Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 May 22;65(29):e7687036. doi: 10.1002/anie.7687036

Unlocking Low‐Energy‐Barrier Pathways via Monoclinic Acceptor Induction for Homogeneous Regeneration of Spent LiMn x Fe1‐ x PO4 Cathodes

Zihao Zeng 1, Hai Lei 1, Chao Zhu 1, Yunpeng Wen 1, Xizhuo Chen 1, Shengming Xu 2, Wei Sun 1, Peng Ge 1, Yue Yang 1,3,✉
PMCID: PMC13360312  PMID: 42171420

ABSTRACT

As the next‐generation olivine‐type cathode, high energy‐density LiMn x Fe1‐ x PO4 (LMFP) would reach end‐of‐life and produces numerous retired materials, thus exploring its effective recycling technology is urgent. However, suffered from the differences in oxidation energy barriers, synchronous regulation of Mn/Fe phases could be hardly achieved in direct regeneration, resulting in spent LMFP difficult to be repaired. For solving the problems above, a monoclinic lattice‐induction strategy is proposed. Supported by introducing monoclinic‐structured Li3Fe2(PO4)3 as acceptors, the reaction energy of phase transition can be effectively reduced, finally alleviating the “rivet” effect of Mn. Benefiting from effort above, atomic‐level homogeneity of Fe/Mn distribution is achieved in regenerated samples, along with the alleviation of strain‐stress concentration. Specially, the lower ratio of anti‐sites defects of repaired LMFP is controlled, accelerating the Li‐diffusion along (010) direction. Importantly, Mn─O bonds are reinforced, suppressing Mn‐dissolving behaviors and improving the structural stability. As lithium‐storage cathode, the as‐optimized sample displays a capacity of 145.5 mAh g−1 at 1.0 C, even achieving 91.61% capacity retention ratio after 1500 loops. Given this, monoclinic lattice‐induction strategy is expected to provide significant guidance for large‐scale LMFP recycling.

Keywords: electrochemical performance, regeneration, rivet effect, spent LiMn x Fe1‐iPO4

For recycling spent LiMn x Fe1‐ x PO4, a monoclinic Li3Fe2(PO4)3 acceptor is utilized to alleviate the Mn “rivet” affect, lowering down the phase transition energy barriers. Benefiting from the evenly distribution of Mn/Fe atoms, the regenerated LiMn x Fe1‐ x PO4 cathode delivers a high capacity of 145.5 mAh g−1 at 1C with 91.61% retention after 1500 cycles.

graphic file with name ANIE-65-e7687036-g002.jpg

1. Introduction

For meeting the goal of “carbon neutrality” and “carbon peak dioxide emissions”, lithium‐ions batteries (LIBs) have attracted extensive research attention and achieved remarkable progress. As a key component of LIBs, the cathode material directly determines their overall energy storage performance, promoting continuous exploration of next‐generation cathode technologies [1, 2]. Among existing cathodes, lithium manganese iron phosphate (LiMn x Fe1‐ x PO4, LMFP) was regarded as a promising next‐generation cathode due to its superior properties [2, 3]. By introducing Mn into the olivine structure, a higher working voltage platform of 3.6–3.7 V could be achieved, significantly surpassing that of traditional lithium iron phosphate (LFP). Benefiting from elevated voltage, energy density of LMFP could be substantially improved, strongly promoted the applications for large‐scale energy storage [4]. Moreover, LMFP retains the inherent safety and cycling stability characteristic of olivine‐type materials, ensuring stable long‐term performance [5]. Afore‐mentioned merits, combined with competitive cost and environmental‐friendly merits, prompting LMFP become a candidate for the new generation cathode material of power batteries [6].

Given the limited‐service life (typically 5–10 years) of LIBs, LMFP would inevitably be retired in large quantities in the near future [7, 8, 9]. Restricted by the low value of Fe and Mn elements, the value of LMFP was lower than other cathode materials. Thus, conventional hydrometallurgical and pyrometallurgical recycling method in straggled in extensive processing, high cost, and severe pollution methods, which will hardly meet the demand of low‐carbon and economic recycling standards in future [10, 11]. In contrast, direct regeneration strategies could repair the cathode without structure‐destroy, displayed considerable economy, and environmental‐friendly traits [12]. However, present LMFP regeneration still suffered from existing challenge. As a solid solution cathode, the introducing of Mn‐containing phases makes LMFP a complex solid solution of Fe and Mn [13]. For diverse bonds energy of Mn─O and Fe─O bonds [14], different phase conversion would be triggered in regeneration process, especially in the high‐temperature sintering process, making it difficult to achieve synchronous regulation of Mn and Fe during regeneration [15]. Therefore, a new strategy that specifically overcomes this Mn‐induced uneven distribution of Fe/Mn phase is urgently required.

In solid‐state chemistry, crystal phase inter‐diffusion could be promoted between two phases with similar crystal framework, which has been widely investigated in the synthesis of phosphate cathode materials [16]. Li3Fe2(PO4)3, as a compounds with monoclinic structure, possessed a similar framework to that of targeted oxidized product about LMFP [17, 18]. Besides, all elements contained in the Li3Fe2(PO4)3 intermediate are also existed in LMFP. Beneficial from the structural and elemental similarity, Li3Fe2(PO4)3 was expected to lower down the Mn diffusion barrier without introducing any extra impurity elements, leading to the regeneration of spent LMFP with high performance. Herein, Mn “rivet” affect during regeneration process was revealed. Based on the “rivet” affect an oxidative regulation strategy is proposed, converting LMFP into a monoclinic phase via the O2‐supported reactions, along with regulation of Mn/Fe phases. However, suffered from high phase conversion energy barrier, Mn‐rich phase could be hardly oxidized, thus the formation of Mn‐rich monoclinic phase could be hardly triggered. Owing to the low energy barrier for Fe migration, it is expected that Fe ions can diffuse readily away from the olivine LFP lattice, suggesting that the high phase conversion energy barrier in LMFP is intrinsically related to the introduction of Mn ions. Thus, extracting Mn atoms from LMFP lattice through solid‐state reaction is expected to solve the rivet effect, promoting a higher proportion of LMFP to transform into uniform monoclinic intermediates. Inspired by this, monoclinic Li3Fe2(PO4)3 was introduced into oxidation process, which serve as an acceptor to extract Mn atoms from LMFP. Benefiting from the reducing of oxidative reaction energy barrier, LMFP completely converted into evenly monoclinic phase, achieving the synchronous regulation of Mn/Fe phase. Owing to the Mn/Fe synchronous regulation, as‐regenerated samples displayed a considerable discharging capacity of 145.5 mAh g−1 at 1.0 C (1.0 C = 170 mAh g−1), maintaining 133.3 mAh g−1 after 1500 loops with 91.61% capacity retention ratio.

2. Results and Discussion

2.1. Revealing Mn “Rivet” Affect in LMFP Phase Conversion

Due to the introduction of Mn atoms in LMFP, experimental validation is necessary to determine the two‐phase transformation mechanism in LMFP. Figure 1a illustrated the XRD patterns of LMFP (commercial LiFe0.4Mn0.6PO4) and HR‐LMFP (commercial LiFe0.4Mn0.6PO4 which treated with 500°C sintered in atmosphere). After sintering process, it could be noted that peaks belong to olivine structure shifted into higher degree, indicating the main framework was maintained in the HR‐LMFP. Figure 1b figured out the reason of the peak shifting behaviors. After sintering, part of Fe ions would diffuse away from the lattice of LMFP, resulting in the formation of Fe2O3 peak located in 32.4° (Figure 1b). From the whole XRD patterns of HR‐LMFP in Figure S1, peaks of manganese oxide could be hardly noted, indicating that Mn could not reacted with O2 at sintering process. Critically, unlike pure‐phase LiFePO4, which decomposes extensively into Fe2O3 and Li3Fe2(PO4)3 under similar conditions, the amount of Fe2O3 formed from LMFP was less, and the olivine structure of LMFP remained intact. This direct comparison provides compelling evidence that Mn ions within the lattice act as pivotal “rivets,” effectively inhibiting the high‐temperature diffusion of Fe ions from the bulk to the surface and thereby anchoring the olivine structural framework (Figure 1c,d).

FIGURE 1.

FIGURE 1

Open in a new tab

XRD patterns of LMFP and H‐LMFP (LMFP headed in 500°C for 2 h): (a)22°–26° and (b)30°–35°. Illustration of oxidation reaction of different cathode: (c) LiFePO4 and (d) LiFe0.5Mn0.5PO4. DFT calculations on (e) the formation energy and (f) reaction energy about 10 models (including LiFePO4 and Li3Fe2(PO4)3, LiMn0.25Fe0.75PO4 and Li3Fe1.5Mn0.5(PO4)3, LiMn0.5Fe0.5PO4 and Li3FeMn(PO4)3, LiMn0.75Fe0.25PO4 and Li3Fe0.5Mn1.5(PO4)3, LiMnPO4 and Li3Mn2(PO4)3). (g) Contour maps of in situ XRD of LiFePO4 and LiFe0.5Mn0.5PO4 mixture in sintering process.

Based on this mechanism, it should be recognized that during subsequent carbothermal reduction, the Fe ions expelled from LMFP could not be easily reduced and re‐inserted into their original lattice sites, which was a feasible process for pure LiFePO4. Thus, before regenerating spent LMFP process, it is necessary to alleviate the Mn “rivet” affect through chemical manners. Considering that Mn ions introducing would bring about Mn “rivet,” detailed DFT calculations were conducted to reveal the relationship between Mn content and the strength of the “rivet” effect.

Figure 1e,f evaluated the formation energy of 10 structures, including 5 olivine structures (such as LiFePO4, LiMn0.25Fe0.75PO4, LiMn0.5Fe0.5PO4, LiMn0.75Fe0.25PO4, LiMnPO4) and 5 monoclinic structures (such as Li3Fe2(PO4)3, Li3Fe1.5Mn0.5(PO4)3, Li3FeMn(PO4)3, Li3Fe0.5Mn1.5(PO4)3, Li3Mn2(PO4)3). The five olivine and five monoclinic structures above correspond one‐to‐one, and the olivine structure can react with oxygen to generate its corresponding monoclinic structure. Therefore, five reactions were existed in these ten structures (Figure 1e), which was further illustrated in Table S1. By calculating the formation energy of each material, the reaction energy barriers for the five reactions could be obtained. It could be noted that, the formation energies of LiFePO4, LiMn0.25Fe0.75PO4, LiMn0.5Fe0.5PO4, LiMn0.75Fe0.25PO4, and LiMnPO4 were −1.963, −1.933, −1.903, −2.443, and −2.146 eV/atom, respectively, and, the formation energies of Li3Fe2(PO4)3, Li3Fe1.5Mn0.5(PO4)3, Li3FeMn(PO4)3, Li3Fe0.5Mn1.5(PO4)3, and Li3Mn2(PO4)3 were −2.608, −2.095, −2.124, −2.154, and −2.187 eV atom−1, respectively. Based on the data above, the reaction energy of five reactions could be evaluated in Figure 1f. When the Mn content increases from 0% to 50%, the reaction energy barrier for the transition from olivine structure to monoclinic structure decreases (from 3.2374 to 3.1938, and finally decreased into 3.1490 eV atom−1), indicating that introducing of low content of Mn atoms will be beneficial for structural transformation. However, when the Mn content exceeded 50%, the energy barrier of the reaction increases significantly. Interestingly, when all iron atoms are replaced by manganese atoms, the energy barrier of the reaction decreases again, reaching the same level as pure LFP. This indicates that the previously mentioned theory regarding the “rivet” effect of Mn atoms in suppressing structural transformation can be further refined. At low concentrations, Mn atoms may not exhibit the so‐called “rivet” effect. The “rivet” effect comes into play only when the Mn content ranges between 50% and 100%, effectively inhibiting the transformation of the olivine cathode material into a monoclinic structure. Combining our experimental data, the Mn content of spent LMFP could up to 60%, which falls within the region where the riveting effect is most pronounced. As a result, during the oxidation process, LMFP material does not fully transform into the monoclinic phase and largely retains its original olivine structure. For further confirming the “rivet” effect of Mn ions, in situ XRD was conducted on LMFP which sintered from 25°C to 500°C. During oxidative heating, if Fe atoms diffuse away from olivine lattice, pronounced shifting behaviors of diffraction peaks (located at approximately 12.5° and 22.5°) would be triggered, which could be ascribed to the extensive extraction of Fe. As noted in Figure S2, compared to the LFP [16], peaks of pure LMFP shift more slowly than of LFP, indicating that Fe migration is significantly hindered in the Mn‐containing system. Therefore, during the LMFP regeneration process, it is necessary to tailoring the Mn/Fe ratio in LMFP to overcome the “rivet” effect of Mn atoms and achieve the transformation from the olivine to the monoclinic structure. Based on our computational findings, when the Mn/Fe ratio is 1:1, the energy barrier for the transformation from the olivine to the monoclinic structure is minimized. Therefore, it is necessary to add iron‐containing materials. Owing to the similar olivine structure, LFP was deemed as the optimal material for this purpose.

Therefore, mixture of LFP and LMFP was treated with 500°C in air, and the in situ XRD contour maps in Figure 1g illustrated the phase conversion of LMFP/LFP mixture. At stage 1 (300–350°C), peaks of samples shifted into a higher degree, indicating that LFP was reacted with O2 to form Li3Fe2(PO4)3, then during stage 2 (350–500°C), peaks of Li3Fe2(PO4)3 were shifted into a lower degree, indicating that Mn ions were diffused into Li3Fe2(PO4)3. Meanwhile, more Mn ions participated in forming monoclinic structure, indicating that extra added LFP could induced the phase conversion from olivine to monoclinic structure, which matched well with the results of DFT calculations in Figure 1e. Furthermore, for reaction products (HR‐ LMFP/LFP) of LMFP/LFP mixture with O2 at 500°C, SEM was conducted to observed the elements distribution. As shown in the Figure S3, Mn is uniformly distributed throughout the material, both in regions where particles are individually dispersed and where they are aggregated. This indicates that during the initial oxidation process, the presence of LFP induces intense solid‐phase diffusion of Mn in LMFP, resulting in the formation of an increased amount of the monoclinic phase. To further verify that the monoclinic intermediate facilitates cation interdiffusion, a diffusion experiment was conducted. Oxidized LFP and oxidized LMFP powders were separately pressed into pellets and further pressed into a combined pellet, followed by heat treatment at 500°C for 2 h. Cross sectional elemental analysis of the resulting interface revealed clear interdiffusion of Fe and Mn across the boundary in Figure S4. The migration of Mn was driven by the steep Mn concentration gradient established between the Mn‐rich side (LMFP) and the Mn‐free side (LFP), providing direct evidence that the monoclinic Li3Fe2(PO4)3 served as an effective acceptor for receive the Mn ions from the LMFP.

2.2. Direct Regeneration Strategy for Spent LMFP Recycling

Based on a detailed investigation of the oxidation‐induced phase transformation mechanism in LMFP, we have correspondingly developed an efficient monoclinic lattice‐induction strategy to recovery LMFP. Herein, spent LFP (S‐LFP) and spent LFP black (S‐LMFP) powder were collected from electrode sheets. As noted in Figure 2a, for S‐LFP and S‐LMFP, several types of impurity carbon could be noted, including binder agent (PVDF), conductive agent (carbon black) and amorphous carbon‐coated layer. For regenerating S‐LMFP and tailoring the carbon content at the same time, three methods were proposed and the resulted samples were detailed compared in this work. As shown in Figure 2b1, for removing the amorphous carbon‐coated layer and promoting the solid solubility of LFP and LMFP phase, the sintering process was utilized at the first stage with the assistance of O2, along with the lattice homogenization of LFP and LMFP. Later, supported by the protection of Ar/H2, olivine‐structure lattice would be reconstructed, meanwhile LFP/LMFP phase would be mutual integrated, finally regenerating LMFP (OR‐LMFP/LFP) with the tailoring of elements‐ratio. In regeneration route 2 (Figure 2b2), mixed S‐LFP/LMFP powder was directly sintering without oxidation process, and in regeneration route 3 (Figure 2b3), only S‐LMFP powder was regenerated with O2‐supported sintering and reduction process without introducing of S‐LFP. The samples originated from regeneration route 2 and 3 were named as DR‐LMFP/LFP and OR‐LMFP, respectively. Spent LMFP cathode was obtained from the spent cathode sheets from industrial (Figures 2c and S5), thus binder and conductive agent would be inevitable introduced. Benefiting from the O2‐supported sintering process, the removing process of binder and conductive agent could be observed in the SEM images (Figure 2d1–4 and Figures S6), resulting in the nano‐scale LMFP particles formation [17]. All the elements, including Li, Fe, P, C, and O, could be detected by XPS full‐spectra (Figure S7), indicating that no impurity was introduced during regeneration process. Based on the regeneration routes above, series of samples could be synthesized and characterized by x‐ray diffraction (XRD) and ICP‐OES. All the regenerated samples were coincided well with LMFP standard PDF card (01‐077‐0178) in Figure S8. It should be noted that, the peaks of DR‐LMFP/LFP moved towards higher degree, indicating that S‐LFP could be hardly integrated with S‐LMFP phase. With the introducing of S‐LFP, the Mn/Fe ratio of LMFP was successfully tailored from 6:4 to 5:5 (Figure 2e). Pre‐oxidation process could effectively remove the carbon layer and conductivity agent, leading to the evolution of D‐peak and G‐peak in Figure S9. Meanwhile, supported by the lattice transformation‐reconstruction process, their particle size could be effectively lowered (Figure S10), resulting in the similar uniform size‐distribution. Furthermore, based on the FTIR results (Figure S11), Li–Fe anti‐sites and Li–Mn anti‐sites could be evaluated, which was related to the [PO4] peak in the 900–1000 cm−1 [19]. According to Figure 2f, the wavenumbers of [PO4] were in order of S‐LMFP >DR‐LMFP/LFP >OR‐LMFP >OR‐LMFP/LFP. It could be concluded that lattice reconstruction process and extra Fe ions introducing could rational eliminate part of Li–Fe anti‐sites and Li–Mn anti‐sites for LMFP, which was conductive to improving the Li‐diffusion rate for (010) direction and finally improving the energy storage capabilities (Figure 2g) [20].

FIGURE 2.

FIGURE 2

Open in a new tab

(a) Illustration of S‐LFP and S‐LMFP. Diverse regeneration routes of LMFP: (b1) sintering‐reduction route of S‐LFP/S‐LMFP mixture (OR‐LMFP/LFP), (b2) direct‐sintering route of S‐LFP/S‐LMFP mixture (DR‐LMFP/LFP), (b3) sintering‐reduction route of S‐LMFP (OR‐LMFP). (c) Spent LMFP electrode sheets utilized in this work. SEM images of: (d1) S‐LFP, (d2) S‐LMFP, (d3) H‐LMFP/LFP, and (d4) OR‐LMFP/LFP (with mapping images). (e) ICP results of samples in this work. (f) Wavenumber of [PO4] for diverse samples. (g) illustration of Li‐Fe anti‐sites elimination.

2.3. Reaction Mechanism of Regeneration Process

For probing the differences between the as‐obtained LMFP samples, XRD Refinement have been carried in Figure 3a1,2. Considering that two phase was co‐existence in DR‐LMFP/LFP, LiM0.6Fe0.4PO4 and LiFePO4 standard cards were both utilized, while only LiM0.5Fe0.5PO4 /LiM0.6Fe0.4PO4 standard cards was used in other samples. Diverse regeneration methods resulted in the different cell parameters and volume of as‐regenerated samples (Table S2), which were in sequence of OR‐LMFP/LFP >DR‐LMFP/LFP >OR‐LMFP. Owing to the LMFP/LFP phase integrating and Fe atoms introducing behaviors, the cell parameters and volume of OR‐LMFP/LFP were obviously lower than that of other samples. It could be concluded that, spent LFP and LMFP mutually fused into a new LMFP phase. Attracted by the evolution of sample‐parameters, the regeneration mechanism of OR‐LMFP/LFP was investigated through SEM, TG and in situ XRD. For promoting the lattice reconstruction process, carbon source was mixed with H‐LMFP/LFP (LMFP/LFP mixture heated in 500°C for 2 h) to obtain the precursor. Among the sintering process under, glucose could be turned into small molecules with reduction capability (such as H2 and CO), promoting to reconstruct the lattice of LMFP. Given this, TG was performed to investigate the weight losing of H‐LMFP/LFP and glucose mixture. As shown in Figure 3c, the weight loss behaviors could be divided into two stages: 150°C–350°C and 350°C–700°C. For probing the phase evolution at different stage, in situ XRD was conducted in Figure 3d, which revealed the phase evolution during the regeneration process. As noted in Figure 3e1–3, peaks belong to Fe2O3 could be observed during 25°C–350°C sintering process, while it was vanished during 350°C–700°C sintering process. Thus, it could be concluded that, Fe2O3 would be generated during the first oxidation step, then it was reacted with the carbon source and later fused into the LMFP lattice. Furthermore, as shown in Figure 3e4, the intensity of peaks belongs to LMFP became stronger, revealing that crystal structure is continuously improved and turned into more stable during high‐temperature calcination process. Based on the analysis above, the reaction could be summarized with following equation (1). For evaluating the fusing results of regeneration process, time‐of‐flight secondary ion mass spectrometry (TOF‐SIMS) was carried out (Figure S12). Based on the negative ions model (Table S3), 2D/3D distribution of elements could be detailed analysis. As shown in (Figures 3f1–3, 2g, and S13), the spatial distribution of three key elements (including Li, Fe, and Mn) was analyzed from 2D/3D and perspectives, indicating that the three materials in the recycled material are evenly distributed without any aggregation behavior.

LiFePO4+5LiMn0.6Fe0.4PO4→6LiMn0.5Fe0.5PO4 (1)

FIGURE 3.

FIGURE 3

Open in a new tab

XRD Rietveld Refinement results of: (a1) OR‐LMFP/LFP, (a2) DR‐LMFP/LFP, and (a3) OR‐LMFP. (b1–b6) SEM images of precursors under different sintering temperatures. (c) TG and DSC curves of regeneration process. (d) In situ XRD contour maps for regeneration process. (e1–e3) Intensity evolution of Fe2O3 peak during regeneration process. (e4) Intensity evolution of LiFeMnPO4 peak during regeneration process. (f1–f3) 3D imaging analysis of Li+, Fe2+, and Mn2+ ions for OR‐LMFP/LFP. (f1–f3) 2D imaging analysis of Fe2+ and Mn2+ ions for OR‐LMFP/LFP.

Detailed structure characterization was carried out in Figure 4. Focused ion beam‐scanning electron microscope (FIB‐SEM) was utilized to probe the interior structure differences of OR‐LMFP/LFP and DR‐LMFP/LFP. As shown in Figure 4a, the elemental distribution was analyzed along straight line of the section. The Fe and Mn concentrations are consistent and evenly distributed along straight line, further indicating that spent LFP have been introduced into the lattice of spent LMFP. However, for DR‐LMFP/LFP, the particles of spent LFP could be captured in the Figure 4b. Mn elements could be hardly detected by FIB‐SEM, and Fe content of edge region was lower than that of the inner region, further revealing that spent LFP and spent LMFP are unable to form solid solutions through direct sintering. Mapping images of SEM/TEM (Figure 4c1–4, d1–4) further proved the phenomenon mentioned above. Furthermore, SEM‐mapping images of HR‐LMFP/LFP (Figure S14) also demonstrating that uneven‐distribution behaviors of Fe and Mn, indicating that the atomic‐level uniform distributed solid solution was achieved in subsequent reduction process. For OR‐LMFP, SEM‐mapping images in Figure S15 proved that Fe and Mn were uniformly distributed, derived from no extra LFP introducing.

FIGURE 4.

FIGURE 4

Open in a new tab

FIB‐SEM images of OR‐LMFP/LFP and DR‐LMFP/LFP. (c1, c2) SEM‐mapping images of DR‐LMFP/LFP. (c3, c4) SEM‐mapping images of OR‐LMFP/LFP. (d1–d4) TEM‐mapping images of OR‐LMFP/LFP. TEM images of (e) OR‐LMFP/LFP, (f) DR‐LMFP/LFP, and (g) OR‐LMFP. (h) Elements content analysis of OR‐LMFP/LFP from surface to bulk. (i1, j1) STEM‐HRTEM images of OR‐LMFP/LFP, (i2, j2) corresponding crustal structure of OR‐LMFP/LFP. Stress distribution of OR‐LMFP/LFP and OR‐LMFP: (k) strain‐e xy for OR‐LMFP/LFP, (l) strain‐e yy for OR‐LMFP/LFP, (m) strain‐e xy for OR‐LMFP, (n) strain‐e yy for OR‐LMFP.

Supported by the TEM in Figure S16, it could be disclosed that the particle size of all three samples was in 100‐nanometer range. Figure 4e–g illustrated the HRTEM images of regenerated samples, all the samples displayed good crystallinity that is to say, the differences between the samples mainly focused on the elemental distributions described above. Figure S17 investigated the plane spacing of OR‐LMFP/LFP, and the as‐observed 0.2379 nm was related to the exposure of (041) plane. HRTEM further revealed that R‐LMFP particles are uniformly encapsulated by ultrathin carbon layer (1–3 nm). More importantly, for OR‐LMFP/LFP, the vital elements were uniformly distributed from surface to bulk, indicating that LFP phase have been evenly introduced into the interior of as‐regenerated samples (Figure 4h). Attracted by the uniform distribution behaviors of OR‐LMFP/LFP, AC‐TEM was carried out to explore the nano‐scale structure of OR‐LMFP/LFP. As illustrated in Figure 4i1, [010] crystal planes could be detected, accompanying with regular arrangement of Fe/Mn atoms, especially in Figure 4i2. Furthermore, [100] crystal planes could be noted, proving that the atoms of OR‐LMFP/LFP was regularly arranged along different crystal orientations, displaying good crystallinity. In the high angle annular dark field images, the intensity of atomic columns scales approximately with the square of the atomic number. Since Fe possesses a slightly higher atomic number than Mn, Fe rich atomic columns appear somewhat brighter while Mn rich columns appear comparatively darker. Thus, as noted in Figure S18, the evenly distribution of Mn/Fe could be observed from the light area (Fe) and dark area (Mn), confirming the atomic‐level uniform distribution of Mn/Fe atoms. Importantly, for evaluating the inner differences about OR‐LMFP/LFP and OR‐LMFP, their position of strain distribution was analyzed from geometric phase analysis (GPA). Both for strain‐e xy and strain‐e yy , the strain distribution of OR‐LMFP/LFP was more uniform than that of OR‐LMFP. It could be concluded that, assisted by well‐designed regeneration strategy, their stress concentration could be effectively alleviated, contributing to the improvement of structural stability.

2.4. Electrochemical Properties of As‐regenerated Samples

Figure 5a illustrated the 10th charging/discharging curves of as‐obtained samples. The discharge capacity of OR‐LMFP/LFP, DR‐LMFP/LFP, and OR‐LMFP were 144.2 111.2 and 99.7 mAh g−1 at 1.0 C, respectively. Besides, the platform from Mn‐redox was shortening about DR‐LMFP/LFP and OR‐LMFP. Figure 5b showed their middle voltage from 0 to 100 cycles, where DR‐LMFP/LFP and OR‐LMFP displayed middle voltage (3.0–3.3 V) for 0th ∼ 100th cycles. In comparation, that of OR‐LMFP/LFP increased up to ∼3.58 V. Thus, the higher energy density was expected for OR‐LMFP/LFP than that of DR‐LMFP/LFP and OR‐LMFP. Besides, homogeneous distribution of Fe and Mn elements could bring about considerable rate performance for OR‐LMFP/LFP. As shown in Figure 5c, OR‐LMFP/LFP displayed a discharging capacity of 155.6, 141.9, 119.7, and 107.6 mAh g−1 at 0.5, 1.0, 2.0, 5.0 C, greater than that of DR‐LMFP/LFP (142.1, 128.6, 87.9, and 19.7 mAh g−1) and OR‐LMFP (142.1, 128.6, 87.9, and 19.7). Besides, for DR‐LMFP/LFP, the stable cycling performance was displayed, achieving 93.3 % capacity retention ratio even over 200 cycles at 0.5 C (Figure 5d). However, that of DR‐LMFP/LFP and OR‐LMFP were only kept about 68.57% and 25.92%, accompanying with the serious capacity fading.

FIGURE 5.

FIGURE 5

Open in a new tab

(a) Charging and discharging curves (10th) at 1.0 C for as‐regenerated samples. (b) Comparison of the middle voltages for as‐regenerated samples. (c) Rate performance of as‐regenerated samples. (d) Cycling performance of as‐regenerated samples at 0.5 C. (e) Total density of state of LM5F5P and LM6F4P model. (f) Comparison of difference value of total d‐band center and O p‐band center. (g) The pristine Nyquist plots of as‐regenerated samples. (h) In situ x‐ray diffraction XRD counter map of OR‐LMFP/LFP and OR‐LMFP. (i) Long‐term cycling performance of as‐regenerated samples at 1.0 C. (j) Comparison of the energy density of LMFP (OR‐LMFP/LFP), commercial LiMnO2 (LMO), LiFePO4 (LFP), and LiCoO2 (LCO). (k) Illustration of Mn dissolvement behaviors in LM5F5P and LM6F4P model. (l) Comparison of Mn‐vacancy and O‐vacancy formation energy of LM5F5P and LM6F4P model.

Further probing the relationship between structure traits and energy‐storage traits, density functional theory (DFT) calculations, electrochemical impedance spectroscopy (EIS) spectra and in situ x‐ray diffraction XRD were carried out as following. For evaluating the effect of element‐ratio tailoring, Li x Fe1‐ x MnPO4 models (Figure S19) were constructed, while detailed calculations about total density of state (DOS) were performed in Figure 5e. Based on the detailed investigation about their DOS in Figure S20. In Figure 5f, their difference value of total d‐band center and O p‐band center was compared, where that of LF5M5P was lower than that of LF6M4P, indicating that extra introduced Fe ions were beneficial for improving the stability of M─O bonds (M═Fe and Mn). As shown in Figure 5g, the initial charge transfer resistances (R ct) of as‐prepared samples were in sequency of OR‐LMFP/LFP < DR‐LMFP/LFP < OR‐LMFP, indicating that OR‐LMFP/LFP possessed the faster ions/electron transferring behaviors. Furthermore, for further exploring the difference of OR‐LMFP/LFP and OR‐LMFP, in situ x ray diffraction was taken place in Figure 5h. From XRD patterns from 27.5° to 32.5°, reversible phase evolution of OR‐LMFP/LFP (LMFP‐MFP‐LMFP) could be noted [21]. However, for OR‐LMFP, irreversible phase evolution could be observed in the in situ x ray diffraction contour maps, indicating that some de‐intercalated Li ions hardly returned into the lattice of MFP. Given this, long‐term cycling test was carried out in Figure 5i, where OR‐LMFP/LFP displayed a high discharging capacity of 145.5 mAh g−1 at 1.0 C, maintaining 133.3 mAh g−1 even after 1500 loops with 91.61% capacity retention ratio. And noted in Figure S21, LMFP/graphite pouch full‐cell, it can still deliver 739.35 mAh at 1.0 C, and maintained 731.39 mAh after 50 cycles. Compared to other work on LMFP [22, 23, 24, 25, 26, 27, 28], as‐obtained OR‐LMFP/LFP displayed better cycling performance (Figure S22)

For DR‐LMFP/LFP, limited by the hardly removed carbon impurity, the relatively low capacity was displayed. About OR‐LMFP, the obvious capacity fading behaviors could be observed, indicating that the structure would be destroyed, resulted from the uncontrolled lattice stress. Thus, its cycling stabilities were less than 300 cycles at 1.0 C. Figure S23 illustrated the XRD patterns of three samples after corresponding cycles at 1.0 C. Obviously, the intensity of peak located at 18.2° (belong to the olive type MFP) was in order of OR‐LMFP/LFP < DR‐LMFP/LFP < OR‐LMFP, indicating that serious irreversible phase conversion was triggered in OR‐LMFP; and, in OR‐LMFP/LFP, the peak of olive type MFP could be hardly found, proving that regeneration process effectively improved the structural stability. Furthermore, for DR‐LMFP/LFP and OR‐LMFP samples, apparent amorphous region could be found in TEM images after corresponding cycles at 1.0C (Figure S24), indicating that Mn dissolution behavior has change part of the lattice. To further quantify the Mn dissolution behavior, EDS analysis was performed on the separators harvested from the cycled half cells. The measured Mn contents are summarized in Table S4. The OR‐LMFP cell exhibited the highest Mn deposition at 2.42 wt%, followed by DR‐LMFP/LFP at 0.82 wt%, while the OR‐LMFP/LFP cell showed the lowest Mn content at only 0.53 wt%, confirming that the OR‐LMFP/LFP sample experiences the least Mn dissolution among the three regenerated cathodes.

Moreover, dQ/dV curves were further used to prove the stability of OR‐LMFP/LFP (Figure S25). Attracted by the high capacity and considerable cycling stability of OR‐LMFP/LFP, energy density comparison of LMFP (OR‐LMFP/LFP), commercial LiMnO2 (LMO), LiFePO4 (LFP), and LiCoO2 (LCO) was displayed in Figure 5j [23]. Supported by the well‐designed regeneration process, spent LFP and LMFP could be upgraded toward new LMFP with improved average middle voltage, resulting in the improvement of energy density. As noted in Figure 5j, the energy density of OR‐LMFP/LFP was much higher than that of commercial LFP, even closing to that of commercial LCO. For Mn‐based cathode, the electrochemical stability was closely related to Mn/O‐ions dissolving behaviors, thus, for probing the in‐depth reason about the differences of the OR‐LMFP/LFP and OR‐LMFP, Mn/O vacancies formation energy was evaluated in Figure 5K,l. Obviously, about Mn/O vacancies, the formation energy of OR‐LMFP/LFP was higher than that of OR‐LMFP, revealing that Mn─O bond would be easily destroyed, deteriorating the stability of lattice structure.

2.5. Economy and Environmental Analysis of Regeneration Process

From spent LMFP/LFP to new LMFP, two traditional strategies have been proposed, such as hydrometallurgical (hydro‐) and pyrometallurgical (pyro‐) recycling manners [29, 30]. However, restricted by the inevitable energy consumption, high recycling cost and extra chemicals introducing, traditional manners could hardly meet the demand of “green chemistry” and “low‐carbon recycling.” In this work, the direct recycling manner was proposed, which could induce the transforming from spent LMFP/LFP to new LMFP. In Figure 6 and Tables S5–S8, the economic and environmental benefits of three recycling methods were compared. In Figure 6a, the detailed steps of hydro‐, pyro‐, and direct‐manners were compared. Compared to traditional strategies, direct‐manners displayed great large‐production potential due to its shortening steps. About their recycling cost, six main parts were collected, such as materials, energy, process, labor, maintenance, and plant overhead. Based on Everbat 2023, it could be evaluated that, the cost of processing 1 kg spent LIBs through hydro‐, pyro‐, and direct‐ was about 0.48, 0.58, and 0.41$ (Figure 6b). However, direct can achieve up to considerable revenue of 3.5$ for processing 1 kg spent LIBs, much higher than that of hydro, pyro‐ (Figure 6c). Thus, considering the lower cost and high revenue, direct recycling strategy could be up to the highest profit which was illustrated in the Figure 6d. Furthermore, for meeting the demand of “green chemistry” and “low‐carbon recycling,” the greenhouse gas and energy emission recycling manners for large‐scale production was also of great importance [31]. Compared to the traditional manners, direct‐ showed less greenhouse gas and consumed lower energy (Figure 6e). Figure 6f summarized the five key indicators for three recycling manners, and direct recycling manner displayed full merits on all the indicators, confirming that direct regeneration showed high environmental friendliness and economic benefits, accompanying with the potential for large‐scale promotion.

FIGURE 6.

FIGURE 6

Open in a new tab

(a) Detailed steps of different manners of recovering spent LFP and LMFP, including hydrometallurgical manners, pyrometallurgical manners, and upcycling manners. For the different manners of recovering spent LFP and LMFP: comparison of (b) recycling cost, (c) recycling revenue, (d) recycling profit, (e) greenhouse gas, (GHG) and energy emission, and (f) comprehensive comparison of the value mentioned above.

3. Conclusions

In summary, with assistance of monoclinic lattice‐induced strategy, spent LMFP cathode materials could be successfully regenerated. Proved by detailed DFT calculations, the phase transition (from olivine to monoclinic structure) energy of LMFP could be effective lower down from 3.6526 to 3.1490 ev atom−1 when the Mn:Fe ratio controlled towards 1:1. Guided by the DFT calculations, suitable content of spent LFP was introduced in this work, successfully regulating the Mn:Fe ratio to 1:1. Simultaneously, supported by oxidative sintering, olivine LMFP/LFP mixture was completely converted to monoclinic‐structure compounds. Importantly, Mn ions would diffuse from LMFP to Li3Fe2(PO4)3 in regeneration process, finally lowered down the oxidative energy of Mn‐rich phase. As a result, Mn “rivet” effect of spent LMFP was mitigated, achieving homogeneity of Fe/Mn distribution and alleviating strain‐stress concentration of regenerated LMFP. Compared to spent samples, the formation energy of Mn‐vacancy and O vacancy in regenerated LMFP was effectively reduced, proving well‐designed regeneration process obviously improved the structural stability. Benefiting from the synchronous regulation of Fe/Mn phase, as‐regenerated OR‐LMFP/LFP cathode exhibits a high initial capacity of 145.5 mAh g− 1 at 1.0 C and outstanding long‐term stability, with 91.61% capacity retention after 1500 cycles. Compared to the traditional recycling methods, monoclinic lattice‐induction strategy showed shortening processing steps with reduces costs, which was expected to shed light on the orientation of LMFP regeneration in the future.

Author Contributions

Zihao Zeng: conceptualization, writing – original draft, writing – review and editing, methodology. Hai Lei: software. Chao Zhu: formal analysis. Yunpeng Wen: validation. Xizhuo Chen: investigation. Shengming Xu: project administration. Wei Sun: formal analysis. Peng Ge: investigation, writing – review and editing. Yue Yang: writing – original draft, writing – review and editing, funding acquisition, conceptualization, methodology, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie72800‐sup‐0001‐SuppMat.docx.

ANIE-65-e7687036-s001.docx (8.4MB, docx)

Acknowledgments

This work was financially supported by the State Key Program of the Major Research Plan of the National Natural Science Foundation of China (No. 92575204), Sichuan Science and Technology Program (2025ZDZX0080).

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  • 1. Yoshino A., “The Birth of the Lithium‐Ion Battery,” Angewandte Chemie 51 (2012): 5798–5800, 10.1002/anie.201105006. [DOI] [PubMed] [Google Scholar]
  • 2. Zhang H., Li C., Eshetu G. G., et al., “From Solid‐Solution Electrodes and the Rocking‐Chair Concept to Today's Batteries,” Angewandte Chemie 59 (2020): 534–538, 10.1002/anie.201913923. [DOI] [PubMed] [Google Scholar]
  • 3. Padhi A. K., Nanjundaswamy K. S., and Goodenough J. B., “Phospho‐Olivines as Positive‐Electrode Materials for Rechargeable Lithium Batteries,” Journal of the Electrochemical Society 144 (1997): 1188–1194, 10.1149/1.1837571. [DOI] [Google Scholar]
  • 4. Choi J. U., Voronina N., Sun Y. K., and Myung S. T., “Recent Progress and Perspective of Advanced High‐Energy Co‐Less Ni‐Rich Cathodes for Li‐Ion Batteries: Yesterday, Today, and Tomorrow,” Advanced Energy Materials 10 (2020): 2002027, 10.1002/aenm.202002027. [DOI] [Google Scholar]
  • 5. Peng Z., Xia X., Xiao Y., and Huang T., “Progress on Lithium Manganese Iron Phosphate Cathode Materials,” Journal of Alloys and Compounds 1016 (2025): 178901, 10.1016/j.jallcom.2025.178901. [DOI] [Google Scholar]
  • 6. Li S., Zhang H., Liu Y., Wang L., and He X., “Comprehensive Understanding of Structure Transition in LiMn y Fe1 y PO4 during Delithiation/Lithiation,” Advanced Functional Materials 34 (2023): 2310057, 10.1002/adfm.202310057. [DOI] [Google Scholar]
  • 7. Xu P., Tan D. H. S., Jiao B., Gao H., Yu X., and Chen Z., “A Materials Perspective on Direct Recycling of Lithium‐Ion Batteries: Principles, Challenges and Opportunities,” Advanced Functional Materials 33 (2023): 2213168, 10.1002/adfm.202213168. [DOI] [Google Scholar]
  • 8. Yu W., Guo Y., Shang Z., Zhang Y., and Xu S., “A Review on Comprehensive Recycling of Spent Power Lithium‐Ion Battery in China,” eTransportation 11 (2022): 100155. [Google Scholar]
  • 9. Roy J. J., Rarotra S., Krikstolaityte V., et al., “Green Recycling Methods to Treat Lithium‐Ion Batteries E‐Waste: A Circular Approach to Sustainability,” Advanced Materials 34 (2021): 2103346, 10.1002/adma.202103346. [DOI] [PubMed] [Google Scholar]
  • 10. Tao S., Li J., Wang L., Hu L., and Zhou H., “A Method for Recovering Li3PO4 From Spent Lithium Iron Phosphate Cathode Material through High‐Temperature Activation,” Ionics 25 (2019): 5643–5653, 10.1007/s11581-019-03070-w. [DOI] [Google Scholar]
  • 11. Yang Y., Meng X., Cao H., et al., “Selective Recovery of Lithium from Spent Lithium Iron Phosphate Batteries: A Sustainable Process,” Green Chemistry 20 (2018): 3121–3133, 10.1039/C7GC03376A. [DOI] [Google Scholar]
  • 12. Li J., Shi R., Wang J., et al., “Interfacial Metal‐Solvent Chelation for Direct Regeneration of LiFePO4 Cathode Black Mass,” Advanced Materials 37 (2024): 2414235, 10.1002/adma.202414235. [DOI] [PubMed] [Google Scholar]
  • 13. Huang Z., Han S., Pang J., et al., “From Modification to Regeneration: Synergistic Advances in Olivine LiMn x Fe1− x PO4 Cathodes and Closed‐Loop Recycling of Spent LiFePO4 ,” Advanced Functional Materials 36 (2026): e29587, 10.1002/adfm.202529587. [DOI] [Google Scholar]
  • 14. Jeong S. Y., Lee S., Lee H., et al., “Thermal Characteristics of LiMnxFe1‐xPO4 (x = 0, 0.6) Cathode Materials for Safe Lithium‐Ion Batteries,” Journal of Power Sources 626 (2025): 235755, 10.1016/j.jpowsour.2024.235755. [DOI] [Google Scholar]
  • 15. Peng W., Liu J., Yuan W., et al., “Unraveling Phase Transition Dynamics in Carbon‐Modulated Synthesis of LMFP Cathodes from Hydrated Phosphate Precursors,” ACS Applied Materials & Interfaces 17 (2025): 47112–47121, 10.1021/acsami.5c12833. [DOI] [PubMed] [Google Scholar]
  • 16. Zeng Z., Lei H., Wen Y., et al., “Closed‐Loop Regenerative Cycling of Spent LiFePO4 Cathodes via Integrated Lattice Reconstruction and Hydrometallurgy,” Advanced Energy Materials 15 (2025): e04682, 10.1002/aenm.202504682. [DOI] [Google Scholar]
  • 17. Zeng Z., Lu X., Lei S., et al., “Purification–lithiation Collaborative Regeneration of Mixed Graphite/LiFePO4: Building 2D Li+‐diffusion Channels Towards Enhanced Energy‐Storage Capabilities,” Journal of Materials Chemistry A 12 (2024): 27712–27723, 10.1039/D4TA04717F. [DOI] [Google Scholar]
  • 18. Zeng Z., Xu P., Li J., et al., “Large‐Scale and Homogenized Strategies of Spent LiFePO4 Recycling: Reconstruction of Targeted Lattice,” Advanced Functional Materials, 34 (2023): 202308671, 10.1002/adfm.202308671. [DOI] [Google Scholar]
  • 19. Zeng Z., Lei H., Lu X., et al., “Li‐Fe Anti‐Sites Defects in LiFePO4: Mechanism, Characterization and Cathode‐regeneration Applications,” Energy Storage Materials 74 (2025): 103947, 10.1016/j.ensm.2024.103947. [DOI] [Google Scholar]
  • 20. Xin Ren J. G., Liu R., Guo H., and Liang G., “Are Fe ‐ Li Antisite Defects Necessarily Detrimental to the Diffusion of Li+ in LiFePO4/C? ” Journal of The Electrochemical Society 169 (2022): 20507. [Google Scholar]
  • 21. Luo C., Jiang Y., Zhang X., Ouyang C., Niu X., and Wang L., “Misfit Strains Inducing Voltage Decay in LiMn Fe1−PO4/C,” Journal of Energy Chemistry 68 (2022): 206–212, 10.1016/j.jechem.2021.11.007. [DOI] [Google Scholar]
  • 22. Xiang W., Zhong Y.‐J., Ji J.‐Y., et al., “Hydrothermal Synthesis, Evolution, and Electrochemical Performance of LiMn0.5Fe0.5PO4 Nanostructures,” Physical Chemistry Chemical Physics 17 (2015): 18629–18637, 10.1039/C5CP02665B. [DOI] [PubMed] [Google Scholar]
  • 23. Ji G., Tang D., Wang J., et al., “Sustainable Upcycling of Mixed Spent Cathodes to a High‐Voltage Polyanionic Cathode Material,” Nature Communications 15 (2024): 4086, 10.1038/s41467-024-48181-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Xiang W., Wu Z.‐G., Wang E.‐H., et al., “Confined Synthesis of Graphene Wrapped LiMn0.5Fe0.5PO4 Composite via Two Step Solution Phase Method as High Performance Cathode for Li‐Ion Batteries,” Journal of Power Sources 329 (2016): 94–103, 10.1016/j.jpowsour.2016.06.108. [DOI] [Google Scholar]
  • 25. Kim D., Lee S., and Choi W., “Boosting both electronic and ionic conductivities via incorporation of molybdenum for LiFe0.5Mn0.5PO4 cathode in lithium‐ion batteries,” Journal of Alloys and Compounds 989 (2024): 174396, 10.1016/j.jallcom.2024.174396. [DOI] [Google Scholar]
  • 26. Liu J., Wu Y., Zhang B., et al., “A Promising Solid‐State Synthesis of LiMn1‐yFeyPO4 Cathode for Lithium‐ion Batteries,” Small 20 (2023):2309629, 10.1002/smll.202309629. [DOI] [PubMed] [Google Scholar]
  • 27. Liu S., Zheng J., Zhang B., et al., “Constructing Robust Cathode/Li Interfaces and Intensifying Ion Transport kinetics For PEO‐Based Solid‐State Lithium‐Sulfur Batteries,” Chemical Engineering Journal 454 (2023):140385, 10.1016/j.cej.2022.140385. [DOI] [Google Scholar]
  • 28. Deng Z., Wang Q., Peng D., Liu H., and Chen Y., “Fast Precipitation‐induced LiFe0.5Mn0.5PO4/C Nanorods With a Fine Size and Large Exposure of the (010) Faces for High‐Performance Lithium‐Ion Batteries,” Journal of Alloys and Compounds 794 (2019): 178–185, 10.1016/j.jallcom.2019.04.184. [DOI] [Google Scholar]
  • 29. Zhang J., Dong Z., Zeng Z., et al., “Upcycling Spent LiFePO 4: Precisely Fe‐Sites Doped by Ti‐Atoms with Expanded Li‐Diffusion Channels Toward High Capacity,” Advanced Functional Materials 36 (2025): 202522632, 10.1002/adfm.202522632. [DOI] [Google Scholar]
  • 30. Huang Z., Han S., Pang J., et al., “From Modification to Regeneration: Synergistic Advances in Olivine LiMnxFe1−xPO4 Cathodes and Closed‐Loop Recycling of Spent LiFePO4,” Advanced Functional Materials 36 (2026): 202529587, 10.1002/adfm.202529587. [DOI] [Google Scholar]
  • 31. Wen Y., Zeng Z., Li J., et al., “Molecularly Tailored Dual‐Function Deep Eutectic Solvent Enhances Spent Lithium‐Ion Battery Cathode Delamination and Regeneration,” Angewandte Chemie International Edition 65 (2026): 8828799, 10.1002/anie.8828799. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting File: anie72800‐sup‐0001‐SuppMat.docx.

ANIE-65-e7687036-s001.docx (8.4MB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Angewandte Chemie (International Ed. in English) are provided here courtesy of Wiley

RESOURCES