Upcycling lithium extraction by-products from spent lithium-ion batteries into high-voltage polyanionic LiMn_xFe_1−xPO₄

Abstract

Lithium iron phosphate (LiFePO₄, LFP) and lithium manganese oxide (LiMn₂O₄, LMO) batteries dominate the energy storage market due to their superior safety, thermal stability, and cost-effectiveness compared to nickel–cobalt–manganese (NCM) oxide-based systems. However, the imminent retirement of these batteries necessitates sustainable recycling methods. Conventional hydrometallurgical and pyrometallurgical processes hold a dominant position in the present lithium recovery industry but generate low-value by-products such as FePO₄ and MnO₂, the majority of which are discarded, causing issues of resource wastage and environmental pollution. Herein, we propose an upcycling protocol to transform these FePO₄ and MnO₂ residues into LiMn_xFe_1−xPO₄, a promising cathode material renowned for its high operating voltage, enhanced energy density, and structural robustness. Specifically, the recovered FePO₄ and MnO₂ were mixed with lithium and carbon sources with designed molar ratios and then subjected to solid-state reaction at high temperatures to produce LiMn_xFe_1−xPO₄. The synthesized LiMn_0.3Fe_0.7PO₄/C cathode material with a homogeneous carbon coating exhibits remarkable electrochemical performance with a large specific capacity of 164.44 mAh g⁻¹ at 0.1 C, high rate capability (121.73 mAh g⁻¹ at 10 C) and great cycling stability (97.46% capacity retention after 500 cycles at 1 C). Notably, the regenerated cathode material after Mn doping delivers about 20% greater energy density than the regenerated LFP. This work establishes a scalable and eco-efficient methodology for closed-loop battery recycling, successfully addressing critical challenges in resource utilization and high-value material synthesis for advanced energy storage systems.

---

Green foundation

1. This work advances the field of green chemistry by converting the lithium extraction by-products FePO₄ and MnO₂ from waste lithium-ion batteries into high-performance LiMn_xFe_1−xPO₄ cathode materials via a simple and scalable pathway to address the issues of resource waste and environmental pollution.

2. The proposed upcycling strategy affords high-performance and high-value LiMn_xFe_1−xPO₄ electrode materials, with nearly a 20% increase of energy density compared with that of the regenerated LiFePO₄ material and with a profit of $5.57 per kilogram, approximately five times higher than that with the traditional method.

3. The upcycling strategy minimizes environmental impact by reducing chemical and energy inputs in the recycling process, enabling sustainable production of high-value LiMn_xFe_1−xPO₄ cathode materials from waste lithium-ion batteries.

1. Introduction

The sustainable recycling of lithium-ion batteries (LIBs) has emerged as a critical research focus in recent years, driven by escalating environmental concerns and resource sustainability imperatives.^1–4 Among LIB components, cathode materials constitute the predominant mass fraction (typically 40–50 wt%) and dominate the overall battery cost structure.^5,6 Among the commercialized cathode materials, lithium iron phosphate (LiFePO₄, LFP) and lithium manganese oxide (LiMn₂O₄, LMO) account for a significant market share. These materials exhibit distinct advantages in thermal safety, cycle life and production costs.^7,8 Consequently, LFP/LMO-based batteries have been extensively deployed in grid-scale energy storage systems, electric vehicles, and portable energy storage applications.⁹ However, the exponential growth in LFP/LMO battery deployment presents a critical end-of-life management challenge. Projections indicate that global LIB waste streams will reach 6.76 million metric tons by 2035, with LFP/LMO batteries constituting approximately 45% of this volume.¹⁰ Conventional pyrometallurgical and hydrometallurgical recycling approaches have been commercially applied to waste electrode materials like nickel–cobalt–manganese (NCM) oxides containing high-value metals like Li, Co, and Ni. However, they are not economically beneficial for recycling spent LFP/LMO cathode materials, which contain relatively cheap Fe, Mn and P elements except for the high-value lithium element.^11,12 Current state-of-the-art approaches for recycling degraded LFP and LMO materials primarily focus on lithium extraction through hydrometallurgical processes.^13,14 Yang et al. demonstrated effective lithium recovery from a degraded LFP material using H₂SO₄ solution leaching, achieving battery-grade Li₂CO₃ with 99.95% purity.¹⁵ Similarly, Zhang et al. reported a 96.5% lithium leaching efficiency from degraded LMO using sulfuric acid under optimized conditions.¹⁶ Zhang et al. proposed the synergistic leaching of S-LFP and S-LMO using sulfuric acid. The leaching efficiencies of lithium and manganese reached as high as 99.99% and 70.02%, respectively.¹⁷ However, current industrial practices often discard the resultant iron phosphate (FP) and manganese dioxide (MnO₂) residues generated after lithium extraction from spent LFP and LMO, respectively. This disposal practice raises significant concerns regarding both resource utilization efficiency and environmental sustainability.

Meanwhile, the inherent disadvantages of LiFePO₄ and LiMn₂O₄ cathode materials also present significant challenges in their application. LiFePO₄ exhibits a suboptimal operating voltage (3.55 V versus Li⁺/Li), relatively low energy density and poor intrinsic electronic conductivity. LiMn₂O₄ exhibits inferior cycling performance due to manganese dissolution and Jahn–Teller distortion under cycling conditions.¹⁸

To address these issues, cationic substitution strategies have emerged as a promising paradigm for engineering olivine solid solutions. Partial replacement of Fe²⁺ in LFP with redox-active transition metals (e.g., Mn, Co, and Ni) to form LiTM_xFe_1−xPO₄ (TM = transition metals, x < 1) enables precise tuning of the operating voltage while preserving structural integrity. For instance, the LiMn_xFe_1−xPO₄ (LMFP, x < 1) system demonstrates a composition-dependent voltage plateau between 3.4 and 4.1 V. LMFP outperforms LFP in energy density and operating voltage while retaining much of LFP's safety and cost advantages. Its tunable chemistry positions it as a versatile alternative to both LFP and high-voltage layered oxides. Zhou et al. used the leached FePO₄ residue from spent LFP batteries to react with chemical reagents MnCO₃ and Li₂CO₃ for the synthesis of LiMn_xFe_1−xPO₄ cathode materials.¹⁹ Mei et al. recycled mixed spent LFP and LMO cathodes toward high energy density LiMn_xFe_1−xPO₄ cathode materials.²⁰ These practices align with the emerging green strategies in lithium-ion battery recycling and are helpful for the sustainable development of lithium-ion batteries.

In line with green and circular principles, here, we propose a different upcycling strategy to synergistically recycle the by-products FePO₄ and MnO₂ from the lithium extraction process of spent lithium-ion batteries into high-voltage polyanionic LiMn_xFe_1−xPO₄ cathode materials with superior electrochemical performance to their LFP counterparts. The upcycling scenario consists of a ball-milling step for element replenishment and mixing, followed by a solid-state annealing step to regenerate LMFP. As expected, the upgraded LiMn_0.3Fe_0.7PO₄/C (simply denoted as R-LM₃F₇P) cathode material exhibits excellent electrochemical properties, delivering an initial discharge capacity of 164.44 mAh g⁻¹ at 0.1 C and 121.73 mAh g⁻¹ at 10 C, while maintaining an impressive capacity retention of 97.46% after 500 cycles at 1 C. The R-LMFP increased the energy density by approximately 20% compared with the regenerated LiFePO₄/C (R-LFP), exhibiting comparable performance to that of commercial LFP materials. Interestingly, the R-LMFP exhibits superior cycling stability to LFP at −20 °C, with a high capacity retention rate of 99.95% after 1000 cycles, which opens up low-temperature application scenarios for the R-LMFP cathode material. This facile upcycling method provides a viable and scalable approach for the production of next-generation phosphate cathode materials for high-performance LIBs and is conducive to their sustainable development.

2. Experimental section

2.1. Materials and reagents

Acetic acid (CH₃COOH, Sinopharm, AR ≥ 99.99%), sulfuric acid (H₂SO₄, Sinopharm, AR ≥ 99.99%), lithium carbonate (Li₂CO₃, Aladdin, AR ≥ 99.99%), lithium dihydrogen phosphate (LiH₂PO₄, Aladdin, AR ≥ 99%), and glucose (C₆H₁₂O₆, Aladdin, AR ≥ 99%) were used as received. The recovered spent electrode materials used in the experiment were from Anhui Nandu Huabo New Materials Technology Co., Ltd, and their pretreatment is described in the following experimental procedures.

2.2 Experimental procedures

2.2.1. Preparation of the lithium extraction by-products. S-LFP and S-LMO cells were discharged to approximately 20% of the charged state. Then, these two cells were carefully disassembled manually to separate the cathode sheets in a dry room (at −60 °C dew point). These cathode sheets were calcined at 600 °C for 1 h under a N₂ atmosphere to eliminate residual binders and conductive carbon black, and this was followed by mechanical oscillation to detach the S-LFP powder and S-LMO powder from their aluminum foil current collectors, respectively, thus obtaining the corresponding waste cathode material powders. Inductively coupled plasma optical emission spectroscopy (ICP-OES) results indicate that spent LFP material contains lithium (4.35 wt%), iron (34.8 wt%), phosphorus (19.26 wt%) and low content of aluminum (0.031 wt%), and spent LMO contains lithium (3.59 wt%), manganese (56.36 wt%) and aluminum (0.087 wt%), respectively (Tables S1 and S2). Besides, both spent electrode materials include minor impurities like Cu, Zn, Fe and other elements.

Then, we conducted the investigation of acid leaching and ball-milling experiments by examining some influencing factors, including acid leaching concentration, leaching time and ball-milling parameters. The detailed experimental conditions are described below.

Leaching of S-LFP. 5 g of S-LFP powder was added to a 250 mL beaker and then mixed with 50 mL of 1.2 mol L⁻¹ acetic acid and 2.5 mL H₂O₂ (30 wt%) through magnetic stirring at room temperature for 1 hour. The remaining solid material was filtered, washed with deionized water, and then dried at 100 °C for 10 h to obtain FePO₄.
Leaching of S-LMO. 2.715 g of S-LMO powder was added to 40 mL of 2 M sulfuric acid solution in a 250 mL beaker and magnetically stirred at room temperature for 3 h. When the color of the mixture turned red and the pH of the mixture stabilized, the remaining solid material was filtered, washed with deionized water, and then dried at 100 °C for 10 h to obtain MnO₂.
Ball-milling experiment. The ball-milling test was conducted using a desk-top high-speed vibrating ball mill (MSK-SFM-3), with 1.039 g of precursor mixture at a ball/powder mass ratio of 20 : 1.

The relevant optimization results for acid leaching concentration, leaching time and ball-milling parameters have been supplemented in the SI (see Fig. S1 and S2). The leached lithium ions can be recovered as lithium carbonate or lithium hydroxide. The compositions of the recovered FePO₄ and MnO₂ by-products were measured by ICP-OES, and the results are listed in Tables S3 and S4, respectively. Compared with the S-LFP and S-LMO, the recovered FePO₄ and MnO₂ samples contain lower aluminum contents of 0.018 wt% and 0.021 wt%, and no copper impurities are detected after the acid-leaching treatment.

2.2.2. Preparation of the R-LMFP and R-LFP materials. The upcycling process consists of two distinct stages. In the first stage, the leached FePO₄ and MnO₂ materials were placed in an agate ball-milling jar at various Fe/Mn ratios (7 : 3, 6.6 : 3.3, 6 : 4, 5 : 5). To compensate for the loss of lithium and phosphorus and to facilitate the formation of a carbon coating, Li₂CO₃, LiH₂PO₄, and 10 wt% glucose were added, with molar ratios of Li₂CO₃ : LiH₂PO₄ being 7/2, 3/1, 3/4, and 5/2, respectively. The designed precursor mixture was subjected to ball milling under optimal conditions (600 rpm for 6 hours, Fig. S2) under an argon atmosphere. After drying at 80 °C, the precursor powder was placed in an alumina crucible and calcined under an argon atmosphere based on a programmed procedure:^19,21 held at 400 °C for 2 hours, followed by 600 °C for 8 hours, with a heating rate of 5 °C min⁻¹. After annealing, the regenerated LiMn_xFe_1−xPO₄ (R-LM_xF_1−xP) cathode material was obtained. For comparison, we also regenerated LiFePO₄ (R-LFP) from the recovered FePO₄. Similarly, we mixed the recovered FePO₄ with lithium carbonate (the molar ratio of FePO₄ to Li₂CO₃ of 1 : 0.525) and 5 wt% glucose through ball milling at 600 rpm for 6 hours. Then the obtained mixture powder was transferred to an alumina crucible and annealed under an argon atmosphere using a programmed procedure: at 400 °C for 2 hours and then at 700 °C for 8 hours with a heating rate of 5 °C min⁻¹, finally yielding R-LFP.

2.3 Materials characterization

An inductively coupled plasma mass spectrometer (ICP-MS, Agilent, A-7900QMS) was employed to analyze the elements including Li, Fe, Mn, P Al and Cu in the electrode materials. The phases and compositions of the LiMn_xFe_1−xPO₄ samples and LiFePO₄ samples were characterized using an X-ray diffraction instrument (XRD, PANalytical X-Pert PRO MPD) with Cu-Kα radiation (λ = 0.154178 nm), operated at 40 kV and 80 mA. X-ray photoelectron spectroscopy (XPS) was performed using a Thermo ESCALAB250Xi instrument.

2.4 Battery assembly and electrochemical tests

The electrochemical properties of the upgraded LiMn_xFe_1−xPO₄ and the regenerated LiFePO₄ electrode materials were characterized using both 2032-type coin cells and pouch cells. For manufacturing the positive electrode with the regenerated LiMn_xFe_1−xPO₄ material, 80 wt% of LiMn_xFe_1−xPO₄ powder, 10 wt% of acetylene black, and 10 wt% of polyvinylidene fluoride (PVDF) were mixed. Then, an appropriate amount of N-methyl-2-pyrrolidone (NMP) was introduced into the above mixture to form a uniform slurry, which was evenly coated onto an aluminum-foil current collector and then dried at 80 °C for 12 h. For assembling R2032 coin cells, the aluminum foil loaded with the positive electrode active material was cut into fixed-size discs and used as the positive electrode, with lithium foil as the counter electrode, Celgard 2400 as the separator, and a mixture of lithium hexafluorophosphate (LiPF₆) and ethylene carbonate (EC), dimethyl carbonate (DMC), and diethyl carbonate (DEC) as the electrolyte solution. The batteries were assembled in an argon-filled glove box (H₂O < 0.5 ppm; O₂ < 0.5 ppm). The assembly procedure for R2032-type coin cells for regenerated LiFePO₄ is similar to that described above for LiMn_xFe_1−xPO_4.

A lamination method was used to prepare the pouch cells. They used the same cathode, separator, and electrolyte as the coin cells, but the anode is made of commercial graphite. The mass loading weights of the cathode and anode were approximately 9.39 mg cm⁻² and 10.41 mg cm⁻², with designed capacities of 42.27 mAh g⁻¹ and 37.86 mAh g⁻¹, respectively. The anode-to-cathode capacity ratio is roughly 1.1. Each pouch cell comprising one layer of anode (42 mm × 56 mm in size, single-side coated) and one layer of cathode (40 mm × 54 mm in size, single-side coated) achieves a designed capacity of about 17.86 mAh. The pouch cells were activated three times at 0.1 C and then cycled at 1 C in the range of 2.5–4.5 V. For electrochemical performance measurements, constant current charge/discharge tests were conducted using a battery testing system (Neware, CT-4008Tn-5V10 mA) in a voltage window of 2.5–4.5 V. Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were performed using an electrochemical workstation (CHI-760E, Chenhua Instruments).

2.5 Techno-economic analysis

We performed a techno-economic analysis (TEA) of the downgrading recycling, direct recycling, and upcycling processes based on the EverBatt 2023 model. We took 10 000 tonnes of spent LFP and 10 000 tonnes of spent LMO end-of-life batteries as the basis for calculation, and the related pretreating and different recycling processes for the separated black powders were included. The detailed costs of raw materials and revenue from the products are listed in the tables of the SI. Since the LMFP material has not been widely commercialized and its market price is unavailable, here we assume that the price of LMFP is 1.2 times that of LFP, considering that the upcycled LMFP exhibits approximately 20% higher energy density than the recovered LFP.

3. Results and discussion

3.1 Experimental design

Due to the structural incompatibility between the olivine-type crystal structure of LFP and the spinel-type structure of LMO, it is unfeasible for direct upcycling synthesis of olivine-structured LMFP from S-LFP and S-LMO. Consequently, we conducted lithium extraction from S-LFP and S-LMO to obtain FePO₄ and MnO₂ by-products, respectively, which were subsequently utilized for upcycling regeneration as illustrated in Scheme 1. In the initial step of the upcycling process, the raw materials including the recovered FePO₄ and MnO₂, as well as Li₂CO₃, LiH₂PO₄, and glucose reagents were mixed through ball milling. Given the hygroscopic nature of lithium hydrogen phosphate (LiH₂PO₄), the ball milling was conducted under an argon atmosphere to prevent moisture absorption. Then, a carbothermal reduction reaction was employed to facilitate structural reconstruction and surface modification, ultimately yielding carbon-coated R-LiMn_xFe_1−xPO₄ materials (see the Experimental section for detailed procedures). The addition of glucose served as a carbon source, enabling the formation of a conductive carbon layer applied on the R-LiMn_xFe_1−xPO₄ particles, thereby enhancing the electrical conductivity of the polyanionic materials. For comparative analysis, the regenerated LiFePO₄ (R-LFP) sample was also synthesized from S-LFP and used as a benchmark for evaluating the performance of the regenerated R-LiMn_xFe_1−xPO₄.

Scheme 1 Schematic illustration of the upcycling process for LiMn_xFe_1−xPO₄.

3.2 Phase and composition analysis

The phase composition and crystal structure of the regenerated R-LMFP samples with four different Mn/Fe molar ratios as well as the regenerated R-LFP sample were characterized by XRD. The diffraction peaks of the four R-LMFP materials are consistent with the olivine structure of R-LFP and can be indexed to the orthorhombic Pnma space group (PDF#83-2092), as shown in Fig. 1a. Additionally, Fig. 1b exhibits the high-resolution XRD patterns of R-LFP and R-LMFP materials with different Mn/Fe molar ratios within the 2θ range of 24° to 38°, revealing that the (111), (211), and (311) peaks of R-LMFP shift toward lower angles compared with those of R-LFP. This shift becomes more pronounced with increasing Mn content, indicating that Mn doping induces lattice expansion in R-LMFP. Fig. 1c and d show the Rietveld refinement results of the XRD patterns for R-LM₃F₇P and R-LFP, with final reliability factors of R_wp = 6.30% and R_wp = 5.25%, respectively. Rietveld refinements were also performed for the R-LMFP materials with other Mn/Fe molar ratios (Fig. S3 and Table S5), and no obvious alien diffraction peaks are observed, suggesting that the obtained materials are free of impurities. We established the simulation models of the R-LFP and R-LMFP materials through DFT calculations (Fig. S4, calculation details are provided in the SI), and the original bond lengths of Fe–O1 and Fe–O2 are determined to be 2.08 Å and 2.06 Å for the R-LFP crystal. After doping with manganese ions, the bond lengths of Mn–O1 and Mn–O2 are 2.11 Å and 2.11 Å for the R-LM₃F₇P crystal. This is because the ionic radius of Mn²⁺ (0.83 Å) is larger than that of Fe²⁺ (0.63 Å), resulting in a slight increase in the transition-metal–oxygen (M–O) bond lengths. The above results indicate that the unit cell volume of R-LM₃F₇P is larger than that of R-LFP. This volume change broadens the lithium-ion diffusion channels and reduces the diffusion energy barrier, thereby enhancing the lithium-ion migration rate and rate performance.

Fig. 1 XRD characterization of the R-LMFP and R-LFP samples: (a) XRD patterns and (b) enlarged XRD patterns at 2θ = 24–38° of the four different R-LMFP materials; (c) and (d) XRD Rietveld refinement of R-LM₃F₇P and R-LFP.

XRD patterns of R-LM₃F₇P and R-LFP exhibit no distinct peaks associated with carbon, signifying the amorphous characteristics of carbon in the upgraded compounds. Raman spectra (Fig. S5) of R-LM₃F₇P and R-LFP exhibit characteristic D and G bands at around 1340 cm⁻¹ and 1580 cm⁻¹, respectively, which correspond to the disorder-induced phonon mode and the E_2g vibration of graphite, respectively. The I_D/I_G ratio (the area ratio of the D-band to the G-band) for R-LM₃F₇P is 0.95, lower than 0.99 for R-LFP (Table S6), and a lower I_D/I_G ratio indicates that the sp² carbon network in the material is more complete, indicating a higher degree of graphitization for carbon content in R-LM₃F₇P and hence a better electronic conductivity for it.²² To precisely identify the chemical compositions of the upcycled materials, inductively coupled plasma-optical emission spectroscopy (ICP-OES) analysis was conducted, with results presented in Table S7. The chemical compositions of R-LM₃F₇P and R-LFP are determined to be LiMn_0.29Fe_0.71P_0.95O_3.832 and LiFe_0.989P_0.980O_3.919, respectively, both closely aligning with the target compositions of LiMn_0.3Fe_0.7PO₄ and LiFePO₄. Additionally, thermogravimetric analysis (TG) was used to investigate the mass changes of R-LM₃F₇P and R-LFP during heating in air. The mass loss of R-LM₃F₇P and R-LFP mainly comes from the oxidation of transition metal elements (Fe/Mn) and the removal of carbon. Therefore, the estimated carbon contents in R-LFP and R-LM₃F₇P were determined to be 3.53% and 7.57%, respectively (Fig. S6).

3.3 Microstructure characterization

The scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HRTEM) images of the recovered FePO₄ and MnO₂ are shown in Fig. 2a and b. It can be seen that the FePO₄ and MnO₂ samples are mainly composed of micro-sized irregular particles, with obvious micro-/nano-cracks resulting from their repeated charge/discharge cycling and the lithium extraction process. These cracks promote fragmentation of particles into smaller sizes during ball-milling treatment, thus beneficial for subsequent element diffusion and regeneration of the final products during the subsequent calcination process. The SEM images in Fig. 2c show that the regenerated R-LFP and R-LM₃F₇P materials consist of nanoscale particles, which are conducive to lithium-ion transfer during charge and discharge.

Fig. 2 Morphology and microstructure characterization of the materials: (a) SEM images of FePO₄ and MnO₂ particles; (b) HRTEM images of FePO₄ and MnO₂ particles; (c) SEM images of R-LFP and R-LM₃F₇P particles; (d) HRTEM images of R-LFP and R-LM₃F₇P particle surface layer and the corresponding lattice-resolved iFFT images; (e) EDS mapping of R-LM₃F₇P particles; and (f) the elemental atomic fractions of R-LM₃F₇P.

The regenerated materials were further characterized by HRTEM and the Fast Inverse Fourier Transform Mode (iFFT), as presented in Fig. 2d. The R-LFP particles are coated with a uniform carbon layer of about 2.5 nm thickness, and the R-LM₃F₇P particles are coated with a carbon layer of about 4.5 nm thickness. A clear olivine crystal structure with a lattice spacing of 0.348 nm is observed for R-LFP, which corresponds to the (111) crystal faces of the Pnma-type structure of LiFePO₄. Likewise, R-LM₃F₇P exhibits a clear olivine structure, suggesting that Mn doping does not disrupt the crystal framework. The planar spacing in iFFT is determined to be 0.363 nm, corresponding to the (101) crystal plane. At the microscopic level, energy-dispersive spectroscopy (EDS) elemental mapping (Fig. 2e) shows a uniform distribution of Fe, Mn, P, and O elements in an R-LM₃F₇P particle. The atomic ratios of elements in R-LM₃F₇P, shown in Fig. 2f, indicate an Mn/Fe ratio of 2.98/7.02, closely aligning with a target ratio of 3/7.

X-ray photoelectron spectroscopy (XPS) analysis was performed to observe the surface information of the upgraded samples, as shown in Fig. S7. The XPS spectra of R-LM₃F₇P and R-LFP clearly show the expected characteristic peaks of the target elements (Fig. S7a). Fig. S7b shows the high-resolution Fe 2p XPS spectra of the two samples. For R-LMFP, the characteristic peaks are centered at binding energies of 710.78 eV (Fe 2p3/2) and 724.33 eV (Fe 2p1/2). For R-LFP, the characteristic peaks are detected at binding energies of 712.18 eV (Fe 2p3/2) and 725.93 eV (Fe 2p1/2). Both materials exhibit the presence of Fe²⁺via the carbothermal reduction treatment. The Mn 2p XPS spectra of the R-LM₃F₇P material (Fig. S7c) show distinctive peaks at binding energies of 642.18 eV (Mn 2p3/2) and 653.98 eV (Mn 2p1/2), showing Mn²⁺ dominance. The XPS high-resolution spectra of C 1s of R-LFP and R-LM₃F₇P materials (Fig. S7d) can each be resolved into three subpeaks at binding energies of 284.8 eV (C–C), 286 eV (C–O–C), and 288.5 eV (O–C=O),²³ and the area ratios of the C–O–C subpeak to the C–C subpeak for R-LM₃F₇P and R-LFP are 0.328 and 0.51 (Table S8), respectively, showing that R-LM₃F₇P contains less carbon defects, consistent with the previous Raman test results.²²

3.4 Electrochemical evaluation

The above compositional and structural analyses imply that R-LMFP materials may exhibit ideal electrochemical performance. First, we evaluated the electrochemical properties of the regenerated R-LMFP with different Mn/Fe molar ratios. The galvanostatic charge/discharge profiles in Fig. 3a indicate that the four cathode materials, namely R-LM₅F₅P, R-LM₄F₆P, R-LM_3.3F_6.6P and R-LM₃F₇P, exhibit similar redox reactions during charge and discharge processes, which are Mn²⁺–Mn³⁺(∼4.0 V) and Fe²⁺–Fe³⁺ (∼3.5 V), respectively. R-LM₅F₅P, R-LM₄F₆P, R-LM_3.3F_6.6P and R-LM₃F₇P present specific discharge capacities of 142.75, 149.91, 156.46, and 164.44 mAh g⁻¹ at a current density of 0.1 C, respectively. At a high rate of 10 C, the four cathode materials are able to deliver high discharge specific capacities of 113.76, 112.93, 121.11 and 121.73 mAh g⁻¹, respectively (Fig. 3b). In particular, R-LM₃F₇P demonstrates the best electrochemical performance at both low and high rates among the four. These testing results reveal that with increasing manganese content, the specific capacity of R-LMFP gradually decreases. This might be because the increase in manganese content leads to an intensification of the disproportionation reaction and manganese dissolution, thus resulting in discharge specific capacity decay with higher manganese doping content.²⁴

Fig. 3 Electrochemical performance of the upgraded R-LMFP with different Mn/Fe molar ratios: (a) galvanostatic charge/discharge profiles at 0.1 C in the voltage range of 2.5–4.5 V; (b) rate performance; (c) Nyquist plots at 4.5 V; (d) cycling performance at 1 C in the voltage range of 2.5–4.5 V.

The electrochemical impedance spectroscopy (EIS) data presented in Fig. 3c validate the enhanced capacitor ability of the R-LMFP materials. In the Nyquist plot, the high-frequency semicircles of the four materials correspond to the ohmic resistance (R_s) and charge transfer resistance (R_ct), respectively, indicating the constraints on electron transport. The linear portion at low frequencies reflects the diffusion limitation. The four R-LMFP materials with different Mn/Fe molar ratios exhibit similar R_s and R_ct values in the high- and low-frequency regions. As shown in Fig. S8, the diffusion coefficient of Li⁺ in R-LM₃F₇P is 1.00 × 10⁻¹¹ S cm⁻², which is greater than that of R-LMFP in other proportions (Table S9). This indicates that Mn doping has little effect on material impedance, but it affects Li⁺ transport kinetics, and the Li⁺ diffusion coefficients decrease with the Mn doping content increasing.

Fig. 3d presents the cyclic testing results of the four R-LMFP cathode materials with different proportions at 1 C. After 500 cycles, R-LM₃F₇P, R-LM_3.3F_6.6P, R-LM₄F₆P and R-LM₅F₅P maintain capacity retention rates of 97.46%, 92.25%, 91.47% and 63.39%, respectively. This shows that the cycling stability of the materials declined with the increase of manganese content in R-LMFP, and especially the R-LM₅F₅P material with the highest Mn content of the four exhibits dramatic attenuation in capacity after about 250 cycles. This is ascribed to the Jahn–Teller effect caused by Mn³⁺, which causes lattice distortion and therefore reduces the long-term cycle performance.

We also made a comparative performance evaluation of the upgraded R-LM₃F₇P with regenerated R-LFP. As shown in Fig. 4a, compared with R-LFP, R-LM₃F₇P exhibits two voltage platforms around 3.5 V and 4.0 V, respectively, corresponding to the oxidation reactions of Fe²⁺/Fe³⁺ and Mn²⁺/Mn³⁺, respectively. Compared with pure LiFePO₄ (≈3.4 V) and LiMnPO₄ (≈4.1 V), the discharge voltage platforms of Fe²⁺/Fe³⁺ and Mn²⁺/Mn³⁺ in R-LM₃F₇P materials are slightly different (≈3.5 V, 4.0 V, respectively), attributable mostly to the ion super-exchange of transition metals within the R-LMFP solid solution.²⁵ The R-LMFP material contains both Fe²⁺ and Mn²⁺ ions that can mutually substitute for each other in the crystal lattice.²⁵ Furthermore, the dQ/dV curves of each of the products are illustrated in Fig. 4b; in comparison with R-LFP, R-LM₃F₇P demonstrates a polarization voltage decrease of 0.02 V (refer to the magnified section inset in Fig. 4b), indicating that the upcycled R-LMFP material has well restored the crystal structure and exhibits excellent reversibility.²⁶ As shown in Fig. 4c, R-LM₃F₇P demonstrates higher specific discharge capacities than R-LFP across various current densities ranging from 0.1 to 10 C. R-LM₃F₇P delivers a high specific capacity of 164.44 mAh g⁻¹ at 0.1 C, compared to 156.50 mAh g⁻¹ for R-LFP. Particularly, at a high rate of 10 C, it still delivers a specific capacity of 121.73 mAh g⁻¹, exhibiting a remarkable capacity enhancement of 26.45 mAh g⁻¹ compared to R-LFP. Notably, the R-LM₃F₇P cathode achieves a significant 20% enhancement in energy density (568.34 Wh kg⁻¹ for R-LM₃F₇P and 454.67 Wh kg⁻¹ for R-LFP, as shown in Fig. S9). This increase in energy density is due to R-LM₃F₇P's improved composition and structural characteristics, which leverage the synergistic effects of manganese and iron in the olivine framework. Cyclic voltammetry (CV) analysis provides valuable insights into the electrochemical behavior and stability of R-LM₃F₇P and R-LFP. Fig. 4d and e show the contour maps of the CV curves of R-LM₃F₇P and R-LFP at 0.1 mV s⁻¹. The iron redox peaks of R-LM₃F₇P have narrow half-peak widths of 0.027 V and 0.045 V, while those of R-LFP have narrow half-peak widths of 0.073 V and 0.071 V. The narrower half-peak widths indicate that R-LM₃F₇P has high crystallinity, low defect density and a uniform crystal structure. These characteristics reduce resistance to lithium-ion diffusion and charge transfer, thereby promoting rapid electrochemical kinetics and efficient lithium-ion intercalation/deintercalation processes.²⁷ With the increase in the number of scanning cycles, the good symmetry and stability of the redox peaks further indicate that R-LM₃F₇P has outstanding reversibility due to its optimized composition and well-defined crystal structure, as well as nanoscale particle size with uniform carbon coating (Fig. S10).

Fig. 4 Electrochemical performance comparison of R-LM₃F₇P and R-LFP materials. (a) Initial charge and discharge curves of R-LM₃F₇P and R-LFP; (b) the associated dQ/dV profiles; (c) rate performance; (d and e) the contour maps of CV curves at different cycles; (f) retention rates after 100 cycles at 1 C and 5 C current densities for R-LFP and R-LM₃F₇P; (g) CV curves of R-LM₃F₇P at different scan rates (0.01–0.08 V s⁻¹); (h) comparison of capacity contributions of diffusion control and pseudo-capacitance; and (i) radar comparison chart of relative cathode materials.

In addition, we compared the cycling performance of coin cells using R-LM₃F₇P and R-LFP cathodes at 1 C and 5 C. As shown in Fig. 4f, the R-LM₃F₇P cathode material exhibits higher specific discharge capacities (148.42 and 125.92 mAh g⁻¹) and higher average coulombic efficiencies (99.62% and 99.53%) at 1 C and 5 C, respectively, compared to the R-LFP cathode material with the corresponding data of 143.52 and 117.15 mAh g⁻¹, 99.34% and 99.03%, respectively. The R-LM₃F₇P cathode also demonstrates excellent cycling stability, with capacity retention rates of 99.93% and 99.86% after 100 cycles at 1 C and 5 C, respectively. In summary, Mn doping significantly improves the specific discharge capacity of the cathode material, while the surface carbon coating and nanoscale particle size promote the lithium ion transport of the R-LM₃F₇P material.

The electrochemical kinetics investigation of R-LM₃F₇P was performed by CV testing at different scan speeds (Fig. 4g and h). The electrochemical process in R-LM₃F₇P is characterized by diffusion at low currents and by pseudocapacitive behavior at high currents. Small reduction peaks with distinct rises are observed between peak 3 and peak 4 as the scan rate increases (Fig. 4g, marked with a red circle). Research indicates that the small peak in the CV profile arises from the Mn³⁺/Mn²⁺ redox couple, distinct from the conventional Mn³⁺/Mn²⁺ reaction due to its lower voltage. As shown in Fig. 4g, this peak becomes more pronounced with increasing current density, reflecting a kinetically controlled transient process. The Mn³⁺/Mn²⁺ reaction exhibits superior kinetics compared to the classic Mn³⁺/Mn²⁺ reaction, with higher current densities amplifying this effect. This aligns with Fig. 3b, where higher manganese content correlates with reduced discharge capacity at high rates. As shown in Fig. 4h, at high current density, owing to the augmented influence of pseudocapacitive effects, the electrode surface may swiftly intercalate and deintercalate lithium ions, consequently sustaining a superior capacity output. This is also in line with the superior performance of R-LM₃F₇P, which delivers 26.45 mAh g⁻¹ higher capacity than R-LFP at a current density of 10 C.

To evaluate the practical application potential of the R-LM₃F₇P material, we also made a pouch cell with the R-LM₃F₇P electrode as the cathode and commercial graphite as the anode. As shown in Fig. S11, the pouch cell can maintain a high retention rate of 87.93% after 600 cycles at 1 C in a voltage range of 2.5–4.5 V at room temperature, and the inset photo demonstrates the LED bulbs powered by the pouch cell. These performances indicate that the R-LM₃F₇P material is promising for practical applications. Furthermore, we also investigated the low-temperature performance of R-LM₃F₇P and R-LFP at −20 °C and found that R-LM₃F₇P delivers a specific capacity of 60.47 mAh g⁻¹ at 1 C, outperforming R-LFP, which shows 44.62 mAh g⁻¹ (Fig. S12). The initial charge–discharge coulombic efficiency (ICE) of R-LM₃F₇P is 99.8%. As shown in Fig. S12b, R-LM₃F₇P still achieves a specific capacity of 25 mAh g⁻¹ at a low temperature of −20 °C and a 10C high rate, while R-LFP delivers almost zero specific capacity. Moreover, R-LM₃F₇P can cycle 1000 times at a 1C rate with a capacity retention of 99.68% (Fig. S12c), exhibiting its outstanding cycling stability under low-temperature conditions.

The electrochemical properties of the R-LM₃F₇P material in this study are compared with the representative regenerated LFP and LM_xF_1−xP materials reported recently (Fig. 4i and Table S10).^19,20,28,29 It can be readily concluded that the electrochemical performance of our R-LM₃F₇P material is comparable, or even marginally superior, to that of previously reported materials in key properties, including specific capacity, energy density, capacity retention rate and cycling stability. This enhancement underscores the efficacy of the upcycling process employed, as well as the structural and compositional optimization of the R-LM₃F₇P material. These results highlight its potential as a high-performance cathode material for advanced lithium-ion battery applications.

Furthermore, through in situ XRD measurements, we clarify the structural evolution mechanism of R-LMFP and reveal its structural changes during the delithiation/lithiation process.^30,31 During the first charge–discharge cycle (Fig. 5a), when the voltage is charged to 3.47 V, the diffraction peaks (211) and (311) gradually move to higher angles, the LMFP phase gradually disappears, and a new phase of Li_1−xMn_0.3Fe_0.7PO₄ (L_1−xMFP) begins to appear. Later, when the voltage is charged to 4.0 V, the MFP phase appears. It can be determined that the two-phase reactions (L_xMFP ↔ MFP) coexist during the process. The above results suggest that R-LM₃F₇P is a dynamic solid solution during charge and discharge, which is consistent with previous studies.³² To sum up, during the charging process, R-LM₃F₇P exhibits two distinct phase transitions: first from R-LM₃F₇P to the intermediate Li_1−xMn_0.3Fe_0.7PO₄ (L_1−xMFP) phase and then to the Mn_0.3Fe_0.7PO₄ (MFP) phase. Furthermore, the in situ XRD patterns of the subsequent discharge process are different from those of the charging process, indicating that the phase transitions of lithium deintercalation/intercalation follow two opposite reactions, respectively, as shown in Fig. 5b.

Fig. 5 (a) Charge and discharge curves and in situ XRD contour map of R-LM₃F₇P; (b) diagram illustrating the alterations in crystal structure during the charge and discharge processes. LiMn_0.3Fe_0.7PO₄, Li_1−xMn_0.3Fe_0.7PO₄ (1 ≥ x ≥ 0) and Mn_0.3Fe_0.7PO₄ are denoted as LMFP, L_1−xMFP and MFP, respectively; (c) unit cell volume change during the charging process of R-LMFP.

Interestingly, through calculation, we find that the unit cell volume undergoes continuous contraction with the phase transformation as the voltage increases (Fig. 5c). The displacement of the (211) peak represents the change in the b-axis of the crystal. As can be seen from Fig. S13, the b-axis is gradually decreasing. This might be because lithium ions mainly migrate along the b-axis direction. When lithium ions are released, the b-axis usually contracts slightly. For lithium iron manganese phosphate, a solid solution of Mn and Fe, the difference in ionic radii between Mn²⁺and Fe²⁺ (approximately 0.83 Å for Mn²⁺ and approximately 0.63 Å for Fe²⁺), as well as the changes in the ratio of Mn³⁺/Fe³⁺ during the oxidation process, will have an impact on the shrinkage degree of the b-axis. Furthermore, the Jahn–Teller effect of Mn³⁺ may lead to local lattice distortion, resulting in a reduction of the b-axis.^32,33

In addition, we conducted a post-process analysis of R-LM₃F₇P and R-LFP after 500 cycles at 1 C. Fig. S14 displays the post-process structural characterization of the R-LM₃F₇P material after 500 cycles at 1 C. The SEM image shown in Fig. S14a exhibits the morphology of the R-LM₃F₇P material after cycling, and it is somewhat different from the original morphology of R-LM₃F₇P (Fig. 2c), due to exposure to various processes including mixing with conductive carbon and organic binder, high-pressure rolling, electrolyte soaking, and charge/discharge cycles. The STEM images (Fig. S14b) show that the thickness of the coating layer on the R-LM₃F₇P particle surface after cycling is greater than before cycling, which may be due to the formation of CEI (i.e., the cathode electrolyte interface). The corresponding lattice-resolved STEM images (Fig. S14c) of R-LM₃F₇P after cycling show clear lattice spacings of 0.205/0.368 nm, corresponding to the interplanar distances of the (321)/(210) crystal planes of the olivine-structured LMFP. Besides, no disordered structures were found in the STEM observation, indicating its good structural stability. In addition, the EDS element mapping of the Fe, Mn, P, and O elements (Fig. S14d) for the cycled R-LM₃F₇P particles still shows a uniform synchronous distribution of these elements. Furthermore, as shown in Fig. S14e–g, XPS analysis was conducted to further investigate the surface properties of the material after cycling. Through fitting, the P 2p XPS spectrum can be divided into three peaks: Li_xPO_yF_z, LiFePO₄, and FePO₄. Among them, Li_xPO_yF_z is ascribed to the CEI layer.³⁴ Comparing the high-resolution P 2p XPS spectra of the two samples (Fig. S14e and S15d), it is found that the relative intensity of R-LFP was higher than that of R-LM₃F₇P after cycling, indicating that the cycled R-LFP electrode has a thicker CEI layer than the cycled R-LFP electrode due to side reactions. This is confirmed by the increase of coating layer thickness of the two samples: for R-LM₃F₇P, it increased from 4.5 nm to 5.5 nm (Fig. S14b), and for R-LFP, it increased from 2.5 nm to 4.5 nm (Fig. S15a). Fig. S14e shows the high-resolution Fe 2p XPS spectrum of R-LM₃F₇P, in which the Fe²⁺ and Fe³⁺ peaks appear at 711.08 eV and 713.88 eV, respectively. For cycled R-LM₃F₇P, the Fe²⁺ peak exhibits higher relative intensity, and the Fe³⁺ peak shows lower relative intensity, compared to the Fe²⁺ peak (710.98 eV) and Fe³⁺ peak (711.78 eV) of R-LFP (Fig. S15e), indicating that Mn doping significantly suppressed Fe²⁺ to Fe³⁺ conversion for R-LM₃F₇P. The peaks centered at 642.88 eV and 654.68 eV are attributed to Mn 2p1/2 and Mn 2p3/2 for Mn²⁺, which show no significant difference compared to fresh R-LM₃F₇P (Fig. S7c and S14g). These characterization studies not only explain the increased surface thickness observed by SEM for cycled samples, but also further demonstrate that the structural design of R-LM₃F₇P makes the material more stable.

For comparison, Fig. S15 exhibits the postmortem structural characterization of the R-LFP material after 500 cycles at 1 C. The lattice-resolved STEM image of the cycled R-LFP material (Fig. S15a) shows a lattice spacing of 0.28 nm, corresponding to the interplanar distance of the (021) crystal planes in olivine-structured LFP. However, in region 2, disordered structures with no diffraction spots are also found in the FFT images, suggesting surface structure degradation (Fig. S15b and c). The above results indicate that the structure of R-LFP is not as stable as that of R-LM₃F₇P.

3.6 Techno-economic analysis

Fig. 6a summarizes the recycling methods for waste lithium-ion batteries, particularly cathodes containing low concentrations of valuable elements. The traditional recycling methods for S-LFP and S-LMO involve using pyrometallurgical, hydrometallurgical, or direct repair methods to convert the by-products obtained after lithium extraction into metal salts or metal oxides. These methods have minimal economic benefit and are inappropriate for recycling S-LFP and S-LMO. Innovative regeneration techniques seek to maintain the material's integrity while concentrating on reinstating the composition and structure, rendering the regenerated materials suitable for new batteries. While these approaches preserve the worth of the recycled materials, their economic advantages are still constrained. The upgrade recycling method established in our research converts low-value recycled by-products into high-performance LiMn_xFe_1−xPO₄ products, hence accomplishing value-added recycling.

Fig. 6 Comparison of different battery recycling technologies and techno-economic analysis. (a) Illustration of downgrading recycling, direct recycling, and upcycling processes for S-LFP and S-LMO cathodes. (b) Cost analysis. (c) Revenue analysis. (d) Profit analysis.

We perform a techno-economic analysis (TEA) of different strategies for recycling with the EverBatt 2023 model to demonstrate the benefits of this upcycling technology. The yearly capital cost is identified as the primary component in the cost analysis for all technologies. In downgrading recycling, since metal salts or metal oxides are directly recovered without the need for additional chemicals for modification or upgrading, the material costs are relatively low. In comparison, material costs are a significant component of the total costs in the other two recycling processes, such as the lithium salt cost in the direct recycling and upcycling processes. Consequently, the recycling costs for downgrading recycling, direct recycling, and upcycling are $0.37, $1.09, and $1.61 per kilogram of raw materials, respectively (Fig. 6b). Revenue analysis is directly linked to the product values. In downgrading recycling, only iron and manganese products are produced, resulting in low revenue. Conversely, recycled materials (LFP, LMO, and LMFP) constitute the principal sources of overall income in the direct recycling and upcycling processes. Additionally, recycled manganese salts further boost the revenue in upcycling. The revenue per kilogram of raw material for downgrading recycling, direct recycling, and upcycling is $1.37, $4.78, and $7.58, respectively (Fig. 6c). Therefore, our proposed upcycling process achieves the highest profit, at $5.57 per kilogram of raw material, as shown in Fig. 6d. Overall, based on the processing scale of 10 000 tons of waste lithium iron phosphate batteries and 10 000 tons of waste lithium manganate batteries annually, a profit of $109 400 000 can be achieved, indicating that this upgrading strategy holds significant market potential.

4. Conclusion

We have successfully developed a sustainable, straightforward, and scalable direct upcycling strategy to efficiently convert lithium extraction by-products: FePO₄ and MnO₂ from spent lithium iron phosphate (S-LFP) and spent lithium manganese oxide (S-LMO) into high-performance lithium manganese iron phosphate (R-LMFP) cathode materials. The resultant R-LM₃F₇P cathode material exhibits outstanding electrochemical performance. Specifically, it delivers high capacities up to 164.44 mAh g⁻¹ at 0.1 C and 121.73 mAh g⁻¹ at 10 C and retains 97.46% of its initial capacity after 500 cycles at 1 C. Furthermore, the pouch cells prepared with this cathode material achieve a capacity retention of 87.93% after 600 cycles. Furthermore, R-LM₃F₇P demonstrates superior low-temperature performance to R-LFP. Additionally, based on economic analysis, the proposed upcycling strategy yields a profit of $5.57 per kg, significantly surpassing the profits from direct recycling ($1.68 per kg) and downgrading recycling methods ($0.8 per kg). This upcycling strategy offers a value-added solution for managing lithium extraction by-products and provides a practical, scalable framework for recycling waste lithium-ion batteries (LIBs). It promotes the sustainable development of LIBs by successfully tackling environmental and economic challenges in LIB recycling.

Author contributions

Fan Xiao: conceptualization, methodology, investigation and original draft writing. Lehan Zhu and Zhangjun Wu: investigation, characterization, data curation, and software. Haotian Zhu, Juan Xia and Jiannan Zhu: methodology, investigation and resources. Zeheng Yang and Weixin Zhang: conceptualization, methodology, supervision, and writing – review and editing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Additional spectral data, crystallographic data, computational details, etc. are provided in the supplementary information offered in this article. See DOI: https://doi.org/10.1039/d5gc03642a.

Acknowledgements

The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (NSFC Grant No. 22278107). This work is also supported by the Science and Technology Key Project of Anhui Province (No. 2022e03020004 and 202423h08050005) and the Collaborative Innovation Project of Colleges and Universities of Anhui Province (GXXT-2023-097).

References

1. G. Harper, R. Sommerville and E. Kendrick, et al., Recycling lithium-ion batteries from electric vehicles, Nature, 2019, 575, 75–86.
2. H. C. Ji, J. X. Wang and J. Ma, et al., Fundamentals, status and challenges of direct recycling technologies for lithium ion batteries, Chem. Soc. Rev., 2023, 52, 8194–8244.
3. Y. Q. Tao, C. D. Rahn and L. A. Archer, et al., Second life and recycling: Energy and environmental sustainability perspectives for high-performance lithium-ion batteries, Sci. Adv., 2021, 7, 16.
4. Y. K. Li, W. G. Lv and H. L. Huang, et al., Recycling of spent lithium-ion batteries in view of green chemistry, Green Chem., 2021, 23, 6139–6171.
5. A. Manthiram, A reflection on lithium-ion battery cathode chemistry, Nat. Commun., 2020, 11, 1550.
6. J. X. Wang, J. Ma and K. Jia, et al., Efficient extraction of lithium from anode for direct regeneration of cathode materials of spent Li-ion batteries, ACS Energy Lett., 2022, 7, 2816–2824.
7. X. G. Yang, T. Liu and C. Y. Wang, Thermally modulated lithium iron phosphate batteries for mass-market electric vehicles, Nat. Energy, 2021, 6, 176–185.
8. C. Y. Zhu, J. X. Liu and X. H. Yu, et al., Boosting the stable Li storage performance in one-dimensional LiLa_xMn_2-xO₄ nanorods at elevated temperature, Ceram. Int., 2019, 45, 19351–19359.
9. E. Fan, L. Li and Z. P. Wang, et al., Sustainable recycling technology for Li-ion batteries and beyond: Challenges and future prospects, Chem. Rev., 2020, 120, 7020–7063.
10. J. Yu, K. Huang and J. Zheng, et al., Advance technology for treatment and recycling of electrolyte and organic matters from spent lithium-ion battery, Curr. Opin. Green Sustainable Chem., 2024, 47, 100914.
11. M. Li, J. Lu and Z. W. Chen, et al., 30 years of lithium–ion batteries, Adv. Mater., 2018, 30, 24.
12. X. X. Zhang, L. Li and E. Fan, et al., Toward sustainable and systematic recycling of spent rechargeable batteries, Chem. Soc. Rev., 2018, 47, 65.
13. F. Jiang, K. Qu and M. S. Wang, et al., Atomic scale insight into the fundamental mechanism of Mn doped LiFePO₄, Sustainable Energy Fuels, 2020, 4, 2741–2751.
14. H. L. Zeng, Y. X. Wan and S. Z. Niu, et al., Lithium manganese iron phosphate materials: Design, progress, and challenges, Energy Mater. Devices, 2025, 3, 9370060.
15. C. Yang, J. L. Zhang and Q. K. Jing, et al., Recovery and regeneration of LiFePO₄ from spent lithium-ion batteries via a novel pretreatment process, Int. J. Miner., Metall. Mater., 2021, 28, 1478–1487.
16. Y. L. Zhang, X. D. Li and X. P. Chen, et al., Tunnel manganese oxides prepared using recovered LiMn₂O₄ from spent lithium-ion batteries: Co adsorption behavior and mechanism, J. Hazard. Mater., 2021, 425, 127957.
17. Z. T. Zhang, R. H. Lu and T. Y. Li, et al., Green and sustainable recycling of spent lithium batteries: Synergistic leaching of SLFP and SLMO for valuable metal extraction and environmental benefits, Green Chem., 2025, 27, 4688–4705.
18. J. Zhang, S. H. Luo and L. L. Sui, et al., Preparation of neodymium–doped LiMnPO₄/C cathode by sol–gel method with excellent electrochemical performance for lithium–ion batteries, Int. J. Energy Res., 2021, 45, 10590–10598.
19. J. Zhou, C. Xing and J. Huang, et al., Direct upcycling of leached FePO4 from spent lithium-ion batteries toward gradient-doped LiMn_xFe_1−xPO₄ cathode material, Adv. Energy Mater., 2023, 14, 2302761.
20. Y. R. Mei, R. Chen and Z. Shao, et al., Novel upcycling of mixed spent cathodes toward high energy density LiMn_xFe_1−xPO₄ cathode material, Adv. Funct. Mater., 2025, 2507185.
21. P. Wang, J. Zhou and J. W. Huang, et al., Unified upcycling of degraded LiFePO₄ materials toward a high-performance LiMn_0.25Fe_0.75PO₄ cathode for high-voltage Lithium-ion batteries, ACS Sustainable Chem. Eng., 2025, 13, 12647–12657.
22. Y. X. Zou, J. W. Cao and H. Li, et al., Large-scale direct regeneration of LiFePO₄@C based on spray drying, Ind. Chem. Mater., 2022, 1, 254–261.
23. A. Shchukarev and D. Korolkov, XPS Study of group IA carbonates, Open Chem., 2004, 2, 347–362.
24. E. H. Xu, T. Wang and J. X. Chen, et al., Stress-induced anomalous lithiation plateau of LiFe_yMn_1−yPO₄ over high-rate discharging, Adv. Energy Mater., 2024, 15, 2404929.
25. Y. K. Hou, G. L. Pan and Y. Y. Sun, et al., LiMn_0.8Fe_0.2PO₄/carbon nanoribbons prepared by the biomineralization process as the cathode for lithium-ion batteries, ACS Appl. Mater. Interfaces, 2018, 10, 16500–16510.
26. C. X. Xing, H. R. Da and P. Yang, et al., Aluminum impurity from current collectors reactivates degraded NCM cathode materials toward superior electrochemical performance, ACS Nano, 2023, 17, 3194–3203.
27. X. X. Zhao, X. T. Wang and J. Z. Guo, et al., Dynamic Li+ capture through ligand-chain interaction for the regeneration of depleted LiFePO₄ cathode, Adv. Mater., 2024, 36, 2308927.
28. G. J. Ji, D. Tang and J. X. Wang, et al., Sustainable upcycling of mixed spent cathodes to a high-voltage polyanionic cathode material, Nat. Commun., 2024, 15, 4086.
29. B. B. Chen, M. Liu and S. Cao, et al., Regeneration and performance of LiFePO₄ with Li₂CO₃ and FePO₄ as raw materials recovered from spent LiFePO₄ batteries, Mater. Chem. Phys., 2022, 279, 125750.
30. D. B. Ravnsbæk, K. Xiang and W. T. Xing, et al., Engineering the transformation strain in LiMn_yFe_1–yPO₄ olivines for ultrahigh rate battery cathodes, Nano Lett., 2016, 16, 2375–2380.
31. Y. Atsuo, K. Yoshihiro and K. Y. Liu, Reaction mechanism of the olivine-type Li_x(Mn_0.6Fe_0.4)PO₄ (0≤x≤1), J. Electrochem. Soc., 2001, 148, A747.
32. Q. Hu, L. Wang and G. M. Han, et al., Revealing the voltage decay of LiMn_0.7Fe_0.3PO₄ cathodes over cycling, Nano Energy, 2024, 123, 109422.
33. S. Z. Li, H. Zhang and Y. Liu, et al., Comprehensive understanding of structure transition in LiMn_yFe_1−yPO₄ during delithiation/lithiation, Adv. Funct. Mater., 2023, 34, 2310057.
34. K. Taehoon, O. Luis and K. Y. B. Qi, Understanding the active formation of a cathode–electrolyte interphase (CEI) layer with energy level band bending for lithium-ion batteries, J. Mater. Chem. A, 2023, 11, 221–231.

---