Carbothermal regeneration of LiFePO4/C cathode material by recycling the precursors from out-of-line LiFePO4 cells Authors: A.R. Abbasi, S.M. Masoudpanah Published: 2026 ABSTRACT SummaryIn this work, the LiFePO4/C (LFP/C) cathode material was regenerated by carbothermal processing using Li2CO3, Li3PO4, and FePO4 precursors recycled from out-of-line (substandard) LFP cells. The separated black mass was converted into the various metal precursors, Li2CO3, Li3PO4, and FePO4. The LFP/C powders were then regenerated by the carbothermal reaction between the metal precursors in various molar ratios of lithium precursor to FePO4 (1:1, 1.05:1, and 1.10:1) and sucrose as a carbon source. The LFP/C powders regenerated from Li2CO3 and FePO4 precursors at a molar ratio of 1:1 exhibited long-term cycling performance, with 90% capacity retention after 2,000 charge/discharge cycles at a current density of 20C. Introduction The demand for lithium-ion technology in electric vehicles (EVs) and energy storage systems has increased significantly over the past decade due to the high specific energy density and long cycling life.1,2 An essential component of lithium-ion batteries (LIBs) is the cathode electrode, which is based on various materials, such as LiCoO2, LiNi0.6Co0.2Mn0.2O2, and LiFePO4 (LFP).3,4 The olivine-structured LFP material is the most common cathode for EVs, due to its low cost, relatively low toxicity, reliability, and long cycling life.5,6,7 There is also rapid growth in spent LIBs, which must be recycled due to their valuable materials, such as lithium, nickel, cobalt, manganese, aluminum, copper, and graphite, and to address environmental issues.8,9,10,11,12 Sorting, stabilization, dismantling, separation, and recovery are the five key steps in recovering valuable materials from LIBs.13,14 The recovery process is traditionally based on pyrometallurgical and hydrometallurgical treatments.15,16,17,18 Furthermore, a new approach to direct recycling is proposed in which the cathode material is regenerated from the recovered precursors without separating the individual elements.19,20 The simplicity, flexibility, low cost, and high time- and energy efficiency are the advantages of the direct regeneration approach.21 For example, Huang et al.22 used a direct regeneration approach to repair spent LFP materials using residual conductive carbon and polyvinylidene fluoride (PVDF) as reductive agents. The regenerated LFP/C cathode materials showed a comparable specific capacity of 113.3 mAh g−1 at a current rate of 0.1C. Wu et al.23 selectively extracted lithium from spent LFP cells and regenerated LFP powders from FePO4 and Li2CO3 using a mechanochemical method. Among various synthesis methods, the carbothermal method has attracted significant attention for the regeneration of LFP powders. The carbothermal reaction is a high-temperature process in which a carbon source is used for reducing the metallic oxides.24,25 The carbothermal treatment for synthesizing the LFP/C material is finely tuned not only to reduce iron but also to simultaneously create a highly desirable carbon coating on the LFP particles, all in one high-temperature step. The carbothermal reaction for preparation the LFP/C using an inexpensive and chemically stable Fe2O3 precursor is as follows26:(Equation 1)Li2CO3+Fe2O3+2NH4H2PO4+0.5C→2LiFePO4+2NH3+3H2O+1.5CO2. The carbon sources, such as glucose, sucrose, citric acid, and phenolic resins are oxidized into CO or CO2, reducing Fe3+ into Fe2+ during high-temperature calcination which facilitates the crystallization of the LFP phase.27 The application of the low-cost Fe3+ precursors and deposition of a conductive carbon network on the LFP particles are major advantages of the carbothermal synthesis method.28 In the carbothermal regeneration, the separated black mass is first converted to lithium and iron salts by acid leaching.29 Then the stoichiometric ratios of the proper precursors are mixed with sucrose or glucose as a carbon source. The mixture is calcined at high temperatures under an inert atmosphere, such as N2 or Ar, to crystallize the LFP/C particles.30,31,32 The type and amount of metal precursors play a significant role in determining the electrochemical properties of the carbothermally regenerated LFP/C powders, as they control the crystallinity and microstructure. Herein, the different precursors, such as Li2CO3, Li3PO4, and FePO4, were obtained from the separated black mass via a chemical process. The well-characterized precursors were used in the carbothermal method to regenerate the LFP/C particles. The procedure was previously reported for recycling the spent LiFePO4 cells, but there was no comparison of the type of lithium precursors and ratio of metal precursors on the electrochemical properties of the LFP/C powders. In other words, our work showed a practical strategy for recycling the spent LiFePO4 cells to regenerate the LFP/C cathode material. Results and discussion Recycling of precursors Figure 1 shows the X-ray diffraction (XRD) patterns of the recovered Li3PO4, Li2CO3, and FePO4 precursors. There is a good match between the reflections and PDF2 cards, confirming the high phase purity of precursors. The elemental purity of FePO4, Li3PO4, and Li2CO3 precursors was mentioned in Table S1 of the supplemental information. Furthermore, their high diffraction peak intensities indicate high crystallinity, suggesting that the proper selection of precipitation temperature and concentration was achieved. The precursors were crystallized by a selective leaching of lithium from the LiFePO4 black mass by H2SO4 and H2O2, and precipitating the FePO4 as follows33:(Equation 2)LiFePO4+H2SO4+H2O2→2FePO4↓+Li2SO4+2H2O.Figure 1XRD patterns of the recovered precursors(A) Li3PO4, (B) Li2CO3, and (C) FePO4. XRD patterns of the recovered precursors (A) Li3PO4, (B) Li2CO3, and (C) FePO4. The Li was precipitated as phosphate and carbonate from leachate by adding a proper amount of Na3PO4 and Na2CO3 agents as follows23:(Equation 3)Li2SO4+2Na3PO4→2LiPO4↓+3Na2SO4(Equation 4)Li2SO4+Na2CO3+Li2CO3↓+Na2SO4 To monitor the reaction progress, samples were collected at various stages (Scheme 1), and the concentrations of Fe and Li elements were measured by inductively coupled plasma (ICP) (Table 1). The analysis after iron phosphate precipitation revealed that all the Fe3+ ions had been selectively precipitated as FePO4, while all the Li+ ions remained in the solution. The hot conditions for lithium precipitation suggest that sodium will remain dissolved due to its higher ionization potential and greater solubility than lithium. To further ensure complete sodium removal, the precipitated lithium precursors are washed with boiling water after filtration (Table S1 of the supplemental information).Scheme 1Recycling and regeneration of the LFP/C powders from the out-of-line (substandard) LFP cellsTable 1ICP results of the filtered solutionsParameterFeLiConcentration after Fe3+ precipitation (ppm)<0.12160.0Concentration after Li+ precipitation (ppm)<0.169.3Recovery percentage (%)99.996.8 Recycling and regeneration of the LFP/C powders from the out-of-line (substandard) LFP cells Table 1ICP results of the filtered solutionsParameterFeLiConcentration after Fe3+ precipitation (ppm)<0.12160.0Concentration after Li+ precipitation (ppm)<0.169.3Recovery percentage (%)99.996.8 ICP results of the filtered solutions Figure 2 presents scanning electron microscopic (SEM) images of the Li3PO4, Li2CO3, and FePO4 precursors. The recycled Li3PO4 precursor exhibits excellent, consistent particle size uniformity and a quasi-spherical shape (Figures 2A1 and A2). In contrast, the Li2CO3 precursor has a strong agglomeration of various particles (Figures 2B1 and B2). Analysis of the FePO4 precipitates (Figures 2C1 and C2) revealed tightly agglomerated small nanoparticles, which can strongly promote the carbothermal reaction rate.8Figure 2SEM images of the recycled precursors(A1) Li3PO4, (A2) magnified view of A1, (B1 and B2) Li2CO3, and (C1) FePO4 and (C2) magnified view of C1. Scale bar of (A1) 5 μm, (A2) 2 μm, (B1) 10 μm, (B2) 5 μm, (C1) 5 μm, and (C2) 2 μm. SEM images of the recycled precursors (A1) Li3PO4, (A2) magnified view of A1, (B1 and B2) Li2CO3, and (C1) FePO4 and (C2) magnified view of C1. Scale bar of (A1) 5 μm, (A2) 2 μm, (B1) 10 μm, (B2) 5 μm, (C1) 5 μm, and (C2) 2 μm. Carbothermal regeneration Figure 3 shows the XRD patterns of the regenerated LFP/C powders from the various amounts of the Li3PO4 and Li2CO3 precursors. The reflections are well matched to the olivine crystal structure of LFP phase, with a space group of Pnma (PDF#96-110-1112), for the lithium phosphate precursor at all molar ratios. However, the Li2CO3 precursor leads to the impurity Fe(0) phase (PDF2#00-006-0696) at a 1:1 ratio (1C) (Table 2). The LFP phase was crystallized by solid-state reaction between the Li2CO3 and Li3PO4 precursors and FePO4 as follows22,29:(Equation 5)Li3PO4+3FePO4→3LiFePO4+PO4,(Equation 6)Li2CO3+2FePO4+2LiFePO4+CO2+0.5O2.In fact, the Fe3+ species in FePO4 were first reduced to Fe2+ species during the carbothermal treatment using sucrose as a carbon source under Ar/H2 atmosphere.34 Simultaneously, the Li precursors are thermally decomposed and react with the Fe2+ and PO43− species to crystallize the LFP phase. The formation of the impure Fe(0) phase in the 1C sample can be attributed to the high reductive atmosphere generated during the carbothermal treatment. However, the further liberation of CO2 decreased the reduction capability of atmosphere at the higher ratio of Li2CO3 precursor. The absence of Fe(0) in the P samples is related to the role of metal precursor on the decomposition rate of carbon source.35,36 The weak interaction of carbonate precursor is a responsible for the development of the reductive atmosphere.35 Furthermore, the molar ratio of lithium precursors plays a crucial role in determining the crystal arrangement by controlling the diffusion pathway for Li+ ions during crystallization of the olivine-structured LFP phase.37 The crystal quality of the LFP phase was usually evaluated by the ratio of various diffraction peaks, (101)/(311), (111)/(311), and (020)/(311), as a criterion of the crystal defects.38 The well-crystalline LFP material shows a higher (111)/(311) ratio because the crystal defects, such as antisite, interstitial, and vacancy defects, increase the intensity of (020) peaks.39 The 1.1P and 1.1C samples exhibit a higher ratio of (111)/(311) peaks, indicating higher crystallinity of the LFP phase due to its facile crystallization at higher Li precursor concentrations. Table 3 compares the lattice parameters and unit cell volume of the 1.1P and 1.1C samples with those of the defect-free LFP material.40 The larger unit cell volume and longer b-axis of the 1.1C sample are due to the lower decomposition temperature of Li2CO3 precursor, promoting the crystal arrangement of the LFP phase.41 Furthermore, the lower lattice microstrains and better refinement parameters of the 1.1C sample confirm its higher crystallinity caused by the lower crystal defects.Figure 3XRD patterns of the regenerated LFP/C material(A) Li3PO4 and (B) Li2CO3 precursors (: Fe).Table 2Codes of the regenerated LFP/C powders versus lithium source and molar ratio of Li:FePO4Li sourceLi:FePO4 molar ratio1:11.05:11.1:1Li3PO41P1.05P1.1PLi2PO41C1.05C1.1CTable 3Comparison of lattice parameters, unit cell volume, lattice strain, and refinement parametersSamplesa (Å)b (Å)c (Å)V (Å3)ε (%)Rwp (%)σ2Ref.10.33286.00434.6977291.4510–––1.1P10.33036.01224.6971291.72660.005191.781.1C10.34466.01534.6997292.44290.010231.65 XRD patterns of the regenerated LFP/C material (A) Li3PO4 and (B) Li2CO3 precursors (: Fe). Table 2Codes of the regenerated LFP/C powders versus lithium source and molar ratio of Li:FePO4Li sourceLi:FePO4 molar ratio1:11.05:11.1:1Li3PO41P1.05P1.1PLi2PO41C1.05C1.1C Codes of the regenerated LFP/C powders versus lithium source and molar ratio of Li:FePO4 Table 3Comparison of lattice parameters, unit cell volume, lattice strain, and refinement parametersSamplesa (Å)b (Å)c (Å)V (Å3)ε (%)Rwp (%)σ2Ref.10.33286.00434.6977291.4510–––1.1P10.33036.01224.6971291.72660.005191.781.1C10.34466.01534.6997292.44290.010231.65 Comparison of lattice parameters, unit cell volume, lattice strain, and refinement parameters Figures 4 and 5 show the SEM micrographs and elemental EDS map of the samples. The powders consist of quasi-spherical LFP particles that are tightly agglomerated into large particles. The LFP powders prepared by the Li3PO4 precursor are composed of the large LFP particles decorated by small particles. Furthermore, the morphology of the Li2CO3 precursor disappears during the crystallization of the LFP/C particles, especially at higher ratios, indicating its thermal decomposition. However, the morphology of the LFP/C is similar to that of the FePO4 precursor in the 1P sample. The 1.1C sample (Figure 5) exhibits a porous microstructure with pores (∼50 nm) between the LFP particles, resulting from the release of gaseous products during carbothermal treatment. The uniform dispersion of Fe, P, O, and C elements confirms the composition homogeneity of powders.Figure 4SEM images and EDS map of regenerated LFP/C materials(A1) 1P, (A2) magnified view of a1, (B1) 1.05P, (B2) magnified view of b1, (C1) 1.1P, (C2) magnified view of A1, and (D) elemental EDS maps of the 1.1P sample. Scale bars of (A1) 5 μm, (A2) 2 μm, (B1) 5 μm, (B2) 2 μm, (C1) 5 μm, and (C2) 2 μm.Figure 5SEM images of and EDS map of regenerated LFP/C materials(A1 and A2) 1C, (B1) 1.05C, (B2) magnified view of B1, (C1) 1.1C, (C2) magnified view of C1, and (D) elemental EDS map of the 1.1C sample. Scale bars of (A1) 5 μm, (A2) 2 μm, (B1) 5 μm, (B2) 2 μm, (C1) 5 μm, and (C2) 2 μm. SEM images and EDS map of regenerated LFP/C materials (A1) 1P, (A2) magnified view of a1, (B1) 1.05P, (B2) magnified view of b1, (C1) 1.1P, (C2) magnified view of A1, and (D) elemental EDS maps of the 1.1P sample. Scale bars of (A1) 5 μm, (A2) 2 μm, (B1) 5 μm, (B2) 2 μm, (C1) 5 μm, and (C2) 2 μm. SEM images of and EDS map of regenerated LFP/C materials (A1 and A2) 1C, (B1) 1.05C, (B2) magnified view of B1, (C1) 1.1C, (C2) magnified view of C1, and (D) elemental EDS map of the 1.1C sample. Scale bars of (A1) 5 μm, (A2) 2 μm, (B1) 5 μm, (B2) 2 μm, (C1) 5 μm, and (C2) 2 μm. Raman spectroscopy was performed to evaluate the degree of graphitization of the carbon layer on the LFP particles. The Raman spectra can be deconvoluted into four peaks using the Lorentzian function. The D1 and G bands are associated with the disordered (A1g symmetry) and ordered (E2g symmetry) graphite carbon layers, respectively. The amorphous carbon moieties such as organic molecules and functional groups and stretching vibrations of C=C/C–C result in the D3 and D4 bands, respectively. There are G peaks at 1,579 and 1,577 cm−1 and the D1 peaks at 1,329 and 1,324 cm−1 for the 1.1P and 1.1C samples, respectively (Figure 6). The calculated IG/ID1 ratios are 1.31 and 1.25 for the 1.1C and 1.1P samples, respectively, indicating a higher degree of structural order in the carbon coating of the 1.1C sample. The carbon layer was deposited on the LFP particles via the thermal decomposition of sucrose during calcination. The lower agglomeration of LFP particles in the carbonate-assisted samples promotes uniform coatings, as the particle surface acts as a catalyst for carbon deposition.42 The enhanced degree of graphitization suggests improved electrical conductivity, positively influencing the material’s electrochemical performance. It was worth noting that the 1.1C and 1.1P samples have similar carbon contents of 7.1 and 8.5 wt. %, respectively, according to the carbon, hydrogen, nitrogen, and sulfur (CHNS) analysis.Figure 6Raman spectra(A) 1.1P and (B) 1.1C samples. Raman spectra (A) 1.1P and (B) 1.1C samples. Figures 7A and 7B show the rate capability and cycling performance of the powders prepared using the Li2CO3 and Li3PO4 precursors. The specific capacity increased with the increase in the amount of Li precursor, irrespective of the precursor type. Furthermore, the carbonate-assisted LFP/C powders exhibit a higher specific capacity than the Li3PO4 precursor at all current rates. The higher specific capacity of the LFP/C powders prepared using the Li2CO3 precursor can be attributed to their higher crystallinity, which facilitates the Li+ de/intercalation process. The specific capacity of the 1.1P sample decreases from 93 to 18 mAh g−1 with an increase in current rate from 0.1C to 20C, corresponding to a rate capability of 22%. However, the 1.1C sample shows specific capacities of 135, 127, 119, 93, 73, 58, 49, and 39 mAh g−1 for 0.1C, 0.2C, 0.5C, 1C, 2C, 5C, 10C, and 20C, respectively, indicating a rate capability of 28%. The higher rate capability of the 1.1C sample can be attributed to its LFP/C nanoparticles and higher degree of graphitization of the carbon layer. Furthermore, the specific capacity was reversibly returned to 135 mAh g−1 by decreasing the current rate to 0.1C. The galvanostatic charge and discharge (GCD) curves of the 1.1C sample at various current rates are shown in Figure 7C. The specific capacity of LFP/C material is strongly related to the many factors, such as crystallinity, particle size and morphology, carbon coating, electronic conductivity, ion diffusivity, electrode formulation, and electrolyte compatibility.28,43,44 The smaller particles with uniform size distribution shorten lithium-ion diffusion paths and avoided the inactive dead zone, improving rate capability and accessible capacity.45 Poor electronic conductivity limits active material utilization, leading to the lower specific capacity.28 The crystal defects, such as oxygen vacancies, antisite defects (Fe on Li sites), and Fe3+ impurity can block the fast diffusion paths of <010> directions, leading to the capacity fade.45 Excess carbon reduces gravimetric capacity by adding inactive mass.46 The difference in the electrochemical properties of our synthesized samples can be attributed to their crystallinity, particle size, and quality of carbon layer, as characterized by XRD, SEM, and Raman spectroscopy results.Figure 7Rate capability of LFP/C powders prepared by(A) Li3PO4 and (B) Li2CO3, and (C) charge/discharge curves of the 1.1C sample at various current rates, and cycling performance of LFP/C powders prepared by (D) Li3PO4 and (E) Li2CO3 at 20C, and (F) charge/discharge curves of the 1.1C sample at various cycles at 20C. Rate capability of LFP/C powders prepared by (A) Li3PO4 and (B) Li2CO3, and (C) charge/discharge curves of the 1.1C sample at various current rates, and cycling performance of LFP/C powders prepared by (D) Li3PO4 and (E) Li2CO3 at 20C, and (F) charge/discharge curves of the 1.1C sample at various cycles at 20C. Charge/discharge plateaus are observed at approximately 3.4 V vs. Li/Li+, with shrinking the plateau as the current rate increases, which is attributed to electrode polarization. The voltage plateaus are due to the Li de/insertion from the olivine crystal structure of the LFP material, accompanied by the following phase transition47,48:(Equation 7)LiFePO4↔FePO4+Li++e−. Figures 7D and 7E show the long-term cycling performance of the 1.1P and 1.1C samples at a current rate of 20C for 2,000 charge/discharge cycles. The phosphate-assisted LFP/C powders exhibit a relatively similar capacity retention of ∼72%, irrespective of the amount of Li3PO4 precursor. However, the values of capacity retention are 90%, 82%, and 55% for the 1C, 1.05C, and 1.1C samples, respectively. The improved cycling performance of the 1C and 1.05C samples can be attributed to their smaller particle size and higher crystallinity, which provide a greater number of electroactive sites. Figure 7F presents the charge/discharge curves of the 1.1C sample at various cycles. The clear plateaus at high current rates confirm the high stability of structural and microstructural features of the regenerated LFP/C powders. Furthermore, the coulombic efficiency remains stable at 95.7% for various cycles, indicating the highly reversible redox reaction for storage Li ions. The reversibility of the electrochemical oxidation and reduction reactions is evaluated using a cyclic voltammetry (CV) test. Figures 8A and 8B compare the CV curves of the P and C samples at a scan rate of 0.2 mV s−1. The anodic peak at ∼3.54 V vs. Li/Li+ and cathodic peak at ∼3.33 V vs. Li/Li+ are due to the Li+ ion deintercalation and intercalation, respectively.49,50 With the increase in lithium contents, the redox peaks became sharp and intense, indicating the presence of a crystal structure with high crystallinity. Furthermore, the potential differences between the anodic and cathodic peaks are 208 and 218 mV at a scan rate of 0.2 mV s−1 for the 1.1C and 1.1P samples, respectively (Figure 8C), indicating lower electrode polarization due to the higher quality of the carbon layer on the LFP particles. Also, the 1.1C electrode shows a higher current response at scan rates of 0.2 and 0.5 mV s−1, confirming its higher specific capacity. The CV curves show a stable shape at high scan rates, suggesting high reversibility (Figure 8D).Figure 8Electrochemical propertiesCV curves of (A) P and (B) C samples at 0.2 mV s−1 and (C) comparison of CV curves of the 1.1C and 1.1P samples at 0.2 mV s−1 and (D) the CV curves of the 1.1C samples at scan rates of 0.2 and 0.5 mV s−1. Electrochemical properties CV curves of (A) P and (B) C samples at 0.2 mV s−1 and (C) comparison of CV curves of the 1.1C and 1.1P samples at 0.2 mV s−1 and (D) the CV curves of the 1.1C samples at scan rates of 0.2 and 0.5 mV s−1. To compare electrode kinetics, electrochemical impedance spectroscopy (EIS) was performed. Figure 9 shows the Nyquist plots and the equivalent circuit. The Nyquist plots are modeled using an equivalent circuit comprising the ohmic resistance (Rs), charge transfer resistance (Rct), constant phase element (CPE), and Warburg impedance (Zw) elements. The charge transfer resistance results in a semicircle in the mid-frequency range. The Warburg impedance, arising from Li+ ion diffusion within the LFP structure, appears as an inclined line in the low-frequency region.51,52 The fitted resistance values are presented in Table 4. The carbonate-assisted LFP/C samples exhibit lower Rct values than those of the Li3PO4 precursor due to their higher crystallinity, which induces higher electrical conductivity. Furthermore, the Rct values decrease with increasing the amount of lithium precursor. The carbon layer on the LFP particles in the 1.1C sample exhibits greater graphitization, thereby promoting electrical conductivity. The diffusion coefficient (D) of Li+ ion can be calculated by the Warburg factor (σ) as the linear slope of Z′ vs. ω−1/2 plot as follows53,54:(Equation 8)D=2R2T2A2n4F4C2σ2,which R and F are constant parameters, T is the temperature, A is the effective area of the electrode, Li+ concentration is shown by C, and n is the number of electrons exchanged. The C electrodes show a lower σ value than the P sample at all lithium ratios. Furthermore, the σ value decreases when a further lithium precursor is used. Accordingly, the C samples have the higher diffusion coefficients than those of the P samples (Table 5). The higher diffusion coefficient is attributed to the lower agglomeration of LFP particles in the C powders, which provides facile accessibility of the electrolyte and a higher number of electroactive sites.57 The larger DLi+ value of the 1.1C sample can be attributed to its smaller LFP nanoparticles. In LFP nanoparticles, the diffusion paths are so short that they are less likely to contain a blocking antisite defect (Fe on Li channels), allowing for much faster, intrinsic mobility (DLi+∼10−11 cm2 s−1) than the micron-sized LFP particles (DLi+∼10−14 cm−2 s−1).45,58Figure 9Electrochemical impedance spectra(A and B) Nyquist plots and equivalent circuit and (C and D) Z′ vs. ω−1/2 plot.Table 4Resistance values and diffusion coefficient of the various samplesSampleRs (Ω)Rct (Ω)σ (Ω s−0.5)D (cm2 s−1)1P10.375.67213.50.61 × 10−141.05P12.4868.69130.41.63 × 10−141.1P14.223.2247.312.3 × 10−141C6.12103.42138.31.45 × 10−141.05C14.4155.34112.42.2 × 10−141.1C9.2116.48.48386.3 × 10−14Table 5Electrochemical performance of the regenerated LFP/C powders via the carbothermal methodPrecursorSynthesis conditionDischarge capacity (mAh g−1) @0.1CCapacity retention (%)ReferencesFePO4.2H2O (recycled)LiOH.H2O (CommercialSucrose 6%wt750 °C - 8 h14398 (25 cycle, 1C)55FePO4.2H2O (recycled)Li2CO3 (Commercial)Glucose 20%wt350 °C - 4 h650 °C - 9 hLi:Fe1.05:115995.4 (500 cycle, 5C)56FePO4 (recycled)Li2CO3 (recycled)Glucose 12%wt300 °C - 4 h700 °C - 10 hLi:Fe1.05:116294 (200 cycle, 5C)81.1C750 °C - 6 hLi:Fe1.1:113555 (2,000 cycle, 20C)This work1.1P750 °C - 6 hLi:Fe1.1:19567 (2000 cycle, 20C) Electrochemical impedance spectra (A and B) Nyquist plots and equivalent circuit and (C and D) Z′ vs. ω−1/2 plot. Table 4Resistance values and diffusion coefficient of the various samplesSampleRs (Ω)Rct (Ω)σ (Ω s−0.5)D (cm2 s−1)1P10.375.67213.50.61 × 10−141.05P12.4868.69130.41.63 × 10−141.1P14.223.2247.312.3 × 10−141C6.12103.42138.31.45 × 10−141.05C14.4155.34112.42.2 × 10−141.1C9.2116.48.48386.3 × 10−14 Resistance values and diffusion coefficient of the various samples Table 5Electrochemical performance of the regenerated LFP/C powders via the carbothermal methodPrecursorSynthesis conditionDischarge capacity (mAh g−1) @0.1CCapacity retention (%)ReferencesFePO4.2H2O (recycled)LiOH.H2O (CommercialSucrose 6%wt750 °C - 8 h14398 (25 cycle, 1C)55FePO4.2H2O (recycled)Li2CO3 (Commercial)Glucose 20%wt350 °C - 4 h650 °C - 9 hLi:Fe1.05:115995.4 (500 cycle, 5C)56FePO4 (recycled)Li2CO3 (recycled)Glucose 12%wt300 °C - 4 h700 °C - 10 hLi:Fe1.05:116294 (200 cycle, 5C)81.1C750 °C - 6 hLi:Fe1.1:113555 (2,000 cycle, 20C)This work1.1P750 °C - 6 hLi:Fe1.1:19567 (2000 cycle, 20C) Electrochemical performance of the regenerated LFP/C powders via the carbothermal method The electrochemical performance of the regenerated LFP cathode materials, obtained via the carbothermal method, is compared with other works in Table 5. The carbothermal-regenerated LFP/C powders prepared from the Li2CO3 precursor exhibited comparable electrochemical properties owing to their high crystallinity. Furthermore, their smaller particle size benefits the quality of the carbon layer on the LFP particles. Higher amounts of lithium precursors improve electrochemical performance by controlling the crystal structure, microstructure, and quality of the carbon layer. Finally, the FePO4, Li2CO3, and Li3PO4 precursors were recovered from the black mass separated from the out-of-line (substandard) LiFePO4 cells. The precursors were mechanically mixed at various molar ratios and were calcined at high temperature to regenerate the LFP/C powders. With increasing lithium precursor concentration, the structural, microstructural, and electrochemical properties improved by promoting crystallinity, reducing particle size, and enhancing the quality of the carbon layer. The different morphologies of LFP/C powders, as compared to their precursors, indicated thermal decomposition of metal precursors during carbothermal treatment. A combination of Li2CO3 and FePO4 precursors at a molar ratio of 1:1 yielded high-performance LFP/C powders with a high capacity retention of 90% after 2,000 charge and discharge cycles at 20C. Limitations of the study This study has two main limitations. First, access to high-resolution transmission electron microscop (TEM) imaging was limited; such data could directly visualize LiFePO4 and carbon interfaces and atomic-scale defects, further validating the proposed mechanisms. Second, there was limited access to experimental facilities for post-cycling characterizations. Resource availability Lead contact Further information and requests for resources, materials, and data should be directed to and will be fulfilled by the lead contact, Dr. Seyyed Morteza Masoudpanah (email: masoodpanah@iust.ac.ir). Materials availability This study did not generate new unique materials, cell lines, or biological resources. All synthesized LFP/C samples are available from the lead contact upon reasonable request. Data and code availability •All data reported in this paper will be shared by the lead contact upon reasonable request.•This paper does not report any original code.•No publicly archived datasets were generated during this study. •All data reported in this paper will be shared by the lead contact upon reasonable request. All data reported in this paper will be shared by the lead contact upon reasonable request. •This paper does not report any original code. This paper does not report any original code. •No publicly archived datasets were generated during this study. No publicly archived datasets were generated during this study.