}\right)\right]\mathrm{dT},& \left(5\right)\end{array}$

in which Q as the required energy, t as time, m as mass, C.sub.p as the specific heat capacity, A as the surface area exposed to air, h as the convective heat transfer coefficient, T as the temperature of the material, T.sub.amb as the ambient temperature, we can calculate the energy needed for both processes. The parameter values for the energy consumption calculation are provided in Table 4.

TABLE-US-00004 TABLE 4 C.sub.P of LiCoO.sub.2 C.sub.P of LiOH C.sub.P of Water h T.sub.amb 71.62 J 49.58 J 75.3 J 12 W 298 K mol.sup.−1 K.sup.−1 mol.sup.−1 K.sup.−1 mol.sup.−1 K.sup.−1 m.sup.−2 K.sup.−1
We take the volume ratio between solid material (cathode particles) and water in hydrothermal reactor as 1:1, which is easily achievable in this solid-liquid two phase reaction. Considering the processing of 1 kg LiCoO.sub.2, A for the mixture of LiCoO.sub.2 and LiOH solution is assumed to be 0.033 m.sup.2, and A for the pure LiCoO.sub.2 powder is assumed to be 0.021 m.sup.2, based on the tap density of 3.5 g/cm.sup.2 for LiCoO.sub.2. For hydrothermal plus short annealing approach, the energy consumption to heat 1 kg LiCoO.sub.2 together with LiOH solution to 220° C. and keep 4 h is calculated to be 1589.4 kJ, and to heat LiCoO.sub.2 to 800° C. and keep 4 h is calculated to be 4287.5 kJ. The total energy consumption is 5876.9 kJ. For solid-state synthesis, the energy consumption to heat 1 kg LiCoO.sub.2 to 850° C. and keep 12 h is calculated to be 10614.1 kJ. Therefore, the energy consumption of hydrothermal treatment is much less than that of the solid-state synthesis, which means the hydrothermal plus short annealing approach is more energy efficient than the solid-state synthesis approach.

[0098] In terms of electrochemical performance, the regenerated active particles from this process can achieve better rate capability than those regenerated through the best condition in solid-state synthesis while offer similarly high capacity and cycling stability. Furthermore, the hydrothermal approach can be used to regenerate mixed cathode materials, which makes it more attractive, considering mixed cathode chemistry is more likely to be used in the LIB industry. Therefore, compared with the state-of-the-art approaches, this work provides a promising strategy to regenerate spent LiCoO.sub.2 cathodes with easy processing and low energy consumption without generating additional wastes.

[0099] In summary, a simple yet efficient non-destructive approach is disclosed to recycle and regenerate LiCoO.sub.2 particles from spent LIBs by combining hydrothermal treatment and short annealing. This approach could fully recover the specific capacity and cycling stability of LiCoO.sub.2 without changing the original morphology and size distribution. Compared with the solid-state synthesis approach, the LiCoO.sub.2 particles regenerated through hydrothermal approach show improved rate capability, which are attributed to the smaller charge transfer resistance and larger Li.sup.+ diffusion coefficient. This strategy represents a simple and energy efficient approach to regenerate spent LiCoO.sub.2 cathodes with high electrochemical performance, and can be applied on industrial-scale operation. Furthermore, the inventive regeneration process is applicable to other types of cathodes in LIBs, such as LiMn.sub.2O.sub.4, LiFePO.sub.4 and Li.sub.xNi.sub.y Mn.sub.z Co.sub.1−y−zO.sub.2 (0<x,y,z<1).

[0100] The aforementioned work on combining hydrothermal treatment with short thermal annealing to regenerate degraded LiCoO.sub.2 (LCO) particles has demonstrated the successful reconstruction of stoichiometry composition and desired crystalline structure from severely degraded LCO cathode materials. However, the cathode reactivity and stability may change dramatically with their original composition and crystal structure. The complex chemistry in layered oxide LiNi.sub.x Co.sub.y Mn.sub.zO.sub.2 (NCM) cathodes can influence the change of crystal structure and local phase after cycling, which further affects the regeneration process. Accordingly, challenges may arise from the different degradation mechanisms of NCM cathode materials compared with simple LCO. More specifically, besides the Li.sup.+ loss due to the thickening of solid electrolyte interface (SEI), the crystal structure and microphase change on the particle surface (or sub-surface) is a major reason for the capacity degradation in layered oxide cathodes. For LCO, spinel phases such as Co.sub.3O.sub.4 and LiCo.sub.2O.sub.4 can form after degradation, while in the case of NCM, the phase change is more complicated. Due to the Li.sup.+ deficiency and migration of Ni.sup.2+ between the layers, the rock salt phase (e.g., NiO) will form at the surface besides the common spinel phase. Both phases increase the charge-transfer resistance and reduce the cathode performance. However, reconstructing rock salt phase into Li.sup.+ conducting layered structure is challenging due to the thermodynamically unfavorable nature of this reaction.

[0101] Despite these challenges, a particle-to-particle approach was developed for successful regeneration of degraded NCM cathodes. Using non-destructive methods, nearly ideal stoichiometry, low cation mixing and high phase purity were achieved in the regenerated NCM particles, which offer high specific capacity, cycling stability and rate capability reaching pristine materials. This work represents a simple yet efficient approach to directly regenerate high-performance NCM cathodes with distinct advantages over traditional hydrometallurgical methods and builds an important foundation for the sustainable manufacturing of energy materials.

[0102] The effort focuses on developing non-destructive approaches to directly regenerate degraded NCM cathode particles by resolving their compositional and structural defects. Specifically, a hydrothermal treatment combined with a short annealing was used in controlled atmospheres to regenerate NCM cathode particles. As a comparison, direct solid-state sintering approach was also examined to understand the activity of degraded NCM particles. The reaction mechanism was carefully investigated during different cathode regeneration processes. LiNi.sub.1/3 Co.sub.1/3Mn.sub.1/3O.sub.2 (NCM111) and LiNi.sub.0.5 Co.sub.0.2Mn.sub.0.3O.sub.2 (NCM523) cathodes were selected as the model materials to study the effect of nickel content on the evolution of particle stoichiometry and microphase. With optimized conditions, both spinel and rock salt phases can be fully converted back to layered phase using these direct regeneration approaches, as confirmed by systematic physicochemical characterizations. The lithium storage capacity and cycling stability of the degraded NCM111 and NCM523 cathode particles can be also recovered to the original levels of the pristine materials.

[0103] Dry pouch cells (220 mAh) with NCM523 as the cathode and graphite as the anode were directly purchased from Li-Fun Technology (Xinma Industry Zone, Golden Dragon Road, Tianyuan District, Zhuzhou City, Hunan Province, PRC, 412000). Electrolyte was filled in and the pouch cell was sealed by the vacuum sealer (MTI corporation). The electrolyte (LP40) was 1 M LiPF.sub.6 in ethylene carbonate (EC) and diethyl carbonate (DEC) with a weight ratio of 1:1. After formation at C/10 (C=150 mA g−1) for the first cycle, the pouch cells were cycled in the voltage range of 3-4.5 V at 1 C for 200 cycles. Commercial LCO pouch cells were purchased from MTI Corporation (2000 mAh, EQ-PL-605060-2 C). NCM111 pouch cells were assembled as described in previous research, with Li metal foil as the anode and a typical electrode area of 20 mm×55 mm. The NCM111 powder was obtained from Toda America. LCO and NCM111 pouch cells were also cycled in the voltage range of 3-4.5 V to gain >20% capacity loss. All pouch cells were discharged to 2 V before disassembly.

[0104] The cathode strips were harvested from the pouch cell, thoroughly rinsed by dimethyl carbonate and soaked in NMP followed by sonication. The active materials, binder and carbon black were removed from the aluminum substrate. The suspension is centrifuged and the active materials were precipitated. The precipitation was washed several times and the active materials were harvested and dried.

[0105] Regeneration of cathode materials: The composition of cycled cathode was measured by an Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES, Perkin Elmer Optima 3000 DV). For hydrothermal treatment, cycled cathode materials were added in a 100 mL Teflon liner of an autoclave filled with 80 mL of 4 M lithium hydroxide (LiOH) solution. In some embodiments, a lithium containing solution can be made by a small concentration (e.g., 0.1 M) of lithium salt (e.g., LiOH, Li.sub.2SO.sub.4, LiCl, LiNO.sub.3) with an alkaline solution from NaOH, KOH, NH.sub.4OH or their mixture. The autoclave was heated (at ambient) for different periods of time and temperatures. The treated powders were washed with deionized water and sintered with 5% excess amount of Li source (Li.sub.2 CO.sub.3) in oxygen at 850° C. for 4 h with a ramping rate of 5° C./min. 5% excess amount of Li was added to compensate the Li evaporation during the sintering process. For solid-state sintering, the cathode powders were mixed with Li.sub.2 CO.sub.3 with agate mortar and pestle. The amount of Li.sub.2 CO.sub.3 was calculated to make the mixture with a Li/Co ratio of 1.05. The mixtures were sintered at 850° C. for 12 h with a ramping rate of 5° C./min. The sintering process was performed in air and oxygen atmosphere, respectively.

[0106] Characterization of regenerated materials: The morphology of the powders was observed by Ultra High Resolution Scanning Electron Microscope (UHR SEM, FEI XL30). The particle size distribution was analyzed with Nano Measurer software. The crystal structure of the powders was examined by X-ray Powder Diffraction (XRD) employing Cu Kα radiation. The crystal structure was also examined by Transmission Electron Microscopy (TEM) (FEI Titan 80-300 kV S/TEM). The XPS measurement was performed with Kratos AXIS Ultra DLD with Al Kα radiation.

[0107] Electrochemical characterization: The pristine, cycled and regenerated cathode materials were mixed with PVDF, and Super P65 in NMP at a mass ratio of 8:1:1. Then the slurries were cast on aluminum foil using a doctor blade and dried in vacuum at 80° C. for 6 h. Circle electrodes were cut and compressed by rolling mill. The active mass loading was about 3 mg/cm.sup.2. Type-2016 coin cells were assembled with Li metal disc (thickness 1.1 mm) as the anode, 1 M LiPF.sub.6 in EC:DEC (1:1 wt) as the electrolyte, and trilayer membrane (Celgard 2320) as the separator. Galvanostatic charge-discharge was carried out using a LAND battery testing system in the potential range of 3-4.3 V at 1 C after C/10 in the initial cycle. The electrochemical impedance spectroscopy (EIS) tests were carried out at discharged state in the frequency range of 106 Hz to 10-2 Hz with a signal amplitude of 10 mV by Metrohm Autolab Potentiostats.

[0108] FIG. 10 illustrates the flow for an exemplary method of regenerating degraded NCM cathode materials using a particle-to-particle approach. In step 1002, the degraded or cycled NCM cathode materials are obtained, and in step 1004, lithium (Li) is pre-dosed into the Li-deficient NCM cathode particles and the NCM cathode particles are treated with a hydrothermal treatment, after which a short thermal annealing is performed in step 1006. Once the NCM cathode particles have been regenerated, in step 1008, the particles may then be processed into a slurry which, in step 1010, can be fabricated into new battery cells.

[0109] Both commercial and home-made cells were used for the demonstration and performance evaluation. The cell assembly and the materials harvesting for LCO and NCM111 pouch cells followed standard procedures. The assembly of NCM523 pouch cells are described in the Experimental section above, with the same materials harvesting procedures as LCO and NCM111 pouch cells. All pouch cells were cycled in the voltage range of 3-4.5 V at 1 C with capacity degradation of more than 20% over approximately 200 cycles, as shown by the graph of cycling performance of the NCM523 pouch cell in FIG. 11. The cell was cycled in the voltage range of 3-4.5 V at 1 C (C=150 mA g.sup.−1). The capacity degradation is 22% after 200 cycles.

[0110] The degraded cathode particles were obtained and subject to hydrothermal treatment with a short thermal annealing step (denoted as HT-SA), or to a direct solid-state sintering treatment by mixing with Li salt (denoted as SS). The regenerated cathode materials were made into slurries to fabricate new coin cells to evaluate their electrochemical performance for the sake of convenience and consistence, as described in the Experimental section. The scanning electron microscopic (SEM) images of pristine NCM523 particles are shown in Panel (a) of FIG. 12, with the size distribution of the pristine NCM523 particles displayed in Panel (b). SEM images of cycled (degraded) NCM523 particles are shown in Panel (c) of FIG. 12, with the size distribution of the cycled NCM523 particles displayed in Panel (d). The left images of the two SEM images in Panels (a) and (c) show secondary particles, while the right images show primary particles.

[0111] The particle morphology and size distribution of NCM111 samples were also monitored, as illustrated in FIG. 13. Panel (a) shows SEM images of the samples, while the corresponding Panel (b) illustrates size distributions of the NCM111 particles. All the particles have similar morphology and sizes, in which the sizes of the secondary particles mainly distribute in approximately 3-5 micrometers (m), with primary grains of approximately 0.2-1.0 m in diameter. The spherical secondary particle was maintained after cycling. Since the micro-phase change is one of the most important reasons for the cathode degradation, high resolution transmission electron microscopic (HR-TEM) images were taken to directly observe the changes in crystal structure after cycling. More than 20 particles were examined for each sample. As expected, the pristine NCM 111 cathodes shown in the HR-TEM and Fast Fourier Transform (FFT) images taken along the [−1-21] zone axis shown in FIG. 14. The pristine NCM523 cathodes shown in the HR-TEM image and FFT image of Panel (e) in FIG. 12 only show layered rhombohedral structure, with hexagonal diffraction pattern showing (012) plane. As shown in FIG. 15, the cycled NCM111 particles have layered phase in the bulk region and show spinel phase near the surface, as shown in the HR-TEM and FFT images of cycled NCM111 along the [−1-21] zone axis. Diffraction spots from the spinel phase were detected along with the layered phase, with the additional diffraction spots indexed as (−220).sub.s. Similarly, diffraction spots from the spinel phase were also observed in cycled (degraded) NCM523 shown by the HR-TEM and FFT images in Panel (f) of FIG. 12, demonstrating co-existence of layered, spinel and rock salt phases. Besides the spinel phase, the rock salt phase appears in the region of ˜3 nanometers (nm) from the surface of the NCM523 particles, with a reduction in the number of diffraction spots due to the high symmetry of rock salt phase, and the diffraction spots are indexed as (−220).sub.c. No rock salt phase was clearly detected in cycled NCM111 cathode in this case. The phase transformation in the surface region could be due to the Li.sup.+ deficiency near the surface, which is generally observed in layered cathode material because it is a thermodynamically favored transformation when the Li contents are reduced to half of that in the original structure.

[0112] With the above understanding, the first step to regenerate the degraded cathode particles is to re-dose lithium using a hydrothermal-based solution impregnation method. FIG. 16A illustrates the lithiation process of various cathode particles during the hydrothermal treatment, in which Li.sup.+ are re-dosed to the Li-deficient sites to recover its desired stoichiometry. Interestingly, significant differences were found for different cathode chemistries, as illustrated in FIG. 16B by the lithiation kinetics of degraded LIB cathode particles during hydrothermal treatment. The Li.sup.+ concentration of different cathode particles changed dramatically with hydrothermal treatment temperature and time. For degraded LCO particles (Li.sub.0.8 CoO.sub.2) 1602, the Li concentration can be recovered to 0.98 after being treated at 180 degrees Celsius (° C.) for 4 hours (h), which is not the perfect stoichiometry, but common for LCO even in the pristine particles. However, for degraded NCM111 and NCM523 particles 1604, the Li stoichiometry cannot be fully recovered (e.g., only 0.95) after being treated at 180° C. for even 24 h. Nevertheless, once the temperature increases to approximately 220° C., the Li concentration can be fully recovered after a treatment of only 4 h. The different lithiation kinetics of NCM and LCO may be related to the higher degree of cation mixing in NCM due to the similar sizes of Ni.sup.2+ (0.69 Angstroms (Å)) and Li.sup.+ (0.72 Å). Since Ni.sup.2+ ions occupy Li.sup.+ sites, the activation energy barrier is higher for the diffusion of Li.sup.+ because of the smaller separations between the transition metal layers. For the purpose of demonstration, hydrothermal treatment at 220° C. for 4 h was selected to lithiate the degraded NCM111 and NCM523 electrode particles for annealing in the next step.

[0113] The compositions of different pristine, degraded and regenerated NCM cathode materials as measured by Inductively-Coupled Plasma (ICP) are listed in Table 5. Compared with their pristine composition, both NCM111 and NCM523 particles had about 22% of Li loss after cycling. With the hydrothermal treatment, these degraded particles can be reconstituted with Li to reach the ideal stoichiometry (˜1.0 Li). A following thermal annealing step needs to be performed to reconstruct their desired microphase and crystallinity and maintain the Li concentration in the particles during the thermal treatment.

TABLE-US-00005 TABLE 5 Sample NCM111 NCM523 Pristine Li.sub.0.995Ni.sub.0.331Co.sub.0.341Mn.sub.0.329O.sub.2.012 Li.sub.1.009Ni.sub.0.492Co.sub.0.209Mn.sub.0.305O.sub.2.015 Cycled Li.sub.0.786Ni.sub.0.328Co.sub.0.340Mn.sub.0.325O.sub.1.996 Li.sub.0.788Ni.sub.0.490Co.sub.0.208Mn.sub.0.302O.sub.1.985 HT only Li.sub.1.012Ni.sub.0.329Co.sub.0.341Mn.sub.0.326O.sub.2.009 Li.sub.1.006Ni.sub.0.491Co.sub.0.209Mn.sub.0.304O.sub.2.012 SS-air Li.sub.1.016Ni.sub.0.331Co.sub.0.340Mn.sub.0.325O.sub.2.010 Li.sub.1.017Ni.sub.0.490Co.sub.0.207Mn.sub.0.304O.sub.2.014 SS-oxygen Li.sub.1.016Ni.sub.0.331Co.sub.0.341Mn.sub.0.326O.sub.2.012 Li.sub.1.018Ni.sub.0.491Co.sub.0.208Mn.sub.0.304O.sub.2.013 HT-SA Li.sub.1.019Ni.sub.0.330Co.sub.0.340Mn.sub.0.327O.sub.2.011 Li.sub.1.021Ni.sub.0.490Co.sub.0.209Mn.sub.0.303O.sub.2.013

[0114] In the LCO cathode, only spinel phases (Co.sub.3O.sub.4 and LiCo.sub.2O.sub.4) are formed after cycling, while in the cycled NCM cathodes, nanoscale domains of rock salt phase often exist besides the spinel phases. To convert the local rock salt MO (M=Ni, Co, Mn) domains back to layered LiMO.sub.2, the following reaction should occur:

MO+0.5Li.sub.2 CO.sub.3+0.25O.sub.2.Math.LiMO.sub.2+0.5 CO.sub.2  (6)

This reaction indicates that oxygen partial pressure may be an important factor for the conversion process. Therefore, for comparison, the degraded NCM111 and NCM523 particles were mixed with pre-determined amount of Li.sup.+ salts to perform direct sintering to reach a target mole ratio between Li and transition metal ions (1.05:1) in both air and oxygen atmosphere. As shown in Table 5, both of the particles can reach desired overall compositions, indicating their non-sensitivity to O.sub.2 partial pressure.

[0115] For the HT particles, a short annealing treatment at 850° C. for 4 h was performed in O.sub.2 to reconstruct the desired crystallinity of the material. Also, considering the possible Li loss during the annealing which can lead to cation mixing, a small amount of excess Li was added to compensate such a Li loss. Similarly, the particles with this short annealing treatment also reached the target stoichiometry.

[0116] It is also critical to investigate the evolution of the microstructure defects. FIG. 17 illustrates X-Ray Diffraction (XRD) patterns of pristine, cycled and regenerated cathode particles, with Panel (a) showing NCM111 particles and Panel (b) showing NCM523 particles by HT-SA, SS-air and SS-oxygen approaches; the insets at right show enlargement of the regions in the range of approximately 18.5-19.5° and approximately 64-66°. For both NCM111 and NCM523, the cycled cathode particles show a larger intensity ratio of 1003/1104, suggesting higher cation mixing, which is consistent with the previous report. The (003) peak shifts to lower angles, corresponding to an increase in c lattice parameter due to the electrostatic repulsion between the oxygen layers along c directions in the Li deficiency state. The spacing between the peaks in the (108)/(110) doublets increases after cycling, corresponding to the decrease in a lattice parameters due to the smaller effective ionic radii of Ni.sup.3+ than Ni.sup.2+ to compensate Li deficiency. After different regeneration processes, the (003) peak shifts back towards higher angles and the spacing between two doublets peaks decreases, which indicates the recovery of the pristine crystal structure. Rietveld refinement was performed on all the XRD patterns (FIG. 18 and FIG. 19), and the lattice parameters for pristine (Panel (a)), cycled (Panel (b)) and regenerated cathode particles (Panels (c)-(e)) are compared in Table 6 below. FIG. 18 illustrates Rietveld refinement of the XRD patterns comparing pristine, cycled and regenerated NCM111 cathodes, while FIG. 19 illustrates Rietveld refinement of the XRD patterns comparing pristine, cycled and regenerated NCM523 cathodes. The refinement results further confirm that the degraded particles have a decreased a lattice parameter and increased c lattice parameters, clearly showing increased Li/Ni cation mixing.

TABLE-US-00006 TABLE 6 Sample a/Å c/Å Li/Ni mixing R.sub.B R.sub.wp NCM111 pristine 2.8631(7) 14.248(3) 2.42% 5.92% 1.47% NCM111 cycled 2.8600(4) 14.258(4) 2.81% 4.9% 1.52% NCM111-HT-SA 2.8629(4) 14.246(2) 2.07% 4.22% 1.43% NCM111-SS-air 2.8628(6) 14.250(2) 2.45% 5.61% 1.58% NCM111-SS-oxygen 2.8624(7) 14.244(7) 2.43% 6.14% 1.51% NCM523 pristine 2.8689(4) 14.240(6) 3.39% 4.41% 1.65% NCM523 cycled 2.8591(0) 14.319(2) 5.10% 4.92% 2.53% NCM523-HT-SA 2.8703(8) 14.249(5) 3.74% 4.99% 2.17% NCM523-SS-air 2.8729(9) 14.255(9) 5.53% 3.93% 1.71% NCM523-SS-oxygen 2.8682(9) 14.244(8) 4.28% 4.96% 1.75%

[0117] For all regeneration conditions, the a and c lattice parameters change to higher and lower values, respectively. By comparing HT-SA and SS approaches, it shows that the I.sub.003/I.sub.104 intensity ratio of the particles regenerated by the former approach is higher than the latter. This indicates smaller Li/Ni cation mixing of the material regenerated by the HT-SA, which is further confirmed by the refinement results in Table 6. It is also noted that the cation mixing of the NCM111-SS-air sample is similar with that of the NCM111-SS-oxygen sample, while the cation mixing of NCM523-SS-air is larger than that of NCM523-SS-oxygen. Since a high nickel content is considered the key factor for the formation of the rock salt phase, it is speculated that the rock salt phase tends to form more easily in cycled NCM523 cathode than that in cycled NCM111 cathode. As oxygen atmosphere is a critical factor that turns the rock salt phase into the layered phase, the added Li source may not effectively react with the rock salt phase when the oxygen partial pressure is low, and the migration of Ni.sup.2+ to Li.sup.+ sites continues to happen in a Li.sup.+ deficient state, which leads to higher cation mixing degree in the NCM523-SS-air. Overall, the HT-SA samples show much smaller Li/Ni mixing, suggesting its advantage of offering a more homogenous lithiation and more effective phase conversion for particle regeneration.

[0118] Even though no obvious changes in the morphology and particle size distribution are observed after direct regeneration—as shown in FIG. 20 (and previously in FIG. 13) by the representative SEM images of NCM523-HT-SA particles in Panel (a) and corresponding size distribution graph in Panel (b), their microstructure needs further examination. To prove that the surface phase change can be recovered, the regenerated cathodes were carefully examined by HR-TEM, as shown in Panels (c), (d) and (e) of FIG. 20. The influence of short annealing after the hydrothermal step was investigated by comparing the images of NCM523-HT and NCM523-HT-SA samples. For NCM523-HT, there remain some amorphous domains on the surface, as shown by the HR-TEM and FFT images of NCM523 in Panel (a) of FIG. 21, but the NCM523-HT-SA sample in Panel (b) of FIG. 21 shows a high degree of crystallinity at the surface for each observed particle. Among the 24 particles that were observed, 7 of them show amorphous phase in the region ˜20 nm from the surface. For NCM111-HT-SA and NCM523-HT-SA samples (as shown in HR-TEM and FFT images in FIG. 22 and Panel (c) of FIG. 20, respectively), only layered phase in both the bulk and the surface regions was observed. Both the bulk and the surface show layered rhombohedral structure with diffraction pattern showing (012) plane. For example, in the zone axis of [−1-21], only layered phase exists with the diffraction spot indexed as (012). This indicates that both the spinel and rock salt phases can be effectively converted back to the layered phase by the HT-SA approach.

[0119] In addition, it was found that the change in microphase of NCM523 shows higher sensitivity to the oxygen partial pressure than NCM111 in SS regeneration. The SS in air can convert the spinel phase to layered phase in cycled NCM 111 cathode, as illustrated by the HR-TEM and FFT images of cycled NCM111-SS-air sample along the [−1-21] zone axis shown in FIG. 23. Both the bulk and the surface show layered rhombohedral structure with diffraction pattern showing (012) plane. However, for the NCM523 cathode, some rock salt phases can still be observed on the particle surface after SS in air (NCM523-SS-air samples in Panel (d) of FIG. 20), which indicates that the oxygen partial pressure is important for such a phase conversion. By comparison, for the SS in oxygen, no spinel or rock salt phase is observed, as shown by the NCM523-SS-oxygen samples in Panel (e) of FIG. 20, which means a successful regeneration of the layered structure in cycled NCM523 cathode. As expected, SS in oxygen can also fully recover the layered structure in cycled NCM111 cathode as well, as shown by the HR-TEM and FFT images of cycled NCM111-SS-oxygen sample along the [−1-21] zone axis shown in FIG. 24. Both the bulk and the surface show layered rhombohedral structure with diffraction pattern showing (012) plane.

[0120] To provide further evidence of spinel/rock salt phase in the cycled cathodes and the successful reconstruction of the layered phase after regeneration, x-ray photoelectron spectroscopy (XPS) measurement was performed on NCM111 (as shown in FIG. 25) and NCM523 (as shown in Panel (f) of FIG. 20). FIG. 25 illustrates XPS patterns of pristine NCM111, cycled NCM111, NCM111-HT-SA, and NCM111-SS-oxygen samples. In Ni 2p spectra, Ni.sup.2+ peaks are observed in all NCM111 samples, while Ni.sup.3+ peaks are only observed in cycled NCM111 sample. In Mn 2p spectra, Mn.sup.4+ peaks are observed in all NCM111 samples, while Mn.sup.3+ peaks are only observed in cycled NCM111 samples. For Ni 2p spectra, all cathodes have two dominant peaks at 854.6 eV (2p3/2) and 872.2 eV (2p1/3) which represent Ni.sup.2+, and the two less dominant shake-up peaks at 860.9 eV and 879.2 eV further confirm the existence of Ni.sup.2+. In cycled NCM111 and all NCM523 cathodes, besides the existence of Ni.sup.2+, the less prominent peaks at 857.2, 864.3, 875.5 and 882.9 eV indicate the existence of Ni.sup.3+. The quantitative analysis shows that the Ni.sup.2+ concentration of NCM111-cycled, NCM523-pristine, NCM523-cycled, NCM523-HT-SA and NCM523-SS-oxygen samples are 55.14%, 60.22%, 72.56%, 60.37%, and 60.44%, respectively. The significantly higher Ni.sup.2+ concentration in the cycled NCM523 cathode is consistent with the existence of NiO rock salt phase on the surface by the TEM observation. For Mn 2p spectra, all samples have two major peaks at 642.3 eV (2p.sub.3/2) and 653.8 (2p.sub.1/3) which represent Mn.sup.4+. The peaks at 640.9 and 652.4 eV found in cycled NCM 111 and NCM523 samples indicate the existence of Mn.sup.3+, while such peaks were not observed in pristine and regenerated samples. The quantitative analysis of Mn 2p spectrum shows that the Mn.sup.3+ concentration is 34.58% and 36.54% in cycled NCM111 and NCM523 cathodes, respectively, which is expected because the layered to spinel/rock salt transformation originates from oxygen loss, and results in the formation of Mn.sup.3+ for charge compensation. Therefore, the existence of Mn.sup.3+ in cycled cathodes and its disappearance in regenerated cathodes further support that the spinel and rock salt phases formed after cycling was recovered into layered structure after regeneration.

[0121] FIG. 26 illustrates results comparing the electrochemical performance of the cathode particles. Panel (a): the cycling performance of pristine, non-treated and regenerated NCM111 samples at 1 C; HT-SA: hydrothermal treatment at 220° C. for 4 h, followed by annealing at 850° C. for 4 h; SS-air: sintering at 850° C. for 12 h in air; SS-oxygen: sintering at 850° C. for 12 h in oxygen. Panel (b): the cycling performance of pristine, non-treated and regenerated NCM523 samples at 1 C. Panel (c): the rate performance of NCM111 samples. Panel (d): the rate performance of NCM523 samples. Panel (e): voltage profiles of NCM111 samples at 5 C. Panel (f): voltage profiles of NCM523 samples at Panel (c). Panel (g): illustration of the crystal structure change of NCM523 after cycling and regeneration. The right scheme in Panel (g) shows the atomic arrangement of layered, spinel and rock salt phases along the [−1-21] zone axis (same as TEM images).

[0122] The electrochemical performance of the NCM111 and NCM523 cathode particles in different conditions were evaluated in the voltage range of 3-4.3 V at 1 C (C=150 mA g−1) after one activation cycle at C/10. For NCM111 in Panel (a), the pristine cathode shows a capacity of 145.1 mAh g.sup.−1 in the first cycle at 1 C and 123.8 mAh g.sup.−1 after 100 cycles. In Panel (b), the non-treated, cycled cathode shows a capacity of 98.4 mAh g.sup.−1 after 100 cycles, which is due to the existence of spinel phase at the surface (see FIG. 15). The electrochemical activity of the regenerated cathodes from HT-SA, SS in air and SS in oxygen was fully recovered, with a capacity of 158.4, 153.3 and 157.4 in the first cycle at 1 C and 122.6, 125.4, and 123.8 mAh g.sup.−1 after 100 cycles, respectively. For NCM523, the pristine cathode shows a capacity of 146.6 mAh g−1 in the first cycle at 1 C and 130.4 mAh g.sup.−1 after 100 cycles (FIG. 26, Panel (b)). The non-treated, cycled cathode shows a poor cycling performance with only 88.6 mAh g.sup.−1 capacity retention after 100 cycles. The fast capacity decay is consistent with the observed spinel and rock salt phase at the surface (FIG. 12, Panel (f)). After regeneration by HT-SA and SS in oxygen, the cycling stability of both cathodes was fully recovered and the regenerated cathodes can maintain a capacity of 128.3 mAh g.sup.−1 (HT-SA) and 127.4 mAh g.sup.−1 (SS) after 100 cycles. However, even though the SS in air increases the specific capacity and capacity retention of the regenerated cathode, it cannot fully recover the cycling performance. This is also consistent with the TEM observation (Panel (d) in FIG. 20) which shows rock salt phase remaining at the surface.

[0123] The rate capability and voltage profiles of the pristine and regenerated cathodes at 5 C are compared in Panels (c), (d), (e) and (f) of FIG. 26. For both NCM111 and NCM523 samples, the HT-SA approach delivers better rate capability (Panels (c) and (d)), as well as smaller voltage drops and polarization (Panels (e) and (f)) than the SS approach. The rate capability of the NCM523-SS-oxygen sample is better than that of the NCM523-SS-air sample (Panel (d)), and the voltage drop and polarization of the former is smaller than the latter. The reason for the better rate capability of HT-SA samples is believed to relate to their lower Li/Ni mixing (Table 6). It was also reported before that hydrothermal treatment is effective to suppress the degree of Li/Ni mixing. Cation mixing blocks the Li.sup.+ transportation channel, and therefore decreases the rate capability. The different rate performance of NCM111-SS-air and NCM523-SS-air samples matches well with their microphase structures (FIG. 23 and Panel (d) of FIG. 20). The layered structure of NCM111-SS-air is recovered, while in NCM523-SS-air, the rock salt phase still exists, which results in the inhibition of Li.sup.+ motion.

[0124] To further understand the rate performance, Electrochemical Impedance Spectroscopy (EIS) measurement was performed on pristine and regenerated NCM523 cathodes after 100 cycles at 1 C. FIG. 27 illustrates Nyquist plots of pristine and regenerated NCM523 cathodes after 100 cycles (Panel (a)), where the inset displays the equivalent circuit to fit the Nyquist plots; and the relationship between the real parts of the complex impedance and ω.sup.−1/2 (Panel (b)). R.sub.s, R.sub.sei and R.sub.et are the ohmic resistance, solid electrolyte interface (SEI) resistance, and charge-transfer (CT) resistance, respectively. W is the Warburg impedance related to the Li.sup.+ diffusion. Table 7 lists the resistance values of pristine and regenerated NCM523 cathodes after 100 cycles. As seen in the table, the NCM523-SS-air sample has much larger charge-transfer resistance (R.sub.et) (367.4Ω) than the NCM523-SS-oxygen sample (198.7Ω); the NCM523-HT-SA sample has the smallest R.sub.ct (142.8Ω).

TABLE-US-00007 TABLE 7 R.sub.s/Ω R.sub.sei/Ω R.sub.ct/Ω D.sub.Li.sup.+/cm.sup.2 s.sup.−1 Pristine 6.374 18.68 151.6 3.66 × 10.sup.−12 HT-SA 6.963 21.72 142.8 4.85 × 10.sup.−12 SS-air 8.855 22.95 367.4 5.55 × 10.sup.−13 SS-oxygen 8.019 21.87 198.7 1.22 × 10.sup.−12

[0125] The superior rate performance of the HT-SA sample is attributed to its lowest R.sub.ct which favors the charge-transfer reaction for Li.sup.+ intercalation. The linear part of Nyquist plot in the low frequency range is directly related to Li.sup.+ diffusion in electrode, and the diffusion coefficient (D.sub.Li.sup.+) could be calculated by the EIS method using Warburg impedance (Table 7). The NCM523-HT-SA sample has the largest D.sub.Li.sup.+ of 4.85×10.sup.−12 cm.sup.2 s.sup.−1, and the NCM523-SS-oxygen sample has a larger DL (1.22×10.sup.−12 cm.sup.2 s.sup.−1) than the NCM523-SS-air sample (5.55×10.sup.−13 cm.sup.2 s.sup.−1). The lower R.sub.ct and higher D.sub.Li.sup.+ of the regenerated cathode by the HT-SA approach explain its better rate capability than the cathode regenerated by the SS approach. The remaining rock salt phase and cation mixing on the surface of the regenerated NCM523 in air (Panel (d) of FIG. 20) leads to the large R.sub.ct and low D.sub.Li.sup.+.

[0126] Overall, degraded NCM particles with different compositional deficiencies and microphase impurities can be effectively regenerated using these direct methods that combine hydrothermal lithiation and short annealing, which leads to ideal stoichiometry, low cation mixing and high phase purity. Panel (g) of FIG. 26 simply illustrates such a reversible process for particles after cycling and regeneration, using the NCM523 cathode as the example. After extensive cycling, scattered rock salt phase forms at the surface of NCM523 cathode together with the spinel phase accompanying Li loss. With direct regeneration, the undesired spinel and rock phases are converted back to the layered phase with lithiation and thermal annealing.

[0127] To summarize, the inventive scheme allows the regeneration of chemical composition and microstructure in degraded NCM cathodes using a novel particle-to-particle approach. Hydrothermal treatment with a short annealing provides the regenerated NCM cathode particles with high capacity, long cycling stability and high rate performance that of pristine materials, even with high Ni content. Due to the higher nickel content in NCM523 than NCM111 cathodes, the approach of direct solid-state sintering in air can restore the cycling stability of the latter but not the former. The oxygen partial pressure should be maintained at a high level to effectively convert the rock salt phase impurities to the layered phase in NCM cathodes with high Ni content.

[0128] The following examples are directed to further enhancements of the inventive regeneration approach of hydrothermal treatment and annealing. Specifically, these improvements are directed to optimization of efficiency from economic and environmental perspectives, with reduced raw materials costs and energy consumption.

Example 1: Direct Regeneration with Optimized Efficiency

[0129] FIGS. 1A-1C, described above, provide overviews of the process in which a lithium-deficient metal oxide was regenerated in two steps, hydrothermal relithiation (HT) and short annealing (SA).

[0130] The effects of hydrothermal temperatures and relithiation time on the electrochemical performance of the regenerated materials were examined first. D-NCM111 was relithiated at 160°, 200°, and 220° C. with 4 M LiOH for 1, 2, 4 and 6 h. The results of this test are plotted in FIG. 28. When D-NCM111 was relithiated at 160° C. (Panel (a)), the sample treated for 1 h showed an initial discharge capacity of 130 mAh/g at a C/3 rate and retained only 81.5% of its initial capacity after 50 cycles. When the relithiation time was increased to 2 h, the initial capacity improved to 135 mAh/g and the capacity retention increased to 83.7%, which was further improved when relithiation time was increased as shown in Table 8. When D-NCM111 was relithiated for 6 h, the capacity reached 147 mAh/g and the capacity retention increased to 85.0%. The treatment temperature was then increased to 200° C. (FIG. 28, Panel (b)). At this temperature, the sample relithiated for 1 h exhibited an initial capacity and capacity retention of 147 mAh/g and 81.6% after 50 cycles, respectively. A similar improvement trend with the increase of relithiation time to that of samples relithiated at 160° C. was also observed, indicating both the relithiation temperature and time are crucial. The temperature was then elevated to 220° C., the results for which are shown in FIG. 28, Panel (c). An initial capacity of 149 mAh/g was produced after only 1 h of relithiation, but only 120 mAh/g (80.5%) was retained after 50 cycles. However, when the relithiation time was increased to 2 h, the capacity at the end of the 50.sup.th cycle reached 131 mAh/g, corresponding to a retention of 87.3%. Longer treatment times, e.g., 4 h and 6 h, did not improve the cycling performance further. Full results for capacity retention of relithiated NCM111 at the C/3 rate are provided in Table 8 below.

TABLE-US-00008 TABLE 8 Discharge capacity (mAh/g) Retention Samples 1.sup.st cycle 50.sup.th cycle (%) HT-160° C.-1 h 130 106 81.5% HT-160° C.-2 h 135 113 83.7% HT-160° C.-4 h 142 119 83.8% HT-160° C.-6 h 147 125 85.0% HT-200° C.-1 h 147 120 81.6% HT-200° C.-2 h 148 124 83.7% HT-200° C.-4 h 148 127 85.8% HT-200° C.-6 h 149 130 87.2% HT-220° C.-1 h 149 120 80.5% HT-220° C.-2 h 150 131 87.3% HT-220° C.-4 h 150 130 86.7% HT-220° C.-6 h 151 130 86.1%

[0131] The temperature, time and initial capacity data in Table 8 were then evaluated for underlying trends using robust, polynomial regression as described below.

[0132] For two independent variables, the general, fitting polynomial was y=a.sub.1x.sub.1+a.sub.2x.sub.2+a.sub.12x.sub.1x.sub.2+b.sub.1x.sub.1(x.sub.1−x.sub.2)+b.sub.2x.sub.2(x.sub.1−x.sub.2), where x.sub.1 represented temperature in ° C., and x.sub.2, time in hours. The a.sub.nx.sub.n terms represented the linear contributions to y, and the others represented the contributions of the interactions between time and temperature. The linear terms, a.sub.1x.sub.1+a.sub.2x.sub.2, were always present in the polynomials.

[0133] A search algorithm for candidate polynomials was implemented in Visual Basic for Applications in Microsoft® Excel®. Robust linear regression was used to minimize the effect of noise in the data and Tukey's iterative, biweight function with a tuning constant of 6 was used to weight the data. The goodness-of-fit, r.sup.2, was calculated using the expressions given in equations (7)-(9).

[00005]$\begin{array}{cc}\mathrm{TSS}=.Math.y^2& \left(7\right)\\ \mathrm{RSS}=.Math.{\left(\hat{y}-y\right)}^2& \left(8\right)\\ r^2=\frac{\mathrm{TSS}-\mathrm{RSS}}{\mathrm{TSS}}& \left(9\right)\end{array}$

where TSS is the total sum of squares, y is the experimentally observed value, RSS is the residual sum of squares and y is calculated from the fit.

[0134] The search process generated many polynomials. Candidate polynomials that represented the data with a value of r.sup.2≥0.95 were selected for further consideration. The following criteria were used to further limit the number of polynomials. (1) The polynomial had the fewest number of terms; the fit was overdetermined. (2) The standard error had to be 40% or less than the value of the parameter so that, at the 95% confidence level, the value of the fitting parameter would still be larger than twice of the standard error.

[0135] Using the fitting polynomial, the values of the fitting parameters are given in Table 9, providing an r.sup.2 greater than 0.99.

TABLE-US-00009 TABLE 9 Parameter Value (s.e.e)* a.sub.1 0.73 (0.02) a.sub.2 26.29 (3.71) b.sub.1 −0.13 (0.02) *standard error of the estimate

[0136] The value of a.sub.2 is much greater than that of a.sub.1, indicating that the hydrothermal process time was the most important factor to obtain high-performing cathode materials. The negative value of b.sub.2, at first glance, is somewhat puzzling, and indicates that the interaction of time and temperature negatively affected the initial capacity of these materials. In the time and temperature range studied, the value of b.sub.2x.sub.2(x.sub.1−x.sub.2) was always negative and tended to decrease (became more negative) with increasing time at a given temperature. Together, these observations indicated that there was an optimum value beyond which additional hydrothermal time was unlikely to be beneficial.

[0137] The samples that showed the highest cycling performance, 160°-6 h, 200° C.-6 h, 220° C.-2 h, were selected for additional characterization. They are denoted as HT-160, HT-200 and HT-220, respectively. Referring to FIG. 29A, XRD was first performed to characterize the crystal structure. Overall, the samples treated at different temperatures as well as the control samples (T-NCM111 and D-NCM111) displayed the typical diffraction peaks associated with the α-NaFeO.sub.2 structure (R3m space group), indicating that the bulk structure of NCM111 is not affected by chemical dilithiation and relithiation. After chemical dilithiation (D-NCM111), the (003) peak clearly shifted to a lower angle, reflecting an increase in the c lattice constant due to the increased electrostatic repulsion between the oxygen layers along the c direction due to Li.sup.+ deficiency. After relithiation at 160° C., 200° C. and 220° C., the (003) peak shifted back to the same position as that of T-NCM111.

[0138] The lattice parameters were determined via Rietveld refinement of the neutron diffraction pattern, where the evolution of the a and c unit cell parameters were plotted in FIG. 29B. After chemical relithiation, the a lattice parameter of D-NCM111 decreased from 2.85761 (7) Å to 2.84370 (7) Å, which originates from the decrease in the average metal-metal distance that occurs upon dilithiation of the structure. The c lattice parameter increased from 14.2165 (4) Å to 14.2953 (4) Å, which is consistent with the XRD result. Overall, the values of the a and c lattice parameters after hydrothermal relithiation were close to those of pristine T-NCM111. In contrast, the a and c lattice parameters of HT-160 sample did not reach the values of the samples treated at higher temperatures (inset of FIG. 29B), likely indicating that Li.sup.+ had not reached its equilibrium position in the lattice due to kinetic limitations at low temperature.

[0139] The valence states of the transition metals are also sensitive to the concentration of Li.sup.+ in cathode materials, which were probed via XPS. The XPS spectra of Co 2p in FIG. 29 C for all the samples were analogous, with two main peaks located at 779.86 eV and 794.99 eV corresponding to Co 2p.sub.3/2 and Co 2p.sub.1/2, respectively. The absence of a satellite peak around 785 eV indicates that the valence state of Co in all the samples was 3+, which is reasonable given that only 10% of the Li.sup.+ was extracted from the D-NMC samples, where only Ni is expected to be oxidized due to its lower redox potential. After chemical dilithiation, the main peak at 854.28 eV related to Ni 2p.sub.3/2 of T-NCM111 shifted to a higher binding energy of 854.79 eV (D-NCM111), confirming this trend and indicating that Ni was oxidized to a higher valence state. Although the peak shifted back slightly after relithiation at 160° C., the binding energy was still higher than that of pristine T-NCM111, implying that some Ni.sup.3+ still existed in the cathode material. Notably, the peak of Ni 2p.sub.3/2 shifted back to the same position as that of the pristine T-NCM111 for the samples treated at 200° C. for 6 h and 220° C. for 2 h, suggesting that the compositional defects were completely repaired under these relithiation conditions.

[0140] To better understand the effect of hydrothermal temperature and time on the properties of the final cathode product after direct regeneration, the electrochemical cycling performance of the samples relithiated at different temperatures was evaluated in half cells at a rate of C/10 for 4 cycles followed by 50 cycles at a rate of C/3. It should be mentioned that the relithiated samples were annealed with an excess 5 mol % of Li.sub.2 CO.sub.3 at 850° C. for 4 h before the electrochemical test for a fair comparison with pristine T-NCM111. The chemical extraction of Li.sup.+ reduced the discharge capacity of the first cycle at the C/10 rate from 156.5 mAh/g to 146.1 mAh/g (T-NCM111 vs. D-NCM111). When the rate was increased to C/3, D-NCM111 showed a capacity of 133.9 mAh/g, and a capacity retention of 105.3 mAh/g (78.6%) was observed after 50 cycles. The samples that underwent hydrothermal relithiation at 160° C. followed by annealing improved the initial capacity to 154.3 mAh/g and 149.5 mAh/g at C/10 and C/3, respectively, suggesting that a large portion of Li deficiencies has been recovered. However, after 50 cycles the discharge capacity was reduced to 133.6 mAh/g, which corresponds to a capacity retention of 89.4%. When the relithiation temperature was increased to 200° C. and 220° C., the C/10 discharge capacities were found to be further increased to 156.1 and 156.3 mAh/g, respectively, which are nearly the same as that of pristine T-NCM111 (156.5 mAh/g). Notably, even at the C/3 rate, an initial capacity of 150.4 mAh/g and 150.6 mAh/g, respectively, was achieved, reaching the level of pristine T-NCM111 (150.5 mAh/g). At the end of the 50.sup.th cycle, the retained capacities were 140.2 and 140.3 mAh/g for the samples relithiated at 200° C. and 220° C., respectively, both of which were comparable to that of T-NCM111 (140.3 mAh/g). Given this performance, it was concluded that the Li deficiencies of D-NCM111 can only be fully refilled at temperatures higher than 200° C. in the limited time frames examined here. Considering the much longer time needed to achieve full relithiation at 200° C. (6 h) as compared to that of 220° C. (2 h), the optimal hydrothermal relithiation temperature and duration are 220° C. and 2 h.

Example 2: Improved Material Economics

[0141] The effect of the lithium-bearing solution composition used in hydrothermal relithiation was then explored. As described in above, 4 M LiOH was used as the relithiation solution. Considering the high cost of LiOH, a more-dilute LiOH solution would be preferable. In order to exclude the effect of pH, a solution consisting of a mixture of 0.1 M LiOH and 3.9 M KOH (to yield 4 M OH.sup.−) was used for relithiation.

[0142] The crystal structure of D-NCM111 which was relithiated with the mixed solution was characterized by XRD (FIG. 30A). The XRD patterns of D-NCM111, and T-NCM111 as well as that treated with 4 M KOH were included as reference points. There was no evidence of any new phases present, suggesting that the layered structure of pristine T-NCM111 was maintained for all the samples. When treated with 4 M KOH, it was noted that the (003) peak presented at a lower-than-expected angle, indicating that the Li deficiencies still existed in the bulk crystal structure. However, when the 0.1 M LiOH 3.9 M KOH solution was employed, the (003) peak returned to a position like that of pristine T-NCM111, which suggests that the structure was fully lithiated.

[0143] XPS was again performed to probe the valence state of Ni in the above samples (FIG. 30B). The Ni 2p spectra of all the samples showed two main peaks corresponding to Ni 2p.sub.3/2 and Ni 2p.sub.1/2, accompanied by a satellite peak (denoted as “Sat.”). The treatment with 4 M KOH did not change the peak position of the Ni 2p.sub.3/2 peak as compared to that in D-NCM111, suggesting that the valence state of Ni was still higher than that of pristine T-NCM111. This indicates that K.sup.+ may not have appreciably intercalated into the lattice vacancies of the layered structure. By comparison, the Ni 2p.sub.3/2 peak from D-NMC111 treated with the mixed solution shifted back to the same position as that for the pristine T-NCM111, indicating that Ni was reduced to 2+. In order to further confirm the exclusion of K.sup.+ in the mixed-solution case, XPS measurements in the K 2p binding energy region were carried out. Interestingly, two noticeable peaks related to K 2p.sub.3/2 and K 2p.sub.1/2 for the sample treated with 4 M KOH were seen (FIG. 30B), which implies the intercalation and possible exchange of K.sup.+ and Li.sup.+. On the contrary, the sample treated with the mixed solution showed a flat XPS spectrum in this binding energy region, suggesting that there was negligible K.sup.+ intercalation or Li.sup.+/K.sup.+ exchange.

[0144] The concentrations of Li.sup.+ and K.sup.+ in the cathode particles were further examined by ICP-MS measurement (FIG. 30C). Note that the original chemical oxidation of T-NCM111 extracted 10% of the Li.sup.+ (D-NCM111) from the pristine cathode particles. After hydrothermal treatment with the 4 M KOH solution, it was found that an additional 4% of Li.sup.+ in the particles were lost. Meanwhile, ˜4% of K.sup.+ were detected in the cathode particles, which supports the hypothesis of Li.sup.+/K.sup.+ exchange proposed above. When 0.1 M LiOH was introduced, the cation exchange was eliminated, and no K.sup.+ was detected in the relithiated sample. Moreover, the Li stoichiometry of the NCM111 was found to be 1.07 after treatment for 2 h, which is close to the value of the T-NCM111 (1.06). This result implies that the existence of a large amount of KOH in the relithiation solution does not affect the relithiation process if a small concentration of Li.sup.+ is present in the solution.

[0145] The electrochemical performance was then evaluated to understand the effect of the mixed solution on the properties of the regenerated sample. Prior to the electrochemical test, the samples were again annealed at 850° C. for 4 h with excess 5 mol % of Li.sub.2 CO.sub.3 after hydrothermal relithiation. As shown in FIG. 30D, the discharge capacity of the sample treated with 4 M KOH only produced a capacity of 134.3 mAh/g, which was even lower than that of D-NCM111 (146.1 mAh/g), indicating the Li.sup.+/K.sup.+ exchange has a detrimental effect on performance (FIG. 30C). After treatment with the mixed solution, the capacity was restored to 156.4 mAh/g, which is nearly the same as that of T-NCM111 (156.5 mAh/g). At a C/3 rate, the sample treated with 4 M KOH only delivered an initial capacity of 115.6 mAh/g, also less than that of D-NCM111 (133.9 mAh/g), which then further decayed to 92.4 mAh/g over 50 cycles. Conversely, after treatment with the mixed solution, the sample delivered an initial capacity of 150.4 mAh/g, which decreased to 140.1 mAh/g after 50 cycles. The obtained capacity retention (93.1%) was comparable to T-NCM111 (93.2%). The effective regeneration of D-NCM111 using the mixed solution to achieve the same level of performance as using 4 M LiOH suggests that the cathode recycling cost can be reduced by replacing a majority of Li in the relithiation solution, leading to a lower raw material cost.

[0146] It should be noted that besides the replacement of 4 M LiOH with a mixture of 0.1 M LiOH and 3.9 M KOH, the cost of hydrothermal relithiation can be also reduced by recycling the concentrated LiOH solution. To briefly evaluate the process effectiveness, a D-NMC111 sample was relithiated with a previously used 4 M LiOH solution at 220° C. for 2 h, annealed at 850° C., and subjected to electrochemical cycling for comparison. The initial discharge capacity sample was found to be 156.6 mAh/g and 150.5 mAh/g at C/10 and C/3, respectively, also retaining 92.8% of this capacity after 50 cycles, comparable to T-NCM111. The viability of continuous recycle/reuse of the spent LiOH during relithiation also represents a promising strategy in reducing the recycling cost based on the hydrothermal relithiation method.

[0147] As described above, after hydrothermal relithiation, a short annealing is required to remedy the remaining structural defects and achieve the desired electrochemical performance. However, during high-temperature annealing, a portion of Li in the crystal structure may be lost, leading to inferior capacity and cycling instability (FIGS. 31A and 31B). Generally, 5 mol % excess of Li is added to compensate for Li loss in this step. The effects of annealing temperatures on the composition and electrochemical performance of the final regenerated NCM111 were explored when Li.sub.2 CO.sub.3 and LiOH was used as the Li source. FIG. 32A shows the XPS spectra of samples annealed with Li.sub.2 CO.sub.3 at different temperatures. In the XPS spectra for Ni, the position of the main Ni 2p.sub.3/2 peak was not affected by the annealing temperature, and remained almost identical to that of the pristine T-NCM111 for all the samples. Nevertheless, the peak intensity relative to O 1s increased significantly with increased annealing temperature. The relatively low intensity of the samples annealed at low temperatures, such as 550° C. and 650° C., might be due to the undecomposed Li.sub.2 CO.sub.3, which reduced the signal from the underlying crystalline NCM111. This was supported by the O 1s XPS spectra, where the intensity of the peak associated with lattice oxygen almost disappeared for the sample annealed at 550° C. and 650° C. Additionally, the cumulative intensity of the peaks related to C—O and C═O was quite high, which suggests that a large portion of Li.sub.2 CO.sub.3 was still present on the surface. When the annealing temperature was increased to 750° C., however, the intensity of the lattice oxygen peak increased and the intensity of the peak associated with Li.sub.2 CO.sub.3 decreased, indicating the additional Li was incorporated into the crystal structure of NCM111. This incorporation was further increased when the annealing temperature was increased to 850° C.

[0148] When LiOH was used as the Li source instead, the relative intensity ratio of Ni 2p to O 1s for the samples annealed at 550° C. and 650° C. was not as low as that of the sample annealed with Li.sub.2 CO.sub.3 under the same conditions (FIG. 32B). This result indicates that some Li.sup.+ was incorporated into the material at lower temperatures, likely due to the lower melting point of LiOH (471° C.) than that of Li.sub.2 CO.sub.3 (723° C.). The typical (003) peak of layered cathode material starts to appear at ˜575° C. when LiOH was used as the Li precursor, whereas it emerged at ˜775° C. when Li.sub.2 CO.sub.3 was used as Li precursor. In the O 1s spectra, the signals of C—O and C═O associated with Li.sub.2 CO.sub.3 are likely due to the reaction of CO.sub.2 and surface LiOH when the sample was exposed to air. When the annealing temperature was increased to 750° C., the intensity of the peak associated with the lattice oxygen was increased while that of Li.sub.2 CO.sub.3 decreased. The relative intensity of the peaks did not change when the relithiated sample was annealed at 850° C., similar to that of pristine T-NCM111.

[0149] The electrochemical performance of the samples annealed at different conditions were then compared with the same protocol described earlier. When the relithiated sample was annealed at 550° C., the capacity of the first charging cycle at C/10 was 231.4 mAh/g, far beyond that of pristine T-NCM111 (182.8 mAh/g). The discharge capacity was only 154.6 mAh/g, corresponding to a Columbic efficiency (CE) of 66.8%, which again is likely from the irreversible Li.sup.+ capacity introduced by residual Li.sub.2 CO.sub.3. However, after washing with water, the charge capacity dropped to 185.4 mAh/g, indicating that the additional capacity originated from the residual Li.sub.2 CO.sub.3. When the annealing temperature was increased to 650° C., the charge and discharge capacities dropped to 212.2 mAh/g and 152.3 mAh/g, respectively, corresponding to a CE of 71.7%, indicating that the inactive Li.sub.2 CO.sub.3 remained in the cathode material. Even when the temperature was further increased to 750° C., the charge capacity was still as high as 206.4 mAh/g but the discharge capacity was only 149.8 mAh/g. When the annealing temperature was increased to 850° C., the charge capacity dropped to 182.2 mAh/g, and the discharge capacity increased to 156.8 mAh/g, achieving a CE of 86.1%, which is even higher than that of pristine T-NCM111 (85.6%).

[0150] The cycling stability was also compared for these samples annealed with Li.sub.2 CO.sub.3 at various temperatures (FIG. 32C) and summarized in Table 10, which lists capacity retention (C/3 rate) of relithiated NCM111 after annealing with Li.sub.2 CO.sub.3 and LiOH.

TABLE-US-00010 TABLE 10 Discharge capacity (mAh/g) Retention Samples 1.sup.st cycle 50.sup.th cycle (%) HT-Li.sub.2CO.sub.3-550° C. 145.3 127.4 87.7% HT-Li.sub.2CO.sub.3-650° C. 146.4 129.4 88.4% HT-Li.sub.2CO.sub.3-750° C. 148.6 134.4 90.4% HT-Li.sub.2CO.sub.3-850° C. 150.6 140.3 93.2% HT-LiOH-550° C. 145.4 124.6 85.7% HT-LiOH-650° C. 148.2 128.3 86.5% HT-LiOH-750° C. 150.5 140.4 93.3% HT-LiOH-850° C. 150.6 140.6 93.4% D-NCM111 133.9 105.3 78.6% T-NCM111 150.8 140.5 93.2%

[0151] The samples annealed at 550° C. and 650° C. showed a similar capacity degradation trend where the capacity retention was improved from 88.2% to 90.4%. Increasing the annealing temperature to 850° C. further increased the initial capacity to 150.4 mAh/g and a capacity of 140.7 mAh/g could be achieved after 50 cycles, (a capacity retention of 93.1%), which is comparable to pristine T-NCM111. Therefore, it was concluded that 850° C. is the minimum temperature required for fully restoring the electrochemical performance of D-NCM111 when Li.sub.2 CO.sub.3 is used as the Li source.

[0152] The optimal annealing time at 850° C. was evaluated by testing the cycling stability of relithiated samples annealed at 850° C. for 1 h, 2 h, 4 h and 6 h. After annealing for 1 h, the sample exhibited a capacity of 149.1 mAh/g, which decreased to 130.2 mAh/g over 50 cycles. With a further increase of annealing time to 2 h, the initial capacity was not affected dramatically (150.2 mAh/g). Nevertheless, the capacity after 50 cycles was improved to 135.6 mAh/g, which increased further to 140.3 mAh/g when the sample was annealed for 4 h. Annealing for 6 h did not improve the capacity or cycling stability, leading to the conclusion that a duration of 4 h is sufficient to attain the desirable electrochemical performance.

[0153] Due to the lower melting point (471° C.), LiOH was proposed as an alternative to Li.sub.2 CO.sub.3. Similarly, different annealing temperatures were explored to identify the optimal temperature for fully recovering the electrochemical properties of the regenerated cathode. When the annealing temperature was 550° C., the charge capacity reached 180.3 mAh/g, which was significantly lower than that of the sample annealed under the same condition with Li.sub.2 CO.sub.3 as the Li source (231.4 mAh/g). The reversible capacity was 154.4 mAh/g, corresponding to a CE of 85.6%. This indicates that Li.sup.+ from LiOH can be more easily incorporated into the layered crystal structure at a low temperature compared to Li.sub.2 CO.sub.3. The charge and discharge capacities did not change when the temperature was increased to 650° C. The increase of temperature to 750° C. improved the reversible capacity to 156.7 mAh/g, almost identical to that of the sample annealed at 850° C. (156.8 mAh/g).

[0154] The cycling stability of all the samples annealed with LiOH is shown in FIG. 32D and the results are summarized in Table 10. The samples annealed at 550 and 650° C. delivered initial capacities of 145.4 and 148.2 mAh/g, respectively at a C/3 rate. After 50 cycles, capacities of 124.6 and 128.3 mAh/g were measured, corresponding to capacity retentions of 85.6% and 86.5%, respectively. When the temperature was increased to 750° C., the capacity increased to 150.5 mAh/g and 93.2% of this capacity was retained after 50 cycles. Notably, the capacity and retention were nearly the same as those of the sample annealed at 850° C. The 750° C. annealing temperature produced a similar performance to that of the pristine T-NCM111, 100° C. less than that required when annealing with Li.sub.2 CO.sub.3. It appears that although a relatively low temperature can trigger the incorporation of LiOH, the temperature must still be high enough to obtain a phase that can withstand long-term charge/discharge cycling. Nevertheless, the optimal, annealing temperature could be decreased from 850° C. to 750° C. when Li.sub.2 CO.sub.3 was replaced by LiOH as a Li source.

[0155] The T-NCM and R-NCM electrodes after 50 cycles were characterized by XPS and XRD to probe the difference in structural durability. The overall XPS survey spectra of cycled T-NCM and R-NCM (FIG. 33A) showed additional signals of F and P compared with that of the pristine T-NCM, which originate from the formed cathode electrolyte interphase (CEI) on the cycled cathode particles after cycling. The high-resolution of Ni 2p XPS spectra (FIG. 33B) displayed a similar right shift for both the cycled T-NCM and R-NCM electrode, indicating a higher valence state of Ni, which is resulted from the loss of the active Li.sup.+ after cycling. The XRD diffraction peaks in FIG. 33C matched well with the layered structure and no major phase impurity was detected for the cycled materials. Overall, no appreciable difference in terms of structure and composition was observed between R-NCM and T-NCM after cycling.

[0156] Additionally, the direct regeneration was scaled up from 1 g to 10 g of cathode per batch to demonstrate the viability of the process. Referring to Table 11, the composition of the product (R-NCM-10 g) was analyzed by ICP, where the Li content was found to be similar to that of the control sample T-NCM111, showing no difference from the sample regenerated at a smaller scale (R-NCM-lg). The composition of NCM111 regenerated at different scales and the control samples (D-NCM111 and T-NCM111).

TABLE-US-00011 TABLE 11 Sample Composition D-NCM111 Li.sub.0.9Ni.sub.0.33Co.sub.0.33Mn.sub.0.33O.sub.2 R-NCM111-1 g Li.sub.1.05Ni.sub.0.33Co.sub.0.33Mn.sub.0.33O.sub.2 R-NCM111-10 g Li.sub.1.07Ni.sub.0.33Co.sub.0.33Mn.sub.0.33O.sub.2 T-NCM111 Li.sub.1.06Ni.sub.0.33Co.sub.0.33Mn.sub.0.33O.sub.2

[0157] The R-NCM-10 g was further characterized by XRD and XPS. Compared with the starting D-NCM, a right shift of the (003) diffraction peak (FIG. 34A) and a left shift of the Ni 2p.sub.3/2 peak (FIG. 34B) similar to that observed for R-NCM-lg sample was also present for the R-NCM-10 g, suggesting that the recovery of the structure and composition is not affected by the increase of the material loading in the relithiation reactor. Moreover, both the capacity and its retention of R-NCM-10 g were found to be identical to the pristine level (T-NCM111), indicating the scalability of this process for industrial application.

[0158] The inventive scheme provides for the regeneration of chemical composition and microstructure in degraded Li-ion battery cathodes using a novel particle-to-particle approach that is economical, energy efficient and more environmentally friendly than existing methods. Hydrothermal treatment with a short annealing step provides the regenerated cathode materials with high capacity, long cycling stability and high rate performance comparable to that of pristine materials.